Any perceived psychological or physical stress, including an injury such as TBI, initiates a physiological response in the body to maintain homeostasis. The hypothalamic-pituitary-adrenal (HPA) axis is a major element of the stress response. Stressful stimulus causes activation of the hypothalamic paraventricular nucleus (PVN). Cells in the PVN secrete corticotropin releasing hormone (CRH), which stimulates the anterior pituitary to produce adrenocorticotropin hormone (ACTH). ACTH is secreted and signals to the adrenal glands, resulting in corticosteroid synthesis (Tapp et al., 2019; Leistner and Menke, 2020). Corticosteroids are steroid hormones that act throughout the body to regulate homeostasis in response to stress, contributing to fundamental biological processes in all major tissue types (HPA axis has been recently reviewed in detail by Leistner and Menke, 2020).

GCs are a major class of corticosteroids released through HPA axis signaling. They are essential stress-response hormones, playing key roles in a variety of physiological processes including metabolism, immunity, and cognition (Fietta et al., 2009; Cain and Cidlowski, 2017). Cortisol is the main glucocorticoid in humans, while corticosterone is the glucocorticoid in most other animals, including rodents (Katsu and Baker, 2021). Primary injury stimulates immediate activation of the stress response through stimulation of the HPA axis, and clinical studies show elevated serum cortisol in the first few days following TBI (Figure 1A; Woolf, 1992; Wagner et al., 2011).

The severe tissue damage sustained at the site of primary injury induces a rapid inflammatory response, with significantly increased cytokine production and microglia activation detected in experimental models 1 day after TBI (Tobin et al., 2014; Witcher et al., 2021). Cellular and vascular damage often leads to breakdown of the blood brain barrier (BBB) in the first several days following TBI (reviewed by Cash and Theus, 2020). This BBB disruption allows peripheral immune cells to enter the injured brain, synthesizing with the already inflamed resident cells in the brain to result in additional cytokine production and heightened inflammatory state (Sabet et al., 2021). Neuroinflammation underlies much of the secondary damage sustained after TBI. In addition to long-term outcomes such as cognitive and behavioral consequences, chronic neuroinflammation can also result in severe clinical symptoms such as intracranial pressure (ICP) and edema (Figure 1B). Importantly, cerebral edema not only causes severe pain and discomfort for patients but is also one of the major factors in survival after TBI (reviewed by Zusman et al., 2020).

The human glucocorticoid receptor gene, Nr3c1, is alternatively spliced into several variants. The two most well-studied isoforms are GRα and GRβ. GRα is the more widely expressed ligand-binding product, regulating several key homeostatic processes (Meduri and Chrousos, 2020). Expression of GRβ is more limited, but it is a component of several intracellular complexes, including a heterodimer with ligand-bound GRα (De Castro et al., 1996). GRβ also plays a key role in repressing some of the transcriptional effects of GRα (Ramos-Ramírez and Tliba, 2021). Interestingly, in vitro experiments suggest that GRβ plays a role in astrocyte-mediated wound healing, but the specific role of GRβ in the context of TBI has not been explored (Yin et al., 2013; Wang et al., 2015). Rodents also produce different isoforms of GR through alternative splicing of Nr3c1, though the precise mechanisms of splicing differ from those in humans. The more limited rodent GRβ isoform was not identified until 2010, but recent evidence suggests that, like its human homolog, it is important for negative regulation of many of the effects of GRα (Hinds et al., 2010; Dubois et al., 2013). In both humans and rodents GRα is the classic form of GR that responds to glucocorticoids (Nicolaides et al., 2010). However, because the discovery of GRβ in rodents is relatively recent, there is rarely a distinction between the α and β isoforms in preclinical literature. Therefore, for the purposes of this review, “GR” will collectively refer to either isoform produced by the Nr3c1 gene. Still, it is important to note that this is an oversimplification. Though GRα is the more prevalent isoform, it is possible that GRβ plays a role in the response to TBI, perhaps through astrocyte-mediated mechanisms similar to those described in the context of wound healing (Yin et al., 2013; Wang et al., 2015). Future studies should be careful to make the distinction between GRα and GRβ and consider that they may have very different injury-related roles.

