Reduced adaptation of glutamatergic stress response is associated with pessimistic expectations in depression

Stress is a major risk factor for MDD. Toxic stressors, characterized by unpredictability and uncontrollability, often involve social threat. A common consequence is stress-induced anhedonia, marked by behavioral inhibition and reward avoidance. Stress reduces reward-related information acquisition and blunts activity in reward processing brain regions like the mPFC. Responses to stress vary; only stress-reactive individuals show diminished reward sensitivity, and high perceived stress increases risk for blunted reward processing. The neural mechanisms underlying stress-induced anhedonia and the interplay between acute and perceived stress remain unclear. The mPFC plays a crucial role in stress-induced anhedonia. Preclinical research shows stress negatively impacts the mPFC, including glutamate-mediated excitotoxicity due to elevated glucocorticoids. Initial stress increases extracellular glutamate, potentiates excitatory currents, and upregulates glutamate receptors in rodent mPFC, leading to short-term improvements in learning and memory. However, repeated stress causes glutamate release habituation to subsequent stressors. Animals exposed to chronic stress show reduced glutamatergic signaling in response to new stressors. This reduction might be a protective mechanism. The mPFC's role in coordinating behavioral and endocrine responses to stress, and in reward valuation and expectation of future outcomes, makes its glutamate function particularly relevant to stress-related psychopathology and learned helplessness. This study investigates the relationship between acute stressor-induced mPFC glutamate changes and perceived stress using MRS, examining healthy adults and MDD patients. The researchers hypothesized that in healthy individuals, mPFC glutamate would increase following acute stress with low perceived stress and decrease with high perceived stress; this adaptive response would be disrupted in MDD.

The existing literature strongly links stress to major depressive disorder (MDD), categorizing stress into "good," "tolerable," and "toxic" stress, with the latter strongly associated with mental illness. Toxic stressors, often involving social threats like isolation, rejection, and exclusion, frequently lead to stress-induced anhedonia—a reduced capacity to experience pleasure. This anhedonia manifests as behavioral inhibition and an inability to pursue rewards. Studies show stress impairs the acquisition of reward-related information and reduces activity in corticostriatal regions vital for reward processing. Individual differences in stress responses are crucial; only stress-reactive individuals exhibit blunted reward processing. Increased perceived stress significantly elevates the risk of this blunted response. Preclinical studies extensively demonstrate the mPFC's sensitivity to stress; stress induces glutamate-mediated excitotoxicity through elevated glucocorticoids. Acute stress initially boosts extracellular glutamate in the mPFC, enhancing postsynaptic excitatory currents and upregulating AMPA and NMDA receptors. However, this effect habituates with repeated stress exposure, and chronically stressed animals show reduced glutamatergic signaling in response to new stressors. This reduced glutamate response is potentially a protective adaptation. The mPFC's role in regulating behavioral and endocrine stress responses, as well as in reward valuation and future outcome expectancy, links its function to depression. Animal studies further link mPFC to learned helplessness, indicating an expectation of inaction's inability to influence future outcomes. Prior studies demonstrate the relationship between perceived stress and blunted neural responses to rewards in the mPFC, however the neural mechanisms and interaction of acute and perceived stress remained unclear, prompting this research.

This study used magnetic resonance spectroscopy (MRS) to measure glutamate metabolites in the mPFC of four groups: three independent samples of healthy adults with varying levels of perceived stress, and one group of unmedicated individuals with MDD. Participants completed self-report measures (Perceived Stress Scale [PSS] and Stress and Adversity Inventory [STRAIN]) and underwent MRI sessions with two mPFC MRS assessments. Two healthy control groups completed the Maastricht Acute Stress Task (MAST), designed to be unpredictable, minimally controllable, and include a social threat component, between MRS scans. A third healthy control group completed a no-stress control manipulation. The MDD group also underwent the MAST. Mood was assessed using the Visual Analog Mood Scale (VAMS). Salivary cortisol was measured before and after the MAST or control manipulation. The study tested the hypothesis that mPFC glutamate would increase after acute stress in healthy individuals with low perceived stress and decrease with high perceived stress, while this adaptive response would be absent in MDD participants. The correlation between PSS scores and percent change in mPFC glutamate (%∆Glu) was analyzed, followed by a replication study on another healthy control sample. The association between %∆Glu and PSS was compared between the stress and no-stress groups using hierarchical linear regression. The researchers also explored the relationship between %∆Glu and STRAIN. Finally, a 4-week ecological momentary assessment (EMA) was used to evaluate the link between mPFC glutamate function and reward processing in daily life in the Emory samples.

