Adaptation as a dynamic construct for studying stress resilience and susceptibility

Resilience is the process of adapting favorably to stress, involving both passive and active strategies (Dantzer et al., 2018; Russo et al., 2012). Susceptibility arises through failure to adapt favorably, and is characterized by maladaptive responses to stress that can increase the risk for stress-related disorders. As stress is not always avoidable and can even be beneficial when it elicits adaptive responses, studying the neurobiology of stress resilience in rodent models may be a viable approach for advancing prevention and treatment of stress-related disorders in humans. However, translating from rodents to humans depends upon the definition of adaptive within each species and addressing the complexity of defining this construct in rodents. Over-operationalizing resilience and susceptibility risks losing this complexity because defining what is adaptive to a rodent requires continual scrutiny since it is difficult to infer this from a human perspective. In the recently published article in BBI by Gururajan et al. (2019), the authors initiate an investigation of signatures of resilience and susceptibility, while demonstrating a need to reevaluate whether each component of the signature is adaptive or not.

Adaptive responses to stress are not permanently defined measures and being resilient is not a fixed characteristic (Dantzer et al., 2018). Any one individual can have a range of both adaptive and maladaptive responses and the same response can be either adaptive or maladaptive, depending on the physiological system assessed, the context evaluated, and timing both after stress and across the lifespan. The repeated social defeat (RSD) stress paradigm studies two distinct populations of rodents, based on characterizing mice with reduced social interaction ratios (the day after the final defeat) as susceptible and mice who maintain social interaction despite defeat as resilient (Golden et al., 2011). However, the ‘susceptible’ group may not always be maladaptive, abnormal, or prone to pathology and the ‘resilient’ group may not always be adaptive or ideal. Even commonly conceded anxiety-like or depressive-like behaviors could be adaptive given the context, duration, and physiological state of the animal (Myers et al., 2015). Gururajan et al. (2019) addresses the complexity of resilience versus susceptibility by studying an impressive range of physiological, central, endocrine, and immune parameters following a modified RSD paradigm. This investigation provides a platform for future work to refine the definition of adaptive in order to understand mechanisms behind stress resilience that translate to human experience.

With susceptible or resilient categorization based on social interaction ratios, resilient mice appear to have failed threat appraisal because they continue approaching the aggressor mouse strain. Reduced social interaction in susceptible mice after RSD suggests that they have learned that the aggressor mouse strain is a real threat. In a primitive sense, this threat appraisal and social avoidance is adaptive for rodents. In a modern sense for humans, excessive social avoidance in the absence of a threat may be a feature of stress-related psychiatric disorders. Susceptible mice also display long-lasting generalized social avoidance to mice of a conspecific strain (Russo et al., 2012). This exaggerated and prolonged social avoidance may indeed be maladaptive, especially in a modern sense where humans frequently suffer from psychological stress in the absence of physical threat that contrasts from primitive physical stress. Accordingly, the modified RSD model (Gururajan et al., 2019) limits attacks, which lessens the impact of physical wounding to highlight the psychological effect of defeat. Future work should investigate threat appraisal and prolonged or generalized social avoidance, along with how they relate to physiological, central, endocrine, and immune outcomes reported in Gururajan et al. (2019).

HPA and immune crosstalk is tightly linked to both adaptive and maladaptive responses to stress, with Gururajan et al. (2019) reporting divergent HPA axis and immune signatures in susceptible versus resilient mice. Glucocorticoids mobilize energy to face real or anticipated environmental demands, while providing context-dependent and coordinated cellular, molecular, and behavioral responses to stress (Myers et al., 2015). Given the context of RSD, a rise in circulating glucocorticoids would seem adaptive to enable the animal to energetically respond to subsequent aggressor mice and facilitate threat appraisal. However, this elevation during circadian nadir and exacerbated adrenal hypertrophy reported in susceptible mice in Gururajan et al. (2019) could indicate excessive HPA axis activation or impaired negative feedback. Previously, both susceptible and resilient mice had exaggerated corticosterone levels following acute stress (Krishnan et al., 2007), indicating that RSD exposure sensitizes the HPA axis to subsequent stressors. This sensitization could be adaptive to respond to concurrent stress while elevated basal levels may indicate a maladaptive response. Interestingly, resilient mice can have elevated basal corticosterone 4 weeks after RSD while susceptible mice instead had a reduction (Krishnan et al., 2007), indicating that future work must address temporal dynamics and functional relevance to differentiate between adaptive versus maladaptive responses. A recent study (Gaelle et al., 2019) found that an exaggerated intra-PFC corticosterone response drives working memory deficits after stress. Paired with the reduced PFC CRH expression in resilient mice seen in Gururajan et al. (2019), the PFC remains a prime target for understanding glucocorticoid signaling after chronic stress and may reveal specific insights into adaptive versus maladaptive responses.

Excessive HPA axis activation can interact with the immune system to cause glucocorticoid resistance in immune cells, which is predominantly considered a maladaptation that can increase the risk for infection upon exposure to pathogens (Dantzer et al., 2018; Niraula et al., 2018). Increased inflammatory monocytes seen in susceptible mice in Gururajan et al. (2019) may reflect a maladaptive response, since increased innate immune cell activation, along with glucocorticoid resistance, can drive behavioral manifestations of RSD that may indicate susceptibility (McKim et al., 2017; Ménard et al., 2017; Niraula et al., 2018). Conversely, stress-induced stimulation of the adaptive immune system can provide inoculation against subsequent stressors to facilitate resilience (Ménard et al., 2017), so evaluating adaptive immune parameters may provide additional insight in future studies. At any rate, within the RDS model, it seems adaptive for the animal to mount some type of immune response to prepare for physical attacks. However, this depends on many variables and Gururajan et al. (2019) highlights a future need to test immune cell responsivity, functionality, temporal dynamics, and the relation to behavior and CNS function.

Given the multifaceted nature of adaptive behavior, defining susceptibility and resilience will require an additional level of research into individual differences across biological systems and across time. The important first steps are obtaining a range of data, as initiated in Gururajan et al. (2019), while future approaches should integrate these data together while taking into account individual variability to inform precision medicine approaches that are becoming increasingly more successful in the clinic.