Glucocorticoids (GCs) have been traditionally considered as pivotal components of the stress response. They constitute the end-product of the hypothalamic-pituitary-adrenal (HPA) axis, a complex neuroendocrine system; they are synthesized by the adrenal cortex and secreted into the systemic circulation, reaching multiple targets within the mammalian body to exert their molecular effects The HPA axis comprises several key nodes that interact reciprocally, including cortical and subcortical brain regions such as the prefrontal cortex, amygdala, and hippocampus. These areas encode stressful cues and regulate the secretion of corticotropin-releasing hormone (CRH) by the hypothalamus, which in turn upregulates the secretion of corticotrophin (ACTH) by the anterior pituitary. Subsequently, ACTH stimulates the biosynthesis and secretion of GCs by the adrenal cortex. GCs feed back to the anterior pituitary and hypothalamus to inhibit the secretion of CRH and ACTH, creating a negative auto-feedback loop (1). High GC levels during a stress response facilitate the proper functioning of cells and tissues within a framework of increased metabolic and cognitive demands, in various cases, also requiring the mobilization of the immune system. Of equal importance, the reduction of GC levels at the post-stress period is crucial for facilitating the reestablishment of homeostasis and the successful, long-term management of the allostatic load (i.e., effective neurobehavioral, immunological and metabolic adaptation). Chronic stress may dysregulate the HPA axis, thus disrupting the fine temporal balance between appropriately high and appropriately low GC levels. This loss of synchronicity between when GC levels should be high/low (based on the bodily demands) and when they actually are, could have a role in the development of stress-related disorders, including post-traumatic stress disorder and Major Depressive Disorder (MDD) (2, 3).

But in the realms of neurobiology and neuroendocrinology, a nuanced understanding of the role of GCs in the brain physiology reveals a complex interplay that goes far beyond their well-characterized role in the stress response. GCs, mainly cortisol in humans and corticosterone in rodents, are steroid hormones that orchestrate a wide array of processes; from interactions at the genomic and molecular level, all the way to the regulation of behavioral responses (4, 5). Within the complexity of neural circuits and cellular interactions, these hormones wield a significant influence.

One of the key aspects that merits attention, is the circadian rhythmicity of GC secretion. Under basal conditions, GCs follow a dynamic circadian pattern with high GC levels right before and during the active phase and low levels at the inactive phase of the day (6). The brain’s circadian clock, centered in the suprachiasmatic nucleus (SCN) of the hypothalamus, orchestrates the diurnal release of GCs (7). This circadian variation in GC levels is vital for optimizing various neural functions, including memory consolidation, mood regulation, and cognitive performance (8–10). Dysregulation of this rhythm, mainly due to chronic stress, has been linked to cognitive impairments and mood disorders (11).

Going deeper into the complex GC biorhythmicity, recent sources of evidence illustrate that -under baseline conditions- GCs are secreted in a pulsatile manner, every hour or few hours (depending on the mammalian species), as a result of the positive feedforward, negative feedback interplay between the anterior pituitary and adrenal glands, which is not dependent on the hypothalamic input. This GC pulsatility also seems to have a neurobiological significance; absence of pulsatility (or changes in the characteristics of pulsatility, like the frequency or duration of pulses) impact for example the temporal pattern of the expression of GC-sensitive genes and synaptic plasticity in rodents (11), as well as the neural processing, cognitive and behavioral responses in man (11).

