Background
Traumatic brain injury (TBI) is a type of physical brain insult caused by blunt mechanical force [
1]. According to the World Health Organization, TBI is a critical public health and socioeconomic problem throughout the world. Considering the incidence of TBI increases annually, TBI will be a major health problem and a substantial cause of disabilities by 2020 [
2,
3]. Treatment of TBI can be very costly due to the need for on-going treatment of long-term effects, which can include psychiatric vulnerability and neurological disorders. The prevalence of neurological dysfunction following TBI has been increasingly appreciated [
4,
5]. For example, the risk of clinical depression is approximately1.5 times higher in TBI survivors compared to the general population. Furthermore, the risk for developing Alzheimer disease is 2.3 to 4.5 times greater, depending on whether the TBI is moderate or severe [
6]. Worse, still no effective treatment options are available, and little is known about the complex cellular response to TBI.
TBI occurs when the brain is exposed to external forces that induce focal and/or diffuse injuries that include axonal shearing, edema, vascular damage, and neuronal death [
7‐
9]. Injuries from TBI can range from mild to severe [
10]. The post-TBI primary insult typically leads to secondary damage and neuronal death caused by inflammation [
8,
9,
11], oxygen radical formation [
12,
13], calcium release [
14], and mitochondrial dysfunction [
15,
16]. To date, TBI research outcomes have not translated into clinical therapeutic approaches [
17]. Thus, there is an increasing need for further research into the signaling cascades activated by TBI.
Neuroinflammatory signaling leads to a specific type of inflammatory necrosis called pyroptosis, which is characterized by cell swelling, lysis, and release of pro-inflammatory cytokines and intracellular contents [
18]. Pyroptosis is involved in pathological processes during viral or bacterial infection (or in the presence of products from pathogens such as lipopolysaccharides (LPS) or viral DNA) [
19,
20], myocardial ischemia, lung and kidney injury, and cerebral stroke [
21‐
24]. This type of inflammatory necrosis occurs predominantly in the phagocytes, macrophages, monocytes, and dendritic cells, as well as various other cell types such as T cells. The pro-inflammatory caspase subfamily, including caspase-1 in humans and mice, caspase-4 and caspase-5 in humans, and caspase-11 in mice, mediate pyroptosis [
19]. Li et al. found that chemical neurotoxicity-induced pyroptosis of the cortical neurons is dependent on caspase signaling through the Erk1/2-Nrf2/Bach1 signal pathway [
25]. However, the specific mechanism of neuronal pyroptosis in TBI has not yet been elucidated.
To better understand the mechanism of brain injury responses, we used an in vivo mouse model of TBI and in vitro cellular models of mechanical stress and inflammation. Caspase-1 knockout (KO) mice (caspase-1−/−), siRNA knockdown, and pharmacologic inhibition were used as tools to assess the involvement of caspase-1 in downstream inflammation following TBI and mechanical stress. We found that genetic ablation, expression reduction, and pharmacologic inhibition of caspase-1 decreased TBI- and mechanical injury-induced behavioral neurological deficits, inflammatory cytokine protein and mRNA expression levels, and neuronal apoptosis. Ultimately, our results lead us to conclude that mechanical stress caused by TBI activates caspase-1 which contributes to TBI-induced pyroptosis.
Methods
Animals and TBI models
In this study, we used caspase1-deficient (caspase1−/−; Jackson Laboratories, Bar Harbor, ME, USA) adult male mice (8–10 weeks old) and age-matched C57Bl/6 (WT) male mice. Mice were housed in the animal care facility under a 12-h light-dark cycle, with ad libitum access to food and water. All surgical procedures were carried out in accordance with protocols approved by the Institutional Animal Care and Use Committee. The animals were anesthetized with intramuscular injection of chloral hydrate (5%, 0.1 ml/10 g). Mice were fixed in a stereotaxic frame, and the scalp was shaved and cleaned with iodophor. The left lateral aspect of the skull was exposed by retracting the skin and the surrounding soft tissue. Traumatic brain injury (TBI) was induced by dropping a 200-g steel weight with a flat end from a height of 5 cm onto the left lateral skull. The free-falling weight produced a moderate contusion injury of the left parietal cortex. Sham-injured control animals underwent a similar anesthesia protocol as the TBI animals with the exception that the scalp was not exposed and no weight was dropped. About 25–30% of the mice died in the first seconds after the trauma, but there was no delayed mortality or prostration in the surviving mice. The body temperatures of the animals were monitored by a rectal probe and fixed in a range of 37 ± 0.5 °C using a heating pad. The animals were returned to their quarters after recovery from anesthesia and were allocated randomly into groups, with 18 mice in each group.
