What Is Glutamate?
Glutamate is the most abundant neurotransmitter in our brain and central nervous system (CNS). It is involved in virtually every major excitatory brain function. While excitatory has a very specific meaning in neuroscience, in general terms, an excitatory neurotransmitter increases the likelihood that the neuron it acts upon will have an action potential (also called a nerve impulse).1 When an action potential occurs the nerve is said to fire, with fire, in this case, being somewhat akin to the completion of an electric circuit that occurs when a light switch is turned on. The result of neurons firing is that a message can be spread throughout the neural circuit. It is estimated that well over half of all synapses in the brain release glutamate, making it the dominant neurotransmitter used for neural circuit communication.
Glutamate is also a metabolic precursor for another neurotransmitter called GABA (gamma-aminobutyric acid). GABA is the main inhibitory neurotransmitter in the central nervous system. Inhibitory neurotransmitters are essentially the flip-side of the coin—they decrease the likelihood that the neuron they act upon will fire.1
What Does Glutamate Do?
In the brain, groups of neurons (nerve cells) form neural circuits to carry out specific small-scale functions (e.g., formation and retrieval of memory). These neural circuits interconnect with each other to form large-scale brain networks, which carry out more complex functions (e.g., hearing, vision, movement). In order to get the individual nerve cells to work together across these networks some type of communication between them is needed and one way it is accomplished is by chemical messenger molecules called neurotransmitters. Glutamate plays a prominent role in neural circuits involved with synaptic plasticity—the ability for strengthening or weakening of signaling between neurons over time to shape learning and memory. It’s a major player in the subset of plasticity called long-term potentiation (LTP).
“The brain doesn't grow new neurons to store memories. It strengthens connections between existing neurons. This process is called long-term potentiation (LTP).”
Because of these and other roles, the glutamatergic system is paramount for fast signaling and information processing in neuronal networks. Glutamate signaling is critical in brain regions, including the cortex and hippocampus, which are fundamental for cognitive function. Glutamate receptors are widely expressed throughout the CNS, not only in neurons, but also in glial cells.
[Note: Glial cells (or neuroglia or simply glia) are non-neuronal brain cells that provide support and protection for neurons.]
Because it is the main molecule promoting neuronal excitation, glutamate is the principal mediator of cognition, emotions, sensory information, and motor coordination, and is linked to the activity of most other neurotransmitter systems.2 But glutamate is not a “more is better” molecule. Glutamatergic communication requires the right concentrations of glutamate be released in the right places for only small amounts of time. Less than this results in poor communication. More than this can be neurotoxic and can damage neurons and neural networks.
The Glutamate Neurotransmitter and the Goldilocks Principle
Glutamate signaling is an example of what’s sometimes referred to as the “Goldilocks Principle.” In the fairy tale story, Goldilocks tastes three different bowls of porridge. The first is too cold; the second is too hot, and the third is just the right temperature. This concept of a "just the right amount" has widespread application, including in cognitive science.
In cognitive science, this principle can refer to a process where the same neurotransmitter (or medication) can have both antagonist (inhibitory) and agonist (excitatory) properties. It can also apply to situations where too little or too much stimulation by the same signaling molecule is linked to sub-optimal performance, but some middle ground amount produces healthy responses.
When thinking about things that follow the Goldilocks Principle it’s important to avoid black-white or good-bad thinking. The key thing to focus on is that there’s a just right amount, often a range, where the best results are produced.
“Glutamate follows the Goldilocks Principle. Too little glutamate excitation can result in difficulty concentrating or mental exhaustion. But too much can result in excitotoxicity, which can damage nerve cells (neurons).”
Glutamate Synthesis, Signaling, and Cleanup
Neurotransmitters have several characteristics in common. The first is that they are synthesized (i.e., made or created) in neurons. After that, they are moved into areas near the end of neurons (synaptic vesicles near the terminal end of nerve cells) where they are stored until needed. This occurs in preparation for signaling, which involves release of the neurotransmitter from the message-sending neuron into the space between neurons (synaptic cleft), so it can activate (i.e., bind to) receptors on message-receiving neurons. After this signal is sent, the space between neurons is cleaned up, so it can be made ready for the next time a message needs to be sent. This is achieved by absorbing the neurotransmitter into a cell so it can be reused (recycling), and/or by degrading (breaking down and inactivating) the neurotransmitter in the space outside cells. Let’s explore how these occur with glutamate.
