The gene expression approach has provided promising insights into the pathophysiology of posttraumatic stress disorder (PTSD). However, few studies used hypothesis-free transcriptome-wide approach to comprehensively understand gene expression underpinning PTSD. Researchers at Stony Brook University conducted a transcriptome-wide expression study using RNA sequencing of whole blood in 324 World Trade Center responders (201 with never, 81 current, 42 past PTSD). Samples from current and never PTSD reponders were randomly split to form discovery (N = 195) and replication (N = 87) cohorts. Differentially expressed genes were used in pathway analysis and to create a polygenic expression score. There were 448 differentially expressed genes in the discovery cohort, of which 99 remained significant in the replication cohort, including FKBP5, which was found to be up-regulated in current PTSD regardless of the genotypes. Several enriched biological pathways were found, including glucocorticoid receptor signaling and immunity-related pathways, but these pathways did not survive FDR correction. The polygenic expression score computed by aggregating 30 differentially expressed genes using the elastic net algorithm achieved sensitivity/specificity of 0.917/0.508, respectively for identifying current PTSD in the replication cohort. Polygenic scores were similar in current and past PTSD, with both groups scoring higher than trauma-exposed controls without any history of PTSD.
a. Receiver operating characteristic curve of the predicted PTSD probability on the replication cohort, trained using the 448 genes significant at FDR 0.05 on the discovery cohort via the elastic net prediction algorithm. b. Boxplot comparing the predicted risk score for current, past and never PTSD in the replication cohort
Together with the pathway analysis results, these findings point to HPA-axis and immune dysregulation as key biological processes underpinning PTSD. A novel polygenic expression aggregate that differentiates PTSD patients from trauma-exposed controls might be a useful screening tool for research and clinical practice, if replicated in other populations.