D-Serine: physiology and pathology

Radzishevsky, Inna; Sason, Hagit; Wolosker, Herman

Current Opinion in Clinical Nutrition & Metabolic Care: January 2013 - Volume 16 - Issue 1 - p 72–75
doi: 10.1097/MCO.0b013e32835a3466
PROTEIN, AMINO ACID METABOLISM AND THERAPY: Edited by Olav Rooyackers and John Brosnan

Purpose of review: Here, we discuss the recent data on the role of different N-methyl D-aspartate receptor (NMDAR) coagonists, D-serine and glycine, in regulating NMDAR activity and neurotoxicity.

Recent findings: D-Serine originates from both neurons and astrocytes, from where it is released by different mechanisms. Recent data indicate that like glial D-serine, neuronal D-serine is required for NMDAR-dependent, long-term potentiation at the hippocampal CA1–CA3 synapses and proper synapse formation in the cerebral cortex. D-serine is the physiological coagonist of synaptic NMDAR, whereas glycine action is restricted to extrasynaptic sites.

Summary: D-Serine is now recognized as the major NMDAR coagonist at the synapse. The data establish D-serine as a key transmitter or neuromodulator that mediates synaptic NMDAR activation and neurotoxicity. In this context, drugs that inhibit D-serine synthesis or release will provide new neuroprotective strategy.

Department of Biochemistry, B. Rappaport Faculty of Medicine, Technion-Israel Institute of Technology, Haifa, Israel

Correspondence to Herman Wolosker, Department of Biochemistry, B. Rappaport Faculty of Medicine, Technion-Israel Institute of Technology, Haifa, 31096, Israel. Tel: +972 4 829 5386; e-mail: hwolosker@tx.technion.ac.il

Article Outline
Back to Top | Article Outline


D-Serine is an endogenous D-amino acid that is synthesized in the brain by the serine racemase enzyme [1]. The levels of D-serine are higher than several essential amino acids and account for a third of L-serine levels. As a D-isomer, D-serine seems to interact with only a few proteins in the brain. The main target of D-serine is the N-methyl D-aspartate receptor (NMDAR), which is essential for neurotransmission, learning, and memory formation [2]. NMDAR operation requires the simultaneous binding of glutamate along with a coagonist (glycine or D-serine) to distinct receptor subunits. Here, we review the recent studies from different laboratories demonstrating that D-serine is a major, if not the main, coagonist for NMDAR-dependent processes ranging from synaptic transmission [3▪▪,4▪], long-term potentiation (LTP) at the hippocampal CA1 [5▪,6▪▪], neurotoxicity [2,7▪,8], behavioral paradigms [9,10], and dendritic architecture [11]. Furthermore, we discuss the roles of altered D-serine dynamics in neurodegeneration and psychiatric disorders.

Back to Top | Article Outline


The purification and cloning of serine racemase established the conservation of D-amino acid metabolism in mammals [1]. Serine racemase knockout (SR-KO) mice display 90% lower D-serine in the brain and deficits in NMDAR activity [1]. Despite the several roles attributed to D-serine, its cellular origin and release mechanisms are still controversial. Early studies indicate that D-serine is exclusively produced and released by astrocytes, a class of glial cells that ensheath the synapses. On the other hand, recent studies using new antibodies, and SR-KO mice as negative controls, demonstrate that serine racemase is predominantly expressed by neuronal structures [2,6▪▪], suggesting that a significant fraction of D-serine comes from neurons [12].

Fresh studies provide interesting, though somewhat conflicting evidence, that D-serine from astrocytes and neurons mediate important brain functions. Mothet and colleagues recently employed pharmacological means to distinguish between D-serine of neuronal and glial origin [4▪]. They report that disruption of astrocyte metabolism with fluoroacetate decreases the isolated NMDAR synaptic potentials. These potentials are restored by adding D-serine, strongly suggesting a role for astrocytes in releasing D-serine [4▪]. On the other hand, as astrocytes provide metabolic support for neurons, inhibition of glial metabolism may indirectly disrupt neuronal homeostasis by other mechanisms.

