A convergent depression vulnerability pathway encoded by emergent spatiotemporal dynamics

Abstract

Fluctuations in brain local field potential (LFP) oscillations reflect emergent network-level signals that mediate behavior. Cracking the code whereby these LFP oscillations coordinate in time and space (spatiotemporal dynamics) to represent complex behaviors would provide fundamental insights into how the brain signals emotional processes at the network level. Here we use machine learning to integrate LFP activity acquired concurrently from seven cortical and subcortical brain regions into an analytical model that predicts the emergence of depression-related behavioral dysfunction across individual mice subjected to chronic social defeat stress. We uncover a spatiotemporal dynamic network in which activity originates in prefrontal cortex (PFC) and nucleus accumbens (NAc, ventral striatum), relays through amygdala and ventral tegmental area (VTA), and converges in ventral hippocampus (VHip). The activity of this network correlates with acute threat responses and brain-wide cellular firing, and it is enhanced in three independent molecular-, physiological-, and behavioral-based models of depression vulnerability. Finally, we use two antidepressant manipulations to demonstrate that this vulnerability network is biologically distinct from the networks that signal behavioral dysfunction after stress. Thus, corticostriatal to VHip-directed spatiotemporal dynamics organized at the network level are a novel convergent depression vulnerability pathway. Main TextMajor depressive disorder is the leading cause of disability in the world [1]. While stress contributes to the onset of depression [2,3], only a fraction of individuals that experience stressful events develop behavioral pathology. Multiple factors including childhood trauma and alterations in several molecular pathways have been shown to increase disease risk [3,4]; nevertheless, the neural pathways on which these factors converge to yield subthreshold changes that render individuals vulnerable to stress are unknown. Knowledge of these neural pathways would facilitate the development of novel diagnostic technologies that stratify disease risk as well as preventative therapeutics to reverse neural circuit endophenotypes that mediate vulnerability to depression. To achieve this aim it is essential to distinguish the neural alterations that confer vulnerability to depression from those that accompany the emergence of behavioral dysfunction.Chronic social defeat stress (cSDS) is a widely validated pre-clinical model of depression [5][6][7]. In this paradigm, test mice are repeatedly exposed to larger aggressive CD1 mice. At the end of these exposures, test mice develop a depression-like behavioral state characterized by social avoidance, anhedonia-and anxiety-like behavior, and sleep/circadian dysregulation [6,7]. Critically, only ~60% of C57 mice subjected to this paradigm exhibit susceptibility to developing this stress-induced syndrome. While the remaining ~40% of mice subjected to cSDS exhibit resilience [6], susceptible and resilient mice experience the...

Non-rapid eye movement sleep determines resilience to social stress

“…Functional magnetic resonance imaging studies reveal that sleep deprivation leads to decreased functional connectivity between the ventromedial prefrontal cortex (vmPFC) and amygdala; furthermore, this decreased functional connectivity is associated with decreased mood (33)(34)(35). This same brain region, the vmPFC, is known to mediate resilience to social defeat stress (21,(36)(37)(38)(39). We reasoned that the sleep deprived state of susceptible mice would lead to vmPFC disfunction; thus, leading to decreased connectivity and an inability to inhibit the limbic circuits responsible for regulating behavioral responses to social stress.…”

Neural Decoding and Feature Selection Techniques for Closed-Loop Control of Defensive Behavior

“…diverse array of machine learning (ML) models has been employed for neuropsychiatric tasks and 42 brain-machine interface applications, including linear 43 regression (LR)[62,103], support vector machine 44 (SVM)[63,67, 77,107,108], random forest (RF) [109,45 110], and artificial neural network (ANN) [111, 112]. 46 Moreover, gradient-boosted decision trees (GBDT) have demonstrated promising performance in previous 48 neurophysiological task studies [90, 91, 97, 107, 113].…”