Systemic Elevation of n-3 Polyunsaturated Fatty Acids (n-3-PUFA) is Associated with Protection against Visual, Motor and Emotional Deficits in Mice following Closed-Head Mild Traumatic Brain Injury

Abstract

Traumatic brain injury (TBI) causes neuroinflammation and neurodegeneration leading to various pathological complications such as motor, and sensory (visual) deficits, cognitive impairment, and depression. N-3 polyunsaturated fatty acid (n-3 PUFA) containing lipids are known to be anti-inflammatory, whereas the sphingolipid, ceramide (Cer) is an inducer of neuroinflammation and degeneration. Using Fat1+-transgenic mice that contain elevated levels of systemic n-3 PUFA, we tested whether they are resistant to mild TBI-mediated sensory-motor and emotional deficits by subjecting Fat1-transgenic mice and their WT littermates to focal cranial air blast (50-psi) or sham blast (0-psi, control). We observed that visual function in WT mice was reduced significantly following TBI but not in Fat1+-blast animals. We also found Fat1+-blast mice were resistant to the decline in motor functions, depression, and fear-producing effects of blast, as well as the reduction in the area of oculomotor nucleus and increase in activated microglia in the optic tract in brain sections seen following blast in WT mice. Lipid and gene expression analyses confirmed an elevated level of the n-3 PUFA eicosapentaenoic acid (EPA) in the plasma and brain, blocking of TBI-mediated increase of Cer in the brain, and decrease in TBI-mediated induction of Cer biosynthetic and inflammatory gene expression in the brain of the Fat1+ mice. Our results demonstrate that suppression of ceramide biosynthesis and inflammatory factors in Fat1+-transgenic mice is associated with significant protection against the visual, motor, and emotional deficits caused by mild TBI. This study suggests that n-3 PUFA (especially, EPA), has a promising therapeutic role in preventing neurodegeneration after TBI.

INTRODUCTION

Mild traumatic brain injury (TBI) is a common occurrence during military combat, recreational activities, sports, and vehicular accidents [1,2]. Mild TBI-related symptoms reported by patients include increased irritability, emotional lability, anxiety, and depression [3]. The injury associated with mild TBI appears to stem from the brain tissue deformation, which results from shock waves transmitted through the brain and CSF by the blast impact, or from the brain compression and expansion during rapid head acceleration-deceleration [4]. In animal models, mice exposed to mild TBI exhibit little evident neuropathology coupled with “diffuse” axonal injury that resembles characteristics of human mild TBI [5–7]. Axonal injury leads to a variety of adverse sensory, motor, cognitive, and emotional outcomes [8], and is exacerbated by neuro-inflammation stemming from the TBI [9]. In animal models of mild TBI, the primary mechanisms that lead to neuronal cell death and dysfunction involve excitotoxicity, oxidative stress, and neuro-inflammation [10]. in which the resident microglia in the brain play a significant role [11,12]. Reduction of microglial activation and inflammation using the cannabinoid receptor-2 (CB2) inverse agonist, SMM-189, has been reported to alleviate motor, visual, and emotional deficits resulting from mild TBI [8]. Oxidative stress, inflammation, and microglial activation have also been observed in other neurodegenerative diseases, such as Alzheimer’s and multiple sclerosis (MS) [13].

There is increasing evidence that supports the role of omega-3 polyunsaturated fatty acids (ω-3 or n-3 PUFAs) in reducing oxidative stress [14] and enhancing anti-inflammatory processes [15] in disease models of Alzheimer’s disease (AD) [16]. Parkinson’s disease (PD) [17], and cerebral ischemia [18]. As essential fatty acids, Linolenic acid (LNA, 18:3n3) and Linoleic acid (LA, 18:2n6), which serve as the precursors of a variety of n-3 and n-6 PUFAs, respectively, are not synthesized in the mammalian body and therefore need to be consumed from the diet. LNA is the precursor for long-chain omega-3 fatty acids such as eicosapentaenoic acid (EPA, 20:5n3) docosahexaenoic acid (DHA, 22:6n3), while long-chain omega-6 fatty acids such as arachidonic acid (AA, 20:4n6) are produced from LA [19]. The brain contains high levels of PUFAs (25–30%), which are mainly DHA (n-3 PUFA; 12–14% of total fatty acids) and AA (n-6 PUFA; 8–10% of total fatty acids) [20]. Transgenic mice harboring the fat1 gene (called Fat1⁺ mice) from C. elegans that codes for an n-3 fatty acid desaturase (which is absent in mammals) are a useful mammalian model to study the role and benefits of n-3 fatty acids in different disease models and pathological conditions [21,22]. The Fat1⁺ mice systemically convert n-6 PUFAs into n-3 PUFAs and thus have higher levels of DHA and EPA in their brain, retina, plasma and in other tissues than wild-type mice fed the same diet [23,24,21]. Studies on Fat1⁺ mice have reported that they are protected against neuronal death in experimental ischemic injury [25] and from cerebral angiogenesis in stroke models [26]. Conversely, animal models with DHA depletion exhibit a significantly slower recovery from motor deficits compared to omega-3 adequate mice after brain injury [27]. Although the normal level of EPA is far less than the level of DHA in the brain, recent advancement in clinical and animal studies support the notion that EPA plays a greater role than DHA against depression, neuroinflammation, and other neurological disorders [28–30].

Among the brain lipids, another major structural and functional class relevant to brain injury and disease includes the sphingolipids (SPLs), which have a backbone of sphingoid bases, a set of aliphatic amino alcohols that include sphingosine. The SPLs were discovered more than a century ago in the brain and are essential for the development and maintenance of the functional integrity of the nervous system [31]. SPLs have recently become an increasingly investigated group of lipids because of their bioactive properties, with the sphingolipid ceramide (Cer) shown to act as a cell death regulator, and sphingosine1-phosphate (S1P) shown to play role in cell survival [32]. Not surprisingly, defects in sphingolipid metabolism are found in numerous neurological diseases, including AD, PD, several types of epilepsy, Krabbe disease, Gaucher disease, and dementia [33]. SPLs have recently been shown to also be associated with brain injury processes, including excitotoxicity, oxidative stress, and inflammation [34,35]. A correlation and functional association between the bioactive PUFAs or their derivatives and SPLs are not well established. However, there are reports that suggest: 1) n3-PUFA attenuates ceramide lipotoxicity in metabolic syndrome and is effective against insulin resistance, and inflammation [36]; and 2) DHA can inhibit both acidic and neutral sphingomyelinase (aSMase and nSMase) in human retinal endothelial cells and supports their involvement in maintaining cellular homeostasis [37]. We hypothesize that n-3 PUFA-mediated neuroprotection may involve modulation of SPL signaling. In this study, we showed that the systemic increase in n-3 PUFA (especially, EPA) in Fat1⁺ mice was associated with a protection from mild TBI-mediated neurological and visual deficits and a reduction in the activation of ceramide signaling.

