Repetitive Mild Traumatic Brain Injury and Transcription Factor Modulation

Abstract

The worldwide incidence of traumatic brain injury (TBI) is ~0.5% per year and the frequency is significantly higher among military personnel and athletes. Repetitive TBIs are associated with military and athletic activities, and typically involve more severe consequences. The majority of TBIs are mild; however, these still can result in long-term cognitive deficits, and there is currently no effective treatment. tert-Butylhydroquinone (tBHQ) and pioglitazone can activate the nuclear factor (erythroid-derived 2)-like 2 (Nrf2) and peroxisome proliferator-activated receptor-gamma (PPAR-γ) transcription factors, respectively, and each has been shown to be neuroprotective in various model systems. We examined behavioral and gene expression changes after repetitive mild TBI followed by simultaneous treatment with both factors. We used a repetitive closed head injury of mice involving five injuries with a 1-week interval between each TBI. We found that memory performance was significantly reduced by the injuries, unless the TBIs were followed by the tBHQ and pioglitazone administrations. Certain genes; for example, growth hormone and osteopontin, were downregulated by the injury, and this was reversed by the treatment, whereas other genes; for example, a tumor necrosis factor receptor, were upregulated by the injury and restored if the post-injury treatment was administered. Analysis of gene expression levels affected by the injury and/or the treatment point to potential mechanisms that could be exploited therapeutically.

Introduction

Approximately 1,700,000 people have a traumatic brain injury (TBI) in the United States every year.¹ In particular, military personnel and athletes have an increased risk, especially as it relates to repeated injuries, which has resulted in increased public interest in recent years.^2,3 Although the vast majority of all TBIs are mild, it is estimated that ~10% involve long-term complications, both cognitive and pathological, and there is no known effective treatment for these deficits.^4–8 Perhaps the most common long-term problem encountered by numerous TBI patients is memory impairment.^9,10 The hippocampus, which plays an important role in the formation of new memories, has been shown to be susceptible to mechanical injury, which may account for the impairment shown with TBI.¹¹ Animal models of closed head injury are able to recapitulate the impact forces, pathology, and cognitive deficits observed in humans with TBI.¹²

Regulation of transcription factor signaling could be useful target in protecting the brain against further damage caused by neuroinflammation and oxidative stress. Pioglitazone is an agonist for the transcription factor peroxisome proliferator-activated receptor-gamma (PPAR-γ) originally developed to treat diabetes. Pioglitazone has been shown to reduce neuroinflammation in rat models of brain injury.^13,14 A single dose of pioglitazone administered after injury has been shown to reduce cortical oxidative damage and microglial responses.¹⁵

Nuclear factor (erythroid-derived 2)-like 2 (Nrf2) is another transcription factor of interest. Nr2 has been shown to activate downstream antioxidant and oxidative stress genes by binding to their antioxidant response element promoter (ARE) region.¹⁶ Nrf2 gene expression increases in the cortex and hippocampus up to 72 h following controlled cortical impact in CF-1 male mice, resulting in the increased expression of ARE-containing genes.¹⁷ In the same injury model, treatment with Nrf2 activator carnosic acid IP 1.0 mg/kg 15 min prior to injury reduced oxidative stress and protected mitochondrial electron transport chain complex I and complex II activity 24 h after the injury.¹⁸ Conversely, In C57BL/6 mice, Nrf2 expression was found to be downregulated 24 h following a fluid percussion injury, and injury was worse in C57BL/6 Nrf2^(−/−) mice than in C57BL/6 Nrf2^(+/+) wild-type (WT) mice.¹⁹ Twenty-four hours after injury, Nrf2 downregulation was accompanied by activation of matrix metallopeptidases (MMPs), transforming growth factor (TGF)-β1, and nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB), resulting in increased neuroinflammation.¹⁹

Using the chemical activator of Nrf2, tert-butylhydroquinone (tBHQ), a common food additive, we have shown previously that post-injury activation of the Nrf2 pathway leads to improved cognitive performance in a mouse model.²⁰ Simultaneous activation of both the PPAR-γ and Nrf2 pathways following brain injury may provide a synergistic effect that leads to further improved outcomes, both behaviorally and pathologically.