Neurobehavioral training and evaluation
Neurological deficits were assessed using well-established, modified neurological severity scoring (mNSS), open-field and Rotarod testing at 12, 24, and 48 h after TBI. Experimenters were blinded to the TBI or sham treatments as well as the genotype of the mice. Each behavioral test was repeated twice with four different trials to validate the data. The mNSS test consists of ten different tasks that can evaluate the motor (muscle status, abnormal movement), sensory (visual, tactile, and proprioceptive), balance, and reflex functions of mice. Neurological function was graded from 0 to 18 (0 = normal function; 18 =maximal deficit). One point was scored for each abnormal behavior or for the lack of a tested reflex. Therefore, higher scores imply greater neurological injury.
The open-field test was used to determine general activity levels and to measure anxiety-like behavior. Animals were monitored under moderate lighting for 15 min in a 40-cm2 open field using video tracking software (ANY-Maze, Stoelting, USA). General activity was evaluated by determining the total of distance traveled. Anxiety-like behavior was assessed based on the pattern of exploration in the open field (center vs. periphery).
Fine motor coordination and learning were assessed using an accelerating Rotarod apparatus (RWD Life Science, China). On the day before the injury, mice were trained on the Rotarod for three consecutive trials at a slow rotational speed (4 rpm/min) for 1 min to adapt to the rod, followed by four additional trials with an accelerating rotational speed (from 4 to 40 rpm in 5 min) to obtain baseline latency. On each testing day, the mice were given four 300-s accelerating Rotarod trials with an inter-trial interval of 30 min. The average latency to the first fall off the rod was recorded. Passive rotation, accompanying the rotating rod without walking, were also considered as a fall.
Primary neuronal culture and injury models
The primary neuronal culture was prepared as follow: Cortex from embryonic day 14 BALB/c mice was used for cultures. After decapitation of the mice and removal of the meninges, the cortex was digested by papayotin, and the cells were treated with DNase I (Sigma) prior to centrifugation at 1000×g for 3 min. Neurons were cultured on poly-l-lysine-coated dishes at a density of 1 × 106 cells per 10-cm dish in minimum essential medium with 10% horse serum. The medium was changed to neuronal base medium (supplemented with 2% B27, 0.2 mM L-glutamine, and 1% penicillin-streptomycin) on the following day. The medium was changed every 3 days, and cultured cells were used for experiments on post-culture days 7–14. Subsequently, the primary neuron was collected and detected the Iba1 (microglia) and GFAP (astrocytes) expression through Western blotting (WB) and NeuN stain to assess neuronal purity.
We then performed sets of three experiments to examine the effects of mechanical and inflammatory stress on neurons in vitro: (1) neurons were incubation for 5 h with LPS and 5 mM ATP for 30 mins; (2) neurons were scratched by a 200-μl yellow gunshot; and (3) equiaxial stretch (12% strain, 1.0 Hz frequency) was applied to cultured neurons for 4 h, by a Flexcell® FX-5000™ Tension System (Flexcell, USA). In the stretch model, neurons were inoculated in 6-well plates (BioFLEX®). At the same time, we studied the role of caspase-1 in neuronal pyroptosis by caspase-1 siRNA knockdown or adding the caspase-1-specific inhibitor, Belnacasan (VX-765, 10 μM, for 24 h). After treatments, cells were used for flow cytometry, immunofluorescence, and protein or RNA extraction.
Transient transfection
The primary neuron was respectively infected with caspase-1 shRNA (m) lentiviral particles (Santa Cruz, sc-29922-V) according to the manufacturer’s protocol (Santa Cruz Biotechnology, Inc.). Plate targets cells in a 24-well plate 48 h prior to viral infection. The cells should be approximately 50% confluent before infection. Add 0.5 ml of complete optimal medium with caspase-1 shRNA lentiviral particles (MOI = 5) (without polybrene), incubate cells for 48 h, and then change complete optimal medium without lentiviral. At 72 h after infection, the RNAi efficiency was determined by qRT-PCR (Additional file
1: Table S1 and Additional file
2: Figure S1).
