Differential impact of stress and environmental enrichment on corticolimbic circuits

Abstract

Stress exposure can produce profound change in physiology and behavior that can impair health and well-being. Of note, stress exposure is linked to anxiety disorders and depression in humans. The widespread impact of these disorders warrants investigation into treatments to mitigate the harmful effects of stress. Pharmacological treatments fail to help many with these disorders, so recent work has focused on non-pharmacological alternatives. One of the most promising of these alternatives is environmental enrichment (EE). In rodents, EE includes social, physical, and cognitive stimulation for the animal, in the form of larger cages, running wheels, and toys. EE successfully reduces the maladaptive effects of various stressors, both as treatment and prophylaxis. While we know that EE can have beneficial effects under stress conditions, the morphological and molecular mechanisms underlying these behavioral effects are still not well understood. EE is known to alter neurogenesis, dendrite development, and expression of neurotrophic growth factors, effects that vary by type of enrichment, age, and sex. To add to this complexity, EE has differential effects in different brain regions. Understanding how EE exerts its protective effects on morphological and molecular levels could hold the key to developing more targeted pharmacological treatments. In this review, we summarize the literature on the morphological and molecular consequences of EE and stress in key emotional regulatory pathways in the brain, the hippocampus, prefrontal cortex, and amygdala. The similarities and differences amongst these regions provide some insight into stress-EE interaction that may be exploited in future efforts toward prevention of, and intervention in, stress-related diseases.

1. Introduction to the Interaction of Environmental Enrichment (EE) and Stress

1.1. EE and Stress: Overview

Environmental enrichment (EE) was first described by Hebb in 1947. Hebb demonstrated that allowing rats to move freely in his house, rather than standard cage housing, improved their performance on cognitive tasks¹. Since then, numerous studies have investigated the effects of EE, both behaviorally and mechanistically. While EE protocols can vary greatly between labs, EE typically consists of improvements to the physical, cognitive, and/or social components of housing^2–6. Enriched cage environments usually house multiple rodents together in the presence of toys and enclosures. For example, our lab uses 1 m³ wire mesh cages to house 10 Sprague-Dawley rats (Figure 1). Rats can interact with each other, climb on the wire mesh, and access a variety of toys and shelters that are rotated weekly^7,8. Animals housed in enriched conditions such as these are often protected from the negative physiological or psychological impact of physical and emotional challenges/stressors, as compared with standard housed controls^2–4. The beneficial effects of EE extend to humans, in that humans who engage in socially, cognitively, and physically stimulating activities often show improvements in mood and cognition^9–11. These improvements are often similar to those achieved with medication but without side effects, so EE presents a unique non-pharmacological opportunity to treat a variety of disorders. Consequently, EE has considerable promise as a means of increasing stress resilience.

Figure 1: Example of EE housing.

Images of the EE apparatus used in our lab. EE cages are 1m³, have wire mesh walls for climbing, contain toys that are rotated weekly, and house 10 rats. This design provides physical, cognitive, and social stimulation, which are all traditional elements of EE. (A) Empty cage. (B) Cage with rats.

It is important to note that for the purposes of this review, we define EE as housing in a complex environment with opportunities for social interaction. This definition does not include paradigms focused purely on exercise, e.g. running wheels alone. EE is not a unilateral manipulation, but involves addition of multiple, diverse elements to stimulate the animal across different domains. Animals may engage in more physical activity as a result of EE housing via play or climbing, but exercise is not the sole or even primary manipulation. We are also not considering standard home cage enrichment (social housing, nylon bones, crinkle paper, huts, etc.) in our definition of EE. While of clear potential benefit to individuals, cage enrichment alone does not afford the spectrum of complexity associated with EE, including exercise opportunities, choice of companion animals, choices of enclosures, and voluntary investigation of familiar and novel stimuli (toys). It is however worth considering that introduction of mandatory enrichment in accordance with regulatory guidelines may well have beneficial effects in and of themselves, which may impact the magnitude of enhanced activity imparted by EE (as well as potentially occlude any number of endpoints previously observed in single-housed, unenriched rats)^12–15. The experiments cited in this review largely abide by this definition of EE, and any exceptions will be noted.

It is similarly important to define what we refer to as ‘stress’, since this term can have different interpretations. Here, we consider stress to be a ‘real or perceived threat to homeostasis or well-being’¹⁶. Note we do not give the term a valence: e.g., vigorous exercise, enjoyed by many, causes engagement of physiological responses (HPA axis activation and autonomic drive) that are indistinguishable from responses to noxious stimuli. Here intense exercise is interpreted as a very real threat (requiring mobilization of physiological responses to mobilize energy resources), but one that can be managed through efficient counter-regulatory adaptations, likely by the brain.

1.2. Conceptual Understanding of EE and Stress Interactions

Multiple hypotheses have been proposed to explain how EE conveys beneficial effects on behavior, some based on the contention that EE acts as a mild stressor. While EE is typically considered a positive experience, over time it is sufficient to generate physiological changes consistent with mild stress, e.g., elevated corticosterone (CORT) secretion, body weight reduction, increased adrenal mass, etc. Such observations support the ‘cross-stressor adaptation hypothesis’, which proposes that exercise and stress act upon similar mechanisms in a way that enables exercise to better prepare an animal to cope with stressors and vice versa^17,18. Exercise is a component of EE, and may provide activation of physiological responses, to which added social and experiential stimuli may synergize. Cross-stressor adaptation is consistent with the well-known phenomenon of stress habituation, whereby repeated exposure to the same stressor decreases subsequent responses^19,20. Thus, it is possible that EE may work via a similar mechanism, serving to temper physiological processes associated with repeated mild stressor exposure.

This notion that EE acts as a mild stressor is further consistent with the ‘inoculation stress hypothesis’, which posits that the variety of morphological and molecular changes initiated by EE can improve stress resilience^3,21–24. Indeed, engagement of stress pathways by EE is suggested by functional connectivity studies in the brains of EE rats, whereby chronic connectivity among stress-processing brain regions is increased by EE, while connectivity between all regions is decreased^25,26. This shift in connectivity was associated with improved cognition, suggesting that EE rearranges circuitry in a way that improves information processing across the brain, strengthening important connections and weakening inefficient ones.

The hypothesis that EE mildly activates stress-related circuitry to improve overall brain function also aligns with the notion that stimulus intensity and behavioral effects typically obey an inverted U-shaped function, with low or high physiological responses driving inferior performance. For example, it is clear that increments in stress hormones (CORT, norepinephrine (NE), epinephrine) can improve performance of avoidance memory tasks, which are then impaired when hormone levels are driven higher. Thus, stress exposure can be adaptive and beneficial when appropriately controlled^27–31. It is thought that EE may enable sufficient stimulation to improve function over that of animals lacking stimuli without pushing the animal into high, maladaptive levels of stress (Figure 2)^20,32. In this way, EE is adaptive, building stress resilience and ameliorating negative behavioral consequences of stress^3,20,27.

Figure 2: The position of EE on the inverted-U shaped function of stress intensity.

Stress intensity and behavioral effects exist on a spectrum where too much or too little can impair function. However, there is an optimum amount of stress that is protective and conveys stress resilience. The positive simulation of EE likely falls in this beneficial part of the spectrum, suggesting that EE provides just enough mild stress to guard against other stressors.

Despite that fact that EE causes mild physiological responses, it is also true that it is perceived as rewarding. Rats will lever-press for access to a running wheel or social counterparts, and they demonstrate conditioned place preference for compartments that contain novel, rather than familiar, objects^33,34. EE also activates reward circuitry and dopamine (DA) production³⁵. These findings suggest that EE is intrinsically rewarding and positive manipulation, making it a potentially important intervention to build stress resilience³⁶. So while EE activates physiological domains of the stress response, it is generally perceived as rewarding, indicating that the context of the stress response matters.

Overall, EE appears to physically and psychologically improve animals’ capacities to cope with stress^3,11,28–3. These benefits act both prophylactically in preparation for future stressors, and as a means of minimizing the impact of previously experienced stressors. This concept of improved function in EE conditions also appears to hold true in humans with depression and anxiety disorders, as exercise, cognitive training, and social contact all assist in amelioration of mood symptoms^9–11,37. We will explore the mechanisms thought to play a role in this conceptual relationship between EE and stress in later sections.

1.3. Behavioral Effects of EE and Stress

On a behavioral level, EE confers anxiolytic, anti-depressive, and pro-cognitive effects^2–5. Animals exposed to EE generally habituate to novelty faster, thereby reducing their anxiety-like behavior and improving their ability to cope with challenges^10,38,39. This faster habituation is seen in open field (OF) tests, in which EE animals show increased center time and less locomotion. These behaviors suggest that the animals are less cautious about the novel environment and more exploratory^24,40,41. Decreased anxiety-like behavior is observed in the elevated plus maze (EPM), where open arm times are increased in animals exposed to EE^41–44. It seems likely that this increased habituation/decreased anxiety is a result of the more constant novelty animals experience in EE, as their environment (toys and shelters) changes frequently^4,45. Similar beneficial effects are also observed in the domains of mood and cognitive processing. Beneficial effects of EE on mood are manifested as decreased anhedonia in the sucrose preference test (SPT)^46–48, increased active coping in the forced swim test (FST)^46–50, and increased resilience to social defeat⁵¹. EE also improves learning and memory in passive avoidance tests⁵², spatial memory in the radial arm maze^53,54 and Morris water maze^{24,39,55–60}, and reversal learning in operant touchscreen tasks⁶¹.

These effects are observed under both baseline and stress conditions. In non-stressed animals, simple addition of EE helps improve overall function. When various stressors are applied, these benefits are magnified, with EE ameliorating or even reversing detrimental effects of stress. For example, EE can rescue anxiety-like behaviors in the OF and EPM and depression-like behaviors in the SPT and FST induced by restraint, chronic variable stress, and social defeat^51,62–64. EE rescues stress-induced increases in freezing during contextual and cued fear conditioning and ameliorates avoidance behaviors related to post-traumatic stress disorder (PTSD), suggesting beneficial effects on fear reactivity and emotional regulation^65–69. In the context of stress, EE can work both prophylactically and therapeutically. EE exposure prior to stress can lessen its maladaptive effects, whereas EE exposure after stress can improve recovery^2,3.

While EE conveys some level of behavioral benefit in most cases, these effects can vary by EE paradigm, as well as the age, strain, and sex of the animal. For example, longer EE exposures (several weeks) typically have a greater impact on more behaviors than acute exposure (hours to days)⁴. EE applied in adolescence also tends to have stronger effects than that initiated in adulthood⁶⁵. Certain strains of mice may exhibit more aggression in group housing than rats or other mouse strains, resulting in attenuation of the benefits of EE⁷⁰. Females and males may respond to certain components of EE differently, resulting in variable behaviors between sexes⁷¹. While these differences make it difficult to draw parallel conclusions across studies, they do not change the fact that EE is primarily beneficial and neuroprotective, regardless of specific technical details. These differential behavioral effects of EE across conditions have been reviewed elsewhere^2,3.

1.4. Physiological Effects of EE and Stress

A variety of physiological responses are affected by EE, including hormones and cytokines. CORT release, which is an important factor in stress-related phenotypes, can be altered by EE^3,19,72. Prior work suggests that EE increases basal CORT and, in some cases, increases adrenal mass, the latter finding reflecting long-term activation of the HPA axis^73–76. In contrast, EE can blunt CORT responses to novel stressors, either by lowering peak responses or quickening return to baseline^44,73,74,77. This suggests increased basal HPA axis drive in EE animals, along with smaller and more quickly resolved stress responses. Such improved stress processing would be in line with the resiliency phenotypes described above. It is important to note that chronic stress exposure typically causes sensitization rather than blunting of CORT responses^19,20,78. These data indicate that while chronic HPA axis drive may contribute to beneficial actions of EE, it is not sufficient to mimic the impact of chronic stress. It is also important to note that while EE can activate the HPA axis, this activation is not perceived as stressful since EE is largely a rewarding, rather than aversive, stimulus.

A causal role of HPA axis activation in effects of EE is called into question by the lack of generalizability of CORT and ACTH elevations between laboratories. While most groups report increased basal and decreased CORT release during stress, contradictory results have also been found. Most studies suggest that basal CORT is increased by EE^73–76, but others have found decreases⁷⁷ or no change^20,79. Similarly, most suggest that CORT responses to stress are decreased by EE^44,73,74,77, but others report increases⁷ or no change^62,75,80. Similar varied results are seen with adrenocorticotropic hormone (ACTH), both basally and in response to stress^60,81,82. While it is important to note that these studies differ in terms of rat strain, EE protocol, and sampling times, CORT and ACTH differences across studies suggest that beneficial actions of EE cannot be directly attributed to HPA axis properties alone. It should be noted that discrete stress hormone sampling may fail to capture the dynamic nature of HPA axis drive. Chronic changes in HPA axis drive can be assessed via increases in adrenal weight; however, studies measured this endpoint again found contradictory results. Some reported increased adrenal weight in response to EE^8,73,83, while others found no change^7,74,84. Thus, while EE experience likely causes episodic or cumulative increases in CORT secretion, the data do not unequivocally support HPA axis drive as a primary cause of the beneficial actions of EE on physiology and behavior.

Stress-buffering by EE extends to sympathoadrenal responses. EE can blunt epinephrine release⁷³ and stress-related increases in heart rate^85–89, again pointing to an improved regulation of stress responses in general.

EE also modulates the immune system in a way that may promote resilience. EE is typically anti-inflammatory under basal conditions, although it can also increase the efficiency of the immune response^4,41,80,86. Many of these immune studies focus specifically on exercise^10,45,90, but cognitive and social EE have shown similar effects^4,91,92. EE reduces production of circulating pro-inflammatory factors, such as interleukin 1 beta (IL-1β), tumor necrosis factor alpha (TNFα), and interferon gamma (IFNγ), and promotes production of circulating anti-inflammatory factors, such as interleukin 10 (IL-10) and interleukin 1 receptor antagonist (IL-1Ra)^4,80,93. While these effects may be secondary to exercise-induced increases in interleukin 6 (IL-6) release from muscles⁹⁴; toys and social EE can also lower plasma IL-1β and TNFα^4,91, making the source of these changes unclear. This shift in cytokine production promotes an anti-inflammatory state at baseline, preventing potentially deleterious effects of excessive inflammatory responses when they are not required to combat a viral or bacterial challenge, while enhancing the body’s ability to deal with these challenges. Beyond cytokines, EE stimulates leukocytosis, increasing the proliferation of macrophages, neutrophils, monocytes, and natural killer cells^41,45,95. Macrophage chemotaxis and phagocytosis are also increased by EE⁴⁵. These changes in cellular immunity better prepare the immune system to respond to and resolve challenges. As elevations in many of these pro-inflammatory cytokines have been implicated in stress and psychiatric disease^96,97, this anti-inflammatory profile with improved responsivity is expected to promote general health and stress resilience, thereby playing a role in the beneficial effects of EE.

2. EE Mechanisms in Stress-Processing Brain Regions

The central nervous system is the primary location for processing EE and stressors, as well as for generating the physiological effects and resiliency discussed above. EE promotes numerous actions in the brain that likely drive behavioral effects. Understanding these actions will be critical for understanding how EE is able to ameliorate maladaptive effects of stress. Therefore, this review focuses on the molecular and morphological effects of EE in various stress-processing brain regions in adult rodents, as well as how these changes may contribute to the interaction between EE and stress. Specifically, we will focus on the hippocampus, prefrontal cortex, and amygdala, three regions which are critical to processing stressful stimuli and are differentially impacted by EE¹⁹. The hippocampus and prefrontal cortex typically exert top-down control to terminate or alleviate responses to stressors, whereas the amygdala enhances stress responses. We will identify mechanisms consistently altered in these regions which present interesting targets for future resilience research, as well as unique mechanisms which could hold the key to understanding the complex effects of EE. Additionally, we will explore potentially negative effects of EE and important future directions for the field.

2.1. Morphological Effects and Synaptic Plasticity

The morphological effects of EE under baseline and stress conditions are summarized in Table 1 and Table 2, respectively.

Table 1:

Morphological Effects of EE under Baseline Conditions

Category	Hippocampus	Prefrontal Cortex	Amygdala	References
Neurogenesis	EE ↑ neurogenesis			5,38,101,102,105
Cell Survival	EE ↑ cell proliferation			43,114,115
	EE ↓ apoptosis, autophagy, oxidative stress			67,68,113,116
		EE ↓ oxidative stress		63,110,116
Dendritic Complexity	EE ↑ dendritic branching and spine density			111,119–121
		EE ↑ spine density in IL/PL		126
		EE ↑ or ↓ complexity in Cg3		124,125
		EE ↑ dendritic branching in OFC		124
			No effects	20,62
Electrophysiology	EE ↑ LTP			130–133
	EE ↑ cell excitability			100
Plasticity Markers	EE ↑ synaptophysin, PSD95, synapsin I			64,136,138–140
		EE ↑ synaptophysin and synapsin I		136,139,145
			EE ↑ synaptophysin	65
Other Cell Types	EE ↑ PV interneurons			153
	EE ↑ microglia and astrocyte #			154–156
	EE ↑ IBA1 and GFAP			99,156
	EE ↑ endothelial cell proliferation and angiogenesis			10,119,162
		EE ↑ endothelial cell proliferation		162

Table 2:

Morphological Effects of EE and Stress Interaction

Target	Hippocampus	Prefrontal Cortex	Amygdala	References
Neurogenesis	EE block stress-induced decreases in neurogenesis			65,97,98,107,206
Cell Survival	EE block stress-induced increases in cell death			56,102,112,113
Dendritic Complexity	EE block stress-induced decreases in dendritic complexity			64,123
			EE block chronic stress-induced dendritic branching	84,129
			EE cannot block acute stress-induced dendritic branching	62
Electrophysiology	EE block stress-induced decreases in LTP and LTD			134,135
Plasticity and Activity Markers	EE block stress-induced decreases in synaptophysin, PSD95, and synapsin I			64,138
	EE ↑ or ↓ Fos responses to stress			51,60,141,142
		EE ↑ or ↓ Fos responses to stress		51,60,142,146,147
			EE ↓ Fos responses to stress	60,141,142,148,149
Other Cell Types	EE block stress-induced loss of PV			67
		EE block stress-induced loss of PV		67
			EE block stress-induced loss of PV	40
	EE block stress-induced decreases in GFAP			63
			EE block stress-induced increases in GFAP	63

2.1.1. Cellular Proliferation and Survival

One of the most widely studied effects of EE is adult neurogenesis in the hippocampus^{5,56,98–100}. This effect was first recognized by Kempermann et al., 1997, using BrdU to identify newborn neurons in the dentate gyrus (DG) of mice exposed to EE¹⁰¹. Since then, many groups have also found this EE-induced increase in neurogenesis under baseline conditions both in mice and rats^{5,38,102–106}. Moreover, EE can block stress-induced decreases in hippocampal neurogenesis, thought to contribute to the negative consequences of stress on behavior^{59,65,97,98,107}. While most often associated with the exercise component of EE^10,108,109, cognitive and social EE can also increase neurogenesis^{66,104,110,111}. Indeed, the more complex forms of EE tend to induce more lasting effects than exercise alone^66,104,110. EE also promotes health and survival of existing neurons, as it reduces spontaneous and stress-induced cell death in the hippocampus^{56,102,112,113}. This protective effect of EE occurs in mice and rats under both baseline and stress conditions^43,114,115, possibly as a result of decreased apoptosis¹¹³, autophagy⁶⁸, and oxidative stress^67,116. Given the role of the hippocampus in inhibiting stress responses, improved cell survival conveyed by EE likely contributes to its ability to improve stress resilience.

Neurogenesis is a phenomenon largely restricted to the DG of the hippocampus¹⁰¹, and is not thought to occur in other stress-regulatory limbic circuits (e.g. prefrontal cortex or amygdala). Direct measures of cell survival in EE have not been measured in the prefrontal cortex or amygdala under basal conditions or following stress. Similar protection from oxidative stress has been described in the prefrontal cortex^63,110,116 and hippocampus, suggesting that benefits of EE on the hippocampus likely extend to cortical neurons as well. We will explore these effects further in later sections.

2.1.2. Dendritic Complexity

In addition to improving cellular health, EE increases the plasticity of hippocampal neurons, manifest as increased dendritic complexity (both branching and spine density)^2,4,117,118. Under baseline conditions, EE increases arborization and spinogenesis in both apical and basal dendrites^{111,119–121}. This effect is opposite to that seen with stress alone, which decreases the branching and length of apical and basal dendrites in the hippocampus^106,122. When applied together, EE is able to block these stress-induced decreases and ultimately increase overall dendritic branching^64,123. This increased connectivity between neurons likely allows for more rapid communication and ultimately more efficient information processing that likely help animals to cope with and recover from stressful situations²⁶.

Three studies examined the effects of EE on dendrites in the prefrontal cortex under baseline conditions. One found that complex housing increased basal dendrite branching in the orbitofrontal cortex (OFC) and decreased basilar branch length in the anterior cingulate cortex (Cg3)¹²⁴. Another found increased spine density without changes in branching in the Cg3¹²⁵. The third found that complex housing increased overall dendritic complexity and spine density in the infralimbic (IL) and prelimbic (PL) cortices¹²⁶. Overall, the data indicates that EE has different effects in individual prefrontal cortex subregions, biased toward increased complexity. Increased dendritic complexity in the prefrontal cortex is expected to oppose the decreased complexity caused by chronic stress¹²⁷, although this effect has yet to be directly investigated.

Interestingly, EE and stress-induced changes in dendritic complexity in the amygdala, specifically the basolateral amygdala (BLA), are in the opposite direction to those seen in the hippocampus and prefrontal cortex. The BLA is the focus of many of these studies, as it is particularly important to processing psychological stressors¹⁹. Chronic stress alone increases dendritic complexity of projection neurons in the BLA^{84,106,122,128}. This finding is in line with the pro-stress nature of the amygdala, namely that stress increases amygdalar activity and drives stress-related behavioral phenotypes¹⁹. EE alone does not appear to alter these endpoints¹²⁹. However, when EE and chronic stress are run concurrently in adulthood, EE is able to block stress-induced increases in BLA dendritic branching and spine density¹²⁹. This finding also occurs with adult EE after adolescent stress⁸⁴. However, EE failed to reverse increases in BLA dendritic arborization caused by acute stress⁶², suggesting that the protective effects of EE in the amygdala only emerge after chronic stress. The differences in these results from those in the hippocampus or prefrontal cortex make the amygdala particularly interesting with respect to the interaction of EE and stress.

2.1.3. Synaptic Plasticity and Activity Markers

These changes in neuronal number and complexity translate to functional differences in EE and stress. Electrophysiological properties of the hippocampus change in response to EE alone, which strengthens long term potentiation (LTP) in pyramidal cells of the DG^5,10,130 and CA1^{100,130–133}. EE alters both the pre- and post-synaptic excitability in these regions to ultimately improve hippocampal neuroplasticity, likely contributing to the improved learning and memory of EE-treated animals¹⁰⁰. EE also blocks stress-induced decreases in LTP and long term depression (LTD)^134,135. These functional consequences of EE and stress have been further demonstrated using molecular markers of synaptic plasticity and neuronal activation. EE increases synaptophysin expression throughout the hippocampus^64,136–139 and blocks stress-induced decreases in this plasticity markers^64,138. Exercise alone stimulates similar changes in post synaptic density protein 95 (PSD95) and synapsin I^137,140, suggesting that these changes may also exist in EE. Studies examining markers of neuronal activity in the hippocampus have yielded less consistent results. Some have found that EE increases the Fos response to stress^51,60, while others have found decreases in Fos and delta-FosB^141,142. These differences likely resulted from variations in stress and EE paradigms, or a species difference between mice and rats, although none of these factors appear to be the sole cause of differential responses. None of these studies conducted further colocalization analysis to identify the Fos-associated cell types, so it is difficult to draw conclusions about the functional implications of these results. However, studies utilizing cytochrome c oxidase (CCO) found that EE decreases the long-term metabolic capacity of the hippocampus, suggesting that EE improves the processing efficiency of hippocampal cells^25,143,144. This heightened activity likely further enhances the ability of this region to adapt to novel stimuli, including stress. Altogether, these EE and stress-induced changes to membrane properties, synaptic plasticity markers, and activity support the above morphological observations. EE improves synaptic plasticity in the hippocampus, both under baseline and stress conditions, putatively guarding against the maladaptive effects of stress.

While fewer studies examined these markers in the prefrontal cortex, this region appears to show similar changes in synaptic plasticity as the hippocampus. EE increases synaptophysin and synapsin I in the prefrontal cortex^136,139,145. Further exploration of electrophysiological properties, other plasticity markers, and stress interactions would be interesting for a more complete comparison. Again, the Fos results are conflicting in this region, with different studies finding that EE can increase^51,146,147 or decrease^60,142 Fos and delta-FosB responses to stress. CCO studies again demonstrate decreased metabolic capacity, and thus improved processing efficiency, in the prefrontal cortex^25,143,144. These similarities further suggest that EE exerts similar effects on the hippocampus and prefrontal cortex.

One study found that exercise increases synaptophysin in the amygdala, suggesting increased synaptic plasticity⁶⁵. However, more studies focusing on EE specifically are required to be confident that the amygdala response is the same as the hippocampus and prefrontal cortex in this case. In terms of neuronal activation, the amygdala Fos studies were relatively more consistent. EE often reduces stress-induced Fos responses in the amygdala^{60,141,142,148,149}, suggesting that EE blocks the hyperactivation of the amygdala during stress. However, other studies found that EE increased Fos responses to stress^147,150, so we cannot definitively make this conclusion. Overall, it appears EE dampens stress via opposite effects in the hippocampus (increased activity) and amygdala (decreased activity). Additionally, CCO studies suggest that EE exerts minimal effects on the metabolic activity of the amygdala under baseline conditions, providing another difference from the other regions^25,143,144. Further studies are warranted to determine if this endpoint changes under stress conditions.

2.1.4. Other Cell Types

EE also influences interneurons, glia, and vasculature. Stress can often disrupt the function of parvalbumin (PV) interneurons, impairing their inhibitory control over excitatory neurons^151,152. On the other hand, EE can increase PV immunoreactivity in the hippocampus under baseline conditions¹⁵³ and reverse stress-induced loss of PV protein levels⁶⁷. This effect is thought to be mediated by the ability of EE to ameliorate oxidative stress^67,116, and it likely improves the inhibitory tone of the hippocampus. Similar EE protection from stress-induced PV loss has been suggested in the prefrontal cortex⁶⁷ and amygdala⁴⁰.

EE can increase the number of microglia and astrocytes in the hippocampus under baseline conditions^154–156, but this has not been explored in the context of stress. This finding is further supported by increased levels of the microglia marker ionized calcium binding adaptor molecule 1 (IBA1) and the astrocyte marker glial fibrillary acidic protein (GFAP) in EE rats^99,156. Additionally, EE rescues stress-induced decreases in hippocampal GFAP, suggesting that EE’s protective effects extend to astrocytes⁶³. This improved support would further enhance the efficiency hippocampal signaling and is expected to contribute to EE-induced neurogenesis^92,157–160. Although some studies have noted increased gliogenesis in the neocortex¹⁶¹, direct measures of microglia and astrocytes have not been made in the prefrontal cortex. In the amygdala, EE blocked stress-induced increases of GFAP, suggesting that EE reduces stress-induced increases in amygdalar astrocytes⁶³. This amygdala effect is again opposite of that seen in the hippocampus.

EE can also increase the vasculature in the hippocampus, via enhanced endothelial cell proliferation and angiogenesis^10,119,162. This effect would increase delivery of nutrients and clearance of waste, adding to the improved efficiency of the hippocampus during EE. Increased endothelial cell proliferation was also noted in the prefrontal cortex¹⁶². While vasculature has not been directly studied during stress or in the amygdala, EE-induced alterations in the neurotrophin vascular endothelial growth factor (VEGF) suggest that this proposed vascular effect is likely plays a role in both^63,66,121. VEGF and other molecular mechanisms underlying all these morphological and plasticity effects will be explored in the next section.

2.2. Molecular Effects

The molecular effects of EE under baseline and stress conditions are summarized in Table 3 and Table 4, respectively.

Table 3:

Molecular Effects of EE under Baseline Conditions

Category	Hippocampus	Prefrontal Cortex	Amygdala	References
Neurotrophins	EE ↑ BDNF and TrkB			99,103,113,163,164,166,266
		EE ↑ BDNF		163,164
			No effects	65,66,129
	EE ↑ VEGF, NGF, NT-3, IGF-1, GDNF			66,79,113,115,121,163,165,170,172
			EE ↑ VEGF, NGF	66
Hormones	EE ↑ GR expression and translocation			63,170,173,185
	EE alters CRH receptor expression			188,189
Neurotransmitters	EE ↑ AMPA, NMDA, glutamate transporter			5,83,135,196–198
		No effects		203,206
		EE ↓ D1 and DAT		204,206
		EE alters 5-HT system		114,200,207
			EE ↓ NE	69
Immune Markers	EX ↑ CX3CR1			212
		EE ↓ TNFα		180
Intracellular Signaling	EE ↑ PKA and MAPK pathway activity (↑LTP)			131,215
	EE ↑ CREB activity (↑ BDNF)			58,113,165,174,176
		EX ↓ ROS, iNOS; ↑HSP70 (↑ antioxidants)		216–218
			No effects	65,129

Table 4:

Molecular Effects of EE and Stress Interaction

Target	Hippocampus	Prefrontal Cortex	Amygdala	References
Neurotrophins	EE block stress-induced decreases in BDNF			50,63,64,66,68,168,267
		EE block stress-induced decreases in BDNF		63,180
			EE block stress-induced increases in BDNF	63,129
	EE block stress-induced decreases in VEGF, NGF, NT-3, IGF-1, GDNF			50,63,171,172,66,79,113,115,121,163,165,170
		EE block stress-induced decreases in VEGF and IGF-1		63,181
			EE cannot block stress-induced increases in VEGF and NGF	63
Hormones	EE block stress-induced decreases in GR expression and translocation			63,186,187
		EE ↑ GR response to stress		53
		EE ↓ CORT response to stress		192
		EE ↑ IL control over HPA axis		51,193
			EE block stress-induced increases in GR expression and translocation	194
	EE ↑ NPY response to stress			184,190,191
			EE ↑ NPY response to stress	191
			EE block stress-induced increases in CRH receptors	191,195
Neurotransmitters	EE alters 5-HT and NE response to stress			69,199–201,227
			EE ↑ 5-HT response to stress	184
		EE blunt DA and ACh release during stress		192,205,206
Immune Markers	EE and EX blunt stress-induced increases IL-1β, TNFα, IL-6, IL-1Rα, CCL2, CCL3, CXCL12			4,93,99,208,209
		EE block stress-induced increases of IL-1β		180
Intracellular Signaling	EE block stress-induced decreases in CREB activity (↑ BDNF)			58,113,165,174,176
			EE block stress-induced activation of ERK-MAPK-CREB pathway (↓ GR)	194
			EE block stress-induced EGR-1 (↓ neuronal activation)	194
	EE blunt stress-induced increases in PKM pathway activity (↑ GR signals)			186
	EE block stress-induced decreases in SIRT1/miR-134 pathway (↑ VEGF)			64

2.2.1. Neurotrophins

Increased neurotrophic activity in the hippocampus of EE animals is proposed as a primary molecular mechanisms for stress-protective effects of EE^2–4,118. Hippocampal brain derived neurotrophic factor (BDNF), in particular, is increased by EE under basal conditions^{99,103,113,140,163–166}. EE prevents stress-induced decreases in BDNF, which are believed to contribute to the maladaptive behavioral effects of stress^{50,63,64,66,68,167,168}. The receptor for BDNF, tropomyosin receptor kinase B (TrkB), is also increased by EE and exercise^50,169. EE stimulates other neurotrophins in the hippocampus as well, including VEGF, nerve growth factor (NGF), neurotrophin-3 (NT-3), insulin-like growth factor-1 (IGF-1) and glial cell-derived neurotrophic factor (GDNF), under basal and stress conditions^{50,63,171–173,66,79,113,115,121,163,165,170}. Neurotrophins generally promote the survival and proliferation of neurons. BDNF and NGF specifically increase dendritic complexity^{163,174–176}, while VEGF increases angiogenesis and spinogenesis^{66,121,177,178}. Therefore, these molecules are expected to contribute to the positive morphological changes induced by EE in stress exposed animals noted above^103,179.

EE-induced alterations of BDNF levels in the prefrontal cortex are similar to those seen in the hippocampus. EE increases BDNF in the prefrontal cortex under basal conditions^163,164, and rescues stress-induced decreases in BDNF^63,180. Similar protective effects have been found with VEGF⁶³ and IGF-1¹⁸¹. However, another study found no change in VEGF⁶⁹ and NGF, NT-3, and GDNF have not been examined in this context. While further studies are needed to fully compare the prefrontal cortex to the hippocampus, it appears that these regions share similar neurotrophin responses to EE and stress.

Neurotrophins in the amygdala are also impacted by EE and stress, but this region is again different from the hippocampus and prefrontal cortex. EE does not impact baseline levels of BDNF in the amygdala^65,66,129, while stress alone increases BDNF here^182,183. When animals experience both, EE blocks stress-induced increases in BDNF^63,129. This directionality is opposite to that of the hippocampus where increased BDNF contributes to stress resilience, suggesting that increased amygdalar BDNF may instead promote, rather than ameliorate, stress. Despite this difference in direction, EE still counteracts this effect to improve resilience to chronic stress. However, one study specifically examining PTSD-related stress found that EE enhanced the stress-induced upregulation of BDNF¹⁸⁴, suggesting that this effect may be sensitive to specific stressors. The other neurotrophins are not well-studied in this region. One study found no change in VEGF⁶³, while another noted increases in VEGF and NGF in response to EE⁶⁶. These conflicting results add to the complexity of the interaction of EE and stress in the amygdala, and suggest that the amygdala is more sensitive to different types of stress than the hippocampus and prefrontal cortex.

2.2.2. Hormones

The HPA axis is a critical element of the stress response, and glucocorticoid receptors (GR) in the brain are key to regulation of this system. As discussed above, CORT, the ligand for GR, is altered by EE and stress, although the directionality of this change is inconsistent between studies. In the hippocampus however, EE consistently increases GR expression under baseline conditions, while stress decreases GR expression^{63,170,173,185}. Together, EE blocks stress-induced loss of GR^63,186,187. This protective effect extends to GR translocation¹⁸⁶, indicating that GR increases not only the number of GR receptors, but also their activation. This effect would increase the responsivity of the hippocampus to CORT stimulation, which would enhance feedback and ultimately dampen HPA responsiveness^19,186. This finding supports the notion of a blunted stress response in EE animals, which agrees with the majority of CORT studies discussed above and further describes the role of the HPA axis in EE-induced stress resilience. Along the same lines, toys can decrease FKBP5 expression in the hippocampus⁶⁶. FKBP5 negatively regulates GR, so a decrease in its expression would support this concept of increased GR activity in EE. Corticotropin releasing hormone (CRH), another element of the HPA axis, can also be altered by EE; although, different studies have found increases¹⁸⁸ or decreases¹⁸⁹ in this hormone and its receptors. The conflicting nature of these studies makes it difficult to draw conclusions about CRH in the hippocampus at this point in time. Another factor that may be important to EE is neuropeptide Y (NPY), a neuropeptide typically associated with PTSD. EE enhances NPY responses to stress and the expression of NPY-Y1 receptors. These receptors are thought to contribute to stress resilience^184,190,191, so NPY may be a relatively unrecognized effector of EE.

While these mechanisms have been less studied in the prefrontal cortex, it appears that this region does share some characteristics with the hippocampus. EE induced a trend toward increased expression of GR in the prefrontal cortex in response to restraint stress, although this effect was not statistically significant⁵³. Microdialysis examining free CORT in the prefrontal cortex showed that EE decreases the CORT response to restraint within this region¹⁹². Both of these results support the idea that EE decreases the responsivity of the HPA axis to stress: more GR suggests better feedback, which, in this region, would more quickly return the CORT response to baseline^19,53,192. Lesion studies suggest that EE enhances the control of the IL over the HPA axis, further supporting this concept¹⁹³. Beyond GR, EE increased CRH and CRHR2 in the prefrontal cortex; however, aggression between mice may have been confounding in this case¹⁸⁹. No studies have examined NPY in this region.

The amygdala again responds to EE and stress differently than the hippocampus or prefrontal cortex. Stress often increases GR translocation in the amygdala^19,183, which drives rather than dampens stress responsiveness. EE can block this stress-induced GR translocation¹⁹⁴, suggesting that EE still opposes stress effects in this region, even though the directionality of the effects are switched. While further studies are needed to replicate this finding, the continued differential effects between regions is intriguing. EE blocks stress-induced increases in CRHR1 in the amygdala^191,195, theoretically decreasing the excitatory output of the amygdala to the HPA axis and promoting resilience^19,183. For NPY, EE increased NPY-Y1 expression in the amygdala, which is unexpectedly in agreement with the findings in the hippocampus. The exact mechanisms involved in NPY’s role have yet to be elucidated but may reveal a rare functional similarity between the hippocampus and amygdala.

2.2.3. Neurotransmitters

EE also alters neurotransmission across the brain, under both basal and stress conditions. In the hippocampus, EE increases the expression of glutamate receptors, specifically α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)^{5,83,135,196,197} and N-methyl-D-aspartate (NMDA) receptors^197,198. EE decreases glutamate transporter expression, which would also increase glutamate neurotransmission via decreased astrocytic uptake of glutamate in the synaptic cleft¹⁹⁸. These changes are consistent with the increases in LTP found in EE. Unfortunately, these glutamate EE studies did not examine stress, so, while these EE effects are expected to improve cognition and resilience, further studies are needed. Hippocampal serotonin (5-HT) and norepinephrine (NE) are also altered in the context of EE and stress. Some found that EE decreased the responsiveness of serotonin to stress^69,199, while others found that this was increased²⁰⁰. Others noted increased expression of serotonin 1A (5-HT1A) receptors and enhanced serotonin turnover^69,201. While these effects are somewhat contradictory, they seem to suggest that EE improves the buffering capacity of the serotonergic system in the hippocampus, which could aid in stress coping⁶⁹. Norepinephrine studies conflict, showing increases²⁰⁰ and decreases⁶⁹ in response to stress, so it is difficult to draw conclusions about this system.

Neurotransmitter systems in EE and stress have received more attention in the prefrontal cortex. It appears that the glutamatergic system is less involved in the prefrontal effects of EE, with more changes being noted in the dopaminergic, cholinergic, and serotonergic systems. Glutamate and gamma aminobutyric acid (GABA) levels and receptors are unchanged by EE under both baseline and stress conditions^202,203. Several changes occur in the dopaminergic system when EE is applied in baseline conditions, including reduced function and expression of dopamine 1 (D1) receptors and decreased dopamine transporter (DAT) surface expression^204,205. EE animals also release less dopamine when they are stressed^192,206. A similar blunted stress response was observed with acetylcholine (ACh)^192,205, suggesting that EE reduces the responsivity of these two systems in the prefrontal cortex. EE effects on serotonin are more uncertain, with most studies showing that EE increases its activity and others showing decreases^114,200,207. Overall, the notion that EE modulates different neurotransmitter systems in the hippocampus and prefrontal cortex is interesting and demonstrates that unique effects exist in every region.

Less information is available regarding the amygdala. One study found that EE increases the serotonergic response to stress¹⁸⁴, and another found that it lowered norepinephrine under basal conditions only⁶⁹. Clearly, more studies will be needed to understand changes in neurotransmission in the amygdala in response to EE and stress.

2.2.4. Immune Markers

Another domain that EE is known to modulate is the immune system. As discussed above, EE has many effects on the peripheral immune system, which likely have reciprocal effects with the central nervous system^4,208. Within the hippocampus, EE further modulates various cytokines and chemokines^4,93,208. EE typically blunts stress-induced increases in immune factors such as IL-1β, TNFα, IL-6, IL-1Ra, C-C motif chemokine ligand 2 (CCL2), C-C motif chemokine ligand 3 (CCL3), and C-X-C motif chemokine 12 (CXCL12) here^{4,93,99,208,209}. One study found increases in these factors, but aggression between mice could have been confounding¹⁸⁹. It is important to note that these findings are associated with either exercise and stress^93,208 or EE and immune challenge^99,209. Together, EE would be expected to reduce stress-induced increases in inflammation-associated markers, which would likely be neuroprotective^210,211. However, more targeted studies are needed to clarify the role of the central immune system in EE and stress interactions specifically. For example, the above morphological results pointed to increased numbers microglia and astrocytes in the hippocampus of EE animals^154–156, yet here these glia appear to release fewer cytokines in response to stress. While a little counterintuitive, this mismatch does align with the decreased basal activity but increased responsiveness of the peripheral immune system noted above, suggesting that EE exerts similar immune effects throughout the body. Along the lines of improved efficiency, exercise can increase the expression of CX3C chemokine receptor 1 (CX3CR1), which would theoretically improve communication between neurons and microglia^211,212. There is likely a complex interplay between a shift in activity and expression of various cytokines that influence the overall activity of the hippocampus in EE and stress that we cannot yet dissect.

Immune-related factors have not been well-studied in the prefrontal cortex or amygdala. In the prefrontal cortex, EE can block stress-induced increases of IL-1β and reduce basal TNFα¹⁸⁰, suggesting that the prefrontal cortex is again behaving like the hippocampus. However, another study found no such changes in the prefrontal cortex¹⁸⁹. These factors have yet to be examined in the amygdala. Given the unclear results in the hippocampus and the growing literature that the immune system plays a key role in stress responses^211,213,214, further studies of EE, stress, and the central immune system would be valuable.

2.2.5. Intracellular Signaling

Going one level deeper, EE can modulate the activity of various intracellular signaling pathways and molecules in the hippocampus. EE increases the sensitivity of the protein kinase A (PKA)-cAMP pathway, increasing its ability to respond to and convey signals, ultimately enhancing hippocampal plasticity and LTP¹³¹. EE also stimulates extracellular signal-regulated kinase (ERK)/mitogen activated protein kinase (MAPK) signaling, specifically the Ras-GRF2-ERK-MAPK pathway in newborn neurons, which aids in their survival²¹⁵. Stress-induced increases in protein kinase M (PKM) activity are rescued by EE, which aids in synaptic remodeling and is believed to result from the increased expression of GR activity discussed earlier¹⁸⁶. EE-induced increases in AMPA receptors also stimulate the expression and activity of serum and glucocorticoid-regulated kinase 1 (SGK), a downstream kinase which mediates learning and memory¹⁹⁷. Downstream of BDNF, the signaling molecule cAMP response element-binding protein (CREB) mediates its pro-survival effects. CREB expression and phosphorylation are increased by EE, suggesting that BDNF stimulated by EE affects downstream targets^{58,113,165,174,176}. The expected pathways involved in VEGF have been even more thoroughly mapped. Here, EE is thought to downregulate miR-107, which upregulates hypoxia-inducible factor 1-alpha (HIF-1α), which then increases VEGF. VEGF then acts through its receptor Flk-1 to stimulate spinogenesis and convey resilience¹²¹. EE rescues stress-induced decreases in the histone deacetylase sirtuin 1 (SIRT1) and increases in miR-134, changes believed to improve synaptic plasticity and cognition⁶⁴. While these mechanistic studies are currently limited in number, they hold the key to understanding the improved efficiency of the hippocampus in EE and revealing key molecular pathways that may be targeted pharmacologically.

Intracellular EE effects in the prefrontal cortex are less well studied. Lesion studies demonstrate that the IL is necessary for the positive behavioral effects of EE, which supports the importance of this region in EE and stress effects (but does not offer a specific mechanism)⁵¹. Exercise decreases reactive oxygen species (ROS) and inducible nitric oxide synthase (iNOS) in the prefrontal cortex²¹⁶, an effect thought to be mediated by increased heat shock protein 70 (HSP70) expression (a chaperone protein known to improve cell survival)^216–218. These antioxidative actions of EE are expected to improve function of prefrontal cortex neurons and combat enhanced reactive oxygen species production associated with chronic stress^{67,110,116,151}.

EE reduces stress-induced activation of the ERK-MAPK-CREB pathway and increases in neuronal activity marker, early growth response protein 1 (EGR-1), expression in the BLA¹⁹⁴. These effects support the idea that EE blunts the response of the amygdala to stress, which would act to dampen pro-stress signals from this region^19,183,194. These stress-related effects do not appear to be present in EE animals under baseline conditions^65,129, mirroring the early observations with dendritic complexity that stress is needed to reveal EE effects in the amygdala. Overall EE has few effects in the amygdala under standard conditions, but its effects emerge when stress is applied, improving the resilience of the animal.

2.3. Effects in Other Brain Regions

This review chose to focus on three major stress-related brain regions; however, EE effects extend beyond these regions and synergistic effects throughout the brain may contribute to the protective nature of EE. Many of the cardinal studies of EE examined the somatosensory cortices. EE consistently increases dendritic complexity^219–222, synaptic plasticity^223–225, and neurotrophins¹⁷¹ in these areas. Similar effects have been noted in the hypothalamus, where EE increases dendritic complexity²²⁶, and cerebellum, where EE increases BDNF¹⁶⁴. In the nucleus accumbens, EE increases DA release, CREB activation, and delta-FosB expression in non-stressed animals^48,227,228. It also attenuates immediate-early gene induction during stress and drug administration²²⁸. These nucleus accumbens alterations are expected to contribute to the anti-addictive effects of EE, which have been reviewed elsewhere²²⁹. In general, these other regions tend to resemble EE effects in the hippocampus and prefrontal cortex, rather than those in the amygdala.

3. Comparison of Brain Regions

All together, these results suggest that the neuroprotective effects of EE against stress arise from heterogeneous mechanisms throughout the brain in adult rodents. EE and stress both modulate a variety of morphological and molecular factors (Figure 3). However, the direction of these changes is dependent upon the region, the EE paradigm, and the presence of stress. This heterogeneity is both intriguing and challenging, for common mechanisms could present novel therapeutic targets and unique mechanisms present new opportunities to truly understand how EE conveys resilience.

