Exploring Molecular Pathways in Exercise-Induced Recovery from Traumatic Brain Injury

Introduction

Traumatic brain injury (TBI) is a functional change or other brain-related pathology caused by external forces and is a leading cause of death and disability in children and adults [1,2]. The annual incidence of TBI is estimated to be 5.48 million new cases various severities. This has increased by more than 50%, from 534.4 per 100 000 people to 787.1 per 100 000 between 2007 and 2013 in the United States [3]. According to survey data from the United States, as many as 40% of survivors with severe TBI will experience long-term disability, preventing and disrupting individual development [4]. The adverse short- and long-term clinical outcomes of TBI include death and disability as well as cognitive and behavioral problems. Even patients with mild TBI may have symptoms, including motor dysfunction, learning disabilities, memory and attention impairment, anxiety, and depression [5].

The primary causes of TBI vary according to age, socioeconomic conditions, and geographic region, ranging from motor vehicle crashes and falls to assaults. Medical interventions must consider this variability as well as TBI categories, with severe TBI patients being assessed for intracranial pressure, cerebral oxygenation, cerebral metabolism, cerebral blood flow, and autoregulation [6,7]. The implementation of effective interventional measures (eg, physical exercise (PE), medication, vitamin D and other nutritional supplementation) is crucial to minimize adverse outcomes and facilitate health recovery. Exercise has various benefits for the body and brain. Clinical and animal model studies have confirmed that PE can be applied as a non-pharmacological therapy to promote post-TBI nerve repair and improve recovery of cognitive ability, mood disorders, and post-concussion syndrome [8,9]. Compared with pharmacological therapies, exercise intervention offers several advantages, including reducing cost, increasing accessibility and flexibility, which can incorporate different physical activity regimes, whether dynamic or static, endurance, or resistance [8]. In this regard, early exercise at low intensity in the beginning of TBI onset can improve behavioral performance without inducing cognitive deficits, while high-intensity exercise can affect cognitive function [10]. Early subthreshold exercise was also demonstrated to be effective to speed recovery and can reduce the incidence of delayed recovery [11]. Exercise before TBI can be used as a neuroprotective method to minimize initial injury after TBI and to reduce secondary injury since it enhances baseline functioning of the brain [12]. Aerobic exercise has been found to increase the connectivity and intactness of microstructure in the corticospinal tract, resulting in increased academic performance and outcomes in middle and high school students [13,14]. This review summarizes the progress in research on the effect of pre-/post-injury exercise on TBI and outlines the molecular mechanisms underlying exercise-induced improvement of TBI. Future prospects for exercise-based TBI therapy are also presented. The information provided in this review will help develop more effective exercise-based intervention for patients with different TBI severity and further unlock molecular details underlying the therapy effect.

Pathogenesis of TBI

BBB DISRUPTION:

The BBB serves as a dynamic interface between the central nervous system (CNS) and peripheral circulation. It selectively filters substances flowing in and out of the brain. Disruption of the BBB occurs widely in mild, moderate, and severe TBI. Once disrupted, macrophages, neutrophils, and lymphocytes at the injury site are released [1]. When the primary injury occurs, the cerebrovascular system vibrates and be torn due to the external force, leading to shedding of vascular endothelial cells and loosening of tight junctions between endothelial cells, resulting in increased BBB permeability, release of activated leukocytes, inflammatory factors, glial cells, and substances toxic to neurons into the brain parenchyma, causing inflammation and immune reactions [20]. Secondary changes, such as neuroinflammation and immune response, further increase the destruction of the BBB and accelerate nerve tissue damage [21]. Recent studies have identified novel intracellular signaling and cell-cell interactions within the BBB niche, including an influx of immune cells such as neutrophils that can exacerbate the inflammatory response [22] and increase transcytosis across endothelial cells [23]. Therefore, BBB dysfunction is an important cause of secondary injuries associated with TBI.

OXIDATIVE STRESS:

Oxidative stress results from disequilibrium between reactive oxygen species (ROS) generation and antioxidant scavenging systems. The processes responsible for removal of ROS include enzymatic and non-enzymatic antioxidant cellular defense systems [24]. The brain consumes approximately 20% of the body’s total amount of oxygen and requires the highest oxygen supply in the body. After brain trauma, oxygen consumption increases, leading to increased production of ROS and induction of a series of oxidative stress responses [25]. The generation of ROS and other free radicals damages lipids, nucleic acids, and proteins, leading to inactivation of downstream enzymes, receptors, and ion channels, eventually causing cellular damage. ROS generation increases BBB permeability, which is a characteristic of TBI.

Oxidative stress is an important cause of secondary injuries such as neuritis, brain edema, sensory-motor dysfunction, and neuronal injury [26], as well as apoptotic and necrotic cellular damage [27]. Studies have shown that superoxide, hydrogen peroxide, peroxyl radicals, and hydroxyl radicals are the major ROS that react with lipids, nucleic acids, and proteins, and lead to downstream protein inactivation, including enzymes, receptors, and ion channels [28] and contribute to neuroinflammation in TBI [29].