To further underscore their complex role in the central nervous system (CNS), GCs have also been implicated in the mediation of both anti-inflammatory and pro-inflammatory effects. Generally, GCs are well-known for their potent anti-inflammatory properties, a facet of their role that is highly exploited pharmaceutically, especially for the management of autoimmune disorders (12). In the context of the nervous system, acting mainly via the glucocorticoid (GR) and mineralocorticoid receptor (MR), GCs demonstrate this opposing effect on the progression of neuroinflammation, in a distinct temporal manner. More specifically, when GC administration is acute, short-termed and in high doses, the development of an anti-inflammatory environment is favored, mainly characterized by reduced microgliosis and astrogliosis, repression of pro-inflammatory cytokine expression (including TNF-α, IL-1α and IL-1β) and neuroprotection (13–15). On the other hand, chronic GC administration or hypersecretion, induces microgliosis, primes microglial activation and promotes pro-inflammatory cytokine overexpression and hippocampal accumulation (16–18).

Interestingly, this temporal distinction in duration and subsequent downstream regulation of pro- or anti-inflammatory action by GCs, perfectly correlates with the “double-edged sword” nature of neuroinflammation itself within the CNS (19); while it plays a protective role in response to injury or infection, chronic or excessive inflammation can lead to neuronal damage and contribute to neurodegenerative disorders.

Brain-derived neurotrophic factor (BDNF) is a potent neuroprotective factor both during nervous system development and in adulthood, promoting neuronal survival, growth and synaptic plasticity, thus playing a crucial role in shaping the brain’s structure and function throughout life. In neurodegenerative conditions, where neurons are progressively lost, deficits in BDNF expression levels and signaling are often observed (20, 21). This deficiency can exacerbate the neurodegenerative process, compromising the brain’s ability to repair and maintain neural circuits. In Alzheimer’s Disease (AD), for instance, reduced BDNF levels are associated with the accumulation of beta-amyloid plaques and tau tangles (22, 23), while in Parkinson’s disease, BDNF stimulation can promote the dopaminergic neuron survival and differentiation in the substantia nigra (24).

BDNF, besides influencing neuronal growth and survival, also exhibits profound effects on mood regulation and emotional well-being (25). Stress is a well-known trigger for regulating BDNF levels in the brain, with the latter varying depending on the duration and the type of the stressor as well as the brain region. Chronic stress can lead to epigenetic alterations of the Bdnf gene, actively promoting its methylation and suppressing its expression, especially in the prefrontal cortex and the hippocampus (26). Furthermore, prolonged exposure to stress in adolescent rats induced an anxiety-like behavior that was directly correlated to their reduced BDNF protein levels (27). However, the relationship between acute stress and BDNF is more complex. Acute physical stress in rodents, induced either by physical exercise or foot shock, positively correlated with the induction of BDNF expression, especially in the hippocampus (28–30), a response that can facilitate synaptic plasticity and promote the formation of new neural connections, which are crucial for learning and memory processes (i.e., adaptation). On the other hand, acute psychological or restraint stress inhibits the expression of BDNF in many brain regions, including the hippocampus and the prefrontal cortex (28, 31).

The aforementioned role of BDNF in many psychiatric and neurodegenerative diseases, is tightly correlated to its reciprocal interaction with neuroinflammation (32). BDNF can activate astrocytes and microglia, promoting proinflammatory responses, thus exacerbating neuroinflammation, a process observed in various neurological disorders (33). On the other hand, it’s worth noting that neuroinflammation itself can enhance BDNF signaling and expression in microglia, creating a positive feedback loop (34). Furthermore, activated microglia, can trigger the release of BDNF, adding another layer of complexity to the interplay between this neurotrophin and inflammatory processes in the brain (35). These intricate interactions underscore the significance of understanding how BDNF and neuroinflammation jointly influence the pathophysiology of numerous brain-related conditions.

Given the role that GCs and BDNF seem to possess in stress, brain physiology and stress-related neuropsychiatric disease, and the close association between the expression of GR, BDNF and TrkB across brain regions (36), a reasonable question arises, whether the molecular biology of these two regulatory molecules interact, and if yes, at what level, under which conditions and impacting which cells/cellular networks/dynamics?

