The N-Methyl-D-Aspartate Receptor Blocker REL-1017 (Esmethadone) Reduces Calcium Influx Induced by Glutamate, Quinolinic Acid, and Gentamicin

REL-1017 (esmethadone) is a novel N-methyl-D-aspartate receptor (NMDAR) antagonist and promising rapid antidepressant candidate. Using fluorometric imaging plate reader (FLIPR) assays, we studied the effects of quinolinic acid (QA) and gentamicin, with or without L-glutamate and REL-1017, on intracellular calcium ([Ca2+]in) in recombinant cell lines expressing human GluN1-GluN2A, GluN1-GluN2B, GluN1-GluN2C, and GluN1-GluN2D NMDAR subtypes. There were no effects of QA on [Ca2+]in in cells expressing GluN1-GluN2C subtypes. QA acted as a low-potency, subtype-selective, NMDAR partial agonist in GluN1-GluN2A, GluN1-GluN2B, and GluN1-GluN2D subtypes. REL-1017 reduced [Ca2+]in induced by QA. In cells expressing the GluN1-GluN2D subtype, QA acted as an agonist in the presence of 0.04 μM L-glutamate and as an antagonist in the presence of 0.2 μM L-glutamate. REL-1017 reduced [Ca2+]in induced by L-glutamate alone and with QA in all cell lines. In the absence of L-glutamate, gentamicin had no effect. Gentamicin was a positive modulator for GluN1-GluN2B subtypes at 10 μM L-glutamate, for GluN1-GluN2A at 0.2 μM L-glutamate, and for GluN1-GluN2A, GluN1-GluN2B, and GluN1-GluN2D at 0.04 μM L-glutamate. No significant changes were observed with GluN1-GluN2C NMDARs. REL-1017 reduced [Ca2+]in induced by the addition of L-glutamate in all NMDAR cell lines in the presence or absence of gentamicin. In conclusion, REL-1017 reduced [Ca2+]in induced by L-glutamate alone and when increased by QA and gentamicin. REL-1017 may protect cells from excessive calcium entry via NMDARs hyperactivated by endogenous and exogenous molecules.


Introduction
REL-1017 (esmethadone; dextromethadone; d-methadone; S-methadone; (+)-methadone) is the opioid-inactive d-isomer of racemic methadone and a low-affinity, low-potency N-methyl-D-aspartate receptor (NMDAR) channel blocker that binds to the MK-801 site of the NMDAR at low micromolar half-maximal inhibitory concentration (IC 50 ) values [1][2][3]. REL-1017 showed efficacy in animal models of depressive-like behavior [4,5] and rapid, robust, and sustained therapeutic effects as adjunctive treatment in patients with major depressive disorder (MDD) [6]. The clinical findings in patients also confirmed the favorable safety, tolerability, and pharmacokinetic profiles observed in Phase 1 studies [7]. NMDAR antagonists cause neuronal vacuoles, known as Olney's lesions, in rats, which may be related to NMDAR toxicity in humans [8]. In a recent experiment, REL-1017 did not cause Olney's lesions, in contrast with MK-801 [9]. In summary, these studies suggest that REL-1017 is a safe and well-tolerated novel NMDAR channel blocker and rapid antidepressant candidate. To gain additional insights into the mechanism of action of REL-1017, we investigated its effects by means of fluorometric imaging plate reader (FLIPR) assays in the presence of putative NMDAR function-associated endogenous and exogenous neurotoxins, such as quinolinic acid (QA) and gentamicin.
QA is a 2,3-pyridine dicarboxylic acid produced by the breakdown of the amino acid tryptophan via the kynurenine pathway. QA is physiologically present in nanomolar concentrations in the brain; however, increased levels of QA can be produced by activated macrophages and microglia in pathological conditions. Accumulation of endogenous QA has been implicated in the etiology of neurological diseases [10,11] and psychiatric disorders, including depression [12][13][14][15][16][17][18][19]. QA has also been implicated in peripheral neurodegenerative disorders, such as hearing loss [20,21], and has recently been implicated in the neuropsychiatric complications of COVID-19 [22,23]. A variety of mechanisms have been proposed to explain QA toxicity, including activation of NMDARs via an increase in ambient glutamate and via direct agonist actions at the glutamate site of the NMDAR [11,[24][25][26][27][28]. Of note, QA-induced lesions differ across multiple brain regions, suggesting that QA could exert regional effects based on the specific expression of NMDAR subtypes at the site of the lesion [29]. Electrophysiological recording in oocytes overexpressing recombinant NMDARs provided direct evidence of NMDAR activation by QA, with low-potency partial agonist activity at L-glutamate binding sites on GluN1-GluN2A, GluN1-GluN2B, GluN1-GluN2D, and GluN1-GluN3B NMDAR subtypes [30][31][32][33][34]. QA has been shown to be inactive at GluN1-GluN2C subtypes [30]. QA activation of NMDARs may also have a causative role in peripheral diseases [35]. QA has been shown to accumulate during liver failure [36] and renal insufficiency [37,38].
