The glutamate hypothesis of depression: the effect of stress and glucocorticoids on glutamate synapse and the action of antidepressants

Abstract

In the last years, a consistent number of clinical and preclinical studies have demonstrated that glutamatergic transmission has a primary role in the pathophysiology of mood and anxiety disorders (MADI). It has been shown that in depressed patients the levels of glutamate and its metabolites are altered in plasma and in selected brain areas and mRNA and protein levels of glutamate receptors are changed in brain areas. A number of preclinical studies on animal models of MADI have shown that different types of environmental stress and glucocorticoid administration affect glutamate transmission and exert structural brain remodeling in the same areas involved in human pathology. These effects of stress and glucocorticoids have been associated with the onset and exacerbation of neuropsychiatric disorders.In previous studies we found that acute footshock (FS)-stress induces an increase of glutamate release from synaptosomes of prefrontal and frontal cortex (PFC/FC), via glucocorticoid receptors (GR) activation and SNARE complex accumulation in synaptic membranes. Furthermore, we have demonstrated that the increase of glutamate release induced by acute stress is prevented by chronic antidepressants (ADs). Additional studies have also shown that ADs can regulate glutamate transmission through glutamate receptors; reducing the function of NMDA receptors, potentiating the function of AMPA receptor and affecting different subtypes of metabotropic glutamate receptors. Together, these findings have identified the glutamate synapse as a target for novel glutamatergic ADs.Considering the importance of stress-induced alteration of presynaptic glutamate release in the pathophysiology of MADI, we aimed to study whether the enhancement of depolarization-evoked glutamate release induced by acute stress was related to an increase of the readily releasable pool (RRP) of vesicles and whether this effect was mediated by a synaptic non-genomic action of corticosterone (CORT).We found that FS-stress increased glutamate release evoked by hypertonic sucrose (which mobilizes exclusively the RRP), suggesting an increase in the RRP size. Then we found that this synaptic effect of stress was dependent on local CORT action. Indeed, CORT was able to directly affect the RRP size through the activation of GR and mineralcorticoid receptors (MR). The preincubation with RU486, selective GR antagonist, and spironolactone, selective MR antagonist, prevented the CORT-induced increase of RRP. Contrary to acute stress, CORT by itself did not promote vesicle fusion, since CORT application in vitro did not increase glutamate release evoked by depolarization of control synaptosomes, and did not affect excitatory post-synaptic potentials and paired pulse facilitation in mPFC slices. Furthermore, we also found that CORT increased vesicle mobilization towards the RRP via GRs and MRs activation, by using total internal fluorescence microscopy, a technique that allows the study of events occurring in a 100 nm-interval below the plasma membrane. Finally we found that stress and CORT modulated synapsin I, a protein involved in vesicle mobilization and in vesicles docking, fusion and recycling at active zones. We found that both FS-stress and CORT induce an increase of synapsin I phosphorylation in synaptic membranes selectively at site 1. The preincubation with both RU486 and spironolactone, prevented the CORT-induced synapsin I phosphorylation at site I, suggesting that this protein is involved in the pathway downstream of activation of the two receptors. Together these results suggest that the increase of the RRP size induced by stress is promoted by a local action of CORT on synaptic receptors. We speculated that CORT is necessary to promote an increase of the RRP, but not sufficient to increase depolarization-dependent glutamate release, suggesting that additional mediators or neurotransmitters, released during the stress response, are necessary to trigger vesicle release