Our mouse model utilizes a repeated 30 g weight drop followed by a final 50 g weight drop that delivers a noninvasive, closed head injury to the skull.^12,21 Using this paradigm, we seek to model the repeated nature and variability of mild TBIs (mTBIs) received by humans. This model is able to accurately mimic the rotational forces, neuropathological changes, and cognitive and behavioral symptoms that are observed in humans with TBI.^12,21 The model delivers a mild injury that does not cause gross morphological changes to the skin, skull, or brain, or cause edema. There are, however, marked molecular changes and changes to the morphology of individual neurons and glial cells.^20,22–24 Using this model, we found that multiple mild injuries to the brain result in bio-chemical changes in the form of gene dysregulation, which may account for the subsequent long-term deficits that were observed in recognition memory. Simultaneous activation of Nrf2 and PPAR-γ transcription factors with treatment by tBHQ and pioglitazone appears to ameliorate these effects.

In this study, we sought to determine the therapeutic potential of combined tBHQ and pioglitazone treatment following a repetitive closed head TBI. We report gene dysregulation resulting from repetitive TBI and also gene expression changes induced by combined tBHQ and pioglitazone treatment, which may be responsible for the improved behavioral outcome. It may be possible that the multifactor effects of TBI can only be countered by modulating more than one potentially neuroprotective pathway.

Methods

Animals

The animal protocol was approved by the Bay Pines VA Institutional Animal Care and Use Committees (IACUC) and performed in accordance with all institutional, agency, and governmental animal welfare regulations. The ethics committee of the Sackler Faculty of Medicine also approved the experimental protocol, in compliance with the guidelines for animal experimentation of the National Institutes of Health.

Male C57BL6/J mice at 5 weeks of age were obtained from Jackson Laboratories (Bar Harbor, ME) and Male Institute of Cancer Research (ICR) mice of ~5 weeks of age were obtained from Harlan Sprague Dawley, Inc. (now Envigo, Jerusalem, Israel). They were housed three to four per cage in a 22°C ± 0.5°C temperature-controlled environment with a 12 h light/dark cycle. Food and water were available ad libitum. All mice were allowed to acclimate to the facility for 1 week prior to experimentation.

Closed head injury

mTBI was induced using a concussive closed head injury weight drop model, described previously.¹² The apparatus consisted of a guide tube which was 80 cm in length with a 13 mm internal diameter. Mice were anesthetized using isoflurane, and placed on a sponge, allowing rotation of the head upon impact, to inflict a diffuse concussive injury.²¹ The sponge was located directly below the apparatus, and the mouse was placed such that the end of the guide tube spanned the right hemisphere caudal to the eye and rostral to the ear. A 30 g (first four injuries) or 50 g (final injury) cylindrical weight was dropped 80 cm through the length of the guide tube, impacting the right side of the head, in injured mice. Sham mice were anesthetized and placed on the sponge for an equivalent length of time, but no weight was dropped. This procedure, which has been demonstrated to cause a mild injury,²² was repeated five times with each injury occurring 1 week apart (Fig. 1). Following injury or sham, mice were monitored in a separate chamber while recovering from anesthesia before being returned to regular housing. The weight used for injury began at 30 g, to accommodate the small size of the C57BL6/J mice at 6 weeks of age (~ 20 g on average). At the time of the final injury, we utilized a 50 g weight to deliver a slightly stronger injury, mimicking the variability present in humans with repeated TBIs.

FIG. 1.

Experimental timeline. Mice were exposed to a closed head injury once per week for 5 weeks, beginning at ~6 weeks of age. The closed head injury involved dropping a weight through an 80 cm tube where it then impacted the right temporal region of the skull. The first four injuries utilized a 30 g weight, with the fifth and final injury using a 50 g weight. Approximately 30 min after each injury, mice were treated with a combination of tert-Butylhydroquinone (tBHQ) (33.4 mg/kg) and pioglitazone (3 mg/kg) or vehicle. In one group, brains were harvested for RNA analysis 1 day post-injury. In a second group, mice underwent behavioral testing 8 weeks post-injury.

Combination treatment

Approximately 30 min after each of the five injuries (or sham), mice were treated via intraperitoneal injection with a combination treatment containing 33.4 mg/kg tBHQ and 3 mg/kg pioglitazone or a 5% dimethyl sulfoxide (DMSO) in saline vehicle.²⁰ Injection solutions were made fresh on the day of injection. tBHQ and pioglitazone were initially dissolved in sterile DMSO separately at 200 mg/mL and 20 mg/mL concentrations, respectively. The solutions were then mixed and further diluted in sterile saline to the desired concentration (5 mg/mL tBHQ, 0.449 mg/mL pioglitazone). Additional DMSO was added to achieve a 5% total concentration. Total injection volume ranged from 134 μL to 234 μL, according to mouse body weight. As mice grew, injection volume increased to maintain the treatment at the same dosage per kilogram.