The current study utilized QA to activate NMDARs in vitro in order to better evaluate the role of REL-1017 in modulating pathologically hyperactive NMDARs. We first verified if QA could modulate NMDAR-mediated intracellular calcium ( [Ca 2+ ] in ) in recombinant cell lines expressing heterodimeric NMDARs with subtype specificity. We then assessed the ability of REL-1017 to antagonize QA effects on NMDARs, alone and in the presence of glutamate.
In order to further assess the role of REL-1017 in modulating NMDAR activity, REL-1017 was evaluated in the presence of the aminoglycoside antibiotic gentamicin. Gentamicin acts mainly on aerobic Gram-negative bacteria, including multidrug-resistant bacteria. However, its clinical usefulness is limited by potentially serious adverse effects, most commonly ototoxicity, including both cochlear and vestibular toxicity [39,40], and nephrotoxicity [39,41]. A variety of mechanisms have been proposed to explain aminoglycoside side effects [39,40], including activation of NMDARs [42][43][44][45], which are expressed in various cell types of the inner ear [46] and kidney [47]. Direct evidence supports activation of NMDARs by gentamicin [45,[48][49][50] and tobramycin, another aminoglycoside antibiotic [48,51]. Therefore, we also directly tested if gentamicin could increase NMDARmediated [Ca 2+ ] in in recombinant cell lines expressing heterodimeric NMDARs, both in the presence and absence of glutamate. We then assessed the ability of REL-1017 to antagonize the effects of gentamicin on NMDARs activated by glutamate and the ability of REL-1017 to antagonize the effects of gentamicin and QA.
These experiments may improve our understanding of the mechanism of action of REL-1017, including its potential to protect cells from excessive calcium entry via NM-DARs hyperactivated by glutamate and endogenous or exogenous substances, such as QA and gentamicin.

QA Acts as a Low-Potency Subtype-Selective NMDAR Partial Agonist, and REL-1017 Reduces QA-Induced Calcium Entry
The effect of QA on NMDAR subtypes is shown in Figure 1A and Table 1. An increase in [Ca 2+ ] in was detected with the addition of QA and in the absence of L-glutamate in GluN1-GluN2A, GluN1-GluN2B, and GluN1-GluN2D cell lines. In addition, 333 µM QA enhanced [Ca 2+ ] in by an average of 18%, 23%, and 21% in cells expressing GluN2A, GluN2B, and GluN2D, respectively (n = 6 for each group), compared to a 100% value obtained with 10 µM L-glutamate. At the highest tested concentration (1000 µM), QA enhanced [Ca 2+ ] in by an average of 40%, 25%, and 50% of the maximum response to L-glutamate alone (n = 6 for each group) in cells expressing GluN2A, GluN2B, and GluN2D, respectively. QA was inactive on the GluN1-GluN2C cell line at all tested concentrations. Therefore, in this FLIPR calcium assay, QA acted as a low-potency subtype-selective NMDAR partial agonist. is a subtype-selective N-methyl-D-aspartate receptor (NMDAR) partial agonist. QA concentration-response curve (CRC) was performed in the absence of L-glutamate by fluorometric imaging plate reader (FLIPR) calcium assay, using the Chinese hamster ovary (CHO) cell lines expressing the indicated heterodimeric NMDAR. QA, at the two highest tested concentrations, 333 µM and 1000 µM, increased intracellular calcium ( [Ca 2+ ] in ) in GluN2A-, GluN2B-, and GluN2D-expressing cell lines, with estimated an EC 50 of 850 µM, 170 µM, and 520 µM for GluN2A, GluN2B, and GluN2D receptors, respectively. Estimated maximal asymptotic response was 75%, 28%, and 71% of that evoked by 10 µM L-glutamate for GluN2A, GluN2B, and GluN2D receptors, respectively. GluN2C-expressing cells were not activated by any tested concentration of QA. Data are mean ± standard error of the mean (SEM), n = 6 wells per group, and were fitted using four-parametric logistic equation with GraphPad Prism v8.0. (B) It was shown that 10 µM REL-1017 reduced calcium entry in the presence of 1000 µM QA. QA 1000 µM increased [Ca 2+ ] in up to 41%, 37%, and 55% in GluN2A, GluN2B, and GluN2D NMDAR subtypes, respectively. In addition, 10 µM REL-1017 significantly (p < 0.0001) decreased these values to 34%, 12%, and 32% in GluN1-GluN2A, GluN1-GluN2B, and GluN1-GluN2D NMDAR subtypes, respectively. In the GluN1-GluN2C cell line, [Ca 2+ ] in was −5.5% and −7.6% after 1000 µM QA and 1000 µM QA plus 10 µM REL-1017, suggesting a reduction in baseline NMDAR-mediated [Ca 2+ ] in , possibly due to an endogenous low-level L-glutamate release by the CHO cell line. Data (n = 7 wells for buffer and REL-1017 groups, n = 42 wells for the remaining groups) are shown as scatter dot plot, and median value is indicated. **** is p < 0.0001; *** is p < 0.001; * is p < 0.05; ns is not statistically significant.

