Abstract
The high rates of traumatic brain injury (TBI) and post-traumatic stress disorder (PTSD) diagnoses encountered in recent years by the United States Veterans Affairs Healthcare System have increased public awareness and research investigation into these conditions. In this review, we analyze the neural mechanisms underlying the TBI/PTSD comorbidity. TBI and PTSD present with common neuropsychiatric symptoms including anxiety, irritability, insomnia, personality changes, and memory problems, and this overlap complicates diagnostic differentiation. Interestingly, both TBI and PTSD can be produced by overlapping pathophysiological changes that disrupt neural connections termed the “connectome.” The neural disruptions shared by PTSD and TBI and the comorbid condition include asymmetrical white matter tract abnormalities and gray matter changes in the basolateral amygdala, hippocampus, and prefrontal cortex. These neural circuitry dysfunctions result in behavioral changes that include executive function and memory impairments, fear retention, fear extinction deficiencies, and other disturbances. Pathophysiological etiologies can be identified using experimental models of TBI, such as fluid percussion or blast injuries, and for PTSD, using models of fear conditioning, retention, and extinction. In both TBI and PTSD, there are discernible signs of neuroinflammation, excitotoxicity, and oxidative damage. These disturbances produce neuronal death and degeneration, axonal injury, and dendritic spine dysregulation and changes in neuronal morphology. In laboratory studies, various forms of pharmacological or psychological treatments are capable of reversing these detrimental processes and promoting axonal repair, dendritic remodeling, and neurocircuitry reorganization, resulting in behavioral and cognitive functional enhancements. Based on these mechanisms, novel neurorestorative therapeutics using anti-inflammatory, antioxidant, and anticonvulsant agents may promote better outcomes for comorbid TBI and PTSD.
Introduction
T
Post-traumatic stress disorder (PTSD) is caused by physical or emotional trauma that produces many signs and symptoms that overlap with TBI. 2 Like TBI, PTSD is a brain disease produced by white and gray matter damage via stress-related pathologies of neuroinflammation, oxidative damage, and excitotoxicity. PTSD develops after a person has experienced a highly disturbing event in which serious physical harm had occurred or was threatened; for example, after seeing someone injured or killed, or after sexual or physical assault, an accident, combat, or a natural disaster. PTSD often develops within a few months of the trauma and can continue for months or even years. During PTSD, psychological and/or physical trauma is followed by intrusion symptoms related to the event including disturbing memories, nightmares, dissociative reactions such as flashbacks, distress, and physical reactions at exposure to cues or contextual reminders of the trauma. Further, there is persistent avoidance of reminders of the trauma including people, places, and activities associated with it. Likewise, there are often cognitive changes such as amnesia and distortions to important trauma-related events such as negative beliefs about oneself and others, guilt, emotional disconnection from others, and loss of interest. PTSD also produces heightened arousal and reactivity such as irritability, self-destructive behaviors, hypervigilance, startle, inattention, and insomnia. To summarize, this review focuses on trauma-related pathologies and neurobiological underpinnings of TBI and PTSD and their comorbid condition in veterans and active duty military. It also describes characteristics of animal models, to study these pathologies and investigate new treatments for this comorbid disorder.
PTSD and mTBI have discernible overlaps in symptomatology which makes it difficult for a clinician to distinguish between these two conditions. For a majority of individuals with mTBI, symptoms resolve in days to months; however, a significant number (∼ 15%) are estimated to have post-concussive syndrome 1 year post-injury. 1 Both mTBI and PTSD present with nonspecific symptoms such as generalized anxiety, irritability, insomnia, and memory problems, which further complicate diagnostic differentiation. 3 Somatic symptoms such as headache, hypersensitivity to light and sound, and dizziness are more specific to mTBI syndrome. Efforts to determine the diagnostic and neurobiological specificity of mTBI and PTSD, and their comorbid condition, are critical for understanding and effectively treating these post-traumatic disorders.
There has been a further enhanced interest in understanding and treating the comorbidity of TBI and PTSD because of its frequent occurrence in soldiers returning from Operation Enduring Freedom (OEF) in Afghanistan and Operation Iraqi Freedom (OIF) and Operation New Dawn (OND) in Iraq. The Armed Forces Health Surveillance reported that >1,800,000 service members were deployed in these conflicts, and that more than a third of these individuals were deployed at least twice. 4 During OIF and OEF, both the use of improvised explosive devices and continuous levels of threat during these operations contributed to the comorbidity of these conditions. In a sample of United States Army soldiers, 9% reported symptoms of mTBI, and >17% of those with mTBI screened positive for PTSD, whereas >31% screened positive for depression. 5 In a large cohort of returning OIF soldiers with TBI, of those reporting loss of consciousness, 43.9% also met criteria for PTSD. 6 In a large study of active duty Marine and Navy servicepersons who served in OIF or OEF, TBI doubled the likelihood of post-deployment PTSD symptoms. 7 In a study of >100,000 OEF/OIF veterans, one quarter received mental health diagnoses and more than half of these had two or more mental health diagnoses. 8 TBI has been linked to the serious consequence of ongoing suicidal ideation, which has been an epidemic for OIF/OEF veterans. 9 Whereas onsets of acute neuropsychological problems following mTBI often occur within 1–3 months after injury, current guidelines emphasize that a clinical diagnosis of PTSD should not be made until 6 months after the traumatic exposure of the patient. This fact suggests that the time courses of the development and recovery of PTSD and their pathophysiology have not been well understood. The delay in diagnosis and ineffective treatments caused by the complex symptomatology and presentation of the comorbid disorder may contribute to this inconsistent course of recovery.
Comorbid TBI/PTSD can present with a wide variety of neuropsychiatric signs and symptoms. The individual may endure a loss of consciousness from the head trauma but may also experience more subtle alterations of consciousness (feeling dazed), amnesia, or a transient neurologic deficit such as headache, weakness, or sensory or gait problems. When TBI does not result in a loss of consciousness, neuropsychiatric recovery is likely, although enduring cognitive changes worsen that prognosis. 10 TBI is also associated with other neuropsychiatric sequelae including depression, mania, or psychosis; substance use disorders; sleep disorders; and chronic pain. These conditions could directly result from the traumatic injury or from an emotional effect of the TBI on daily functioning. 11 In another study, patients with comorbid TBI and PTSD demonstrated increased health risks and cognitive impairments when compared with patients without PTSD. 12 The comorbidity of TBI and PTSD is associated with increased rates of neuropsychiatric health problems that complicate recovery. 13
Both TBI and PTSD can be represented as brain disorders with disruptions in neural networks that communicate via long axonal pathways through white matter tracts, termed the “connectome.” 14 We propose that both mTBI and PTSD are produced by insults to the connectome underlying neural circuits that subserve executive function, attentional control, motivational, behavioral inhibition, and fear responses. In both TBI and PTSD, the pathological mechanisms occur via a wide range of neuroinflammatory changes, excitotoxicity, and oxidative changes and result in neurodegeneration. Our central hypothesis is that TBI and PTSD result from a variety of pathophysiological mechanisms, resulting in dysfunction in overlapping neural circuits, and some common symptomatology. Further, we propose that the comorbid condition of TBI/PTSD produces more severe health consequences and is more resistant to treatment. A better understanding of the pathophysiological mechanisms and network dysfunctions for this devastating comorbid condition will lead to novel neurorestorative therapeutics.
