Keywords

1 Introduction

Even if only a minority of trauma-exposed individuals develop posttraumatic stress disorder (PTSD) [1], it is one of the most prevalent psychiatric diseases worldwide [2]. The prevalence of PTSD increases with increasing numbers of lifetime traumatic events [3, 4]. In the US general population, lifetime PTSD prevalence is 6.8% [5], in Germany 2.8% [6] and in The Netherlands 7.4% [7]. Core symptoms of PTSD include aversive re-experiencing of traumatic events, avoidance anxiety, nervous hyperarousal, and emotional numbing [8]. PTSD significantly reduces the patients’ quality of life and increases their unemployment, mortality, morbidity, and suicidality rates [9, 10]. The fact that PTSD was not officially recognized as a diagnosis earlier than 1980 [11] probably contributes to the persistent deficit in drugs targeting PTSD-specific symptoms. Undoubtedly, there is an urgent need for new treatments, since only 60% of PTSD patients respond to the first line drug treatment of PTSD [12], i.e. to serotonin reuptake inhibitors (SSRI). Furthermore, a substantial number of patients suffering from PTSD do not benefit from exposure-based interventions [13], the current gold standard for PTSD psychotherapy [14].

Aiming to overcome this unsatisfactory situation in PTSD treatment, large numbers of preclinical and clinical studies have been performed. To date, it is widely accepted that genetic polymorphisms predispose some individuals to trauma-mediated changes in the dynamic epigenome [15, 16] and probably also in the miRNome [17,18,19]. These trauma-elicited miRNomic and epigenetic changes can in turn alter the expression of distinct proteins and consequently lead to disturbances in organ function, e.g. endocrine dysfunction and amygdala overactivity. Thus, an individual’s trauma and stress tolerance level depends on the interplay of environmental and predefined biological factors, in other words on gene–environment interaction (G × E). Of course, there is no single gene or polymorphism that conveys strong risk for PTSD; instead, psychiatric diseases are, in general, considered to be multigenic [20,21,22].

Besides the function and regulation of various brain regions such as the amygdala, the hippocampus, and the prefrontal cortex (PFC), the two major stress hormone systems, namely the hypothalamic–pituitary–adrenal (HPA) axis and the sympathetic nervous system (SNS) have repeatedly been shown to be altered in PTSD [23,24,25, 26]. A meta-analysis indicates that the ventromedial PFC (vmPFC) is the most consistently reported hypoactive, and the amygdala the most consistently reported hyperactive brain region in PTSD [27]. The vmPFC putatively fails to constrain the amygdala thereby leading to, among other symptoms, increased fear responses and impaired extinction of trauma reminders in PTSD patients [23]. Dysfunction of the hippocampus, which, like the vmPFC [23], was repeatedly reported to be smaller in various populations of PTSD patients [28, 29], may contribute to the PTSD-associated impairment in the recognition of safe contexts and memory deficits for neutral stimuli [23, 30]. Animal studies suggest that this PTSD-associated hippocampal volume loss might be mediated by a trauma-elicited transient increase in glucocorticoids [31].

Glucocorticoid homeostasis is regulated by the HPA axis which is widely, but not unequivocally [32], accepted to be hypofunctional in PTSD [33]. The inconsistent findings on HPA axis function in PTSD patients may be explained by different HPA axis responder types which were recently identified in a population of female PTSD patients [34]: HPA axis responder and non-responder PTSD patients differed in the prevalence of combined adult and early life trauma (ELT), the intensity of trauma-related dissociative symptoms, as well as in post-stress expression levels of peripheral FK506 binding protein 51 (FKBP51) [34], an inhibitor of glucocorticoid receptor (GR) signaling [35]. FKBP51 and its gene FKBP5 are well-characterized candidate molecules for affective disorders such as major depression, anxiety disorders and PTSD [24, 36, 37]. Genetic polymorphisms in FKBP5 [37, 38] but, interestingly, not in the gene of its target GR [39], are associated with PTSD. In mice, null mutation of Fkbp5 significantly improved endocrine and behavioral stress coping [40] and, in parallel, alleviated the stress-induced loss of hippocampal synapsin [41], thereby probably counteracting stress-induced hippocampal shrinkage [42].