Novel object recognition (NOR)

NOR was performed on ICR mice 8 weeks following the final closed head injury. Three 5 min sessions, separated by 24 h, were performed in a 60 cm × 60 cm square arena. On the 1st day, each individual mouse was allowed to explore the empty arena. On the 2nd day, two identical objects were placed in the arena and each mouse was allowed to explore the objects. On the 3rd day, one of the objects was replaced with a novel object, and the time spent near each object was recorded. The arena was thoroughly cleaned with 70% ethanol between mice, and the order of mice undergoing behavioral testing alternated between experimental groups. A preference index was calculated as the time the animal spent near the novel object minus the time spent near the familiar object, divided by the sum of the time near the novel and familiar objects. These methods have been described previously.^25–29

Elevated plus maze

Anxiety level was assessed using the elevated plus maze for ICR mice 8 weeks following the final closed head injury. The apparatus consisted of two open arms and two closed arms which were enclosed on all sides with the roof open (30 × 15 × 5 cm). The arms were arranged such that matching arms were opposite each other in a “+” shape. The maze was elevated above the floor level by 60 cm.³⁰ During testing, mice were placed in the center of the apparatus facing one of the open arms. The mouse was then allowed to explore the maze for 5 min, and time spent in each arm was recorded. Between animals, the apparatus was cleaned with a 70% ethanol solution. These methods have been described previously.³¹

RNA extraction

One day following the final closed head injury, C57BL6/J mice were euthanized by cervical dislocation and brains were harvested. C57BL6/J mice were chosen for the genetic analysis as they are an inbred strain that would produce less variation in gene expression between experimental animals. The hippocampus was dissected out and snap-frozen using liquid nitrogen. The tissue was pulverized and underwent TRIzol (Qiagen, Hilden, Germany) separation to extract RNA. Pulverized tissue was resuspended in 1 mL TRIzol and homogenized using a polytron homogenizer (PRO Scientific, Oxford, CT). The homogenized tissue was incubated at 23°C for 5 min and then stored on ice. Next, 200 μL CHCl₃ was added, vortexed, and incubated at room temperature for 2–3 min. The tubes were then spun in a refrigerated centrifuge set to 4°C for 10 min at 12,000 relative centrifugal force (RCF) to separate phases. The top layer was removed, and an equal volume of 70% ethanol was added. The Qiagen RNeasy kit (Qiagen, Hilden, Germany) was used to further purify the RNA. The samples were stored at −80°C until use.

Gene expression analysis

Following RNA extraction, samples containing RNA from the ipsilateral hippocampus were confirmed as intact with an Agilent Bioanalyzer and hybridized to Affymetrix GeneChip MOE430 2.0 arrays (Affymetrix, Santa Clara, CA) at the Moffitt Cancer Center Molecular Genomics Core Facility (Tampa, FL). Although earlier unpublished results suggest that there may be some partial effects in the contralateral hippocampus, this analysis focuses only on the ipsilateral hippocampus in injured versus sham and combination-treated versus vehicle-treated mice. We then analyzed the array data with the aid of GeneSpring software version 14.9.1 (Agilent Technologies, Santa Clara, CA). Each group contained samples from six different brains for a total of 24 different mice examined.

Statistical analysis

Mean values are depicted ± standard deviation with p < 0.05 indicating significance. Analysis of variance (ANOVA) was used for analysis of the behavioral data. The Mann–Whitney U Test was used with median expression values to determine significant fold changes in RNA expression.

Results

Behavioral changes after multiple closed head injuries

Eight weeks after the final injury, recognition memory was tested using the NOR behavioral test. This tests recognition memory by utilizing a mouse’s innate preference toward exploring a novel object. Over the course of 3 days, mice were first exposed to a bare arena and then to two identical objects. Then one of the objects was replaced with a novel object. A preference toward the novel object indicated that a mouse recognized that the object was new, by remembering the objects from the previous day. We express a preference toward the novel object as a preference index, which is the time spent with the novel object minus the time spent near the familiar object divided by the total time spent with either object. In our experiment, sham mice, treated or untreated, had a preference index of ~0.2–0.25, indicating healthy recognition memory. Mice that had received the closed head injuries with only the vehicle treatment showed no preference for the novel object, indicating a loss of recognition memory 8 weeks post-injury, which was statistically significant. When injured mice were treated with tBHQ and pioglitazone, the preference index was ~0.2, not statistically different from the sham preference indices (Fig. 2). This suggests that the treatment, administered 30 min post-injury, protected against long term memory loss following injury. Elevated plus maze revealed no significant differences in anxiety levels between groups (data not shown).