In this review, relevant articles published between 1990 and 2017 were identified by searching MEDLINE®, using the following MeSH search terms that were cross-referenced:
Neuroimaging and Functional Circuitry of TBI and PTSD
Neuroimaging of TBI in military/veteran populations
The neurocircuitry disruptions in TBI, PTSD, and the comorbid disorder have been evaluated from clinical biospecimens. There are a variety of modalities used to assess the damage to these important circuits in TBI and PTSD, including structural and functional neuroimaging. 15,16 These modalities include CT, structural and functional MRI (sMRI and fMRI, respectively), transcranial Doppler (TCD) (which measures vascular abnormalities), positron emission tomography (PET), single photon emission CT, arterial spin labeling (ASL) MRI, magnetic resonance spectroscopy (MRS), and electrophysiological techniques (magnetoencephalography [MEG] and electroencephalography [EEG]). CT or sMRI scanning have limited sensitivity to identify white matter lesions in TBI, and often have negative findings. Experts have concluded that CT or sMRI also have inadequate sensitivity for diffuse axonal injury (DAI) and associated vascular injuries. 15 Advanced MRI methods such as susceptibility weighted imaging (SWI) are able to detect DAI and are accessible in clinical practice. In one MRI study, chronic blast-induced mild and moderate TBI produced greater structural damage and cortical thinning in the right hemispheric insula, inferior temporal, and frontal lobes. 17 CT, sMRI, SWI, and TCD are the most accessible imaging tools in the clinical setting. No single imaging modality standard is sufficient for all patients, because of the heterogeneity of brain injury in TBI. 15 It is necessary to consider all these imaging techniques to gain a comprehensive visualization of the physical damage caused by TBI.
Diffusion tensor imaging (DTI) is a research tool utilized to identify DAI by measuring the diffusion properties of water to generate measures related to the structural integrity of specific white matter structures. DTI fractal anisotropy (FA) is used to measure neuronal fiber density, axonal diameter, and myelination in white matter. A lower FA measure suggests disruption of both white matter and neural circuitry connectivity. In a recent study of veteran cohorts from OIF, OEF, and/or OND, blast exposure effects on white matter tissue integrity using DTI was examined. 18 Interestingly, this study found a blast dose-response relationship between the number of blast exposures and the loss of white matter integrity. In addition, there was a negative relationship between FA and the length of time after injury, indicating greater disruption of white matter connectivity over time. In another DTI study of OIF/OEF/OND veterans exposed to a blast at close range (< 10 m), there was decreased connectivity of bilateral primary somatosensory and motor cortices compared with veterans with blast injury occurring beyond this distance. 19 Another DTI study compared soldiers with TBI with military controls and demonstrated white matter abnormalities in frontostriatal and frontolimbic circuits, fronto-parieto-occipital association tracts, brainstem and corpus callosum fibers. 20 In this study, there were asymmetries along the superior-inferior, anterior-posterior, and left and right hemispheric axes. This last study highlights the presence of diffuse and focal and asymmetrical white matter abnormalities in TBI. In another group of OIF/OEF veterans with TBI, lower FA was found in the corpus callosum and cingulum bundle, and this measure correlated with cognitive processing speed and executive functions. 21 Further, OIF/OEF veterans with blast injury and mTBI, and also a loss of consciousness, showed more spatially heterogeneous white matter abnormalities than veteran controls. 22 In this last study, decreasing FA was demonstrated in the left internal capsule as the number of blast exposures increased. In summary, DTI studies have shown that TBI results in asymmetrical cortical and subcortical tract abnormalities resulting in multiple circuit disruptions, and that DTI could be one of several important tools in assessing TBI in the clinical setting.
fMRI and electrophysiological studies using MEG have been performed in OIF/OEF veterans with TBI. An fMRI study was performed comparing blast-related TBI in a military population (OIF/OEF/OND veterans) to blunt force-induced TBI in civilians. 23 The study found that chronic TBI may cause abnormal patterns of brain activation to prevent mobilizing “response inhibition,” which describes a participant's capability to suppress an intended or already initiated manual action. It was concluded that such altered patterns of brain activation responsible for failure of executing “response inhibition” could be used to delineate military from civilian TBI. Other studies have examined MEG to detect neural changes in mTBI in OIF/OEF/OND veterans with blast injury. 24 This last study showed that abnormalities in the prefrontal cortex (PFC) were positively correlated with changes in mood lability, personality changes, concentration, and depression. It was hypothesized that MEG abnormalities may result from disconnection of multiple PFC circuits via axonal injury in TBI.
During the brain's resting state there are fluctuations in regional brain activity. The default mode network (DMN) involves discrete areas in the medial prefrontal, medial and lateral parietal, and medial and lateral temporal cortices. 25 Activation, deactivation, and connectivity within the DMN were shown to be altered in TBI and associated with cognitive impairment. 14 Patients with TBI have demonstrated sustained attention impairments that were associated with an increase in DMN activation, especially within the precuneus and posterior cingulate cortex. 26 These disruptions in DMN function in TBI were predicted by the amount of white matter damage in a tract connecting the right anterior insulae to the pre-supplementary motor area (SMA) and the dorsal anterior cingulate cortex. 27 In summary, these findings suggest that TBI produces DAI, as demonstrated by cortical and subcortical white matter disruptions, and alterations in DMN functional circuits associated with neurocognitive, as well as emotional and behavioral, deficits that are characteristic of this condition.
Neuroimaging of PTSD in military/veteran populations
Largely based on findings from basic science investigations, PTSD can be conceptualized as a cue- and context-associated fear conditioning (FC) process that is associated with amygdalar hyper-responsivity and prefrontal cortical impairments in fear inhibition. 28 It has been shown that fear-related sensory information is initially transmitted from cortical regions to the amygdala through the basolateral amygdala (BLA), which relays to the central nucleus of the amygdala (CeA) to activate fear responses through outputs to the hypothalamus and brainstem. By contrast, the infralimbic region within the PFC appears to be the primary candidate pathway to suppress fear responses via extinction learning. 29 The PFC inhibits the function of BLA by suppressing the conditioned fear responses after extinction training. 28 –30 Although executive functions are maintained by widely distributed pathways, one key circuit in both disorders is the fronto-cingulo-parietal cognitive control network. Important hubs in this network have been associated with various executive function domains such as the anterior cingulate cortex involving cognitive control, the dorsolateral PFC mediating working memory, the inferior frontal gyrus and (pre-) SMA regulating response inhibition, and the parietal lobes involving attention and its control. Disruptions to this fronto-cingulo-parietal network could result in impairment in cognitive control, memory, attention, and inhibition of fear processing, which are relevant to PTSD and TBI. 31
PTSD is associated with white matter changes and regional atrophy in the brain, resulting in emotional, behavioral, and functional deficits. Various subregions of the PFC, hippocampus, and amygdala that are involved in cognitive and behavioral regulation are important in the pathophysiology of PTSD. Neural changes in PTSD include a reduction in medial PFC gray matter volume, which may mediate behavioral disruption, personality changes, and impairments in cognitive control. 32,33 In an sMRI study, OIF/OEF veterans with more severe PTSD symptoms showed gray matter volume loss in the temporal cortex, including the inferior temporal and parahippocampal gyrus regions. 34 Additionally, gray matter morphology in trauma-exposed individuals was compared in patients with and without PTSD. 35 The results of this meta-analysis demonstrated that there were consistent areas of gray matter volume reduction in the cingulate cortex, ventromedial PFC, temporal regions, and hippocampus in PTSD patients. In another study of structural and functional imaging, veterans with PTSD underwent ASL perfusion using MRI and DTI, in order to measure regional cerebral blood flow abnormalities and structural changes. 36 ASL-MRI findings demonstrated that veterans with PTSD had increased perfusion in the right parietal and superior temporal cortices, suggesting compensatory responses to regional volume losses. Using DTI, these individuals with PTSD showed reductions in white matter regions of the PFC, anterior cingulate cortex, and angular gyrus. In summary, these studies highlight the roles of the frontal, parietal, and temporal lobes in the pathophysiology of PTSD.