Glutamate does not cross the blood-brain barrier and must be synthesized in neurons from building block molecules (i.e., precursors) that can get into the brain. In the brain, glutamine is the fundamental building block for glutamate. The most prevalent biosynthetic pathway synthesizes glutamate from glutamine using an enzyme called glutaminase.2
[Note: Enzymes are catalysts used to produce specific biochemical reactions: They usually have names that end in “ase.” Coenzymes are parts of certain enzymes. Many coenzymes are derived from vitamins.]
Glutamine is the most abundant of the twenty amino acids the body uses to build proteins. It can be produced in the body (so is categorized as non-essential). Most glutamine is made and stored in muscle. Under certain circumstances, such as severe stress, the body can require more than it can make. This has led many scientists to consider glutamine as being a conditionally essential amino acid. It is one of the few amino acids that can directly cross the blood–brain barrier, so the glutamine pool in muscle can be used to support the brain.
“The blood-brain barrier acts a bit like a doorman, choosing what goes in (like nutrients) and out (such as metabolic waste products) of the brain. It also protects the brain against the entry of potentially harmful things (like bacteria).”
Glutamate can also be produced from glucose through a metabolic pathway that begins with the conversion of glucose to pyruvate (a process called glycolysis). Pyruvate then ethers the tricarboxylic acid (TCA) cycle (also called the Krebs cycle or citric acid cycle). The TCA cycle forms multiple important intermediates. One of these intermediates is α-ketoglutarate (α-KG). α-KG can be used to produce glutamate. An enzyme called glutamate dehydrogenase, which uses vitamin B3 (NAD+) as a coenzyme, is responsible for this reaction. This same enzyme can reconvert glutamate back into α-KG. Because of this enzyme, glutamate and α-KG can be continuously converted into each other.2 This dynamic equilibrium is a key intersection between anabolic and catabolic pathways and allows the body to shift resources in whichever direction is required.
[Note: Anabolic pathways construct molecules from smaller units. Catabolic pathways break molecules down into smaller units.]
Neurotransmitters, including glutamate, convey information from one neuron (message sender) to other "target" neurons (message recipients) within neural circuits. After synthesis, glutamate is transported into synaptic vesicles by vesicular glutamate transporters. This transport (and storage) occurs in the message-sending neuron in anticipation of needing to send glutamate messages in the future. Glutamate is stored in these vesicles until a nerve impulse triggers the release of glutamate into the synaptic cleft (i.e.,the space between neurons) and starts a receptor-mediated signaling process.3
“It’s estimated that about 99.99% of all the glutamate in the brain is stored inside cells (intracellular). Intracellular glutamate is inactive. It’s only the glutamate in the extracellular space between cells that causes excitation.”
Neurons with glutamate receptor proteins (i.e., glutamate receivers) respond to glutamate in the synaptic cleft. There are two general types of glutamate receptors. One type are called ionotropic receptors: Glutamate binding to these receptors allows the entry of ions (i.e., electrically charged minerals such as sodium or calcium) into the cell. There are three classes of ionotropic glutamate receptors: (1) N-methyl-D-aspartate (NMDA), (2) α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA), and (3) kainate receptors.
The second type of receptors are linked to molecules that will activate intracellular signaling pathways subsequent to glutamate binding. These are called G protein-coupled or metabotropic receptors. Metabotropic glutamate receptors (mGluR) modulate synaptic transmission (i.e. neuronal communication) by regulating the activity of a wide variety of ion channels, including ionotropic glutamate receptors, as well as receptors for other neurotransmitters.1,4
The next phase of neurotransmission is cleanup. Signaling is based on relative changes, not absolute amounts. In a quiet room the human ear might detect a whisper. In a loud nightclub it might not hear a shout. Neurotransmitters work on a similar principle. Short bursts of glutamate produce responses. But, for the best response to occur with the smallest amount of glutamate, the space between neurons needs to be the equivalent of a quiet room. Glutamate also follows the Goldilocks principle—too little and signaling within neural networks is subpar, but too much can be neurotoxic. For these reasons, the glutamate in the extracellular space between neurons needs to be continuously removed.
