| Abstract: |
Background: After experiencing life-threatening events, one may develop posttraumatic stress disorder (PTSD), which is notably characterized by abnormalities in fear memory. Studies have highlighted a correlation between theta (4 Hz) brain wave activity and fear expression in animal models. The medial prefrontal cortex (mPFC) is a key region of top-down fear regulation, elucidating the temporal dynamics of phenotypic manifestations and gene expression patterns is crucial for a comprehensive comprehension of PTSD pathogenesis. Aims & Objectives: The objective of this study was to further understand the neurobiological underpinnings of PTSD, particularly focusing on (1) The association between pathological fear memory and brain wave characteristics, also the effects of transcranial direct current stimulation (tDCS) on brain wave modulation. (2) The utilization of machine learning methods to detect and predict fear expression. (3) Characterize the genetic temporal alternation and stage-specific pathways and compare them with the non-pathological state of learned fears. Method: Using a modified single-prolonged stress and footshock (SPS&FS) rat model simulating PTSD, animals underwent a sequence of stressors: 2 hours of restraint, 20 minutes of forced swimming, diethyl ether exposure until unconsciousness, and a footshock upon awakening for fear conditioning. Brain wave patterns were measured during context re-exposure in the early (10 mins, 30 mins, 2,4, 6 hrs) and late phases (day 1, 3, 7, and 14) after extreme stress exposure. The local field potentials (LFPs) were recorded from the bilateral mPFC, the amygdala (AMY), and the ventral hippocampus (vHipp), and signals were further analyzed with machine learning methods. Lastly, RNA was extracted from the mPFC for transcriptome-level gene sequencing to compare genetic profiling between the SPS&FS model and a footshock fear conditioning model. Results: SPS&FS rats exhibited phenotypes simulating PTSD with higher anxiety, depression, and impaired fear extinction. Timedependent brain wave activities were observed, with delta activities correlating with fear levels. Among the machine learning methods, the bagged trees method was superior in detecting freezing behavior across all brain regions. And we found the AMY showed the highest accuracy in predicting freezing behavior. Gene sequencing revealed temporal differences in expression patterns between the two models. Discussion & Conclusion: The studies highlight the significance of brain wave measurements in PTSD research, emphasizing their role as a key indicator of fear expression and tracking temporal changes. Machine learning, especially the bagged trees method combined with AMY signals, offers promise in detecting and predicting fear behaviors in the SPS&FS model. Additionally, genetic profiling reveals stage-specific pathways in the SPS&FS model, distinguishing it from non-pathological learned fears. These findings pave the way for a more nuanced understanding of PTSD and potential intervention strategies. [ABSTRACT FROM AUTHOR] |