In order to avoid the caveats of pleiotropic pharmacological approaches, a more straightforward strategy can be used to genetically knock-down serine racemase expression in a cell-specific manner. Recently, Coyle and colleagues produced cell-specific SR-KOs that shed additional light on the cellular origin of D-serine. Conditional neuronal SR-KO mice exhibit lower brain D-serine levels along with deficits in the LTP in the hippocampal CA1–CA3 synapse, reduced isolated NMDAR synaptic potentials, and altered dendritic architecture in the somatosensory cortex in vivo[6▪▪,11]. On the other hand, the same group reports that conditional astrocytic SR-KO mice display only a marginal decrease in brain serine racemase expression, with no changes in brain D-serine or NMDAR synaptic potentials. These data, along with the previous studies showing predominant expression of serine racemase in neurons [2], provide compelling genetic demonstration that D-serine produced by glutamatergic neurons is required for optimal NMDAR activity.

The relative contribution of glia and neurons for D-serine dynamics is still uncertain. Furuya and colleagues showed that L-serine produced by astrocytes is essential for brain D-serine synthesis, leading us to propose a serine shuttle between astrocytes and neurons [2,13]. According to this model, neurons take up L-serine exported by astrocytes to make their own D-serine. Part of the neuron-derived D-serine might accumulate in astrocytes by uptake mechanism.

In summary, there is now genetic and biochemical evidence indicating that neurons play a role in D-serine signaling, but further studies are required to clarify the roles of glia–neuron crosstalk in D-serine dynamics.

Back to Top | Article Outline


Johnson and Ascher first described glycine as an NMDAR coagonist. However, evidence that endogenous glycine indeed regulates NMDARs has been lacking. In the last decade, several studies demonstrated that D-serine is the dominant NMDAR coagonist for synaptic plasticity and neurotoxicity [2]. A recent report by Oliet and colleagues extends the previous findings, showing that D-serine is the major coagonist for synaptic NMDARs composed of GluN2a subunits, whereas glycine seems to play a role only in extrasynaptic receptors composed of GluN2b [3▪▪]. This study employed enzymes to destroy selectively D-serine or glycine, a strategy that does not affect basal glutamatergic neurotransmission. D-Serine removal inhibits the LTP and isolated NMDAR synaptic potentials at the CA1–CA3 hippocampal synapses, whereas glycine removal has no effect. On the other hand, another form of plasticity, the long-term depression, depends on both synaptic and extrasynaptic NMDARs, and hence also requires glycine [3▪▪]. By isolating the extrasynaptic component upon direct application of NMDA to slices, the authors show that glycine, rather than D-serine, is the main co-activator of extrasynaptic NMDARs. Another important observation is that the synaptic NMDARs regulated by D-serine alone are the culprits in neurotoxicity. This notion challenges the widespread belief that NMDAR neurotoxicity depends on extrasynaptic NMDAR activation, whereas synaptic NMDAR activation is thought to transduce neuroprotective signals [14].

Does the activation of synaptic NMDARs depend on glial or neuronal D-serine? It is possible that D-serine produced by neurons mediates most of the effects described by Oliet and coworkers [3▪▪]. Although the LTP and the synaptic NMDAR responses in hippocampal CA1–CA3 synapses are decreased in conditional neuronal SR-KO mice, they are unaffected in astrocytic SR-KOs [6▪▪]. Another unanswered question refers to the nature of the relatively large fraction of the synaptic NMDAR responses (40–70%) that is resistant to endogenous amino acid removal observed by Oliet and colleagues. Is this related to residual endogenous D-serine, glycine, or a third coagonist? The use of different types of SR-KO mice should help to clarify these issues.

