As indicated above, MDD pathophysiology involves a wide array of diverse factors and processes, including two gut processes, namely: (1) gut permeability-induced circulating LPS and activation of the TLR4/ Nuclear factor kappa-light-chain-enhancer of activated B cells (NF-kB) /yin yang (YY)1 pathway; and (2) decreases in the short-chain fatty acid, butyrate and therefore attenuation of butyrate’s epigenetic regulation via histone deacetylase (HDAC) inhibition as well as the loss of butyrate’s mitochondria-optimizing effects via upregulation of the sirtuin-3 deacetylation and disinhibition of PDC. The conversion of pyruvate to acetyl-CoA by PDC not only increases ATP production from the tricarboxylic acid (TCA) cycle and OXPHOS, but also makes acetyl-CoA available as a necessary co-substrate for the initiation of the melatonergic pathway. The latter may be of some importance, given that the antioxidant and glutathione (GSH)-inducing effects of intracrine and autocrine melatonin suppress the PDC induction of ROS [113]. Increased LPS and suppressed butyrate levels are two ways in which the gut microbiome/permeability act to regulate cellular and mitochondrial function across systemic and central cells, with important consequences for reactive cells, including glia and immune cells. Within the CNS such gut microbiome-driven effects have been mostly characterized, to date, as occurring in microglia, rather than astrocytes [23]. As a HDAC inhibitor, butyrate also increases the astrocyte melatonin MT1 receptor [114], indicating that variations in gut microbiome-derived butyrate will modulate the levels and effects of autocrine melatonin in astrocytes and other cells across the body.

The butyrate optimization of mitochondrial function requires the melatonergic pathway [115], via butyrate/sirtuin-3/PDC providing the acetyl-CoA that is a necessary co-substrate for the conversion of serotonin to N-acetylserotonin (NAS) and thereafter melatonin. The loss of butyrate-induced melatonin or the suppression of the melatonergic pathway via a decrease in tryptophan availability, tryptophan uptake, 14-3-3ε stabilization of tryptophan hydroxylase (TPH)2, 14-3-3ζ stabilization of arylalkylamine N-acetyltransferase (AANAT) and/or acetyl-CoA will decrease mitochondrial antioxidant status, including via the loss of melatonin/GSH induction [113]. The suppression of endogenous antioxidants markedly increases oxidants produced by the PDC [116]. Given that mitochondrial ROS is a major driver of oxidant regulation of patterned microRNAs (miRNAs) via miRNA biogenesis, transcription, and epigenetic modifications, a significant change in the patterning of miRNAs and associated changes in gene patterning will occur [50,117]. Gut microbiome/permeability factors therefore have significant impacts on patterned gene expressions, and their intra- and inter-cellular consequences, via the mitochondria melatonin/ROS/miRNAs pathway, forming the underpinnings of why variations in the gut microbiome/permeability significantly impacts on seemingly all medical conditions, including MDD.

Key sites where such changes will have significant impacts in the development of mood dysregulation, include in the mitochondria of enteric glial cells, intestinal epithelial cells, CNS astrocytes and microglia, as well as in systemic immune cells and circulating platelets. This gut microbiome/permeability influence does not emerge in adulthood, but seems evident from early development, as indicated by the data showing the gut microbiome/permeability to regulate amygdala development, and therefore contributing to developmental variations in emotional regulation of brain area interactions underpinning mood disorders. It is by such processes that the gut modulates astrocyte-neuronal interactions in the pathophysiology of MDD. Recent data highlighting the role of the opioidergic system in MDD may be linked to this [13].

Alterations in the opioidergic system, especially the μ-/κ-opioid receptor ratio and their ligands, β-endorphin and dynorphin, respectively, are a cutting-edge area of research in MDD pathophysiology [13,118]. The gut microbiome significantly modulates the opioidergic system, including via butyrate’s upregulation of the μ-opioid receptor [119]. The μ-opioid receptor is intimately linked to social bonding, including from oxytocin’s partial μ-opioid receptor agonist effects [120]. The butyrate induction of mitochondrial melatonin upregulates the endogenous μ-opioid receptor ligand, β-endorphin [121], indicating that some of the beneficial effects of butyrate may be dependent upon its capacity to upregulate cellular melatonin. In contrast, κ-opioid receptor activation in the amygdala induces dysphoria [122]. These regulatory effects of butyrate on the opioidergic system are intimately associated with its capacity to optimize mitochondrial function and upregulate the glia mitochondrial melatonergic pathway, including in the amygdala and associated paracapsular cells of the intercalated masses [38,123], thereby regulating amygdala-PFC interactions.

