Post-traumatic stress disorder (PTSD) is caused by the experience of a traumatic event and is associated with four clusters of symptoms, including intrusive phenomena, avoidance, changes in mood and cognition, as well as hyper-arousal (1). While most individuals have experienced at least one traumatic event during their lifetime (2), the prevalence of the disorder in the general population is approximately 9–12% (3). This raises the question of why some trauma-exposed individuals develop PTSD and others don’t. Genetic factors can confer an increased vulnerability to PTSD following trauma exposure (4, 5). While other risk factors that contribute to the development of the disorder (6), epigenetics may provide a framework for understanding how gene expression is influenced by traumatic experiences to produce individual differences in the prevalence of PTSD. Considering that PTSD treatment outcome may also be quite variable (7, 8), our knowledge of these mechanisms may potentially serve as biomarkers of treatment response in individuals diagnosed with the disorder. In this review, we introduce some general mechanisms of genomic transcription and provide an overview of studies that have established epigenetic associations with PTSD. Assuming that these associations may not only be relevant to the development of the disorder but also potentially contribute as mediators of treatment response in patients, we review human epigenetic findings pertaining to PTSD treatment. We also discuss some of the challenges and future directions for epigenetic research in this field.

Epigenetics refers to the chemical changes to chromatin that control genomic transcription. These changes in gene expression occur through alterations in chromatin structure rather than changes in DNA sequence per se (9). DNA methylation, histone modification and non-coding RNA modification are the three most common epigenetic mechanisms. Epigenetic patterns are responsible for cell-type specific gene expression patterns (10), which bestow cellular identity to DNA. They are formed during development by a highly regulated and orderly process.

DNA methylation has been the most frequently investigated epigenetic mechanism. The covalent methylation modification of the DNA (11), catalyzed by DNA methyltransferases (DNMT), at the 5′ position of the cytosine ring, occurs without changing the DNA sequence. DNA methylation is involved in gene regulation primarily by inhibiting promoter activity, either by interfering with binding of transcription factors or by recruiting methylated DNA binding proteins (MBD) that recruit histone deacetylases (HDACs) and histone methyltransferases (HMTase) that induce chromatin inactivation (12, 13).

Histone modification includes acetylation, methylation, phosphorylation, ubiquitination and sumoylation. The most studied histone modification is acetylation. Acetylation results from the effect of the enzyme histone acetyltransferases (HATs), causing a loss in the histone’s chromatin structure and promoting transcription (14). Histone methylation is a highly complex modification process. Depending on the position of methylation and the number of methyl groups transferred to the histone tail, it can stimulate or repress transcription. Histone phosphorylation is involved in transcriptional activation (15), but this modification process is less studied than the two others.

Published studies in the last decade have provided evidence that epigenetic patterns can be altered in response to environmental stimuli (16–18), leading to long-term alterations in gene activity by regulating gene expression, which further plays a critical role in disease susceptibility, etiology, and progression. The study of epigenetic mechanisms that emerge out of traumatic experiences is part of this rapidly growing scientific literature. A number of human-based epigenetic studies of PTSD have been published (19–22), with the majority of them focusing on candidate genes which are chosen based on animal models or genetic association findings, such as NR3C1 (23, 24), FKBP5 (25, 26), SLC6A3 (27), and BDNF (28). For instance, BDNF (brain-derived neurotrophic factor) has been associated with memory, stress, and neuropsychiatric disorders. In a predator stress model of PTSD in rodents (30) the methylation of Bdnf exon-IV was specifically modified in hippocampus, whereas no changes were found in the prefrontal cortex or amygdala (28). The related physiological systems involved in PTSD have included the hypothalamic-pituitary-adrenal (HPA) axis, immune function, serotonergic system, catecholaminergic system and others [see (19) for a comprehensive review]. Unbiased epigenome-wide association approaches have also been employed in PTSD research. We will briefly review main findings of candidate gene and epigenome-wide research as it pertains to PTSD.