In contrast to HPA axis function, SNS function was found to be elevated in PTSD [43]. Peripheral and cerebrospinal fluid (CSF) levels of norepinephrine, a major effector hormone of the SNS, are increased in PTSD patients [44,45,46] and correlate positively with PTSD symptom severity [43]. Consistent with these findings, several studies showed that both alpha- and beta-adrenoceptor blockers are effective in the treatment of PTSD-associated psychopathology [47, 48]. Besides the SNS, the HPA axis and various brain regions, a variety of other systems and networks such as neurotransmitters and neuropeptides have been extensively studied for their role in PTSD [49], among them the serotonin and the dopamine systems [49] as well as anxiolytic neuropeptides [50, 51]. In addition to molecular and fMRI studies, a plethora of (neuro)psychological studies has been performed with PTSD patients. These studies revealed that PTSD patients suffer, inter alia, from deficits in attention [52], regulation of the stress response [53], emotional processing [54, 55], and cognition [56, 57].

2 The RDoC System

It is disappointing that, although urgently needed, no novel drug treatment has yet sprung from this increased knowledge of the biological and psychological mechanisms of PTSD. Since this problem is the same with any other psychiatric disorder, it is compellingly logical that there is either a lack of implementation of psychiatric research findings into clinical practice and/or that the design of the studies hitherto performed is not suitable for drug development. The latter was suggested by many authors and resulted in the proposition of novel research concepts for psychiatry, for instance in symptom-based approaches [58,59,60] as well as in the Research Domain Criteria (RDoC) concept framed by the US National Institute of Mental Health (NIMH) [61].

Symptom-based approaches suggest assembling patient study cohorts according to major complaints rather than according to diagnoses classified in the two leading psychiatric classification manuals, i.e. the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) [8] and the 10th edition of the International Classification of Diseases (ICD10).[62]. The RDoC project instead proposes to integrate neuroscientific findings with research in psychopathology in order to identify neurobiological and behavioral dimensions across the current disease categories [63]. These dimensions were created to promote the establishment of a novel biologically informed psychiatric nosology [63, 64], thereby addressing the “comorbidity problem” (p.9) of DSM-5 and ICD10 [65].

RDoC defines psychiatric disorders as pathobiological conditions that involve brain circuits implicated in specific domains of behavior, cognition, and emotion. RDoC does not concentrate on pathological conditions, instead its dimensions cover the range from pathological to non-pathological. The two-dimensional matrix of the RDoC framework comprises five research domains which are proposed to be analyzed with seven units of analysis [66] that are all weighted equally. The following five research domains were defined so far: (1) negative valence systems (fear, anxiety, loss, frustrative nonreward), (2) positive valence systems (reward learning, reward valuation, habits), (3) cognitive systems (attention, perception, declarative memory, working memory, cognitive control), (4) systems for social processes (attachment formation, social communication, perception of self, perception of others), and (5) arousal/modulatory systems (arousal, circadian rhythm, sleep and wakefulness) [66]. The RDoC framework proposes to analyze these research domains by taking the following units of analysis into account: genes, molecules, cells, neural circuits, physiology, behaviors, and self-reports [66].

There are several manuscripts proposing a reconceptualization of a variety of categorical psychiatric diagnoses and of psychiatric symptoms according to the RDoC system, for example of the diagnoses major depression [67], panic disorder [68], and schizophrenia [69] and of the symptoms inhibitory control [70], auditory hallucinations [70], and pediatric disinhibited eating [71]. Other research teams aim to optimize the RDoC framework by suggesting novel research domains such as the domain of social cognition [72].