FIG. 2.

Behavioral effects. Mice underwent testing on the novel object recognition task 8 weeks following the fifth closed head injury. Mice receiving an injury, but only vehicle treatment, did not show a preference for a novel object. When injured mice were treated with tert-Butylhydroquinone (tBHQ) and pioglitazone, the preference for the novel object was comparable to that of uninjured sham mice.

Gene dysregulation after multiple closed head injuries and treatment

Twenty-four hours after the fifth injury, brains were harvested, and the ipsilateral hippocampus was dissected for RNA analysis. Analysis of the data identified numerous genes that were dysregulated 1.3-fold at the mRNA level in response to injury or treatment. Interestingly, along with genes exhibiting changed gene expression as a result of injury or treatment alone, there were several genes that changed both when injured mice were compared with sham mice and when injured treated mice were compared with injured vehicle mice. We found 24 genes that were downregulated with injury and subsequently upregulated by the tBHQ and pioglitazone combination treatment, and 6 genes that were upregulated with injury and downregulated with the treatment (Fig. 3). A full list of the named dysregulated genes is shown in Table 1.

FIG. 3.

Dysregulated genes. RNA harvested from the ipsilateral hippocampus 1 day post-injury was evaluated using the Affymetrix GeneChip MOE430 2.0 array. Numerous genes were found to be dysregulated in response to injury and/or treatment. In particular, there were 24 genes that were downregulated with injury and subsequently upregulated with treatment, and 6 genes that showed the opposite regulation pattern.

Table 1.