In PTSD, there are also distortions in traumatic memories that relate to dysfunction in the hippocampus. 37 –40 Moreover, there have been findings of smaller hippocampal volumes in individuals with PTSD than in controls. 41 In male veterans with combat trauma, PTSD was associated with both smaller total hippocampal volumes and CA3/dentate gyrus subfield volumes, which may produce distorted contextual information. 40 There are also reports of hypothesized abnormalities in fear regulation from the BLA in PTSD. Following FC, veterans with PTSD (vs. combat controls) showed impairments in extinction learning and heightened amygdala activity as measured by fMRI in the safety context (i.e., the context in which extinction took place). 42 In contrast, in a fear danger context, these same veterans demonstrated lower activity in the amygdala than combat controls, suggesting dysregulation of fear in both safety and danger settings. In an fMRI study, decreases in ventral anterior cingulate cortical activation with repeated exposure to fearful stimuli predicted an increase in symptoms. 43 In summary, PTSD is associated with hyperactivity of the BLA producing fear responses, hypoactivity of the cingulate cortex resulting in mood regulation problems, and also smaller hippocampi associated with memory dysfunction.
PTSD severity was also associated with the extent of exposure to combat events, suggesting a potential dose-response relationship related to trauma. 44 In this last study of OEF/OIF combat veterans with PTSD, DTI and structural MRI and a variety of symptom scales were used. PTSD severity was associated with abnormal MRI findings and high FA on DTI. During the cognitive appraisal paradigm, the dorsolateral PFC appeared to be impaired in veterans with PTSD, who showed a lesser ability to activate this region, as measured by fMRI, than did combat controls. 45 Antidepressant treatment is known to enhance PFC function in PTSD. Veterans from OEF/OIF, with and without PTSD, were treated with the serotonin uptake inhibitor paroxetine, and underwent fMRI scans that were 12 weeks apart. An emotion regulation task was performed during each scan. Paroxetine treatment increased activation in both the left dorsolateral PFC and the SMA during emotion regulation and, therefore, was an effective treatment in this regard. 46 In summary, impairments in function and volumes in the PFC, temporal cortex, and hippocampal regions may produce cognitive and memory changes in PTSD. These deficits along with cingulate cortical dysfunction may result in decreased top-down control of the amygdala, leading to increased sensitivity to fearful stimuli. 28
Only a few studies have examined neuroimaging alterations in the comorbid condition of TBI and PTSD. Davenport and coworkers used DTI to assess blast-related TBI injury and post-deployment PTSD in OIF/OEF service members. 47 The diagnosis of PTSD (with or without comorbid TBI) was associated with higher general FA and greater integrity of white matter fibers. In another study, OIF/OEF military veterans with mTBI and comorbid PTSD and depression were compared with non-TBI participants using high-angular resolution diffusion imaging. 48 There was a loss of white matter integrity that was associated with mTBI in a distributed pattern of major fiber bundles and smaller tracts. The diffuse loss of white matter integrity appears to be a consistent mechanism of damage specifically produced by blast injury. Amen and colleagues conducted a single photon emission computed tomography (SPECT) study of patients with either PTSD, TBI, or both conditions, compared with controls. 49 The study revealed that subjects with PTSD (compared with those with TBI) exhibited greater activity in the PFC and temporal lobes, cingulum, basal ganglia, insula, thalamus, and limbic regions. One interpretation of the relatively greater activation of these regions in PTSD patients may relate to the need to maintain normal function in damaged circuits from the TBI, producing a different signature of injury from that of either condition alone.
Other studies of comorbid PTSD/TBI have shown an array of functional network disruptions. In subjects with comorbid mTBI and PTSD, fMRI imaging showed that decoupling of hippocampal and PFC circuits might produce the disturbances of traumatic memories. 50 Another finding in this study was that the caudate/putamen was less connected to the PFC. These two network decouplings in PTSD/TBI may interact to regulate trauma memories, and any disconnection could contribute to this comorbid condition. In a SPECT study of veterans with comorbid TBI and PTSD, the DMN is differentially affected in those with PTSD and those with TBI. 51 This study showed that the DMN is hyperperfused in individuals diagnosed with PTSD and hypoperfused in patients diagnosed with TBI, providing a distinction between the two conditions. Further, individuals with the comorbid condition demonstrate perfusion to these regions that appears intermediate between PTSD and TBI. In another study of OEF/OIF/OND veterans with comorbid PTSD and mTBI who underwent sMRI, there were reductions in volume in the bilateral anterior amygdala in the comorbid group, compared with veterans with no history of either condition. 52 Amygdala volume reductions were predictive of poorer inhibitory control of behavior. The comorbid condition of TBI/PTSD appears to have a unique signature of neuroimaging findings.
To understand how these various gray and white matter deficits in mTBI and PTSD may produce their clinical impact, it is important to recognize the role of the “connectome” in brain injury. This term relates to the mapping and organization of neural pathways throughout the brain. These highly centralized long and short pathways integrate brain signals through gateway hubs from one region to another. 14 Even a small structural or cellular lesion that affects a major hub can be very disruptive to the functioning of a neural network, and thereby can affect associated behaviors and cognition. Wolf and Koch 14 hypothesized that damage to axons in TBI disrupts the timing of neural signals within and between brain networks in cortical and hippocampal cognitive circuits and limbic emotional regions. Loss of white matter integrity in mTBI in major fiber bundles and in smaller tracts, and structural and functional deficits in gray matter in frontal cortical, cingulate, hippocampal, and parietal cortices may produce the behavioral and cognitive dysfunction in comorbid TBI/PTSD. Additionally, there are perfusion deficits that are intermediate between PTSD and TBI, which may account for the overlapping and mixed clinical presentation between the two conditions. 51
Translational Models in TBI, PTSD, and their Comorbidity
Using animal models enables better understanding of the pathophysiological mechanisms in TBI, PTSD and the comorbid condition. There is no single animal model for either TBI or PTSD that expresses the complexity of the human condition. However, widely utilized PTSD models, including fear conditioning/retention/extinction, inescapable stress, learned helplessness, prolonged immobilization, and others and have been well reviewed. 53 However, various stress or fear paradigms produce the behavioral and biological aspects of the PTSD phenotype including avoidant behavior, anxiety-like behavior, hyperarousal, cue- or context-induced fear responses, and alterations in brain or hypothalamic-pituitary-adrenal axis stress responses. FC and retention are often employed as an animal model for studying PTSD. FC has strong face validity in that fear cues and contexts become aversive and produce persistent fear behaviors, such as freezing and increased heartrate, startle, and avoidance, similar to PTSD symptoms. FC has good predictive validity as exposure/extinction treatments reduce clinical anxiety by exposing a patient to conditioned stimuli and extinguishing fear. The FC model has strong construct validity in that the theory of the disorder and the model converge through the formation of an association between a neutral and a threatening stimulus that produces cue and contextual fear responses. In the FC model, the fear cues and contexts recruit activation of a similar neurocircuitry, as seen in PTSD patients. 54
There are several well-utilized models for blast and blunt trauma injury in rodent models that enable the study of their pathophysiology. 55 Blast models use the application of pressure waves to approximate blast TBI in humans. A blunt trauma is produced in the controlled cortical impact (CCI) model, in which an air- or electromagnetic-driven piston is used to penetrate the brain. The rodent fluid percussion injury (FPI) model is the most extensively utilized animal model of TBI that produces focal and diffuse brain injury dependent on the intensity of force applied to a cortical site of injury. Blast injury, CCI, and FPI models have strong face value, as they replicate the injury and behaviors of the human condition. They provide good construct validity, because the theory of the brain disorder and the model converge where the neurocircuitry disruption occurs in a fashion similarly to in TBI patients.