NEUROINFLAMMATION AND PPOPTOSIS:

TBI triggers a systemic inflammatory response in the CNS, which activates glial cells (microglia) and astrocytes and causes the aggregation of immune cells in the peripheral blood and the release of inflammatory factors in thein tissue. Activation of the endothelium and a neuroinflammatory response can occur within minutes to hours after injury, resulting in recruitment and upregulation of cytokines, chemokines, neutrophils, and other pro-inflammatory mediators [30]. For example, the release of inflammatory mediators such as interleukin-6 (IL-6), interleukin-1 β (IL-1 β), and tumor necrosis factor-α (TNF-α) increased at approximately 6 h after trauma in the cerebrospinal fluid of TBI patients and rodent models, which in turn increased BBB permeability, leading to brain edema and nerve injury. Among these inflammatory mediators, TNF-α can bind to the proapoptotic Fas ligand to activate caspase (cysteinyl aspartate-specific proteinase), leading to apoptosis [31]. In the CNS, where neurons cannot regenerate, TBI can activate the neuronal apoptosis system, leading to neuronal loss [32]. Its molecular mechanisms mainly include activation of a family of cysteine proteases, release of apoptosis factor Bcl-2, translocation of apoptosis-inducing factor (AIF) to nuclei, and reduction of ubiquinol-cytochrome C reductase and complex III subunit XI (Uqcr11) expression [33].

EXCITOTOXICITY:

Excitotoxicity can cause neural death caused by massive secretion of excitatory neurotransmitters under many neuropathological conditions, including TBI [34,35], where increased concentrations of glutamate are secreted and released into the extracellular space owing to cellular lysis and a subsequent increase in positive charges within the cell [36]. As the major excitatory neurotransmitter in the CNS of mammals, glutamate plays a key role in maintaining learning, memory, cognition, and other functions. When the concentration of glutamate is too high, neurons become overexcited, resulting in excitotoxic reactions [37]. When the brain is exposed to external forces, as in the case of TBI, glutamate is overly released, leading to activation of the N-methyl-D-aspartate receptor, influx of Ca2+, and efflux of Na+ in the synapses, disturbing the ionic homeostasis in the synapses. The increased internal Ca2+ can, in turn, promote release of glutamate in synapses, forming a vicious cycle that aggravates destruction of nerve tissue. Dysregulation of extracellular glutamate and glutamate uptake in the acute stage of TBI and failure to resolve acute disruptions in glutamate homeostatic mechanisms can result in chronic cognitive symptoms [38,39].

MITOCHONDRIAL DYSFUNCTION:

In addition to being an organelle for energy supply, mitochondria also have important functions in maintaining neuronal homeostasis via regulating autophagy and immune reactions to maintain ROS balance and Ca2+ homeostasis. These functions are achieved through the mitochondrial permeability transition pore (MPTP) [40]. Since Sirtuin 3 (SIRT3), the major NAD+-dependent deacetylase, is primarily expressed in neurons [31] and is exclusively located in the mitochondria [41], it plays an important role in regulating mitochondrial function, metabolism, antioxidative defense systems, and energy metabolism [42]. After TBI, SIRT3 levels were suppressed and can be restored by APN/AdipoR1 signaling activation, suggesting that SIRT3 is a key regulator of APN/AdipoR1 signaling essential to maintain mitochondrial morphology and function [43]. In rat models with focal cerebral ischemia and global cerebral ischemia, inhibiting the opening of MPTP reduces infarct size and improves neurological symptoms, suggesting that MPTP may be an important target for neuroprotection [44]. Post-injury administration of MPTP inhibitor NIM811 has been found to be neuroprotective, improving cognition, and reducing oxidative damage in TBI rats [45].

Effect of Pre-TBI Exercise on TBI

ALLEVIATION OF TBI-INDUCED OXIDATIVE STRESS:

Studies have found that pre-injury PE can effectively reduce TBI-induced ROS surges and oxidative stress and has a protective effect on the BBB. These effects prevent or minimize secondary injuries that increase BBB permeability, neutrophil infiltration into the brain, inhibition of Na+/K+-ATPase activity, and generation of thrombin to mediate neurovascular dysregulation after TBI [53].