Multiple Comparisons Test
Tukey's GluN2A p-Value

Gentamicin Is a Selective Positive Modulator at NMDAR GluN1-GluN2B Subtypes for Saturating L-Glutamate Concentration
We tested if gentamicin could activate NMDARs by acting as an agonist in the absence of L-glutamate. The ability of gentamicin to increase [Ca 2+ ] in was first assessed in the absence of L-glutamate but in the presence of 10 µM glycine. A gentamicin CRC using eleven different concentrations of gentamicin ranging from 1.7 ng/mL up to 100 µg/mL was tested in a FLIPR calcium assay using four CHO cell lines, each expressing a different heterodimeric NMDAR, including GluN1-GluN2A, GluN1-GluN2B, GluN1-GluN2C, and GluN1-GluN2D. No increase in [Ca 2+ ] in was detected with the addition of gentamicin in the absence of L-glutamate at any concentration and in any cell line ( Figure 3A). At the tested concentrations, gentamicin had no effect as an agonist on NMDAR by substituting L-glutamate. Gentamicin CRC was performed in absence of L-glutamate by FLIPR calcium assay, using the CHO cell lines expressing the indicated heterodimeric NMDAR. Gentamicin did not increase [Ca 2+ ] in at any tested concentration and in any cell line, indicating that it cannot ne substituted for L-glutamate in the activation of NMDARs. Data are mean ± SEM, n = 6. (B) Gentamicin was a positive modulator only at GluN1-GluN2B subtypes in the presence of 10 µM L-glutamate. A second gentamicin CRC was run in similar FLIPR calcium assay conditions as those for the experiment illustrated in the text, but in the presence of 10 µM L-glutamate. Gentamicin increased the [Ca 2+ ] in only in the cell line expressing the GluN1-GluN2B subtype, eliciting a maximal response of 128% at 100 µg/mL gentamicin, with a calculated EC 50 of 21 µg/mL. Data are mean ± SEM, n = 6.
We tested if gentamicin could act as a positive modulator on NMDARs by enhancing the L-glutamate-induced maximal calcium response. A FLIPR calcium assay was designed to evaluate changes to the gentamicin CRC with the addition of 10 µM L-glutamate, using the same NMDAR cell lines used for the agonist assay. In the cell line expressing the GluN1-GluN2B NMDAR subtype, 100 µg/mL gentamicin significantly (p < 0.0001, unpaired t-test versus control) increased [Ca 2+ ] in compared to levels elicited by 10 µM L-glutamate alone ( Figure 3B). The increase with added gentamicin elicited a 128% change in [Ca 2+ ] in relative to the maximum response recorded in the presence of 10 µM L-glutamate alone. The calculated EC 50 of gentamicin for the GluN1-GluN2B subtype was 21 µM. Considering that [Ca 2+ ] in reaches a plateau following induction by 10 µM L-glutamate, the further change to [Ca 2+ ] in induced by the additional gentamicin can be considered a greater than maximal response on GluN1-GluN2B NMDARs. In contrast, 100 µg/mL gentamicin induced a significant (p < 0.0001) reduction in [Ca 2+ ] in of 80%, 65%, and 78% in the remaining cell lines of GluN1-GluN2A, GluN1-GluN2C, and GluN1-GluN2D subtypes, respectively, acting as a negative allosteric modulator on NMDAR function in these cell lines exposed to 10 µM L-glutamate.

