Identity of the NMDA receptor coagonist is synapse specific and developmentally regulated in the hippocampus.

NMDA receptors (NMDARs) require the coagonists D-serine or glycine for their activation, but whether the identity of the coagonist could be synapse specific and developmentally regulated remains elusive. We therefore investigated the contribution of D-serine and glycine by recording NMDAR-mediated responses at hippocampal Schaffer collaterals (SC)-CA1 and medial perforant path-dentate gyrus (mPP-DG) synapses in juvenile and adult rats. Selective depletion of endogenous coagonists with enzymatic scavengers as well as pharmacological inhibition of endogenous D-amino acid oxidase activity revealed that D-serine is the preferred coagonist at SC-CA1 mature synapses, whereas, unexpectedly, glycine is mainly involved at mPP-DG synapses. Nevertheless, both coagonist functions are driven by the levels of synaptic activity as inferred by recording long-term potentiation generated at both connections. This regional compartmentalization in the coagonist identity is associated to different GluN1/GluN2A to GluN1/GluN2B subunit composition of synaptic NMDARs. During postnatal development, the replacement of GluN2B- by GluN2A-containing NMDARs at SC-CA1 synapses parallels a change in the identity of the coagonist from glycine to D-serine. In contrast, NMDARs subunit composition at mPP-DG synapses is not altered and glycine remains the main coagonist throughout postnatal development. Altogether, our observations disclose an unprecedented relationship in the identity of the coagonist not only with the GluN2 subunit composition at synaptic NMDARs but also with astrocyte activity in the developing and mature hippocampus that reconciles the complementary functions of D-serine And Glycine In Modulating Nmdars During The Maturation Of Tripartite Glutamatergic Synapses.

A CSF and postmortem brain study of D-serine metabolic parameters in schizophrenia.

Clinical trials demonstrated that D-serine administration improves schizophrenia symptoms, raising the possibility that altered levels of endogenous D-serine may contribute to the N-methyl D-aspartate receptor hypofunction thought to play a role in the disease. We hypothesized that cerebro-spinal fluid (CSF) D-serine levels are decreased in the patients due to reduced synthesis and/or increased degradation in brain. We now monitored amino acid levels in CSF from 12 schizophrenia patients vs. 12 controls and in postmortem parietal-cortex from 15 control subjects and 15 each of schizophrenia, major-depression and bipolar patients. In addition, we monitored postmortem brain serine racemase and D-amino acid oxidase protein levels by Western-blot analysis. We found a 25% decrease in D-serine levels and D/L-serine ratio in CSF of schizophrenia patients, while parietal-cortex D-serine was unaltered. Levels of L-serine, L-glutamine and L-glutamate were unaffected. Frontal-cortex (39%) and hippocampal (21%) serine racemase protein levels and hippocampal serine racemase/D-amino acid oxidase ratio (34%) were reduced. Hippocampal D-amino-acid-oxidase protein levels significantly correlated with duration of illness (r=0.6, p=0.019) but not age. D-amino acid oxidase levels in patients with DOI>20 years were 77% significantly higher than in the other patients and controls. Our results suggest that reduced brain serine racemase and elevated D-amino acid oxidase protein levels may contribute to the lower CSF D-serine levels in schizophrenia.

Slow-binding human serine racemase inhibitors from high-throughput screening of combinatorial libraries.

One-bead one-compound combinatorial chemistry together with a high-throughput screen based on fluorescently labeled enzyme allowed the identification of slow binding inhibitors of human serine racemase (hSR). A peptide library of topographically segregated encoded resin beads was synthesized, and several hSR-binding compounds were isolated, identified, and resynthesized for further kinetic study. Of these, several showed inhibitory effects with moderate potency (high micromolar K(I)s) toward hSR. A clear structural motif was identified consisting of 3-phenylpropionic acid and histidine moieties. Importantly, the inhibitors identified showed no structural similarities to the natural substrate, L-serine. Detailed kinetic analyses of the properties of selected inhibitors show that the screening protocol used here selectively identifies slow binding inhibitors. They provide a pharmacophore for the future isolation of more potent ligands that may prove useful in probing and understanding the biological role of hSR.