Antidepressant action of ketamine via mTOR is mediated by inhibition of nitrergic Rheb degradation

Abstract

As traditional antidepressants act only after weeks/months, the discovery that ketamine, an antagonist of glutamate/NMDA receptors, elicits antidepressant actions in hours has been transformative. Its mechanism of action has been elusive, though enhanced mTOR signaling is a major feature. We report a novel signaling pathway wherein NMDA receptor activation stimulates generation of nitric oxide (NO), which S-nitrosylates glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Nitrosylated GAPDH complexes with the ubiquitin-E3-ligase Siah1 and Rheb, a small G protein that activates mTOR. Siah1 degrades Rheb leading to reduced mTOR signaling, while ketamine, conversely, stabilizes Rheb which enhances mTOR signaling. Drugs selectively targeting components of this pathway may offer novel approaches to the treatment of depression.

Introduction

A major limitation of antidepressant therapy with classical antidepressant drugs is the substantial lag of weeks – months before attaining therapeutic efficacy. Glutamatergic neurotransmission, especially the N-methyl-D-aspartate (NMDA) subtype of glutamate receptor, has been implicated in the pharmacotherapy of depression ^1–7. The glutamate-NMDA receptor antagonist drug ketamine relieves clinical depressive symptoms in as little as two hours with therapeutic effects lasting for a period of weeks, a notable advance ^8–14. Mechanisms underlying ketamine’s beneficial effects have been elusive. Duman and associates ¹⁵ reported that ketamine’s antidepressant actions reflect enhanced mammalian target of rapamycin (mTOR) signaling. Thus, in animal models antidepressant effects of ketamine are reversed by the mTOR antagonist drug rapamycin as well as by inhibitors of protein kinases in the mTOR pathway, such as Akt and Erk ¹⁵. However, mechanisms whereby NMDA neurotransmission influences mTOR have not been established. One possibility might involve a pathway which we earlier elucidated whereby NMDA neurotransmission signals via nitric oxide (NO) to influence diverse intracellular events. Thus, NMDA transmission triggers the activation of neuronal NO synthase (nNOS) by eliciting cellular entry of calcium, which binds to and activates calmodulin associated with nNOS. The generated NO nitrosylates glyceraldehyde-3-phosphate-dehydrogenase (GAPDH) which in turn binds the ubiquitin-E3-ligase Siah1 ^{16, 17}. The Siah1-nitrosylated GAPDH complex translocates to the nucleus where, in response to apoptotic stimuli, it binds to the acetylating enzyme CREB-binding protein (CBP/p300) activating it and enhancing diverse cell death proteins ¹⁸. The signaling pathway can also exert trophic actions ¹⁹. Thus, neurotrophins also activate nNOS forming NO that nitrosylates GAPDH, which binds to Siah1 and translocates to the nucleus. In the nucleus the neurotrophin pathway differs from the apoptotic system by associating with the histone methylating enzyme Suppressor of Variegation 3–9 Homologue-1 (SUV39H1) in a ternary complex. In this complex, Siah1 ubiquitylates SUV39H1 and elicits its degradation. Loss of the methylating activity of SUV39H1 is associated with diminished methylation of histone-3 on lysine 9 thereby leading to increased acetylation of the histone which, in turn, activates CREB target genes and increases neuronal process growth.

The NO/GAPDH/Siah1 pathway has been elucidated by the use of the drug CGP3466B, an extremely potent inhibitor of GAPDH nitrosylation, acting at as little as 0.1 nM ^{20, 21}. By inhibiting GAPDH nitrosylation CGP3466B prevents cytotoxicity associated with NO activation and hence is neuroprotective ^{20, 22, 23}.