There are two distinct categories of GR actions: fast (non-genomic) mechanisms and delayed (genomic) mechanisms (Figure 2, right panel). Non-genomic, rapid effects of GR work through interactions with membrane-associated G protein-coupled receptors (GPCRs; Tasker et al., 2006). Many of these rapid effects rely on receptor-dependent kinase activity (Oakley and Cidlowski, 2013). Non-genomic mechanisms have a broad range of functional outcomes throughout multiple tissues, from contraction of smooth muscle in the trachea to pancreatic insulin release (Se Sutter-Dub, 2002; Sun et al., 2006). In the brain, GR rapidly affects neuronal synapse plasticity and transmission, with detectable changes in neuronal activity occurring in just a few minutes of GR activation (summarized in Myers et al., 2014). Synthetic glucocorticoids such as dexamethasone are frequently used in experimental studies to activate GR and are instrumental in elucidating mechanisms of GR activity. Previous experimental studies show that a single injection of synthetic GC in rats causes a rapid increase (within 7.5–15 min) in the locomotor response to a novel environment (Sandi, 1996). Patch clamp studies in mice suggest dexamethasone induces rapid GR-mediated effects on both excitatory and inhibitory synapses (Nahar et al., 2015).

GR is primarily studied in the context of its delayed transcriptional effects, though these are complex and multifaceted. Upon ligand binding, cytoplasmic GR translocates into the nucleus, where it regulates transcription of hundreds of anti- and pro-inflammatory genes. GR can both activate and repress transcription of its downstream targets, either through direct binding of glucocorticoid response elements (GREs) in DNA or by tethering to other regulatory DNA-binding proteins. In the nucleus, GR can form a heterodimer with MR, or it can homodimerize and bind to other GREs to affect stress-related gene expression (Figure 1, right panel). GR may act as an anti- or pro-inflammatory regulator under different conditions, but it is notorious its powerful anti-inflammatory immunosuppressive effects in response to stress. Many of these anti-inflammatory transcriptional effects occur through tethering of GR to pro-inflammatory factors, such as Nuclear Factor κB (NFκB) or Activator Protein 1 (AP-1), inhibiting their transcription to promote a return to homeostasis (reviewed in Chinenov et al., 2013).

Importantly, of the limited experimental studies on the effects of TBI on Nr3c1/GR, the majority do not dissect the effects on specific downstream mechanisms of GR. Some studies quantify nuclear GR as a measure of transcriptional activity (Zhang et al., 2021). Others measure GC-dependent downstream targets, such as Sgk1 or FKBP5, as markers of transcriptional activation of GR (Aminyavari et al., 2019; Lengel et al., 2022). Still, the use of these methods to measure genomic GR activity in the context of TBI have been very limited to date. Furthermore, there is a stark lack of studies exploring the rapid, surface receptor-mediated mechanisms of GR after TBI, and these may be worth a closer look. For example, previous work has linked GR to N-methyl-D-aspartate receptor (NMDAR) activity in the hippocampus, which is known to contribute to TBI-induced excitotoxicity and cognitive deficits (Zhang et al., 2012; Baracaldo-Santamaría et al., 2022). NMDAR agonists have shown promise in improving cognition after TBI, but extensive issues regarding their safety and efficacy have limited their therapeutic potential and stalled clinical studies (Khormali et al., 2022; Hanson et al., 2024). A closer look at the role of GR in activity of NMDAR and other surface receptors may shed light on previously unexplored mechanisms of excitotoxicity after TBI.

Even after the initial recovery from TBI, many survivors experience lasting symptoms which can continue for several months or even years after injury. These include behavioral dysfunction such as depression or anxiety, and cognitive issues such as attention disorders and memory impairment (reviewed by Karr et al., 2014; Howlett et al., 2022). Given the robust effects of TBI on the physiological stress response it comes as no surprise that stress also profoundly impacts long-term recovery after TBI, but the relationship between injury and stress is complex. Persistent or intense stress is harmful and can exacerbate neurological dysfunction. However, the mechanisms through which stress affects post-injury recovery depend on the timing and type of stressor, and some studies suggest that stress can even improve outcomes after TBI (reviewed by Houle and Kokiko-Cochran, 2022; Zheng et al., 2022).