3 Integration of the RDoC System into PTSD Research

3.1 State of the Literature

We have found three manuscripts on RDoC and PTSD [73,74,75]. In their narrative review, Montalvo-Ortiz and colleagues welcomed the RDoC approach for psychiatric genetics and epigenetics since genes are linked to distinct phenotypes and behaviors rather than to multifaceted syndromes such as DSM-5 or ICD10 diagnoses [74]. Another excellent narrative review on PTSD pathobiology was provided by Gerald Young. He suggested to enrich the RDoC framework by adding further candidate endophenotypes [75]. Bauer and colleagues proposed the psychophysiological posterior probability score (PPPS), a composite measure of psychophysiological reactivity to script-driven imagery, as a disease marker for PTSD and categorized it as the “sort of quantifiable physiological or biological trait” (p. 1042) defined by the RDoC system [73].

To the best of our knowledge, there is so far no manuscript integrating PTSD research into the RDoC framework. In the following, we provide a narrative review of neurobiological findings and concomitantly analyze the relevance of the RDoC concept to PTSD research. A summary of our synthesis is provided in Table 1. We conclude this paper with suggestions for future directions in the application of the RDoC framework to PTSD research.

Table 1 Integration of findings on PTSD pathopsychobiology in the RDoC framework
Full size table

3.2 Negative Valence Systems

The relevance of negative valence systems to PTSD is supported by multiple lines of research. PTSD patients report and show fear and anxiety symptoms, in particular anxious avoidance of trauma-related cues [8]. This specific avoidance anxiety tends to generalize as neutral trauma-unrelated environmental cues can turn into trigger cues when the stimuli occur during a flashback or intrusion. The core physiological mechanisms related to avoidance anxiety are fear conditioning and fear extinction. Fear conditioning in PTSD is biased toward stimuli with higher emotional intensity than the original conditioned-fear stimulus [76]. The overgeneralized fear usually becomes harmful even though the learning of fear is an evolutionarily beneficial response mechanism [77]. Extinction deficits and cue generalization appears to be predominantly associated with PTSD symptoms but not general anxiety or depression, suggesting this is an important dimension for understanding PTSD pathology [78]. The neural circuits underlying fear conditioning and extinction and their pathological alterations have been extensively studied. The basolateral amygdala (BLA) is widely accepted as the main neural structure in which information of unconditioned and conditioned stimuli are integrated [79]. Impairment of the function of the BLA [80] and of the prelimbic division of the medial PFC (mPFC) disrupts the learning of fear [79]. Upon presentation of an extinguished cue during extinction training, the hippocampus activates the infralimbic cortex that stimulates inhibitory interneurons in the BLA which, in turn, prevent conditioned responding by inhibiting the output neurons in the central amygdala [79, 81]. In PTSD, the vmPFC is hypoactive and presumably fails to constrain the amygdala, thereby leading to increased fear responses and impaired extinction of trauma reminders [23, 28]. Furthermore, presumably, upon remembering and reconsolidating the trauma, the hippocampus fails to utilize environmental contextual cues to signal safety [23, 82]. Impairments in fear expression and extinction have been widely accepted to be central to PTSD pathobiology and thus have been analyzed in much more detail than given here – for review see [83]. Although the two major stress hormone systems, the HPA axis and the SNS, are of undisputed importance in psychiatric disorders, their roles in fear extinction, and in particular in PTSD-associated fear extinction learning deficits, have not yet been fully investigated [83].

The amygdala, the PFC, and the hippocampus harbor high concentrations of brain derived neurotrophic factor (BDNF), a protein implicated in synaptic plasticity [84]. Carriers of a distinct BDNF polymorphism, the BDNF val66met-allele, were shown to exhibit impaired extinction learning together with an elevated activity of the vmPFC and the amygdala during extinction trials [83, 84]. Accordingly, in comparison to val66val carriers, PTSD patients carrying the val66met-allele showed a poorer response to (fear extinction based) exposure therapy [85]. Besides BDNF val66met, polymorphisms, genes encoding for the catechol-O-methyltransferase (COMT) and for the serotonin transporter have also been associated with impaired fear extinction learning in PTSD patients [83].