Extensive animal and human studies have been conducted to determine how altered HPA axis functioning contributes to the development and maintenance of PTSD (31). In rodent studies, significant disruptions were observed in the functioning of the HPA axis, including reduced basal glucocorticoid levels and increased dexamethasone-induced inhibition of cortisol levels (29, 32, 33). Some (but not all) PTSD patients have been reported to have lower levels of the stress hormone cortisol (34–38)–a finding attributed to glucocorticoid receptor (GR) hypersensitivity, as well as an increased negative feedback inhibition of the HPA axis (39, 40). HPA axis-related genes have been widely studied in animals (31, 41–43) and significant epigenetic changes that occur in conjunction with PTSD have been reported in humans (23, 24). The HPA-axis related gene NR3C1, which encodes GR, is a candidate gene most extensively studied in stress and PTSD research (44–46). As shown in a recent review (47) a growing body of studies on early life adversity in both rodents and humans have reported increased methylation levels of the exon 1F promoter at the NR3C1 (analog of Nr3c1 1₇ in rodents). However, evidence from PTSD research suggests that there may be lower DNA methylation levels at this gene. For instance, Yehuda et al. (24) compared a cohort of combat veterans diagnosed with PTSD to a sample of trauma-exposed individuals without PTSD and found that the veterans with PTSD displayed lower DNA methylation levels of the NR3C1 exon 1F promoter from their peripheral blood samples (24). Likewise, Labonté et al. (23) reported lower T-cell levels of NR3C1 exon 1B and 1C and higher GR expression in civilians with PTSD compared to participants with current or remitted PTSD.

Although previous studies have targeted candidate genes in PTSD research, large-scale, unbiased epigenome-wide association approaches are increasingly used to search for methylation transformations across a variety of newly discovered genes. Epigenome-wide studies have focused primarily on DNA methylation. There has been a number of such studies investigating DNA methylation profiling in PTSD and reviews have summarized these findings (19, 20, 48). Uddin et al. (49) first identified methylation profiles underlying immune system changes associated with lifetime PTSD. This study not only revealed differences among PTSD-affected individuals with respect to the number of uniquely methylated genes, but also indicated that uniquely unmethylated genes were associated with a strong signature of immune system involvement.

Another group (50) working on immune system and DNA methylation reported differentially methylated CpGs in five genes (TPR, CLEC9A, APC5, ANXA2, and TLR8) associated with PTSD (50). Furthermore, in a PTSD study on epigenome-wide gene expression and DNA methylation profiles, eight odorant/olfactory receptor-associated genes, as well as genes related to immunological activation, the Gamma-Aminobutyric Acid A (GABAA) receptor, and vitamin D synthesis, were found to be up-regulated with PTSD (51). In a meta-analysis of PTSD epigenome-wide association studies in three trauma-exposed civilian cohorts, the NRG1 and HGS were identified as blood-based biomarkers associated with PTSD (52). Moreover, a DNA methylation study of PTSD in World Trade Center first-responders (53) identified a set of novel genes (ZDHHC11, CSMD2, COL9A3, PDCD6IP, TBC1D24, and FAM164A) associated with PTSD. Moreover, the gene BRSK1, LCN8, NFG, DOCK2 (54) and ZFP57, RNF39, HIST1H2APS2 (55) were found to be linked to the severity of PTSD symptoms.

Several comprehensive reviews describing epigenetic mechanisms and PTSD with the candidate gene and epigenome-wide approaches have been recently published (19, 20, 48). A detailed review of these studies is beyond the scope of the current paper.

Although research has identified epigenetic markers in PTSD, only a few studies have explored whether these markers affect treatment response (see Table 1).

Yehuda et al. (56) were the first to investigate the role of epigenetics in psychotherapy interventions. Their study involved a small sample of veterans with PTSD receiving a 12-week prolonged exposure (PE) psychotherapy intervention. Blood samples were taken at pre-treatment, after 12 weeks of psychotherapy and at a 3-month follow-up visit. Blood-based DNA methylation at the NR3C1 promoter region was found to predict positive treatment response, (higher DNA methylation assessed at pre-treatment was associated with lower PTSD symptoms at follow-up). In contrast, the FK506 binding protein 5 (FKBP5) promoter did not predict treatment response, while decreased DNA methylation was found to be associated with PTSD recovery. Moreover, higher gene expression of FKBP5 at post-treatment was observed in treatment responders. The authors suggested that these epigenetic markers may be associated with prognosis (NR3C1 methylation) and symptom severity (FKBP5 methylation). This study represents an important initial first step in establishing relevant molecular markers for PTSD therapies.