Apoptosis analysis by Annexin V-PI double staining
Neurons were stained with Annexin-V and propidium iodide (PI) using the Annexin-V-FITC kit (BD Biosciences) according to the protocol of the company. Briefly, cell culture medium was collected separately, and the treated cells were digested and washed twice with ice-cold PBS and re-suspended in the binding buffer at a concentration of 1 × 105 per 100 μl. Then, the cells were supplied with 5 μl of Annexin V-FITC and 5 μl of PI and incubated in the dark at RT for 10 min. Next, 400 μl of binding buffer was added into each sample for flow cytometry analysis (BD Accuri. C6 Plus, BD Biosciences).
LDH release detection
Mice were sacrificed under deep anesthesia for blood collection (centrifugation at 3000g, 10 min), and the blood serum was measured for lactate dehydrogenase (LDH) release. For the primary neuron, supernatant from serum-free media was filtered using 0.2-μm syringe filters to use for LDH release detection. The detection was using a commercially available kit (Solarbio, Beijing). One hundred microliters of the blood serum or supernatant was transferred to 96-well plates, and then the reaction mixture was added and incubated in the dark for 30 min at RT. LDH concentration was quantified by measuring the absorbance at 490 nm.
Enzyme-linked immunosorbent assay
Total protein concentrations were measured using a BCA Protein Assay Kit (Thermo Fisher Scientific, Carlsbad, CA, USA). The levels of TGFβ, IFN-γ, IL-6, IL-10, IL-1β, and IL-18 were measured using ELISA kits (Anoric-Bio, Tianjin, China) according to the manufacturer’s instructions.
Caspase-1 activity detection
The caspase-1 activity was detected with Caspase-1 Activity Assay Kit (Solarbio, Beijing). The specific steps were conducted according to the manufacturer’s instructions: First, 2–10 × 106 cells were lysed in 50–100 μl lysis buffer on ice for 10 min. For tissue samples, 3–10 mg tissues were added to 100 μl lysis and were homogenated with a tissue homogenizer and centrifuged. The supernatant was retained. Protein concentrations were detected using the Bradford method, ensuring that the protein concentration was 1–3 μg/μl. A standard curve was prepared using the pNA standard. Then, the optical density of specimen was read on a microplate reader (Molecular Device) at 405 nm. The percentage of caspase-1 activity changes was calculated by the radio of OD405 of the experimental wells to that of the normal wells.
TUNEL staining
The cerebral cortex was harvested, and apoptosis was determined by terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick end labeling (TUNEL) staining, using a TUNEL Apoptosis Assay Kit (R&D, Switzerland). TUNEL staining was performed with fluorescein-dUTP for apoptotic cell nuclei and 4′,6-diamidino-2-phenylindole (DAPI) to stain all cell nuclei. apoptosis index (AI), the number of TUNEL-positive cells divided by the total cells per field, was examined. Each AI was assessed in 20 randomly selected fields.
Immunohistochemical analysis
Mice were euthanized by injection of 5% chloralic hydras in accordance with our approved animal protocol, and the brain tissue was removed. The brain tissues were immediately perfused with 4% paraformaldehyde in PBS after removal and embedded in paraffin for sectioning. Following rehydration of the paraffin section and two washes in PBS, endogenous peroxidase activity was blocked using 3% H2O2 for 10 min. Antigen crosslinking was conducted in an autoclave for 3 min, and then the samples were washed twice in PBS. All preparations were then treated with goat serum blocking reagent (Abcam, ab7481) for 45 min. Excess reagent was removed with a quick rinse with PBS. Sections were incubated with primary antibodies overnight. Following two washes in PBS, the secondary antibody (Abgent, ASS3403) were added to all slides for 30 min. Excess reagent was removed, and the slides were washed in PBS and incubated with DAB (Solarbio, DA1010) for 5–10 min until the desired color appeared. All preparations were counterstained with hematoxylin (Solarbio, G1120) for 30 s. Protein expression was determined in five adjacent sections per sample.