A number of animal and clinical studies aimed at establishing a pharmacotherapy for PTSD-associated deficits in memory reconsolidation and fear extinction learning and suggested D-cycloserine as a potential therapeutic option [86, 87]. However, reports of the efficacy of D-cycloserine augmentation of fear extinction on PTSD psychotherapy revealed inconsistent results [88,89,90]. Other drugs or potential drugs, such as neuropeptide S, which has been shown to enhance the effects of D-cycloserine [87], L-DOPA and the indole alkaloid yohimbine [91], and the neuropeptide oxytocin [92, 93] have been suggested as augmentative treatment for PTSD psychotherapy. Since the extinction-optimizing drugs hitherto tested are far from being recommendable therapeutic options, future research should place particular emphasis on bridging this development gap. Although the neural circuits and molecular pathways of memory reconsolidation, fear learning, and extinction deficits associated with pathological anxiety are already quite well understood, this level of understanding does not yet appear sufficient for systematic drug development. One potential challenge to current study designs for pharmacotherapeutic enhancement of exposure therapy is that cognitive enhancers may strengthen contextual learning during extinction, limiting generalization of recall, thus treatments may need to be administered in multiple settings or during in vivo exposure to support general symptom reduction [94, 95].

3.3 Positive Valence Systems

In relation to negative valence systems, positive valence systems such as reward learning and reward valuation are understudied topics in PTSD research. The first systematic review on reward processing in PTSD was published in 2015 [96]. Reward processing is thought to underlie anhedonia [96], a symptom which is not pathognomonic for PTSD since it can be observed in a variety of other psychiatric disorders as well, predominantly in major depression [97]. Anhedonia is related to the symptom of emotional numbing [97] which is more characteristic for PTSD patients and thus belongs to the DSM-5 core symptoms of PTSD [8]. Both anhedonia and emotional numbing appear to reflect emotional inexpressiveness and insensitivity to emotional stimulation [97]. However, in accordance with the hypothesis of a bivariate regulation of aversion and appetition [98], emotional numbing reflects a diminished goal-oriented behavior “in response to incentives and positive stimuli at the cognitive-experiential level” (p.464) while anhedonia reflects a loss of pleasure and interest in previously pleasurable activities [97]. As, to the best of our knowledge and to our surprise, the neurobiology of emotional numbing is still elusive, we concentrate here on integrating anhedonia into the RDoC framework.

Reward processing comprises two steps, namely reward motivation (“wanting”) and reward consumption (“liking”) [96]. The anticipation of a reward stimulates a feeling of desire (“wanting”) which motivates a behavior directed at consumption of the reward. In detail, reward cues stimulate the meso-corticolimbic reward pathway leading finally, through dopaminergic activation of the motor cortices and the dorsal striatum, to an approach behavior. This mesolimbic reward pathway comprises a bundle of dopaminergic fibers originating from the ventral tegmental area (VTA). These fibers innervate limbic structures including the BLA, the mPFC, and the nucleus accumbens [99]. This mesolimbic pathway is also, at least partly, involved in the sense of pleasure in response to a reward (“liking”). However, this reward consumption-associated feeling of joy is mainly mediated by GABA and opioid receptors [96]. Both motivational and consummatory anhedonia have been reported in PTSD patients [96]. PTSD-associated reward deficits were reported more often in female PTSD patients and in studies analyzing social stimuli [96].