The possible routes and targets of D-serine are now summarized in Fig. 1. Regardless of whether D-serine comes from neurons or glia, most studies now agree that this D-isomer is a major regulator of synaptic NMDAR activity, with important implications for neurotransmission and pathological states encompassing NMDAR overactivation.

Back to Top | Article Outline


As an endogenous coagonist of NMDARs, D-serine is likely to be involved in pathologies encompassing NMDAR dysfunction. Previous data demonstrated a role of D-serine in NMDAR neurotoxicity, as well as alterations in D-serine metabolism in diverse disorders, such as amyotrophic lateral sclerosis (ALS) and schizophrenia [2]. D-Serine degradation is carried out by the enzyme D-amino acid oxidase (DAO), which is enriched in the cerebellum, brain stem, and spinal cord. A recent study highlights the role of endogenous D-serine in motor neuron degeneration that takes place in ALS [7▪]. This study follows the previous findings by De Belleroche's group who described mutations in the DAO gene as the cause of some classical, adult-onset, familial ALS cases. Sasabe and coworkers now found that inactivation of DAO in mice augments spinal D-serine levels and leads to motor neuron degeneration through overactivation of NMDARs. These data underscore the role of NMDAR in ALS, though the disease has been generally thought to involve overactivation of calcium-permeable 2-amino-3-(3-hydroxy-5-methyl-isoxazol-4-yl)propanoic acid receptors.

D-Serine also seems to play a role in epileptic seizures. Mori and colleagues found that SR-KO mice are partially resistant to seizures induced by pentylenetetrazol, a GABA antagonist [15▪]. The data indicate that serine racemase inhibitors may be useful to ameliorate epilepsy.

In addition to its involvement in neurodegeneration, altered D-serine synthesis or metabolism may play a role in schizophrenia. Lower D-serine levels may underlie the NMDAR hypofunction thought to play a role in the disease, but the molecular mechanisms involved are not known. A recent study led by Snyder's and Pletnikov's groups shed new light on this issue. They report that mice harboring mutant Disc-1, a gene truncated in a large family with major psychiatric disorders, exhibit lower brain D-serine and abnormalities that are reminiscent of NMDAR hypofunction [16▪▪]. Disc-1 interacts with serine racemase, stabilizing its levels, and inhibiting serine racemase degradation by the ubiquitin–proteasome system [16▪▪]. Truncated Disc-1 substantially lowers serine racemase levels, possibly by a dominant negative effect. Strikingly, only the selective expression of mutant Disc-1 in astrocytes impacts the serine racemase expression. Overexpression of truncated Disc-1 in neurons has no effect, indicating that Disc-1–SR interaction takes place solely in glia. The data indicate that glial and neuronal serine racemase are regulated by different mechanisms. Most importantly, this study provides compelling evidence linking glial dysfunction and schizophrenia.

Back to Top | Article Outline


Data obtained in the last year indicate that D-serine is involved in several physiological and pathological processes. D-Serine is the major regulator of synaptic NMDARs, LTP, and neurotoxicity, whereas glycine plays a role in extrasynaptic NMDARs. In parallel, fresh data using cell-specific SR-KO mice demonstrate new roles for neuronal D-serine, suggesting that neurons are important for D-serine dynamics and NMDAR activity. Pathologically elevated D-serine may cause neurodegeneration, whereas lower D-serine is implicated in the NMDAR hypofunction thought to play a role in schizophrenia.

Back to Top | Article Outline


This study was supported by the Israel Science Foundation.

Back to Top | Article Outline

Conflicts of interest

There are no conflicts of interest.

Back to Top | Article Outline


Papers of particular interest, published within the annual period of review, have been highlighted as:

▪ of special interest

▪▪ of outstanding interest

Additional references related to this topic can also be found in the Current World Literature section in this issue (p. 116).

