D-Serine is an amino acid that plays a role in cognitive enhancement and schizophrenia treatment.
Sources and Structure
D-Serine is known to be a glial cell-derived neuromodulator, being produced in glial cells to support transmission from one neuron to another, being one of the first D-isomer amino acids to be found to have biological relevance in the human body (shortly followed by D-Aspartic Acid). Due to being derived from glial cells, it is sometimes called a gliotransmitter or gliomodulator.
D-Serine is an endogenous ligand at the glycine binding sites of NMDA receptors, and despite being named after glycine it is not sure which ligand is more biologically relevant in vivo; in vitro, D-serine appears to have similar binding potency yet is more effective at signalling (possible due to more exposure time), being active at 1μM. Furthermore, D-serine seems to have its action localized at synaptic NMDA receptors whereas glycine is an agonist at extrasynaptic although it still seems possible to induce excitotoxicity (which is historically thought to be due to extrasynaptic receptors due to the N2B subunit and synaptic having more N2A).
D-Serine is a neuromodulator that is secreted from the support cells of the nervous system (glial cells) which then modulates transmission from one neuron to another. It appears to be an endogenous ligand of the glycine binding site of NMDA receptors
While not commonly found in the diet, it is synthesized from dietary Glycine (an amino acid)
L-Serine (a dietary amino acid) is racemized into D-serine via the enzyme serine racemase, present in neurons and glial cells although relatively speaking the glial cells known as astrocytes express the highest levels of serine racemase. most notably in the forebrain; serine racemase expression correlates well with D-serine location. The rate of D-serine synthesis from serine racemase requires the cofactors of ATP and Magnesium and is positively affecte by calcium while it is inhibited by both glycine and L-aspartic acid.
Activation of AMPA receptors, possibly through encouraging glutamate-receptor-interacting protein (GRIP) to interact with serine racemase, causes a 5-fold increase in D-serine concentrations when activated. Finally, this enzyme is not specific for this conversion as it also converts L-serine into pyruvate (3:1 ratio relative to D-serine production) and ammonia.
D-Serine is produced mostly in astrocytes (some in neurons) via the serine racemase enzyme, produced from L-serine
D-serine is degraded by the d-amino acid oxidase (DAAO) enzyme which is exclusive to astrocytes. D-Serine concentrations are inversely related to expression/activity of the DAAO enzyme and ablating the enzyme causes a rise in D-serine in all tested brain regions. D-serine can be converted back into L-serine via the serine racemase enzyme as well, but this reaction has lower affinity than the opposite.
The primary mechanism that degrades D-serine is its reaccumulation in astrocytes and subsequent degradation by DAAO (major pathway) or conversion back into L-serine (minor pathway)
When looking at the degree of symptom reduction in schizophrenia, the general 17-30% range seen with 30mg/kg D-Serine is somewhat comparable to glycine when used at the dosage of 800mg/kg when investigated under similar conditions, suggesting that D-serine is more potent on a weight basis.
Supplementation of similar doses of D-serine and sarcosine (2,000mg) daily for six weeks noted that D-serine failed to be significantly better than placebo while sarcosine was, and concluded sarcosine more efficacious. This has been noted elsewhere, with sarcosine at the same dose of D-serine causing significantly more symptom reduction.
While D-serine appears to be more potent than glycine at the same signalling properties and interventions, the limited comparative evidence suggests that D-serine underperforms relative to sarcosine (a glycine transport inhibitor)
Supplementation of 30-120mg/kg (in schizophrenic patients) has been noted to increase serum D-serine concentrations with a Tmax of 1-2 hours and Cmax values have been reported at 120.6+/-34.6nmol/mL (30mg/kg), 272.3+/-62nmol/mL (60mg/kg), and 530.3+/-266.8nmol/mL (120mg/kg).
D-Serine supplementation peaks in the blood 1-2 hours after oral ingestion and follows linear dose-dependence up to 120mg/kg oral intake (highest tested dose)
Six weeks supplementation of D-Serine (30mg/kg) in persons with Parkinson's disease has been noted to increase serum D-serine from less than 10µM to 120.0+/-52.4µM and in persons with PTSD a similarly large increase (10-fold) has been seen with the same oral dose, reaching serum concentrations of 146+/-126.26µM.