The MAST successfully induced a mood and salivary cortisol response, validating the stress manipulation. In healthy control participants, %∆Glu was inversely correlated with PSS scores (r = -0.457, p = 0.022 in the first sample and r = -0.517, p = 0.014 in the replication sample), indicating an adaptive response: individuals with low PSS showed increased glutamate, while those with high PSS showed no change or a decrease. This relationship was not observed in the no-stress control group, confirming the specificity of the effect. Hierarchical linear regression showed a significant interaction between acute stress and PSS in predicting %∆Glu. In contrast, no significant correlation was found between PSS and %∆Glu in the MDD group (r = 0.115, p = 0.602). Moderation analysis confirmed a significant difference in how PSS predicted mPFC %∆Glu response to acute stress as a function of MDD. A "maladaptive glutamate response" (MGR) metric was created to represent the difference between actual %∆Glu and expected %∆Glu based on PSS. In the EMA, MDD participants reported higher negative affect, lower positive affect, and lower expected and experienced outcomes than controls. They showed less accurate expectations and a higher proportion of pessimistic expectations. MGR was positively correlated with negative affect (r_partial = 0.367, p = 0.030) and pessimistic expectations (r_partial = 0.457, p = 0.006), but not optimistic expectations. The association between MGR and pessimistic expectations remained significant even after controlling for PSS and frequency of pessimistic expectations.

This study reveals an adaptive glutamatergic response to acute stress in the human mPFC, where healthy individuals show reduced mPFC glutamate response to a new stressor as recent perceived stress increases. This adaptation was absent in unmedicated individuals with MDD, potentially contributing to stress-related psychopathology. The lack of adaptive glutamate response correlated with negative daily life functioning, particularly pessimistic expectations. The inverse relationship between perceived stress and mPFC glutamate in healthy individuals suggests a beneficial adaptation, especially considering that participants had no psychiatric history. This supports the allostatic regulation model, predicting that responses to acute stressors are modulated by recent perceived stress. Preclinical studies support this, showing that acute stress potentiates glutamate transmission, an effect reversed by prior stress exposure. The absence of this adaptation in MDD participants indicates a disruption in stress response regulation. The MGR metric's association with pessimistic expectations underscores the importance of anticipatory anhedonia in MDD. This aligns with previous research implicating the mPFC in evaluating the expected value of future outcomes. While the PSS showed a strong association with mPFC glutamate, STRAIN (assessing lifetime stress exposure) did not, possibly due to differing timescales or stress components measured. No differences in baseline glutamate levels were observed between MDD and control groups, suggesting that cross-sectional comparisons may be insufficient. The study acknowledges limitations including sample size, age range, and PSS distribution. The observed age-related glutamate changes also need further investigation.

This study is the first to identify the attenuation of mPFC glutamate as an adaptive response to acute stress in the context of perceived stress and shows that this response is impaired in individuals with depression. The findings highlight the importance of considering contextual factors like perceived stress when interpreting mPFC glutamate levels in the context of stress responses and suggest that the absence of this adaptive response may be a potential biomarker for or contributor to stress-related psychiatric disorders. Further research should explore this adaptation's relationship to various types of stressors, and the development of targeted interventions to restore this adaptive response in MDD.

The study's limitations include a moderately sized sample, a limited age range potentially impacting the generalizability of age-related findings, and a lack of complete overlap in PSS scores between MDD and healthy control participants. The MRS methodology provides a measure of intracellular glutamate, not synaptic glutamate directly, potentially limiting inferences on neurotransmission. Additionally, the specific temporal resolution of the MRS and salivary cortisol measurements might have limited the ability to find relationships between them. Finally, the cross-sectional nature of the study limits causal inferences.