TBI often occurs in high stress environments, including domestic violence, combat, and sporting events (Brand et al., 2023). Under these conditions, GC levels are already elevated prior to injury, and MRs become saturated while GRs are overactivated (Fox et al., 2016). Multiple previous studies have indicated that pre-injury stress can significantly impact outcomes after TBI. Early life stress (ELS) paradigms are often used in preclinical studies to model physically and psychologically stressful life events. Maternal separation, in which pups are temporarily separated from their mothers for an extended period, is a common method of ELS, and it has been shown to exacerbate TBI-induced cortical atrophy and learning deficits in a rodent FPI model (Sanchez et al., 2021). Other experimental stress paradigms, such as chronic unpredictable stress (CUS), have also been shown to exacerbate subsequent injury. Mice exposed to 5 weeks of mild CUS followed by CCI exhibit exaggerated learning and memory deficits compared to mice receiving CCI in the absence of additional stress (Park et al., 2023). One proposed mechanism for this stress-induced aggravation of TBI pathology is that existing stress primes the immune system, leading to an exaggerated inflammatory response to a subsequent TBI (Brand et al., 2023). Indeed, ELS is correlated with higher TBI-induced microglial activation and production of inflammatory cytokines (Lajud et al., 2021; Sanchez et al., 2021). These studies indicate that pre-injury stress exacerbates cognitive and inflammatory consequences of TBI, though the mechanisms remain unclear. A likely explanation is that pre-existing stress leads to a dysfunctional HPA response to TBI. Indeed, clinical evidence has shown that patients exposed to pre-injury stress or disease have significantly lower cortisol levels following TBI compared to patients who did not report stress (Sörbo et al., 2020). This could be indicative of a stress-induced reduction in HPA axis functionality, but future studies will need to investigate the precise mechanisms through which pre-injury stress affects TBI outcome.

Interestingly, recent experimental evidence suggests that prior exposure to stress before TBI can also be neuroprotective. When adolescent rodents exposed to early-life CUS are given several weeks to recover from stress before receiving TBI in adulthood, they exhibit decreased behavioral deficits compared to rodents receiving TBI with no stress (de la Tremblaye et al., 2021). In general, ELS seems to exacerbate the damage and cognitive consequences sustained by TBI but may provide some protection if followed by adequate recovery time (de la Tremblaye et al., 2021; Lajud et al., 2021; Sanchez et al., 2021). The relationship between TBI and stress is complex, and it is important to point out that chronic stress during development could have a unique impact on the response to TBI compared to pre-injury stress during adulthood. More work is needed to elucidate the mechanisms through which pre-injury stress may be either detrimental or beneficial to long term outcomes.

After injury, TBI survivors are highly susceptible to multiple forms of secondary stress, such as chronic pain and insomnia. Importantly, many of these post-injury stressors can also be exacerbated by recovery-related environmental stress, such as sleep loss, isolation, and medical anxiety (Jain et al., 2014; Agtarap et al., 2021). When the HPA axis has recently been activated by TBI, subsequent activation by other stressors can result in an impaired stress response (reviewed by Komoltsev and Gulyaeva, 2022). This impaired response during the vulnerable post-injury period can result in loss of essential immediate anti-inflammatory actions of GC, exacerbating the chronic inflammatory state (Komoltsev and Gulyaeva, 2022). Additionally, survivors of TBI have an increased likelihood of developing post-traumatic stress disorder (PTSD), which can amplify the stress and trauma experienced post-TBI (Spadoni et al., 2018). Chronic variable stress (CVS) is frequently used to model PTSD in rodents, and CVS followed by TBI causes greater cognitive deficits than either CVS or TBI alone (Fesharaki-Zadeh et al., 2020). A recent review of clinical PTSD literature estimated that 13.5% of non-military mild TBI survivors develop PTSD, and in many cases these symptoms progressively worsen over time (Van Praag et al., 2019). TBI survivors also have increased probability of insomnia, and clinical evidence suggests that post-injury insomnia is correlated with HPA dysfunction in patients with mTBI (Zhou et al., 2016; Zhou and Greenwald, 2018). Experimental studies in mice demonstrate that 3 days of post-injury sleep disruption increases TBI-induced neuroinflammation (Tapp et al., 2020). When post-TBI sleep fragmentation is extended to 30 days, mice exhibited hippocampal dysfunction and deficits in memory acquisition (Tapp et al., 2022). Notably, this was also associated with increased cortical expression of stress- and inflammation-related gene and inhibition of upstream regulation by Nr3c1, suggesting that post-TBI sleep fragmentation stress suppresses anti-inflammatory actions of GR (Tapp et al., 2022).