In schizophrenics, the same dose has been noted to increase serum levels from 102.0+/-30.6nmol/mL to 226.8+/-72.8nmol/mL (122% increase) and mean serum levels at rest appear to increase in a dose-depedent fashion between 30-120mg/kg oral D-serine intake over 4 weeks.
Basal levels of D-serine appear to be increased following supplementation of D-serine, with 30mg/kg causing around a 10-fold increase in some studies but a lesser (and still present) spike in schizophrenics. There is quite a fairly bit of unreliability in the degree of increase
Supplemental D-serine does not appear to affect serum glycine, glutamate, or alanine concentrations and serum L-serine also appears unaffected.
D-Serine does not appear to significantly influence other amino acids in serum that may be related to serine metabolism
D-serine concentrations in the brain appear to be between 66+/-41nmol/g wet weight or 2.18+/-0.12nmol/mg which is somewhere between 10-15% of the total serine pool (being outnumbered by L-serine). D-serine is at higher concentrations in the prefrontal and parietal cortex relative to cerebellum and spinal cord. D-serine also possesses a half-life of 16 hours in the brain and increases in brain D-serine concentrations are seen with doses as low as 58mg/kg (mice).
D-serine has been detected in the cerebrospinal fluid (CSF) of control persons (2.72+/-0.32μM) and varying amounts have been found in persons with postherpetic neuralgia (1.85+/-0.21μM) and degenerative osteoarthritis (3.97+/-0.44μM) while in schizophrenics D-serine concentrations are lower than control (median value of 1.26μM versus 1.43μM; ranges not significantly different) despite L-serine being higher (22.8+/-8.01μM versus 18.2+/-4.78μM) and the ratio of L-serine to D-serine.
D-Serine is detected in cerebrospinal fluid (CSF) and the brain, where it is in lower concentrations in the CSF than in the blood and has a longer half-life in the brain than in the blood
Chronic supplementation of D-Serine may increase levels of L-serine in the cortex of mice.
Regulation and Distribution
It is thought that D-serine is stored alongside glutamate in neurons and astrocytes since D-serine is released when stimuli that are known to release glutamate are used while D-serine has been detected in neurons expressing a glutamate transporter. This colocalization and release has been noted elsewhere with glutamate and glycine.
D-Serine is possible released along with glutamate in neurons to activate NDMA receptors, as activation of two distinct sites (one of which requires either glycine or serine) is co-required. Neuronal release contibutes some, but not a majority, of D-serine to the synapse
D-Serine is known as a gliotransmitter, a neuromodulatory agent that is release from glial cells. It is thought that D-serine is release via some form of vesicular exocytosis (as these vesicles have been detected in glia). Synaptic vesicles have been found to coexpress glycine, glutamate, and GABA (but not D-serine) whereas D-serine is thought to have its own vesicular storage; vesicles may not be the only method of D-serine release, as the Asc-1 and TRPA1 transporters have been implicated (direct transport and supporting calcium influx, respectively) and inhibiting vesicular storage does not abolish D-serine release.
Astrocyte release of D-serine appears to be vital to NDMA-dependent processes (removing astrocytes from hippocampal cultures suppress long term potentiation, and this is restored with D-serine)
D-Serine is release from astrocytes (glial cells) appears to be the predominant form of D-serine release in the brain (neuronal release being a smaller total amount) and the mechanisms mediating the release are not well known at this point in time. It is known to be vital to glutaminergic signalling, however
It has been noted that nitric oxide (NO) can suppress serine racemase activity and enhance DAAO activity and thus negatively regulates D-serine concentrations (which may be mutual, as D-serine is a Nitric oxide synthase (NOS) enzyme inhibitor). This is potentially a negative feedback loop as activation of NMDA receptors causes NOS activation and elevates NO concentrations.