TBI-induced oxidative stress is closely related to changes in in vivo signal transduction. For example, nuclear factor erythroid 2-related factor 2 (Nrf2) is a basic leucine zipper protein transcriptional factor that plays an important role in regulating transcription of antioxidant proteins such as hemeoxygenase-1 (HO-1) and glutathione peroxidase (GPX), and influences the levels of biomarkers of oxidative stress, such as 4-hydroxynonenal (4-HNE) and malondialdehyde (MDA) [54]. Normally, Nrf2 is associated with a group of proteins, including cytoplasmic Kelch-like ECH-associated protein1 (Keap1) and Cul3 (Cullin3), and is localized in the cytoplasm. Under TBI-induced oxidative stress, Nrf2 detaches from Keap1 and translocates to the nucleus, where it heterodimerizes with small Maf proteins in the ubiquitin-proteasome system [55]. The dimer binds to the antioxidant response element (ARE) to initiate the transcription of a variety of antioxidant genes, leading to production of antioxidant proteins. Therefore, activation of the Keap1-Nrf2-ARE system is an important protective mechanism against TBI [56]. Activation of the Keap1-Nrf2-ARE pathway is dependent on exercise duration, with longer duration associated with higher expression of proteins in the pathway [57,58]. Furthermore, the signaling pathway can be activated in a blast-induced traumatic brain injury (bTBI) mouse model with significantly increased nuclear Nrf2, heme oxygenase-1 (HO-1), and NAD(P)H: quinone oxidoreductase-1 (NQO1) levels 3 days after bTBI [59]. A loss-of-function study showed that Nrf2 knockout (Nrf2(−/−)) mice had more severe neurological deficits, brain edema, and neuronal cell apoptosis compared with Nrf2(+/+) mice and were more susceptible to diseases related to oxidative stress, with higher expression of genes in the ER stress and ER stress-induced apoptotic pathways [60]. In contrast, a study on Nrf2 knockout mice showed that deletion of Nrf2 worsened post-TBI motor function recovery and lesions after TBI, and treatment with either apocynin (NOX2 inhibitor) or TBHQ (Nrf2 activator) alone significantly improved motor function after TBI [61]. Several studies have demonstrated that exercise can affect the expression of Nrf2 both in humans and animals. Regular exercise increases the level of Nrf2 in the cerebrospinal fluid and enhances the body’s ability to cope with oxidative stress [62]. Therefore, regular pretraumatic exercise can reduce the effect of TBI-induced oxidative stress on nerve tissues [63]. Accumulating evidence shows that in both animals and humans, exercise can increase Nrf2 protein and RNA expression in the brain, muscle, liver, and heart, and promote its binding with ARE in an exercise duration- and intensity-dependent manner, although further study is needed to elucidate the relationship between the expression of Nrf2 and exercise duration, intensity, and age [64–66]. Previous swimming training protected against glutamate uptake inhibition and oxidative stress, and increased the expression of hippocampal phosphorylated Nrf2, suggesting that previous PE can decrease lesion progression in this model of brain damage [67].

UPREGULATION OF HSP:

HSPs are a large family of conserved proteins that act as molecular chaperones and play a key role in protein homeostasis, regulation of apoptosis, and protection from various stresses (such as hypoxia, thermal stress, and oxidative stress) [68]. Hsp70 proteins are central components of the cellular network of molecular chaperones and play important roles in oxidative stress. They activate the molecular defense response in the triage of damaged and aberrant proteins for degradation via the ubiquitin-proteasome pathway in various neurological diseases, including TBI [69]. A study found that through previous swimming training, the expression of Hsp70 in the hippocampus of trained injured rats was significantly increased after fluid-percussion injury (FPI) as compared to the control rats, and inhibition of selected targets for free radicals (eg, Na(+), K(+)-ATPase) 24 h after neuronal injury was reduced, suggesting that previous PE can modulate protein expression with an antioxidant defense to prevent toxicity induced by TBI [67].

In another study, it was found that after 6 weeks of swimming exercise, the level of HSP70 in the hippocampus and prefrontal cortex was upregulated and the behavioral changes and the function of the HPA axis were improved in chronic unpredicted mild stress (CUMS) animals, while the level of inducible nitric oxide synthase (iNOS) was downregulated [70], suggesting that previous exercise can delay the disease progression characterized by TBI-induced movement defects and memory impairment through the activation of endogenous antioxidant molecules such as Nrf2 and HSP70 to protect the targets of oxidative stress (such as Na+- K+- ATPase) after TBI.

ALLEVIATION OF TBI-INDUCED INFLAMMATION:

Neuroinflammation is an important pathological cause of secondary TBI, which induces instant activation of innate immunity in brain tissue associated with impairment of the BBB, increases water content in the brain, and promotes the secretion of excitatory neurotransmitters, causing an imbalance in ion homeostasis and apoptosis, eventually leading to brain dysfunction [71]. Following TBI, a neuroinflammatory response can persist for years and contribute to the development of chronic neurological manifestations, including infiltration of peripheral immune cells into the brain after injury [72].

