Given the persistence of stress-related disorders, it is critical to identify the circuits, cells and mechanisms that fail to recover from stressor exposure. The stress hormone glucocorticoid causes reversible remodeling of brain circuits involved in cognitive, executive and hedonic behaviors. While small amounts of glucocorticoids support cognition and spine synapse turnover, excess stimulation by glucocorticoids predisposes to stress disorders via the loss of dendritic spines. To date, however, because of the paucity of knowledge of underlying mechanisms, deleterious effects of glucocorticoids are not prevented following extreme stress. My model predicts a fine balance exists between the spine formation and spine elimination needed for behavioral adaptation to stress. Which spines survive, which spines are eliminated may encode the structural basis of learning adaptive behavior. There is hope that promoting dendritic spine turnover facilitates recovery from stressful experience, and paves the road for the discovery of better drug treatments for disorders featuring disrupted glucocorticoid circadian rhythms. One hypothesis to account for the onset and severity of these disorders is that glucocorticoid actions diverge as a function of brain-derived neurotrophic factor (BDNF) signaling. Using mouse genetics, behavior, in vivo imaging and transcriptomic, the “SPICED” project aims at charactering a novel neurotrophic-sensing mechanism of glucocorticoid actions on spine turnover in changing environment. The SPICED project capitalizes on robust preliminary data arguing that BDNF signaling employs the phosphorylated glucocorticoid receptor (GR) as transcription factor to produce a unique gene expression signature. Should BDNF-induced GR phosphorylation become impaired, appropriate turnover of dendritic spines to durable stress or excess glucocorticoids may be inhibited thus setting the stage for neural network wiring defects and development of neuropsychiatric disorders.