Perhaps the most extensively investigated aspect of the effects stemming from the interaction between GCs and BDNF in neurons, pertains to their influence on neuronal plasticity and memory consolidation. The initial evidence for this interplay was provided by Zhou et al., who elucidated corticosterone’s inhibitory impact on the generation of long-term potentiation (LTP) in rat hippocampal slices, an effect that could be counteracted by BDNF administration (62). Presently, it has become evident that the well-defined positive modulatory role of GC-BDNF interaction in several facets of memory formation and learning, necessitates the maintenance of a delicate equilibrium between GC administration and GR activation on one side and BDNF expression and TrkB activation on the other. Prolonged elevation of GC levels and/or compromised neurotrophic support, lead to the development of deficits in spatial memory, contextual fear memory, or fear extinction memory (10, 63, 64).

Mechanistically, recent sources of evidence have unveiled several pathways and molecular processes implemented in GC-BDNF signaling integration that promote synaptic plasticity and memory consolidation. In an in vitro setting, the excessive administration of a synthetic GC, dexamethasone (Dex) (a potent GR activator), to immature neurons, exerted inhibitory effects on BDNF-mediated synaptic maturation through two distinct mechanisms: transcriptional repression of synaptic proteins and constriction of their dendritic outgrowth (65). In a study conducted in rats investigating the role of hippocampal GR in memory retention, researchers identified a GR-mediated, non-genomic activation of the CaMKIIα-BDNF-CREB pathway, which was found to be indispensable for rescuing rats from amnesia, induced by GR inhibition (66). Further elucidating on this mechanism, Arango-Lievano et al. demonstrated that neurotrophin priming of GR, preserves GR phosphorylation at BDNF-sensitive sites, guiding GR’s transcriptional machinery and promoting neuronal plasticity (67). Subsequent pioneering work by the same research team, showed that both chronic stress and BDNF deficiency significantly inhibit this BDNF-dependent GR phosphorylation, resulting in impairments in memory retention and post-training dendritic spine preservation (68), increased Tau phosphorylation and development of Tau neuropathology (69), as well as exacerbation of the neuropathological manifestations of the APP/PS1 mouse model of AD (70). Interestingly, in the latter study, using postmortem human prefrontal cortex samples from AD patients, the team was able to positively correlate BDNF-dependent GR phosphorylation with increased cognitive performance, in contrast to the cortisol-dependent GR phosphorylation which was associated with cognitive deficiencies. These results reveal that imbalance between BDNF-mediated and GC-mediated GR phosphorylation in favor of GCs, can contribute to the development and progression of AD dementia.

Beyond its role in synaptic plasticity, an excessive imbalance of GC-BDNF interactions, can elicit significantly more deleterious effects on neurons, potentially even threatening their survival. The neuroprotective importance of GC-BDNF interactions has been extensively researched. In vitro studies conducted on primary hippocampal neurons and mouse hippocampal cell lines have delineated the neuroprotective potential of BDNF/TrkB/CREB pathway activation in mitigating GC-induced neurotoxicity (71, 72). Additionally, compelling evidence by Zhang et al., experimenting in rats following traumatic brain injury, revealed the contrasting outcomes associated with GR and MR activation (73). High GC levels, inducing GR activation, were found to promote neuronal death in the PVN, while low GC levels, promoting MR activation, supported neuronal survival. These findings underscore the paramount importance of maintaining a balanced GR/MR expression and activation ratio, reflective of a healthy circadian and pulsatile secretory pattern of GCs, as a critical factor for neuroprotection in the face of neurotoxic challenges.