Western blot analysis
Cells were lysed in lysis buffer containing Proteinase Inhibitor Cocktail (Roche Applied Sciences) and Halt Phosphatase Inhibitor Cocktail (Thermo Scientific). Protein concentrations were quantified by using a BCA protein kit (Thermo Scientific). Samples (50 μg protein per lane) were loaded on 12% SDS-PAGE gels. After electrophoresis, the proteins were transferred to nitrocellulose membranes. Then, the membranes were blocked with 5% BSA and incubated at 4 °C overnight with the appropriate primary antibodies: caspase-1 (Abcam, 1:1000), caspase1 (p10) (AdipoGen, 1:1000), caspase-1 (p20) (AdipoGen, 1:1000), caspase-11 (Abcam, 1:1000), GSDMD (Santa Cruz, 1:100), Iba1 (Abcam, 1:1000), and GFAP (Abcam, 1:1000). Immunoreactivity was detected by incubation with horseradish peroxidase-conjugated secondary antibodies (Abgent, 1: 20,000) followed by chemiluminescent substrate development (Thermo Scientific, USA). Optical densities of the bands were calculated using a MiVnt image analysis system (Bio-Rad, CA, USA). All samples were run in parallel with four replicates.
For supernatant detection, supernatant from neurons in serum-free media was filtered using 0.2-μm syringe filters and concentrated using Centricon Plus-20 centrifugal filter devices (Millipore). Concentrated supernatant (50 μg protein per lane) was separated by 15% SDS-PAGE and transferred to nitrocellulose membranes, and processing was assessed by Western blotting using anti-caspase-1 (Abcam, 1:1000), anti-caspase1 (p10) (AdipoGen, 1:1000), and caspase-1 (p20) (AdipoGen, 1:1000). Uncropped blots for the pro-form and cleaved fragments of caspase-1 in figures were supplemented in Additional file
3: Figure S2.
To extract protein from the cortical brain tissues, the frozen brain samples were homogenized in ice-cold buffer containing 20 mM HEPES (pH 7.0), 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, Proteinase Inhibitor Cocktail (Roche Applied Sciences), and Halt Phosphatase Inhibitor Cocktail (Thermo Scientific) for 30 s. The samples were sonicated through the Cell Ultrasound Cruizer (Thermo Scientific) for 10 s and centrifuged at 15,000 rpm for 10 min, and total proteins were recovered in the supernatants.
Immunofluorescence assays
Primary neurons were subjected to immunofluorescence by the protocol described below. Cells were fixed by 4% PFA for 20 min at RT followed by PBS wash for three times (5 min each). The cells were then incubated by the blocking buffer (0.2% TritonX-100 and 5% BSA dissolved in PBS) for 1 h. After PBS wash for three times, the cells were then incubated with the primary antibodies of NeuN (Abcam, 1:300,), dissolved in the blocking buffer at 4 °C for 24 h. PBS wash was carried out followed by the treatments with the Alexa Fluor 488/568-conjugated secondary antibody (Abgent, 1:1000) dissolved in the blocking buffer for 2 h at RT. After incubation with secondary antibodies, cells were washed and counterstained with DAPI solution (1 μg/ml in the blocking buffer) for 10 min at RT. Cells were analyzed by fluorescence microscope (Leica, IX71).
Quantitative real-time PCR
Total RNA was isolated from the cortical brain tissue or primary cultured neurons, lysed in Trizol reagent (Invitrogen) to recover the total RNA according to the manufacturer’s protocol (Trizol™ Reagent, Invitrogen). The total RNA was reversely transcripted to cDNA using a HiFi-MMLV cDNA First Strand Synthesis Kit (CW Bio, China). Quantitative real-time PCR was performed using GoTaq qPCR Master Mix (Promega) on the CFX96TM Real-Time System (Bio-Rad). GAPDH was amplified in parallel as an internal control. Primer sequences as follow, TLR4 F: 5′- TCACAACTCGCCCAAGGAGGAA -3′, R: 5′- AAGAGACCACGGCAGAAGCTAG -3′; MyD88 F: 5′- CCACCTGTAAAGGCTTCTCG -3′, R: 5′- CTAGAGCTGCTGGCCTTGTT -3′; NLRP3 F: 5′- GCTAAGAAGGACCAGCCAGAGT- 3′, R: 5′- GAACCTGCTTCTCACATGTCGT -3′; GAPDH: F: 5′- AACTTTGGCATTGTGGAAGG -3′ R: 5′- GGATGCAGGGATGATGTTCT -3′.