Back to Top | Article Outline


1. Wolosker H, Mori H. Serine racemase: an unconventional enzyme for an unconventional transmitter. Amino Acids 2012.
2. Wolosker H. Serine racemase and the serine shuttle between neurons and astrocytes. Biochim Biophys Acta 2011; 1814:1558–1566.
3▪▪. Papouin T, Ladepeche L, Ruel Jrm, et al. Synaptic and extrasynaptic NMDA receptors are gated by different endogenous coagonists. Cell 2012; 150:633–646.

This study reports that D-serine acts on synaptic NMDARs, whereas glycine is restricted to extrasynaptic receptors. The authors also propose that synaptic NMDARs mediate neurotoxicity.

4▪. Fossat P, Turpin FR, Sacchi S, et al. Glial D-serine gates NMDA receptors at excitatory synapses in prefrontal cortex. Cereb Cortex 2011; 22:595–606.

This study shows that disrupting glia metabolism leads to inhibition of NMDAR potentials in the prefrontal cortex.

5▪. Turpin FR, Potier B, Dulong JR, et al. Reduced serine racemase expression contributes to age-related deficits in hippocampal cognitive function. Neurobiol Aging 2012; 32:1495–1504.

This report demonstrates that lowered D-serine underlies deficits in LTP observed in aged rats.

6▪▪. Benneyworth MA, Li Y, Basu AC, et al. Cell selective conditional null mutations of serine racemase demonstrate a predominate localization in cortical glutamatergic neurons. Cell Mol Neurobiol 2012; 32:613–624.

This study provides compelling genetic evidence that neuronal D-serine regulates NMDARs.

7▪. Sasabe J, Miyoshi Y, Suzuki M, et al. D-amino acid oxidase controls motoneuron degeneration through D-serine. Proc Natl Acad Sci US A 2012; 109:627–632.

This report implicates pathological D-serine increase in ALS pathogenesis.

8. Esposito S, Pristera A, Maresca G, et al. Contribution of serine racemase/D-serine pathway to neuronal apoptosis. Aging Cell 2012; 11:588–598.
9. Benneyworth MA, Coyle JT. Altered acquisition and extinction of amphetamine-paired context conditioning in genetic mouse models of altered NMDA receptor function. Neuropsychopharmacology 2012; 37:2496–2504.
10. DeVito LM, Balu DT, Kanter BR, et al. Serine racemase deletion disrupts memory for order and alters cortical dendritic morphology. Genes Brain Behav 2011; 10:210–222.
11. Balu DT, Coyle JT. Neuronal D-serine regulates dendritic architecture in the somatosensory cortex. Neurosci Lett 2012; 517:77–81.
12. Ding X, Ma N, Nagahama M, et al. Localization of D-serine and serine racemase in neurons and neuroglias in mouse brain. Neurol Sci 2011; 32:263–267.
13. Yang JH, Wada A, Yoshida K, et al. Brain-specific Phgdh deletion reveals a pivotal role for L-serine biosynthesis in controlling the level of D-serine, an N-methyl-D-aspartate receptor co-agonist, in adult brain. J Biol Chem 2010; 285:41380–41390.
14. Martel MA, Ryan TJ, Bell KF, et al. The subtype of GluN2 C-terminal domain determines the response to excitotoxic insults. Neuron 2012; 74:543–556.
15▪. Harai T, Inoue R, Fujita Y, et al. Decreased susceptibility to seizures induced by pentylenetetrazole in serine racemase knockout mice. Epilepsy Res 2012.

This study suggests that serine racemase inhibitors may be useful in ameliorating epileptic seizures.

16▪▪. Ma TM, Abazyan S, Abazyan B, et al. Pathogenic disruption of DISC1-serine racemase binding elicits schizophrenia-like behavior via D-serine depletion. Mol Psychiatry 2012.

This study links D-serine deficits to mutant Disc-1 expression in astrocytes and schizophrenia.


D-serine; gliotransmission; NMDA receptors

© 2013 Lippincott Williams & Wilkins, Inc.