The amygdala inputs to the ventral tegmental area (VTA)-N.Acc junction are an important determinant of affect-driven motivated behaviors, with the opioidergic system also a significant direct regulator of VTA-N.Acc associated motivated behaviors [124]. The HDACi effects of butyrate in the N.Acc significantly upregulates motivated behaviors, driven by μ-opioid receptor activation [125]. The κ-opioid receptor in these brain reward regions correlates negatively with social status [126]. Such data highlights how the gut can act on opioidergic processes in different brain regions in the modulation of MDD pathophysiology. It is important to note that such gut regulation of the opioidergic system in the amygdala and VTA-N.Acc will be determined by effects on astrocyte-neuronal interactions, involving alterations in the mitochondrial melatonergic pathway in these different brain regions, and will be important aspects of the early developmental processes regulating such inter-area communication across brain regions [127,128]. Importantly, the amygdala can over-ride cortex and hippocampal inputs into the VTA-N.Acc junction, as shown in rodents, highlighting the relative strength of affect over higher order cognition in the regulation of motivated behavioral outputs [129,130].

The amygdala and VTA-N.Acc junction are intimately linked to MDD pathophysiology. However, it is the regulation of astrocytes and the astrocyte mitochondrial melatonergic pathway that is a crucial determinant of astrocyte-neuronal interactions at these important hubs in MDD pathophysiology. Gut-derived factors such as LPS and butyrate act on MDD pathophysiology at these brain sites via their impact on astrocyte mitochondrial function and thereby on astrocyte-neuronal interactions.

The significance of circulating ceramide in MDD will be important to clarify in future research [93,94]. Ceramide can be derived from many cellular sources [96,97,131], including astrocytes [98], contributing to the association of MDD with neurodegenerative conditions via suppressed mitochondrial function [98,132]. ROS and proinflammatory cytokines induce ceramide-producing enzymes, leading to ceramide’s inhibition of the mitochondrial respiratory chain and suboptimal mitochondrial function [132]. This allows ready links to alterations in the gut microbiome/permeability and the mitochondrial melatonergic pathway, given their regulation of ROS and pro-inflammatory cytokines.

Gut dysbiosis/permeability and decreased butyrate in MDD [133] raises levels of gut-derived, circulating trimethylamine N-oxide (TMAO). Heightened TMAO is evident in MDD [134] and activates platelets, which further contribute to ceramide production and release, and to the association of MDD with other medical conditions [135]. Importantly, gut microbiome-derived butyrate suppresses ceramide levels via butyrate converting ceramide to glucosylceramide, which then acts as a precursor for an array of gangliosides [136]. Butyrate, like melatonin, also suppresses platelet activation [98] and microglia activation [137], indicating the importance of the gut microbiome in regulating the ceramide levels via a number of different routes across different cell types. It will be important to determine as to whether the effects of butyrate in suppressing ceramide require the induction of the mitochondrial melatonergic pathway, including in platelets, endothelial cells and systemic immune cells as well as glia, and the consequences that this has on astrocyte-neuronal interactions. Such data will contribute to an understanding of the heterogeneity of MDD pathophysiology, and how systemic cells and their interactions act to modulate astrocyte-neuronal interactions.

Preclinical data shows that exposure to the gut microbiome-derived short-chain fatty acid, propionate, in adolescence produces significant changes in the mitochondria of amygdala astrocytes, microglia and neurons with immediate changes in social behavior [138], with relevance to high addiction comorbidity [139] and the development of personality traits linked to MDD susceptibility [140,141]. Such data highlights the diversity and complexity of data pertaining to MDD and the challenge, and opportunity, that such data provides in finding a parsimonious physiological model of human affect and trait development.

Suppressed neurogenesis is a classical aspect of MDD pathophysiology. Neurogenesis involves the interactions of astrocytes and neuronal progenitor cells in the dentate gyrus, including via astrocyte BDNF activation of TrkB. Notably, TrkB is also activated by NAS [142], indicating that the backward conversion of melatonin to NAS, including by activation of the purinergic receptor P2Y1, metabotropic glutamate receptor, mGluR5, and the AhR [143], will increase NAS production and release [63], with NAS activation of TrkB increasing neurogenesis [144]. NAS also induces hippocampal BDNF [145]. As such, the astrocyte mitochondrial melatonergic pathway provides response plasticity that can be influenced by local and systemic processes. As to how circulating ceramide suppression of neurogenesis in MDD [96] interacts with the astrocyte mitochondrial melatonergic pathway requires investigation.