The molecular pathology of PTSD-associated impairments in reward processing is still elusive. Studies on this topic are rare. However, one of the few reported an increase in striatal dopamine transporter (DAT) availability in PTSD patients [100]. Elevation of DAT availability leads to a reduction in dopaminergic transmission and was hence suggested to underlie deficits in motivational reward processing [96]. Accordingly, dopamine agonists were proposed to overcome PTSD-associated impairments in reward processing [101]. Selective serotonin reuptake inhibitors (SSRIs) are known to enhance striatal function [96] and lead to full remission in about a third of PTSD patients [12]. Thus, in SSRI-sensitive PTSD patients, SSRIs may exert their therapeutic effects, at least in part, through enhancement of motivational reward processing [96, 102]. Besides SSRIs, the anxiolytic neuropeptide oxytocin has been recently shown to augment the sensitivity of the reward pathway during reward anticipation in PTSD patients versus trauma-exposed controls [103]. This finding is in accordance with the fact that oxytocin is proposed as a potential cognitive enhancer in PTSD treatment [91, 104].

3.4 Cognitive Systems

Cognitive systems are undoubtedly affected in PTSD. PTSD patients were repeatedly reported to exhibit deficits in attention [52] and planning [105] as well as in declarative (explicit) [106] and working memory [107]. Moreover, memory deficits, in particular impairments in fear extinction memory, also play a role in the above-discussed pathobiology of avoidance anxiety (see section “Negative Valence Systems”). PTSD patients are overengaged in scanning for potential environmental threats. They have an attentional bias [52] and a memory bias [108, 109] towards threat and negative stimuli at the expense of other cognitive processes [105]. The amygdala, the dorsal anterior cingulate cortex (ACC), the insula, and possibly also the vmPFC (mixed findings) were reported to be active in PTSD patients during tasks of negative attention [105].

PTSD-associated cognition deficits are promoted by negative emotionality through interactions between the amygdala and the hippocampus [105]. Most studies agree on the presence of amygdala overactivity in PTSD [27] whereas reports on hippocampal activity are mixed, possibly due to the fact that some studies employed general negative stimuli while others used trauma-specific cues [105]. The latter can induce false memories and a reduction in the activity of the hippocampus [105].

In contrast to explicit memory, which mediates the encoding and recall of facts [110], implicit or nondeclarative memory refers to the unconscious recall of encoded items. One facet of implicit memory is repetition priming – it refers to a bias or facilitation in retrieval of an encoded stimulus due to prior processing of a related or the same stimulus. “Repetition priming is perceptual when it reflects prior processing of stimulus form” and “conceptual when it reflects prior processing of stimulus meaning” (p. 494, [110]). Many, but not all, studies on this topic have reported that PTSD is associated with an increase in perceptional priming [111]. Together with diminished fear extinction and enhanced conditioning (see section “Negative Valence Systems”), the increase in perceptual priming of threat cues might be a powerful etiological combination in maintenance and pathogenesis of PTSD. Cognitive deficits in PTSD patients were reported to improve through successful PTSD treatment [105].

In comparison to fMRI studies, studies on the molecular basis of memory deficits in PTSD are scarce. One of the few studies on this topic found that increased methylation of the promoter of the gene encoding for the GR was linked to PTSD risk in genocide survivors as well as to reduced picture recognition in healthy men [112]. In fear consolidation, the most intensely studied epigenetic mechanism is histone acetylation. Animal models revealed that drugs that block histone acetylation (histone acetyltransferase (HAT) inhibitors) disrupt fear consolidation whereas the prevention of histone deacetylation by histone deacetylase (HDAC) inhibitors was found to increase it [113]. Epigenetic modifications in fear consolidation and extinction have been excellently reviewed by [113] and seem to be promising drug targets for PTSD, at least for PTSD-associated avoidance and possibly also for the aversive recall of traumatic memories.

