Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE (www.oxfordhandbooks.com). © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

date: 18 October 2019

Dysregulated Protein Synthesis in Major Depressive Disorder

Abstract and Keywords

Major depressive disorder is a debilitating disorder with a lifetime prevalence of 17% in the adult population. By reverse engineering how antidepressants work at the cellular level, significant progress has been made within the last decade regarding the underlying etiology of depression. Unexpectedly, dysregulation of protein synthesis pathways is at the core of depression. Activation of one or more mRNA translation, initiation, or elongation pathways (including mammalian target of rapamycin [mTOR] kinase, extracellular regulated kinase, and eukaryotic elongation factor 2) is central to symptomatic relief. In preclinical models of stress and/or depression, co-administration of antidepressants and pharmacological inhibitors of these pathways block hallmark characteristics of antidepressant efficacy, including upregulation of key synaptic proteins, increased dendritic and spine complexity, and antidepressant-like behaviors. In this chapter, we review studies demonstrating altered translational pathways in animal models, treated and untreated patients, with a focus on mTOR-regulated protein synthesis.

Keywords: mRNA translation, RNA, mammalian target of rapamycin, rapid antidepressant, NMDAR antagonist, protein synthesis, depression, major depressive disorder, chronic mild stress

Introduction

Major depressive disorder (MDD) is a heterogeneous disorder, wherein the symptoms and treatment efficacy are highly variable among individuals. Current treatments range from typical antidepressants, such as selective serotonin reuptake inhibitors (SSRIs), to electroconvulsive therapy. While these therapies have shown promise in some cases, only 35–50% of patients undergoing treatment experience relief from depressive symptoms, and the rate of relapse is high (Murrough, 2012; Rush et al., 2011; Undurraga & Baldessarini, 2012). Moreover, the onset of efficacy of typical antidepressants is delayed, thus increasing the risk of suicide (Hieronymus, Nilsson, & Eriksson, 2016). Within the last decade, glutamate N-methyl-D-aspartate receptor (NMDAR) antagonists, such as ketamine, have emerged as viable treatments for MDD due to their rapid onset of symptom relief, thus bridging the gap between the administration of typical antidepressants and the onset of their efficacy (Trullas et al., 1991). In addition, NMDAR antagonists show promise in treatment-resistant patients, for whom typical antidepressants are ineffective (Murrough, 2012). While ketamine appears to be the Holy Grail to treat MDD, its abuse potential is high and it can induce psychosis at high doses (Y. Liu et al., 2016). Thus, the discovery of a safe pharmaceutical that elicits the same rapid-acting antidepressant effects of ketamine is necessary.

MDD Is a Disease of the Synapse

Patients with MDD exhibit alterations in neural tissue volume and connectivity within the frontal cortex and hippocampus, regions and circuits responsible for cognition, emotional regulation, and stress-responsiveness (Koolschijn et al., 2009; Bearden et al., 2009; Genzel et al., 2015). Specifically, these regions exhibit decreased volume and activity (Drevets, Price, & Furey 2008; Murrough et al., 2011), and patients with MDD also exhibit fewer synapses (Kang et al., 2012) and decreased synaptic proteins (Feyissa et al., 2009) within these regions. Preclinical rodent models of depression display similar neural alterations, including significant loss of dendrites, dendritic spines, and synaptic excitation (Liston et al., 2006; Radley et al., 2008; Yuen et al., 2012).

Together, these data suggest that depression is a disease of synapses, leading to a breakdown of communication between neurons. Thus, drugs that increase synaptic efficacy and new synapse formation are likely to also ameliorate MDD-associated symptoms. Research regarding the molecular basis of learning and memory provides insight into mechanisms that strengthen synapses and induce structural plasticity; namely, the necessity of local dendritic protein synthesis and repression (for review, see Graber, McCamphill, & Sossin, 2013). These studies provide a new understanding of and insight into the molecular bases and synaptic dysregulations underlying MDD.

MDD Is a mTORopathy

mTOR (mammalian target of rapamycin), a serine/threonine kinase, affects several signaling pathways, including those regulating cell growth, autophagy, and mRNA translation. mTOR is composed of two complexes, complex 1 (C1) and complex 2 (C2). mTORC1 activity regulates local protein synthesis in dendrites, and this specific role is required for long-term increases in synaptic strength, structural plasticity, and cognition (Graber, McCamphill, & Sossin, 2013), all domains that are affected in patients with MDD (Millan et al., 2012; Drevets, Price, & Furey, 2008; Murrough et al., 2011; Kang et al., 2012). For our purposes, we will focus solely on mTORC1.

mTORC1 has two primary targets that affect translation, S6 kinase (S6K1) and eukaryotic initiation factor 4E-binding protein (4E-BP1; see Figure 1, green). S6K1 is activated by phosphorylation, which in turn leads to increased translation. S6K1 regulates ribosomal protein S6, eukaryotic translation initiation factor 4B (eIF4B), and eukaryotic elongation factor 2 kinase (eEF2K; see Figure 1, dark blue). Conversely, 4E-BP1 phosphorylation by mTORC1 prevents it from binding to eIF4E, which enables cap-dependent translation (Hay & Sonenberg, 2004; Laplante & Sabatini, 2009), and animals with disrupted phosphorylation of eIF4E exhibit depressive-like behaviors and dysregulated serotonin (Aguilar-Valles et al., 2018). Previous studies demonstrate that serotonin increases the phosphorylation of 4EBP and activates S6K in an mTORC1-dependent manner in Aplysia neurons (Khan, Pepio, & Sossin, 2001; Carroll, Dyer, & Sossin 2006). Together, these data suggest that serotonin, eIF4E, and 4E-BP1 may participate in a feedback loop (see Figure 1, dark cyan). Not surprisingly, downregulation of the mTORC1 pathway—including mTOR and its downstream targets, p-S6K1, eIF4B, and p-eIF4B—as well as dysregulated serotonergic mechanisms are observed in patients with MDD (Jernigan et al., 2011; Ressler & Nemeroff, 2000). Thus, mTORC1 itself is a promising candidate to target as a means to treat depression.

Dysregulated Protein Synthesis in Major Depressive DisorderClick to view larger

Figure 1. Several pathways influence the activity of the mTORC1 signaling pathway. mTORC1 affects protein synthesis via the inhibition or activation of several pathways. mTORC1 is activated by amino acids but can be inhibited by the compound rapamycin or the tuberous sclerosis complex (TSC). TSC is activated by REDD1 and GSK3. ERK and Akt inhibit TSC. Akt also inhibits GSK3. Akt is activated by PI3K via BDNF-activated TrkB receptors. Here, NMDAR antagonism is being depicted as inhibiting FMRP expression, which enables the translation of GABABR mRNA and de novo protein synthesis of GABABRs. New GABABRs are coupled to L-type calcium channels, which leads to mTORC1 activity. mTORC1 has two primary targets that affect translation, S6 kinase (S6K1 and downstream ribosomal protein S6) and eukaryotic initiation factor 4E-binding protein (4E-BP1), regulating translation initiation (eIF4E and eIF4B) and elongation (eEF2K/eEF2). Serotonin (5HT) can activate S6K in an mTORC1-dependent manner and phosphorylate 4EBP; affecting 4E-BP1 phosphorylation of eIF4E dysregulates 5HT. The mTORC1 pathway is implicated in promoting the synthesis of the synaptic proteins PSD95, Synapsin1 (Syn1), and GluA1, as well as BDNF, which can then feedforward to activate mTORC1 via TrkB receptors.

Red lines with diamond ends indicate inhibition; green lines with arrow heads indicate activation. Dashed lines indicate pathways that have not yet been fully determined.