Gentamicin Is Not a Positive Modulator at NMDARs Activated by Quinolinic Acid
We then tested the effect of gentamicin on QA-induced changes to [Ca 2+ ] in , as it is known that positive modulator effects might be agonist dependent. Additionally, QA is an endogenous metabolite that acts as an agonist at the glutamate site of the NMDAR and has neurotoxic and ototoxic potential [21]. QA is a weak, subtype-selective, partial agonist at the L-glutamate site of NMDARs; 1000 µM QA induced changes in [Ca 2+ ] in of 42%, 37%, and 55% of the 10 µM L-glutamate response at GluN1-GluN2A, -GluN2B, and -GluN2D NMDARs (p < 0.0001), respectively, and was inactive on GluN1-GluN2C subtypes ( Figure 5; Table 6). When tested in combination with 1000 µM QA, 10 µg/mL gentamicin was inactive on all tested NMDARs. Figure 5. The 1000 µM QA was a partial agonist at GluN1-GluN2A, GluN1-GluN2B, GluN1-GluN2C, or GluN1-GluN2D subtypes and 10 µg/mL gentamicin did not increase [Ca 2+ ] in in the presence of 1000 µM QA. REL-1017 significantly decreased [Ca 2+ ] in elicited by 1000 µM QA or 1000 µM QA plus 10 µM gentamicin. Data (n = 7 wells for buffer and REL-1017 groups, n = 42 wells for the remaining groups) are shown as scatter dot plot, and median value is indicated. **** is p < 0.0001; * is p < 0.05; ns, not statistically significant. Table 6. Statistical significance of values reported in Figure 5.

REL-1017 Lowered Intracellular Calcium Levels in Presence of 10 µM Gentamicin and Low L-Glutamate
After showing that REL-1017 reduced [Ca 2+ ] in that was increased after NMDAR activation by QA (Figure 1), we investigated if REL-1017 was also effective in reducing [Ca 2+ ] in increased by gentamicin, a positive modulator at the NMDAR, in the presence of low glutamate concentrations. It was found that 10 µM REL-1017 reduced [Ca 2+ ] in induced by the addition of 0.04 or 0.02 µM L-glutamate in all NMDAR cell lines, in the presence or absence of 10 µg/mL gentamicin (Figure 4, p < 0.0001 or p < 0.001). In addition, 10 µM REL-1017 also reduced [Ca 2+ ] in elicited by 1000 µM QA in the presence and absence of 10 µg/mL gentamicin in all NMDAR cell lines ( Figure 5, p < 0.0001).

Discussion
The central role of NMDARs in neural plasticity has been recognized for several decades [52,53]. The distribution of NMDAR subtypes is genetically, regionally, and developmentally determined [53]. Diseases may be triggered, maintained, or worsened by dysregulated calcium currents through NMDARs [54], including peripheral NMDARs [55]. The effect of QA on NMDAR in this intracellular FLIPR calcium assay confirms and extends the data obtained by electrophysiological recordings in transduced oocytes expressing heterodimeric NMDARs [30][31][32][33][34]. In particular, we confirmed that QA is a weak partial agonist at GluN1-GluN2A, GluN1-GluN2B, and GluN1-GluN2D subtypes. QA significantly increased [Ca 2+ ] in only at the two highest tested concentrations, 333 µM and 1000 µM (Figure 1). QA alone was inactive at every tested concentration up to 1000 µM on heterodimeric GluN1-GluN2C subtypes (Figure 1), as previously described [30]. Our results support the hypothesis that brain region-specific patterns of QA-induced lesions may be caused by preferential agonist activity at regionally and developmentally specific NMDAR subtypes [29].