Activation of nitric oxide metabolism due to glutaminergic signalling may negatively regulate D-serine production and thus limit the signal enhancement from D-serine
Peripheral injections of D-serine (50mg/kg) to mice have been noted to increase hippocampal D-serine concentrations from 96.9nmol/g to 159.4nmol/g (64.5% increase) which correlates with improvements in memory. These changes were without influence on glutamate or L-serine concentrations.
D-Serine in systemic circulation is known to increase hippocampal concentrations of D-Serine, suggesting that it crosses the blood brain barrier
It has been noted that while glycine is likely the main agonist of the glycine binding site of NMDA receptors in spinal cord and in the hindbrain, the forebrain is likely most influenced by D-serine due to higher serine racemase expression (and thus more D-serine synthesis) and higher expression of glycine transporters that take glycine up into astrocytes. D-Serine has directly been measured to be higher in the forebrain area in association with where NMDA receptors are expressed.
D-Serine is likely more biologically relevant in the forebrain region than glycine is
D-Serine shares many mechanisms similar to Glycine, in the sense that it can bind to the glycine-dependent binding site on NMDA receptors (the subtype of NR1, as NR2 binds glutamate and any one NMDA receptor is a tetramer made of usually two of each) which then potentiates signalling through these NMDA receptors initially caused by glutamate or other agonists. Unlike glycine, D-Serine appears to be more effective and is active at a concentration as low as 1µM (Glycine requiring 10µM) which may not be related to their acting on the receptor itself (the two are quite comparable) but may be due to less glial cell reuptake of serine than there is with glycine.
Similar to glycine (or anything that can activate the glycine binding site), any increase in synaptic concentrations of D-serine is also accompanied by an increase in NMDAergic signalling which is thought to be due to activity at the glycine binding site being rate-limiting. Several brain regions such as the hippocampus, thalamus, neocortex, brain stem and retina have been noted to have the glycine binding site not saturated and thus respond to additional glycine or D-serine.
D-Serine, similar to Glycine, is a ligand at the NMDA receptor's glycine binding site and has the ability to potentiate glutaminergic signalling through these NMDA receptors. They are equally potent at the level of the receptor, but D-Serine appears to be more biologically relevant and overall potent
D-Serine appears to have a concentration-dependent inhibitory effect on AMPA receptors (induced by kainate acid) with an IC50 of 3.7+/-0.1mM. L-Serine does not have this effect, but this concentration may be too high to be relevant.
Although there may be inhibitory effects on AMPA receptors, the concentration appears to be very high and may not be practically relevant
In regards to excitotoxicity, D-serine and glycine appear to potentiate glutamate induced excitotoxicity with ED50 values of 47µM and 27µM (respectively) which are concentrations and 50-100fold higher than the dosages required to activate the glycine binding sites on NMDA receptors.
This enhancement of excitotoxicity is mediated not by NMDA receptors, but by activation of glycinergic receptors. Since GABA (acting via GABAA receptors) was also able to potentiate NMDA-induced excitotoxicity, it was thought to be due to the chloride influx in the neuron.
Excessive signalling through glycincergic receptors appears to potentiate NDMA induced toxicity, although the concentrations of which this occurs seem to be higher than those required to activate NMDA receptors. Practical significance of this data is unknown
Glycinergic signalling (inhibitory and via glycine receptors and causes chloride influx into neurons) can also occur with D-serine supplementation
In studies that compare glycinergic signalling between the two amino acids, glycine tends to be more potent as evidenced by lower EC50 values (27µM v. 47µM).
The transporters that mediate glycine reuptake (Glycine transporter-1 and 2) as well as the more general alanine–serine–cysteine transporter-1 (AscT1) can mediate both glycine and serine (both isomers). Due to this, both D-serine and glycine are affected by sarcosine.
D-Serine also shares signalling properties on glycinergic receptors and is subject to the same transporters that glycine is
D-Serine can be used experimentally to induce oxidation secondary to overexciting NMDA receptors of which hyperexcitation and the resulting calcium influx induce oxidative damage, and this oxidative damage from D-serine induced hyperexcitation has been noted in vitro and in vivo at 50-200mg/kg (rats) by a mechanism that is attenuated with COX2 inhibitors.
