Biological studies of post-traumatic stress disorder

Abstract

Post-traumatic stress disorder (PTSD) is the only major mental disorder for which a cause is considered to be known: that is, an event that involves threat to the physical integrity of oneself or others and induces a response of intense fear, helplessness or horror. Although PTSD is still largely regarded as a psychological phenomenon, over the past three decades the growth of the biological PTSD literature has been explosive, and thousands of references now exist. Ultimately, the impact of an environmental event, such as a psychological trauma, must be understood at organic, cellular and molecular levels. This Review attempts to present the current state of this understanding on the basis of psychophysiological, structural and functional neuroimaging, and endocrinological, genetic and molecular biological studies in humans and in animal models.

Figure 1. Assessing structural abnormalities in PTSD using a combat-discordant identical-twin design

Sample coronal structural magnetic resonance images of right (red) and left (blue) hippocampi in a twin pair consisting of a combat veteran with PTSD (1) and his combat-unexposed co-twin (2), who has no PTSD but is considered “high risk” because his identical twin developed PTSD when exposed to trauma; as well as a control twin pair consisting of a combat-exposed veteran without PTSD (3) and his “low-risk” combat-unexposed co-twin (4), who also has no PTSD. Contrast A provides a replication test of studies demonstrating smaller hippocampal volume in combat veterans with vs. without PTSD. Two additional contrasts can shed light on whether this abnormality is a result of combat exposure or of having PTSD, or whether it represents a pre-existing vulnerability factor. Contrast B compares hippocampal volumes in combat-exposed PTSD veterans with their own high-risk co-twins. If the twin with PTSD (1) has smaller hippocampal volume than his co-twin (2), the trait has likely been acquired. Contrast C compares hippocampal volumes in high vs. low-risk co-twins. If the trauma unexposed, non-PTSD twin of the veteran with PTSD (2) has smaller hippocampal volume than the unexposed, non-PTSD twin of the veteran without PTSD (4), it is likely that the trait represents a pre-existing vulnerability factor. This type of design can also be used to assess the origin of other abnormalities observed in patients with PTSD.

Figure 2. Brain regions implicated in PTSD functional neuroimaging studies

The amygdala (shown in Panels A and C) is involved in recognizing both conditioned and unconditioned stimuli signaling danger, as well as in expressing the fear response. Amygdala reactivity is exaggerated in individuals with PTSD and is positively correlated with symptom severity. The insular cortex (Panel A) and dorsal anterior cingulate cortex (Panel B) are also hyperreactive in PTSD; these structures may modulate (in these cases enhance) the amygdala’s expression of fear. In contrast, activation in the ventral medial prefrontal cortex (Panel B), which also modulates (in this case reduces) the amygdala’s expression of the fear response, is diminished in PTSD; vmPFC activity is also negatively correlated with symptom severity. Functional neuroimaging findings in the hippocampus (Panel C), which is involved in recognizing both safe and dangerous contexts, have been mixed in PTSD, with both hypo- and hyperreactivity observed.

Figure 3. Contribution of prefrontal regions to fear regulation and expression

A. Activation in the human brain during fear conditioning and extinction can be investigated in a Pavlovian fear conditioning and extinction paradigm. During conditioning, a conditioned stimulus (a colored light) is paired with a mild electric shock to the fingers in a particular context (context A). The acquisition of conditioned fear in this paradigm can be measured by the skin conductance response to the light. During extinction learning, the light is subsequently repeatedly presented without the shock in a different context (context B). This leads to extinction of the conditioned response. The next day, the light is again presented in the absence of the shock in Context B (extinction recall), and then again in Context A (fear renewal). Greater retention of extinction learning is associated with a lower fear response during extinction recall. Extinction retention is context-dependent: the hippocampus is thought to recognize whether a context is safe (B) or dangerous (A) and to communicate this information to other structures in the fear network. B. fMRI studies have shown that activation of ventromedial prefrontal cortex (vmPFC, shown in green) during extinction recall is positively correlated with extinction retention, whereas activation of dorsal anterior cingulate cortex (dACC, shown in red) is negatively correlated with extinction retention. Ventromedial PFC sends excitatory glutamatergic projections to gamma-aminobutyric acid (GABA) -ergic intercalated cells (ITC) in amygdala, which in turn inhibit the expression of the fear response by the amygdala’s central nucleus (Ce). In contrast, dACC sends excitatory glutamatergic projections to amygdala’s basolateral nucleus, which in turns activates the expression of the fear response by Ce. LA=lateral amygdala nucleus. C. Non-PTSD subjects (left) have a relatively high vmPFC/dACC activation ratio during extinction recall, which tips the balance toward better extinction retention and less fear expression. By contrast, PTSD subjects (right) have a lower vmPFC/dACC activation ratio, which tips the balance toward less extinction retention and more fear expression.