Neurotransmitter cleanup is commonly a combination of (1) transporting some of the neurotransmitter back into cells, and (2) inactivating the neurotransmitter that’s left floating in the space between cells. While the first of these processes applies to glutamate, there is no enzymatic inactivation system for glutamate in the extracellular space. This means that glutamate can interact with its receptors continuously until it diffuses away or is taken up by cellular transporters for reuse/recycling.
“Because there are no enzyme systems in the spaces between nerve cells to inactivate glutamate, it’s important to support nerve cells against excessive glutamate excitation. Astrocytes provide part of this support.”
Some glutamate can be taken up into neurons. This is done by excitatory amino acid transporters (i.e., glutamate transporters), but much of the released glutamate is taken up by a type of glial cell called astroglia or astrocytes. Astroglia surround synapses and play important roles in areas including nervous system repair, metabolic support of neurons, and neurotransmitter cleanup. The combination of neurons and supporting astroglia are responsible for emptying the synaptic cleft of glutamate to turn off the signal and reset the system for generation and propagation of the next glutamate signal. In this cleanup role, astroglia act to protect neurons from glutamate excitotoxicity.
Once glutamate has been taken up by astrocytes, it reacts with ammonia to form glutamine through the activity of glutamine synthetase. Glutamine is then exported to the extracellular fluid where it’s taken up by neurons, starting the glutamate synthesis process again. This sequence of events is referred to as the glutamate-glutamine cycle: It is how the nervous system ensures it maintains an adequate supply of glutamate.3,5
Glutamate, and its receptors, are central elements in memory formation and retrieval because of their role in the key cellular mechanism of memory and learning called long-term potentiation (LTP).2,6
LTP is a form of synaptic plasticity, a term that refers to the biochemical processes through which synapses respond to patterns of activity, either by strengthening in response to increased activity, or by weakening in response to decreased activity. LTP is the persistent strengthening component of plasticity. It is one of the major cellular mechanisms that underlies how the brain encodes memories.
LTP occurs in many brain regions involved in memory processes, including the neocortex, the amygdala, and the striatum, but it is in the hippocampus that it is most extensively studied.7 Different areas of the brain can have varying molecular mechanisms of LTP, but overall glutamate signaling plays the largest role.
“Learning and memory are thought to be fundamentally associative. This means we are able to recall new information better if we associate it with something we already know. Glutamate signaling has unique properties that allows neurons to build associations, so is at the foundation of memory building.”
Note: The next few paragraphs will go into greater detail of the mechanisms that allows for synaptic coincidence and associative properties for readers who are interested. For those who just want the big picture, the key point is that glutamate allows neurons to associate information.
The glutamate NMDA receptor is one of the most prominent intermediaries in LTP. It is both a ligand-dependent (i.e., responds to chemical properties of the glutamate molecule) and a voltage-dependent ion channel (i.e., responds to changes in electrical signaling). Glutamate binding to neurons affects both of these processes simultaneously. This unique characteristic allows the NMDA receptor to (1) detect synaptic coincidence (i.e., the simultaneous activation of a presynaptic and a postsynaptic neuron), and (2) have associative properties (i.e., be able to associate two events together at the synaptic level).8,9 On a larger scale this results in the brain being able, through LTP, to continuously change neural circuits and larger networks as it learns new information.