Nearly 20 years after the discovery of the antidepressant effects of NMDAR antagonists (Trullas et al., 1991), low, subanesthetic doses of ketamine, which rapidly decrease MDD symptoms (Berman et al., 2000; Diazgranados et al., 2010), were discovered to activate mTOR and mTORC1-dependent protein synthesis of several synaptic proteins within 30 minutes of administration (N. Li et al., 2010). Ketamine’s proposed antidepressant mechanism of action helps elucidate the importance of protein translation in treating MDD. Within 30 minutes of exposure, ketamine decreases the phosphorylation of eEF2 (effectively promoting translation) within the hippocampus, and a corresponding increase in brain derived neurotrophic factor (BDNF) expression is also observed (Autry et al., 2011). When animals are pretreated with a protein synthesis inhibitor, ketamine fails to produce an antidepressant-like behavioral effect, suggesting that eEF2-driven translation of BDNF is crucial to the antidepressant-like effects of ketamine (Autry et al., 2011). Interestingly, mTORC1-S6K activity leads to phosphorylation of eEF2K, inhibiting its action of repressing elongation through eEF2 (McCamphill, Ferguson, & Sossin, 2017). These data strongly point toward a role for mTOR relieving repression of elongation of BDNF mRNA. Transgenic animals with altered MAPK-extracellular regulated kinase (ERK)1/2-eIF4E phosphorylation respond similarly to control animals in a number of behavioral domains when treated with ketamine, demonstrating that the mechanism of action for ketamine is MAPK-ERK1/2-independent (Aguilar-Valles et al., 2018). Ketamine also strengthens synaptic connections and induces synaptic complexity within the prefrontal cortex (PFC) of an animal model. Specifically, at a dose that produced antidepressant-like behavioral effects, ketamine increased dendritic tufting, maturation, and excitatory responsiveness of PFC neurons. These crucial findings demonstrate that activation of the mTORC1 protein synthesis pathway underlies the antidepressant efficacy of ketamine, which further emphasizes the critical role of mTORC1-regulated protein synthesis in treating MDD, thus possibly rectifying the mTORC1-related disruptions observed in MDD.

Preclinical Models Provide Further Evidence for mTORC1-Regulated Protein Synthesis as an Effective Treatment of MDD

Preclinical models of depression are extremely valuable, as they allow scientists to probe for (1) the underlying neurobiological substrates of depression and (2) the efficacy of antidepressant medications at the molecular, cellular, and behavioral level. Animal models of depression typically involve exposure to chronic mild stress (CMS), repeated social defeat, or chronic administration of stress hormones. Stressed animals exhibit neuronal abnormalities such as altered dendritic morphology, decreased spine number, and decreased evoked excitatory responsiveness, similar to human patients with MDD (N. Li et al.,; Radley et al., 2006; Radley et al., 2008; R.-J. Liu & Aghajanian, 2008). Further, stress decreases the phosphorylation and activity of the mTORC1 signaling pathway with a corresponding reduction in the expression of synaptic proteins regulated by mTORC1, including PSD95 (postsynaptic density protein 95), Synapsin1, and GluR1 (α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid [AMPA] glutamate receptor subunit 1), similar to what is seen in postmortem tissue from patients with MDD (Feyissa et al., 2009). Importantly, stress-induced reduced protein expression and dendritic/synaptic abnormalities are reversed with the administration of ketamine (N. Li et al., 2011). Moreover, when mTORC1 signaling is blocked with rapamycin (see Figure 1, pink), the positive effects of ketamine at the molecular, cellular, and behavioral level are blocked.

Despite the many data collected by many laboratories, ketamine’s ability to activate mTORC1 was puzzling. These data seemed paradoxical because NMDAR activation is usually required for mTORC1-dependent signaling, synaptic strengthening, and increased spine density (Raab-Graham & Niere, 2017). NMDARs are tetrameric ligand- and voltage-gated cation channels and consist of several subunit types: GluN1, GluN2 (GluN2A-D), and GluN3 (GluN3A–B). Functional receptors require two GluN1 and two GluN2 or GluN3 subunits, with different configurations affecting region specificity, kinetics, open probability, and deactivation time (Paoletti, Bellone, & Zhou, 2013). Interestingly, GluN2A- and GluN2B-containing receptors are most prominent within the hippocampus and PFC (Paoletti, Bellone, & Zhou, 2013). NMDARs are considered “coincidence detectors,” since both conditions of ligand-binding and membrane depolarization must occur to displace a magnesium ion that blocks the ion channel pore so that calcium may enter the cell via the receptor. This calcium then activates a number of signaling cascades, including initiating the mTORC1 translational pathway and synaptic strengthening. Although both GluN2A- and GluN2B-containing receptors generate large calcium currents when activated, GluN2B-mediated currents are longer lasting (Bloodgood, Sabatini, & Van Dongen 2009). In developing synapses, NMDAR-mediated calcium influx activates calcium-dependent eEF2K, which then phosphorylates eEF2 (Scheetz, Nairn, & Constantine-Paton, 2000). This phosphorylation may slow the local rate of protein translation, and GluN2B-mediated calcium may keep eEF2K active longer than GluN2A-mediated conductance, thus leading to greater inhibition of global translation. Elongation, then, instead of initiation, consequently would become the rate-limiting step in protein synthesis.

Ketamine acts to block the channel pore and prevents calcium from entering the cell. The best evidence, however, that the activation of mTORC1 signaling mediates the antidepressant-like effects of ketamine was provided by studies that demonstrated that not all NMDAR antagonists exhibit antidepressant-like effects. For example, memantine, a partial NMDAR antagonist that also blocks the channel pore and which is used to treat Alzheimer’s disease, does not promote antidepressant-like behavioral effects (Zhang et al., 2017; Gideons, Kavalali, & Monteggia, 2014). Memantine appears to more effectively affect GluN2A-containing NMDARs, whereas ketamine is more effective at GluN2B-containing NMDARs (Glasgow et al., 2017), thus suggesting that ketamine acts to decrease the GluN2B-mediated eEF2 phosphorylation, whereas memantine might increase or not alter this effect. Transgenic animals with cortical GluN2B subunits removed exhibit decreased behavioral despair in the tail suspension test (TST) similar to ketamine-treated animals, and these transgenic animals are not affected by ketamine treatment (Miller et al., 2014). Further, other GluN2B-specific antagonists are effective at treating both animal models of MDD and human patients with MDD (Maeng et al., 2008; Preskorn et al., 2008). What is more, memantine does not increase phosphorylation of mTORC1 (Zhang et al., 2017), nor does it affect the phosphorylation of eEF2K like ketamine, thus rendering memantine unable to promote activity of the mTORC1 signaling pathway and protein synthesis (Zhang et al., 2017; Gideons, Kavalali, & Monteggia, 2014). Thus, crucial differences between memantine and ketamine include subunit specificity and the ability to initiate mTORC1-dependent protein synthesis. Together, these data provide strong support for the hypothesis that the antidepressant-like effects of NMDAR antagonists rely on activity of the mTORC1 signaling pathway.

If this hypothesis is correct, then one can predict how altering genes downstream of mTORC1 signaling will affect animal models of MDD. For instance, knocking out one such gene would correspond to dendritic and spine deficits, as well as depressive-like behaviors. Indeed, similar to patients with MDD (Jernigan et al., 2011), genetically knocking down S6K1 in the medial PFC of animals generates depressive-like behaviors in the absence of stress (Dwyer et al., 2015). Reduced S6K1 activity (and thus, decreased activity of the mTORC1 signaling pathway) also blocks the antidepressant-like effects of ketamine on the behavior of these animals. Conversely, overexpression of S6K1 (and, thus, increased activity of the mTORC1 signaling pathway) produces antidepressant-like behaviors and is protective against stress-induced behavioral changes. Also, increased S6K1 activity increases neuronal complexity via neuronal branch number compared to controls. Collectively, these data strongly support the hypothesis that mTORC1 signaling pathway and, thus, mTORC1-dependent protein synthesis is disrupted in MDD. Therefore, endogenous upstream targets that regulate mTORC1 may also be good candidates for pharmacological treatment of MDD.

One such potential target is REDD1 (regulated in development and DNA damage responses 1), an upstream inhibitor of mTORC1 (see Figure 1, red). REDD1 is increased in the dorsolateral PFC of patients with MDD, and CMS increases mRNA and protein expression of REDD1 in the PFC of rats (Ota et al., 2014). These increases are also associated with decreased expression of p-S6K, p-4EBP1, and p-Akt. Further, REDD1 knockout (KO) mice display a stress resilient phenotype wherein stress exposure does not decrease sucrose preference in these animals, nor does stress affect p-S6K or p-4EBP expression in the PFC (Ota et al., 2014). In addition, while stress exposure promotes decreased spine density, this effect is not seen in REDD1 KO mice. In contrast, REDD1 overexpression decreases p-S6K and p-mTOR in the PFC and leads to decreased spine density (Ota et al., 2014). Also, overexpression of REDD1 leads to a pro-depressive phenotype even in the absence of stress. Thus, inhibiting REDD1, and disinhibiting the mTORC1 signaling pathway, may help treat depressive-like symptoms.

Importantly, there are reports of NMDAR antagonists improving performance in behavioral tests of despair, motivation, and self-care, without prior stress. A single administration of ketamine (N. Li et al., 2010) or Ro 25-6981, a GluN2B-specific antagonist (Workman, Niere, & Raab-Graham, 2013), decreases immobility in the forced swim test (FST) and TST, indicating less behavioral despair. Workman et al. (2015) also demonstrated that a single administration of Ro 25-6981 increases grooming frequency in the splash test, a behavior that is a measurement of self-care. Importantly, co-administration of rapamycin, an mTORC1 inhibitor, blocks these behavioral effects. These drugs improve performance in tasks without chronic administration or without prior exposure to stress, which may be a result of rapidly boosting proteins that mediate antidepressant-like effects through mTORC1 activation.