In the present study we have investigated mechanisms whereby the NO/GAPDH/Siah1 pathway interfaces with the small G protein Rheb to influence mTOR. Specifically, we show that nitroyslated GAPDH and Siah1 form a ternary complex with Rheb. Rheb is well established as the proximal stimulus to mTOR activation. In the GAPDH/Siah1/Rheb complex, the ubiquitin-E3-ligase activity of Siah1 leads to degradation of Rheb, which accordingly loses its ability to activate mTOR. Thus, NMDA signaling, acting through NO/GAPDH/Siah1, degrades Rheb and diminishes mTOR signaling. As an NMDA antagonist, ketamine reverses this process leading to the activation of mTOR and presumably accounting for the antidepressant actions of ketamine. Conceivably CGP3466B and other drugs that influence various steps in this pathway will provide novel approaches to the therapy of depression.

Materials and Methods

Reagents

Neurobasal-A medium (no glucose, no sodium pyruvate), B-27 supplement and B-27 supplement minus antioxidants were purchased from Life Technologies (Grand Island, NY). Clasto-lactacystin β-lactone (CLβL) and MG132 were purchased from Cayman Chemical (Ann Arbor, Michigan). Deubiquitinase inhibitor bAP15 was purchased from Ubiquitin-Proteasome Biotechnologies (Aurora, CO). CGP3466B was purchased from Tocris Bioscience (Bristol, United Kingdom). N^ω-nitro-l-arginine methyl ester (L-NAME), 7-nitroindazole, protein G plus-agarose and n-ethylmaleimide were purchased from Santa Cruz Biotechnology (Paso Robles, CA). Ketamine hydrochloride was purchased from Spectrum Chemicals & Laboratory Products (Gardena, CA). N-methyl-d-aspartic acid (NMDA) and Siah1 shRNA (verified) lentiviral plasmid were purchased from Sigma-Aldrich (St. Louis, MO). Rapamycin was purchased from Cell Signaling Technology (Danvers, MA).

Antibodies

Anti-GAPDH-HRP was purchased from GenScript USA Inc. (Piscataway, NJ). Anti-Rheb was purchased from Santa Cruz Biotechnology (Paso Robles, CA). Anti-Siah1 was purchased from AbCam (Cambridge, MA). Anti-phospho-p70S6 Kinase (Thr389), anti-p70S6 Kinase, anti-beta-actin-HRP, anti-phospho-mTOR, anti-mTOR, anti-phospho-ASK, anti-ASK, anti-phospho-MKK4, anti-MKK4, anti-phospho-AMPK and anti-ubiquitin were purchased from Cell Signaling Technology (Danvers, MA). Anti-mono- and polyubiquitylated conjugates-HRP was purchased from Enzo Life Sciences (Farmingdale, NY). Anti-GST-HRP and Anti-HA-HRP were purchased from Sigma-Aldrich (St. Louis, MO).

Transfections

HEK 293 cells transfections were performed using the polyfect transfection reagent following the manufacturer’s instructions (Qiagen, Valencia, CA).

Animals

All experiments involving animals were conducted in accordance with the Johns Hopkins Medical Institutions Animal Care and Use Committee guidelines. Inducible nitric oxide synthase knockout (iNOS KO), neuronal nitric oxide synthase knockout (nNOS KO), C57Bl/6j adult and timed pregnant mice were purchased from The Jackson Laboratory (Bar Harbor, ME). CD-1 timed pregnant mice were purchased from Charles River Laboratories. Animal handling and procedures were conducted in accordance with the National Institutes of Health guidelines for use of experimental animals and the Johns Hopkins animal care and use guidelines.

S-nitrosocysteine (CysNO) treatment

CysNO was prepared by mixing equimolar concentrations of l-cysteine and sodium nitrite. Then quickly using sodium hydroxide to bring the mixture into neutral pH. Freshly prepared CysNO was used immediately.