Exercise can inhibit neuroinflammation and its anti-inflammatory effect may be related to myogenic IL-6 [73]. The physiological concentration of IL-6 after exercise has been shown to stimulate the expression of the anti-inflammatory cytokines interleukin-1 receptor antagonist (IL-1ra) and interleukin-10 (IL-10) to increase the levels of a number of anti-inflammatory cytokines, naturally occurring cytokine inhibitors, and chemokines, and inhibit the production of pro-inflammatory factors such as TNF-α [74,75]. In clinical studies, the effect of IL-1ra as an anti-inflammatory drug in the treatment of TBI has also been confirmed [76], further proving that exercise has an anti-inflammatory effect. Mota et al found that 4 weeks of running exercise prevented FPI-induced motor impairment in rats, as indicated by fluorescein extravasation. Molecular analysis showed that the expression of anti-inflammatory IL-10 was upregulated and pro-inflammatory cytokines (IL-1 β) and tumor necrosis factor-α (TNF-α), as well as the infiltration of neutrophils in the brain, were downregulated, leading to reduced BBB damage and Na+- K+- ATPase inhibition, demonstrating that exercise preconditioning can improve the inflammatory status in brain tissue, reduce primary damage, and long-term secondary degeneration after TBI, with better prognosis. Compared to sedentary rats, previous physical training was effective against myeloperoxidase (MPO) activity increase and Na+,K+-ATPase activity inhibition after FPI [77]. Another study also showed that 6 weeks of swimming training protects rats against oxidative damage and neurochemical alterations, represented by immunodetection of the alpha subunit and activity of Na(+)/K(+)-ATPase after FPI in the cerebral cortex of rats [78]. Taken together, the effective protection of selected targets, such as Na(+)/K(+)-ATPase, by physical training supports the hypothesis that physical training has prophylactic effects on neuronal cell dysfunction and injury derived from TBI.

TBI-induced injuries are not restricted to the CNS and can also cause inflammatory reactions and damage other tissues and organs [79]. Furthermore, the TBI-induced oxidative-inflammatory response in peripheral organs can in turn lead to neuronal dysfunction as a result of a series of signal transduction pathways, such as the release of inflammatory chemokines in the liver and the production of molecules related to oxidative stress [80]. It was also found that 6 weeks of swimming training before trauma could regulate the oxidative-inflammatory pathway in the liver after TBI, resulting in elevated mRNA levels of X receptor alpha and ATP-binding cassette transporter, and decreased iNOS, cyclooxygenase-2 (COX-2), TNF-α, and IL-6 expression in the liver. In addition, exercise training can also prevent dysglycemia and impaired hepatic signaling after TBI (such as an increase in phosphorylated c-Jun NH2-terminal kinase, decrease in insulin receptor substrate, and phosphorylated AKT expression), and protect against hepatic inflammation (COX-2, iNOS, TNF-α, and IL-6), and oxidative stress (decreases in non-protein sulfhydryl and glutathione, as well as increases in 2′,7′-dichlorofluorescein diacetate oxidation and protein carbonyl), demonstrating that exercise training alters the profile of oxidative-inflammatory status in the liver and protects against acute hyperglycemia and a cerebral inflammatory response after TBI [81]. Therefore, exercise training may have therapeutic potential for TBI through modulation of responses by metabolic organs, such as the liver [82].

PROMOTING SECRETION OF NEUROTROPHIC FACTORS:

Growth factor brain-derived neurotrophic factor (BDNF) and its receptor tropomyosin-related kinase receptor type B (TrkB) regulate neuronal activity, function, and survival in the brain. The regulation of BDNF expression is critical for the physiology of neuronal circuits and functioning in the brain [83]. A meta-analysis showed that aerobic exercise, aquatic exercise, and multimodal regimens were significantly associated with improved BDNF levels, a proxy for cognitive function [84]. Acute bouts of exercise transiently improve cognitive function, whereas long-term exercise training stimulates brain plasticity and improves brain function due to increased circulating levels of BDNF [85].

LTP is a key mechanism of learning and memory. Because BDNF plays a very important role in promoting LTP, it can enhance the plasticity of synapses and participates in the formation and maintenance of learning and memory. After short-term voluntary wheel running, BDNF expression in the rat hippocampus increased in an exercise strength-dependent manner [86], which is beneficial for recovery of hippocampus-related cognitive functions [87]. Although other growth factors also increase with exercise, BDNF is the most durable and stable. For example, after 4 weeks of wheel running, the expression of BDNF in the ipsilateral cerebral cortex was significantly increased compared with the control group [88], resulting in increased synaptic plasticity, improved cognitive function and motor function after TBI, reduced TBI-induced damage to the hippocampus, cortex, and thalamus, and decreased neuronal apoptosis, suggesting that exercise preconditioning can increase the secretion of BDNF in brain tissue to exert protective effects on neurons after TBI.

In addition to BDNF, vascular endothelial growth factor (VEGF) and erythropoietin (EPO) also have important neurotrophic effects [89,90] and are promising therapeutic targets for a variety of neurological injuries, including TBI, owing to their neuroprotective activity [91]. In an in vitro cellular model, erythropoietin expression was found to be 70% higher in exercised skeletal muscles [92]. High-intensity interval exercise (5 days weekly for 7 weeks) leads to increased brain VEGFA protein and capillary density in wild-type mice, suggesting that PE contributes to improved brain function and delayed neurodegeneration [93]. Taylor et al found that compared with the controls, mice that exercised for 6 weeks of self-wheel running showed elevated VEGF-A and EPO mRNA levels in the cortex and hippocampus and improved TBI outcomes. Increased VEGF-A levels may promote angiogenesis to promote functional recovery from TBI [94,95]. Other benefits of preconditioning exercise include reduced systemic inflammation, largely via decreased serum TNF-α, boosted autophagic flux, and mitigated lesions [96].