MDD pathophysiology has been linked to diverse bodies of data, including discrimination stress, at least in part by its association to a wide range of environmental and social factors that ultimately act on physiological interactions across the body over the course of development. Such a ‘holistic’ perspective also has relevance for the early developmental pathoetiology of a host of adult-onset medical conditions, including Parkinson’s disease, amyotrophic lateral sclerosis (ALS), multiple sclerosis and rheumatoid arthritis, as well as to how MDD interacts with, and may trigger exacerbations in, such conditions [146,147,148,149]. Engaging the complexity of MDD therefore has consequences for the conceptualization and treatment of diverse medical conditions that are currently poorly managed.

The wide range of data reviewed above and their impact on glia-neuronal interactions provide an integrated framework for understanding MDD pathophysiology within a developmental frame of reference, upon which many genetic and epigenetic factors may act. This framework is derived from an assumption that the evolutionary modified bacteria in the form of mitochondria that drive energy provision in almost all body cells are important determinants of cellular function, including patterned miRNA and gene expressions. This allows intercellular interactions to be conceived, at least partly, as a form on mitochondria-to-mitochondria communication, with the diverse and dynamic changes occurring in cellular fluxes forming the means of such communication. The intimate association of the melatonergic pathway with mitochondria, seemingly from the very first multicellular organism [150], and its maintained presence across the three kingdoms of life (animals, plants and fungi) over the course of 2–2.5 billion years of evolution strongly indicates that the mitochondrial melatonergic pathway is a core physiological process. The suboptimal mitochondrial function evident across systemic and central cells in MDD and the long-standing association of decreased serotonin in MDD may be framed within this context, given the necessity of serotonin as a precursor for the initiation of the melatonergic pathway. Cellular interactions may thereby be conceived, at least partly, as a form of modified bacteria communication with homeostasis being dependent upon all interacting cells being able to optimize their mitochondrial melatonergic pathway activity. Ageing, and factors that dysregulate the capacity of cells to maintain more optimal function of their melatonergic pathway, lead to alterations in the homeostatic status quo, which is the essence of neuroprogression across psychiatric conditions. Higher order body and brain systems develop within this basic framework and are shaped by the specific cell types in which mitochondria reside.

Within this context, the bacteria of the gut, lung, skin, and placental microbiomes are in interaction with their evolutionary modified distant cousins in the form of mitochondria. This has parallels to recent thinking on the tumor microenvironment, where cancer metabolism acts to regulate mitochondrial function, and the melatonergic pathways, in the other tumor microenvironment cells [151], allowing cancer cells to dominate the nature of local processes and therefore form a new homeostasis. As to whether such mitochondria/metabolic dominance occurs in other collections of body cells will be interesting to determine, including as to whether this drives a new homeostatic regulation, such as in subregions of the anterior cingulate in MDD. This provides a different conceptualization of aging, and how aging positively associates with almost all medical conditions. Within the tumor microenvironment such ‘metabolic dominance’ is achieved, at least partly, by IDO induction and the release of kynurenine, which activates the AhR in natural killer cells and CD8+ t cells, leading to the induction of a metabolic state of ‘exhaustion’, whereby these cells lose their capacity to kill cancer cells and virus-infected cells [152]. Another readily achievable means by which ‘metabolic dominance’ may be achieved is via the induction or exosomal/vesicular transfer of miRNAs that suppress the melatonergic pathway, such as the suppression of 14-3-3 by miR-7, miR-375, miR-451 and miR-709 [153,154,155], with consequences for the cells in which these miRNAs are induced, including melatonergic pathway suppression and alterations in the regulation of ROS, ROS-driven miRNAs and therefore patterned gene transcriptions. miR-709 also suppresses mitochondrial transcriptional factor A (TFAM) [156], although with some distinct effects in different cell types, indicating that wider aspects of mitochondrial function will be coordinated with alterations in melatonergic pathway regulation [157]. Importantly, such miRNA-suppression of 14-3-3 isoforms and the melatonergic pathway will have detrimental effects on the capacity of butyrate to optimize mitochondrial function, indicating that the beneficial effects of butyrate are intimately intertwined with the capacity of a given cell to regulate the melatonergic pathway. In the case of a suppressed mitochondrial melatonergic pathway, butyrate may not then be able to restore the homeostatic status quo. Aging-associated increases in miR-709 in the murine liver suppresses hepatic function [158], suggesting that the suppression of the mitochondrial melatonergic pathway may be a significant aspect of organ dysfunction over aging, including in the liver but also in other organs.