Immunoprecipitation (IP)

Cultured cells were washed in ice-cold PBS then harvested on ice in IP buffer containing: 10 mM sodium phosphate pH 7.2, 1% NP-40, 100 mM sodium chloride, 2 mM EDTA, 50 mM sodium fluoride, 200 μM sodium orthovanadate, 0.5 μg/ml antipain, 1 μg/ml leupeptin, 1 μg/ml aprotinin, 1 μg/ml chymostatin and 1 μg/ml pepstatin A. For ubiquitylation detection, deubiquitinase inhibitors: n-ethylmaleimide (NEM, 10 mM) and bAP15 (2 μM) were added to the IP buffer. Cell lysates were incubated on ice for 15 min then cleared by centrifugation at 10,000 × g for 10 min at 4°C. Equal amounts of total protein were incubated with the primary antibody for 1 hour at 4°C with rotation. Then, protein G plus agarose beads were added to the lysate-antibody mixture and rotated at 4°C for 1 hour. The beads were washed 5 times by series of centrifugation (at 735 × g) and resuspension using IP buffer. The beads were then resuspended in 1 X reducing SDS-loading buffer, boiled and pelleted by centrifugation. The supernatant was separated by western blot.

Primary cortical neurons (PCNs) culture

Unless otherwise specified, PCNs were isolated from E16-E18 pregnant CD-1 mice. The pregnant mouse was sacrificed by decapitation then the uterus was dissected out immediately. Working in sterile conditions, the uterus was opened, the pups were decapitated and their brains were dissected out and placed in dissection media containing DMEM/F12 1:1 supplanted with 10% horse serum. The cerebral cortices were detached from the rest of the brain then the meninges were removed. The cortices were incubated in 0.025% trypsin for 15 min at 37°C. The trypsin was washed with dissection media. The cortices were disrupted into single cell suspension by pipetting up and down 10 times then strained through a 40 μm sterile mesh. The single cell suspension was cultured in dissection media overnight. Then the media was replaced by PCNs plating media containing neurobasal-A medium (no glucose, no sodium pyruvate) supplemented with 12.5 mM glucose, 2 mM l-glutamine and 2% B-27. On day-4 in vitro (DIV4) the media was changed with PCNs maintenance media containing neurobasal-A medium (no glucose, no sodium pyruvate) supplemented with 12.5 mM glucose, 2 mM l-glutamine and 2% B-27 minus antioxidants. Every 3 days thereafter, 50% of the media was changed with PCNs maintenance media.

Forced swimming test (FST)

C57Bl/6j mice (12–16 weeks old) were subjected to FST 30 min after treatments. The mice were placed in a transparent glass cylinder (17.5 cm across) containing 20 cm deep-water at 25°C. The mice were videotaped for 6 min. The mice were euthanized after the test. The water was changed between subjects. The last 4 min were scored for freezing time in a blinded fashion. Freezing time was calculated by subtracting the active attempts to escape total time from the 4 min.

Novelty-suppressed feeding test (NSFT)

NSFT was performed as previously described ²⁵. Briefly, mice were starved for 24 hours with liberal access to water only. Then the mice were placed in an open field arena under bright light conditions with one pellet of food immobilized in the center of the arena. Mice were left in the arena for a 10 min session. Each session was videotaped. An investigator blinded to the study groups scored the video files for the latency to feed. Immediately after the open arena session, each mouse was placed in its home cage alone with one food pellet for 5 min. The food pellet was weighed before and after the 5 min to quantify the amount consumed by each mouse as an index for appetite.

Lentivirus production

Second-generation packaging was used for lentivirus production. Lentiviral plasmid expressing the Siah1 shRNA was co-transfected into HEK-293T cells with packaging system psPAX2 and envelope plasmid pMD2. The media was changed after 16–18 hours of transfection. The lentivirus containing supernatant was collected 72 hours after transfection. Floating cells and cell debris were pelleted by centrifugation of the supernatant at 2,000 × g for 30 min at 4°C. The supernatant was filtered through a low protein binding 0.45 μm filter. Lentivirus particles were pelleted by ultracentrifugation at 100,000 × g for 2 hours at 4°C. The lentiviral pellet was incubated in 1/100th volume of OPTI-MEM overnight at 4°C. Then the viral particles were resuspended, aliquoted and frozen at −80°C.