Recently, an intriguing facet of the GC-BDNF interactions has come to light, offering promising prospects for the development of novel therapeutic strategies targeting stress-related disorders. Following translation, BDNF protein within neurons is encapsulated in secretory vesicles and subsequently transported anterogradely to dendrites (74). Adachi et al. characterized the ability of GCs to modulate this transportation process through the upregulation of Huntingtin protein via GR activation (75). In cortical neuronal cultures subjected to a single Dex treatment, HTT-mediated transportation of BDNF-containing vesicles to dendrites was enhanced. Conversely, Agasse et al., while investigating the mechanisms underlying the reduction of adult hippocampal neurogenesis in the context of chronic stress, observed a contrasting effect on GC-mediated BDNF vesicle transportation (76). Specifically, the administration of a very high dose of Dex, mirroring GC circulating levels during chronic stress, within a cortico-hippocampal network configuration, impeded BDNF vesicle transportation. This effect was directly correlated with hyperphosphorylation of HTT. Furthermore, in vivo experiments conducted in mice by the same research team replicated the GC-induced HTT phosphorylation and demonstrated a significant suppression of hippocampal neurogenesis as a consequence. The contradictory results observed in these studies are likely attributable to the significant difference in the Dexamethasone dose concentration used in each experiment, highlighting the dose-dependent nature of GC’s influence on BDNF.

Physiological secretion of GCs has been associated with the regulation of neuronal autophagy, acting as a mediator of neuroprotection and initiating the stress-coping response. However, prolonged or excessive exposure to GCs can disrupt the equilibrium of neuronal autophagy, potentially contributing to neuronal dysfunction and neurodegenerative processes (77). In a recently published study conducted by Zhang et al., the research team elucidated how chronic corticosterone exposure induced hyperactivation of the autophagic machinery in neuronal cells within the DG region of the brain (78). This led to an augmented lysosomal degradation of BDNF, which, in conjunction with decreased neuronal proliferation, migration, and survival, ultimately culminated in the development of depressive-like behavior in mice. Additionally, it is noteworthy that GCs have also been implicated in the regulation of the recently characterized pathway known as secretory autophagy, a non-degradative pathway for the extracellular secretion of cytosolic cargo. Specifically, GC-mediated stress has the capability to induce the secretion of matrix metalloproteinase 9 (MMP9) through secretory autophagy, consequently promoting the maturation of BDNF in the extracellular milieu (79).

Altogether, a potential neuronal interplay between the GC/GR and the BDNF/TrkB systems emerges. The previously reviewed studies unveil both the ability of GCs to interfere and alter (mainly through GR) TrkB phosphorylation and BDNF expression and trafficking (37, 45, 48, 49, 75, 76), as well as the ability of BDNF (mainly through TrkB signaling) to prime GR, enabling the maintenance of its phosphorylation even after BDNF is no longer available (67, 68), thus creating a positive feedback loop with two arms, that can promote neuronal plasticity.

The intricate role of stress-induced neuroinflammation in the development and progression of mood disorders has been the subject of profound scientific inquiry, leading to the accumulation of compelling research evidence (80). From the increased levels of proinflammatory cytokines (81, 82), such as interleukin-6 (IL-6) and tumor necrosis factor-alpha (TNF-α), measured in individuals diagnosed with depression, to the PET scan visualization of neuroinflammation in individuals experiencing depressive episodes (83–85). But the connection between stress and neuroinflammation is not unidirectional; neuroinflammation itself can sensitize the brain to future stressors, perpetuating this vicious cycle, in many ways. Increased levels of the pro-inflammatory cytokine IL-6 can increase the activity of brain regions implicated in the etiology of depression and impair brain connectivity causing mood reduction (86). Furthermore, neuroinflammation can disrupt the balance of neurotransmitters, such as serotonin, in mice, leading to increased vulnerability to stress and mood disorders (87). Another way to impede coping with and recovering from the effects of a future stressor, is through suppression of neuroplasticity, a trait that is well-correlated to chronic inflammation (88).