Church et al. (57) conducted a pilot randomized controlled trial in which emotional freedom techniques (EFT) were applied to treat veterans suffering from PTSD. The method incorporates elements exposure and cognitive therapies, as well as somatic stimulation, including acupressure (58). Participants received ten one-hour long sections of EFT and blood samples were drawn before and after treatment. The authors examined a panel of 92 gene expressions using blood RNA before and after treatment. The pre- vs. post-treatment comparisons revealed six significant differential gene expressions (IL-10RB, SELL, TNFAIP6, CXCR3, IL-18, and IFITM1) associated with treatment response to EFT.

Using mindfulness-based stress reduction (MBSR) techniques (i.e., meditation and yoga) to treat PTSD in veterans, Bishop et al. (59) used a case-control study design that involved psychoeducation, followed by 7 weekly 2.5 h group-therapy sessions and a 6.5 h retreat, for a total of nine sessions spanning a 9 week period. Methylation analyses were based on blood draws obtained at post-treatment. The methylation of two candidate genes, serotonin transporter (SLC6A4) and FKBP5, which are involved in monoamine and HPA axis function (60–63), were examined in treatment responders and non-responders. The authors reported a time-by-responder group interaction in which responders exhibited an increase in FKBP5 methylation following treatment, with non-responders showing a decrease in methylation after treatment.

Vinkers et al. (64) investigated epigenome-wide DNA methylation changes associated with symptomatic remission after psychosocial interventions that included trauma-focused cognitive behavioral therapy (tf-CBT) and eye movement desensitization and reprocessing (EMDR) within military participants with PTSD. Participants received 8–10 treatment sessions and underwent genotyping and methylation quantification 6–8 months after treatment. The findings revealed 12 differentially methylated regions (DMRs) in participants with reduced PTSD symptoms following treatment. Furthermore, the authors examined whether these DMRs were associated with the development of PTSD in an independent prospective cohort with deployment-related PTSD. Interestingly, the methylation of the Zinc finger protein 57 (ZFP57), a transcriptional regulator of genomic imprinting, was found to be related to PTSD symptom reduction in this cohort.

Recently, Carleial et al. (65) investigated methylome-wide changes in saliva samples of 84 female former child soldiers from Eastern Congo having a DSM-5 diagnosis of PTSD related to experiences of abduction and other traumatic events. The participants received either Narrative Exposure Therapy (NET) or treatment-as-usual. The NET group received six individual sessions of 90–120 min each in length. The saliva samples were collected at baseline and at a 6-month follow-up visit. DNA methylation of different CpGs were associated with treatment response, and methylated positions (DMPs) were related to four genes (ALCAM, RIPOR2, AFAP1, and MOCOS). Treatment-related DMPs were also found to be involved in memory formation and in biochemical pathways that are typically affected by trauma-related disorders. Furthermore, the findings were partially replicated in a separate group of 53 female former child soldiers.

Xulu et al. (66) applied NET as a forensic offender rehabilitation (FORNET) intervention in order to investigate DNA methylation changes associated PTSD treatment response and appetitive aggression symptoms among South African men. Participants with chronic trauma exposure received one of three interventions: FORNET, CBT or no intervention (waitlist). The FORNET intervention involved 82-h sessions that comprised psychoeducation. Saliva samples were collected at baseline, 8- and 16-months following treatment. DNA methylation of promoter regions were analyzed for a disease-focused gene panel (22 genes). The study revealed that increased gene methylation involved in dopaminergic neurotransmission (NR4A2) and synaptic plasticity (AUTS2) was linked to a reduction in PTSD symptoms in the FORNET group.