Statistical analysis

Statistical analysis was performed using Graphpad software using the alpha power level of 0.05. Two-tailed t test was used to perform two group comparisons. One-way analysis of variance (ANOVA) was used to perform multiple comparisons. Data were graphed as means ± SEM.

Results

NMDA-glutamate signaling elicits Rheb degradation

Suggestive evidence in the literature led us to the hypothesis that the immediate early gene Rheb might relay NMDA signaling to mTOR. Rheb mRNA expression is rapidly induced following NMDA stimulation in the hippocampus ²⁶. In addition, it was previously demonstrated that Rheb binds to the NMDA receptor subunit NR3A ²⁷. Finally, Rheb is a well-documented upstream activator of mTOR. Surprisingly, Rheb levels are rapidly reduced in cortical cultures following NMDA treatment. NMDA stimulation reduces levels of Rheb as well as phospho-S6 kinase with roughly 50% decreases in Rheb levels and phospho-p70S6 kinase, an indicator of mTOR signaling, at about 5 μM NMDA (Fig. 1a). Conversely, ketamine and the NR2B selective inhibitor CERC301 (MK0657) rapidly increase Rheb levels in cortical cultures (Supplementary Figure 1a, c), which correlates with increased phospho-p70S6 kinase (Supplementary Figure 1b, d). On the other hand, the NMDA receptor antagonist memantine does not change Rheb protein levels and mTOR activity as determined by phospho-p70S6 kinase levels (Supplementary Figure 1e). The differential regulation of NMDA receptors by ketamine and memantine might explain these results ^{28, 29}. Injecting mice with ketamine induces a rapid upregulation of Rheb levels in the hippocampus correlating with increased mTOR activation as determined by the phosphorylation status of p70S6 kinase (Fig. 1b). We explored the role of ubiquitylation in mediating downregulation of Rheb levels in response to NMDA. In untreated cortical cultures immunoprecipitated Rheb is mono- and poly-ubiquitylated. Treatment of cortical cultures with the proteasome inhibitor MG132 selectively increases poly-ubiquitylated Rheb (Fig. 1c). Similarly, the proteasome inhibitor bortezomib rapidly increases Rheb and phospho-p70S6 kinase levels in cortical cultures (Fig. 1d). We directly demonstrate both mono- and poly-ubiquitylation of Rheb in response to NMDA stimulation in cortical cultures (Supplementary Figure 1f). Depletion by NMDA of Rheb reflects its proteasomal degradation, as the proteasome inhibitor clasto-lactacystin β-lactone (CLβL) prevents NMDA-associated depletion of Rheb and the associated decline of phospho-p70S6 kinase in cortical cultures (Fig. 1e, f).

Fig. 1. NMDA signaling induces proteasome-mediated Rheb degradation.

Means ± SEM are shown. a, NMDA dose-response in cortical cultures showing a dose-dependent decrease in Rheb levels and mTOR activity. b, Injecting mice with 10 mg/kg ketamine i.p. rapidly upregulates Rheb and activates mTOR in the hippocampus. Graphs show quantification of western blot band intensities normalized to loading control. n = 6–8 mice/group. [* P < 0.05, ** P < 0.01; two-tailed t test]. c, Immunoprecipitation of Rheb from cortical cultures treated with the proteasome inhibitor MG132 (2 μM) shows mono- and polyubiquitylation of Rheb. MG132 treatment leads to accumulation of polyubiquitylated Rheb. d, WB and quantification of Rheb and phospho-p70S6 kinase levels following treatment of cortical culture with the proteasome inhibitor bortezomib [* P < 0.05; two-tailed t test]. e, NMDA stimulation of cortical cultures for 15 min decreases Rheb and phospho-p70S6 kinase levels, which is reversed by treatment with the proteasome inhibitor clasto-lactacystin β-lactone (CLβL) as demonstrated by WB. f, Graph shows quantification of band intensities normalized to loading control from 3 replicates [* P < 0.05; two-tailed t test].

