The Group III metabotropic glutamate receptors (mGluR) are located presynaptically on axon terminals and act as autoreceptors and heteroreceptors by inhibiting neurotransmitter release. Much has been learned about these receptors through exogenous application of L-serine-O-phosphate (L-SOP), an endogenous amino acid derivative and known activator of the Group III mGluRs. We hypothesized that L-SOP is the endogenous co-agonist at the high affinity Group III mGluR, mGluR4. We found the EC50 of L-SOP at mGluR4 was 0.5 μM, and determined that the concentration of L-SOP in whole brain was approximately 5 μM. An immunocytochemical survey revealed that cells containing the enzymatic machinery necessary for L-SOP synthesis and metabolism were observed in two brain regions known to express mGluR4, namely, cerebellum and hippocampus. In the cerebellum, the L-SOP synthetic and metabolic enzymes were found in Bergmann glia and Purkinje cells, two cells which form a tripartite synapse with parallel fiber axon terminals where the mGluR4 subtype is exclusively expressed at high levels. In the hippocampus, the L-SOP metabolic enzyme was detected in young neurons emanating from the neurogenic subventricular zone. Attempts to raise endogenous levels of L-SOP by crippling the L-SOP metabolizing enzyme (phosphoserine phosphatase), over-expressing the L-SOP synthesizing enzyme (phosphoserine aminotransferase), or through dietary protein restriction, to study the effects on neurotransmission and neurodevelopment in the central nervous system (CNS) were unsuccessful, suggesting that the production of L-SOP remains stable despite manipulation of the synthetic and metabolic enzymes. Finally, the ability of L-SOP to modulate glutamate release from presynaptic terminals was examined in cerebellar synaptosomes. Co-incident activation of presynaptic mGluR4 and presynaptic GABAA receptors facilitated glutamate release, suggesting that simultaneous activation of parallel fibers and Bergmann glia may serve to enhance synaptic transmission. This observation expands the traditional view of Group III mGluRs acting solely as inhibitory autoreceptors. Taken together, these results provide compelling evidence to support the hypothesis that L-SOP is the endogenous agonist at mGluR4, and possibly other Group III mGluRs.

The Group III metabotropic glutamate receptors (mGluR) are located presynaptically on axon terminals and act as autoreceptors and heteroreceptors by inhibiting neurotransmitter release. Much has been learned about these receptors through exogenous application of L-serine-O-phosphate (L-SOP), an endogenous amino acid derivative and known activator of the Group III mGluRs. We hypothesized that L-SOP is the endogenous co-agonist at the high affinity Group III mGluR, mGluR4. We found the EC50 of L-SOP at mGluR4 was 0.5 μM, and determined that the concentration of L-SOP in whole brain was approximately 5 μM. An immunocytochemical survey revealed that cells containing the enzymatic machinery necessary for L-SOP synthesis and metabolism were observed in two brain regions known to express mGluR4, namely, cerebellum and hippocampus. In the cerebellum, the L-SOP synthetic and metabolic enzymes were found in Bergmann glia and Purkinje cells, two cells which form a tripartite synapse with parallel fiber axon terminals where the mGluR4 subtype is exclusively expressed at high levels. In the hippocampus, the L-SOP metabolic enzyme was detected in young neurons emanating from the neurogenic subventricular zone. Attempts to raise endogenous levels of L-SOP by crippling the L-SOP metabolizing enzyme (phosphoserine phosphatase), over-expressing the L-SOP synthesizing enzyme (phosphoserine aminotransferase), or through dietary protein restriction, to study the effects on neurotransmission and neurodevelopment in the central nervous system (CNS) were unsuccessful, suggesting that the production of L-SOP remains stable despite manipulation of the synthetic and metabolic enzymes. Finally, the ability of L-SOP to modulate glutamate release from presynaptic terminals was examined in cerebellar synaptosomes. Co-incident activation of presynaptic mGluR4 and presynaptic GABAA receptors facilitated glutamate release, suggesting that simultaneous activation of parallel fibers and Bergmann glia may serve to enhance synaptic transmission. This observation expands the traditional view of Group III mGluRs acting solely as inhibitory autoreceptors. Taken together, these results provide compelling evidence to support the hypothesis that L-SOP is the endogenous agonist at mGluR4, and possibly other Group III mGluRs.