Only a few animal studies have modeled co-occurring disorders of TBI and PTSD. In one model, the PTSD paradigm used a combination of exposures to predator odor, repeated restraint, and inescapable footshock, whereas the mTBI component was concurrent CCI. 56 This study showed that these interventions produced unique signature of behavioral, inflammatory, pathological, and biochemical responses at acute time points. Another CCI model in rats produced loss of hippocampal and amygdala neurons, and also increased conditioned fear and anxiety-like behaviors. 57 This research group further developed a TBI model using CCI and a PTSD model of social defeat that produced greater anxiety-related behavior and fear behaviors within extinction sessions, compared with TBI and PTSD-only controls. 58 Further development of such animal models is essential for a greater understanding of the neurobiology, pathophysiological mechanisms, and treatment approaches to comorbid TBI/PTSD.
Pathophysiological Bases of TBI and PTSD
Primary and secondary injuries after TBI
TBI is initiated by a series of events beginning with a primary injury from the direct trauma, and results in a disruption of brain tissue. 59 The primary injury is produced by an external force; for example, a blast wave, a body or cranial impact, a projectile, and/or rapid acceleration-deceleration within the cranium. The primary brain injury produces cortical or subcortical contusions and lacerations, white matter hemorrhages, intracranial bleeding (subarachnoid hemorrhage or subdural hematoma), venous engorgement, vascular space enlargement, edema, and blood–brain barrier disruption. 59 DAI is the signature injury of TBI, is often caused by the acceleration-deceleration and angular forces on the brain, and is produced by stretching and shearing of axons. Shearing forces from the blast injury primarily produce lesions in the deep frontal white matter and subcortical structures while the common tensile effects produce axonal stretching. 60 DAI has been hypothesized to be an interactive consequence of neuroinflammation and neurodegeneration, which is responsible for producing certain long-term complications of TBI. 61,62
TBI also produces cognitive and behavioral impairments through a process of secondary brain injuries to the critical neurocircuitry. Secondary molecular events follow primary injuries and represent metabolic, ischemic, cellular, oxidative, excitatory, and neuroinflammatory processes that produce neuronal injury and loss (see Fig. 1). These secondary injuries can occur within minutes, and can extend to months and perhaps years after the primary injury. 55 Additionally, these pathophysiological changes are associated with the synthesis and release of various neurochemicals that affect brain metabolism, cerebral blood flow, and ion homeostasis. 59,63 Mechanisms of neuronal and vascular damage include inflammation, calcium-mediated cell toxicity via proteolytic pathways, glutamate-mediated excitotoxicity, mitochondrial rupture, production of oxygen-free radicals, and release of apoptotic substances. 59,60 Notably, intra-axonal calcium increases activate the cysteine proteases, calpain and cathepsin, which produce protein degradation that damages the axon cytoskeleton. These changes further injure axons by producing swelling and disconnection and the signature pathology, DAI. Calpain in TBI and other neurodegenerative diseases is dysregulated, and produces neuronal injury and cell death, and, therefore, calpain inhibitors are possible neurotherapeutics for TBI. 64 These pathophysiological changes that produce axonal degeneration and disconnection and dendritic loss contribute to the associated cognitive and behavioral impairments. The apparent delays in TBI-induced disconnection via secondary injury mechanisms suggests that the process is amenable to therapeutic intervention. 60
Pathophysiological mechanisms leading to the brain circuit disruption and symptomatology of comorbid traumatic brain injury (TBI) and post-traumatic stress disorder (PTSD). Physical trauma may produce psychological trauma (broken line) and results in primary and secondary brain injuries. Primary injuries include brain contusions, hemorrhages, and vascular damage. Secondary injuries follow primary brain injuries and represent metabolic, ischemic, cellular, and molecular events, and produce neuronal injury. Secondary injury from TBI and from psychological trauma produces at least three major neuropathologies: oxidative stress, neuroinflammation, and glutamatergic excitotoxicity. These neural changes from physical and psychological trauma result in neurodegeneration, synaptic plasticity, dendritic remodeling, and neuronal injury and cell death in the prefrontal cortex (PFC), hippocampal, amygdalar, and other circuits. These changes in neural pathways mediating executive function, attentional and cognitive control, fear responses, and behavioral inhibition produce resultant neural dysfunction and symptoms characteristic of comorbid TBI and PTSD.
Neuroinflammatory pathologies in TBI
It has been established that neuroinflammation is a leading secondary injury mechanism following TBI. 65,66 Microglia are the innate immune cells of the CNS, and following TBI, these cells are activated. In addition, there is a breakdown of the blood–brain barrier, allowing further infiltration of neutrophils, macrophages, and lymphocytes into brain tissues. These inflammatory cells participate in phagocytosis and release pro- and/or anti-inflammatory cytokines and complement proteins that enclose foreign substances and limit local bleeding. As a result, there is an enhanced clearance of damaged cells and reduction of toxic effects accumulated by cellular degradation. 67 In addition to the well-known destructive and scavenging functions of the inflammatory process, recent studies have delineated a constructive role of neuroinflammatory cell responses in promoting brain tissue healing, restorative neuroplasticity, and cellular repair processes after TBI. 68 –70 Specifically, distinct microglia/macrophage subpopulations can be defined by unique surface antigens that correlate with different functions; that is, classically activated cells with cytotoxic and pro-inflammatory cytokine-releasing properties (M1 phenotype), alternative activated cells with pro-repair functions (M2A phenotype), immune regulatory cells (M2B phenotype), and deactivated cells (M2C phenotype). 71 The polarization continuum suggests that therapeutic strategies should aim to prevent over-polarization toward the M1 phenotype, but should enhance evolvement of the anti-inflammatory M2 phenotypes, to eventually install an optimal ratio of M1/M2 macrophages/microglia. Conversely, if TBI triggers chronically activated microglia, this causes damage to the neural cells and metabolism in the brain. 72,73 In a rat model of TBI, blood–brain barrier damage was confirmed by measuring immunoglobulin G (IgG) leakage, endothelial damage, activation of microglia (CD-68 expression), and reactive gliosis with heightened production of glial fibrillary acidic protein (GFAP). In a rat chronic TBI model (up to 3 months), myelin loss and microscopic bleeding that was co-localized with glial scars, and CD68 and IgG puncta staining of IgG and CD68, were observed. 74 In the last study, there were delayed increases in microvascular pathology after TBI that were associated with prolonged inflammation, blood–brain barrier disruption and progressive white matter degeneration. The cell surface marker of activated microglia, CD68, was also detected in the expanding cortical lesion 1 year after CCI in C57BL/6 mice. 75 It appears that in chronic TBI, there is long-term microglial activation that could be correlated with neural lesion expansion, continued neurodegeneration, and chronic demyelination.
TBI-initiated chronic activation of microglia has been shown to cause neural damage through the release of toxic substances such as pro-inflammatory cytokines, complement proteins, and proteases. 72 TBI generates time-dependent elevations of pro-inflammatory cytokines of interleukin (IL)-1β, IL-6, and tumor necrosis factor-alpha (TNFα). 76 –78 TNFα is involved in systemic inflammation, is produced by activated macrophage, and is highly relevant to TBI. A rodent model of FPI increased TNF-α mRNA expression at the site of injury in parietal cortex and its adjacent cortical regions, 1 and 6 h post-injury. 79 The increases in TNF-α mRNA and protein levels appeared to be dependent on the extent of injury in the ipsilateral cortex and hippocampus. 80 Complement proteins from neuroinflammatory processes can induce secondary neuropathology in TBI. Therefore, efforts to inhibit complements may reduce inflammation, neurodegeneration, and the resulting neurocircuitry disruptions. In experimental TBI, a novel anti-complement treatment was given after experimental TBI and applied to sites of injury. 81 Compared with placebo control treatment, this therapeutic approach reduced cortical neural cell membrane attack complexes, microglial proliferation, mitochondrial stress, cytokine production, and axonal damage, significantly improving neurological function.
Neuroinflammatory pathologies of PTSD
Neuroinflammation represents an important mechanism that causes neuronal damage and neurocircuitry disturbance in PTSD. Published literature shows evidence of neuroinflammatory responses in stress models and in PTSD models. Both acute and chronic stress increased microglial proliferation in the brain. 82 Stress-induced activation of microglial cells was produced in the hypothalamus, hippocampus, thalamus, substantia nigra, and central gray areas. 83 Moreover, chronic stress increased the number of activated microglia in multiple stress-sensitive brain regions possibly by augmenting their proliferation. 84 Predator stress exposure heightened levels of pro-inflammatory cytokine and NALP3 inflammasome proteins in the hippocampus and PFC. 85 Early life stressors were shown to sensitize microglial cells to induce exaggerated responses to stress later, which may contribute to the development of PTSD in the adulthood. 86 Another study reported that immobilization stress increased hypothalamic IL-1β mRNA expression in rats. 87
Interleukins can promote neurodegeneration, disruption of neural networks, and consequent behavior disturbances, and have been investigated in stress and PTSD models. In a rat PTSD model, interleukin-beta (IL-1β) mRNA and protein levels were elevated in the hippocampus, PFC, and amygdala. 85 Intermittent footshock in rats caused changes in IL-6 in the hippocampus and cortex along with marked surges of IL-1β quantity in the hypothalamus. 88 Interestingly, psychosocial stressors also elevated IL-1β expression in the hypothalamus. 89 Conversely, IL-1β receptor knockout mice demonstrated a reduction of anxiety-related behaviors. 90 Clinically, cerebrospinal fluid (CSF) levels of IL-6 were found to be increased in veterans diagnosed with PTSD after returning from combat. 91 Further, compared with healthy controls, spontaneous production of IL-6 in white blood cells was detected in patients with PTSD and was correlated with PTSD symptom severity. 92 Lastly, levels of plasma IL-6 and norepinephrine were positively correlated in the PTSD group but not in controls. 91
In one clinical 93 and two pre-clinical 94,95 studies, inflammatory responses were observed in the comorbid TBI/PTSD condition. 93,94 The clinical study evaluated 110 deployed, military personnel presenting with sleep disturbances related to TBI, PTSD, and depression. Logistic regression models were used to determine associations among blood plasma IL-6 level and quality of life ratings and service-related disorders. Quality of life was lower and plasma IL-6 concentrations were higher in military personnel with PTSD or depression. Military personnel with PTSD and depression were at higher risk for lower quality of life and higher plasma IL-6 concentrations. In a pre-clinical study of TBI/PTSD comorbidity using CCI and a predator stress model, rats demonstrated greater increases in activated microglial cells in the striatum, thalamus, and cerebral peduncle, compared with in those with PTSD alone or sham surgery controls. 94 Comparisons of levels of Ki-67-positive proliferating cells and DCX-positive migrating neural progenitor cells, as indicators of neuroinflammation and neurogenesis respectively, were made between TBI alone and TBI plus comorbid PTSD. This study revealed that PTSD did not further exacerbate the scales of the aforementioned neuropathological parameters relative to those of TBI-alone group. 94 Another study 95 examined anxiety and cognition changes in rats after repeated exposures to predator stress challenge alone or in combination with a single mild blast TBI. Rats with both TBI and stress exposures showed anxiety-related behaviors and spatial memory impairments that lasted up to 2 months. This TBI/PTSD group had a unique biochemical profile of elevated serum levels of stress and inflammatory biomarkers. The combination of pre-clinical and clinical evidence indicate that the comorbidity of TBI/PTSD lowers quality of life of the patients, and that this may be partially caused by a shared pathophysiological process of neuroinflammation involving molecular, cellular, and network abnormalities that can be measured by specific biomarkers (e.g., IL-1β, IL-6, TNFα, GFAP).
Glutamatergic pathologies of TBI
Glutamatergic dysregulation has been implicated in the pathophysiology of TBI. The most common of these include excitotoxicity from surges of extracellular glutamate and over-stimulation of glutamate receptors, and long-term potentiation (LTP)-related cerebral changes that impact cognition. Excitotoxicity during TBI results from an excessive stimulation of the glutamate N-methyl-
Several animal and clinical studies have reported that TBI produces central glutamate-induced changes in extracellular glutamate, glutamate receptors, and related second messenger systems. In a human trauma study, transient increases in cerebral glutamate were seen daily after TBI, and these increases were associated with seizure activity. 96 Another study used a CCI model coupled with tissue microdialysis after moderate and severe TBI. This study also found several-fold increases in the level of frontal cortical glutamate. 97 In addition, a rat FPI model produced a short-term (4 day) decrease in glutamate NMDA NR2 subtype receptors, which was associated with alterations in calcium accumulation. 98 In other TBI models, there were decreases in AMPA receptor GluR2 subunit internalization, which enhanced calcium-permeable AMPA receptors and subsequently increased cellular vulnerability to excitotoxicity in the hippocampus. 99
Glutamatergic effects resulting from TBI were also found to alter LTP, a cellular correlate of learning. Following FPI, LTP was blocked in brain slices from the hippocampal subregion CA1. This was associated with decreased NMDA potentials and glutamate-induced excitatory currents, and decreased expression of calcium calmodulin kinase II. 100 Patel and colleagues revealed that neuronal connectivity was altered by glutamatergic effects through specific NMDA receptors in TBI. 101 In primary cortical neurons from rats after a TBI, individual neurons whose calcium oscillations were mostly caused by NMDA receptor subunit GluN2B, lost many of their functional targets after injury. After damage, activation of GluN2B receptors limited neural remodeling in response to a plasticity stimulus. In contrast, neurons with large GluN2A contributions or with high functional connectivity were protected against loss of connectivity. Thus, altered cortical glutamate NMDA receptor signaling appears to reshape neural networks in a receptor subtype-specific fashion in TBI models, and these cortical changes may impact cognition.
Another major mechanism for eliminating excessive glutamate is through uptake by excitatory amino acid transporters (EAAT) in astrocytes. Following TBI in humans, there was a decrease in the number of EAAT1-positive, 102 and EAAT2-positive cells. 103 The resultant reduction in glutamate uptake by astrocytes could be a major contributor to increased levels of extracellular glutamate and excitotoxic damage following TBI. There are many functional impacts of these glutamate-induced changes, including epileptiform activity. In a CCI model, cortical glutamate signaling were measured using glutamate biosensors along with cortical field potentials in brain slices. 104 In CCI-injured cortex, glutamate signaling increased electrical stimulation-evoked epileptiform field potentials, whereas markers for GABAergic interneurons were decreased. In summary, TBI produces increases in cortical and hippocampal extracellular glutamate, reductions in glutamate transporters, and regulation of multiple glutamate receptors linked to increasing calcium flux. These alterations result in reduced GABAergic control, impairments in LTP, and increased epileptiform activity resulting in cell death. These neuronal changes may be linked to the cognitive impairments and increased seizure activity that follow TBI.
Glutamate pathologies of PTSD
The glutamatergic system plays an important role in the pathophysiology of stress and PTSD. Restraint stress was reported to increase extracellular levels of glutamate in the medial PFC, hippocampus, striatum, and nucleus accumbens. 105 This glutamatergic enhancement was significantly higher in the PFC than in other brain regions. Subsequently, stress-induced production of glutamate in the PFC and hippocampus was reversed by adrenalectomy, whereas glucocorticoid replacement reversed this effect, demonstrating the role of stress hormones. 106 Acute restraint stress increased extracellular glutamate levels in amygdala subregions of the BLA and CeA, as measured by in vivo microdialysis. 107 Interestingly, stress-induced increases in BLA glutamate were reversed by the antidepressant tianeptine. 108 Additional studies found effects of stress on the phosphorylation states of specific serine residues on the AMPA receptor subunits GluA1 and GluA2, which regulate subcellular trafficking in the amygdala and medial PFC. 109 In this last study, stress increased GluA1 subunit phosphorylation, and post-stress administration of synthetic corticosteroid receptor antagonists reversed these effects. These stress-induced changes in glutamatergic transmission are hypothesized to mediate neuronal morphology remodeling and alter functional neurocircuitry and behavioral changes. 110
In a recent study using a mouse model of comorbid mTBI (using CCI) and PTSD (using FC), there was enhanced acquisition of FC and delayed extinction in subjects with mTBI. 111 In this study, glutamate levels were increased, and the GABA/glutamate ratio was decreased in the ventral hippocampus. Further, GABA levels and GABA/glutamate ratios were both decreased in dorsal hippocampus. In a recent study using a murine modeling of mTBI and PTSD, it was demonstrated that BLA expression of NMDA receptors was significantly enhanced, coupling with decreases of GABAergic inhibitory neurotransmission in BLA and hippocampus. 112 These results support a notion that TBI may result in a loss of top-down control of PFC influence over the fear learning circuit and response to fear stimuli. Therefore, glutamate/GABA balance in PFC, hippocampal and BLA circuits may be related to hippocampal memory function. A complete understanding of these neurochemical changes and how they relate to the fear circuit and memory is critical for the treatment of this comorbid condition.
Oxidative stress and TBI
Free radicals can cause damage to proteins, DNA, and cell membranes by taking unpaired electrons from other molecules through oxidative processes. Oxidative damage results when toxic free radicals and reactive oxygen species (ROS) exceed the neural capacity to produce antioxidants to reverse these effects. The effects of oxidative damage on neuronal function following TBI are well documented in the literature (see Toklu and Tumer 113 for review). Additionally, oxidative stress increases blood–brain barrier permeability, and produces changes in synaptic plasticity, neurotransmission, and neural morphology. 114 Oxidative stress results in the reduction of a host of antioxidant enzymes, which contributes to the pathophysiology of TBI. 115 TBI produces ROS, such as the superoxide radicals and nitric oxide, which impair cerebral vascular function and produce ischemia. By enhancing ROS production, TBI reduces mitochondrial respiration and induces DNA deterioration, lipid peroxidation, protein and enzyme oxidation, and dysfunction. In one report, ROS changes produced impairments in the mitochondrial electron transport system and apoptosis, and all of these effects result in neuronal necrosis. 116 In this study, TBI resulted in mitochondrial fission, and treatment with a fission inhibitor improved hippocampal-dependent learning and memory, highlighting the utility of a mitochondrial approach. Because of the high rate of oxidative metabolism in the brain and its elevated levels of polyunsaturated lipids, which are the target of lipid peroxidation, neural circuits become vulnerable to oxidative stress. TBI-induced oxidative stress stimulates activation of neuroinflammatory cytokines and growth factors such as IL-1β, TNF-α, and transforming growth factor-beta (TGF-β). 117 In C57BL/6 mice with chronic TBI, long-term oxidative damage was found 1 year after injury, and was associated with lesion expansion, neural neurodegeneration, and demyelination. 75
As mentioned, oxidative stress often overwhelms antioxidant responses involving the enzymes glutathione peroxidase, catalase, and superoxide dismutase. In a rat CCI model, hippocampal glutathione activity demonstrated a biphasic response that peaked at 12 h and again at 7 days. 118 Increased hippocampal catalase activity occurred steadily through the 1st week after injury, whereas hippocampal superoxide dismutase activity decreased initially after 6 h following injury and continued to decline through 14 days. These findings demonstrate a complex signature of antioxidant enzyme activity and synaptic proteins following TBI. In a CCI model, cortical tissues were analyzed for several antioxidants. 115 Antioxidant enzyme activity demonstrated a time-dependent increase in oxidative stress over hours and days. Because reduction in pre- and post-synaptic proteins (synapsin-I and PSD-95, respectively) also occurred early in this study, any concurrent depletion of antioxidant systems appeared to adversely affect synaptic function and plasticity. Oxidative stress appears relevant in human TBI, as serial plasma oxidative and antioxidant markers were altered in patients with acute TBI versus controls. 119 In this last study, TBI patients showed increased serum glutathione levels and decreased erythrocyte superoxide dismutase levels. Using the Glascow Coma Scale as a behavioral outcome measure, serum glutathione and erythrocyte superoxide dismutase levels were higher in a subgroup with better functional outcomes than in a poor outcome subgroup. These findings highlight the pertinence and predictive nature of oxidative markers following human TBI.
Because the role of oxidative stress in TBI has been demonstrated in different models, there are a number of antioxidant therapeutic trials involving both experimental and clinical TBI. These trials utilize the ability of antioxidants to scavenge free radicals. A critical evaluation of evidence-based studies showed that antioxidant therapies such as amino acids, vitamins C and E, progesterone, N-acetylcysteine, and Enzogenol may serve as safe and effective adjunctive therapies in adult patients with TBI. 120 Although the impact of these treatments was limited, antioxidant therapies can be restorative, perhaps in combinatory therapies, because of their reversal of oxidative stress mechanisms.
Oxidative stress and PTSD
There is experimental evidence that increased oxidative stress is a factor in PTSD and in other stress models. In a rat model of inescapable footshocks, subjects displayed anxiety-like behavior, and showed enhanced fear acquisition and spatial memory deficits. 121 These behaviors were associated with increases in oxidation markers, nicotinamide adenosine dinucleotide phosphate (NADPH), oxidase 2 (NOX2), and 8-hydroxy-2-deoxyguanosine (8-OH-dG) in the hippocampus and PFC. In a model of single prolonged stress, rats displayed anxiety-like behaviors and enhancements in fear learning behavior that were associated with the increased expression of oxidation biomarkers of malondialdehyde, NOX2, and 4-hydroxynonenal in the hippocampus. 122 Additionally, in a predator stress model, there were increases in hippocampal, amygdala, and PFC levels of oxidative stress biomarkers such as superoxide, peroxynitrite, and total ROS. 85 Similarly, in human PTSD, a novel oxidative stress-related gene at the ALOX 12 locus moderated the association between PTSD and thinning of the right PFC. 123 This effect appeared to be localized to subregions including the middle frontal gyrus, superior frontal gyrus, anterior cingulate cortex, and medial orbitofrontal cortex. In summary, animal and human models of stress and PTSD show increases in oxidative stress biomarkers in the PFC, hippocampus, and amygdala. This oxidative damage in key neural circuits, along with neuroinflammation and excitotoxicity, results in changes in neural plasticity and morphology and disruption of neurotransmission.
One study examined neural oxidative damage in a comorbid model of repeated stress and mTBI. 124 Rats were divided into four groups that included naïve, repeated tail-shock stress, mTBI (using LFP), and a comorbid model of repeated stress followed by mTBI. After 1 week, repeated stress increased the expression of mitochondrial electron transport chain (ETC) complex protein subunits in the PFC, decreased pyruvate dehydrogenase (PDHE1α1) protein in the PFC and cerebellum, and decreased an ETC protein in the hippocampus. LFP alone decreased an ETC protein level in the ipsilateral hippocampus to the injury. The stress-mTBI group had its own synergistic profile of oxidative and mitochondrial damage. In summary, the comorbid TBI/PTSD condition has its own a unique profile of neuroinflammatory, oxidative damage and glutamatergic excitotoxicity damage in critical brain regions.
Neurodegeneration, Axonal Injury, and Dendritic Remodeling in TBI
Given these pathophysiological changes in PTSD, TBI, and their comorbid condition, we now discuss how neural structural changes develop in these disorders. TBI is associated with DAI, remodeling of dendritic connections, and spine loss. Dendritic spines provide the structural context for neuronal signaling and functional organization of synaptic connections. Therefore, any changes in neuronal morphology and dendritic spines have functional implications, as they reflect changes in synaptic strength. 125 Secondary molecular changes that are initiated in TBI, such as oxidative damage, glutamate excitotoxicity, and neuroinflammatory processes, interact with each other and produce neuronal damage and neurocircuitry dysfunction. Production of free oxygen radicals, superoxides, hydrogen peroxide, nitric oxide, and peroxynitrite impair the energy metabolism of the cells, impact membrane lipids, produce DNA fragmentation, and inhibit the mitochondrial electron transport system. This process induces apoptosis or necrosis of neural cells and neurites. 113 Activated microglia secrete various cytotoxic and neurotoxic factors that can lead to neuronal death. Microglia also release pro-inflammatory factors such as cytokines, and react to injury by secreting proteolytic enzymes that degrade the neuronal extracellular matrix and produce cellular debris (reviewed by Kou and VanderVord 126 ). Finally, glutamate excitotoxicity with excessive activation of NMDA receptors produces neuronal damage. NMDA activation of sodium and calcium pathways, proteolytic enzymes, and initiation of cell death pathways are all mechanisms that produce cell damage and neurodegeneration. 127
The loss of neurites is a consistent finding after TBI. In a CCI model in C57BL/6 mice, TBI acutely produced a 32% reduction of dendritic spines in the ipsilateral cortex and a 20% reduction in spines in the ipsilateral dentate gyrus. Additionally, there were >20% reductions in spines in the contralateral cortex and in the hippocampus. 128 These findings were replicated in another study in which spine density was reduced in ipsilateral cortical layers II and III and the dorsal dentate gyrus of the hippocampus. 129 TBI produced longer-term memory impairments accompanied by the shortening of PFC basal arbors at 4 months post-injury, and this occurred with findings of reduced density of both basal and apical spines in these neurons. 130 In contrast, another study indicated that at post-injury days 1, 7, and 28, brain-injured rats showed enhanced dendritic branch intersections in the BLA, as evidenced by Sholl analysis. 131
There are physical and neurochemical mechanisms related to changes in neuronal morphology, neurites, and spine loss. An in vitro model of axonal stretch injury produced evidence of dendritic beading, and suggested that DAI alters dendrite structure and plasticity. 132 Another study used an electromechanical cell shearing device to produce strain injury in neurons or astrocytes in an extracellular matrix scaffold. 133 This study produced a range of strains and tracked fluorescent microbeads in an acellular matrix. Neuronal cell death and loss of neurites correlated with higher strain. Dendritic changes were also impacted indirectly through other mechanisms such as calcium dysregulation in TBI. 134 In this last study, TBI altered the calcium-sensitive phosphatase calcineurin and its effector cofilin, an actin-depolymerizing protein, and these changes coincided with dendritic synapse degeneration. In summary, there are direct physical insults and neurochemical factors that contribute to TBI-induced dendritic and axonal injury. Cell death and dendrite degeneration in the cortex contribute to neural circuitry and behavioral dysfunctions in this condition.
DAI is the signature injury of TBI, and it results from the acute stretching and shearing of axons. This leads to disrupted axonal transport, axonal swelling, and secondary neuronal disconnection. 135 Data from studies of brain trauma in humans and on experimental brain trauma in animals indicate that DAI is a long-term process in which axons continue to degenerate and swell during an extended period. In the disconnected axons, both the β-amyloid precursor protein (β-APP) and other key enzymes (such as presenilin) for related amyloid-beta (Aβ) peptide generation accumulate in the swollen axonal bulbs. 136,137 After β-APP and Aβ are released from injured axons, accumulation occurs in the extracellular space as diffuse plaques. 137,138 Patients with TBI-related axonal injury produce an accumulation of β-amyloid proteins including β-APP and Aβ, and these are used as diagnostic markers of TBI. 139 Even a single brain trauma can result in the development of amyloid plaques, similar to those found in the brains of patients with other neurodegenerative diseases including Alzheimer's disease (AD), chronic traumatic encephalopathy (CTE), Dementia pugilistica, and amyotrophic lateral sclerosis (ALS). 140,136
Neural Remodeling in PTSD
PTSD is a brain disorder that results from alterations in neural plasticity and dendritic remodeling, impacting neurocircuitry function and associated behaviors. Functionally, the amygdala has a central role in fear regulation, and PTSD may result from hyperactivity of neurons in the amygdala. This limbic hyperactivity may result from negative feedback impairments from the PFC, and atrophy of the hippocampal region. 141 In stress models, the aversive stimulus represented in sensory cortical circuits activates the CeA, which projects to brainstem monoaminergic cell body regions. This results in activation of widely disseminated serotoninergic, dopaminergic, and noradrenergic pathways that project to cortical and subcortical regions. In PTSD, cues and contexts associated with the aversive stimuli are represented in the PFC. Presentation of these cues and contexts activates PFC glutamatergic outputs, projecting to neurons in the BLA regulating the activation or inhibition of the amygdala. 142 Repeated FC and fear retention results in impairment of PFC inhibition of the BLA, producing long-term cue- and context-induced fear responses and stress responses of defensive behaviors, autonomic activity, hypothalamus–pituitary–adrenal axis activation, and limits on cognitive processes. 142,143
Chronic stressors along with fear cues and contextual responses alter synaptic plasticity in rodent models. Electrophysiological studies have shown that stress modifies synaptic structure and function in the BLA, hippocampus, and PFC. Specifically, inescapable stress affects connections between the medial PFC and the BLA, and modifies both long-term depression (LTD) and LTP. 144 LTP persistently increases synaptic strength following repeated stimulation of a synapse, whereas LTD decreases synaptic strength. LTP is known to promote the formation of new spines and enhance dendritic complexity, whereas LTD reduces dendritic complexity or spines. 145 Exposure to inescapable stress reverses plasticity in the BLA, resulting in the promotion of LTP and the inhibition of LTD. After inescapable stress, LTP in the BLA is favored to encode memories of fear. 144 Moreover, stressed animals exhibit increased BLA plasticity as measured by neuronal firing rates and responsiveness. 146,147 In the PFC, some neurons respond during the occurrence of a stressor, whereas other neurons fire after the stress stops. 148
There are neuronal morphological and dendritic changes associated with these plasticity changes in PTSD models. Pre-clinical studies have demonstrated that FC increased the rate of spine elimination in the frontal association cortex whereas fear extinction increased spine formation. 149 In contrast, FC increased dendritic spine density and dendritic morphology in the BLA, whereas subsequent extinction training reversed these effects. 150 Another study found that repeated stress increased spine numbers in lateral and basal amygdala subregions in distinct patterns, and increased dendritic length in the basal regions. 151 Repeated stress increased fear retention while it upregulated the density of total and mature dendritic spines in the dorsal hippocampus; this effect was reversed by anxiolytic treatment. 152 In summary, increases in BLA and hippocampal dendritic plasticity have been demonstrated in chronic stress or conditioned fear along with decreases in plasticity in PFC regions. This is hypothesized to produce increases in fear behaviors and fear memories and impairments in their extinction in PTSD. The molecular mechanisms underlying spine loss and dendritic retraction in chronically stressed animals have been reviewed recently. 153 Synaptic proteins including glutamatergic proteins, vesicular glutamate transporter 2, and GluR1, GluR2, NR1, NR2A, NR2B receptors, and many others play a role in the regulation of spine alterations and synaptic plasticity. Expression of these proteins in dendritic spines is influenced by chronic stress and fear conditioning.