Rheb-GAPDH binding is triggered by the NMDA/NO pathway

Previously, we showed that cell stress activates isoforms of NO synthase (NOS) with the generated NO nitrosylating GAPDH, eliminating its catalytic activity and conferring the ability to bind to Siah1 which mediates nuclear translocation of the complex. In the nucleus this complex binds and activates the acetylating enzyme CREB-binding protein/p300 to augment apoptotic proteins leading to cell death ¹⁸. Ryu and colleagues ³⁰ recently reported that GAPDH binds Rheb, preventing its association with mTOR and thereby decreasing mTOR signaling. We wondered whether glutamate-NMDA neurotransmission acting via NO might impact GAPDH-Rheb interactions. In cortical cultures we monitored the influence of NMDA upon GAPDH-Rheb binding. NMDA substantially increases such binding as does the NO donor S-nitrosocysteine (CysNO) (Fig. 2a, b). The drug CGP3466B, an extremely potent and selective inhibitor of GAPDH nitrosylation ^{20, 21}, abolishes the stimulation by NMDA of GAPDH-Rheb binding. The non-specific NOS inhibitor n^ω-nitro-l-arginine methyl ester (L-NAME) and the neuronal NOS (nNOS) specific inhibitor 7-nitroindazole (7-NI) both prevent these influences of NMDA, establishing that NMDA-glutamate signaling stimulates GAPDH-Rheb binding via NO (Fig. 2a, b). The nitrosylation of GAPDH that triggers its binding to Siah1 takes place on cysteine-150 ²⁴. Mutating cysteine-150 greatly reduces binding of exogenously expressed Rheb to GAPDH in HEK 293 cells, (Fig. 2c, d). We explored the impact of GAPDH-Rheb binding upon mTOR signaling. The stimulation by NMDA of this binding is accompanied by a major decrease in phospho-p70S6 kinase (Supplementary Figure 2a). Conversely, ketamine markedly diminishes GAPDH-Rheb binding (Supplementary Figure 2b).

Fig. 2. Nitric oxide and NMDA receptor regulate GAPDH-Rheb binding.

Representative western blots (WB) are shown. Means ± SEM are shown. a, GAPDH co-immunoprecipitation with Rheb in cortical cultures is increased after 5 min treatment with NO donor CysNO and NMDA determined by western blot (WB). NMDA-induced increase in GAPDH-Rheb binding is blunted by the selective potent GAPDH nitrosylation inhibitor CGP3466B (2 nM) and by NOS inhibitors L-NAME & 7-NI (10 μM). b, Graph shows quantification of band intensities normalized to loading control from 3 replicates [ * P < 0.05; *** P < 0.001; ANOVA (Fisher’s LSD)]. c, WB showing GST pull down assay in HEK 293 cells. HA-C150S mutant of GAPDH binds to GST-Rheb with less affinity compared to HA-WT GAPDH. d, Graph shows quantification of band intensities normalized to loading control from 3 replicates [* P < 0.05; two-tailed t test]. e, WB showing increased levels of GST Rheb expression in HEK 293 cells co-transfected with increasing levels of GAPDH shRNA plasmid. f, WB showing decreased levels of GST Rheb expression in HEK 293 cells co-transfected with WT but not C150S mutant of HA-GAPDH. g, Graph shows quantification of band intensities normalized to loading control from 3 replicates [* P < 0.05; two-tailed t test]. h, NMDA stimulation of cortical cultures elicits a time-dependent decrease in Rheb levels in WT and iNOS KO but not in nNOS KO mice-derived neurons.

In HEK-293 cells we established the importance of endogenous GAPDH in lowering Rheb levels, which are increased following GAPDH depletion by shRNA (Fig. 2e and Supplementary Figure 2c). Regulation of Rheb levels by GAPDH requires its cysteine-150, whose mutation prevents the GAPDH-elicited depletion of Rheb (Fig. 2f, g). The interaction of GAPDH and Rheb is direct, as in vitro GAPDH binds Rheb with the binding increased dramatically by treatment with the NO donor CysNO, further supporting the importance of GAPDH nitrosylation in mediating its binding to Rheb (Supplementary Figure 2d, e). The regulation of mTOR activity by GAPDH is not likely the result of a global effect on protein kinases, since GAPDH overexpression in HEK-293 cells decreases phosphorylated p70S6 kinase levels but does not alter phosphorylated apoptosis signal-regulating kinase (pASK), mitogen activated kinase kinase 4 (pMKK4) or 5′ AMP-activated protein kinase (pAMPK) levels (Supplementary Figure 2f).