INHIBITION OF TBI-INDUCED APOPTOSIS:

Apoptotic neuronal death may serve as a mechanism to remove unnecessary neurons with minimal activation of the immune system, and is therefore a physiological and protective response to TBI. However, excessive activation of apoptosis-related pathways could be harmful and lead to nerve function defects, resulting in a variety of cognitive dysfunctions, neurobehavioral disturbances, and emotional disorders [97]. The spatiotemporal distribution of neuronal apoptosis is related to the type and severity of TBI [98]. For example, in the moderate lateral fluid-percussion brain injury model, apoptosis was observed as early as 24 h in the injured cortex, whereas in the hippocampus and thalamus, the apoptotic response was delayed, peaking at 48 h and 2 weeks after injury, respectively [99]. In the brains of TBI mice, increased levels of Bax (B-cell lymphoma-2 associated X protein), caspase-3, and caspase-8 were detected due to activation of the apoptosis pathway [100]. Several studies have shown that regular exercise before injury can reduce TBI-induced neuronal apoptosis in the brain, thereby improving the prognosis. Zhao et al found that 4-week voluntary exercise preconditioning using a running wheel can reduce TBI-induced brain injury and neuronal death in the hippocampus, cortex, and thalamus and attenuate the activation of neuronal apoptosis pathways, resulting in improved recovery of sensorimotor performance in the beam walk task, as well as cognitive/affective functions in the Morris water maze, novel object recognition, and tail-suspension tests. Compared with controls, preconditioned mice had decreased microglial activation in the cortex, increased HSP70 expression, lower levels of BH3-interacting domain-only death agonist (Bid), and p53 upregulated modulator of apoptosis (PUMA), reduced AIF and caspases, and reduced AIF translocation into the nucleus [101,102]. BH3-only proteins (eg, BID, BIM) can activate the effector proteins BAX/BAK by neutralizing anti-apoptotic Bcl-2 proteins to initiate mitochondria-dependent apoptosis [103]. Under normal circumstances, AIF is secluded behind the outer mitochondrial membrane and contains the flavoprotein NADH oxidase, with the unique capacity to induce caspase-independent peripheral chromatin condensation and large-scale DNA fragmentation. Upon apoptosis induction, it is translocated to the cytosol and nucleus to regulate caspase-independent cell death [104]. AIF can be neutralized by HSP70 to inhibit its translocation to the cytosol and the nucleus. Zhao et al showed that running for 4 weeks before TBI also increased the expression of HSP70 to inhibit apoptosis activated by AIF, leading to reduced neuronal apoptosis and better prognosis after TBI [101,102].

STABILIZING MITOCHONDRIAL FUNCTION:

Mitochondria are important organelles that mediate apoptosis through the release of caspases, activation of the Bcl-2 protein family, and changes in the electron transport chain [105]. TBI was found to reduce the levels of mitochondrial antioxidants, such as glutathione, peroxiredoxin (PRX-3), thioredoxin (TRX), nicotinamide adenine dinucleotide phosphate (NADPH), superoxide dismutase (SOD), and catalase (CAT), and increase the expression of glyceraldehyde 3-phosphate dehydrogenase (GAPDH) [106]. Gu et al showed that pre-injury voluntary exercise programs improved the recovery of cognitive function after TBI in mice, and that this improvement was related to COX. After 3 weeks of running wheel exercise, traumatically injured mice had significantly higher levels of COXI, COXII, COXIII, BDNF, synapsin I, and ATP in the hippocampus than the controls, suggesting that exercise increases oxidative phosphorylation and ATP production in mitochondria, leading to reduced loss of neurons and better recovery of cognitive function after TBI [107]. Furthermore, sex appears to be associated with exercise-induced TBI improvement. Compared with male mice, low-moderate exercise is less effective in cognitive recovery but more effective in brain mitochondrial function improvement in female mice [108]. Sex difference was also found after controlled cortical impact (CCI), where PE appears to be more effective in females than males, who were more responsive to low-intensity PE, suggesting that PE protocols need to be tailored based on sex to enhance therapeutic outcomes [109]. Since the major cause of TBI-associated brain damage is secondary injury, mainly from mitochondrial dysfunction, which leads to oxidative stress and apoptosis and decreased cellular energy production, exercise has been demonstrated as an effective therapy to prevent and alleviate complex mitochondrial dysfunction after TBI [110].