Does Inhibition of mTORC1-Regulated Protein Synthesis Promote MDD?

Rapamycin and rapamycin analogs, like everolimus, are FDA approved treatments for disorders where mTORC1 signaling is overactive (see Figure 1, pink). However, the hypothesis that the mTORC1 signaling pathway is disrupted in MDD predicts that chronic administration of mTORC1 inhibitors may cause depressive-like behaviors. A recent study by Russo et al. (2016) addresses this particular prediction. The authors treated mice with synthetic glucocorticoids (a common method used to activate and increase endogenous stress mechanisms) with or without everolimus in their drinking water and examined performance in measures of cognition, despair, and anxiety. While mice treated with everolimus exhibited enhanced cognitive performance, and rescued the cognitive impairment from the glucocorticoid treatment, the mice also exhibited increased anxiety- and depressive-like behaviors. These data suggest that caution should be exercised with the use of mTOR inhibitors in neurological disorders such as autism spectrum disorder and Alzheimer’s disease, two disorders with comorbid depression (Starkstein et al., 2005; Simonoff et al., 2008).

Where in the Brain Should mTORC1 Be Activated to Treat MDD?

mTOR is ubiquitously expressed throughout the body and overactive mTORC1 is implicated in many diseases, including cancer, autism spectrum disorder, and Alzheimer’s disease (for review, see Siddiqui & Sonenberg, 2015 and Raab-Graham & Niere, 2017). Thus, a potential challenge is to design pharmaceuticals that adjust mTORC1 activity within an optimal range without causing somatic problems. A necessary first step is to determine the brain regions that exhibit decreased mTORC1 activity due to MDD. Preclinical animal models of depression are extremely useful because they exhibit alterations to mTORC1 activity within brain regions that parallel those affected in human patients with MDD, as detailed earlier. However, not all of the data agree on which regions are affected. For example, Chandran et al. (2013) examined whether CMS induces changes to the mTORC1 signaling pathway within the frontal cortex, hippocampus, amygdala, and dorsal raphe. The authors found that only the amygdala showed stress-related effects. In addition to decreased expression of phosphorylated mTOR, p-70S6K, and S6, stress-exposed animals also displayed decreased ERK1/2, and Akt/protein kinase B (Akt; see Figure 1, light blue). Moreover, the authors measured p-GluR1, as an indicator for surface-expressed and thus active glutamate receptors, and observed a parallel reduction in the stressed animals. These data imply that the stress-induced reduction in AMPAR signaling and synaptic efficacy are specific to the amygdala (Chandran et al., 2013). In contrast, animals exposed to the FST exhibit less hippocampal p-mTOR and BDNF compared to animals treated with ketamine, which also decreased immobility in the FST (Yang, Hu, et al., 2013). Consistent with these data, other studies indicate that the PFC and hippocampus are sensitive to the effects of NMDAR antagonist treatment. For instance, in a study that examined proteomic changes between non-stressed rats treated with ketamine, ketamine treatment increased hippocampal expression of mTOR in treated animals compared to controls (Wesseling et al., 2015). Similar to ketamine, treatment with Ro 25-6981 also increases the phosphorylation of mTOR, mTOR-dependent synthesis of BDNF, and GluA1 in the PFC in parallel with reducing behavioral despair, as measured by the FST (Workman, Niere, & Raab-Graham, 2013).

Niere et al. (2016) showed that the largest effect of mTORC1 activity is remodeling protein composition at the synapse. One difference between the studies discussed in the previous paragraph is which neuronal fractions were assayed. Chandran et al. (2013) assayed total cellular lysates, rather than synaptic fractions, of the brain regions they examined. Assessment of the total cell lysate may not be sensitive enough to detect changes to mTORC1 activity apparent at the synapse. These data argue that targeting mTORC1 in specific brain regions, and perhaps only at synapses, may provide the most effective results.

It should also be noted that the target mRNAs translated upon mTORC1 activation may vary across brain regions. For example, chronic inhibition of calcineurin, a calcium and calmodulin-dependent serine/threonine phosphatase, in the PFC promotes depressive-like behaviors in animals and downregulates the mTORC1 signaling pathway (Yu et al., 2013). Ketamine treatment increases the expression of the regulatory subunit of calcineurin, Ppp3cb, in the frontal cortex of rats; however, this treatment also downregulates Ppp3cb in the hippocampus (Wesseling et al., 2015). Further, others have demonstrated that Ppp3cb expression increases in the cortex with mTOR activity (Niere et al., 2016). These data add to the complexity of understanding MDD therapies because, even if mTORC1 is activated both in the PFC and hippocampus, mRNAs may be differentially regulated in MDD and by ketamine.

Targeting Protein Synthesis Pathways via Traditional Antidepressants and Other Treatments

Other drugs with antidepressant-like effects may produce their effects via the mTORC1 signaling pathway and mTORC1-dependent protein synthesis. However, as in the case of typical antidepressants, little is known about their mechanisms of action. Thus, in order to determine whether several typical antidepressants utilize the mTORC1 signaling pathway, Park et al. (2014) examined the effect of chronic treatment on primary rat hippocampal neurons. The authors found that, similar to ketamine, chronic exposure to escitalopram (SSRI), paroxetine (SSRI), and tranylcypromine (MAO inhibitor) increases the phosphorylation of several mTORC1 pathway markers, including mTOR, 4EBP1, and p70S6K. Further, phosphorylation of upstream activators of the mTOR pathway, Akt and ERK, also increased with chronic treatment of these antidepressants. The ability for the antidepressants to increase expression of p-mTOR was blocked with PI3K (phosphoinositide 3 kinase), MEK (mitogen-activated protein kinase kinase, also known as MAP2K), and mTOR inhibitors (see Figure 1, light purple). Similar to ketamine, these antidepressants also increase dendritic outgrowth and branching in an mTORC1-dependent manner. Increases in the synaptic markers PSD95 and synaptophysin were also found to be mTORC1-dependent.

In agreement with these studies, proteomic studies conducted by Shen et al. (2017) sought to determine the molecular mechanisms associated with venlafaxine (SNRI) treatment. The authors administered venlafaxine to mice and then compared differences in hippocampal metabolites between treated and control animals. The authors identified 27 significantly different metabolites via gas chromatography-mass spectrometry (GC-MS), for which they then developed a molecular interaction network. Of the pathways identified, the MAPK-ERK1/2 and PI3K-Akt pathways were highly correlated with the differentially expressed metabolites. Finally, the authors verified via western blotting that targets within each pathway were differentially affected between the control and treated groups. They found significant increases in hippocampal ERK1/2, p-Akt, CREB, BDNF, p-C-Raf, and p-MEK in the venlafaxine-treated animals (Shen et al., 2017).

Several other typical antidepressants, however, exhibit different effects. The antidepressants fluoxetine (SSRI), sertraline (SSRI), and imipramine (tricyclic) increase the phosphorylation of Akt and ERK without affecting mTORC1 or its downstream targets in primary rat hippocampal neurons (Park et al. 2014), agreeing with previous research showing no effect on the phosphorylation of mTOR, 4E-BP1, or p-S6K in the PFC (N. Li et al., 2010). It should be noted that these antidepressants also increase dendritic complexity and synaptic markers, but these results are unaffected by mTORC1 inhibition (Park et al., 2014). Also, there may be discrete regional effects of these antidepressants. For example, chronic in vivo fluoxetine administration (21 days) increases phosphorylation of both eIF4E and eEF2 within the dentate gyrus. eEF2 phosphorylation was also increased within the hippocampus and PFC (Dagestad et al., 2006). While the conclusions from Park et al. (2014) are limited because they only investigated the effects of these antidepressants on the mTORC1 pathway in cultured hippocampal neurons, the authors suggest that these three antidepressants may activate CREB and BDNF via the ERK pathway and eEF2 in an mTORC1-independent manner to increase synaptic markers and dendritic complexity. However, transgenic animals with altered MAPK-ERK1/2-eIF4E phosphorylation do not exhibit a different response to fluoxetine treatment compared to controls, thus suggesting that fluoxetine does not affect eIF4E through the MAPK-ERK1/2 pathway (Aguilar-Valles et al., 2018). As mentioned earlier, because ketamine has a high abuse potential, it is not prudent to utilize this drug to treat a disorder that is marked with high comorbid substance use disorders (Kessler et al., 2003). Thus, researchers are continuously investigating other drugs that may similarly affect the mTORC1 signaling pathway and produce antidepressant-like effects. Dwyer et al. (2012) demonstrate that the mGluR2/3 antagonist LY 341495 also increases mTORC1 signaling targets and synaptic proteins in PFC synaptoneurosomes similar to ketamine. Further, LY 341495 decreases immobility in the FST in an mTORC1-dependent manner. Similar increases to synaptic proteins within hippocampal synaptoneurosomes are reported; however, not all of the same effects seen in the PFC are evident in this hippocampal preparation (Dwyer, Lepack, &Duman, 2012). These differences could reflect either region-specific effects of the stress or drug treatment.