COX2 tends to be overexpressed following stressors that cause NMDA hyperexcitation such as ischemia, traumatic brain injury, and Alzheimer's and as this induction mediates the production of reactive oxygen species it is thought that COX2 inhibitors can be neuroprotective against NDMA toxicity.
Although these mechanisms are thought to be related to some pathological conditions such as Alzheimer's, the interactions of supplemental D-serine and oxidative damage is uncertain.
Overdoses or excessive administration of D-serine can cause oxidative damage (enhancing NMDA signalling too much, resulting in excitotoxicity) and it is thought that overactive D-serine metabolism also plays a role in some disease states.
The effect of nutritional supplementation is not ascertained, but oxidative damage has been induced at doses that are not too much over the standard supplemental dosages
Memory and Learning
Glutaminergic signalling is known to enhance memory formation, as activation of the NMDA receptors causes calcium influx and mobilization of calmodulin dependent kinase (CaMK) and CREB binding protein which work to induce long term potentiation (LTP) that is known as the mechanisitic basis of memory formation and causing an increase in NMDA signalling (particularly via the NR2B subunit) causes an increase in memory and LTP (This is also the memory enhancing mechanism seen with Magnesium L-Threonate). Due to the ability for D-serine to enhance signalling via the NMDA receptor (52+/-16% enhancement at 1μM and increasing activity up to 30μM) paired with the vitality of D-serine in this process and known sensitivity of hippocampal cells to D-Serine stimulation, it is thought that supplemental D-serine can promote memory and learning.
There is another phenomena known as long term depression (LTD; not the opposite of LTP) that mediates synaptic plasticity and can influence LTP indirectly, and D-serine can activate at 600-1000mg/kg injections with in vitro studies noting it effective at increasing LTD magnitude at 5μM (from 19.3% in control to 58.3%) which was more effective than both 3μM and 10μM concentrations. It seems that LTD from D-serine is mediated via its glutaminergic actions as well, and during LTD more D-serine is released from astrocytes.
D-Serine may play a role in promoting memory formation, and does so secondary to augmenting glutaminergic neurotransmission via the NMDA receptors (as D-serine can activate the glycine binding site)
The aging related process, at least as it pertains to the hippocampus, is related to a reduced ability to drive calcium-dependent neuronal plasticity which seems to be related to subactive glutaminergic receptor signalling (specifically NMDA). Due to the reduced D-serine concentrations in the brain during the aging process (possibly related to reduced expression of serine racemase) and the failure of the previous theory (the age-related decline in NMDA receptor expression may not be relevant as reduced NMDA receptor expression does not per se cause cognitive decline,) it is now thought that reduced activity at the glycinergic binding site of the NMDA receptor contributes to age-related decline (by reducing NMDA signalling and thus synaptic plasticity). This is further supported by studies noting that age-related cognitive impairment is preserved with D-serine and that NMDA-dependent plasticity is D-serine dependent.reference
The aging process causes a reduction in D-serine production (not exactly known why) and the lower levels of D-serine reducing NMDA signalling and thus may contribute to age-related cognitive decline
In regards to interventions, a study in otherwise normal mice given 50mg/kg D-serine daily noted that D-serine was able to improve memory formation after a single dose and after multiple doses. The potency of 50mg/kg D-Serine appears to be comparable to 20mg/kg D-Cycloserine which is known to be a cognitive enhancer.
D-serine is effective when given 30 minutes after training, suggesting it aids in memory consolidation. It was ineffective when given 6 hours later.
The impairment of MK-801 induced amnesia appears to be attenuated with D-serine.
It is possible that D-Serine supplementation can enhance memory formation in otherwise healthy rodents, but the studies are currently those using injections and some with quite large dosages (although the 50mg/kg has a human equivalent of 3mg/kg and is quite reasonable)
One study using 2.1g D-serine in otherwise healthy adult subjects acutely (two hours prior to cognitive testing) found improved performance in the continous performance test (CPT-IP) for sustained attention and an improvement in immediate word recall; there was an improvement in the digits forward task, but not digits backwards.
There may be minor improvements in cognition when otherwise healthy adults subjects are given D-serine supplementation
Genetic overexpression of D-serine synthesis or chronic supplementation to mice (58mg/kg over five weeks) appears to confer antidepressant effects in mice that were otherwise normal at baseline.