3.5 Arousal Systems

Nervous hyperarousal belongs to the PTSD core symptoms [8] and shows up inter alia in enhanced nervousness, sleeping problems including nightmares and enhanced jumpiness. It is broadly accepted that SNS overdrive plays a core etiological role in PTSD, in particular in PTSD-associated hyperarousal [23, 24] and may be a risk factor for developing PTSD [114]. Both animal and clinical studies strongly suggest that the major effector hormones of the SNS, adrenaline and noradrenaline, enhance memory storage and that, consequently, excessive SNS activity at the time of trauma exposure might foster the consolidation of traumatic memory thereby promoting it to become intrusive [115]. A deletion variant of the gene encoding the α2B adrenergic receptor (ADRA2B) was reported to be linked to enhanced emotional memory both in survivors of the Rwandan genocide survivors and in healthy Swiss control cohort [116]. A polymorphism of another adrenoceptor, β2-adrenergic receptor (ADRB2), was found to be associated with PTSD both in male European Americans and in female African Americans [117]. In the latter study, the polymorphism in ADRB2 interacted with childhood adversity to predict adult PTSD symptoms. In a recent review article, PTSD patients were reported to have elevated peripheral and cerebrospinal fluid (CSF) noradrenaline and adrenaline levels [48]. Accordingly, adrenoreceptor blockers such as the α-1 adrenoreceptor blocker prazosin [118] and the beta-blocker propranolol [47] have been found by several authors to be effective in PTSD treatment. However, promoting a systemic attenuation of SNS activity can bring various side effects such as arterial hypotonia and loss in motivational drive.

In the brain, adrenergic transmission is regulated by the locus coeruleus (LC)-noradrenaline arousal system [119]. The LC is the principal site for brain synthesis of norepinephrine [120]; it mediates arousal and primes neurons to stimulus activation in widespread central regions such as the cerebellum, the hypothalamus, the thalamic relay nuclei, and the amygdala [120]. The LC was shown to mediate cognition through arousal [120]. Enhanced noradrenergic postsynaptic responsiveness, in particular in the circuit spanning from the LC to the BLA, was suggested as a major factor in the pathophysiology of PTSD and other stress-related disorders [121].

3.6 Systems for Social Processes

Social processes such as attachment formation, social communication and perception of self and others are clearly affected in PTSD patients, especially in patients with complex PTSD and in those having suffered an interpersonal trauma. However, as the body of literature on this topic comprises mainly psychological experiments and data, we did not review or summarize it here in this paper, which is focusing on neurobiological findings. However, we suggest putting particular emphasis on the study of the concepts of shame, guilt, and paranoid distrust since all of them are particularly frequent in interpersonally traumatized PTSD patients.

4 Proposed Novel Domains for PTSD Research

Table 1 demonstrates that many core findings on PTSD vulnerability and pathogenesis can be easily integrated into the RDoC framework. Hence, the RDoC system is unquestionably useful for PTSD research. However, there are some facets of PTSD pathobiology that do not easily fit into the proposed RDoC research domains, in particular PTSD-associated impairments in emotion processing and dissociative symptoms. For this reason, we propose two novel RDoC domains for PTSD research, the “emotional processes” and the “maintenance of consciousness” domains that we describe in the following and outline in Table 2.

Table 2 Suggested novel domains for PTSD research
Full size table

4.1 Stress and Emotion Regulation

In our eyes, stress and emotion regulation neither fits properly in the RDoC domain “negative valence systems” nor in any other of the RDoC research domains suggested so far. Accordingly, del Rio-Casanova and colleagues also stated that “emotion regulation should be considered a core domain when constructing clinical phenotypes in trauma spectrum disorders” [122].

Gross and colleagues defined stress and emotion regulation as “processes by which individuals influence which emotions they have, when they have them, and how they experience and express these emotions” [123]. However, definitions vary due to inadequate distinctions of character traits versus cognitive distortions and automatic versus voluntary processes [122].