Another avenue researchers are exploring is augmenting the effects of low, sub-effective doses of ketamine with other drugs that lack abuse potential. Co-administration of GSK-3β (a serine/threonine kinase upstream of mTORC1) inhibitors with sub-effective doses of ketamine decreases immobility in the FST, similar to the effects of higher ketamine doses (R.-J. Liu et al., 2013; see Figure 1, dark red). In addition, lithium administered with ketamine increases phosphorylation of PFC synaptosomal GSK-3β, mTORC1 signaling pathway markers (mTOR, p-S6K), as well as Akt and ERK, with some changes lasting up to 24 hours post-administration. Lithium and ketamine co-administration also increases synaptic responses and dendritic complexity (R.-J. Liu et al., 2013).

Interestingly, while most accept that ketamine drives its antidepressant-like effect via NMDAR antagonism, some data suggest that ketamine’s active metabolite, (2S,6S;2R,6R)-hydroxynorketamine (HNK), mediates the drug’s antidepressant properties (Kavalali & Monteggia, 2018). Zanos et al. (2016) demonstrate that administration of HNK promotes antidepressant-like behavioral effects in mice in an NMDAR inhibition-independent manner. The authors also report that neither ketamine nor HNK administration affected phosphorylation of mTOR within hippocampal or PFC synaptosomes, unlike previous reports detailed earlier. Further, they demonstrated that both ketamine and HNK decrease hippocampal phosphorylation of eEF2, but increase expression of BDNF, GluA1, and GluA2. Again, in contrast to others’ findings, increased expression of these proteins occurred only in hippocampal synaptosomes and not PFC, and only at 24 hours post-administration (Zanos et al., 2016). Importantly, the authors suggest that HNK does not have the adverse side effects of ketamine. However, independent groups suggest that HNK can block NMDARs and have been unable to replicate the antidepressant-like effects of HNK on behavior (Suzuki et al., 2017; Yang et al., 2017; Shirayama & Hashimoto, 2018).

The mTORC1 pathway is also activated by amino acids (see Figure 1, dark purple). Baranyi et al. (2016) investigated whether patients with MDD exhibited any changes to branched chain amino acids (BCAA) in their plasma. They found that compared to controls, patients with MDD have significantly lower levels of the BCAAs valine, leucine, and isoleucine, which would lead to decreased activation of the mTORC1 signaling pathway and protein synthesis. Further, the authors report a negative correlation between the concentration of these BCAAs and scores on self-report and observer-based psychiatric assessments. These data corroborate previous research that identified an increase in amino acids after chronic antidepressant treatment in animals (Webhofer et al., 2011).

Exercise, which is often touted as a non-pharmaceutical treatment of MDD, also rescues deficits in the mTORC1 signaling pathway after stress. Fang et al. (2013) demonstrate that immobilization stress reduces BDNF, phosphorylation of TrkB, Akt, GSK-3β, mTOR, and p-70S6K within the hippocampus. Additionally, the authors found decreased expression of the synaptic markers synaptophysin, PSD95, b-neurexin, and neuroligin in response to stress. However, treadmill exercise reversed the effect of stress on these markers, providing some evidence for the effectiveness of exercise as an MDD treatment.

Collectively, these data argue that, regardless of the upstream signal, the key unifying target for relief of MDD symptoms is the activation of protein synthesis pathways that increase synaptic proteins and synaptic strength.

Targeting mRNA Translational Repression Factors to Mitigate MDD

RNA-binding proteins (RBPs) and microRNAs (miRs), small noncoding RNAs, scan the transcriptome to silence mRNA translation of specific transcripts (Raab-Graham & Niere, 2017). Thus, manipulating levels of translational repressors may be an additional avenue to alleviate depression. The following sections describe recent reports indicating the roles of translational repressors in MDD and how they may be targeted to reverse depressive-like symptoms.

Establishing the Function of the RBP Fragile X Mental Retardation Protein in MDD

Fragile X mental retardation protein (FMRP) is one of the most well characterized RBPs. FMR1 premutation carriers, people without fragile X syndrome (FXS) but who are likely to have children with FXS, have higher rates of MDD and depressed mood (Johnston et al., 2001; Roberts et al., 2009). These data suggest that insufficient levels of FMRP can lead to MDD. Consistent with this finding, patients with MDD express reduced levels of FMRP in the lateral cerebellum (Fatemi et al., 2013). Recent work characterizing the antidepressant pathway of NMDAR antagonists demonstrates that de novo protein synthesis of the metabotropic inhibitory neurotransmitter receptor gamma-aminobutyric acid B (GABABR) drives the activation of mTORC1 necessary for antidepressant effects (Workman, Niere, & Raab-Graham, 2013; Workman et al., 2015; see Figure 1). GABABR mRNA is a target of FMRP (Darnell et al., 2011) and Fmr1 KO mice exhibit increased basal levels of GABABR protein in hippocampal synaptosomes (Wolfe et al., 2016). NMDAR antagonists promote the synthesis of new GABABRs and their surface expression (Workman, Niere, & Raab-Graham, 2013; Workman et al., 2015; Wolfe et al., 2016), however, this antidepressant-mediated effect is absent in cultured primary hippocampal neurons from Fmr1 KO mice (Wolfe et al., 2016). Further, the antidepressant-like effects associated with NMDAR antagonists require GABABR activation (Workman, Niere, & Raab-Graham, 2013; Workman et al., 2015) and is absent in Fmr1 KO mice (Wolfe et al., 2016). Thus, understanding how NMDAR antagonism promotes the translation of FMRP-targeted mRNAs may lead to new avenues of MDD treatments.

Targeting the miR Landscape and Downstream Protein Synthesis in the Brain to Treat MDD

Many drugs alter the expression of miRs, usually resulting in the inverse expression of their mRNA targets. Importantly, many miRs are specific to the brain, and specifically targeting them avoids the problem of somatic side effects associated with mTOR activity. One of the best examples of this interaction is demonstrated by Lopez et al. (2014). First, the authors demonstrated that the expression of miR-1202 is decreased in the ventrolateral PFC in patients with MDD compared to controls. Next, they showed that miR-1202 is primate-specific and enriched in the brain compared to other tissues. Further, the mRNA of a predicted target of miR-1202, metabotropic glutamate receptor 4 (GRM4), was demonstrated to be negatively correlated with the expression of miR-1202 within the PFC. The authors then investigated the effect of antidepressants on the expression of miR-1202 and GRM4 in patients with MDD. Regardless of antidepressant use, miR-1202 was decreased in patients with MDD compared to controls; however, within the MDD group, miR-1202 was increased in patients being treated with antidepressants. Further, GRM4 was upregulated in brain tissue from patients with MDD who were not being treated with antidepressants. Importantly, however, antidepressant use mitigated differences in GRM4 expression between controls and MDD patients. The authors went on to demonstrate a bidirectional interaction between miR-1202 and GRM4, and found that only antidepressants that have direct effects on serotonin or serotonin transporters regulate miR-1202 expression. Finally, they demonstrated that miR-1202 expression was decreased in MDD patients who were treatment naïve compared to controls. Importantly, in patients who were treated and whose depression was alleviated, miR-1202 expression was increased. They further found a negative correlation between miR-1202 expression and severity of MDD symptoms. All together, these data strongly support targeting miR-1202 as a treatment for MDD.

Higuchi et al. (2016) demonstrated that chronic ultra-mild stress (similar to chronic mild stress, but without food or water deprivation, and no nociceptive events) decreased expression of miR-124, which was rescued with chronic treatment of imipramine. Hippocampal overexpression of miR-124 increased resilience to stress, whereas inhibition led to greater reactivity to mild stress. To further demonstrate the protective role of miR-124, overexpression of miR-124 prevented stress-induced increases in GSK-3β mRNA and protein. In contrast, Bahi et al. (2014) found that socially defeated rats express increased hippocampal miR-124 and downregulated BDNF, a direct target of miR-124. These authors also found that overexpression of miR-124 in the hippocampus increased depressive-like behaviors, whereas inhibition or infusion of BDNF decreased these behaviors. Such findings emphasize the heterogeneity of MDD and suggest that finding specific biomarkers that would indicate effectiveness of specific antidepressant treatment is desperately needed.