MATERIALS AND METHODS

Animals and TBI Methods:

All procedures were performed according to the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research, the National Institutes of Health, the Society for Neuroscience, and the University of Tennessee Health Science Center Guidelines for Animals in Research. All procedures, tissue harvests, and euthanasia methods were reviewed and approved by the UTHSC Institutional Animal Care and Use Committee (UTHSC IACUC). C57BL/6J mice, used as wild type (WT) control, and fat1 transgenic mice on C57BL/6J background (Fat1⁺ mice) were born and raised in the UTHSC LACU (Laboratory Animal Care Unit) vivarium and maintained from birth under cyclic light (50-100 lux, 12 h on/off, 7 a.m. to 7 p.m. CST). Fat1⁺ mice founders were a gift from Dr. Robert E. Anderson (Dean McGee Eye Institute, University of Oklahoma Health Sciences Center, Oklahoma City, OK). Fat1⁺ transgenic mice and WT mice for the present study were generated by breeding Fat1⁺ males on a C57/BL6 background and WT C57/BL6 females. The resulting offspring comprised ~50% Fat1⁺ transgenic and ~50% WT. The parents and the offspring were on a regular mouse chow diet.

Mice were exposed to mild TBI created by generating an air blast as described previously [5,8]. The over-pressure air blast was delivered by a small horizontally mounted air canon system using a modified paintball gun (Invert Mini, Empire Paintball, Sewell, NJ). A 0-psi blast was referred to as ‘sham blast’ and all experimental outcomes were compared between sham and 50-psi blasts. Mice were divided into four groups: WT-Sham (WT-0), WT-Blast (WT-50), Fat1⁺-Sham (Fat1-0), and Fat1⁺-Blast (Fat1-50) (Fig. 1). Animals were anesthetized with Avertin (400mg/kg body weight) and exposed to the blast on the left side of the cranium between ear and eye. Before the blast, the targeted head region was shaved, and a white dot was painted in the middle of the region. A foam rubber sleeve was used to place the animal inside a Plexiglass tube in such a way the targeted head region was positioned in the center of a 7.5 mm diameter hole of the tube, facing the blast cannon barrel tip. The system is arranged to expose the parietal region of the mouse head between the ear and eye to be blasted with specific air pressure, which is set by the transducer and analyzed by a Labview software (National Instruments, Austin, TX).

Fig. 1: Schemetic diagram of the experimental timeline, study groups, treatments, observations and tests, and tissue harvest.

Experimental mice were divided in 4 groups and exposed to mild TBI with a 50-psi blast (Blast) or 0-psi blast (Sham) on the left side of the cranium between ear and eye. The groups include, wild type (WT)-Sham (WT-0), WT-Blast (WT-50), Fat1⁺-Sham (Fat1-0) and Fat1⁺-Blast (Fat1-50). Mice were tested for motor, visual, cognitive function and depression at various timepoints before (Pre-Blast) and after blast (Post-Blast). OKN, Optokinetic nystagmus reflex; ERG, Electroretinography.

Optokinetic Reflex Measurement:

Mild TBI is known to cause visual deficits in mice [38]. To evaluate visual function after blast injury, optokinetic reflex measurements were conducted in WT and Fat1⁺ mice prior to and post blast using the OptoMotry system of Cerebral Mechanics, Lethbridge, AB, Canada. Visual acuity was assessed at 100% contrast by varying spatial frequency threshold, and contrast sensitivity was measured at a 0.042 cycles per degree (c/d) spatial frequency.

Electroretinogram:

The scotopic (dark-adapted) threshold electroretinogram (ERG) recordings were obtained using the Diagnosys Espion E2 ERG system (Diagnosys LLC, Lowell, MA). Mice were kept in total darkness overnight and prepared for ERG recording under dim red light. Animals were anesthetized with ketamine (100mg/kg body weight) and xylazine (5mg/kg body weight) via intraperitoneal injection. For dilating the pupil, one drop of 10% (v/v) phenylephrine was applied on the cornea, and one drop of 0.5% (v/v) proparacaine HCl was applied for local anesthesia. To keep mice warm, a heating pad at 37°C was used while recording. A gold electrode was placed on the cornea, while a reference electrode was positioned on the head, a ground electrode was placed at the tail, and the animal was placed in a Ganzfeld illuminating hemisphere. Responses were differentially amplified, averaged, and stored. The scotopic ERG assessed rod photoreceptor function with three strobe flash stimuli presented at intensities of 0.1, 1 and 10 cd.s/m². The amplitude of the a-wave was measured from the pre-stimulus baseline to the a-wave trough and the amplitude of b-wave was measured from the trough of the a-wave to the peak of the b-wave. We measured the a- and b-wave amplitudes before blast and one-month after blast.

Open-field Behavior Assessment:

To evaluate motor function open-field behavior, an automated 30-min open-field behavior assessment was conducted using a Noldus EthoVision video tracking system to record and digitize the mouse movements (Noldus Information Technology, The Netherlands), and the SEE software of Drai and Golani [39] to analyze mouse motor behavior. Open field sessions were conducted one day before the blast and 8 days after the blast following previously published protocols [12].

Tail Suspension Depression Test:

To assess depression-like behavior in mice, a tail suspension test was performed as described previously [5,8]. For this experiment, each mouse was suspended by its tail from a solid surface using adhesive tape so that its body dangled in the air, facing downward. The test was for 5 mins duration and a video tracking system interfaced to a computer and automated software (FreezeFrame, Coulbourn, Whitehall, PA, USA) was used to quantify responses. The measured behavioral score was immobility, with a depressive-like state indicated by a longer duration of immobility over the test period. Data are plotted as cumulative immobility per minute over the five-minute test.

Fear Acquisition and Extinction Tests:

Pavlovian fear conditioning is used to evaluate pathological fear responsiveness in animals [40]. Our studies employed a fear-conditioning chamber made up of polycarbonate walls and a stainless-steel grid floor (MED Associates, Model ENV-008), fitted with a video camera interfaced with a computer. After being placed in the training chamber for five minutes, mice received five fear training trials, each consisting of a 30-s tone (12kH) conditioned stimulus (CS) co-terminating with a 0.250-s, 0.4mA foot shock (unconditioned stimulus), with an intertrial interval of two minutes. Next, on each of the following three days, mice were given extinction sessions, each consisting of 15 tone-alone test trials, with freezing again the measure of fear, again at an intertrial of two minutes. The CS presentations were preceded by a three-minute assessment of responding to the fear context. Freezing responses were collected and analyzed using automated software (FreezeFrame, Coulbourn, Whitehall, PA, USA).

Immunohistochemical Analysis:

Immunohistochemical analyses were carried out on fixed brain tissue to determine the effect of TBI on: 1) the motoneurons of the oculomotor nucleus as detected by immunolabeling for the cholinergic neuron marker choline acetyltransferase (ChAT) after ~60 days post-TBI; and 2) the activation state of microglia in the optic tract as detected by immunostaining for ionized calcium-binding adapter molecule-1 (IBA1) ~60 days post TBI. Mice were deeply anesthetized with Avertin @ 0.2mL/g body weight, the chest opened, and 0.1 ml of heparinized saline (800 U.S.P units/ml) injected into the heart. They were then perfused transcardially with 40ml of 0.9% NaCl in 0.1 M sodium phosphate buffer at pH 7.4 (PB), followed by 200ml of 4% paraformaldehyde, 0.1 M lysine-0.1 M sodium periodate in 0.1 M PB at pH 7.4 (PLP). The brains were removed and stored for at least 24h in a 20% sucrose/10% glycerol solution at 4° C. A pin was inserted longitudinally into the right side of the brain to distinguish left from right. The fixed brains were sectioned frozen on a sliding microtome in the transverse plane at 35μm. Sections of each brain were collected as 12 separate series in 0.1 M PB with 0.02% sodium azide and stored until processed for immunohistochemistry. In the case of oculomotor nucleus, we immunolabeled brain sections for choline acetyltransferase (ChAT), using a goat polyclonal antibody (Chemicon #AB144) and by the peroxidase-antiperoxidase method as in prior studies [38]. Images were captured at a standardized level through the midbrain, the area of the oculomotor nucleus determined for that section using Image J, and ChAT+ perikarya counted and the nucleus area measured blinded as to group. Similarly, sections through the level of the optic tract were immunolabeled with rabbit anti-IBA1 (Wako Chemicals #019-19741) to visualize microglia by the peroxidase-antiperoxidase method as in prior studies [41,38]. The number of microglia showing transformation to an activated phenotype by morphology (rounded cell body, retracted processes, increased intensity of IBA1 immunolabeling) were counted per unit area of optic tract blinded as to group.

Analysis of Polyunsaturated Fatty Acids (PUFA):

Polyunsaturated fatty acids were analyzed at the Lipidomic facility at University of South Florida, Tampa, FL. PUFAs were measured and analyzed from animal brain, liver, and plasma after 30 days post TBI from WT-Sham, WT-Blast, Fat1⁺-Sham, and Fat1⁺-Blast animals following previously published procedures [42].

Analysis of Sphingolipids:

Sphingolipids from animal brain 30 days post TBI were analyzed at the Lipidomic core at the Virginia Commonwealth University, Richmond, VA following a previously published protocol [43]. Samples were dissolved in a cocktail of CH₃OH:CHCl₃ (2:1). Internal standards for sphingolipid metabolites were purchased from Avanti Polar Lipids. The internal standards of 500 pmol each were dissolved in the cocktail of C₂H₅OH:CH₃OH:H₂O (7:2:1) and added to the sample having the final volume of 20 μl. The detailed method of sample preparation and processing was followed as in our previously published article [43]. Sphingolipids were separated by reverse phase liquid chromatography (LC) using a Supelco 2.1 (i.d.) x 50mm Ascentis Express C18 column (Sigma, St. Louis, MO) and a binary solvent system at a flow rate of 0.5 mL/min. Before running and eluting the samples in the LC, the column was equilibrated for 30 s with a solvent mixture of 95% Mobile phase A1 (CH₃OH/H₂O/HCOOH, 58/41/1, v/v/v, with 5mM ammonium formate) and 5% Mobile phase B1 (CH₃OH/HCOOH, 99/1, v/v, with 5mM ammonium formate). After injecting the sample in the column, the A1/B1 ratio was maintained at 95/5 for 2.25 min, followed by a linear gradient to 100% B1 over 1.5 min, which was held at 100% B1 for 5.5 min, followed by a 0.5 min gradient return to 95/5 A1/B1. The column was re-equilibrated with 95:5 A1/B1 for 30 s before the next run. The species of ceramide (Cer), hexosyl-ceramide (Hex-Cer), sphingomyelin (SM), sphingoid lipids such as sphingosine (Sph), dihydro-sphingosine (DH-Sph), S1P, and DHS1P were identified based on their retention time and m/z ratio (mass-to-charge ratio), and quantified as described in the aforementioned article [43].

Sphingomyelinase Assay:

Acidic and neutral sphingomyelinase (aSMase and nSMase, respectively) activity was measured from the protein isolated from brain tissues 30 days post TBI from WT-Sham, WT-Blast, Fat1⁺-Sham, and Fat1⁺-Blast mice using the Amplex Red Sphingomyelinase Assay kit (Invitrogen, Carlsbad, CA) following a previously published protocol from our group [44].

RT-PCR Analysis:

Total RNA was extracted from the brain tissues after 30 days post-TBI (n=4) using Life Technologies RNA extraction kit and quantitative RT-PCR was performed accordingly with our previously published procedure [45]. Gene (mRNA) expression was measured from the brain tissues of WT-Sham, WT-Blast, Fat1⁺-Sham, and Fat1⁺-Blast mice. The transcripts include, inflammatory markers, such as Tumor Necrosis factor-α (TNF-α), Interleukin 6 (Il-6), Inducible Nitric oxide synthase (iNos), Interferon-γ (Ifn-γ), Il-4, Interleukin-4 (Il-4), Interleukin −10 (Il-10), Nucleotide-binding oligomerization domain, leucine rich repeat and pyrin domain containing 1 (NLRP1), NLRP-3, NLRP-4, C-X-C motif chemokine ligand 1 (Cxcl1), and C-X-C motif chemokine ligand 10 (Cxcl10); as well as sphingolipid metabolic markers, such as Serine palmitoyl transferase 1 (Spt1), Spt2, Ceramide synthase 2 (CerS2), CerS4; CerS5; CerS6;, Acyl-sphingosine amido-hydrolase-1(acid-Ceramidase) (Asah1), N-acyl sphingosine amidohydrolase 2 (Asah2), Sphingomyelin phosphodiesterase 1 (Smpd1), Smpd2, Ceramide kinase (Cerk), and Sphingosine kinase 1 (Sphk1), Sphk2. The sequence of the primers is listed in the Supplementary Table 1.

Statistical Analysis:

Data analysis was performed using Graph pad prism 7 software or SPSS. Statistical analyses were performed by Student’s t-test or one-or two-way ANOVA based on the experimental design with application of appropriate correction measures. Effects on the tail suspension test and for contextual fear and fear extinction were assessed by chi-square. Statistical significance was considered p < 0.05.

RESULTS

Assessment of PUFAs in Fat1⁺ transgenic mice:

We compared the levels of PUFA in Fat1⁺ transgenic mice relative to their WT littermates to determine the systemic effect of fat1 gene by analyzing plasma, liver, and brain tissue. Using LC-MS/MS analyses, we observed a significant increase in the relative mole percentage of EPA in the plasma of Fat1⁺-transgenic mice compared to their wild type counterpart (Fig. 2a), although we did not see any significant changes in AA and DHA. The plasma of transgenic mice showed a decreased level of 22:5(n-6) DPA compared to the plasma of WT mice (Fig. 2a). In the liver, we observed a significant decrease of relative mole percentage of 20:4(n-6) AA, 22:4(n-6) ADA (Adrenic acid), and 22:5(n-6) DPA in Fat1⁺-transgenic mice compared to wild type mice (Fig. 2b).

Fig. 2: Analysis of different Polyunsaturated fatty acids (PUFAs) in the tissues of wild type (WT) and Fat1⁺-transgenic (Fat1) mice.

Relative mole percentages of PUFAs (20:5n3, Eicosapentaenoic acid (EPA); 22:6n3, Docosahexaenoic acid (DHA); 20:4n6 Arachidonic acid (AA); 22:4n6 Adrenic acid (ADA); 22:5n6 Docosapentaenoic acid (DPA) in the plasma (a) and liver (b) of WT and Fat1⁺-transgenic mice. c. Relative mole percentages of EPA, DHA, AA, ADA, DPA and DGLA (20:3n6, Dihomo-gamma-linolenic acid DGLA) in the brain treated with 0-psi (sham) and 50-psi (blast) focal cranial air blast. Data are presented as Mean ± SEM and levels of significance are shown as p values (n=5; *p < 0.05, **p < 0.01; T-Test). WT-Sham (WT-0), WT-Blast (WT-50), Fat1⁺-Sham (Fat1-0) and Fat1⁺-Blast (Fat1-50).

In the brain, the tissue of greatest present interest, we found significantly increased levels in the relative mole percentages of EPA (20:5 n-3) and DPA (22:5 n-3) in Fat1⁺-transgenic mice compared to WT controls (Fig. 2c), and significantly decreased level of 22:5(n-6) DPA (Fig. 2c). The transgenic mice also showed a slight increase in the relative mole percentage of 20:3 (n-6) dihomo-gamma-linolenic acid (DGLA) in their brain compared to wild type (Fig. 2c).

Our data clearly indicated that Fat1⁺-transgenic mice born from the same parent and that had the same diets have higher levels of n-3 PUFA and lower levels of n-6 PUFA in their system than littermate WT mice and thus provide a useful model to study the effect of elevated n-3 PUFA in various disease and experimental conditions including mild TBI.

Fat1⁺-transgenic mice were protected from visual deficits in mild TBI:

We measured visual function by optokinetic reflex tests in WT and Fat1⁺-transgenic mice before and 10- and 30-days after 50-psi blast compared to a 0-psi sham for each genotype (Fig. 1). We detected a decrease in visual acuity (VA) in WT animals (both eye combined) after blast injury compared to sham-blast WT mice (Fig. 3a). Interestingly, we observed minimal reduction of VA in the blast treated animals Fat1⁺-transgenic mice compared to Fat1⁺-transgenic sham mice relative to their pre-blast levels (Fig. 3a). Similarly, along with the reduction in VA, we found a decrease in contrast sensitivity (CS) (i.e. an increase in threshold) in WT-blast animals compared to sham-blast mice, relative to their pre-blast levels (Fig. 3b). No reduction of CS was observed in the transgenic mice (Fig. 3b). Our results show that visual function was adversely affected in WT mice exposed to mild TBI, but this was not the case in the Fat1⁺-transgenic mice.

Fig. 3: Visual function test by Optokinetic reflex measurements of wild type (WT) and Fat1⁺-transgenic (Fat1) mice.

Line diagram shows Visual Acuity (a) and Contrast Sensitivity (b) at different time points-Pre-Blast, 10D-Post Blast, and 30D-Post Blast. Data are presented as Mean ± SEM and levels of significance are shown as p values (n=5; *p < 0.05, ANOVA). WT-Sham (WT-0), WT-Blast (WT-50), Fat1⁺-Sham (Fat1-0) and Fat1⁺-Blast (Fat1-50).

Fat1⁺-transgenic mice were protected from retinal functional deficit caused by mild TBI:

The scotopic ERG responses showed a significant reduction in a- and b-wave amplitude in blast-injured WT mice compared to their pre-blast levels (Fig. 4a, b). This decrease was absent in Fat1⁺-transgenic mice (Fig. 4a, b). We conclude that left cranium specific mild focal blast TBI causes an ERG deficit in WT mice, as also reported by Honig et al., (2019), which was prevented in Fat1⁺-transgenic mice [38].

Fig. 4: Retinal function assay by Electroretinogram of wild type (WT) and Fat1⁺-transgenic (Fat1) animals treated with 0-psi (sham) or 50-psi (blast) TBI.

Pre-blast a-wave and b-wave amplitude values served as reference for each group. Data are presented as percentages value of pre-blast in a-wave (a) and b-wave (b) (n = 5 WT-0, 3 WT-50, 5 Fat1-0, 4 Fat1-50; *p < 0.05; ANOVA). WT-Sham (WT-0), WT-Blast (WT-50), Fat1⁺-Sham (Fat1-0) and Fat1⁺-Blast (Fat1-50).

Open Field Test:

We performed pre- and post-blast open field (OF) testing to assay several parameters related to general locomotor activity. Before the blast, no significant differences were observed between WT and Fat1⁺-transgenic mice in the 30-min open field session. At 8 days after blast, WT mice with blast generally showed poorer performance relative to sham WT mice in OF, measuring change in performance from preblast. Sham WT mice improved pre to post blast, while the WT blast mice generally did worse. By contrast, the sham mutant mice generally did worse postblast than preblast, and the blast mutant mice improved from pre to postblast. Thus, the mutants appeared to be less adversely affected by blast in open field than were the WT. To evaluate this statistically, performance was normalized to pre-blast sham test performance, which showed that several unique motor endpoints were notably worsened post-blast relative to sham in the WT mice, but significantly better or nearly so in the mutant mice, including distance traveled, length of progression segment, progression segment maximum speed, progression segment mean speed, turn rate, turn radius, and number of stops per distance. For example, length of progression segment was reduced by 41% in WT-blast mice, but actually increased by 27% in Fat1⁺-transgenic mice (Fig. 5a). Similarly, progression segment speed was reduced by 20% in WT blast mice but increased 17% in blast Fat1⁺-transgenic mice. Particularly noteworthy was an 103% increase in number of stops per unit distance in WT blast mice, compared to 10% reduction in stopping in Fat1⁺-blast mice (Fig. 5a). In summary, the Fat1⁺-transgenic mice appeared to be rescued from the OF deficit seen in WT mice with mild TBI.

Fig. 5: Analysis of open field behavior, depression and fearfulness in wild type (WT) and Fat1⁺-transgenic animals with (50-psi, blast) or without (0-psi, sham) mild TBI.

Mice were divided into four groups, WT-0, WT-50, Fat1-0 and Fat1-50, each containing 8-12 mice. Data are presented as Mean ± SEM and levels of significance are shown as p (*) values. Open-field motor testing was conducted before and after the blast injury and data are represented in comparison to pre-blast values normalized to WT-sham mice (n = 10 WT-0, 10 WT-50, 10 Fat1-0, and 11 Fat1-50; *p < 0.05; **p < 0.01; one-way SPSS ANOVA) (a). Tail suspension testing was conducted in the blast injured animal and data presented as line graphs of cumulative immobility per minute over the 5 min test, in which more immobility reflects more depression-like behavior (n = 9 WT-0ham, 10 WT-50, 9 Fat1-0, 10 Fat1-50; **p < 0.01; ***p < 0.001; chi-square) (b). We also assessed the effect of blast per genotype on depression-like behavior by subtracting sham performance from blast performance per genotype, as depicted in this graph (n = 9 WT-0, 10 WT-50, 9 Fat1-0, 10 Fat1-50; **p < 0.01; ***p < 0.001; chi-sq; ***p < 0.001; chi-square) (c). Fear acquisition to a tone CS paired with shock was assessed (n = 9 WT-0, 10 WT-50, 9 Fat1-0, 10 Fat1-50; **p < 0.01; ANOVA) (d). Contextual fear in response to the fear conditioning chamber was assessed (n = 9 WT-0, 10 WT-50, 9 Fat1-0, 10 Fat1-50; ***p < 0.001; chi-square) (e). Conditioned fear retention in WT and Fat1⁺-transgenic mice, with males and females combined was assessed (n = 9 WT-0, 10 WT-50, 9 Fat1-0, 10 Fat1-50; *p < 0.05; ***p < 0.001; chi-square) (f). Conditioned fear for male mice alone is shown (n = 5 WT-0, 6 WT-50, 6 Fat1-0, 6 Fat1-50) (g). Conditioned fear for female mice alone is shown (n = 4 WT-0, 4 WT-50, 3 Fat1-0, 4 Fat1-50; *p < 0.05; **p < 0.01; chi-square) (h).

Tail Suspension Depression Assessment:

All mice were tested for depression by the tail suspension test about one month after the blast. Overall, mutants and WT did not differ in their depression, when TBI and sham were combined within genotype. Thus, the mutation by itself did not make mice more or less depressed. The depression test did, however, show a clear differential effect of blast for the two genotypes, with the cumulative immobility score significantly higher after TBI in WT mice compare to WT-sham than after TBI in Fat1⁺ transgenic mice compared in sham mutant mice (Fig. 5b). In Fig. 5c, we have simplified the depiction of the sham versus blast difference for the WT and Fat1⁺ transgenic mice. Sham cumulative immobility per minute is subtracted from blast immobility, showing how much depression is elevated above sham for WT mice but not for Fat1⁺ transgenic mice.

Fear Conditioning - Fear Acquisition:

The transgenic mice were in general more fearful even before fear conditioning than the WT mice. The graph shows fear responses during each of the eight 30-second blocks before the first CS-US (tone-shock) pairing, and then during the tone for each of the five 30-second tone presentations (Fig. 5d). Shock occurs at the end of the tone, so fear responses are lower during the first CS than during subsequent CS presentations, as mice learn to fear the tone. We thus examined the results by group, correcting each animal’s fear learning for its basal fear level by subtracting out its mean pre-CS basal responding, and compared cumulative responses per CS over the five CS presentations by ANOVA. This revealed that blast WT mice showed significantly greater fear responding to the CS than WT sham mice during acquisition. They also showed enhanced fear responding over the acquisition trials compared to the Fat1⁺ blast mice.

Fear Conditioning - Contextual Fear:

Fear responding to the context in which the fear conditioning occurred, but in the absence of CS presentation, is called contextual fear. In our paradigm, this is tested over 3 minutes before the CS-alone presentations. Because the Fat1⁺-transgenic mice had shown themselves to be more fearful than the WT mice as a basal condition during the fear acquisition test prior to CS presentation, we corrected contextual fear performance by subtracting from it the pre-CS fear level during fear acquisition for each mouse. Taking this approach, we found that the Fat1⁺-blast mice showed a significantly lesser contextual fear increase (fear in response to the chamber alone) compared to Fat1⁺-sham mice than was the case for WT-blast mice compared to WT-sham mice. This trend was prominent on the second and third days of testing, but not the first. Thus, Fat1⁺-transgenic mice showed a lesser increase in contextual fear after TBI than did WT mice (Fig. 5e).

Fear Conditioning - Learned Fear:

The results for the extinction testing of fear learning showed consistently and significantly less fearfulness in Fat1⁺ mice across extinction sessions than in WT mice, combining sham and blast mice for each genotype. For analysis of fear responding to the CS-alone, we again corrected for the differing levels of basal fear by subtracting out the mean pre-acquisition fearfulness. We then plotted the mean fearfulness across the 15 CS-only fear extinction (FE) sessions on each of days FE1, FE2, and FE3 for the WT-blast and Fat1⁺-blast mice, normalizing for differences in contextual fear prior to CS presentations. Sham Fat1⁺ mice showed significantly less fearfulness than sham WT mice to the CS presentations over the three extinction sessions, but blast Fat1⁺ mice were largely comparable in fearfulness to blast WT mice during the CS presentations (Fig. 5f). Combining groups within genotype, Fat1⁺ mice were significantly less fearful than WT mice over the three extinction sessions. Although males and females did not differ during fear acquisition or contextual fear testing, they did differ in the extent to which blast caused increased fear responding to the CS during extinction testing. WT blast males showed greater fear responding than WT sham males, especially during FE1, but Fat1⁺ males with blast were comparable to sham Fat1⁺ males during FE1 (Fig. 5g). By contrast, WT blast females showed lesser fear responding than WT sham females during FE1-FE3, while Fat1⁺ females with blast showed increased fear during FE1 and FE2 (Fig. 5h). Nonetheless, combining groups within genotype, Fat1⁺ males and females each showed less fear responding to the CS than gender-matched WT mice, significantly so in the case of females (Fig. 5f and 5h).

Immunohistochemical Assessment of Brain Neurons and Microglia:

To further evaluate the benefit of the Fat1⁺ genotype for resisting the deleterious outcome from TBI, we investigated its impact on the adverse effects of oculomotor motoneurons and on microglial activation in the optic tract in mouse brains harvested about 60 days post TBI and processed by immunohistochemical labeling for ChAT for motoneurons and IBA1 for optic tract microglia. As shown in Figs. 6a–d, the representative images of oculomotor nucleus in the four group of mice and Fig. 6e, the area of the oculomotor nucleus was reduced significantly in the WT-blast brains compared to WT-sham. By contrast, it remained unaltered in both sham and blast animal brain of Fat1⁺-transgenic mice compared to WT-sham mice (Figs. 6c, 6d, 6e). We previously reported that mild TBI induces microglial activation in the optic tract in mice [12], Figs. 7a–7d and 7a’–7d’ show representative images of activated microglia in WT-blast and Fat1⁺-blast mice right optic tracts. The spatial density of activated microglia in the right optic tract of WT-blast brain was significantly increased 3-fold compared to the WT-sham brain, whereas no changes were observed in activated microglial numbers in the sham or blast Fat1⁺-transgenic mice compared to WT-sham mice (Fig. 7e).

Fig. 6: Immunohistochemical analysis of oculomotor neurons with choline acetyltransferase (ChAT) immunostaining at midbrain level in sections from wild type (WT) and Fat1⁺-transgenic animals with and without mild TBI.

Representative images of ChAT-immunolabeling of oculomotor nucleus of WT-blast and Fat1⁺-blast littermates, which shows a lesser size of the oculomotor nucleus in WT-blast one month after TBI (a and b). No changes were observed in Fat1-transgenic mice (c and d). Quantification of oculomotor nucleus area (μm²) shows a significant decrease in the area of the oculomotor nucleus in the wild-type blast (WT-50) group compared to wild-type sham (WT-0) (n = 14 WT-0, 12 WT-50, 10 Fat1-0, 11 Fat1-50; *p < 0.05; ANOVA) (e).

Fig. 7: Analysis of microglial activation in the optic tract by immunostaining for IBA1 in wild type (WT) and Fat1⁺-transgenic mice with or without mild TBI.

Representative images of the brain sections showing the right optic tracts with activated microglia (thick, bushy appearance). a: WT-Sham (WT-0); a’: WT-Sham, boxed area enlarged; b: WT-Blast (WT-50); b’: WT-Blast, boxed area enlarged; c: Fat1⁺-Sham (Fat1-0) c’: Fat1⁺-Sham, boxed area enlarged; d: Fat1⁺-Blast (Fat1-50); d’: Fat1⁺-Blast, boxed area enlarged. Histogram shows the number of microglia/mm² of the section (e). Data are presented as Mean ± SEM and levels of significance are shown as p values (n = 14 WT-0, 12 WT-50, 10 Fat1-0, 11 Fat1-50: n=11; *p < 0.05; ANOVA). Arrows indicate bushy microglia.

Expression of Inflammatory Marker Genes:

We tested expression of different cytokines, chemokines, and inflammasome markers in brain tissue of the four groups of mice after 30 days post-TBI. The mRNA level of proinflammatory cytokines, Il-6, Tnf-α, and iNos was elevated significantly in WT-blast brain compared to their sham counterpart (Fig. 8a). In comparison, the Fat1⁺-blast animal brain showed no upregulation of these mRNAs (Fig. 8a). Like proinflammatory cytokines, the expression of proinflammatory chemokines such as Cxcl1 and Cxcl10 was increased significantly in the WT-blast brain, but no change was detected in Fat1⁺-blast brain (Fig. 8b). On the contrary, anti-inflammatory Il-10, though increased in the blast exposed brain, was found to be yet more increased and at its highest levels in the Fat1⁺-blast brain (Fig. 8a). Inflammasome protein complex, Nucleotide-binding oligomerization domain, Leucine rich repeat and Pyrin domain containing (NLRP) proteins play important roles in the process of inflammation during TBI [46]. We found elevated gene expression of Nlrp1 and Nlrp3 in WT-Blast brain, compared to their sham littermates. By contrast, their expression level in Fat1⁺-transgenic mice in both sham and blast animals remained unaltered (Fig. 8a).

Fig. 8: Expression analysis of inflammatory marker genes in the brain tissues of wild type (WT) and Fat1*-transgenic animals treated with 0-psi (sham) or 50-psi (blast) TBI. a: Cytokines and inflammasome markers; b: Chemokine markers.

mRNA expression levels are presented over wild type sham (WT-0) (=1.0) after normalization with two housekeeping genes, Ribosomal protein L19 (Rpl19) and Glyceraldehyde-3-phosphate dehydrogenase (Gapdh) (n = 4; *p < 0.05; *** p < 0.001; ANOVA). WT-Sham (WT-0), WT-Blast (WT-50), Fat1⁺-Sham (Fat1-0) and Fat1⁺-Blast (Fat1-50). TNF-α, Tumor Necrosis factor-α; Il-6, Interleukin 6; iNos, Inducible Nitric oxide synthase; ifn-γ, Interferon-γ, Il-4, Interleukin-4; Il-10, Interleukin -10; NLRP1, Nucleotide-binding oligomerization domain, leucin rich repeat and pyrin domain containing; NLRP-3, NLRP-4, Cxcl1, C-X-C motif chemokine ligand 1; Cxcl10, C-X-C motif chemokine ligand 10.

PUFA Levels in the Brain after TBI:

As described earlier, the brain of the Fat1⁺-transgenic mice had higher levels of n-3 PUFA and lower levels of n-6 PUFA compared to their WT littermates (Fig. 2c). We did not see any significant alteration in n-3 or n-6 PUFA species due to blast exposure at one month after blast. Instead, these brains maintained their original levels of these fatty acids as observed without TBI (Fig. 2c).

Assessment of Sphingolipid Metabolism in the Brain:

No study has so far reported on the sphingolipid profile in Fat1⁺-transgenic mice brain and none on the brains from mild TBI mouse models. We determined the levels of the major cellular sphingolipids (e.g., ceramide, Cer; hexosyl-ceramide, HexCer; sphingomyelin, SM) and long-chain sphingoid base containing species (sphingosine, Sph; dihydro-sphingosine, DHSph; sphingosine 1-phosphate, S1P and dihydro sphingosine 1-phosphate, DHS1P) in the brain tissue harvested one month after blast. We found no difference in the total Cer levels between WT and Fat1⁺-transgenic mice brain without blast injury (Fig. 9a), but mild TBI caused a significant increase in Cer levels in WT-blast mice compared to their sham WT controls (Fig. 9a). Interestingly, we found a significant reduction in the Cer levels in Fat1⁺-blast mice one month after blast compared to Fat1⁺-sham as well as WT-sham controls (Fig. 9a). No alteration was observed in the levels of total HexCer and SM in WT and Fat1⁺-transgenic mice exposed to either sham or blast treatments (data not shown). All sphingolipid classes contain many sub-species with varied chain-lengths and saturation states. Among the ceramides, we observed mild TBI caused a significant increase in the long-chain ceramides (C20:0, C24:0, and C24:1) in WT animals. On the other hand, the Fat1⁺-blast mice showed a significant decrease in the long-chain ceramides C20:0 and C24:0 and the short-chain ceramide C18:0 compared to their sham mutant controls. Moreover, compared to WT-blast, the Fat1⁺-blast mice had significantly reduced levels of short-chain ceramides (C14:0, C16:0, and C18:0), as well as long-chain ceramides (C20:0, C24:0, and C24:1) (Fig. 9b). We did not find any alterations among HexCer species between WT and transgenic animals (data not shown). Like for C20:0 and C24:0 Cer, we observed an increased level of C20:0 and C24:0 SM in wild-type blast animals compared to their Fat1⁺ blast counterparts (Supplementary Fig. S1a). Long chain base SPL remained unaltered among the four groups of animals (Supplementary Fig. S1b). Our data suggest mild TBI caused alteration in sphingolipid metabolism in the brain with a specific increase in the bioactive Cer levels, which was prevented in Fat1⁺-transgenic mice, indicating Cer downregulation as a possible mechanism of neuroprotection in these mice from blast injury.

Fig. 9: Analysis of Ceramide (Cer) in the brain tissues of wild type (WT) and Fat1*-transgenic animals treated with 0-psi (sham) or 50-psi (blast) TBI.

a: Ceramide levels (pmole/mg of tissue) in the brain of WT-Sham (WT-0), WT-Blast (WT-50), Fat1⁺-Sham (Fat1-0), and Fat1⁺-Blast (Fat1-50) mice one month after TBI. b: Ceramide species in the brain of four groups of animals as shown in A. (n = 5 WT-0, 4 WT-50, 3 Fat1-0, 7 Fat1-50; *p < 0.05, **p < 0.01, ***p < 0.001; ANOVA).

Expression and Activity of Ceramide Pathway Genes and Enzymes:

In several prior studies, we have assessed activation of Cer generating pathways in different disease or experimental conditions [47,45,44]. Using a similar strategy, here we studied the expression level of different sphingolipid metabolic genes in the brain of sham and blast exposed Fat1⁺-transgenic mice and their WT counterparts one month after the blast. We detected that the mRNA level of the de novo ceramide biosynthetic gene, Spt1, in Fat1⁺-transgenic mice remained half of that in their WT counterparts for both sham and blast groups (Fig. 10). The expression of the other SPT gene, Spt2 remained unaltered (Fig. 10). Then we studied the mRNA level of different ceramide synthases (CerSs) that catalyze the immediate next step of de novo ceramide biosynthesis and found a significant decrease in Cers2 and CerS4 in Fat1⁺-transgenic mice compared to their WT counterpart (Fig. 10). Other CerS genes (CerS1, CerS5, and CerS6) remained unaltered, though a decreasing trend was noted in Fat1⁺-transgenic mice (Fig. 10). We also detected decreased levels of acidic ceramidase, Asah1 mRNA in the transgenic animal brain compared to the WT brain (Fig. 10). However, the levels of neutral ceramidase, Asah2 remained unaltered (Fig. 10). No effect was noticed for either acidic and neutral sphingomyelinase genes Smpd1 and Smpd2 (Fig. 10). Overall, we observed reduction in the Cer metabolic genes in Fat1⁺-transgenic mice brain with or without exposure to TBI. The effect of blast was only significant in the case of Cers2, which was lower in WT-blast mice compared to WT-sham mice.

Fig. 10: Expression analysis of sphingolipid metabolic genes in the brain tissues of wild type (WT) and Fat1*-transgenic animals treated with 0-psi (sham) or 50-psi (blast) TBI.

mRNA expression levels are presented over wild type sham (WT-0) (=1.0) after normalization with two housekeeping genes, Ribosomal protein L19 (Rpl19) and Glyceraldehyde-3-phosphate dehydrogenase (Gapdh) (n = 4; *p < 0.05; *** p < 0.001; ANOVA). Spt1, Serine palmitoyl transferase 1; Spt2; CerS2, Ceramide synthase 2; CerS4; CerS5; CerS6; Asah1, Acyl-sphingosine amido-hydrolase-1(acid-Ceramidase); Asah2; Smpd1, Sphingomyelin phosphodiesterase 1); Smpd2; Cerk, Ceramide kinase; Sphk1, Sphingosine kinase 1; Sphk2.

Along with gene expression, we also conducted activity assays for the major Cer generating enzymes sphingomyelinases (SMase, both neutral and acidic) in the brain tissue one month after TBI. We found an increase in neutral SMase (nSMase) activity in WT-blast brain compared to WT-sham; whereas no changes were found in Fat1⁺-transgenic mice either exposed or unexposed to TBI (Supplementary Fig. S2). We did not detect any changes in acidic SMase (aSMase) activity in these brain tissues (data not shown). Though this study provides some clues regarding Cer activation in the blast exposed brain, the gene expression and enzyme activity may be temporary and a detailed characterization of the brain at different timepoints after the TBI needs to be undertaken to determine the specific series of events that lead to higher Cer levels in TBI brains.

DISCUSSION

In the present study, we used the Fat1⁺-transgenic mice to understand the role of elevated levels of systemic n-3 PUFAs in the prevention of motor, sensory (including visual), and behavioral deficits resulting from mild TBI, and to determine the involvement of another group of lipid, namely sphingolipids, in this pathophysiology. The Fat1⁺-transgenic mice express the fat-1 (n-3 fatty acid desaturase) gene from C. elegans in every tissue. Since mammals are deficient in the ability to generate n-3 PUFAs and n-3 fatty acid desaturase converts n-6 PUFAs to n-3 PUFAs, Fat1⁺-transgenic mice have elevated n-3 PUFAs in their tissues compared to WT mice. As the mammalian system cannot generate n-3 PUFAs from n-6 PUFAs, which are abundant in our diet, Fat1⁺-transgenic mice have been generated to help understand the benefit n-3 PUFAs may provide in different diseases and experimental models [48]. It is well established that mild TBI in experimental animal models causes motor, visual, and emotional deficits, which resemble the characteristics of mild TBI in humans [5,8]. We have used the Fat1⁺-transgenic model to determine whether the elevated levels of systemic n3-PUFAs in these mice can prevent neuropathological disorders occurring from mild TBI. For the past few decades, n-3 PUFAs have displayed promising benefits in a variety of neurological disorders including ischemic stroke and traumatic brain injury. Our study provides evidence that 1) Fat1⁺-transgenic mice exhibit a beneficial effect on behavioral deficits following mild TBI, 2) Fat1⁺-transgenic mice are protected from microglial activation and neurodegeneration, and 3) Fat1⁺-transgenic mice brains have reduced production of ceramides, and inflammatory and cell death factors after mild TBI in comparison to their WT littermates.

Fat1⁺-transgenic mice have been used in various studies since their development in 2004 by Kang et ai. [49]. A common way to make a major difference in their n-3 PUFA content compared to WT littermate is to feed with only an n-6 diet [50] (i.e. Safflower oil). In this study, we fed them with normal lab chow and depended on the transgene to modify their n-3 PUFA content. By analyzing their lipids, we found 2-fold increase of EPA (20:5 n-3) in Fat1⁺-transgenic mice plasma and brain tissue but no change in the most abundant n-3 PUFA, DHA (22:6 n-3) (Fig. 2a,c). In the liver and brain tissue, we saw a concomitant decrease of various n-6 PUFAs such as n-6 AA (20:4 n-6), n-6 ADA (22:4 n-6), and n-6 DPA (22:5 n-6) (Fig. 2b,c). Along with EPA, a significant increase of n-3 DPA (22:5 n-3) was also evident in the Fat1⁺-transgenic mice brain (Fig. 2c). Since mammals are incapable of converting n-6 to n-3 fatty acids, these increases in n-3 PUFAs and decreases in n-6 PUFAs Fat1⁺ mice are purely the result of the action of the Fat-1 transgene protein product n-3 fatty acid desaturase [51]. Fat1⁺ mice have been shown to generate increased levels of EPA and DPA as major n-3 products in previous studies [24].