Effect of Post-Traumatic Exercise on TBI

PROMOTING RECOVERY FROM TBI:

Although physician-prescribed rest after TBI remains the standard of patient care for TBI, it has been increasingly scrutinized. Animal studies have shown that moderate-intensity treadmill exercise initiated after CCI reduces anxiety-like behavior, improves hippocampus-dependent spatial memory, and promotes hippocampal proliferation and newborn neuronal survival in mice, although lesion volume and axon degeneration are not altered by exercise, suggesting that carefully titrated and implemented physical activity may be a safe and effective approach that benefits recovery after TBI [120].

Using a translational rodent model of concussion/mild traumatic brain injury (mTBI), the effects of voluntary exercise on were investigated, assessing post-concussive syndrome (PCS) and gene (bdnf, dnmt1, Igf-1, pgc1-a, Tert) expression changes in the prefrontal cortex and hippocampus. The results indicated that exercise initiated within 1–3 days after concussion significantly improved the recovery of motor and cognitive function and restored mTBI-induced changes in gene expression, suggesting that exercise is valuable for treatment of concussion [121]. Voluntary and forced exercises following TBI were found to have different effects on stress responses. Forced exercise continuously elevated corticosterone (CORT), while voluntary exercise elevated adrenocorticotropic hormone (ACTH), suggesting that exercise regimens with strong stress responses may not be beneficial during the early post-injury period [122]. Exercise intensity has a strong impact on the recovery of cognitive function after severe TBI. Rats in the low-intensity exercise group showed a significantly better improvement in spatial memory and learning in the Morris water maze (MWM) test compared to rats in the high-intensity exercise group. Consistent with this, BDNF and p-CREB protein levels in the contralateral hippocampus were significantly increased in the low-intensity exercise group, suggesting that 2 weeks of low-intensity treadmill exercise helped improve cognitive function and increased hippocampal BDNF expression after severe TBI [123]. This is consistent with results from an early study showing that exercise can endogenously upregulate BDNF and associated proteins involved in synaptic function and both acute and delayed exercise following experimental TBI enhance recovery, but the acutely exercised FPI rats failed to show activity-dependent BDNF upregulation and had significantly decreased phosphorylated synapsin I and total CREB; therefore, administration of excise after TBI should carefully consider the type and severity of TBI [124]. Using a well-characterized mouse controlled cortical impact model, it was found that late exercise beginning 5 weeks after injury significantly reduced working and retention memory impairment at 3 months and decreased lesion volume compared to non-exercise injury controls. However, such beneficial effects were not observed in the early exercise group (initiated at 1week) [125]. These results underscore the critical importance of the timing of exercise initiation and disagree with the widely held view that early intervention results in improved neuroprotection.

MECHANISMS RELATED TO RECOVERY OF TBI BY EXERCISE:

Post-injury exercise can promote the endogenous repair of brain injury and increase neuroplasticity to promote functional recovery after TBI via various mechanisms (Figure 2). Delayed wheel running (14–20 days after FPI) significantly upregulated the expression of BDNF, cAMP-response element binding protein (CREB), and synaptic protein I. Compared with the controls, delayed exercise improved the learning and memory abilities of rats. In contrast, when BDNF was blocked with TrkB-specific antibodies, the improvement derived from exercise in the recovery of cognitive function was significantly reduced [124]. Therefore, the time window for exercise following injury is critical to achieve a desirable therapeutic effect in TBI, which is highly dependent on the severity of TBI. For example, after a mild FPI, when voluntary exercise was initiated 2 weeks later, BDNF expression increased with an improvement in behavioral outcomes in rats. Furthermore, in more severely injured animals, a longer time to initiate exercise after injury is needed to increase BDNF, synapsin I, and CREB levels. In addition, both less and more exercise increase hippocampal pCREBSer133-CREB-proBDNF-BDNF signaling and mitochondrial coupling efficiency and could build up long-lasting CREB-BDNF and bioenergetic neural reserves that preserve memory fitness after severe TBI [126]. Therefore, it is likely that exercise-mediated improvement in cognitive function may be associated with activation of the CREB/BDNF/TrkB signaling pathway. HSP also plays an important role in the exercise-mediated improvement of TBI through its effect on BDNF. Chou et al investigated whether physical exercise (PE) could reduce cognitive deficits in rats after TBI, and determined the effect of hippocampal and cortical HSP20 on PE-induced cognitive recovery. They found that the expression of HSP20, BDNF, and TrkB increased after PE, and brain contusion and cognitive deficits were significantly reduced. However, if HSP expression was knocked down by intracerebral injection of pSUPER containing an hsp20 small interfering RNA, the PE-induced overexpression of BDNF and the TrkB ratio was reversed, and the beneficial effect on neurotrauma and cognitive deficits disappeared, highlighting the importance of HSP in stimulating the cerebral HSP20/BDNF/TrkB signaling axis [127]. Stimulation of the HSP70/NF-κB/IL-6/synapsin I axis by exercising led to reduced neurological deficits, while knockout of cerebral HSP70 significantly reversed the beneficial effects of exercise. A positive correlation was found between IL-6 and HSP70 in peripheral blood, suggesting that HSP70 is related to anti-inflammatory responses in injured brains [128].