Detection of Peripheral Biomarkers

Biomarkers for MDD are needed, but how and what to assess in patients with MDD has remained elusive. However, in a letter to the editor, Yang et al. (2013) describe peripheral changes to mTORC1 signaling targets after administering ketamine to three patients with MDD. Over the course of 2 hours, p-mTOR, p-GSK-3β, and p-eEF2 increased above baseline, whereas scores on self-reported and observer-based measures of depressive symptoms steadily decreased. Interestingly, Li et al. (2013) found a significant increase in GSK-3β mRNA in blood from patients with MDD prior to antidepressant treatment, compared to controls. This increase was brought to control baseline levels with 8 weeks of escitalopram (SSRI) treatment. These results are unexpected and promising. The phosphorylation state of these proteins is a good readout for mTORC1 activity and suggests that increased activity of mTORC1 in the brain can be also observed in the blood. In contrast, mRNA levels do not always correlate with protein levels, perhaps explaining why Li et al. (2013) detected less GSK-3β mRNA in the blood while Yang et al. (2013) detected increased levels of the active form of the protein. Nonetheless, these data are promising and thus provide potential markers to screen for antidepressant efficacy in humans.

Importantly, miRs that are dysregulated in the brain can often be detected peripherally. miR-132 represses the expression of BDNF and MeCP2, an important enzyme that methylates DNA and silences transcription (Su et al., 2015; Zimmermann et al., 2015). Thus, changes in miR-132 in the brain are likely to have short-term (via BDNF mRNA repression) and long-term effects (MeCP2) on depression. A recent study examined the relationship between the expression of BDNF, MeCP2, and miR-132 in the blood of patients with MDD (Su et al., 2015). They found an increase in miR-132 and a corresponding decreased BDNF and MeCP2, as would be predicted. To further verify the relationship between BDNF, MeCP2, and miR-132, Su et al. (2015) examined changes to these targets by using the chronic unpredictable stress animal model. They found that stress increased miR-132 expression in the hippocampus but decreased total protein expression of MeCP2 and BDNF, similar to what was observed in the blood from patients with MDD. Capitalizing on an animal model, they overexpressed miR-132, which further decreased MeCP2 and BDNF expression. In contrast, when they downregulated miR-132, there was a corresponding increase in MeCP2 and BDNF.

More recently, Lopez et al. (2017) examined the effects of treating MDD with desvenlafaxine (SNRI) on peripheral miR-1202 and neural activity associated with a Go/NoGo task. They compared performance and miR-1202 expression prior to antidepressant treatment and after 8 weeks of treatment. The researchers found that after treatment, peripheral miR-1202 levels were correlated with altered network connectivity in task-based and resting-state activity, suggesting that desvenlafaxine may modulate network activity through the glutamatergic system.

Assessment of peripheral miRs may help design personalized treatment for depression. For example, Issler et al. (2014) found that overexpression of miR-135a in serotonergic neurons promotes stress resiliency and altered serotonin metabolism, but also prevents stress-related reductions to serotonin. Further, they found that miR-135a is decreased peripherally, as well as in the dorsal raphe nucleus in patients with MDD and in suicide victims. While SSRI treatment did not affect peripheral miR-135, the authors demonstrate that cognitive behavioral therapy increased miR-135 expression. The authors did not report efficacy of either treatment in decreasing symptoms associated with MDD. However, these studies do raise the question of whether peripheral detection of miRs can predict which treatment course will be most effective for individuals with MDD. Together, these data suggest that miRs are useful peripheral biomarkers for MDD, and that further validation of miRs that are dysregulated in MDD will help researchers identify potential treatments for MDD. In addition, identifying miRs and their targets that are affected in MDD will help researchers further understand the underlying etiology of MDD.

Conclusion

Together, the data presented here demonstrate that effective antidepressant treatments influence protein synthesis, either by directly targeting protein synthesis pathways (such as the mTORC1 signaling pathway) or by targeting translation repressors (such as RNA binding proteins or microRNAs). What is more, these treatments aim to increase synaptic functioning and appear to combat the loss of neural volume and disrupted circuitry. One remaining issue, however, is the delayed onset of efficacy of traditional antidepressants. Researchers are currently studying the mechanisms behind ketamine’s rapid effects in order to develop new therapeutics that are both fast-acting and non-addictive.

Future Directions

  1. 1. Developing effective targets for treatment that avoid abuse potential and negative side effects by specifically targeting brain regions and mapping how antidepressants work in specific regions and on which transcripts.

  2. 2. Because MDD is heterogenous, developing personalized therapeutics with programs like the Connectivity Map (CMAP) may help optimize treatment based on mRNA profiles rather than general chemical transduction.

  3. 3. Using membrane potential as a guide for treatment, since many ion channels are expressed only in the brain. Pharmaceutical companies have designed many drugs that target specific ion channels; could these drugs be repurposed?

  4. 4. Some populations with comorbid MDD may be unresponsive to current treatments. For example, people with FXS lack FMRP, which is required for the effects of rapid antidepressants with NMDAR antagonists. The neurobiology of MDD in these populations need to be studied independently.

  5. 5. MDD that appears during adolescence may arise differently than in adults. This population may also need to be studied independently in order to determine how the etiology of MDD is different, and whether different drug treatment strategies are needed.

  6. 6. People with MDD are more likely to also develop a substance abuse disorder. Is this comorbidity due to people with MDD self-medicating with alcohol and/or illicit drugs, or does MDD, the disorder itself, affect neurobiology in such a way to make people more susceptible to addiction? Can mechanisms that underlie antidepressant efficacy provide insight into treating substance abuse disorders?

References

Aguilar-Valles, A., Haji, N., De Gregorio, D., Matta-Camacho, E., Eslamizade, M. J., Popic, J., Sharma, V., et al. (2018). Translational control of depression-like behavior via phosphorylation of eukaryotic translation initiation factor 4E. Nature Communications 9 (1): 1–14. Article No. 2459. doi:10.1038/s41467-018-04883-5.Find this resource:

Autry, A. E., Adachi, M., Nosyreva, E., Na, E. S., Los, M. F., Cheng, P. F., … Monteggia, L. M. (2011). NMDA receptor blockade at rest triggers rapid behavioural antidepressant responses. Nature 475 (7354): 91–95. doi:10.1038/nature10130.Find this resource:

Bahi, A., Chandrasekar, V., & Dreyer, J.-L. (2014). Selective lentiviral-mediated suppression of microRNA124a in the hippocampus evokes antidepressants-like effects in rats. Psychoneuroendocrinology 46 (August): 78–87. doi:10.1016/j.psyneuen.2014.04.009.Find this resource:

Baranyi, A., Amouzadeh-Ghadikolai, O., von Lewinski, D., Rothenhäusler, H.-B., Theokas, S., Robier, C., … Meinitzer, A. (2016). Branched-chain amino acids as new biomarkers of major depression: A novel neurobiology of mood disorder. PLoS One 11 (8): e0160542. doi:10.1371/journal.pone.0160542.Find this resource:

Bearden, C. E., Thompson, P. M., Avedissian, C., Klunder, A. D., Nicoletti, M., Dierschke, N., … Soares, J. C. (2009). Altered hippocampal morphology in unmedicated patients with major depressive illness. ASN Neuro 1 (4): AN20090026. doi:10.1042/AN20090026.Find this resource:

Berman, R. M., Cappiello, A., Anand, A., Oren, D. A., Heninger, G. R., Charney, D. S., & Krystal, J. H. (2000). Antidepressant effects of ketamine in depressed patients. Biological Psychiatry 47 (4): 351–354. doi:10.1016/S0006-3223(99)00230-9.Find this resource:

Bloodgood, B. L., Bernardo, L., Sabatini, & Van Dongen, A. M. (2009). NMDA receptor-mediated calcium transients in dendritic spines. In Antonius M. Van Dongen (Ed.), Biology of the NMDAR Receptor. Boca Raton, FL: CRC Press/Taylor & Francis.Find this resource:

Carroll, M., Dyer, J., & Sossin, W. S. (2006). Serotonin increases phosphorylation of synaptic 4EBP through TOR, but eukaryotic initiation factor 4E levels do not limit somatic cap-dependent translation in Aplysia neurons. Molecular and Cellular Biology 26 (22): 8586–8598. doi:10.1128/MCB.00955–06.Find this resource:

Chandran, A., Iyo, A. H., Jernigan, C. S., Legutko, B., Austin, M. C., & Karolewicz, B. (2013). Reduced phosphorylation of the mTOR signaling pathway components in the amygdala of rats exposed to chronic stress. Progress in Neuro-Psychopharmacology and Biological Psychiatry 40 (January): 240–245. doi:10.1016/j.pnpbp.2012.08.001.Find this resource:

Dagestad, G., Kuipers, S. D., Messaoudi, E., & Bramham, C. R. (2006). Chronic fluoxetine induces region-specific changes in translation factor eIF4E and eEF2 activity in the rat brain. European Journal of Neuroscience 23 (10).: 2814–2818. doi:10.1111/j.1460–9568.2006.04817.x.Find this resource:

Darnell, J. C., Van Driesche, S. J., Zhang, C., Hung, K. Y. S., Mele, A., Fraser, C. E., Stone, E. F., et al. (2011). FMRP stalls ribosomal translocation on mRNAs linked to synaptic function and autism. Cell 146 (2): 247–261. doi:10.1016/j.cell.2011.06.013.Find this resource:

Diazgranados, N., Lobna, A. I., Brutsche, N. E, Ameli, R., Henter, I. D., Luckenbaugh, D. A., Machado-Vieira, R., & Zarate, C. A. (2010). Rapid resolution of suicidal ideation after a single infusion of an N-Methyl-D-Aspartate antagonist in patients with treatment-resistant major depressive disorder. Journal of Clinical Psychiatry 71 (12): 1605–1611. doi:10.4088/JCP.09m05327blu.Find this resource:

Drevets, W. C., Price, J. L., & Furey, M. L. (2008). Brain structural and functional abnormalities in mood disorders: implications for neurocircuitry models of depression. Brain Structure & Function 213 (1–2): 93–118. doi:10.1007/s00429-008-0189-x.Find this resource:

Dwyer, J. M., Lepack, A. E., & Duman, R. S. (2012). mTOR activation is required for the antidepressant effects of mGluR2/3 blockade. International Journal of Neuropsychopharmacology 15 (4): 429–434. doi:10.1017/S1461145711001702.Find this resource:

Dwyer, J. M., Maldonado-Avilés, J. G., Lepack, A. E. DiLeone, R. J., & Duman, R. S. (2015). “Ribosomal Protein S6 Kinase 1 signaling in prefrontal cortex controls depressive behavior.” Proceedings of the National Academy of Sciences 112 (19): 6188–6193. doi:10.1073/pnas.1505289112.Find this resource:

Fang, Z. H., Lee, C. H., Seo, M. K. Cho, H., Lee, J. G., Lee, J. B., Park, S. W., & Kim, Y. H. (2013). Effect of treadmill exercise on the BDNF-mediated pathway in the hippocampus of stressed rats. Neuroscience Research 76 (4): 187–194. doi:10.1016/j.neures.2013.04.005.Find this resource:

Fatemi, S. H., Folsom, T. D., Rooney, R. J., & Thuras, P. D. (2013). mRNA and protein expression for novel GABAA receptors Θ and Ρ2 are altered in schizophrenia and mood disorders: Relevance to FMRP-mGluR5 signaling pathway. Translational Psychiatry 3 (6): e271. doi:10.1038/tp.2013.46.Find this resource:

Feyissa, A. M., Chandran, A., Stockmeier, C. A., & Karolewicz, B. (2009). Reduced levels of NR2A and NR2B subunits of NMDA receptor and PSD-95 in the prefrontal cortex in major depression. Progress in Neuro-Psychopharmacology and Biological Psychiatry 33 (1): 70–75. doi:10.1016/j.pnpbp.2008.10.005.Find this resource:

Genzel, L., Dresler, M., Cornu, M., Jäger, E., Konrad, B., Adamczyk, M., Friess, E., … Goya-Maldonado, R. (2015). Medial prefrontal-hippocampal connectivity and motor memory consolidation in depression and schizophrenia. Biological Psychiatry 77 (2): 177–186. doi:10.1016/j.biopsych.2014.06.004.Find this resource:

Gideons, E. S., Kavalali, E. T., & Monteggia, L. M. (2014). Mechanisms underlying differential effectiveness of memantine and ketamine in rapid antidepressant responses. Proceedings of the National Academy of Sciences of the United States of America 111 (23): 8649–8654. doi:10.1073/pnas.1323920111.Find this resource:

Glasgow, N. G., Povysheva, N. V., Azofeifa, A. M., & Johnson, J. W. (2017). Memantine and ketamine differentially alter NMDA receptor desensitization. Journal of Neuroscience 37 (40): 9686–9704. doi:10.1523/JNEUROSCI.1173–17.2017.Find this resource:

Graber, T. E., McCamphill, P. K., & Sossin, W. S. (2013). A recollection of mTOR signaling in learning and memory. Learning & Memory 20 (10): 518–530. doi:10.1101/lm.027664.112.Find this resource:

Hay, N., & Sonenberg, N. (2004). Upstream and downstream of mTOR. Genes & Development 18 (16): 1926–1945. doi:10.1101/gad.1212704.Find this resource:

Hieronymus, F., Nilsson, S., & Eriksson, E. (2016). A mega-analysis of fixed-dose trials reveals dose-dependency and a rapid onset of action for the antidepressant effect of three selective serotonin reuptake inhibitors. Translational Psychiatry 6 (6): e834–34. doi:10.1038/tp.2016.104.Find this resource:

Higuchi, F., Uchida, S., Yamagata, H., Abe-Higuchi, N., Hobara, T., Hara, K., Kobayashi, A., et al. (2016). Hippocampal microRNA-124 enhances chronic stress resilience in mice. Journal of Neuroscience 36 (27): 7253–7267. doi:10.1523/JNEUROSCI.0319–16.2016.Find this resource:

Issler, O., Haramati, S., Paul, E. D., Maeno, H., Navon, I., Zwang, R., Gil, S., et al. (2014). MicroRNA 135 is essential for chronic stress resiliency, antidepressant efficacy, and intact serotonergic activity. Neuron 83 (2): 344–360. doi:10.1016/j.neuron.2014.05.042.Find this resource:

Jernigan, C. S., Goswami, D. B., Austin, M. C., Iyo, A. H., Chandran, A., Stockmeier, C. A., & Karolewicz, B. (2011). The mTOR signaling pathway in the prefrontal cortex is compromised in major depressive disorder. Progress in Neuro-Psychopharmacology and Biological Psychiatry 35 (7): 1774–1779. doi:10.1016/j.pnpbp.2011.05.010.Find this resource:

Johnston, C., Eliez, S., Dyer-Friedman, J., Hessl, D., Glaser, B., Blasey, C., Taylor, A., & Reiss, A. (2001). Neurobehavioral phenotype in carriers of the Fragile X premutation. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics 103 (4): 314–319.Find this resource:

Kang, H.-J., Voleti, B., Hajszan, T., Rajkowska, G., Stockmeier, C. A., Licznerski, P., Lepack, A., et al. (2012). Decreased expression of synapse-related genes and loss of synapses in major depressive disorder. Nature Medicine 18 (9): 1413–1417. doi:10.1038/nm.2886.Find this resource:

Kavalali, E. T., & Monteggia, L. M. (2018). The ketamine metabolite 2R,6R-hydroxynorketamine blocks NMDA receptors and impacts downstream signaling linked to antidepressant effects. Neuropsychopharmacology 43 (1): 221–222. doi:10.1038/npp.2017.210.Find this resource:

Kessler, R. C., Berglund, P., Demler, O., Jin, R., Koretz, D., Merikangas, K. R., … Wang, S. (2003). The epidemiology of major depressive disorder: Results from the National Comorbidity Survey Replication (NCS-R). Journal of the American Medical Association 289 (23): 3095–3105. doi:10.1001/jama.289.23.3095.Find this resource:

Khan, A., Pepio, A. M., & Sossin, W. S. (2001). Serotonin activates S6 kinase in a rapamycin-sensitive manner in Aplysia Synaptosomes. Journal of Neuroscience 21 (2): 382–391. doi:10.1523/JNEUROSCI.21-02-00382.2001.Find this resource:

Koolschijn, P. C. M. P., van Haren, N. E. M., Lensvelt-Mulders, G., Hulshoff Pol, H. E., & Kahn, R. S. (2009). Brain volume abnormalities in major depressive disorder: A meta-analysis of magnetic resonance imaging studies. Human Brain Mapping 30 (11): 3719–3735. doi:10.1002/hbm.20801.Find this resource:

Laplante, M., & Sabatini, D. M. (2009). mTOR signaling at a glance. Journal of Cell Science 122 (Pt 20): 3589–3594. doi:10.1242/jcs.051011.Find this resource:

Li, N., Lee, B., Liu, R.-J., Banasr, M., Dwyer, J. M., Iwata, M., Li, X. Y., … Duman, R. S. (2010). mTOR-dependent synapse formation underlies the rapid antidepressant effects of NMDA antagonists. Science 329 (5994): 959–964. doi:10.1126/science.1190287.Find this resource:

Li, N., Liu, R. J., Dwyer, J. M., Banasr, M., Lee, B., Son, H.,& … Duman, R. S. (2011). Glutamate N-Methyl-D-Aspartate receptor antagonists rapidly reverse behavioral and synaptic deficits caused by chronic stress exposure. Biological Psychiatry 69 (8): 754–761. doi:10.1016/j.biopsych.2010.12.015.Find this resource:

Li, S.-X., Liu, L.-J., Xu, L.-Z., Gao, L., Wang, X.-F., Zhang, J.-T., & Lu, L. (2013). Diurnal alterations in circadian genes and peptides in major depressive disorder before and after escitalopram treatment. Psychoneuroendocrinology 38 (11): 2789–2799. doi:10.1016/j.psyneuen.2013.07.009.Find this resource:

Liston, C., Miller, M. M., Goldwater, D. S., Radley, J. J., Rocher, A. B., Hof, P. R, … Bruce S McEwen. 2006. Stress-induced alterations in prefrontal cortical dendritic morphology predict selective impairments in perceptual attentional set-shifting. Journal of Neuroscience 26 (30): 7870–7874. doi:10.1523/JNEUROSCI.1184–06.2006.Find this resource:

Liu, R.-J., & Aghajanian, G. K. (2008). Stress blunts serotonin- and hypocretin-evoked EPSCs in prefrontal cortex: Role of corticosterone-mediated apical dendritic atrophy. Proceedings of the National Academy of Sciences of the United States of America 105 (1): 359–364. doi:10.1073/pnas.0706679105.Find this resource:

Liu, R.-J., Fuchikami, M., Dwyer, J. M., Lepack, A. E., Duman, R. S., & Aghajanian, G. K. (2013). GSK-3 inhibition potentiates the synaptogenic and antidepressant-like effects of subthreshold doses of ketamine. Neuropsychopharmacology 38 (11): 2268–2277. doi:10.1038/npp.2013.128.Find this resource:

Liu, Y., Lin, D., Wu, B., & Zhou, W. (2016). Ketamine abuse potential and use disorder. Brain Research Bulletin 126 (Pt 1): 68–73. doi:10.1016/j.brainresbull.2016.05.016.Find this resource:

Lopez, J. P., Lim, R., Cruceanu, C., Crapper, L., Fasano, C., Labonte, B., Maussion, G., et al. 2014. miR-1202 is a primate-specific and brain-enriched microRNA involved in major depression and antidepressant treatment. Nature Medicine 20 (7): 764–768. doi:10.1038/nm.3582.Find this resource:

Lopez, J. P., Pereira, F., Richard-Devantoy, S., Berlim, M., Chachamovich, E., Fiori, L. M., … Jollant, F. (2017). Co-variation of peripheral levels of mir-1202 and brain activity and connectivity during antidepressant treatment. Neuropsychopharmacology 42 (10): 2043–2051. doi:10.1038/npp.2017.9.Find this resource:

Maeng, S., Zarate, C. A., Du, J., Schloesser, R. J., McCammon, J., Chen, G., & Manji, H. K. (2008). Cellular mechanisms underlying the antidepressant effects of ketamine: role of alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid receptors. Biological Psychiatry 63 (4): 349–352. doi:10.1016/j.biopsych.2007.05.028.Find this resource:

McCamphill, P. K., Ferguson, L., & Sossin, W. S. (2017). A decrease in eukaryotic elongation factor 2 phosphorylation is required for local translation of sensorin and long-term facilitation in Aplysia.” Journal of Neurochemistry 142 (2): 246–259. doi:10.1111/jnc.14030.Find this resource:

Millan, M. J., Agid, Y., Brüne, M., Bullmore, E. T., Carter, C. S., Clayton, N. S., Connor, R. et al. (2012). Cognitive dysfunction in psychiatric disorders: Characteristics, causes and the quest for improved therapy. Nature Reviews Drug Discovery 11 (2): 141–168. doi:10.1038/nrd3628.Find this resource:

Miller, O. H., Yang, L., Wang, C.-C., Hargroder, E. A., Zhang, Y., Delpire, E., & Hall, B. J. (2014). GluN2B-containing NMDA receptors regulate depression-like behavior and are critical for the rapid antidepressant actions of ketamine. eLife 3 (October): 513. doi:10.7554/eLife.03581.Find this resource:

Murrough, J. W. (2012). Ketamine as a novel antidepressant: From synapse to behavior. Clinical Pharmacology & Therapeutics 91 (2): 303–309. doi:10.1038/clpt.2011.244.Find this resource:

Murrough, J. W., Iacoviello, B., Neumeister, A., Charney, D. S., & Iosifescu, D. V. (2011). Cognitive dysfunction in depression: Neurocircuitry and new therapeutic strategies. Neurobiology of Learning and Memory 96 (4): 553–563. doi:10.1016/j.nlm.2011.06.006.Find this resource:

Niere, F., Namjoshi, S., Song, E., Dilly, G. A., Schoenhard, G., Zemelman, B. V., … Raab-Graham, K. F. (2016). Analysis of proteins that rapidly change upon mechanistic/mammalian target of rapamycin complex 1 (mTORC1) repression identifies Parkinson protein 7 (PARK7) as a novel protein aberrantly expressed in tuberous sclerosis complex (TSC). Molecular & Cellular Proteomics 15 (2): 426–444. doi:10.1074/mcp.M115.055079.Find this resource:

Ota, K. T., Liu, R.-J., Voleti, B., Maldonado-Avilés, J. G., Duric, V., Iwata, M., Dutheil, S., et al. 2014. REDD1 is essential for stress-induced synaptic loss and depressive behavior. Nature Medicine 20 (5): 531–535. doi:10.1038/nm.3513.Find this resource:

Paoletti, P., Bellone, C., & Zhou, Q. (2013). NMDA receptor subunit diversity: Impact on receptor properties, synaptic plasticity and disease. Nature Reviews Neuroscience 14 (6): 383–400. doi:10.1038/nrn3504.Find this resource:

Park, S. W., Lee, J. G., Seo, M. K., Lee, C. H., Cho, H. Y., Lee, B. J., Seol, W., & Kim, Y. H. (2014). Differential effects of antidepressant drugs on mTOR Signalling in rat hippocampal neurons. International Journal of Neuropsychopharmacology 17 (11): 1831–1846. doi:10.1017/S1461145714000534.Find this resource:

Preskorn, S. H., Baker, B., Kolluri, S., Menniti, F. S., Krams, M., & Landen, J. W. (2008). An innovative design to establish proof of concept of the antidepressant effects of the NR2B subunit selective N-Methyl-D-Aspartate antagonist, CP-101,606, in patients with treatment-refractory major depressive disorder. Journal of Clinical Psychopharmacology 28 (6): 631–637. doi:10.1097/JCP.0b013e31818a6cea.Find this resource:

Raab-Graham, K. F., & Niere, F. (2017). mTOR referees memory and disease through mRNA repression and competition. FEBS Letters 591 (11): 1540–1554. doi:10.1002/1873-3468.12675.Find this resource:

Radley, J. J., Rocher, A. B., Miller, M., Janssen, W. G. M., Liston, C., Hof, P. R., … Morrison, J. H. (2006). Repeated stress induces dendritic spine loss in the rat medial prefrontal cortex. Cerebral Cortex 16 (3): 313–320. doi:10.1093/cercor/bhi104.Find this resource:

Radley, J. J., Rocher, A. B., Rodriguez, A., Ehlenberger, D. B., Dammann, M., McEwen, B. S., … Hof, P. R. (2008). Repeated stress alters dendritic spine morphology in the rat medial prefrontal cortex. Journal of Comparative Neurology 507 (1): 1141–50. doi:10.1002/cne.21588.Find this resource:

Ressler, K. J., & Nemeroff, C. B. (2000). Role of serotonergic and noradrenergic systems in the pathophysiology of depression and anxiety disorders. Depression and Anxiety 12 (Suppl. 1): 2–19. doi:10.1002/1520-6394(2000)12:1+<2::AID-DA2>3.0.CO;2-4.Find this resource:

Roberts, J. E., Bailey, D. B., Mankowski, J., Ford, A., Sideris, J., Weisenfeld, L. A., … Golden, R. N. (2009). Mood and anxiety disorders in females with the FMR1 premutation. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics 150B (1): 130–139. doi:10.1002/ajmg.b.30786.Find this resource:

Rush, A. J., Trivedi, M. H., Stewart, J. W., Nierenberg, A. A., Fava, M., Kurian, B. T., Warden, D., et al. (2011). Combining medications to enhance depression outcomes (co-med): acute and long-term outcomes of a single-blind randomized study. American Journal of Psychiatry 168 (7): 689–701. doi:10.1176/appi.ajp.2011.10111645.Find this resource:

Russo, E., Leo, A., Crupi, R., Aiello, R., Lippiello, P., Spiga, R., et al. (2016). Everolimus improves memory and learning while worsening depressive- and anxiety-like behavior in an animal model of depression. Journal of Psychiatric Research 78 (July): 1–10. doi:10.1016/j.jpsychires.2016.03.008.Find this resource:

Scheetz, A. J., Nairn, A. C., & Constantine-Paton, M. (2000). NMDA receptor-mediated control of protein synthesis at developing synapses. Nature Neuroscience 3 (3): 211–216. doi:10.1038/72915.Find this resource:

Shen, P., Hu, Q., Dong, M., Bai, S., Liang, Z., Chen, Z., Li, P., et al. (2017). Venlafaxine exerts antidepressant effects possibly by activating MAPK-ERK1/2 and P13K-AKT pathways in the hippocampus. Behavioural Brain Research 335 (August): 63–70. doi:10.1016/j.bbr.2017.08.011.Find this resource:

Shirayama, Y., & Hashimoto, K. (2018). Lack of antidepressant effects of (2R,6R)-hydroxynorketamine in a rat learned helplessness model: Comparison with (R)-ketamine. International Journal of Neuropsychopharmacology 21 (1): 84–88. doi:10.1093/ijnp/pyx108.Find this resource:

Siddiqui, N., & Sonenberg, N. (2015). Signalling to eIF4E in Cancer. Biochemical Society Transactions 43 (5): 763–772. doi:10.1042/BST20150126.Find this resource:

Simonoff, E., Pickles, A., Charman, T., Chandler, S., Loucas, T., & Baird, G. (2008). Psychiatric disorders in children with autism spectrum disorders: Prevalence, comorbidity, and associated factors in a population-derived sample. Journal of the American Academy of Child and Adolescent Psychiatry 47 (8): 921–929. doi:10.1097/CHI.0b013e318179964f.Find this resource:

Starkstein, S. E., Jorge, R., Mizrahi, R., & Robinson, R. G. (2005). The construct of minor and major depression in Alzheimer’s disease. American Journal of Psychiatry 162 (11): 2086–2093. doi:10.1176/appi.ajp.162.11.2086.Find this resource:

Su, M., Hong, J., Zhao, Y., Liu, S., & Xue, X. (2015). MeCP2 controls hippocampal brain-derived neurotrophic factor expression via homeostatic interactions with microRNA‑132 in rats with depression. Molecular Medicine Reports 12 (4): 5399–5406. doi:10.3892/mmr.2015.4104.Find this resource:

Suzuki, K., Nosyreva, E., Hunt, K. W., Kavalali, E. T., & Monteggia, L. M. (2017). Effects of a ketamine metabolite on synaptic NMDAR function. Nature 546 (7659): E1–E3. doi:10.1038/nature22084.Find this resource:

Trullas, R., Folio, T., Young, A., Miller, R., Boje, K., & Skolnick, P. (1991). 1-Aminocyclopropanecarboxylates exhibit antidepressant and anxiolytic actions in animal models. European Journal of Pharmacology 203 (3): 379–385. doi:10.1016/0014-2999(91)90894-V.Find this resource:

Undurraga, J., & Baldessarini, R. J. (2012). Randomized, placebo-controlled trials of antidepressants for acute major depression: Thirty-year meta-analytic review. Neuropsychopharmacology 37 (4): 851–864. doi:10.1038/npp.2011.306.Find this resource:

Webhofer, C., Gormanns, P., Tolstikov, V., Zieglgänsberger, W., Sillaber, I., Holsboer, F., & Turck, C. W. (2011). Metabolite profiling of antidepressant drug action reveals novel drug targets beyond monoamine elevation. Translational Psychiatry 1 (12): e58. doi:10.1038/tp.2011.56.Find this resource:

Wesseling, H., Rahmoune, H., Tricklebank, M., Guest, P. C., & Bahn, S. (2015). A targeted multiplexed proteomic investigation identifies ketamine-induced changes in immune markers in rat serum and expression changes in protein kinases/phosphatases in rat brain. Journal of Proteome Research 14 (1): 411–421. doi:10.1021/pr5009493.Find this resource:

Wolfe, S. A., Workman, E. R., Heaney, C. F., Niere, F., Namjoshi, S., Cacheaux, L. P., Farris, S. P., et al. (2016). FMRP regulates an ethanol-dependent shift in GABABR function and expression with rapid antidepressant properties. Nature Communications 7 (September): 12867. doi:10.1038/ncomms12867.Find this resource:

Workman, E. R., Haddick, P. C. G., Bush, K., Dilly, G. A., Niere, F., Zemelman, B. V., Raab-Graham, K. F. (2015). Rapid antidepressants stimulate the decoupling of GABAB receptors from GIRK/Kir3 channels through increased protein stability of 14-3-3η. Molecular Psychiatry 20 (3): 298–310. doi:10.1038/mp.2014.165.Find this resource:

Workman, E. R., Niere, F., & Raab-Graham, K. F. (2013). mTORC1-dependent protein synthesis underlying rapid antidepressant effect requires GABABR signaling. Neuropharmacology 73: 192–203. doi:10.1016/j.neuropharm.2013.05.037.Find this resource:

Yang, C., Hu, Y.-M., Zhou, Z.-Q., Zhang, G.-F., & Yang, J.-J. (2013). Acute administration of ketamine in rats increases hippocampal BDNF and mTOR levels during forced swimming test. Upsala Journal of Medical Sciences 118 (1): 3–8. doi:10.3109/03009734.2012.724118.Find this resource:

Yang, C., Qu, Y., Abe, M., Nozawa, D., Chaki, S., & Hashimoto, K. (2017). (R)-Ketamine shows greater potency and longer lasting antidepressant effects than its metabolite (2R,6R)-hydroxynorketamine. Biological Psychiatry 82 (5): e43–e44. doi:10.1016/j.biopsych.2016.12.020.Find this resource:

Yang, C., Zhou, Z.-Q., Gao, Z.-Q. Shi, J.-Y., & Yang, J.-J. (2013). Acute increases in plasma mammalian target of rapamycin, glycogen synthase kinase-3β, and eukaryotic elongation factor 2 phosphorylation after ketamine treatment in three depressed patients. Biological Psychiatry 73 (12): e35–e36. doi:10.1016/j.biopsych.2012.07.022.Find this resource:

Yu, J.-J., Zhang, Y., Wang, Y., Wen, Z.-Y., Liu, X.-H., Qin, J., & Yang, J.-L. (2013). Inhibition of calcineurin in the prefrontal cortex induced depressive-like behavior through mTOR signaling pathway. Psychopharmacology 225 (2): 361–72. doi:10.1007/s00213-012-2823-9.Find this resource:

Yuen, E. Y., Wei, J., Liu, W., Zhong, P., Li, X., & Yan, Z. (2012). Repeated stress causes cognitive impairment by suppressing glutamate receptor expression and function in prefrontal cortex. Neuron 73 (5): 962–977. doi:10.1016/j.neuron.2011.12.033.Find this resource:

Zanos, P., Moaddel, R., Morris, P. J., Georgiou, P., Fischell, J., Elmer,G. I., Alkondon, M., et al. 2016. NMDAR inhibition-independent antidepressant actions of ketamine metabolites. Nature 533 (7604): 481–486. doi:10.1038/nature17998.Find this resource:

Zhang, K., Yamaki, V. N., Wei, Z., Zheng, Y., & Cai, X. (2017). Differential regulation of GluA1 expression by ketamine and memantine. Behavioural Brain Research 316 (January): 152–159. doi:10.1016/j.bbr.2016.09.002.Find this resource:

Zimmermann, C. A., Hoffmann, A., Raabe, F., & Spengler, D. (2015). Role of Mecp2 in experience-dependent epigenetic programming. Genes 6 (1): 60–86. doi:10.3390/genes6010060.Find this resource: