Inhibitory corticotropin-releasing factor neurons in the dorsomedial prefrontal cortex promote stress-resilient behavior in male rodents

Abstract

Adaptive behavior in response to adverse experiences facilitates faster recovery and less time spent engaging in maladaptive behaviors that contribute to psychopathology, including anxiety, depression, and posttraumatic stress disorder. Dysregulation of activity in the prefrontal cortex (PFC) has been implicated in these disorders, with stress further exacerbating these negative effects. Corticotropin-releasing factor is an important regulator of the stress-response system and may have a differential involvement in individuals who are especially susceptible to negative stress-related outcomes. A recent study by Chen et al. (Chen P, Lou S, Huang ZH, Wang Z, Shan QH, Wang Y, Yang Y, Li X, Gong H, Jin Y, Zhang Z, Zhou Z. Neuron 106: 301–315, 2020) has identified a novel subtype of GABAergic CRF interneurons in the dorsomedial PFC that, upon activation, can promote active responses and resilient behavior in response to stress.

INTRODUCTION

Although everyone experiences stress in response to environmental adversity, some individuals are exposed more often and more severely than others. Exposure to stressful life events has been associated with increased risk of developing psychiatric disorders, including posttraumatic stress disorder (PTSD), major depressive disorder (MDD), and substance use disorder (SUD; reviewed in 1). However, greater exposure to stressors does not always equate to more detrimental effects, as research has shown that some individuals are particularly susceptible, whereas others are resilient to these negative outcomes (reviewed in 1, 2). Understanding the neurobiological mechanisms of individual differences in stress responses and resiliency can facilitate the development of novel therapeutics for preventing and treating psychiatric disorders.

The behavioral repertoire that an individual exhibits while coping with stress can influence the likelihood of consequent psychopathology (1). For example, avoidance can be an adaptive response to a direct threat, but maladaptive after the acute threat has passed. The inability to utilize adaptive responses can lead to emotion reactivity and maladaptive emotion regulation strategies such as avoidance, leading to both anxiety and depression (reviewed in 1, 3). Dysregulated activity in the medial prefrontal cortex (mPFC) has been reported in individuals with anxiety, depression, and PTSD, with increased top-down control expected to improve cognitive appraisal of an adverse event and facilitate quicker recovery (2). Thus, it is possible that mPFC activity serves as a resilience factor in response to stressful life events.

One of the key players in regulating the stress response is corticotropin-releasing factor (CRF), which is released in prefrontal regions in addition to the hypothalamus as part of the stress-response system (4). Release of CRF triggers an increase in cortisol, followed by a decrease in CRF, thereby maintaining a negative feedback loop under healthy conditions. However, this loop, known as the hypothalamic-pituitary-adrenal axis, can malfunction in individuals exposed to chronic stress, and abnormal CRF levels may mediate or moderate the link between stress and depression (4).

To summarize, the PFC plays an important role in the ability to utilize adaptive responses to environmental stress. CRF is released from the PFC, has a demonstrated role in cognition and reward-associated behaviors, and is an important regulator of the stress response (4, 5). Therefore, chronic dysregulation of the CRF stress-response system in the PFC could play a key role in sensitization to stress. However, the differential cell-circuit involvement of CRF in coping styles and in susceptible versus resilient phenotypes in response to stress remains unknown.

The use of animal models of stress allows researchers to investigate neural pathways that may be responsible for the effects of CRF in these stress-related disorders. With chemogenetic tools like designer receptors exclusively activated by designer drugs (DREADDs), cells can be manipulated to either increase or decrease their activity. DREADDs can be used in conjunction with mutated genes so that specific neuronal populations are targeted in select brain regions, allowing researchers to observe and measure behavioral responses that are unique to the cell circuit being manipulated. A recent study published in Neuron by Chen et al. (6) used this approach in conjunction with established behavioral models of anxiety and depression in mice to characterize mPFC CRF neurons as a unique subtype of GABAergic inhibitory interneurons that modulate inactive versus active behavioral responses and susceptible versus resilient phenotypes.

METHODS

Medial PFC CRF neurons were targeted in adult male CRF transgenic mice that were bred so that the expression of CRF and/or GABA could be visualized by distinct fluorescent proteins. Since subtypes of GABAergic neurons can be differentiated by their protein composition, the authors labeled this population of CRF neurons with molecular markers for established GABAergic cell types and quantified the colocalization with CRF to determine whether this population reflects a unique subtype. The authors then evaluated whether these CRF neurons project to and inhibit local pyramidal neurons by using optogenetics to selectively turn on and off the CRF neurons while using whole patch clamp recording of pyramidal neurons in layer 5 of the mPFC to determine excitability.

After the authors isolated this CRF neuron subtype, they investigated the extent to which they are recruited during the active or inactive behavioral responses to stress. The mice first underwent a tail suspension test (TST) where they were suspended by their tail from a horizontal bar and their movement was recorded as either an inactive, “immobile” behavior or an active, “struggling” behavior. When placed in water, a rodent will typically struggle to escape, whereas exhibiting an inactive, immobile response is considered to be depression-like behavior (7). Using Ca²⁺ imaging, the activity of the CRF neurons in the dorsomedial PFC (dmPFC) was recorded during this task to quantify their involvement relative to the behavioral response.

Next, the authors determined whether activity of these dmPFC neurons was necessary for expression of immobile behavior in the TST and forced swim test (FST), less social interaction in the social approach test, and less exploratory time in the elevated plus maze (EPM) and open field (OF) test. The FSW and TST are considered rodent models of depression, giving insight into coping strategy to acute, inescapable stressors. The FST involves placing a mouse in a cylinder filled with water while measuring struggling as active movement and immobility time as inactive movement. In the EPM, the mice are placed on an elevated platform surrounded by open and closed arms and can explore. Exploratory behavior is quantified as entries and time spent in the open arms of the EPM. In the OF test, the animal is placed in a box with four corners, and more time spent in the center of the field was classified as exploratory behavior, whereas total distance moved served as an important control evaluating changes in overall locomotion. The OF test and EPM are both considered animal models of anxiety, as rodents naturally avoid open and lit areas. Finally, the social interaction test consisted of two parts: a social defeat stress test and a social approach test. In the social defeat stress test, the mice were physically exposed to two aggressor mice on two separate occasions. This was followed by the social approach test where their interaction behavior, measured by the amount of time they spent approaching a zone with an unfamiliar CD-1 mouse, was recorded. The authors tested the effects of manipulating mPFC CRF neurons on these behaviors by 1) injecting a viral vector bilaterally in the dmPFC to induce cell death in the CRF interneurons before challenging the mice with several stressors, and 2) using designer receptor exclusively activated by designer drugs (DREADDS) as a chemogenetic approach to inhibit and activate (with Gi- and Gq-DREADDS; respectively) the CRF neurons before each behavioral test. To confirm the chemogenetic results, the authors additionally used an optogenetic approach to increase the activity of dmPFC CRF neurons. They also assessed whether activating CRF neurons was sufficient to induce resiliency by repeating the social defeat test across 10 days, while activating CRF neurons using DREADDS. Chronic social defeat is considered a rodent model of depression (4, 7), with social interaction being a measure of avoidance behavior. In this case, the time the animals engaged in stress-coping behaviors and social interaction after CRF neuronal activation was measured.

RESULTS

Using this approach, Chen et al. (6) identified a unique population of GABAergic interneurons in the dmPFC that is responsible for differentiating active versus inactive coping behaviors in response to stress, as well as characterizing a stress-resilient phenotype. Confocal imaging revealed a majority of the CRF neurons targeted in the mPFC as GABAergic. Histology revealed that coexpression of CRF neurons never exceeded 50% for any individual marker of GABAergic interneuron—with the greatest coexpression seen with vasoactive intestinal polypeptide and calretinin at ∼41% and 33%, respectively. Although this does not completely rule out the involvement of other GABAergic interneuron types in the behavioral results, it does support the existence of a unique CRF GABAergic interneuron subtype. In addition, these neurons were shown to project to and inhibit local layer 5 pyramidal neurons. By ablating the CRF neurons in the dmPFC, the authors demonstrated their essential role in increasing active stress coping and exploratory behaviors. The absence of CRF neuronal activity led to a significant increase in immobility time during the TST, a nonsignificant increase in immobility time during the FST, and a significant decrease in open arm time and entries (i.e., exploration) in the EPM. Decreasing CRF neuronal activity through inhibition significantly increased immobility time in both the FST and TST, and decreased open arm time and entries in the EPM. Ablation and inhibition also decreased the time the mice spent in the social interaction zone when exposed to an unfamiliar CD-1 mouse.

In contrast, increased activity of the dmPFC CRF neurons had the opposite effect on behavior. Activation before the OF test and EPM increased exploratory time, and activation before the TST and FST decreased immobility time. Moreover, activation before each bout of social defeat during 10 days of persistent social stress increased time spent approaching the social interaction zone when an unfamiliar CD-1 mouse was present. The authors identified a subset of “stress-resilient” mice that spent more time approaching the location of the unfamiliar mouse when it was present versus absent. Interestingly, there were more of these stress-resilient mice in the group that had dmPFC CRF neurons activated before each bout of social defeat, suggesting this manipulation led to more stress-resilience. Importantly, no differences were observed in locomotor activity throughout the experiments, supporting the specific involvement of the dmPFC CRF neurons in promoting active behavioral responses to stress, as opposed to altered locomotion.

DISCUSSION

These findings illustrate the role of dmPFC CRF neurons in modulating behavioral responses to stressful challenges in rodents, and highlight the insights that could be gained by investigating individual differences in behavioral response to stress in humans. Chen et al. (6) showed that CRF GABAergic interneurons in the dmPFC are necessary for exhibiting resilient behavior when faced with an acute stressor, as shown by increased social interaction, increased exploration, and decreased immobility. Resilient behavior was also demonstrated in response to persistent social stress in a model, which uses social conflict to inflict emotional and psychological stress (7). Although a subset of mice was identified as stress-resilient based on the chronic social defeat test, it is unclear whether they also demonstrated the lowest immobility time (i.e., less depression-like behavior) and highest exploratory time (i.e., less anxiety-like behavior) in the other tests. Evaluating the relationship between these behaviors could elucidate whether resiliency carries over to different types of stressors and associated responses.

Although the findings reported by Chen et al. (6) have important implications for understanding individual differences in susceptibility to negative impacts of stress, it is important to note that the experiments described therein were performed exclusively in male mice. Notably, there are established sex differences in the prevalence of anxiety disorders and depression, with higher rates in females than in males (reviewed in 8). As such, females may be especially well suited to study susceptibility and resilience to stress effects. Bangasser and Wiersielis (8) reviewed sex differences in CRF function and identified several that reflect enhanced effects of CRF in females compared with males, including greater CRF expression and differences in receptor expression, distribution, trafficking, and signaling in several brain regions. Further, there is evidence that these differences impact CRF-mediated responses to stress, including a study by Li et al. (9) that showed that CRF-binding protein (CRFBP) released by oxytocin interneurons in the mPFC was effective at reducing anxiety in males, but not in females, likely due to higher CRF expression in females that exceeded the ability of CRFBP to bind and reduce the activity of CRF on its receptors. This highlights the importance of carrying out experiments in both males and females to fully characterize CRF-effects on stress and potential novel approaches to treatment.

In addition to evaluating sex differences, efforts to identify a complete cortical-to-subcortical circuit for stress-resilient behavior could facilitate the discovery of therapeutic interventions. Chen et al. (6) discovered a population of CRF interneurons in the dmPFC, which project to and inhibit local pyramidal neurons. Future research can further explore this mechanism by investigating the output regions of these layer 5 pyramidal neurons after mice are exposed to a chronic social defeat paradigm and/or the FST and TST. Layer 5/6 pyramidal neurons project subcortically to the ipsilateral striatum, thalamus, pons, and brainstem (10), and it has previously been shown that mPFC layer 5/6 pyramidal neurons expressing dopamine 1 receptors (D1-PYR) have increased excitability following 4 wk of chronic unpredictable stress (11). Furthermore, since mPFC glutamatergic projections to the NAc (nucleus accumbens, main component of ventral striatum) have been implicated in depressive-like behavior (12), it is possible that in stress-resilient mice activation of dmPFC CRF neurons inhibits activity of D1-PYR, and subsequent activity in the NAc.

Another important avenue worth investigating is whether this population of CRF GABAergic interneurons is involved in a broad range of disorders that share stress-induced negative affect. Stress plays a large role in relapse among those with substance use disorder, and CRF₁ receptor activity is upregulated in preclinical models of alcohol binge drinking (5). Blocking CRF₁ receptors in the ventral tegmental area, a subcortical dopamine hub responsible for reward-related behavior, decreases alcohol drinking in mice and rats. Moreover, a recent review (13) suggests a role of infralimbic (rodent analog to vmPFC in humans) CRF in compulsive alcohol seeking behavior. Despite this evidence, CRF receptor antagonists have been unsuccessful in treating alcohol use disorder in clinical trials (5). One possible explanation is that systemic CRF antagonist drugs may be preventing the critical action of CRF neurons in brain regions responsible for promoting resilience to stress. These current findings by Chen et al. (6) offer preliminary evidence for a cell-circuit mechanism that may be a target for therapeutics used by individuals who are most susceptible to stress-induced psychopathology. It is unclear whether CRF release from this CRF-containing population of interneurons is responsible for any of the reported behavioral outcomes, and future studies should look into this. In addition, an essential next step before clinical drug development will involve investigating whether homologous PFC CRF GABAergic interneurons exist in non-human primates.

In summary, Chen et al. (6) identified a unique subtype of CRF GABAergic interneurons in the dmPFC that is recruited during stress and is able to promote active coping behaviors and resiliency upon activation. This resiliency was demonstrated as increased exploratory time and social interaction in established rodent models. The discovery of this subtype of inhibitory interneuron in the cerebral cortex offers novel insight into stress-related pathology and paves the way for future research to investigate subcortical projections. This is an important step toward facilitating the growth of personalized medicine through the development of therapeutic interventions that address individual differences in response to stress.

GRANTS

This work was supported by National Institute on Drug Abuse Grant T32DA724430 and National Institute on Alcohol Abuse and Alcoholism Grant P60AA011605.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

E.M.V. and M.M.R. drafted manuscript; E.M.V. and M.M.R. edited and revised manuscript; E.M.V. and M.M.R. approved final version of manuscript.