To investigate the role of NO in the NMDA-induced decrease of Rheb levels, we examined the time course of NMDA’s actions in cortical cultures from wild type (WT), nNOS or inducible NOS (iNOS) knockout mice. The influence of NMDA is rapid, with maximal effects by 10–15 min (Fig. 2h), and involves nNOS selectively, as the decrease of Rheb levels elicited by NMDA is abolished in cortical cultures from nNOS knockout mice but not altered in cultures from iNOS knockouts (Fig. 2h).

Siah1 binds and degrades Rheb

NMDA stimulates binding of Siah1, an ubiquitin-E3-ligase, to Rheb robustly and rapidly, with detectable effects at 2.5 min and associated with increased Rheb-GAPDH binding (Fig. 3a, b). The binding of Rheb to Siah1 is direct, as it is demonstrable with purified proteins (Fig. 3c). GAPDH enhances this binding in the presence of NO, reflecting the ternary complex of GAPDH with Rheb and Siah1 (Fig. 3c and Supplementary Figure 3a). Siah1’s role as an ubiquitin-E3-ligase for Rheb is supported by the increased levels of Rheb in cortical cultures following depletion of Siah1 by shRNA (Fig. 3d, e and Supplementary Figure 3b).

Fig. 3. Siah1 directly binds Rheb and acts as its E3 ubiquitin ligase.

Means ± SEM are shown. a, NMDA treatment time-course in cortical cultures shows Siah1 and GAPDH co-immunoprecipitation with Rheb increases following NMDA stimulation as determined by WB. b, Graph shows quantification of band intensities normalized to loading control from 3 replicates [ * P < 0.05; ** P < 0.01; *** P < 0.001; **** P < 0.0001; ANOVA (Fisher’s LSD test). c, Western blot showing GST pull down assay using purified proteins. GST-Siah1 binds directly to Rheb and GAPDH. CysNO increases the binding between GST-Siah1, Rheb and GAPDH. e, Knockdown of Siah1 in cortical cultures using lentiviral shRNA leads to up-regulation of Rheb protein levels. f, Graph shows quantification of band intensities normalized to loading control from 5 replicates [* P < 0.05; two-tailed t test].

CGP3466B, which inhibits GAPDH S-nitrosylation, displays rapid antidepressant actions

Our findings thus far reflect a signaling cascade wherein NMDA-glutamate neurotransmission activates the formation of NO which nitrosylates GAPDH enabling it to bind to Siah1 together with Rheb. In this complex the ubiquitin-E3-ligase activity of Siah1 leads to degradation of Rheb and a resulting diminution of mTOR signaling. By blocking NMDA receptors, ketamine does the reverse, stimulating mTOR signaling and presumably leading to the antidepressant actions of ketamine. We speculated that CGP3466B, which selectively inhibits GAPDH nitrosylation at concentrations as low as 0.1 nM ^{20, 21, 31}, might, like ketamine, interrupt the NMDA-mTOR signaling pathway and exert antidepressant actions. Accordingly, we administered CGP3466B to mice in a swimming paradigm which predicts the actions of acutely administered antidepressants ^{32, 33}. In this model mice subjected to “forced swimming” appear to “give up” and stop swimming or “freeze” ³⁴. Treatment with antidepressant drugs decreases the freezing/immobility time. CGP3466B substantially decreases immobility time, an effect reversed by prior treatment with rapamycin (Fig. 4a, b). CGP3466B does not alter locomotor activity of mice (Supplementary Figure 4a). We also explored potential rapid antidepressant activity of CGP3466B in the novelty-suppressed feeding test (NSFT). Traditional antidepressants are only effective in NSFT following chronic (21 days) administration ³⁵. By contrast, CGP3466B is acutely (30 min) effective in NSFT, with effects reversed by rapamycin (Fig. 4a, c). CGP3466B does not alter food intake (Supplementary Figure 4b).

