Genetic risk mechanisms of posttraumatic stress disorder in the human brain
To circumvent such confounds, a viable alternative approach would be to initiate studies with a molecular biomarker that is tractable in that it is unaffected by treatment and substance abuse, such as a genetic sequence variant. Recently, genome‐wide association studies (GWAS) in psychiatry have identified genetic risk variants in schizophrenia (Ripke et al., 2013; Schizophrenia Working Group of the Psychiatric Genomics Consortium, 2014); autism (Weiss et al., 2009); MDD (Ripke et al., 2013; CONVERGE Consortium, 2015); and PTSD (Logue et al., 2015). Evidence reveals that a genetic risk variant can modulate the expression of transcripts within a genetic locus (GTEx Consortium, 2013, 2015; Ramasamy et al., 2014), thereby potentially altering phenotypes that form the mechanisms underlying disease. Therefore, deriving quantitative associations between genotype and gene expression (expression quantitative trait loci [eQTL] associations) in normal controls with minimal confounding influences of treatment and/or substance abuse would be the first important step in elucidating mechanisms by which genetic variants increase risk for disease. Once the risk‐specific transcripts are identified, we can then proceed to evaluate and better understand the mechanisms of clinical risk in diseased subjects.
Insofar as PTSD is a human brain disorder, such an approach is likely best implemented in postmortem human brain tissue. PTSD is a complex psychiatric disorder involving multiple brain regions including the amygdala, hippocampus, and prefrontal cortex (PFC) (Grillon et al., 1996; Rauch et al., 2000; Liberzon and Sripada, 2008; Cisler et al., 2014). The amygdala has long been known to play a critical role in recurrent fearful memory‐associated stress (Bremner et al., 2008; Cisler et al., 2014), with the hippocampus and ventromedial PFC also being associated with stress‐related changes in neuroanatomy and physiology (Admon et al., 2013; Li et al., 2014). Evidence from recent studies also implicates dorsolateral prefrontal cortex (DLPFC) dysfunction in PTSD. One study reported a significantly increased DLPFC response to emotional anticipation in women diagnosed with PTSD (Aupperle et al., 2012). Another study showed significantly decreased DLPFC efficiency during a working memory task in subjects with PTSD (Honzel et al., 2014), adding to evidence for DLPFC dysfunction in PTSD. Similar studies reporting significant deficiencies in executive cognitive function and attention domains further substantiate DLPFC dysfunction in PTSD (Leskin and White, 2007; Pannu et al., 2009). Similarly, evidence from functional neuroimaging studies also implicates hippocampal and medial prefrontal dysfunction in the neurobiology of PTSD (Nutt and Malizia, 2004; Shin et al., 2006). Therefore, a thorough characterization of the posttraumatic influence on the transcriptome using RNA sequencing technology in multiple brain regions, including the amygdala, hippocampus, and DLPFC and medial PFC, would be both timely and potentially insightful for understanding PTSD biology.
To begin with this approach, we analyzed normal brain DLPFC (n = 237, Lieber Institute for Brain Development [LIBD], Table Ia) and amygdalae (n = 67, Gene‐Tissue Expression (GTEx) Consortium, Table Ib) RNA sequencing data to identify eQTL associations for a set of published PTSD risk variants I.