There may be some minor anti-depressive effects of D-serine supplementation which need to be further investigated
Alzheimer& #039;s and Dementia
NMDA-mediated neurotransmission appears to be perturbed in Alzheimer's disease and is thought to be related to the decline in memory and synaptic formation resulting in behavioural deficits. Unlike schizophrenic patients, the signalling in Alzheimer's may be hyperfunctional as Beta-amyloid peptides (β-amyloid) can accumulate both glutamate and D-serine in the synapse and both encourage their release while promoting serine racemase expression, all of which may contribute to excitotoxicity (excessive glutaminergic signalling resulting in cellular damage).
D-serine concentrations do not appear to be significantly perturbed in persons with Alzheimer's relative to control.
D-serine may be involved in the pathology of Alzheimer's disease secondary to beta-amyloid pigmentation
Symptoms of schizophrenia (particularly negative symptoms) are currently thought to be related to glutaminergic hypofunction (a reduced total signalling capacity via glutamate receptors), and recent pharmacological therapies that aim to restore glutaminergic firing include serine/glycine restoration (as despite total elevated serine and glycine being higher in schizophrenics versus control, D-serine itself is reduced suggesting problems with serine racemase) since impairing glycine binding to NMDA receptors causes negative symptoms of schizophrenia and serine racemase knockout mice (or anything to reduce production of D-serine) have schizophrenic symptoms as well yet knocking out D-amino acid oxidase (and preventing degradation of D-serine) is rehabilitative. Finally, clinical remission of schizophrenia is accompanied by an increase in D-serine concentrations independent of supplementation.
Other possible therapeutic options include AMPAkines that enhance signalling via AMPA receptors (this includes Piracetam and Aniracetam), and indirectly supporting the aforementioned NMDA signalling via inhibiting glycine uptake into cells and promoting their synaptic effects (seen with sarcosine). Enhancing AMPA signalling will inherently increase glutaminergic signalling and can remove the magnesium block from NMDA receptors
Negative symptoms of schizophrenia refer to affective flattening and social isolation whereas hallucinations, delusions, and thought disorder are referred to as 'positive' symptoms and the cognitive impairment that is associated with schizophrenia in neither category.
D-Serine administration, via affecting the NMDA glycine binding site and thus positively regulating signalling through NDMA receptors, is thought to be able to reduce symptoms of schizophrenia. This is somewhat supported by the fact that schizophrenia appears to be a D-serine deficiency state (role in cause or effect not established)
Limited positive studies tend to note 17-30% improvements in negative symptoms of schizophrenia with 30mg/kg (2.12+/-0.6g overall) D-serine as assessed by the Positive and Negative Syndrome Scale (PNSS) which is a potency similar to 800mg/kg Glycine under similar research conditions. Studies that measure negative symptom progression over time note beneficial effects within 2 weeks of supplementation, which increases in potency over 6 weeks of observation and is more apparent with higher doses in the 60-120mg/kg range. When looking specifically at positive symptoms, 30mg/kg (2.12+/-0.6g overall) D-serine has noted significant improvement after 6 weeks although the trend noted at weeks 2 and 4 was not statistically significant. Improvements have been noted with 60-120mg/kg to a higher degree than 30mg/kg, and benefits on both positive and negative symptoms are correlated with serum exposure to D-serine.
2,000mg of D-serine daily for 16 weeks in schizophrenics in addition to standard anti-psychotic medication failed to find a significant benefit of supplementation over placebo, although the authors cautioned that the larger than normal placebo response could in part explain the results although D-serine has elsewhere failed to be any significantly different than placebo at the 30mg/kg or 2,000mg dosage. These studies do note that individuals get benefit albeit not consistenly enough to reach statistical significance, and this paired with the correlation between D-serine in the blood and the benefits to symptoms suggest that the established variance in serum D-serine from oral therapy may underlie the null effects observed.