Statistical analysis
All the data were presented as the mean ± SEM and were analyzed using SPSS statistical software (version 22.0, IBM). Non-parametric data from the mNSS test were analyzed using the Kruskal-Wallis H analysis followed by a Mann-Whitney U test. One-way analysis of variance (ANOVA) with repeated measures was used to analyze Rotarod data. The remaining biochemical data were analyzed using a two-way ANOVA. Each experiment was repeated three times, and significance was determined with two-tailed Student’s t test or one-way ANOVA. Error bars represent ± SEM. p value lower than 0.05 was considered as significant.
Discussion
Results of this study demonstrate that pyroptosis modulates responses to the acute phase of TBI through caspase-1. Caspase-1 is expressed and activated in neurons and therefore is likely to be activated via the mechanical damage. During the acute post-traumatic period (0–48 h), neurological deficits, inflammatory cytokine, and pyroptosis are greater in the brain-injured mice. To date, there are scant data regarding the regulating of pyroptosis during the acute phase of TBI. Therefore, we performed a time-effect study for damage on post-TBI sensory-motor deficits and motor dysfunctions, which has significantly higher mNSS score and minimum Rotarod test score at 24 h. We found that acute injury significantly reduced neuron functional deficits as early as 24 h post-TBI. For instance, a previous study using this murine TBI model observed higher neurological deficits at 24 h than at 72 h and found that brain injury is most pronounced after 24 h [
26]. Inflammatory correlation factors are also detected in the acute phase. A range of anti-inflammatory cytokines and pro-inflammatory cytokine secretion were all increased and continued high secretion from 24 to 48 h, such as TGF-β, IL-10, IFN-γ, IL-1β, and IL-18. Simultaneously, mRNA of the upstream inflammatory regulator TLR4, MyD88, and NLRP3 are also highly expressed at 24 h post-TBI. Apoptosis and immunostaining of the damaged cortex further validated our murine TBI model. There was significant apoptosis, expression of caspase-1, caspase-11, and GSDMD proteins around the contusion site, and blood LDH release at 24 h post-TBI. These data suggest that pyroptosis is likely involved in neuroinflammation with TBI. Apoptoses mediated by caspase-induced cleavage events are key features of pyroptosis [
27]. Adamczak et al. verify neuronal pyroptosis which is mediated by the AIM2 inflammasome, is an important cell death mechanism during the CNS infection and injury [
28]. NLRP3 inflammasome plays a key role in the secondary phase of TBI, and lack of the NLRP3 inflammasome will improve recovery following TBI [
29‐
31]. AIM2 and NLRP3 as sensors, they are involved in activation of caspase-1 which catalyze neuroinflammation and pyroptosis. However, we directly demonstrate whether pyroptosis involves the acute phase of TBI and whether caspase-1 deficiency will reduce neuroinflammation and neuron injury in vivo and in vitro.
Pyroptosis drives by noncanonical and canonical inflammasomes [
32]. In the process of pyroptosis, Toll-like receptor (TLR)- and/or interferon (IFN)-mediated priming upregulates the expression of guanylate binding proteins critical for bacterial vacuole lysis and/or pattern-associated molecular pattern (PAMP, including LPS and bacterial DNA) exposure, sensors (NLRP3, caspase-1, caspase-11), and cytokine precursor (pro-IL-1β, pro-IL-18) [
33,
34]. The ability of inflammatory caspases to promote cell lysis involves the cleavage of GSDMD to promote the formation of membrane pores [
35]. Membrane disruption leads to pyroptosis lytic cell death, releasing alarmins and pro-inflammatory cytokines [
36,
37]. Noncanonical inflammasome also controls caspase-1 activation mediated by the NLRP3 canonical inflammasome through pannexin-1 and GSDMD cleavages in a cell-intrinsic manner involving K
+ efflux [
38‐
40]. Furthermore, some studies have shown that the pyroptosis is usually along with the released cytoplasmic LDH [
41,
42]. All those suggest that pyroptosis involves in the acute phase of TBI, through increasing caspase-1 activity, caspase-11, GSDMD expression, and blood LDH release. While in caspase-1
−/−-TBI mice, caspase-1 deficiency impels pyroptosis decay, less LDH release, neuroinflammation, neurological deficits, and neuronal death in the acute phase.