We obtained additional information regarding the QA interactions with NMDARs by studying the effect of 1000 µM QA added to cell culture media containing low L-glutamate concentrations. It was found that 1000 µM QA potentiated the effect of 0.04 µM and 0.2 µM L-glutamate on GluN1-GluN2A or GluN1-GluN2B NMDARs (Figure 2). This potentiation was observed with L-glutamate concentrations well below the EC 50 for NMDAR heterodimeric GluN1-GluN2A or GluN1-GluN2B subtypes. In addition, 1000 µM QA is expected instead to decrease the response to saturating L-glutamate concentrations on GluN1-GluN2A or GluN1-GluN2B, but this condition remains to be tested. Divergent effects of 1000 µM QA were instead observed with [Ca 2+ ] in elicited by 0.04 µM and 0.2 µM L-glutamate at the GluN1-GluN2D subtype. Approximately 18% of the maximum [Ca 2+ ] in induction was elicited by 0.04 µM L-glutamate alone, which increased to 55% in the presence of 1000 µM QA (Figure 2A, p < 0.0001). In contrast, an average of 92% of the maximum [Ca 2+ ] in induction was elicited by 0.2 µM L-glutamate alone, while the combination with 1000 µM QA significantly decreased the 0.2 µM L-glutamate effect on [Ca 2+ ] in response in GluN1-GluN2D heterodimeric NMDARs ( Figure 2B, p < 0.0001). Interestingly, 0.04 µM and 0.2 µM L-glutamate concentrations represent values below and above the L-glutamate EC 50 for the GluN1-GluN2D subtype in this assay, respectively; this observation may help explain the contrasting outcomes. Moreover, 1000 µM QA decreased [Ca 2+ ] in elicited by 0.2 µM L-glutamate in GluN1-GluN2C NMDAR subtypes (p < 0.0001; Figure 2B). This result suggests that QA can bind to the L-glutamate site of GluN1-GluN2C subtypes and compete with L-glutamate without a pharmacological effect on this receptor subtype, as QA alone is unable to activate the GluN1-GluN2C NMDAR subtypes.
Considering that sub-micromolar concentrations of ambient glutamate tonically activate NMDARs in select cell types and/or brain regions, physiologically and pathologically [54,[56][57][58][59], our observed QA potentiating effect of sub-saturating L-glutamate concentrations may help explain the quinolinic-induced neurotoxicity by GluN1-GluN2A, GluN1-GluN2B, or GluN1-GluN2D NMDAR subtype hyperactivity ( Figure 2). With regard to the GluN1-GluN2D subtypes, potentiation was evident only at 0.04 µM and not at 0.2 µM, suggesting that QA-induced calcium dysregulation at this receptor subtype may be preferentially relevant in the presence of very low ambient glutamate. As ambient glutamate in physiological states is expected to be negligible, the importance of this receptor subtype as the most vulnerable target in the pathophysiology of disorders triggered or maintained by excess ambient glutamate, even at very low concentrations, may deserve more attention. Of note, this receptor subtype is the preferential target for REL-1017 in the presence of physiological magnesium [60]. Finally, the effect of REL-1017 on the reduction in [Ca 2+ ] in induced by QA, alone (Figure 1) or in combination with sub-saturating L-glutamate concentrations (Figure 2), suggests potential neuroprotective effects when QA might overstimulate NMDARs directly, alone and in the presence of low-concentration ambient glutamate.
The effects of gentamicin on NMDARs in our intracellular calcium assay confirm and extend the data obtained through electrophysiological recordings from transduced oocytes expressing heterodimeric NMDARs [48][49][50]. In particular, we confirmed the selectivity of gentamicin for the GluN1-GluN2B receptor at 10 µM L-glutamate concentration (Figure 3), with an EC 50 of 21 µg/mL (30.8 µM). Exposure to 100 µg/mL (146.7 µM) gentamicin in addition to 10 µM L-glutamate induced a [Ca 2+ ] in increase of 128% of the previously measured plateau from saturating L-glutamate concentrations alone. This finding is in agreement with the 38 µM EC 50 and 132% plateau maximal response with 300 µM gentamicin de-scribed in oocytes [48]. In addition, we observed a potentiation by 10 µg/mL gentamicin of the sub-saturating 0.04 µM glutamate effect on GluN1-GluN2A, GluN1-GluN2B, and GluN1-GluN2D NMDAR subtypes ( Figure 4A), while potentiation of a slightly higher but still sub-saturating 0.2 µM glutamate effect was observed only with GluN1-GluN2A NMDAR subtypes ( Figure 4B).
Recently, NMDAR positive allosteric modulators (PAMs) have been grouped into the following two classes [61]: type I, which modulates the maximal glutamate effect, and type II, which increases agonist affinity. Our reported results suggest that gentamicin can be considered a selective GluN1-GluN2B type I PAM. Further studies should be performed to establish if gentamicin could also be considered a GluN2A, GluN2B, or GluN2D type II PAM. Considering that sub-micromolar concentrations of ambient glutamate might tonically activate NMDARs in regional-and developmentally-specific neuronal populations [57], the observed gentamicin potentiating effect of sub-saturating L-glutamate concentrations might indicate a previously unappreciated mechanism of gentamicin neurotoxicity. The effect of REL-1017 in reducing [Ca 2+ ] in induced by 10 µg/mL gentamicin and sub-saturating L-glutamate concentrations suggests the potential neuroprotective effects of REL-1017 against gentamicin-associated NMDAR excitotoxicity in the context of low ambient glutamate.
