New evidence of specific synaptic remodelling by corticosterone-treated astrocytes conditioned medium

Major depressive disorder (MDD) is a debilitating multifactorial neuropsychiatric syndrome, affecting about 20% of the population and representing the leading cause of disability worldwide with severe social and economic consequences [1]. Stress exposure has been recognized as the main risk factor and more in detail, an individual’s ability to cope with stress in an adaptive way can determine their resilience or vulnerability to MDD development [2]. MDD is characterized by many alterations among which, in most patients, hyperactivity of the hypothalamic-pituitary-adrenal (HPA) axis and glucocorticoid resistance. In patients with severe depression, increased levels of serum and salivary cortisol have been found [3]. Astrocytes are glial cells fundamental for the central nervous system, functioning as neuronal support and participating in the regulation of ion homeostasis, neurotransmission, synaptic plasticity and neuroinflammation. Evidence from both human post-mortem brains and animal models indicate the involvement of astrocytes in MDD pathophysiology [4]. In literature the impact of glucocorticoids on astrocyte alone is still understudied and even less is known about how treated astrocytes can affect neuronal function. Thus, to shed light on this topic, we treated mouse primary hippocampal astrocytes with corticosterone (CORT) at low concentration or DMSO twice a day for 3 days and then used their conditioned medium (ACM) to treat mouse primary hippocampal neurons at DIV 17 for either 1 or 24 hours. We investigated the impact of such treatment on neurons by (i) recording miniature post-synaptic currents (mEPSCs and mIPSCs) in whole-cell patch clamp configuration, and by (ii) performing immunofluorescence experiments targeting vGLUT1 and vGAT, markers of the excitatory and inhibitory pre-synaptic compartments respectively. Interestingly, ACM-CORT-treated neurons displayed specific and persistent changes of both mEPSCs and mIPSCs frequencies compared to ACM-DMSO-treated neurons at both 1 and 24 hours whereas only minimal changes appeared in terms of amplitude. In parallel and in line, ACM-CORT-treated neurons showed changes in the expression of excitatory and inhibitory synaptic puncta. The TUNEL assay showed no changes in the number of apoptotic nuclei, indicating that the functional and structural variations in ACM-CORT-treated neurons were not related to neuronal loss. We could conclude that the medium conditioned by astrocytes treated chronically with CORT has a toxic effect on neuron synapses, specifically affecting the pre-synaptic and post-synaptic compartment at 1 hour and then at 24 hours. In order to pinpoint the mechanisms by which ACM-CORT and ACM-DMSO have different effects on neurons, ACMs were analyzed through liquid chromatography-tandem mass spectrometry (LC-MS/MS) with label-free quantification. Of all the 1193 proteins present in both conditions, a total of 636 were significantly varied: 451 were decreased and 221 were increased in ACM-CORT compared to ACM-DMSO. 16 proteins were present only in ACM-DMSO and 20 were present only in ACM-CORT. In the next future, understanding exactly which factors released from CORT-treated astrocytes are impacting on neuronal function will be important to better understand how astrocytes and glucocorticoids can be involved in MDD pathophysiology and potentially be targeted to discover new treatments.