Fig. 4. Rapid antidepressant-like behavioral effects of the GAPDH nitrosylation inhibitor CGP3466B require mTOR.

Means ± SEM are shown. Behavioral experiments were scored in a blinded fashion. a, Schematic diagram of experimental time line. FST forced swimming test; NSFT, novelty suppressed feeding test. b, FST in mice showing reduction in immobility or freezing time 30 min after 0.6 mg/kg of CGP3466B reversed by 5 mg/kg rapamycin. Control n = 36; Rapamycin n = 15, CGP3466B n = 35; Rapamycin-CGP3466B n = 20; ANOVA; P = 0.0003; Newman-Keuls multiple comparisons test: *** P < 0.001; * P < 0.05. c, NSFT in mice showing reduction in latency to feed 30 min after 0.6 mg/kg of CGP3466B reversed by 5 mg/kg rapamycin. Control n = 31; Rapamycin n = 15, CGP3466B n = 20; Rapamycin-CGP3466B n = 15; ANOVA; P = 0.008; Newman-Keuls multiple comparisons test: * P < 0.05; ** P < 0.01.

Discussion

In summary, our findings provide a molecular mechanism whereby NMDA receptor antagonists such as ketamine can increase mTOR signaling to elicit antidepressant actions. By blocking NMDA receptors, ketamine decreases neural formation of NO. The reported antidepressant phenotypes associated with NOS inhibition ³⁶ may reflect the pathway described here. In this cascade NO nitrosylates GAPDH at cysteine-150. Conceivably, drugs, such as CGP3466B, which inhibit such nitrosylation, may exert antidepressant influences. Though CGP3466B displays prominent neuroprotective actions ²⁰, we are not aware of studies of CGP3466B in animal models of depression. In the ternary complex of nitrosylated GAPDH/Rheb/Siah1 the ubiquitin-E3-ligase activity of Siah1 degrades Rheb leading to decreased mTOR activation.

What is most striking about the antidepressant actions of ketamine is their rapidity and persistence with clinical improvement of mood in depressed patients evident within 2 hours and with these effects lasting at least for a week ³⁷. Whether or not the immediate action of ketamine involves the same or different mechanisms than the longer acting effects is not established, but it is reasonable to suppose that the initial improvement in mood is persistent at least for several days and involves the same mechanism. Accordingly, the influence of ketamine upon Rheb disposition might account both for the immediate and the longer lasting influences of ketamine and other NMDA antagonists.

Neurotransmission, especially via glutamate acting upon NMDA receptors, is known to impact protein translation ^38–41. Underlying molecular mechanisms have been elusive. The regulation by glutamate/NMDA signaling of Rheb and mTOR may well mediate these actions. Presumably AMPA signaling will influence protein translation by augmenting NMDA transmission.

Memantine is an NMDA antagonist that has been employed clinically in Alzheimer’s Disease ⁴². In our experiments memantine did not impact Rheb/mTOR disposition. This finding may relate to studies reporting that memantine, unlike ketamine, fails to elicit rapid antidepressant responses ⁴³. Lipton and associates ⁴⁴ provided evidence that memantine may differ from other NMDA antagonists in preferentially blocking extra-synaptic rather than synaptic NMDA receptors. It is postulated that the antidepressant effects require blockade of synaptic NMDA receptors.

Ketamine’s blockade of NMDA transmission stimulates mTOR signaling, which presumably mediates antidepressant influences. Other therapeutic possibilities might emerge from further exploitation of this novel signaling pathway whereby NMDA neurotransmission, via GAPDH, Siah1, and Rheb, impacts mTOR.