Dysregulated Genes

		Fold change
Gene symbol	Gene title	Injured vehicle vs. sham vehicle	Injured treated vs. inured vehicle
Zmyml	Zinc finger, MYM domain containing 1	1.59	−1.34
Mfsdl2	Major facilitator superfamily domain containing 12	1.59
Sdf2l1	Stromal cell-derived factor 2-like 1	1.51
Tgfblil	Transforming growth factor beta 1 induced transcript 1	1.49
Slc39a2	Solute carrier family 39 (zinc transporter), member 2	1.46	−1.36
Rrp15	Ribosomal RNA processing 15 homolog (Saccharomyces cerevisiae)	1.46
Myrf	Myelin regulatory factor	1.45
Tnfrsf25	Tumor necrosis factor receptor superfamily, member 25	1.44	−1.31
Hspa1b	Heat shock protein 1B	1.43
Cry1	Cryptochrome 1 (photolyase-like)	1.40
Manf	Mesencephalic astrocyte-derived neurotrophic factor	1.36
C2cd4c	C2 calcium-dependent domain containing 4C	1.35	−1.42
Zbtb40	Zinc finger and Broad-Complex, Tramtrack and Bric a brac (BTB) domain containing 40	1.34
Gpr17	G protein-coupled receptor 17	1.33
Xbp1	X-box binding protein 1	1.33
Abca7	Adenosine triphosphate (ATP)-binding cassette, sub-family A (ABC1), member 7	1.32
Hspb1	Heat shock protein 1	1.31
Cckbr	Cholecystokinin B receptor	1.31	−1.57
Tmem200a	Transmembrane protein 200A	1.31
Cys1	Cystin 1	1.30
Stard8	Steroidogenic acute regulatory protein (StAR)-related lipid transfer (START) domain containing 8	1.30
Hip1r	Huntingtin interacting protein 1 related	−1.31	1.39
Coch	Coagulation factor C homolog (Limulus polyphemus)	−1.31	1.57
Zfhx4	Zinc finger homeodomain 4	−1.31	1.36
Ppp1r1b	Protein phosphatase 1, regulatory (inhibitor) subunit 1B	−1.31	1.46
Fam84b	Family with sequence similarity 84, member B	−1.32	1.35
Lum	Lumican	−1.33	1.31
Slc25a24	Solute carrier family 25 (mitochondrial carrier, phosphate carrier), member 24	−1.33	1.45
Slc6a20a	Solute carrier family 6 (neurotransmitter transporter), member 20A	−1.35	1.45
Folr1	Folate receptor 1 (adult)	−1.41	−1.51
Col1a2	Collagen, type I, alpha 2	−1.41	1.64
Sowahc	Sosondowah ankyrin repeat domain family member C	−1.42	1.33
Mgp	Matrix Gla protein	−1.46	1.37
Prlr	Prolactin receptor	−1.47	1.32
Lce1i	Late cornified envelope 1I	−1.48	1.33
Enox1	Ecto-NOX disulfide-thiol exchanger 1	−1.55	1.60
Ogn	Osteoglycin	−1.55	1.56
Oca2	Oculocutaneous albinism II	−1.59	1.41
Cldn2	Claudin 2	−1.62	1.47
SPP1	Secreted phosphoprotein 1	−1.67	3.07
Gm5802	Predicted gene 5802	−1.68	2.21
Slc13a4	Solute carrier family 13 (sodium/sulfate symporters), member 4	−1.68	1.70
Vat1l	Vesicle amine transport protein 1 homolog-like (T. californica)	−1.95	1.58
Gh	Growth hormone	−2.83	2.48
Mob3b	MOB kinase activator 3B		1.90
Gjb2	Gap junction protein, beta 2		1.79
Zic1	Zinc finger protein of the cerebellum 1		1.59
Tcf7l2	Transcription factor 7 like 2, T cell specific, high mobility group (HMG) box		1.54
Ifitm3	Interferon induced transmembrane protein 3		1.54
Nefh	Neurofilament, heavy polypeptide		1.53
Slc8a3	Solute carrier family 8 (sodium/calcium exchanger), member 3		1.52
Ptgds	Prostaglandin D2 synthase (brain)		1.51
Pik3cd	Phosphatidylinositol 3-kinase catalytic delta polypeptide		1.51
A2m	Alpha-2-macroglobulin		1.49
Nek2	Never in mitosis gene a (NIMA)-related expressed kinase 2		1.46
Serping1	Serine (or cysteine) peptidase inhibitor, clade G, member 1		1.44
Agt	Angiotensinogen (serpin peptidase inhibitor, clade A, member 8)		1.42
Ppif	Peptidylprolyl isomerase F (cyclophilin F)		1.42
Kin	Antigenic determinant of rec-A protein		1.40
Rpap3	RNA polymerase II associated protein 3		1.38
LOC102638850///Rps21	40S ribosomal protein S21 pseudogene///ribosomal protein S21		1.37
Usmg5	Upregulated during skeletal muscle growth 5		1.37
Slc13a3	Solute carrier family 13 (sodium-dependent dicarboxylate transporter), member 3		1.37
Cpox	Coproporphyrinogen oxidase		1.36
Pcolce	Procollagen C-endopeptidase enhancer protein		1.36
Gm13889	Predicted gene 13889		1.35
Vasn	Vasorin		1.35
Slc31a2	Solute carrier family 31, member 2		1.34
Ahnak	AHNAK nucleoprotein (desmoyokin)		1.34
Ptrf	Polymerase I and transcript release factor		1.33
Nbn	Nibrin		1.33
AW112010	Expressed sequence AW112010		1.32
Klhl4	Kelch-like 4		1.32
Efemp1	Epidermal growth factor-containing fibulin-like extracellular matrix protein 1		1.32
Sipa1l3	Signal-induced proliferation-associated 1 like 3		1.31
Vim	Vimentin		1.31
Adgrg1	Adhesion G protein-coupled receptor G1		1.31
B630019K06Rik	Novel protein similar to F-box and leucine-rich repeat protein 17 (Fbxl17)		1.31
Acta2	Actin, alpha 2, smooth muscle, aorta		1.31
Ndn	Necdin		1.31
Cnksr2	Connector enhancer of kinase suppressor of Ras 2		1.31
Mex3b	Mex3 homolog B (Caenorhabditis elegans)		1.31
9430047L24Rik	RIKEN cDNA 9430047L24 gene		1.30
Rubcn	RUN domain and cysteine-rich domain containing, Beclin 1-interacting protein		−1.30
Vwc2l	von Willebrand factor C domain-containing protein 2-like		−1.31
Rxra	Retinoid X receptor alpha		−1.31
Sema4b	Sema domain, immunoglobulin domain (Ig), transmembrane domain (TM) and short cytoplasmic domain, (semaphorin) 4B		−1.31
Plagl1	Pleiomorphic adenoma gene-like 1		−1.32
Cpne7	Copine VII		−1.34
Gtpbp3///LOC102642476	Guanosine-5’-triphosphate (GTP) binding protein 3///tRNA modification GTPase GTPBP3, mitochondrial		−1.35
Gtpbp3///LOC102642476	GTP binding protein 3///tRNA modification GTPase GTPBP3, mitochondrial		−1.35
Rpp25	Ribonuclease P/mitochondrial RNA processing (MRP) 25 subunit		−1.36
Ctgf	Connective tissue growth factor		−1.37
Tshz3	Teashirt zinc finger family member 3		−1.38
Satb2	Special AT-rich sequence binding protein 2		−1.39
Copg2os2	Coatomer protein complex, subunit gamma 2, opposite strand 2		−1.41
Gtpbp2	GTP binding protein 2		−1.41
Klf5	Kruppel-like factor 5		−1.43
Gm3764	Predicted gene 3764		−1.43
Fxyd6	FXYD domain-containing ion transport regulator 6		−1.44
Glra3	Glycine receptor, alpha 3 subunit		−1.45
Tshz2	Teashirt zinc finger family member 2		−1.45
Nts	Neurotensin		−1.47
Mid1	Midline 1		−1.48
Cpne2	Copine II		−1.50
St7	Suppression of tumorigenicity 7		−1.55
Trhr	Thyrotropin releasing hormone receptor		−1.77
G530011O06Rik	RIKEN cDNA G530011O06 gene		−1.96
Erdr1	Erythroid differentiation regulator 1		−2.38

Discussion

Our closed head injury model of repeated mild TBI investigated the behavioral and molecular changes following TBI, as well as the potential neuroprotective effects of a combination treatment targeting the Nrf2 and PPAR-γ transcription factors. We delivered a mild TBI to mice once per week for 5 weeks, followed by treatment 30 min post-injury. In a previous study of single injury using this weight-drop model of mTBI, recognition memory deficits were observed 7 days post-injury.³² In this study, we see injured mice continuing to display these deficits in recognition memory when compared with uninjured sham mice, even 8 weeks after the final injury. When mice were treated with our combination of tBHQ and pioglitazone, these memory deficits were not observed. This suggests that our treatment protected against the longer-term memory loss induced by repeated mild TBIs.

To better understand the molecular impacts of TBI and our treatment following injury, especially as they related to memory loss, we utilized an RNA array to investigate transcriptional changes within the ipsilateral hippocampus. We identified numerous genes that were dysregulated either in response to injury or in response to treatment, most of which were not identified in a previous study using this model.³² In particular, we identified several genes of interest whose expression appeared to be changed by both injury and treatment. One of these genes is secreted phosphoprotein 1 (SPP1) or osteopontin, which was downregulated 1.7-fold with injury when compared with sham controls, and upregulated 3-fold by the treatment following injury when compared with injured and vehicle-treated controls. Osteopontin has been identified in previous studies as having a neuroprotective role in models of stroke,³³ cerebral hemorrhage,³⁴ and focal cerebral ischemia.³⁵ In TBI models, osteopontin has been shown to mediate hippocampal synaptogenesis and recovery.³⁶ One recent study using a rat model of TBI showed that intranasal treatment with osteopontin activated the microglial response following TBI, and may protect against secondary consequences of TBI at later time points.³⁷ Growth hormone (GH) was downregulated 2.8-fold with injury, and upregulated 2.5-fold with treatment. Studies have shown increased incidence of GH deficiency following chronic TBI, and that this deficiency is associated with poorer clinical outcomes.³⁸ Moreover, GH replacement therapy has been shown to improve cognitive outcomes following TBI in human patients.³⁹

Alternately, we observed a 1.4-fold increase with injury and a 1.3-fold decrease with treatment in expression of tumor necrosis factor (TNF) receptor subfamily 25 (TBFRSF25). One recent mouse model showed that TNF exacerbates the effects of TBI, and that inhibition of the TNF receptors could be effective therapeutic targets following injury.⁴⁰ Our mouse model was resulted in decreases in osteopontin and GH expression, and increases in a TNF receptor following injury. Our treatment ameliorated these effects, which suggests a possible mechanism of action for the improved behavioral outcomes observed with treatment.

Our mouse model of repeated mild TBI was able to recapitulate the long-term memory deficits observed in humans with TBI.^9,10 Our hippocampal RNA investigation provides insight into the initial molecular changes that could lead to these long-term deficits, as well as how our treatment is able to modulate these changes to protect against memory deficits. We have identified several important factors that are modulated by both injury and our treatment, which may be important in the search for more effective therapeutic targets for the treatment of TBI. Future studies could further elucidate the mechanisms by which injury, as well as tBHQ and pioglitazone, influence the genetic factors that we have identified. Additional studies could also investigate differences in gene changes in the contralateral versus the ipsilateral hippocampus.