As a pharmacological treatment, a novel CRF1 receptor antagonist GSK561679 has been shown to be effective in the treatment of PTSD (67, 68). Based on these findings, Pape et al. (69) tested the hypothesis that DNA methylation could serve as a potential marker for this treatment. The authors analyzed the association between blood-based DNA methylation levels of CRHR1, NR3C1, and FKBP5 with treatment response in a cohort of women with a history of childhood abuse, as well as in a subgroup of these participants who showed high CRF activity. The treatment protocol involved a 6-week double-blind treatment design and blood draws were taken for genotyping after 5 weeks of study treatment. The findings showed significant differences in CRHR1 methylation levels following GSK561679 treatment in the subgroup with high CRF activity and NR3C1 methylation at baseline interacted with child abuse status (none-to-mild abuse vs. moderate-to-severe abuse) in predicting PTSD symptom change after GSK561679 treatment.

More recently, Yehuda’s team (70) combined prolonged-exposure (PE) psychotherapy (10 weekly sessions) with a pharmacological drug (hydrocortisone) (taken 45 min prior to each PE session; 8 total) to explore epigenetic markers involved in the treatment of military deployment PTSD through epigenome-wide analyses in a longitudinal study. The CREB–BDNF signaling pathway, which is involved in fear learning and memory formation (71), was identified as a marker that predicted both recovery and symptom change, based on blood samples taken at 3 months post-treatment. FKBP5, NR3C1, SDK1, and MAD1L1 were also found to be associated with PTSD recovery, especially decreased FKBP5 was associated with PTSD symptom improvement.

In this narrative review, we provided a brief overview of key concepts in epigenetic research as they pertain to the study of PTSD. While genetic predispositions may interact with traumatic experiences to account for individual differences stress response (21, 130), experience induced epigenetic changes may also result in brain modifications that give rise to behavioral manifestations of PTSD (20). These changes may represent susceptibility factors for the development of PTSD, arise from traumatic experiences themselves, or represent markers of untreated symptom change over time. Because of the widespread use of accessible peripheral tissues in human PTSD research, it will be important to identify peripheral epigenetic markers that can characterize individuals as being particularly susceptible or resistant to developing PTSD. However, alternative post-mortem and neuroimaging methods need to be given further consideration as complementary methods of investigation.

In addition, animal models of PTSD are hoped to shed light on the cellular and molecular pathways that underpin PTSD in humans. However, research has thus far revealed few common epigenetic markers across animal and human studies. There may obviously be potential constraints that influence the degree to which the results derived from animal models may be translated to humans. These include differences in the heterogeneity of the stressors (ex., conditioned stimuli, such as shocks, vs. the myriad of traumatic events that humans experience), complexity of the behavioral stress responses, temporality of symptom onset (ex., stress responses in rodent models normally occur relatively rapidly while PTSD symptoms in humans may be delayed), probability of developing stress reactions following a stressor or traumatic experience (ex., most humans do not develop PTSD following a traumatic event), as well as effects of life history. These differences may need to be given more consideration in animal research if the non-human models are to aid us in our understanding of the human cellular and molecular mechanisms of PTSD.

From a neurological perspective, investigating the underlying epigenetic mechanisms involved in PTSD treatment may hold promise for identifying epigenetic markers of therapy. Although epigenetic markers have been associated with treatment-related changes, it remains to be empirically determined whether there are common epigenetic mechanisms that underly PTSD susceptibility, diagnosis, and treatment change. Moreover, the clinical studies (56, 57, 59) discussed in this review indicate that there may be several epigenetic biomarkers involved in treatment response in patients suffering from PTSD. However, due to the varying methodologies employed, it is difficult to draw consistent conclusions from published literature regarding how epigenetic biomarkers relate to specific therapeutic (psychotherapy or pharmacotherapy) processes of change. We suggest that future clinical trials target specific components of interventions, such as exposure-based therapies, coping strategies, cognitive processing/restructuring (131), as well as reconsolidation-based techniques, to establish their underlying epigenetic mechanisms more firmly using either candidate gene or epigenome-wide approach.

LC-L and JF drafted an earlier version of the manuscript. LC-L and DS prepared the final version for submission. AB reviewed the final version. All authors listed on this manuscript has seen and approved this submission.