Figure 4. Putative brain-state relevant to PTSD

Panel A (resilience): The response of either a) previously nontraumatized individuals exposed to mild or moderately arousing unconditioned threat stimuli, b) resilient individuals who are resistant to the arousing effects of more extreme unconditioned threat, or c) individuals with PTSD after undergoing extinction and recovery so that conditioned threat stimuli are no longer highly arousing. Panel B (risk): The response of a) previously nontraumatized individuals exposed to unconditioned threat that is highly arousing, b) individuals with PTSD who are re-exposed to trauma-related cues (conditioned threat) prior to extinction and recovery, or c) individuals with PTSD who are resistant to recovery. A. Neuromodulation contributing to relative prefrontal cortical dominance (resilience). Mild to moderately arousing sensory stimuli activate the central nucleus (CE) of the amygdala, which projects both directly and indirectly to brainstem monoaminergic cell body regions to activate mesocorticolimbic dopamine (DA) pathways emanating from the ventral tegmental area (VTA), as well as more widely disseminating NE and serotonergic (5-HT) pathways emanating from the locus coeruleus (LC) and median/dorsal raphe nuclei, respectively. In the prefrontal cortex (PFC), the resulting mild to moderate increases in synaptic levels of these monoamines engage high-affinity (e.g., noradrenergic alpha-2) receptors to enhance working memory and activate glutamatergic pyramidal output neurons that project back to the amygdala. There, glutamatergic activation of GABAergic interneurons in the basolateral (BLA) and/or intercalated nuclei suppresses associative learning and inhibits excitatory BLA pyramidal cell projections to the CE and the expression of the species-specific defense response (see below). B. Neuromodulation contributing to relative amygdala dominance (risk). Strongly arousing unconditioned threat stimuli (due to objective threat characteristics or an individual’s increased sensitivity to objective threat) or conditioned threat stimuli in persons with PTSD activate the amygdala to a greater degree, which in turn excites brainstem mesocorticolimbic monoaminergic neurons more vigorously^, —a process likely facilitated by reductions in GABAergic neurotransmission within the amygdala.^, In the PFC, higher synaptic levels of these monoamines engage low-affinity noradrenergic alpha-1, as well as DA1 and 5HT2 receptors, resulting in working memory impairment and a reduction in PFC inhibition of amygdala. Consequent lifting of the PFC “brake” reduces GABAergic tone within the BLA and intercalated nuclei of the amygdala to enable associative pairing of unconditioned threat stimuli (US) and convergent contextual stimuli (CS) or later reconsolidation of conditioned stimuli, as well as activation of the species-specific defense response (SSDR(, which includes increases in blood pressure and heart rate, HPA axis activation, engagement in reflexive defensive behaviors (fight, flight, freezing), and restriction of high-level information processing to enable efficient focus on survival-relevant phenomena. Direct catecholamine effects in the amygdala also facilitate defensive responding: activation of D1 receptors on PFC projection neuron terminals inhibits glutamate activation of GABAergic interneurons, whereas activation of post-synaptic D2 receptors on BLA to CE pyramidal projection neurons increases their excitation by convergent US-CS inputs, as well as adventitious sensory stimuli. This likely facilitates or maintains associative fear conditioning and may contribute to generalization of fear. In contrast, neuronal activity in the BLA is generally suppressed by NE alpha-2 and alpha-1 receptor stimulation, though the latter effect, mediated by enhanced terminal release of GABA, is reduced by chronic stress. In contrast NE beta-receptor activation is excitatory and enhances US-CS pairing.

Figure 5. Behavioral and physiological changes in a PTSD animal model

A. A prototypic rodent model of PTSD: (single prolonged stress (SPS). B. Behavioral effects of exposure to SPS include increased arousal as reflected in startle response (left) and decreased extinction retention in contexts that are either consistent (con) or inconsistent (incon) with the extinction context (right). C. Physiological effects of exposure to SPS include enhanced glucocorticoid receptor (GR) expression in the hippocampus (Hipp) and frontal cortex (FC) and negative feedback in the HPA axis (left), decreased excitatory tone in the medial prefrontal cortex (mPFC) (middle), and decreases in tonic as well as increases in phasic responses to stimulation (arrow) of locus coeruleus (LC) neurons (right). Note: data are illustrative.