This indicates a resetting of local interactions that is coupled to a suppressed capacity of gut microbiome-derived butyrate to return these interactions to their previous homeostatic state. This would also suggest a suppressed capacity of circadian pineal melatonin to reset mitochondrial function at night, which would be further contributed to by the dramatic decrease in pineal melatonin over the course of aging. It is generally assumed that over the course of recurrent depression episodes, there is a change in the homeostatic regulation, whereby even in the successful resolution of a depressive episode there is no return to the previous homeostatic state. This is the essence of the concept of neuroprogression in many psychiatric conditions, including MDD [159], and requires investigation as to pathophysiological alterations occurring in ‘treatment-resistant’ depression.

Although alterations in the mitochondrial function of many cells, including immune, glial, platelets and intestinal epithelial cells, as well as different neurons may occur in MDD, it is proposed here that astrocytes may be of particular relevance and act as an important hub onto which many of the above developmental, genetic, environmental and social processes may ultimately act.

For decades astrocytes were regarded primarily as providers of antioxidants and energy substrates for neurons. This has gradually changed over recent decades with astrocytes now recognized as powerful determinants of neuronal function and survival as well as of neurotransmitter release [160]. An astrocyte’s processes make contact with many synapses across different neurons, whilst also being part of an astrocyte network that communicates via a number of factors, including Ca2+ and ATP transfer across connexin (Cx)-43 gap junctions. Cx-43 also exist as hemichannels, being a route for astrocyte fluxes. Astrocyte end-feet are an integral part of the blood–brain barrier (BBB) in association primarily with vascular endothelial cells and pericytes. Astrocytes therefore provide a powerful interface for neuronal activity, coordinated blood flow regulation and systemic factors crossing the BBB. The astrocyte-neuronal interface would seem to be an important hub in all neuropsychiatric disorders. However, the specificity of astrocyte-neuronal interactions in MDD would seem to be determined by particular interactions across brain regions, such as amygdala and cortex, over the course of development, and the influence of the gut microbiome on such interactions.

Astrocytes have a number of immune-like qualities, being reactive cells with some antigen presenting capacity [161]. Maintained activation of a reactive state in astrocytes leads to the retraction of their processes from synapses and their isolation from the astrocytic network, with a maintained reactivated astrocytic state dysregulating wider coordinated patterned neuronal activity. Astrocytes release a wide array of factors that can be protective or damaging to neurons [97]. Astrocytes may not only regulate but also drive neuronal activity, including via the release of glutamate, which may be targeted to presynaptic and postsynaptic glutamatergic receptors. Astrocytes can release glutamate via a number of mechanisms, including Cx-43 hemichannels, vesicles and via the cystine-glutamate antiporter (System Xc), which effluxes glutamate in exchange for cystine uptake in the course of glutathione (GSH) synthesis, suggesting that increased antioxidant synthesis demand in astrocytes is intimately linked to glutamate efflux and the regulation of neuronal function.

GSH is an important inhibitor of neutral sphingomyelinase (nSMase)2-induced ceramide and pro-inflammatory cytokine release from astrocytes, with powerful consequences for neuronal function and associated cognition, as well as neuronal survival, as shown in preclinical models [162]. The relevance of nSMase2-induced ceramide is highlighted by the clinical and preclinical data showing stress-induced MDD to be significantly determined by circulating levels of nSMase2-induced ceramide [96]. Although these authors indicate brain endothelial cells as the major cell to be impacted by blood-derived ceramide in MDD, such data suggests the potential importance of variations in astrocyte oxidant status and fluxes on neuronal activity and interarea neuronal patterning, as well as on endothelial function, including via the capacity of GSH to suppress nSMase2-induced ceramide.