On the cell level, we stimulated primary neuron through scratch, stretch, or LPS/ATP, to discuss the mechanism of pyroptosis on TBI. Mechanical scratch can directly damage the neurons; inflammatory cytokines in damaged neurons may release rapidly and thus affect the activity of other normal neurons. As for the mechanical stretch, it is commonly used to study the biomechanical stimulation of cardiovascular disease and bone-related diseases [
43,
44]. The neuron/axles may be damaged under equiaxial stretch (12% strain, 1.0 Hz frequency) for 4 h. In some reports, LPS usually has been used to induce inflammation and pyroptosis of macrophages [
23,
36]. In this work, we tried to incur pyroptosis by LPS/ATP in the primary neuron. Scratch or stretch and LPS/ATP, all caused neuron damage, neuron apoptosis, LDH leakage, and a lot of inflammatory cytokine release, and pyroptosis-related proteins were also triggered. Caspase is a protease involved in cell apoptosis, with more than ten subtypes. Caspase-1 regulates cell apoptosis by shearing Bcl-XL and the inflammatory response of related cytokines which are mediated by the shear of the precursors of cytokines [
45]. Caspase-1 is the unique caspase that produces active cytokines in the precursors of IL-1β and IL-18 [
19]. Caspase-11 shears 45kd-caspase-1 precursor protein to produce p10 and p20 fragments, which form a heterogenous dimer and then form a tetrameric active enzyme [
46]. In the case of caspase-1 suppression, caspase-1 activity was significantly inhibited, it weakens the injure and LDH leakage of neurons which induced by scratch, stretch, or LPS/ATP, and it decreased the expression or the release of inflammatory factors. Although previous studies had suggested that caspase-1 was the main enzyme implicated in the cleavage of pro-IL-1β and pro-IL-18 into the biologically active cytokine, caspase-1 and its fragments (p10 and p20) became less and mature of IL-1β and IL-18 reduced when caspase-1 was suppressed in the primary neuron. In the meantime, caspase-1 suppression aroused decrease of GSDMD, which formed large pores in the membrane that drive swelling and membrane rupture. Ultimately, caspase-1 deficiency reduced secretion of mature IL-1β and IL-18, which abated neuroinflammation [
47].
Our results align with what is known about the role of caspase-1 in other brain injury responses. Previous researches show that caspase-1 KO mice and transgenic mice expressing a dominant-negative caspase-1 construct are partially resistant to cerebral disease and stroke. In the permanent stroke model, the proteolytic activity of caspase-1 is increased 30 min after occlusion with a second wave of activation 12 h later [
48]. The increase of caspase-1 expression has been described in neurons and astrocytes after thromboembolic stroke and observed later in microglia (24 h post-stroke) [
49]. Other studies have demonstrated the role of caspase-1 after stroke using transgenic mice. Caspase-1 KO mice and mice expressing dominant-negative caspase-1 exhibit a reduction of brain damage relative to wild type after experimental stroke [
50]. Moreover, intracerebroventricular administration of caspase-1 inhibitors provides protective effects in experimental stroke models [
51,
52]. Beneficial effects of caspase-1 inhibition are also observed in a model of oxygen and glucose deprivation in rat organotypic hippocampal slices [
53]. Some studies indicate that the caspase-1 mRNA expression in the cortex was significantly increased in AD patients [
54]. Altogether, these studies as well as our study show that caspase-1 is a key regulator of underlying mechanisms of brain injury and disease.
Our study experimentally demonstrates that TBI induces pyroptosis, and caspase-1 deficiency reduces neuroinflammation and neuronal damage in the acute phase. These findings may provide strategies for clinical treatment of TBI and other neuroinflammatory conditions. There is an increasing realization that neuroinflammation occurs within the whole gamut of the central nervous system pathologies [
55]. Regulating inflammation to control TBI damage is possible through drug treatments [
56]. NLRP3 inflammasome and interleukin-1 receptor antagonist treatments have been used to mitigate neuroinflammation after TBI [
30,
57]. These treatments were able to alleviate the nerve damage and promote neurological recovery after TBI. Electro-acupuncture-induced TLR4 signaling inhibition has been used to promote hippocampal neurogenesis and neurological recovery post-trauma [
58].