REL-1017 might, therefore, provide neuroprotective channel blocking effects in conditions of NMDAR hyperactivation from potential endogenous and exogenous neurotoxins. REL-1017 is currently in Phase 3 trials for the treatment of MDD (ClinicalTrials.gov Identifiers: NCT04855747; NCT04688164; NCT05081167) and is under investigation for the treatment of other neuropsychiatric and neurodegenerative disorders. Dysregulated NMDAR-mediated calcium currents have been implicated in the pathophysiology of neuropsychiatric and central neurodegenerative disorders, in addition to peripheral disorders [62], including COVID-19 [22,23]. The genetic and environmental paradigm, stating that genetically predisposed individuals exposed to select triggers may develop diseases, is generally accepted [63]. QA is an endogenous substance with the potential to elicit neurotoxicity via excessive NMDAR activity, which has also been implicated in the development of neurodegenerative disorders. Gentamicin is an antibiotic with toxicity, which is also potentially mediated via NMDAR hyperactivation.
Uncompetitive NMDAR channel blockers have been proven effective for the treatment of neuropsychiatric diseases. Memantine, esketamine, and the combination dextromethorphanquinidine are FDA approved for Alzheimer's disease, treatment-resistant major depressive disorder, and for pseudobulbar palsy, respectively. Improvements in the knowledge and characterization of NMDARs' function, their agonists, their allosteric modulators, and their antagonists, combined with the recent availability of clinical data for NMDAR-modulating drugs, are likely to progress our understanding of NMDAR functions in health and disease and are likely to improve our ability to understand and treat diseases associated with this receptor system. The results of this study suggest that REL-1017 could be a novel treatment for diseases triggered, maintained, or worsened by excessive NMDAR-dependent calcium currents, particularly via GluN1-GluN2D subtypes.

Statistical Analysis
The study was not preregistered, and Institutional Ethics Committee approval was not necessary as the studies were performed on cell lines. Sample size was determined based on similar cell culture studies [64]. As these were in vitro cell culture experiments, no randomization procedures were applied and no blinding, sample calculation, or test for outliers were performed.
FLIPR data were expressed as a percentage of response, considering the [Ca 2+ ] in in the presence of 10 µM L-glutamate and 10 µM glycine as the 100% response (H, high control) and [Ca 2+ ] in in assay buffer alone as the 0% response (L, low control). Data were normalized according to the following formula: (1) Statistical analyses were performed with GraphPad Prism v8.0 software (GraphPad Software, Inc. La Jolla, CA, USA). Concentration-response fittings were performed using the four-parameter logistic equation and plotted as mean ± standard error of the mean (SEM). Additional data were shown as dot plots with medians (center line). Since the normal distribution of data, checked by the Kolmogorov-Smirnov test, could not be rejected, statistical comparisons between groups were performed by Student's t-test for unpaired data. Comparisons between more than two groups were performed by one-way analysis of variance (ANOVA), followed by the Tukey's multiple comparison post hoc test or by the nonparametric Kruskal-Wallis test, followed by Dunn's multiple comparison test for non-normally distributed data. For all experiments, p < 0.05 was considered statistically significant. Statistical tests were two-tailed. The need for nonparametric analysis was assessed by the Shapiro-Wilk normality test (alpha = 0.05).

Conclusions
Exposure to endogenous or exogeneous substances acting as agonists or PAMs at NMDARs may dysregulate calcium currents, with downstream consequences that depend on the cells and circuits affected. Here, we show that REL-1017 may counteract the increases in [Ca 2+ ] in caused by endogenous and exogenous molecules in the presence and absence of glutamate. REL-1017 may counteract different mechanisms of QA neurotoxicity, including QA-induced increases in ambient glutamate and the direct agonist effect at the glutamate site of NMDARs. These experiments confirm that the GluN1-2D subtype is the most sensitive to low-concentration glutamate (0.04 µM and 0.2 µM), suggesting that uncompetitive NMDAR blockers with a preference for this subtype, such as REL-1017, may be of particular therapeutic relevance.