Astrocytes can be activated by many factors, the most extensively investigated being LPS via TLR4 activation, leading to NF-kB and YY1 transcription factors induction, as in many other cell types [163,164,165]. NF-kB is the transcription factor most strongly associated with astrocyte reactivation. NF-kB induction typically leads to a transient astrocyte activation state, including the induction of beta-site amyloid precursor protein (APP) cleaving enzyme-1 (BACE1) and subsequent production of amyloid-β [166]. Although classically associated with Alzheimer’s disease, BACE1 and amyloid-β are increased in a wide range of inflammatory conditions, including glioblastoma, breast cancers, amyotrophic lateral sclerosis and Parkinson’s disease with Lewy Bodies [167,168], as well as possibly in later-life MDD [169]. Given that amyloid-β is an endogenous antimicrobial, it’s release following NF-kB and YY1 induction seems predominantly to be an attempt to dampen microbial signaling, including as arising from TLR4 receptor activation by LPS, and perhaps endogenous ligands such as HMGB1 and hsp70. As such, processes such as gut permeability-derived LPS/TLR4/NF-kB signaling in MDD are intimately linked to changes occurring in astrocytes during neurodegenerative conditions. Alterations in astrocyte reactive states in MDD are therefore linked to a diverse array of medical conditions, including the risk and exacerbation of neurodegenerative conditions, although this is highly likely to be confounded by MDD pathophysiological heterogeneity.

In 2007, Liu and colleagues showed astrocytes to produce melatonin in a non-circadian manner [170]. In microglia and macrophages, the NF-kB induction of a reactive M1-like phenotype is concurrently linked to the induction of the melatonergic pathway and subsequent melatonin releases, leading to autocrine effects that induce an M2-like pro-phagocytic phenotype [171,172]. This allows the initial activation of these immune cells to be quickly followed by the production, release and autocrine effects of melatonin, thereby time-limiting the damaging effects that would arise from prolonged activation of these reactive cells. As to whether the NF-kB and YY1 induction of a reactive state in astrocytes is likewise sequentially associated with the upregulation of the melatonergic pathway, as indicated by NF-kB [165] and YY1 [173] in other cell types, requires investigation, including as to the relevance of this in different brain regions. Exogenous melatonin clearly dampens inflammatory activity in astrocytes [173], as in other reactive cells, with effects that seem, at least in part, via the capacity of melatonin to upregulate the mitochondrial melatonergic pathway, thereby better optimizing mitochondrial function and decreasing mitochondrial oxidants [174], with associated impacts on ROS-dependent miRNAs and gene patterning. It should also be noted that astrocyte YY1 has a number of important physiological effects, including over the course of normal development in mice, with some differential effects on gene induction in different brain regions [175]. It requires investigation as to whether these regulatory effects of YY1 are dependent upon its capacity to induce the melatonergic pathway. This is parsimonious with the detrimental effects of YY1 in a number of diverse medical conditions, including cancers [176], where the melatonergic pathway is dysregulated [145].

Melatonin seems to be produced in all mitochondria-containing cells across the three kingdoms of life on earth [150], with the majority of melatonin being produced within mitochondria [177,178]. This is likely to be of significance in MDD pathophysiology.

In the absence of the direct availability of serotonin from its release by serotonergic neurons and uptake into glia, the activation of the melatonergic pathway requires tryptophan uptake via the large amino acid transporter (LAT)-1 (SLC7A5), as well as SLC7A7 and SLC7A8. In astrocytes, as in many other cell types, the conversion of tryptophan to serotonin requires the induction of tryptophan hydroxylase 2 (TPH2), with TPH2 requiring the presence of the 14-3-3ε (YWHAE) isoform in order to be stabilized in an enzymatically active form [179]. The serotonin produced can then be converted by 14-3-3ζ (YWHAZ)-stabilized AANAT to NAS, which is then converted by acetylserotonin methyltransferase (ASMT) to melatonin [155]. For 14-3-3ζ-stabilized AANAT to initiate this pathway requires the presence of acetyl-CoA as a necessary co-substrate. The exclusivity of these 14-3-3 isoforms in the regulation of the melatonergic pathway in glia requires investigation. An array of diverse factors and processes may interact with these components of the melatonergic pathway, leading to a prolonged astrocyte activation state, dysregulated synaptic activity and neuronal oxidant challenge, as well as impacting on BBB homeostasis and the mitochondrial function of astrocytes, leading to changes in ROS-dependent miRNAs and gene patterning, with consequences for alterations in neuronal function, neurotransmitter release and thereby changes in inter-area communication.