We also wanted to establish whether a systemic elevation of n-3 PUFA in Fat1⁺-transgenic mice was associated with prevention of the adverse effects of traumatic brain injury. Cognitive, motor, sensory, visual, and emotional deficits are the common pathological features associated with mild TBI in humans and in animal models [12]. Those adverse effects were previously demonstrated by some members of our group as outcomes in an experimental closed-head injury model [12,38,8]. Consistent with our previous findings for visual function, we found visual acuity and contrast sensitivity were impaired by blast injury (Fig. 3a, b) in WT animals. Of note, we found that Fat1⁺-transgenic animals resist these blast-related deficits in visual function. These transgenic animals also do not show the retinal functional ERG deficit evidenced by reduced amplitude of the a- and b-waves in WT-blast animals (Fig. 4). Our open field testing shows that Fat1⁺-transgenic animals are similarly protected against motor deficits after TBI injury (Fig. 5a). Our prior studies suggest the motor deficits in TBI could be due to damage to the pyramidal tract and corticospinal tract [52] or to the basal ganglia [6]. These regions in the brain carry motor signals and regulate voluntary movement. Similar to our observation, Desai et al (2014) showed that a dietary supplement of n-3 PUFA restricts motor deficits in traumatic brain injury [27]. Additionally, supplementation of DHA in the feed of Sprague-Dawley rats counteracts the effects of the concussive injury on neural function after brain trauma [53]. The lower level of depression-like behavior in Fat1⁺-blast transgenic mice compared to WT-blast animals in the tail suspension test (Fig. 5c) could be due to an endogenously elevated level of EPA in the brain tissue. A clinical trial with ethyl-eicosapentaenoate showing better performance in overcoming depression than standard therapy supports the role of EPA in alleviating depression [28]. Transgenic blast animals also showed less contextual and conditioned fear than the WT-blast animals (Figs. 5e–5g). Depression and fearfulness are controlled by the basolateral amygdala in the brain, where TBI reduces the abundance of a subpopulation of Thy1+ fear-suppressing neurons [40]. The effectiveness against depression and fear in the Fat1⁺-transgenic mice might be an action of EPA, n-3 PUFA, in protecting Thy1+ neuron against blast injury. Consistent with this, several reports indicate that diet-supplemented DHA preserves cognitive capacity after TBI [27,53]. Transgenic Fat1⁺ mice also exhibit long term behavioral protection against transient focal cerebral ischemia by overproduction of n-3 PUFAs [26]. Overall, transgenic Fat1⁺ mice showed promising evidence of reducing or preventing post-traumatic brain disturbance in that the increased n-3 PUFA levels appeared to help maintain sensory and motor function, ameliorated emotional disturbances, and reduced brain injury, as evidenced by reversal of atrophy of oculomotor nucleus and optic tract microglial activation (Figs. 6 and 7). These results and prior data suggest that high n-3 PUFA in the brain is beneficial in warding off the adverse effects of TBI.

Microglial activation and inflammation lead to several secondary pathological complications in TBI [54]. In this context, our previous studies showed microglial activation in the optic tract in mild TBI [12,8,38]. It has been reported that n-3 PUFA supplementation in a multiple sclerosis (MS) model reduces the motor and cognitive deficit by the polarization of microglial phenotype toward a more beneficial state [55]. We found about a three-fold increase in activated microglial abundance in the optic tract of WT-blast animals compared to WT-sham but not in the Fat1⁺-blast animals’ optic tract (Fig. 7). Activation of microglia contributes to inflammation in blast injury. Neuroinflammation and neurodegeneration lead to confusion, memory loss, and depression in TBI [56]. Consistent with these data, we found significant upregulation of pro-inflammatory marker genes such as Tnf-α, Il-6, iNos, and Ifn-γ in WT-blast brains but not in the Fat1⁺-blast brains (Fig. 8). The inflammatory factors TNF-α, IL6, and IL1β induce neuronal changes leading to neuronal dysfunction and damage [20]. Interestingly, among n3-PUFAs, EPA plays a vital role in the reduction of neuroinflammation and neurodegeneration [29]. Although the level of EPA is less than the level of AA and DHA in brain, in microglia the EPA level is about twofold higher than the level of DHA [30]. Thus, the inflammation and neuronal damage might be associated with the cognitive and behavioral changes in WT-blast mice, which is prevented in the Fat1⁺-transgenic mice due to strong anti-inflammatory properties of n3-PUFAs [57].

There is not much evidence on the role and involvement of bioactive sphingolipids in n-3-PUFAs-mediated neuroprotection, although bioactive sphingolipids are well known for their role in neuroinflammation and neurodegeneration [58–60]. Interestingly, we observed Cer level increases in the WT brain following TBI but not in Fat1⁺-transgenic brain (Fig. 9). In fact, Cer is reduced in Fat1⁺-transgenic brain following TBI (Fig. 9). We do not know whether the ceramide increase in the brain following TBI is the cause of neuroinflammation or a consequence of the neuronal damage. While there is a possibility for either case, ceramide’s role in driving inflammation led us to hypothesize that if not the primary, ceramides can be a secondary cause of neuroinflammation and neurodegeneration. Reducing the levels of ceramide and blocking their activation could therefore be one of the mechanisms of n-3-PUFAs- mediated neuroprotection. Previous studies in wobbler mice, a murine model of motor neuron disease, found a higher level of Cer and other SPL metabolites is associated with the neurodegeneration these mice exhibit, and reduction of ceramide synthesis by treating with myriocin improved wellness index, grip strength, muscular atrophy, and neuropathology associated with motor neuron dysfunction [35]. Smpd1 (acid Sphingomyelinase) knock out mice have been shown to recover faster from TBI, which could be associated with their reduced ability in forming ceramide from sphingomyelin and so have reduced ceramide-toxicity [34]. There are various pathways for Cer generation or activation: the de novo pathway, the SMase pathway, or the salvage pathway. We noticed higher levels of nSMase (neutral SMase) activity in the WT brain one month after blast (Supplementary Fig. S2). By contrast, our gene expression studies suggested lower levels of expression of de novo ceramide biosynthetic genes Spt1, Cers2 and CerS4 in Fat1⁺-transgenic mice brain compared to their WT counterpart (Fig. 10). This important information suggests the n-3 PUFAs (or EPA) can regulate ceramide levels in the cells either affecting expression of ceramide biosynthetic genes or activity of ceramide metabolic enzymes. The downstream factors involved in ceramide-driven pathology are not well established in TBI. There is a report that Cer disrupts mitochondrial function in brain injury and induces excitotoxicity, oxidative stress, and inflammation [34,59]. Modulation of an inflammatory pathway through activation of NLRP3 by post-transcriptional activation of the Smpd1 gene was noticed in TBI [34]. It is also reported that Cer in macrophages induces activation of the inflammasome marker NLRP3 and subsequently thereby induces proinflammatory cytokines [61]. However, on the involvement of n-3 PUFA in ceramide regulation, an in-vitro study of inhibition of Cer synthesis by DHA suggests the involvement of n-3 PUFA in restricting Cer-mediated lipotoxicity [37]. DHA is also shown to inhibit SMase activity in human retinal endothelial cells [37]. A detailed study is therefore needed analyzing samples at different time points after TBI to understand the interrelationships of n-3 PUFA and sphingolipids for neuroprotection, neuroinflammation and neurodegeneration.

Overall, we can conclude that an increased level of n-3 PUFAs in Fat1⁺-transgenic mice may be causal to reduced Cer-toxicity, and neuroinflammation. Thus, microglial activation was ameliorated, and oculomotor nucleus homeostasis and neuronal integrity were maintained. Understanding the n-3 PUFA and ceramide relationship in TBI will advance our knowledge on the mechanisms of neuroinflammation and neurodegeneration in many forms of human neurological disorders and thus may lead development of novel therapeutic strategies.