The activation of a presynaptic neuron causes the release of glutamate, which then binds to postsynaptic glutamate ionotropic receptors—NMDA and AMPA. At a resting membrane potential, the ion channel of NMDA receptors is blocked by magnesium (Mg2+), which prevents the flow of calcium ions (Ca2+) through the channel. When the correct electrical charge across the cell membrane occurs, this Mg2+ blockade is relieved and Ca2+ is allowed to enter. This process of withdrawing Mg2+ ions to allow Ca2+ entry is called membrane depolarization and is achieved by glutamate activation of AMPA receptors. Activation of the AMPA receptors allows ion channels to become permeable to the influx of sodium (Na+). This influx changes the electrical characteristics of the cell membrane which causes an excitatory postsynaptic potential.6 The NMDA receptor is permeable to the influx of Na+, the efflux of potassium (K+), and, importantly, to the influx of calcium (Ca2+). Upon activation and opening of NMDA ion channels, Ca2+ influx into the postsynaptic neuron (an essential step in LTP) will initiate a cascade of biochemical events that modifies synaptic strength. Through the activation of Ca2+/calmodulin-dependent protein kinase II (CaMKII), Ca2+ promotes the insertion of additional AMPA receptors into the postsynaptic membrane and increases their channel conductance. The additional AMPA receptors increase the responsiveness of postsynaptic neurons to glutamate. Consequently, the next release of glutamate by the presynaptic neuron will cause a larger excitatory postsynaptic potential—the synapse has become stronger.6,10
“A neuron changes when it responds to a glutamate signal. Some of these changes makes the neuron more likely to respond to glutamate in the future, acting to strengthen connections between neurons.”
The NMDA receptor is a molecular switch for the induction of synaptic plasticity. Its unique properties help account for some important characteristics of LTP that allow it to support memory formation. For example, activation of NMDA receptors requires postsynaptic depolarization. This occurs through the activation of AMPA receptors after glutamate is released by the presynaptic neurons. This combination of cellular events ensures that LTP is only induced when the presynaptic cell fires before the postsynaptic cell (temporal specificity) and that LTP is only induced at synapses receiving stimulation (input-specificity). These act to guarantee the specificity of strengthened connections.9,11
LTP is promoted by multiple neurobiological pathways that underlie memory formation and consolidation. While glutamate signaling is critical for LTP, it does not work in isolation. For example, cholinergic projections (nerve cells that use acetylcholine as a neurotransmitter) to the hippocampus enhance glutamatergic synaptic transmission and LTP.12,13
A seemingly paradoxical aspect of glutamatergic neurotransmission is that it’s both essential for brain development and function, and also toxic when it occurs in excess, an effect known as glutamate excitotoxicity.2
In a healthy brain, almost all of the glutamate (99.99%) is stored within cells and only released in small amounts when needed to produce a signaling response. Because signaling is a change-based event, neurons and astrocytes empty the extracellular space from glutamate between these signals. But, if circumstances result in high levels of glutamate in the synaptic cleft, this can cause an excessive activation of NMDA receptors (i.e., too much glutamate excitation).
“Glutamate excitotoxicity (sometimes called a glutamatergic storm) refers to the damage to nerve cells caused by excessive stimulation of NMDA and AMPA receptors by glutamate.”
Continued excitation of NMDA receptors can lead to an excessive entry of Ca2+ into neurons. NMDA overstimulation also increases intracellular Ca2+ levels by affecting mechanisms of calcium homeostasis within cells.2,14 Intracellular levels of free calcium are normally maintained at very low concentrations relative to extracellular levels. The excessive rise in the intracellular levels of Ca2+ triggers a number of cell-damaging processes that ultimately lead to cell death through a process called apoptosis.14,15 Imbalances in calcium can activate enzymes that disrupt the integrity and function of DNA, cell membranes, and intracellular organelles, particularly mitochondria. Calcium can cause the swelling of mitochondria, leading to the release of reactive oxygen species (ROS) and to the interruption of cell energy production (i.e., the generation of ATP). These processes, while detrimental on their own, produce positive feedback loops that accelerate cell damage, rapidly leading to neuronal self-digestion by protein breakdown, free radical formation, and lipid peroxidation.14,16
Glutamate Signaling Stack
The most important design factor for a glutamate signaling stack is to promote receptor sensitivity to glutamate, while protecting against excessive glutamate signaling. Other considerations are support of (1) enzyme function involved in glutamate synthesis, signaling, and cleanup; and (2) endogenous neuroprotective systems.* Let’s put these pieces together.