Figure 3: Independent effects of EE and stress alone in corticolimbic regions.

Summary of the effects of (A) stress and (B) EE alone on various morphological and molecular endpoints. Coloring reflects the intensity of EE and stress effects in each region. ↑ = increased activity. ↓ = decreased activity. − = no change. na = no data available.

3.1. Common Mechanisms

If we can identify a mechanism of EE that is consistent throughout the brain, it may be possible to target that molecule/pathway pharmacologically. EE in humans is even more heterogeneous than it is in rodents¹¹. There are various ways for humans to attain greater stimulation from their surroundings, in a way that resembles EE in rodents^11,230. For example, people can exercise or spend time with friends or perform cognitively stimulating activities, but these conditions are difficult to control. Human equivalents of EE have been shown to provide numerous benefits, in terms of mood and cognition. EE-like activities including exercise, cognitive training, social stimulation, and video games can help people to cope with stress and convey anti-depressive effects similar to those seen in rodents^9,10,37,231. Some of the molecular mechanisms discussed here have also been recapitulated in humans. For example, cognitive training can prevent hippocampal atrophy, and exercise can increase BDNF in the blood^9,11. Ideally, people would select to engage in a range of these EE-like activities, protecting them from future stressors and helping them recover from past struggles. However, this is often not practical for a variety of circumstances, both internal and external. These challenges introduce the need for development of “enviromimetic” drugs, which are drugs designed to mimic and enhance EE-induced therapeutic effects^230,232. For example, exercise and antidepressants can have synergistic effects to ameliorate maladaptive behaviors^{138,233–235}. In order to maximize beneficial effects and develop more targeted drugs, we need to understand how EE works and identify targets that change consistently throughout the brain. This effort would avoid unintended negative side effects in one region from opposing the expected benefits in other regions. While brain-wide data is not available, we can use the present information regarding highly stress-responsive regions to begin identifying molecular targets that could be manipulated to recapitulate the stress resilience of EE.

One mechanism that is mostly consistent across regions is increased synaptic plasticity. EE increased synaptophysin expression in the hippocampus, prefrontal cortex, and amygdala^65,136,139. While this is just one measure, it suggests that EE increases the overall plasticity of brain, allowing it to adapt to novel (and possibly stressful) stimuli more rapidly. While directly targeting synaptophysin is unlikely in humans, stimulating neurotrophic factors that promote synaptic plasticity could be one potential approach to mimic EE effects¹⁷⁴. Another consistent effect of EE was reduction of stress-induced oxidative stress at the cellular level. While presently these effects are linked to PV interneurons, the improved overall cell survival in EE animals suggests that other cells benefit from this action^{40,67,112,218}. These results suggest that providing stressed individuals with supplemental antioxidants could improve their stress resilience in a similar manner as EE. Further studies are required to identify more consistent mechanisms in EE, as many of the other endpoints examined here showed regional differences.

3.2. Unique Mechanisms

Unfortunately for enviromimetics, the hippocampus, prefrontal cortex, and amygdala showed more differences in their responses to EE and stress than similarities. However, these differences present interesting opportunities to truly dissect the mechanisms underlying EE effects and their role in promoting stress resilience (Figure 4).

Figure 4: Differential interaction of EE and stress in corticolimbic regions.

(A) Summary of the ability of EE to block stress effects on various endpoints. (B) Overall comparison of changes within each region and observations regarding similarities and differences. Coloring reflects the intensity of EE and stress effects in each region. ✓ = EE block stress effects. ✗ = EE does not block stress effects. ↑ = increased activity. ↓ = decreased activity. − = no change. na = no data available.

While the hippocampus is the most studied of these regions, the prefrontal cortex and amygdala are also critical to stress responsivity, with the hippocampus and prefrontal cortex inhibiting stress-related endpoints and the amygdala enhancing them¹⁹. In most cases, EE serves to increase the activity and functionality of the hippocampus and prefrontal cortex, while decreasing that of the amygdala. In the context of stress, these effects would all be predicted to ameliorate negative physiological or psychological consequences, even though they are in different directions. So even though these differential effects make recapitulating EE difficult, common functional effects may promote resiliency in EE animals.

While many endpoints demonstrated this pattern of resilience, the molecular mechanisms at play in each region were often different. Between the hippocampus and prefrontal cortex, the primary differences were the neurotransmitter systems that respond to EE and stress. Minor differences were found in dendritic complexity and intracellular signaling, but studies directly comparing the two regions using the same paradigm are lacking. Other than these endpoints, many of the EE and stress effects were shared between the hippocampus and prefrontal cortex. The amygdala was certainly the most unique of these regions, often showing opposite EE and stress effects. For example, EE blocked stress-induced increases in dendritic branching, neuronal activation, BDNF expression, and GR activity in the amygdala. In contrast, EE blocked stress-induced decreases in the same endpoints in the hippocampus. Again, these opposing functions are consistent with EE conveying resilience against stress in both regions. However, the fact that so many endpoints from so many domains exhibit the same differences is impressive and demonstrates the truly unique nature of the amygdala. It will be interesting to see if these stark differences hold as more region-specific studies are conducted.

It is notable that the amygdala appears to be especially susceptible to stress and important to the stress-responsive elements of EE. Many of the EE changes in the hippocampus were found under baseline conditions, suggesting that the hippocampus benefits the most from EE alone. On the other hand, the EE effects in the amygdala only became evident in the presence of stress, with the effects growing stronger with more chronic stressors. This points to the amygdala as a particularly important target when an animal is stressed. The prefrontal cortex tends to fall in the middle of this stress-responsive spectrum. These observations point to the amygdala as a key region governing the interaction between EE and stress that should be the focus of future mechanistic research.

4. Challenges of EE

4.1. Heterogeneous Nature of Current Research

While the ability of EE to oppose stress-related behavioral problems is exciting, we need to recognize that EE is a complex paradigm with multiple elements, each of which may have different effects on the present endpoints. Variations in exercise, social, and cognitive components can impact the effectiveness of EE^236,237, making it difficult to derive the precise causal element of the molecular effects with our current definition of EE that encompasses these diverse components. The heterogenous nature of EE paradigms likely causes different studies to fall at different points along the inverted U-shaped curve of stress experience, eliciting differential levels of stress resilience. Paradigms that are too weak may fail to exert any effect, while paradigms that are too strong may cause excessive “stress” that is itself harmful. Theoretically, EE that achieves an optimum level of stimulation that trains the circuitry but is not deleterious in itself, should have maximum beneficial effects on an animal’s stress resilience. Finding this precise balance between the diverse components of EE is an important next step in this field, and studies directly comparing different elements of EE will help to delineate the source of different beneficial effects^66,109,111.

While the present review focused on EE in adults, EE actions vary greatly by the age of the animal, with adolescent EE conveying stronger neuroplasticity effects and aged EE magnifying increases in neurotrophins and neurotransmitters^192,238. EE and stress interactions can also vary by the age at which each stimulus is presented¹⁴⁴. These age differences have been reviewed elsewhere^202,239, but should be kept in mind when contemplating translational implications of EE. Differences in species and strain can also influence how EE and stress interact. For example, Wistar rats often exhibit stronger EE responses, both positive and negative, than other rat strains⁷³. Additionally, certain strains of mice sometimes show aggression and establish stronger dominance hierarchies. This can make EE stressful for subordinate mice and may dampen beneficial effects^189,240. However, this aggression is not seen in a majority of rat studies and is often limited to particular strains of mice¹⁸⁰.