A substantial body of literature documents the role of stress and emotion regulation in PTSD, complex PTSD, and borderline personality disorder (BPS) [124, 125]. BPS is seen as a trauma spectrum disorder by many researchers due to its comorbidity with PTSD and the high prevalence of traumatic events in the early life of BPS patients [126]. Moreover, BPS has a significant overlap with complex PTSD, but is, nevertheless, considered a separate diagnostic entity [127]. Here we will therefore concentrate on emotion regulation in PTSD which was repeatedly found to be dysfunctional in various facets [125].

The PFC plays a key role in the central nervous emotion regulation network [128]. This network was repeatedly described to comprise a dorsal and ventral subpathway. The dorsal pathway includes brain regions such as the lateral parietal cortex and the dorsolateral prefrontal cortex (dlPFC), and is involved in executive control of emotions. The ventral pathway includes the mPFC, the amygdala, and the ventrolateral PFC (vlPFC) and is mainly involved in in processing of emotions [122]. Both subpathways are key for the cognitive control of emotions and “compete for attentional resources” [122]. Upon evaluation of the emotional component of a stimulus, the activity of the orbitofrontal cortex and the amygdala increases, whereas the activity of the right vlPFC decreases. Conversely, upon linguistic labeling of the same stimulus, limbic area activity decreases [122]. Hence, emotion regulation processing can be understood as the result of an opposition between limbic and prefrontal areas.

One study demonstrated that PTSD patients, in relation to trauma-exposed controls, exhibit increased connectivity of the prefrontal/parietal region with the subgenual cingulate [122]. Findings on mPFC activity in PTSD are mixed. However, studies agree on the presence of amygdala hyperactivity in PTSD [23] and suggest PTSD to be associated with an impaired top-down-attentional control of emotions [122]. In general, in comparison to BPS, studies researching functional and molecular underpinnings of emotion regulation deficits in PTSD are scarce. To the best of our knowledge, there are so far no studies on the genetic or epigenetic basis of emotional regulation dysfunction in PTSD. However, according to the RDoC system, pathways should no longer be studied solely in relation to traditional diagnoses such as BPS and PTSD – it might well be that the molecular alterations associated with emotional dysregulation in BPS also underlie emotional dysfunction in PTSD. Interestingly, in a cohort of patients suffering from chronic PTSD, impairments in the regulation of emotions improved in response to prolonged exposure-based psychotherapy as well as to treatment with the SSRI sertraline [129].

As stress and emotion regulation not only play a role in PTSD, complex PTSD and BPS but also in a variety of other psychiatric diseases, for instance in major depression, eating disorders [77], and bipolar disorder, in concurrence with Fernandez and colleagues [130], we suggest to add this novel domain to the RDoC framework not only for PTSD research, but also for psychiatric research in general.

4.2 Maintenance of Consciousness

The aversive recall of traumatic memories is pathognomonic for PTSD and always goes along with dissociation and usually also with a change in the state of consciousness. An intrusion is a brief recollection of traumatic memories while a flashback can last for hours and flow into various forms of dissociation, for instance into psychogenic non-epileptic seizures (PNES). However, PTSD is not the only psychiatric disorder associated with dissociative symptoms. Derealization and depersonalization, which belong also to the spectrum of dissociative symptoms [131], occur for instance during panic attacks. Hence, the novel domain “maintenance of consciousness” might not be relevant to PTSD research only, but for various psychiatric disorders. During the experience of dissociation there is a change in consciousness [132] which is the result of, but cannot be fully explained by, cognitive processes. Thus, in our eyes, the research domain “cognitive processes” covers only parts of the phenomenon of psychological dissociation.