Avant sa libération, le glutamate est accumulé dans des vésicules synaptiques par trois transporteurs vésiculaires (VGLUT1-3). Les cellules ciliées internes (CCI) de la cochlée n'expriment que VGLUT3. Pour étudier son rôle dans la physiologie cochléaire, nous avons utilisé une lignée de souris dont le gène Slc17a8, qui code pour VGLUT3, a été invalidé par recombinaison homologue. Les mutants ne présentaient pas de réponse nerveuse à une stimulation sonore. Les mécanismes d'exocytose des CCI étaient normaux et leurs synapses normales en microscopie électronique. Des immunoblots montraient que le transporteur membranaire du glutamate GLAST, ainsi que les sous-unités GLUR2 et NR1 des récepteurs AMPA et NMDA étaient toujours exprimées. Enfin, des potentiels auditifs du tronc cérébral étaient enregistrés après une stimulation électrique au niveau de la fenêtre ronde. Toutefois, nos résultats indiquent des diminutions de ~50% des synapses afférentes et de ~40% des neurones auditifs primaires ainsi qu'une réduction importante des terminaisons efférentes latérales sous les CCI.SLC17A8 est responsable de la surdité de perception non syndromique dominante DFNA25. Nous avons identifié une mutation dans l'exon 5 conduisant au remplacement de l'Alanine211 en Valine. Cette Alanine est conservée dans les VGLUT3 de différentes espèces ainsi que dans les VGLUT1-3 humains, suggérant un rôle fonctionnel important pour cet acide aminé. Nous avons caractérisé les propriétés biochimiques de la mutation A211V en culture de cellules. Le transporteur muté était correctement adressé aux boutons présynaptiques. Cependant, la mutation pA211V entraîne un défaut d'expression important en partie expliqué par le fait que le codon codant la valine est un codon rare. De plus, les études du transport de glutamate ont montré que la forme mutée est hyperactive par rapport à la forme native. L'ensemble de ces résultats montre que la mutation entraine un phénotype cellulaire complexe. / Before its release, glutamate is accumulated into synaptic vesicles by three vesicular glutamate transporters (VGLUT1-3). Only VGLUT3 is expressed in the inner hair cells (IHCs) of the cochlea. To study its role in the hearing physiology, we used a mouse in which the Slc17a8 gene, which encodes VGLUT3, has been null-mutated. In this VGLUT3-/- mouse, no auditory nerve response to acoustic stimuli could be recorded. All the others cochlear potentials were normal. The genetic deletion of Slc17a8 in mice resulted in a profound deafness, without altering the IHCs synapse morphology and the synaptic vesicles turnover. Using western blot, we then observed that the glutamate-aspartate transporter GLAST and the GLUR2 and NR1 subunits of AMPA and NMDA receptors were always expressed. Finally, auditory brainstem responses could be elicited by electrical stimuli on the round window. However, VGLUT3-/- IHCs presented a ~50% loss of IHCs synapses and a ~40% loss of primary auditory neurons. The number of lateral olivocochlear synapses with primary auditory neurons dendrites was strongly reduced.The SLC17A8 gene is responsible for DFNA25, an autosomal dominant progressive, high-frequency nonsyndromic deafness. We identified a heterozygous non-synonymous missense mutation in exon 5, leading to the amino acid change p.A211V. The A211 residue is conserved in VGLUT3 across species and in all the human VGLUT subtypes (VGLUT1-3), suggesting an important functional role. We characterized the biochemical properties of the A211V mutation in cell culture. Our results suggest that the mutated VGLUT3 was correctly addressed at the presynaptic boutons. However, the pA211V mutation induced an expression decrease because the valine codon is a rare codon. Moreover, the glutamate uptake is increased with the mutated VGLUT3. All these results shows that this mutation involves a complex cellular phenotype.