YYI is also an important negative regulator of the excitatory amino acid transporter (EAAT)2, and therefore of glutamate uptake at the synaptic cleft. It requires investigation as to how a suppressed capacity of YY1 to upregulate the mitochondrial melatonergic pathway impacts on such neurotransmitter regulation. Alterations in glutamate regulation and the excitatory/inhibitory balance are an area of extensive research in MDD [180] and many other classical CNS conditions, such as autism, dementia, multiple sclerosis, ALS and schizophrenia [181,182,183,184]. YY1 is highly regulated by HDAC effects at the promotor of many YY1-induced genes, indicating that the loss of gut microbiome-derived butyrate’s HDACi capacity will have significant consequence for astrocyte YY1 regulation of the excitatory/inhibitory balance, which may be further dysregulated by suppression of the melatonergic pathway. It is still unknown as to whether the ten-fold decrease in pineal melatonin release at night between the ages of 18 years and 80 years are replicated in other cell types over the course of aging. Given the consequences that this can have on mitochondrial function and intercellular communication, as highlighted above, this should be a priority area of research in astrocytes, as well as other cell types.

Numerous studies have linked MDD, especially treatment-resistant MDD, to an array of autoimmune disorders [185], with MDD also showing evidence of autoimmune-linked processes [186]. Investigation of brain tissue indicates that genetic susceptibility to MDD is associated with a number of genes located in the major histocompatibility complex (MHC) locus of chromosome 6, providing a ready link of MDD to autoimmunity [187,188]. Neuroticism, a trait-like aspect of MDD susceptibility, is similarly associated with MHC-I [189]. MHC-I is increased by oxidative stress in brain cells, including astrocytes, neurons and oligodendrocytes, thereby increasing autoimmune activation, which can include the chemoattraction of CD8+ t cells [190]. The suppression of the mitochondrial membrane-located PTEN-induced kinase 1 (PINK1), not only attenuates mitophagy of dysfunctional mitochondria, but also increases oxidative stress and MHC-1, indicating that alterations in mitochondrial function may be an important mediator of the association of MDD and autoimmunity. Suppressed PINK1 levels are evident in MDD and MDD-associated pathophysiological changes, being an aspect of suppressed neurogenesis [191]. Importantly, exogenous melatonin promotes PINK1 accumulation on the mitochondrial membrane, as shown in neurons [192], indicating that melatonin not only upregulates the beneficial effects of PINK1 regarding mitophagy and protection against oxidative stress, but would also prevent the consequences of suppressed PINK1, including MHC-1 upregulation and associated induction of autoimmune-linked processes [193]. Astrocyte MHC-I is upregulated by pro-inflammatory cytokines and therefore may be intimately linked to astrocyte reactivity [194]. As astrocyte reactivity is strongly integrated with NF-kB and YY1 upregulation, this would indicate that the suppressed capacity of these transcription factors to upregulate the astrocyte mitochondrial melatonergic pathway, will contribute to mitochondrial dysfunction, nuclear factor erythroid 2-related factor 2 (Nrf2)/GSH suppression, neuronal dysregulation, prolonged astrocyte reactivity and MHC-I linked autoimmune processes in both neurons and astrocytes, arising from suppressed paracrine and autocrine astrocyte melatonin. This may be important, as different cytokines act on astrocytes to determine their morphology and function, including the phenotypes of different immune cells chemoattracted into the CNS [195]. As depression is associated with a host of neurodegenerative and autoimmune conditions, processes acting to regulate the glia mitochondrial melatonergic pathway and its consequences for glia-neuronal interactions, form the basis of MDD comorbidities.

The emerging importance of astrocytes in determining coordinated CNS function and the interface with wider systemic processes via the BBB, would indicate that core processes in astrocyte function are likely to be an integral aspect of understanding brain function. This article highlights how many of the diverse pathophysiological changes in many cells in MDD patients can involve alterations the regulation of the mitochondrial melatonergic pathway, including in CNS and systemic cells, with an array of diverse consequences. However, it is likely that astrocytes, as a core hub that integrates diverse CNS processes, coupled to their capacity to regulate neuronal activity and survival, as well as neurotransmitter release, oxidant and inflammatory challenges, will be an important treatment target for MDD and an array of wider medical conditions to which MDD is pathophysiologically associated.