There are few studies that have examined synaptic plasticity in comorbid models of TBI/PTSD. One study used footshock stress prior to TBI injury in C57BL/6 mice and examined synaptic plasticity in the lateral amygdala using whole-cell patch clamp electrophysiology. 154 In unstressed mice, TBI increased membrane excitability and spontaneous excitatory postsynaptic currents (sEPSCs) in lateral amygdala neurons, up to 3 months post-injury. Stress alone, in the absence of TBI, also increased sEPSC activity in these neurons. In the comorbid TBI/PTSD condition, sEPSC activity was reduced compared with in either condition alone, demonstrating different synaptic mechanisms in the comorbid condition. Another comorbid study showed that mice with mTBI demonstrated increased fear acquisition and delayed fear extinction. 111 In this study, mTBI/FC mice, showed a reduced GABA/glutamate ratio in the hippocampus at 25 days post-injury. These findings suggest that reductions in PFC plasticity and increases in BLA and hippocampal plasticity contribute to long-term fear retention and incomplete extinction. In another study examining TBI and post-injury behaviors at 7 and 30 days after a mild CCI injury, rats displayed greater anxiety-like behavior. 155 This combination of mTBI plus anxiety was associated with a loss of BLA GABAergic interneurons and reductions in spontaneous and miniature GABAA-receptor-mediated inhibitory postsynaptic currents (IPSCs). These synaptic plasticity changes were associated with reduced BLA expression of inhibitory α1, β2, and γ2 GABAA receptor subunits, and with increases in the expression of excitation α7-nicotinic subunits. This suggests that reduction in inhibitory responses and increases in excitatory responses in the BLA may contribute to its hyperexcitability and to the development of fear-based disorders in comorbid TBI/PTSD. In this limited number of studies, comorbid TBI/PTSD produced a more unique synaptic and electrophysiological signature than either condition alone.
Translational Treatment Approaches and Conclusion
Despite intensive and extensive investment and the effort by basic or translational research to devise neuroprotective therapies for brain trauma, the field has not been able to establish any standard treatment via randomized clinical trials. A meta-analysis examined the behavioral and cognitive effects of pharmacological treatments from 30 randomized controlled trials and open-label trials. 156 Only two trials showed some efficacy, one trial demonstrating that dopaminergic agonist methylphenidate produced improvements in irritability and aggression, whereas another trial showed that the cholinergic agent donepezil improved memory and attention. Certain other medications may benefit both conditions, such as antidepressants, stimulants, anticonvulsants, and hypnotics. In the clinic, great attention must be paid to the risks of each medication, and it is useful to start at a low dose, titrate slowly, and monitor side effects and benefits closely.
Based on the pathophysiological mechanisms reviewed in this article, there are several novel approaches that can be tested for TBI/PTSD, including the use of antioxidants, anti-inflammatory agents, and agents such as anticonvulsants that block excitoxicity. Antioxidants could be used to reduce inflammation, edema, the breakdown of the blood–brain barrier, expanding lesion size, and the associated neurological and behavioral deficits. Anti-inflammatory agents can suppress damaging response to astrocytes, microglia, monocytes or macrophages, neutrophils, and lymphocytes. In combination with other treatment agents targeting angiogenesis and neuroplasticity, they may synergistically increase therapeutic efficacy. 157 Anticonvulsant agents can be tested in order to block the epileptiform activity and seizures associated with TBI.
An example of a medication that reduces epileptiform activity and enhances the progression of neuroplasticity is the anticonvulsant agent valproic acid (VPA) (also known as sodium valproate and divalproex sodium). This commercially available histone deacetylase inhibitor (HDACi) produces transcriptional activation and neuronal dendritic sprouting, and enhances synaptic connectivity. 158,159 VPA also reduces the activation of microglia and other inflammatory markers after brain injury. 160 Several studies have demonstrated that VPA treatment is neuroprotective against glutamate excitotoxicity, 161 and ameliorates TBI-induced hippocampal damage. 162,163 Preliminary clinical studies suggest that VPA might be a useful treatment approach for the mood symptoms of TBI patients. 164,165 VPA reinforces fear extinction learning in human studies. 166 Our group has been studying VPA for its capacity to enhance learning, neuroprotection, neuroinflammation regulation, and synaptic plasticity. We demonstrated that VPA enhances extinction of fear responses in FC models, 150 and may be useful in a comorbid TBI/PTSD model.
In the clinic, the combination of TBI and PTSD produces persistent symptomatology, behavioral problems, and dysfunction for many veterans and civilians, and there is a need to develop better treatment approaches. Using animal models, investigators can examine the cognitive, behavioral, and neurobiological changes in this comorbid condition, and test the efficacy of new pharmacological and behavioral treatments. PTSD animal models can include fear conditioning, chronic unpredictable stress, immobilization stress, and others. Dependent measures of PTSD can include cue- and context-induced fear responses, and related measures of anxiety (e.g., elevated plus-maze, dark-light box), anhedonia (e.g., sucrose self-administration), depression (e.g., swim immobility), novelty responses, aggression, and other measures. TBI models can include FPI, CCI, blast injury, and others, and require some measurement of functional outcome (e.g., cognitive impairment). The temporal positioning of TBI or PTSD as the initial trauma event does not seem to make a significant difference in animal models in terms of triggering comorbid behavioral signs. 167 However, the impact of the sequential order of TBI and PTSD encountering on their comorbidity has not been systemically investigated in human cases. The comorbid modeling of TBI and PTSD can be quite complex, and decisions have to be made about the relevance of the models to the human condition and their sequence of onset as well as multiple behavioral, cognitive, and neurobiological outcomes and their time course.
The efficacies of therapeutic agents may be synergistically or additively enhanced by newly emerged combinatorial tactics that include multipotential drugs and even stem cell-based therapies. 168 –170 Pharmacological trials have used pre-clinical models to suppress destructive neuroinflammatory responses, reverse oxidative damage and excitotoxicity, produce axonal and dendritic remodeling, initiate neurogenesis, and restore synapses in comorbid TBI/PTSD. Various forms of pharmacological and/or psychological training (extinction learning) reverse pathophysiological changes and produce axonal growth, dendritic remodeling, and neurocircuitry repair in TBI/PTSD. With further investigation, we hope to gain a better understanding of this devastating comorbid disorder in order to provide more efficacious, novel, and specific treatments.
Footnotes
Acknowledgments
This work was supported by Merit Review Award I01RX001144 from the United States Department of Veterans Affairs Rehabilitation Research and Development Service. The contents do not represent the views of the United States Department of Veterans Affairs or the United States Government. L.W., X.Z., and L.W. of Teng Lab were additionally supported by the Department of Defense (SC140194 to Y.D.T.), the Gordon Project of HMS, and China Scholarship Council, University of Wuhan Scholarship Fund, and Huazhang University of Science & Technology/Wuhan Union Hospital fellowships sponsored by Y.D.T. We thank Dr. David A. Williamson of Nexus Health Systems in Houston, Texas for his expertise in discussions of the psychiatric presentations of TBI and PTSD.
Author Disclosure Statement
No competing financial interests exist.