Because glutamate signaling follows the Goldilocks Principle, it’s important to support endogenous regulatory processes that allow for LTP to occur. This means providing support that will help glutamate receptors (like NMDA and AMPA) be better equipped to detect low levels of glutamate, while not overreacting to high levels (i.e., a just the right amount response).17
Magnesium plays a key role in the regulation of NMDA receptor excitability. At a resting membrane potential (i.e., pre-glutamate excitation), Mg2+ ions act to block the ion channel of NMDA receptors. Glutamate excitation relieves the Mg2+ blockade and calcium (Ca2+) influx through NMDA receptors is allowed. Low levels of Mg2+ in the CNS (i.e., a relative magnesium insufficiency) contribute to the hyperexcitability of NMDA receptors. As a consequence, fewer NMDA channels will be blocked, more NMDA channels will be opened at lower membrane potentials, and an excessive calcium influx can occur at a lower threshold of glutamate signaling, leading to excitotoxicity. Neuronal Mg2+ concentrations are therefore of major neurophysiological importance. However, magnesium insufficiency is common.18 We included magnesium threonate to address this potential insufficiency, and chose the threonate salt form because it is highly bioavailable and crosses the blood-brain barrier, so it can efficiently elevate brain magnesium levels.*19
Several ingredients play roles in promoting balanced glutamate signaling and/or supporting receptors. Celastrus paniculatus might be neuroprotective against excitotoxicity, possibly through modulation of NMDA receptor activity.*20,21 Huperzine A appears to support balanced NDMA receptor binding and may prevent glutamate excitotoxicity by reducing glutamate-induced calcium mobilization.*22,23 Vitamin C and pyrroloquinoline quinone (PQQ) might support NMDA glutamate receptors, which could provide protection against excitotoxicity.*24,25 Taurine may decrease the affinity of NMDA glutamate receptors to glycine, which is needed for their activation,26 while still inducing LTP.*27 Likewise, Lion’s Mane has actions that suggest it might decrease glutamatergic excitotoxicity, while increasing LTP.*28 Curcumin was reported to have neuroprotective effects against glutamate excitotoxicity in animals.29
In later phases of LTP, the reinforcement of synaptic connections requires intracellular gene transcription and protein synthesis.30 Calcium entry into the postsynaptic neuron activates adenylate cyclase and leads to the production of cyclic AMP (cAMP). cAMP is a second messenger. It acts as an important signaling molecule inside cells. In essence, it listens for hormone signals that register on the outside of cells (and can’t get inside) and then delivers these messages inside the cell. One of the messages cAMP delivers activates signaling pathways that lead to an upregulation of mRNA translation that sustains late LTP.31 LTP is also regulated by the activity of phosphodiesterase-4 (PDE4), the enzyme that hydrolyzes cAMP.32 Both adenylate cyclase activation and phosphodiesterase inhibition have been shown to promote long-lasting LTP.33,34
cAMP is an example of a seesaw effect in a pathway. Adenylate cyclase is on one end, pushing cAMP production up, but PDE4 is on the other end, pushing it back down. Coleus forskohlii and Artichoke extract were included in the stack to synergistically target synaptic plasticity by impacting the push-pull of the cAMP pathway. Forskolin (found in Coleus forskohlii) may induce late LTP by supporting adenylate cyclase activity, leading to increased intracellular levels of cAMP,*35 whereas Artichoke might increases cAMP levels indirectly via effects on PDE4, further increasing intracellular levels of cAMP.*36,37 This pulling effect is further supported by caffeine and theobromine, which also exert effects on PDE4.*38
Given the important role of calcium in the mechanisms of LTP, the maintenance of adequate levels of calcium in the brain is essential for synaptic strengthening. The included Vitamin B5 in its calcium salt form (calcium pantothenate) acts as a minor calcium donor to augment calcium pools.* Calcium availability is supported by Vitamin D, which facilitates calcium absorption from the diet.*39
Noopept (found in Qualia but not Mind) appears to enhance glutamate release and increases the activation of NMDA receptors after acute or short-term administration. But Noopept is an example of biphasic (or adaptive effects): After repeated administration, Noopept increases the activation of AMPA receptors and decreases the release of glutamate (i.e., less glutamate might be needed to produce a signal).*40 Noopept may modulates voltage-gated calcium and potassium ion flows in the neuronal membrane: It regulates (both up and down) Ca2+ influx into neurons, and induces both hyperpolarizing and depolarizing membrane responses, an effect that may be linked to the pattern of glutamatergic modulation.41–43 Overall, Noopept seems to support the mechanisms of LTP41–43 while also shifting signaling in ways that would be expected to reduce risk of glutamate-induced excitotoxicity.40,44,45
Other ingredients in our stack that support aspects of glutamate signaling include:
(1) Vitamin B3 (niacinamide) is used as coenzyme in glutamate dehydrogenase enzyme;2*
(2) Carnitine (from acetyl-L-carnitine) might support and protect metabotropic glutamate receptors;46*
(3) Neuroadaptogens including Ginkgo biloba and Rhodiola rosea may protect neurons from excessive glutamate excitation;47-49*
(4) Neurolipid compounds phosphatidylserine and DHA are used in cell membranes and appears to support balanced glutamate receptor signaling and function.50,51*
Nutrients to support glutamate neurotransmission infographic:
Why Should You Support Glutamate Pathways and Processes?