Glial cells, specifically microglia and astrocytes, play pivotal roles as primary mediators of the aforementioned effects of stress-induced neuroinflammation. Both microgliosis and astrogliosis represent common responses to various stressors, observed in both animal models and individuals with MDD (85, 89). The activation of these glial cells significantly amplifies the release of pro-inflammatory cytokines, intensifying neuroinflammation by disrupting the blood-brain barrier and facilitating the entry of peripheral immune cells and molecules into the brain (90, 91). Concurrently, activated glial cells contribute to neurodegeneration and hinder synaptic plasticity. Clinical and postmortem studies have well-documented the association between astrocytic dysfunction and depressive-like phenotypes (92). Moreover, morphological and functional atrophy of astroglial cells has been shown to induce depressive-like phenotypes in animal models.

Although it becomes evident that potential effects in non-neuronal cells could be in existence, with crucial role in maintaining brain homeostasis and responding to neuroinflammatory challenges, the predominant focus of the existing literature on GC-BDNF interactions, as previously demonstrated, relates to their effects on neurons. This trend in the field is inconsistent with the impressively high GR/MR ratio in glial cells compared to neurons (93–96), with single-cell RNA-seq data extending this observation in gene expression to known GR and MR target genes (97). As a result, non-neuronal cell populations are significantly more sensitive to both normal (i.e. at the peaks of ultradian GC secretion) and abnormal GC alterations (e.g. under stress). Similar cell type specific expression profiles have been characterized in the BDNF/TrkB system as well, with BDNF basal expression levels significantly elevated in pyramidal neurons compared to TrkB, while inhibitory neuronal and glial populations mainly expressing the TrkB receptor (98). Such a variety in the proportionate expression between the components of the two systems, suggests potential differences in the glucocorticoid response not just between neuronal and non-neuronal populations but even within the different glial cell types.

Several recent studies have characterized a co-occurrence of repressed BDNF expression and altered levels of microglia and astrocytes in the hippocampus of rats exposed to chronic stress, accompanied by increased secretion of pro-inflammatory cytokines (99–101). However, these studies have not showcased any evidence of a direct correlation between BDNF and GCs. In another work by Chen et al., adrenalectomized rats exposed to chronic mild stress exhibited astrocytic dysfunction in prefrontal cortex. This impairment was associated with reduced BDNF levels in the hippocampus and correlated with anxiety-like behavior displayed by the animals (102). Nevertheless, neither this study delved into investigating the potential linkage between the impaired GC-BDNF interaction and astrocyte dysfunction.

Recent work in our laboratory attempted to decipher the acute effects of GC administration on BDNF expression in primary astrocytic cultures (103). Astrocytes were able to instantly respond to a physiological GC pulse in their environment in a biphasic manner, in which BDNF expression was upregulated within the first hour, followed by a GR-dependent, downregulated BDNF expression for the rest of the 24-hour cycle. Additionally, in an interesting demonstration of GC-mediated homeostatic capabilities, astrocytes exposed to a second identical GC pulse within the same circadian cycle, were able to normalize their BDNF expression levels. These findings highlight the neuroprotective importance of maintaining a normal ultradian rhythmicity and the potential devastating effects that stress-related deregulation of it, can have in the astrocyte-mediated trophic support of the surrounding neurons. Furthermore, our results demonstrate astrocyte’s ability to alter their responses to increased GC secretion based on whether this increase is transient or consistent, an effect that can be attributed to their increased, GR-mediated GC sensitivity.

Overall, the aforementioned evidence of stress-induced neuroinflammation mediated through glial cells, the co-occurrence of irregularities in glial pathology and BDNF repression in chronically stressed mice and the existence of an astrocytic, GC-induced, biphasic response in BDNF expression and TrkB activation (103), unveils a new dimension in the interplay between the two systems as effectors of glial physiology. Such glial effects could subsequently deteriorate or even mediate the detrimental effects described in neuronal populations ( Figure 1 ). Combining this with the emerging role of non-neuronal cells in neuronal support, plasticity and synapse formation, arises the need to further investigate and fully unravel the GC-BDNF interaction in these cell populations. The results of these studies could pave the way for new therapeutic approaches targeting stress-related neurodegenerative disorders.