There are various definitions for psychological dissociation. According to DSM-IV, it is “a disruption in the usually integrated functions of consciousness, memory, identity, or perception of the environment” [133]. Among other reasons, the fact that both frequency and intensity of dissociative symptoms differ significantly among PTSD patients has motivated the definition of the dissociative subtype of PTSD [134]. The assessment of the functional and molecular underpinnings of dissociative phenomena such as PNES is still in its early stages [135] – defining maintenance of consciousness as a novel research domain in the RDoC system might hopefully remedy this gap of knowledge and promote substantial neurobiological research on this important topic, in particular on dissociative symptoms. Studies on the drug treatment of dissociative symptoms are also scarce. Naltrexone, a partial opiate antagonist, is one of the few drugs tested regarding its efficacy in the treatment of dissociative symptoms [136]. Currently, a clinical study tests cognitive behavioral therapy versus standardized medical care in adults suffering from PNES in a multicenter randomized controlled trial (RCT) protocol [137]. To the best of our knowledge, there is so far no RCT that has assessed the effect of any drug treatment on dissociative symptoms. With regard to the suffering of patients experiencing pathological dissociation this is a significant omission.

Thirty percent of PTSD patients have a blunted heart rate response to trauma narrative exposure [138]. Accordingly, in a Trier Social Stress Test (TSST) paradigm, one group of PTSD patients exhibited a blunted HPA reactivity together with an increased prevalence of trauma-related dissociative symptoms (HPA non-responder PTSD patients) [34]. These non-responder patients showed alterations in the peripheral expression levels on the mineralocorticoid receptor (MR) and of FKBP5 [34]. These studies suggest that a blunted reactivity of the HPA axis and of the SNS might be associated with the propensity for dissociative symptoms in PTSD patients. Lanius and colleagues state that only patients suffering from the dissociative subtype of PTSD show a frequent and severe overactivity of the prefrontal area with consecutive hypersuppression of limbic regions upon exposure to trauma-narrative scripts (emotional overmodulation) [138]. Emotional under- and overmodulation is present in all PTSD patients at different time-points [138]. A study by Felmingham and colleagues [139] and studies on dissociative amnesia support this corticolimbic model of dissociation [138].

In Table 2, the various brain networks and regions that have been implicated in PNES pathobiology and hence probably play a role in other dissociative symptoms also are summarized from the review by Perez and colleagues [135]. The fact that the networks and regions mentioned in that review comprise the majority of brain areas might hint at a lack of specificity of the results and again stresses the urgent need for further studies on the neurobiology of the maintenance of consciousness, in particular of psychological dissociation.

5 Summary and Conclusions

In summary, decades of PTSD research have considerably advanced our understanding of PTSD pathobiology (see, e.g., [140, 141]). However, none of the potential biomarkers and none of the proposed drugs, except for prazosin, or drug targets have yet progressed to clinical use [141]. Our synthesis reveals that several core findings in PTSD such as amygdala/BLA overactivity can be linked to all RDoC domains for PTSD research but lack further specification of their exact role in the pathways associated with these domains (Tables 1 and 2). This circumstance indicates that the cellular and molecular processes finally decisive for regulation of psychic processes and hence for the expression of psychopathological symptoms have not yet been identified.

The RDoC framework was conceptualized to overcome the translational gap in psychiatric research by detaching it from the concept of traditional psychiatric diagnoses. The currently described integration of important neurobiological findings of PTSD research into the RDoC system put fear processing, reward functioning, explicit/implicit memory pathways and the SNS in the spotlight of PTSD vulnerability and pathogenesis (Table 1) and, furthermore, revealed that PTSD-associated emotional instability and dissociative symptoms are not adequately represented in current RDoC domains. For this reason, we suggest two novel domains, i.e. the domains “maintenance of consciousness” and “stress and emotion regulation” (Table 2) – the latter has recently also been suggested by others to be indispensable for general psychiatric research [130]. Integrating PTSD research findings into these two novel domains revealed large gaps of knowledge in the associated units of analysis “cells,” “molecules,” and “genes” (Table 2), i.e. in the molecular and cellular processes underlying the regulation of stress and emotion and in the maintenance of consciousness. We hope that the gaps of knowledge in PTSD pathobiology identified here stimulate studies aiming to close them – such studies will certainly profit from the adoption of the RDoC principle.