Exercise has been shown to improve the neuronal microenvironment and help restore recognition and behavioral functions in patients with TBI. This process can result from the effective suppression of neuroinflammation and oxidative stress induced by brain injury, resulting in reduced neuronal degeneration and apoptosis around the damaged area, including increased fetuin-A expression and reduced stress response through increasing adrenal sensitivity (AS), brain corticosterone (CORT), and corticotropin-releasing factor (CRF) [129,130]. Exercise training has also been demonstrated to have anti-apoptotic effects by increasing IGF-1, IGF-1 receptor, p-PI3K, Bcl-2, HSP 72, and p-AKT, and decreasing Bid, t-Bid, Bad, Bak, Bax, TNF, and FADD [131]. Ko et al found that 8 weeks of treadmill exercise beginning 2 weeks after TBI suppressed the expression of the pro-apoptotic protein Bax and increased the expression of the anti-apoptotic protein Bcl-2 in the hippocampal dentate gyrus of rats and improved spatial learning ability, indicating that treadmill exercise after TBI can inhibit neuronal apoptosis in the brain by suppressing TBI-induced activation of the CREB/BDNF/TrkB signaling pathway after TBI [132].

Mitochondria perform diverse yet interconnected functions, generating ATP and many biosynthetic intermediates that support neuronal homeostasis and reduce cellular stress responses such as autophagy and apoptosis [133]. Exercise training is considered an effective stimulus for stabilizing mitochondria-associated endoplasmic reticulum membranes to maintain intracellular Ca2+ homeostasis and metabolic flexibility [134]. It was found that 6 weeks of swimming stabilized the opening of MPTP in rats and reduced the Bcl-2/Bax ratio to inhibit apoptosis. In addition, exercise can induce autophagy in mitochondria through the AMPK-ULK1 pathway to delay symptom progression (Figure 2) [135].

MicroRNAs (miRNAs) are a class of small non-coding RNA molecules that regulate gene expression through post-transcriptional modifications. In TBI, miRNAs such as miR-21, miR-23b, miR-532-5p, and miR-182-5p have been shown to participate in inflammation, neuronal apoptosis, reactive gliosis, disruption of the blood-brain barrier, and influence PE-induced functional and cognitive recovery, and may be developed as diagnostic markers and therapeutic targets for TBI [136,137]. Bao et al found that after TBI, running wheel exercise altered the hippocampal expression of miRNAs in both sham and TBI mice; hippocampal miR-21 and miR-34a were increased after 15 days running wheel started 2 weeks after the injury and were significantly associated with improved cognitive function (Figure 2) [138].

TIME TO START EXERCISE AFTER TBI:

The optimal time to start exercise after TBI depends on many factors such as the type and severity of TBI and the patient’s condition, including age, sex, premorbid conditioning, neuropsychological health, physical strength, and health. It is often case-specific and controversial, and there are no general guidelines for the starting time. Early studies with rats showed that after TBI, the levels of BDNF in the brain undergo dynamic restorative processes and energetic changes that may influence the outcome of exercise, and exercise starting within 1 week after injury may disrupt the brain’s autonomous recovery and aggravate the injury. On the other hand, exercise delayed until the third week could be more effective for recovery of cognitive and other neurological functions, resulting in a significant decrease in the number of trials to locate the platform in 7 s or less for 4 consecutive trials, compared with sedentary rats. In contrast, cognitive performance in acute FPI-RW rats was significantly impaired, and exercise can endogenously upregulate BDNF and enhance recovery when delayed after TBI [124]. Furthermore, they found that the therapeutic window for the implementation of voluntary exercise was dependent on injury. In more severely injured animals, more time was needed after TBI to start voluntary exercise to increase neuroplasticity and exhibited significant increases in synapsin I and CREB, in addition to BDNF [139]. Compared with the sedentary group, rats with access to a running wheel for a 25-day period from post-injury day 11 had restored “when” and “where” memories, with increased concentration of BDNF in the hippocampus, which is associated with “when” memory [140].

However, current ‘return-to-play’ guidelines are conservative, and Mychasiuk et al found that initiating exercise within 1–3 days post-concussion significantly improved motor and cognitive functioning and increased BDNF expression, although it had limited effect on emotional impairments. If deprived of social interaction and exercise, animals do not recover, and exhibit impairments similar to typical mTBI animals [121].