D-Serine appears to be effective at reducing all symptoms of schizophrenia (mostly negative and cognitive, with less effects on positive) but the standard recommended dose of 30mg/kg seems unreliable. This may be due to a large variability in how much D-serine reaches the blood, and taking higher doses seems to be more reliable based on limited evidence
Parkinson& #039;s Disease
Parkinson's disease seems to have some symptoms (impairments in motivation, drive, and initiation/emotional reactivity) that are similar to the negative symptoms in schizophrenia (apathy, flat affect and isolation), and due to the similarities it is thought that D-serine could be useful. Furthermore, dopaminergic neurons in the striatum are involved in NMDA signalling while NMDA receptors in persons with Parkinson's are known to be altered.
Supplementation of D-serine at 30mg/kg for six weeks in a small pilot study with 13 persons with Parkinson's disease (ended up being in the range of 1,600-2,600mg daily) was able to reduce symptoms as assessed by the Unified Parkinson’s Disease Rating Scale, Simpson-Angus scale, and Positive and Negative Syndrome Scale. This study noted that when looking at subjects with a 20% improvement in symptoms, 50-70% of persons in the D-serine group met this criteria while 10-20% in placbo did.
Preliminary evidence suggests that D-serine could be useful for treatment of Parkinson's disease
Stress and Trauma
NMDA receptors appear to be somewhat involved in some symptoms of PTSD including dissociation and perceptual alterations and since ketamine (NMDA antagonist) can also cause these particular symptoms it is thought they are caused by NMDA underfiring, particularly in the hippocampus or amygdala.
D-Cycloserine (partial agonist at the glycine binding site of NMDA receptors, whereas D-serine is a full agonist) has previously shown benefit in a pilot study for reducing symptoms of post-traumatic stress disorder (PTSD; numbing, avoidance, and anxiety symptoms mostly) and following that study one using D-serine at 30mg/kg for six weeks noted improvements on anxiety symptoms (HAMA; 95% CI of 13.4–46.7% symptom reduction), depressive (HAMD; 95% CI of 2.0–43.3% symptom reduction) and CAPS score (95% CI of 10.9-31% reduction).
Preliminary evidence suggests benefit in treating symptoms of Post Traumatic Stress Disorder (PTSD), although the observed benefits seem fairly unreliable
Amyotrophic lateral sclerosis
Amyotrophic lateral sclerosis (ALS) in mice (mSOD1 strain) is associated with 50-100% elevation in D-serine concentrations of spinal fluid which can predict how susceptible neurons from ALS are to NMDA-mediated excitotoxic damage. Although this suggests that elevated D-serine contributes to ALS pathology, it has also been noted that hindering the serine racemase enzyme accelerates disease onset (described as paradoxical) but slowed disease progression which was mimicked by D-serine supplementation in the chow.
D-Serine has unclear influences on the pathology and onset of ALS
Cocaine addiction is known to result in alterations to glutamatergic synaptic plasticity which precedes addictive behaviour which are thought to be related to NMDA receptors (both long term potentiation (LTP) and long term depression (LTD) are implicated).
D-Serine has been noted to be reduced in the nuclear accumbens core (where it is the coagonist of synaptic receptors) of rats undergoing cocaine withdrawal, which is thought to contribute to NMDA hypoactivity and relapse as incubation with their neuronal slices with D-serine normalizes the cocaine-induced changes in LTP and LTD. This was confirmed when 10-100mg/kg of D-serine oral ingestion (or 100mg/kg injections) to cocaine dependent rats reduced their drug-seeking behaviour.
The study to assess how D-serine affects sucrose preference failed to find a significant benefit of supplementation.
Cocaine addiction is associated with alterations in synaptic plasticity from alterations in NMDA function, and D-serine has been noted to be reduced in rats undergoing cocaine withdrawal. Preliminary evidence suggests that D-serine has anti-addictive properties
Safety and Toxicity
Human studies that tend to use 30mg/kg of D-serine (around 2,000mg total) daily for periods of up to six weeks do not tend to notice any side-effects with one preliminary study using 120mg/kg (around 8,000mg) also failing to find significant side-effects relative to control.
The standard supplemental dosages of D-serine do not note any significant side-effects associated with treatment in the standard dosage range