Preclinical evidence from studies coupling multiple experimental TBI models with various post-injury stress paradigms similarly support that post-injury stress exacerbates both cognitive and physiological consequences of TBI. In a repetitive concussive TBI model in rats, post-injury foot shock stress resulted in significantly worsened depressive-like behavior than TBI alone (Klemenhagen et al., 2013). Several weeks of post-injury social isolation exacerbated cognitive outcomes and hippocampal apoptosis in a rat penetrating TBI model (Khodaie et al., 2015). A week of restraint stress after moderate TBI in a CCI mouse model caused increased endoplasmic reticulum (ER) stress and autophagy, resulting in increased neuronal loss (Gao et al., 2022).

These studies suggest that post-injury stress exacerbates TBI-induced dysfunction, inhibiting functional recovery. Together, this emphasizes the need for TBI therapeutics that take into consideration the effects of secondary stress, though much more work is needed to dissect the mechanisms through which stress affects long-term recovery after TBI.

Early clinical reports noted the efficacy of GCs in reducing ICP and edema, resulting in significant neurological improvements in patients with brain tumors (F'rench and Galicich, 1964). Soon, GCs were more widely used to relieve ICP in patients with other neurological conditions, including severe head trauma (Pickard and Czosnyka, 1993; Cook et al., 2020). Unfortunately, the ability of GCs to affect ICP in patients suffering TBI seems to be more limited. Despite promising preclinical results, multiple clinical studies have failed to demonstrate significant effects on functional recovery with glucocorticoid treatment after TBI (Braakman et al., 1983; Hoshide et al., 2016). Early clinical trials suggested that GCs increased survival after TBI, but the results of multiple randomized control trials suggest no significant effect of the drugs (Grumme et al., 1995; Olldashi et al., 2004; Edwards et al., 2005). In fact, the largest clinical trial, known as the corticosteroid randomization after significant head injury, or CRASH trial, reported increased mortality with corticosteroid treatment compared to controls (randomized control trials are summarized in Hoshide et al., 2016 and indicated with an asterisk in Table 1; Edwards et al., 2005; Hoshide et al., 2016). The inability of many of these studies to reach statistical significance leads to speculation that the timing and dose of GC administration may be key (Hoshide et al., 2016). Additionally, this raises the question of whether the methods of GC administration may be highly case-dependent, as severity and type of TBI vary from patient to patient. This would explain why preclinical studies, with model organisms in a controlled environment and consistent injury severity between individuals, generate significant results while clinical studies consistently fail to reach significance.

More recently, GCs have been studied in preclinical models for other post-injury injury applications, such as stabilization of the BBB and alleviation of neuroinflammatory edema (Hue et al., 2015; Moll et al., 2020). A clinical case study from 2021 examined nine patients who were given steroids to manage delayed cerebral edema after mild TBI (Prasad, 2021). All patients exhibited improvement of symptoms over a 2 year period, suggesting steroids can be safely used at lower doses in some cases. However, outcomes of steroid use after TBI remain inconsistent, and some studies suggest that high doses aggravate injury and impair cognition (Chen et al., 2009, 2011). Additionally, long-term use of GCs is associated with many negative side effects, including diabetes and osteoporosis (Marzbani and Bhimaraj, 2022). Many of these side effects are due to systemic immunosuppression that occurs over time with GC use, and systemic infection has been reported in clinical studies even at lower GC doses (Grumme et al., 1995). Interestingly, early GC use after TBI is also associated with increased risk of post-traumatic epilepsy, which may be due to the association of GR with calcium channels in the hippocampus (Karst et al., 1999; Watson et al., 2004). This emphasizes the need to better understand both the non-genomic and genomic mechanisms of GR after TBI and highlights the many issues with previous GC therapeutics.

With the mixed success and potential risks associated with corticosteroid treatment, experimental studies have explored other avenues of manipulating endogenous corticosteroid signaling. Furthermore, excessive GC concentrations have been shown to be neurotoxic, resulting in increased neuronal sensitivity to injury through overactivation of GR (Sapolsky, 1985; Mccullers et al., 2002a). As such, GR has been the subject of many recent studies, in a variety of different stress and injury contexts, with the goal of manipulating GR functionality without triggering these negative consequences. The GR antagonist mifepristone (RU486) is frequently used in rodent models of TBI and has shown some promise in improving post-injury outcomes. Pretreatment with mifepristone was shown to protect against hippocampal neuronal loss in a rat model of CCI (Mccullers et al., 2002a). Hippocampal damage has previously been implicated in development of psychiatric symptoms after TBI, and more recent work has explored the role of GR in post-injury cognitive and psychological outcomes (Meyer et al., 2012). Indeed, mifepristone pretreatment has been shown to prevent anxiety-like symptoms after mild TBI in rats (Fox et al., 2016). Mifepristone also counteracts some of the negative effects of treatment with the corticosteroid dexamethasone after TBI (Zhang et al., 2020). These studies provide evidence that neurotoxic effects of high GC levels act through GR and suggest that blocking GR signaling could improve recovery. However, the progesterone receptor-mediated effects of mifepristone on pregnancy, as well as additional adverse side effects such as liver toxicity, limit its translational potential and create a need for alternative approaches to GR manipulation (Meyer et al., 1990; reviewed by Nieman, 2022).

Transgenic rodent models are powerful tools for studying the functions of conserved genes. However, Nr3c1 is an essential gene in mammals, so there are no viable GR-null rodent strains (Cole et al., 1995). As an alternative approach, antisense RNA was used to create transgenic mice with decreased expression of GR, resulting in significant HPA axis dysfunction and cognitive impairment (Pepin et al., 1992; Montkowski et al., 1995; Barden et al., 1997). A major caveat of this model is that it only creates a partial loss of GR function, limiting its applications for molecular studies of GR signaling. Additionally, the global and constitutive knockdown of GR makes it difficult to distinguish developmental effects of GR loss from acute GR functions in adult mice. The use of this antisense RNA model has declined, and it is rarely used today.

More recently, tissue-specific functions of GR have been investigated using Cre-loxp recombination. In this system, Cre recombinase is expressed under a tissue-specific promoter, and a gene of interest is flanked by loxP sites. In cells expressing Cre recombinase, the loxP flanked (“floxed”) gene is excised, generating a tissue-specific deletion. This Cre-loxp system has been used to delete GR in the brain and peripheral cells, revealing novel insight into the tissue-specific role of GR (Tronche et al., 1999; Arnett et al., 2011; Liu et al., 2021). We will focus on the function of GR in the brain.

GR is abundantly expressed throughout the brain, crucial for maintaining homeostasis and facilitating recovery in response to stress (De Kloet et al., 2005; Joëls, 2018). Previous studies also suggest that GR in the brain plays a key role in learning and memory (Oitzl, 1997; Oitzl et al., 2001; Steckler et al., 2001). The role of GR in memory-related tasks has especially been demonstrated in the context of experimental studies associated with stress. For example, the Morris water maze task is commonly used to assess spatial memory (Vorhees and Williams, 2006). The water maze task has been shown to evoke a stress response in mice, and this significantly affects learning and memory when glucocorticoid signaling is impaired (Aguilar-Valles et al., 2005; Lengel et al., 2022). Cre-loxp deletion of GR in mouse neurons and glial cells leads to dysfunctional HPA axis feedback regulation and significantly elevated HPA activity, similar to symptoms of Cushing syndrome in humans (Tronche et al., 1999). This highlights that nervous system expression of GR is required for normal stress response functionality.

GR is highly enriched in structures of the limbic system, which regulates multiple processes to maintain homeostasis such as emotion, learning, memory, and motivation (Torrico and Abdijadid, 2023). This includes the hippocampus, amygdala, and hypothalamus (Erdmann et al., 2008). One of the first localized GR knockout mouse models was a forebrain-specific GR knockout, which disrupted GR in multiple essential limbic system structures (Boyle et al., 2004, 2006). Importantly, this model resulted in delayed loss of GR completing around 4–6th months, allowing for exclusion of developmental effects of GR deletion. This deletion resulted in disruption of circadian HPA activity, as well as an increase in depressive-like behavior (Boyle et al., 2004). An amygdala-specific deletion of GR was later generated by injecting lentiviral vector expressing Cre recombinase directly into the central amygdala in GR loxP mice (Kolber et al., 2008). These mice exhibited significant deficits in contextual and auditory-cued fear conditioning, suggesting a role for amygdala GR in learning and memory (Kolber et al., 2008; Arnett et al., 2011).

Within the hippocampus, GR is highly expressed in the CA1, CA2, and dentate gyrus, with lower expression in the CA3 region (Fuxe et al., 1985; Sarabdjitsingh et al., 2009). One of the major functions of GR in the hippocampus involves its association with membrane receptors, such as voltage-dependent calcium channels and hyperpolarization-activated cyclic nucleotide-gated channels (Kerr et al., 1992; Kim et al., 2022). Activated GR modulates currents through these membrane-channels, a key component of regulating neuronal excitability and maintaining homeostasis under stress (Kerr et al., 1992; Zü and Reiser, 2011; Lyman et al., 2021; Kim et al., 2022). Region specificity is seen in GR activity even within the hippocampus. Previous studies on GR-dependent calcium channel activity show differential effects of corticosterone on calcium currents in the CA1 and dentate gyrus (Chameau et al., 2007; Van Gemert et al., 2009). Interestingly, transcriptional analysis showed no difference in GR-associated transcriptional control of calcium channels (Van Gemert et al., 2009). This highlights the complex region specificity of GR signaling, and demonstrates the importance of considering both genomic and non-genomic effects of GR.

The pathophysiology of TBI is highly varied. The location and severity of the initial impact, as well as the degree of diffuse vs. focal injury, can result in differing levels of damage to multiple different brain regions (McGinn and Povlishock, 2016). Clinical and experimental evidence shows that the hippocampus and other limbic system tissues are particularly vulnerable to structural and inflammatory damage following TBI, though the precise region-specific consequences vary with injury type and severity (Christensen et al., 2020; Drieu et al., 2022). The effects of TBI on calcium channel-related activity of GR in the hippocampus in particular are worth investigating, as this could relate to mechanisms of post-TBI excitotoxicity (Chameau et al., 2007). Still, it would be interesting to study how the role of GR differs in other brain regions affected by TBI, such as the cortex. Ostensibly, GR-mediated effects of TBI are heavily dependent on the brain regions affected by injury, though future studies will need to more closely explore these region-dependent effects.

In addition to region- and cell-specific GR functions, the GR-mediated response to stress is also heavily influenced by the type and severity of the stressor. This is important because the primary and secondary phases of injury, as well as the extended period of chronic recovery, exert very different physiological and stress-related effects. To date, few studies have directly analyzed GR/Nr3c1 response to TBI. Many of the studies that report effects of TBI on stress signaling through GR and MR make use of receptor antagonists like mifepristone to assess receptor function. The results of a 2002 preclinical CCI study in rats indicated inhibition of GR mRNA 24 h after injury in the hippocampus and dentate gyrus (Mccullers et al., 2002b). Pretreatment with mifepristone prior to injury did not result in increased GR mRNA, suggesting that TBI impaired negative feedback regulation of GR (Mccullers et al., 2002b). A 2022 study involving FPI in mice reported that sleep fragmentation for 30 days after injury led to inhibition of Nr3c1 regulation (Tapp et al., 2022). An experimental study from last year reported a decrease in hippocampal Nr3c1, quantified using qPCR, in male rats more than 80 days after FPI (Ju et al., 2023). These recent studies point to chronic inhibition of GR/Nr3c1 after TBI. The earlier results from Mccullers et al. suggest impaired negative feedback after TBI, which resulted in decreased GR mRNA 24 h after injury. Still, more work is needed to dissect the mechanisms of GR that contribute to altered stress signaling at both acute and chronic time points. Future studies that distinguish between the different isoforms of GR, as well as the genomic and non-genomic effects, will help shed light on the role of GR in HPA dysfunction after TBI. GR has been studied more extensively in the contexts of other stressors, and there are distinct responses to acute and chronic stress (Li et al., 2019). The response of GR in these other stress contexts could inform future studies on the effects of primary and secondary brain injury on GR signaling.

Preclinical studies have shown GR activation in response to acute psychological stress, where acute stress is defined as a rapid intense exposure to stress. Restraint stress has long been used as a model of psychological stress, and studies suggest distinct GR-mediated effects that vary by timing and duration of restraint. Within the hippocampus, a single 20-min exposure to acute restraint in mice induces a small but significant transient decrease in total mRNA expression of Nr3c1, the gene encoding GR (Viudez-Martínez et al., 2018). Another study showed increased phosphorylation of GR following 1 h of acute restraint, which may affect downstream activity of GR (Papadopoulou et al., 2015). A similar acute restraint paradigm in rats showed a decrease in cytosolic GR but increase in nuclear GR in hippocampus following restraint stress (Green et al., 2016). GR translocates to the nucleus upon activation, and these studies together suggest a decrease in total expression but increase in activated GR following acute restraint. Forced swim is another psychological stress paradigm in rodents, and a single 15-min forced swim trial results in significant upregulation of several GR target genes in the hippocampus, indicative of increased GR activity (Mifsud and Reul, 2016). Together, these studies indicate increased activation of GR following acute stress.

The inflammatory stimulus lipopolysaccharide (LPS) has been widely used to induce acute immune challenge and neuroinflammatory response. Microglia are highly sensitive to stress-induced changes in the brain environment, and studies using microglial GR-depleted mice suggest that GR in microglia is neuroprotective following LPS stimulus, preventing LPS-induced neurodegeneration and repressing expression of Kv1.3 calcium channels (Carrillo-De Sauvage et al., 2013).

Chronic exposure to stress, where the stressor endures for an extended period or occurs repeatedly over multiple days, induces changes in HPA axis functionality, altering the role of GR as stress persists. Preclinical and clinical evidence indicates that prolonged stress or injury can cause HPA axis dysfunction, resulting in functional differences in GR signaling. Endocrine disorders such as Cushing's disease and diabetes mellitus can directly affect GR expression and function (Mu et al., 1998; Panagiotou et al., 2021). Other conditions like cardiovascular disease and obesity are associated with chronic disruption of HPA feedback regulation and disrupted GR sensitivity (Ljung et al., 2002). Preclinical models of chronic stress show distinct differences between effects of acute and chronic stress on HPA function and GR signaling. Chronic restraint stress, for example, has been shown to cause downregulation (Chiba et al., 2012; Viudez-Martínez et al., 2018). Chronic unpredictable stress, in which rodents are exposed to various factors such as forced swimming, physical restraint, and food or water deprivation, induces progressive downregulation of hippocampal Nr3c1 (Li et al., 2020). This effect is more pronounced over time, with Nr3c1 expression decreasing over several weeks as mice are continually exposed to stress (Li et al., 2020).

When mice receive an additional LPS challenge following chronic stress, inhibition of GR with the receptor antagonist RU486 significantly prevented LPS-induced neuronal damage. This suggests that in the context of chronic stress, GR is a key mediator of inflammation and neurodegeneration. This is in stark contrast to the neuroprotective role of GR that has been observed in response to acute LPS exposure alone (Espinosa-Oliva et al., 2011; Carrillo-De Sauvage et al., 2013). Taken together, this highlights that even within the brain, the downstream activity of GR signaling is highly context dependent.

As GR is ubiquitously expressed and has highly context- and tissue-specific roles throughout the body, global manipulation with a receptor antagonist such as mifepristone has variable effects in different cell types throughout the body. This could complicate interpreting results from studies like these and make it difficult to generate clear results on cognitive and behavioral outcomes. Genetic approaches for tissue- or cell-specific GR manipulation will likely a better option going forward.

Using GR-expressing lentivirus, a recent preclinical study found that GR overexpression in the dorsal hippocampus improved cognitive outcomes after pediatric TBI (in a CCI model) in rats (Lengel et al., 2022). Additionally, it was found that TBI caused decreased mRNA expression of the GR target Sgk1 compared to controls (Lengel et al., 2022). This suggests that TBI-induced impairments in hippocampal GR functionality may underlie post-injury cognitive impairment. Interestingly, this seems to contradict what has previously been reported with mifepristone treatment, where blocking GR was beneficial to TBI outcome. It could be that pediatric and adult TBI have differential effects on GR signaling. With the availability of Cre-loxp rodent models, and advancements in gene therapy techniques such as lentiviruses, it is worth looking at tissue- and developmental stage-specific effects of GR on TBI outcomes.

As the resident immune cells of the nervous system, microglia play a fundamental role in sensing and responding to changes in the brain environment to maintain homeostasis. Microglia undergo morphological and functional changes when exposed to stress and are an essential component in the brain's response to injury. However, prolonged activation of microglia perpetuates neuroinflammation and contributes to neurodegeneration. Clinical and preclinical evidence suggest chronic post-injury changes in microglia, including the development of primed microglia, after TBI (Fenn et al., 2014; Krukowski et al., 2021; Witcher et al., 2021; reviewed by Wangler and Godbout, 2023). Primed microglia are hypersensitive to secondary stress or immune challenge, mounting an exaggerated inflammatory response associated with cognitive deficits and psychiatric dysfunction (Muccigrosso et al., 2016). Depletion of injury-associated microglia after TBI reduces neurodegeneration and neurological deficits, highlighting the important role of microglia in mediating post-injury outcomes (Witcher et al., 2018; Henry et al., 2020; Bray et al., 2022).

GR is highly expressed on microglia and is a key link between the stress response in the brain and the microglia-mediated inflammatory response (Sierra et al., 2008). Given the instrumental role of microglia in chronic neuroinflammation, microglia-related GR signaling has been a key subject of study in recent preclinical models of stress and injury.

Healthy aging is associated with heightened neuroinflammation and exaggerated response to stress or immune challenge, and experimental evidence has shown that this is mediated by primed microglia (reviewed by Norden and Godbout, 2013). When rats are administered intracranial mifepristone, microglial pro-inflammatory responses are significantly diminished. Furthermore, the immune response to E. coli infection was reduced, and infection-associated memory deficits were rescued (Barrientos et al., 2015). These findings suggest that GR facilitates microglial priming in the context of aging. In another study, GR antagonism with mifepristone prevented microglia-mediated neuronal remodeling and behavioral despair following chronic unpredictable stress (CUS) in mice (Horchar and Wohleb, 2019). This suggests that GR facilitates phagocytic behavior and neuronal interactions of microglia after CUS and supports the notion that GR may mediate microglia activation.

Genetic deletion of GR in myeloid cells, including microglia, improves recovery in a mouse model of spinal cord injury (Madalena et al., 2022). This was correlated with impaired microglia and macrophage infiltration of the injury site, as well as reduced activation of these cells (Madalena et al., 2022). It is worth pointing out that these results are seemingly contradictory to the improvements in cognition that were reported with hippocampal overexpression of GR in rat TBI (Lengel et al., 2022). This emphasizes the highly context dependent role of GR. The type of injury or stress, as well as the cell and tissue type, all play a role in determining the downstream outcomes of GR signaling. Furthermore, Madalena and colleagues pointed out that depletion of GR from myeloid cells could result in compensatory mechanisms in GR signaling from other cell types (Madalena et al., 2022). Future studies are needed to determine these mechanisms, and it would be interesting to see if the role of GR is similar in the context of other forms of CNS trauma.

A 2022 study uncovered a role for GR in modulating outcomes in a model of rodent food restriction (FR) prior to TBI (Perović et al., 2022). Pre-injury FR was previously shown to increase levels of circulating GC and suppress TBI-induced microglial activation (Lončarević-Vasiljković et al., 2009; Loncarevic-Vasiljkovic et al., 2012). In this newer study, it was found that FR enhances nuclear localization of GR and induces upregulation of downstream GR targets (Perović et al., 2022). This suggests that the neuroprotective effects of pre-TBI FR may act through increased GR signaling. Further analysis is needed to determine the cell-specific role of GR in this model, and to confirm whether this increased GR affects microglia. Nevertheless, these recent studies together suggest GR may be an important link between the stress response and microglia-mediated pathophysiology in multiple contexts.