To determine the optimal administration schedules for PE, Amoros et al compared the effect of different protocols of voluntary wheel running on improvement in object recognition memory (ORM), neuroprotection (NeuN+ cells), microglial reactivity (Iba1 staining), and neurogenesis (DCX+ cells) after controlled cortical impact injury (CCI). They found that the effects of exercise on memory and neurogenesis appeared to depend on the specific temporal schedule and the amount of daily exercise. There were significantly more mature neurons in the hippocampal dentate gyrus in the delayed exercise group than in the early exercise group, suggesting that delaying exercise (by approximately 3 weeks) after TBI is more beneficial for improving the prognosis of TBI, although early exercise also promotes functional recovery and reduces the loss of NeuN+ cells in the hilus [141]. Hu et al also found that early voluntary wheel running (VWR) 2 days after injury enhanced neurofunctional recovery and alleviated anxiety in mice, and this change was associated with reduced pro-inflammatory factor levels, increased anti-inflammatory factor levels, and reduced microglial activation [142].

EXERCISE MODALITIES:

At present there are no standardized recommendations concerning physiotherapy of individuals with TBI, and choosing exercise modalities for TBI is often challenging because it is generally not known what therapies are effective. Therapeutic methods used for treatment vary among therapists and institutions [143,144]. While it is difficult to compare the efficacy of various treatment modalities, a few examples are given below to show the variability of exercise modalities for TBI.

Ambulatory participants with TBI for more than 6 months had better outcomes in the 6-min walk test and peak treadmill speed after performed high-intensity training (HIT) stepping practice versus conventional training for up to 15 sessions over 5 weeks. Greater gains were also obtained in aerobic capacity and efficiency after HIT, with additional improvements in cognitive functions, suggesting that the amount and intensity of stepping practice may be important in improving locomotor outcomes in patients with chronic TBI [145].

Bland et al reviewed the effectiveness of non-aerobic exercise interventions for improving gait and balance in patients with mild-to-moderate TBI, finding limited evidence of the positive effects of balance, gait, or the combination of both interventions, in TBI rehabilitation. However, the sample sizes with heterogeneous groups were smaller in these studies, and more consideration and conformity in the choice of outcome measures in design and standardization treatment approaches are essential in future research to advance practice [146].

Ballistic resistance training in patients recovering from TBI was found to be equally or more effective in improving mobility, strength, and balance than non-ballistic training, which included balance exercises, lower-limb stretching, conventional strengthening exercises, cardiovascular fitness training and gait training, suggesting that ballistic resistance may be better targeted towards those with more severe mobility limitations [147,148].

Since a sedentary lifestyle and lack of endurance are common characteristics of individuals with TBI, increasing physical activity and exercise training would improve cardiorespiratory fitness. However, there are very few well-designed studies of physiologic and psychological adaptations of fitness training [149]. In several studies that evaluated the efficacy of endurance training in patients with a TBI, it was found that physical work capacity can be improved with endurance training, and endurance training should be considered a necessary component of rehabilitation programs [150]. Moreover, aerobic training 3 times per week for 16 consecutive weeks improved endurance capacity with increased oxidative capacity and abdominal muscular endurance, but failed to reduce the oxygen cost of walking, suggesting that moderate and prolonged activity are beneficial in the comprehensive rehabilitation of patients with TBI [151]. Rapid-resisted elliptical training was applied to ambulatory individuals with TBI to increase motor and thereby cognitive processing speed. While it was a small trial with 12 patients, the results showed that balance reaction times improved and were correlated with gains in the High-Level Mobility Assessment Tool (HiMAT) and dual-task (DT) performance. Sleep quality also improved and was correlated with improved depression and learning. This study suggests that intensive rapid-resisted training is effective in improving dynamic balance and sleep quality in TBI [152].

Conclusions and Prospects

PE is a potentially inexpensive, accessible, and optimizable rehabilitation approach that facilitates TBI recovery. As a non-pharmacological treatment, regular exercise before and after TBI can promote the recovery of TBI patients and prevent the aggravation of injury due to secondary damage. However, the outcome of PE is influenced by several factors, including the disease severity, sex, and timing and intensity of PE. Several mechanisms underlying these benefits have been investigated, which interact with and influence each other. Although the improvements obtained from exercise suggest that it can be used as an important preventive and therapeutic option for TBI, it should also be noted that owing to the diversity and complexity of secondary injuries in TBI, combined therapy methods should be considered for optimal efficacy, including the use of anti-inflammatory drugs and dietary intake of DHA.

There is significant heterogeneity in human TBI cases. The injury type, severity, interval time, and frequencies differ, and risk factors such as genes, sex, age, and medication history are highly variable. More studies are needed to investigate the type of exercise best suited to treat TBI under various circumstances, to elucidate the detailed pathogenesis mechanisms of TBI, particularly chronic traumatic encephalopathy (CTE), and to develop both diagnostic and prognostic biomarkers for TBI diagnosis and treatment, such as blood-based biomarkers, and clinically well-accepted therapeutic methods tailored to various types of TBI. To facilitate the application of exercise-based therapy for TBI, development of animal models could help mimic the process of TBI, and research on uncontrollable complications such as skull fractures and cerebral hemorrhage, brain contusions, and even brain tissue loss is needed for a better understanding of the molecular pathways involved in TBI.