Glutamate is one of the most important neurotransmitters in the brain and the main neuronal activator. Through its binding to NMDA and AMPA receptors, glutamate is a key element in the cellular mechanisms that support memory and learning. Because of these and other roles, glutamate is critical for brain development and cognitive performance. But glutamate is also subject to the Goldilocks Principle: excessive glutamate signaling can be neurotoxic. Under ideal circumstances the brain’s self-regulatory abilities allows it to have a “just the right amount” of glutamate signaling. The right amounts are released, in the right places, for the right amount of time. But circumstances are not always ideal.
Most of us can use some extra support in times when demands are high or outpacing our ability to cope. The body (and brain are no different). They have tremendous capabilities to adapt. But these capabilities are not infinite. And they are more effective when supported.
By supporting the glutamatergic system, the goal is to (1) promote the optimization of general brain function, and (2) provide nutritional resources that can be used as needed by the fundamental pathways that underlie the brain’s remarkable capacities for synaptic plasticity—the ability to adapt and strengthen neural circuits and networks in response to cognitive demands.
1. Purves D. Neuroscience. 5th edition. Sinauer Associates; 2011.
2. Hassel B, Dingledine R. In: Brady ST, Siegel GJ, eds. Basic Neurochemistry. Elsevier; 2012: 342-366. doi:10.1016/B978-0-12-374947-5.00017-1.