Sex differences have also been noted. Some groups have suggested that females are generally more sensitive to EE than males, exhibiting stronger changes in synaptic plasticity and protective effects against stress-induced anxiety-like behaviors, BDNF expression, and hormone responses^{39,81,111,116,151}. Females also appear to be most responsive to the social components of EE, while males are more impacted by cognitive and physical EE^71,241,242. However, a lack of studies directly comparing adult males and females makes it difficult to draw conclusions about on the extent of these differences. More studies comparing sex differences in early-life EE have suggested further differences in cognition, sociability, and HPA axis responses, but developmental confounds limit the current application of these results^{188,238,243,244}. While further studies directly comparing males and females are certainly warranted, EE still appears to increase resilience in females^71,116,136, suggesting that EE’s beneficial actions can present differently and arise from a variety of mechanisms.

4.2. Maladaptive Effects of EE Removal – Relevance to Loss

An additional challenge that is emerging in EE research is the potentially negative consequences of EE removal (ER). While some of the beneficial effects of EE appear to remain if animals are removed from EE and placed into standard housing^117,245, there can be maladaptive consequences from the loss of EE or exercise itself^7,8,246,247. These harmful effects are largely related to mood and appear to be particularly relevant to stress-related domains. For example, ER increases depression-like and anxiety-like behaviors when compared to control animals that never experienced EE^7,8,247. These ER effects should be considered when designing EE experiments, since removing animals from EE prior to behavioral testing could be confounding.

Beyond complicating EE studies, ER presents an interesting stressor in and of itself. Typically, stressing an animal involves exposure to an externally imposed stimulus. The stressful nature of ER revolves around loss of positive reinforcers and is internally generated⁷. Loss is recognized by the Research Domain Criteria as a negative valence construct, and is broadly described as deprivation of something perceived as valuable²⁴⁸. In humans, loss can result from the death of a loved one, losing a job, a health crisis, or financial struggles^248–251. While loss is recognized in humans and can even precipitate the development of depression^252–254, it has received little attention in clinical or basic stress research. This largely results from difficulties in tracking loss in humans and modeling loss in animals. However, studies that have examined loss have yielded interesting similarities between human loss and ER. For example, both cause weight gain and generate hypoactive HPA axis responses to stress. These phenotypes differ from those observed with major depression and chronic stress, setting loss apart as a unique phenomenon^7,255–257. As such, ER presents an interesting opportunity to study the mechanisms underlying loss and could be of great translational value for numerous people suffering from loss. This novel use of EE demonstrates the flexibility of the field and points to the opportunity for new discoveries in future research.

5. Future Directions

5.1. Links to the Inoculation Stress Hypothesis of EE and Importance of the Microenvironment

Despite the heterogeneous nature of EE and its mechanisms, one consistent pattern that emerged from all of this work is that EE improves the efficiency of stress circuitry and fine-tunes its ability to respond appropriately to various challenges. This observation is in line with the conceptual framework of the inoculation stress hypothesis, which proposes that repeated exposure to novel, diverse stimuli in EE prepares an individual to cope with future stress³. It also aligns well with several related theories including the cross-stressor adaptation hypothesis, stress habituation, and the inverted-U theory of stress resilience^18,19,27. The variety of EE-related changes in the brain suggests that benefits are conferred by multiple mechanisms. EE induces changes in neurons, glia, endothelial cells, and numerous communication molecules that are shared between them. Consequently, EE is capable of altering all aspects of synaptic microenvironments across multiple brain areas, including neurons, glia, and blood vessels, as well as the extracellular matrix and extracellular space surrounding the so-called “tetrapartite” synapse^258–261. By modulating the neurotrophins, hormones, neurotransmitters, immune factors, etc. in this microenvironment, EE may be able to promote holistic changes that impact all of the cells in the system. This broad control enables EE to increase the overall efficiency of signaling within each region; ultimately fine-tuning entire microenvironments of neurons to function more effectively. All together, these changes may provide for improved intercellular communications that can withstand energetic challenges promoted by stress exposure, a perspective supported by studies demonstrating that EE improves the efficiency of connectivity between brain regions by weakening unnecessary connections and strengthening important ones^25,26. Overall, the stress-resiliency conveyed by EE appears to be the sum of many morphological and molecular changes working in tandem to improve the efficiency of stress circuitry.

5.2. Need for More Holistic Analyses

Future studies of EE and stress should consider this holistic nature of EE when trying to elucidate the interaction of the two. There is a particular need for more comprehensive omics studies of EE and stress¹¹⁷. Only a few groups have performed such studies, and their application is limited here by a lack of region-specificity or focus on addiction and adolescence^262–265. These studies have noted EE effects similar to those reviewed here and then identified more specific molecules involved in each. Similar studies focused on EE and stress in adults, particularly examining regional differences, would rapidly advance the field. This type of information will be invaluable both in efforts to develop novel therapies for stress-related disorders and in deepening our understanding of the complex relationship between EE and stress.

HIGHLIGHTS.

• Environmental enrichment (EE) can improve stress resilience in adulthood.

• EE improves the overall efficiency of stress-processing corticolimbic regions.

• EE differentially impacts the hippocampus, prefrontal cortex, and amygdala.

• These effects may reveal novel therapeutic targets in stress-related disorders.

ABBREVIATIONS

5-HT

Serotonin

5-HT1A

Serotonin 1A Receptor

ACh

Acetylcholine

ACTH

Adrenocorticotrophic Hormone

AMPA

α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid

BDNF

Brain Derived Neurotrophic Factor

BLA

Basolateral Amygdala

CCL2

C-C Motif Chemokine Ligand 2

CCL3

C-C Motif Chemokine Ligand 3

CCO

Cytochrome C Oxidase

Cg3

Anterior Cingulate Cortex

CORT

Corticosterone

CREB

cAMP Response Element-Binding protein

CRH

Corticotropin Releasing Hormone

CX3CR1

CX3C Chemokine Receptor 1

CXCL12

C-X-C Motif Chemokine 12

DA

Dopamine

D1

Dopamine Receptor 1

DAT

Dopamine Transporter

DG

Dentate Gyrus

EE

Environmental Enrichment

EGR-1

Early Growth Response protein 1

EPM

Elevated Plus Maze

ER

Enrichment Removal

ERK

Extracellular signal-Regulated Kinase

FST

Forced Swim Test

GABA

Gamma Aminobutyric Acid

GDNF

Glial cell-Derived Neurotrophic Factor

GFAP

Glial Fibrillary Acidic Protein

GR

Glucocorticoid Receptor

HIF-1α

Hypoxia-Inducible Factor 1-alpha

HPA

Hypothalamic-Pituitary-Adrenal

HSP70

Heat Shock Protein 70

IBA1

Ionized calcium Binding Adaptor molecule 1

IFNγ

Interferon gamma

IGF-1

Insulin-like Growth Factor 1

IL

Infralimbic

IL-10

Interleukin 10

IL-1Ra

Interleukin 1 Receptor antagonist

IL-1β

Interleukin 1 Beta

IL-6

Interleukin 6

iNOS

inducible nitric oxide synthase

LTD

Long-Term Depression

LTP

Long-Term Potentiation

MAPK

Mitogen Activated Protein Kinase

NE

Norepinephrine

NGF

Nerve Growth Factor

NMDA

N-methyl-D-aspartate

NPY

Neuropeptide Y

NT-3

Neurotrophin-3

OF

Open Field

OFC

Orbitofrontal Cortex

PKA

Protein Kinase A

PKM

Protein Kinase M

PL

Prelimbic

PSD95

Post Synaptic Density protein 95

PTSD

Post-Traumatic Stress Disorder

PV

Parvalbumin

ROS

Reactive Oxygen Species

SGK

Serum and Glucocorticoid-Regulated Kinase 1

SIRT 1

Sirtuin 1

SPT

Sucrose Preference Test

TNFα

Tumor Necrosis Factor alpha

TrkB

Tropomyosin receptor kinase B

VEGF

Vascular Endothelial Growth Factor