3. Zhou Y, Danbolt NC. J Neural Transm. 2014; 121(8):799-817. doi:10.1007/s00702-014-1180-8.
4. Meldrum BS. J Nutr. 2000; 130(4):1007S-1015S. doi:10.1093/jn/130.4.1007S.
5. Featherstone DE. ACS Chem Neurosci. 2010; 1(1):4-12. doi:10.1021/cn900006n.
6. Tsien JZ. In: Basic Neurochemistry. Elsevier; 2012: 963-981. doi:10.1016/B978-0-12-374947-5.00056-0.
7. Morris RGM, et al. Philos Trans R Soc B Biol Sci. 2003;358(1432):773-786. doi:10.1098/rstb.2002.1264.
8. Wigström H, Gustafsson B. Acta Physiol Scand. 1985; 123(4):519-522. doi:10.1111/j.1748-1716.1985.tb07621.x.
9. Cooke SF. Brain. 2006; 129(7):1659-1673. doi:10.1093/brain/awl082.
10. Carlson NR, Birkett MA. In: Physiology of Behavior. 12th edition. Pearson Education Limited; 2017.
11. Volianskis A, et al. Brain Res. 2015; 1621:5-16. doi:10.1016/j.brainres.2015.01.016.
12. Hasselmo ME. Curr Opin Neurobiol. 2006; 16(6):710-715. doi:10.1016/j.conb.2006.09.002.
13. Ge S, Dani JA. J Neurosci. 2005; 25(26):6084-6091. doi:10.1523/JNEUROSCI.0542-05.2005.
14. Mark LP, et al. AJNR Am J Neuroradiol. 2001; 22(10):1813-1824.
15. Mody I, MacDonald JF. Trends Pharmacol Sci. 1995; 16(10):356-359.
16. Wang Y, Qin Z. Apoptosis. 2010; 15(11):1382-1402. doi:10.1007/s10495-010-0481-0.
17. Lipton SA. Nat Rev Drug Discov. 2006; 5(2):160-170. doi:10.1038/nrd1958.
18. Rosanoff A, et al. Nutr Rev. 2012; 70(3):153-164. doi:10.1111/j.1753-4887.2011.00465.x.
19. Slutsky I, et al. Neuron. 2010; 65(2):165-177. doi:10.1016/j.neuron.2009.12.026.
20. Godkar PB, et al. J Ethnopharmacol. 2004; 93(2-3):213-219. doi:10.1016/j.jep.2004.03.051.
21. Godkar PB, et al. Phytomedicine. 2006; 13(1-2):29-36. doi:10.1016/j.phymed.2003.11.011.
22. Zhang JM, Hu GY. Neuroscience. 2001; 105(3):663-669.
23. Ved HS, et al. Neuroreport. 1997; 8(4):963-968.
24. Majewska MD, Bell JA. Neuroreport. 1990; 1(3-4):194-196.
25. Aizenman E, et al. J Neurosci. 1992; 12(6):2362-2369.
26. Chan CY, et al. Adv Exp Med Biol. 2013; 775:45-52. doi:10.1007/978-1-4614-6130-2_4.
27. del Olmo N, et al. Eur J Neurosci. 2004; 19(7):1875-1886. doi:10.1111/j.1460-9568.2004.03309.x.
28. Phan C-W, et al. Crit Rev Biotechnol. 2015; 35(3):355-368. doi:10.3109/07388551.2014.887649.
29. Wang R, et al. Brain Res. 2008; 1210:84-91. doi:10.1016/j.brainres.2008.01.104.
30. Nguyen P, et al. Science. 1994; 265(5175):1104-1107. doi:10.1126/science.8066450.
31. Wong ST, et al. Neuron. 1999; 23(4):787-798.
32. Ahmed T, Frey JU. Neuroscience. 2003; 117(3):627-638.
33. Navakkode S, et al. J Neurosci. 2004; 24(35):7740-7744. doi:10.1523/JNEUROSCI.1796-04.2004.
34. Barad M, et al. Proc Natl Acad Sci U S A. 1998; 95(25):15020-15025.
35. Gobert D, et al. J Neurochem. 2008; 106(3):1160-1174. doi:10.1111/j.1471-4159.2008.05470.x.
36. Xu B, et al. Eur J Pharmacol. 2010; 627(1-3):99-105. doi:10.1016/j.ejphar.2009.10.038.
37. Röhrig T, et al. Med (Basel, Switzerland). 2017; 4(4). doi:10.3390/medicines4040080.
38. Francis SH, et al. Handb Exp Pharmacol. 2011: 93-133. doi:10.1007/978-3-642-13443-2_4.
39. Christakos S, et al. Mol Cell Endocrinol. 2011; 347(1-2):25-29. doi:10.1016/j.mce.2011.05.038.
40. Vorobyov V, et al. Brain Res Bull. 2011; 85(3-4):123-132. doi:10.1016/j.brainresbull.2011.02.011.
41. Solntseva EI, et al. Gen Pharmacol. 1997; 29(1):85-89. doi:10.1016/S0306-3623(96)00529-0.
42. Bukanova JV, et al. Int J Neuropsychopharmacol. 2002; 5(3):229-237. doi:10.1017/S1461145702002997.
43. Lutsenko VK, et al. Bull Exp Biol Med. 2003; 135(6):559-562.
44. Andreeva NA, et al. Bull Exp Biol Med. 2000; 130(10):969-972. doi:10.1023/A:1002828707337.
45. Antipova TA, et al. Bull Exp Biol Med. 2016; 161(1):58-60. doi:10.1007/s10517-016-3344-z.
46. Llansola M, et al. Metab Brain Dis. 2002;17(4):389-397.
47. Zhu L, et al. Zhongguo Yao Li Xue Bao. 1997; 18(4):344-347.
48. Cho KS, et al. Phytother Res. 2016; 30(1):58-65. doi:10.1002/ptr.5500.
49. Palumbo DR, et al. Phytother Res. 2012; 26(6):878-883. doi:10.1002/ptr.3662.
50. Gagné J, et al. Brain Res. 1996; 740(1-2):337-345.
51. Wang X, et al. Neuroreport. 2003; 14(18):2457-2461. doi:10.1097/01.wnr.0000093483.65585.86.