Mild traumatic brain injury (mTBI), often referred to as concussion, is a serious medical issue with a 10-fold higher burden on the healthcare system than both moderate and severe traumatic brain injuries (TBI) (Corwin et al., 2017; Dewan et al., 2019). mTBIs are heterogeneous in nature and can result in a variety of neurobehavioral sequalae. Acute symptoms of mTBI include but are not limited to disorientation, dizziness, headaches, and emotional, sensorimotor, and cognitive deficits (Schimmel et al., 2017). These symptoms are thought to be caused by an acute and typically transient neuropathophysiological cascade (Pavlova et al., 2018). Evidence suggests that mTBI induces glutamatergic-driven excitotoxicity, neurometabolic and lipid imbalances, and a mitochondrial energy crisis (Ng and Lee, 2019; Romeu-Mejia et al., 2019; Khatri et al., 2021). Furthermore, pharmacological modulators of metabolic derangement and mitochondrial dysfunction, such as excitotoxicity-dampening compounds or antioxidants, are protective against brain trauma (Hwabejire et al., 2013; Ismail et al., 2020). Repeated mTBIs that occur in short succession exacerbate neuropathology and symptomology (Miller et al., 2013; Dretsch et al., 2015), and there is growing evidence that repeated mTBIs increase the risk of developing progressive neurodegenerative and emotional disorders, including Alzheimer’s disease, amyotrophic lateral sclerosis, chronic traumatic encephalopathy, depression, anxiety, and epilepsy (Abel, 2007; Graham and Sharp, 2019; Asken and Rabinovici, 2021). However, despite increased research efforts on the effects of mTBI, there are still important knowledge gaps surrounding the cellular consequences of singular and repeated mTBIs. Uncovering specific cellular alterations after mTBI (i.e., biomarkers) could inform personalized treatment approaches.

One recognised and potentially prominent neuropathological aspect of the mTBI cascade is neurometabolic dysfunction (Kim et al., 2017; Veeramuthu et al., 2018), which can be measured in brain tissue via ¹H-MRS. Several neurometabolites, such as N-acetylasparate (NAA), Creatine (Cr), phosphocreatine (PCr), glucose (Glc), glutamate (Glu), and glutamine (Gln), have been identified to have putative utility in assisting mTBI diagnosis and prognosis (Eisele et al., 2020). Cellular damage induced by mTBI can lead to the excessive release of Glu and thereafter neuronal excitotoxicity and potentially cell death (Baracaldo-Santamaría et al., 2022). Alterations in the concentrations of Glu and Gln, the storage form of Glu (Dorsett et al., 2017; Baracaldo-Santamaría et al., 2022), are indicative of altered brain metabolism, but changes in their abundance have previously been reported as a sum (Ashwal et al., 2004; Shutter et al., 2004; Yeo et al., 2011). Using advanced ¹H-MRS, one can gather more accurate measurements by evaluating Glu and Gln levels separately (Ramadan et al., 2013). Alterations in the levels of Glc, which could lead to hyper-glycolysis (Vagnozzi et al., 2008), has been postulated to be involved in TBI (Yuan et al., 2022), but investigations to answer this question with the use of ¹H-MRS are also lacking. NAA has been flagged as a potential biomarker of neuronal integrity and mitochondrial function, and alterations in NAA/Cr may represent a diagnostic biomarker for mTBI (Vagnozzi et al., 2005). A clinical study found that NAA/Cr alterations and recovery post-mTBI worsened when a second injury had occurred shortly after the initial injury; notably, metabolic alterations persisted beyond symptom resolution (Vagnozzi et al., 2008). In addition, Cr and PCr are thought to be an indicator of cellular energetics (Wallimann et al., 2011) and are found abundantly in the brain (Stovell et al., 2017), where they are significantly altered in response to TBI (Gasparovic et al., 2009; Signoretti et al., 2010; Yeo et al., 2011). Another metabolite, choline, is a major component of phospholipid membranes, and the elevation of glycerolphosphocholine and phosphocholine (GPC and PCh) after TBI are thought to indicate cell membrane breakdown (Shutter et al., 2004). Altogether, it has been speculated that mTBI-induced metabolic changes are, at least partially, responsible for a transient epoch of enhanced vulnerability to cerebral insults – with shorter inter-injury intervals leading to increased neurometabolic deficits (Vagnozzi et al., 2007, 2008).

Mitochondria are vital for the regulation of cellular metabolism due to their roles in ATP production and intracellular calcium buffering (Severo et al., 2020). mTBI can lead to the excessive influx of calcium into cells, which is sequestered into mitochondria and instigates mitochondrial allostatic overload and consequently, ionic imbalances, an accumulation of harmful reactive oxygen species, loss of membrane potential, and deficient ATP production (Hubbard et al., 2018). Furthermore, mitochondrial dysfunction is observed in patients with disorders that are co-morbid with mTBI, like depression, anxiety, cognitive deficits, and epilepsy, suggesting that mitochondrial impairments represent a vulnerability factor for subsequent pathological disturbances (Singh et al., 2021). Nevertheless, it is unclear how mitochondrial deficits after mTBIs are linked to ongoing or progressive neurometabolic and lipidomic alterations.

Another potentially prominent neuropathological characteristic of mTBI is the dysregulation of lipids, which are crucial to the structural and functional integrity of neuronal and non-neuronal cells, as well as cellular organelles, such as mitochondria (Ojo et al., 2019). Examples of lipids that may be implicated in mTBI include phosphatidylethanolamine (PE), Bis(monoacylglycero)phosphate (BMP), acylcarnitine (AC & AC-OH), cardiolipin (CL), lysophosphatidylcholines (LPC), lysophosphatidylethanolamine (LPE), and lysophosphatidylinositol (LPI). PE is a major component of cell membranes that is implicated in membrane stability, trafficking, permeability, and fluidity (Yoon et al., 2022), and it has also been linked to neurodegenerative disorders (Anyaegbu et al., 2022). BMP is enriched in the lysosomal network and is suspected to represent a biomarker of phagocytizing macrophages/microglia in brain tissue and is implicated in various physiological cellular functions, such as cell signaling and the regulation of neuronal growth (Nielsen et al., 2016). Acylcarnitines are generated from the conjugation of acyl-coenzyme A (CoA) with carnitine, to facilitate the transport of fatty acids across the mitochondrial membrane from the cytosol into the mitochondrial matrix, for β oxidation and cellular energy production, and are available as supplements with putative neuroprotective effects (Allen et al., 2021). CL makes up 20% of the inner mitochondrial membrane and are essential for the optimization of enzymes involved in mitochondrial bioenergetics, as well as fusion and fission processes, and apoptosis, with CL oxidization being observed before the appearance of apoptotic biomarkers after TBI in rats (Bayir et al., 2007; Paradies et al., 2014). LPCs are molecules derived from the degradation of phospholipids in the cell membrane, and, once released by apoptotic cells, are thought to act as a “find-me” signal for phagocytes (Park and Kim, 2017) and may be an indicator of injury severity (Pasvogel et al., 2008, 2010). Another minor component of the cell membrane, LPE, is involved in the activation of enzymes and the mediation of cell signaling, and alterations in its expression in a mouse model of mTBI paralleled those detected in soldiers that had endured chronic mTBI (Emmerich et al., 2017), but alterations may be region-specific and time-sensitive (Ojo et al., 2018). LPI is a bioactive lipid that targets the GRP55 endocannabinoid receptor and influences cell growth, differentiation, and motility in various cell types (Calvillo-Robledo et al., 2022), and it was shown to be protective against glutamate-induced excitotoxicity (Blondeau et al., 2002). However, few research endeavours have evaluated the trajectory of lipid alterations after single and repeated mTBIs, and the relationship between lipid abnormalities and other pathological characteristics of mTBI, such as neurometabolic and mitochondrial dysfunction.

Therefore, the present study investigated the metabolic and lipidomic changes after single and repeated mTBI, and whether mitochondrial impairments are associated with mTBI-induced metabolic and lipidomic alterations.

The main findings of this report provide evidence that repeated mTBIs produce worse neurobehavioral outcomes than a single mTBI, and that these outcomes are accompanied by alterations in multiple neurometabolites and lipids. First, we confirmed concussive-like injuries for all ACHIs with a clinically relevant video sign observation protocol; however, only rats given repeated mTBIs had deficits in the sensorimotor beam-walk task at 1- and 4 h post-final injury. Secondly, with ¹H-MRS at 24 h, we found that concentrations of Gln and Glc were altered by repeated mTBI. Although Gln and Glc are involved in mitochondrial bioenergetics, we did not find evidence of impaired hippocampal mitochondrial respiration in these conditions. Finally, although single and repeated mTBI did alter lipid levels in the hippocampus, more notable repeated mTBI effects were found in the cortex.

In the present study, we used the well-characterized ACHI mTBI model, which consistently produces neurobehavioral impairments in the absence of skull fracture, significant focal contusion, or other complications that can occur in other animal models of mTBI (Shultz et al., 2017; Christie et al., 2019). Video observations confirmed that ACHI was associated with signs of brain trauma (e.g., lying motionless, seizures, absence of reflexes, etc.). We also found that sensorimotor deficits, as measured by the time it took to cross a beam and the number of hindlimb slips while crossing the beam, persisted 1- and 4 h after the final injury in rats that received repetitive injuries but not a single injury. These findings provide construct validity for the ACHI model considering that clinical studies show that mTBIs can cause prolonged subtle gait and balance impairments (Howell et al., 2017; Lynall et al., 2019).

Mounting evidence suggests that alterations in neurometabolite concentrations, due to factors induced by trauma, such as disruption of neuronal membranes, axonal stretching, and ion channel dysregulation, are implicated in the neurobehavioral sequalae of mTBI (Giza and Hovda, 2014; Kierans et al., 2014). As such, the evaluation of a panel of neurometabolites with ¹H-MRS has been postulated to have clinical utility as a diagnostic and prognostic biomarker for individuals with suspected concussion (Gasparovic et al., 2009; Vagnozzi et al., 2010; Chamard et al., 2013; Gardner et al., 2014; Veeramuthu et al., 2018; Thomas et al., 2022). Corroborating this notion, pre-clinical studies show that single (Xu et al., 2011; Lyons et al., 2018) and repeated (Wright et al., 2016) brain injuries can acutely alter some neurometabolite concentrations observed with ¹H-MRS. Here we evaluated the effect of single and repeated ACHIs on the concentration of several neurometabolites that have been implicated in mTBI, namely Glu, Gln, Glc, NAA, Cr + PCr, and GPC + PCh (Shutter et al., 2004; Yeo et al., 2011; Veeramuthu et al., 2018).

Due to the similarity of the molecular structures of Glu and Gln, resulting in similar peaks on the spectra, early clinical studies traditionally combined the absolute concentrations of Glu and Gln (Ashwal et al., 2004; Shutter et al., 2004; Yeo et al., 2011; Wright et al., 2016). This method does not allow researchers to distinguish between Glu and Gln, which may undergo opposite directions of concentration change, but recent technological improvements now make it possible to evaluate Glu and Gln separately (Ramadan et al., 2013). We did not find significant group differences in the concentration of Glu with ¹H-MRS, but we did observe a marked increase of Gln in the hippocampus 24 h after repeated but not singular mTBI. This may be due to the astrocytic uptake of excess Glu that is released acutely in response to TBI (Pellerin and Magistretti, 1994; Dorsett et al., 2017) and subsequently converted and stored as Gln to help maintain metabolite homeostasis (Ramadan et al., 2013; Guerriero et al., 2015). Future studies are required to test this hypothesis and to further understand the implications that single and repeated mTBI have on altered glutamatergic metabolism, which is linked to various pathological conditions (Allen et al., 2021). This study, to the best of our knowledge, was the first to detect a reduction in hippocampal Glc using ¹H-MRS following repeated mTBI in rodents. This decrease may be explained by a reduction in Glc uptake, as observed in several animal and human TBI studies (Yoshino et al., 1991; O’Connell et al., 2005; Peskind et al., 2011), or perhaps due to hyper-glycolysis (i.e., increased breakdown of cellular stores of Glc to produce ATP) to compensate for potential ATP depletion that can occur after TBI (Yoshino et al., 1991). However, additional research is required to confirm the presence of hyper-glycolysis after single and repeated mTBI and elucidate whether this contributes to the reductions in Glc levels observed in mTBI models.

We did not detect alterations in the hippocampal concentrations of NAA, Cr + PCr, or GPC + PCh between injury groups. These findings are consistent with some clinical (Maugans et al., 2012; Schranz et al., 2018) and rodent (Xu et al., 2011; Wright et al., 2016) studies; however, NAA may represent a potential biomarker of injury severity (Vagnozzi et al., 2008; Veeramuthu et al., 2018). Our results could differ from clinical studies that report a reduction in NAA/Cr following single mTBI (Veeramuthu et al., 2018), repeated mTBI (Vagnozzi et al., 2008), and in athletes with Persistent Post-Concussive Symptoms (Bartnik-Olson et al., 2014) for several reasons, including differences in species, sex, voxel size and positioning, protocol and analysis method. The current study focused on hippocampal gray matter, but other preclinical and clinical studies that evaluated white matter regions (e.g., corpus callosum and frontal lobes) have found lower levels of Cr + PCr in mTBI patients compared to controls with no differences between groups in gray matter (Gasparovic et al., 2009), indicating that metabolic response to mTBI may vary between brain regions. We also found no differences in GPC + PCh, which was consistent with a previous study of repeated mTBI in rats (Wright et al., 2016). There are many other possible metabolites that could be involved in mTBI neurobehavioral sequalae that should be evaluated in this and other brain injury models, and in males and females, who respond differently to mTBI (Wright et al., 2017, 2022). The underlying neuropathology causing potential metabolite alterations identified with ¹H-MRS is poorly understood; further research efforts are required if we are to better understand these mechanisms, which could lead to the discovery of therapeutic targets and biomarkers of brain injury.

Although mTBI is thought to lead to a period of increased cerebral vulnerability that may be contributed by a dysregulation of bioenergetics (Lyons et al., 2018), the association between deficient mitochondrial respiration and metabolic alterations induced by mTBI is poorly understood. Considering that Gln and Glc, along with other metabolites we evaluated, are highly involved in mitochondrial respiration (Yetkin-Arik et al., 2019; Shaghaghi et al., 2021), and that targeting the mitochondria provides protection against secondary damage (Stelmashook et al., 2019), we investigated whether impairments in mitochondrial bioenergetics were present in these conditions. We focused on the ipsilateral hippocampus as this region has been found to be affected in the ACHI model, with evidence of increased glial reactivity and alterations in several proteins associated with energy metabolism (Pham et al., 2019). Moreover, other mTBI studies have reported changes in hippocampal proteins related to energy metabolism (Opii et al., 2007; Bolton and Saatman, 2014; Girgis et al., 2016). Mitochondrial dysfunction is also observed in disorders that are highly comorbid with mTBI, such as emotional and cognitive disorders (Allen et al., 2018; Lejri et al., 2019). However, we found no changes in OCRs of ETC complexes I, II, and IV, or mitochondrial respiration States 2, 3, 3u, and 4o, 24 h after single and repeated mTBI. Furthermore, there were no significant findings in the RCR, indicating that mitochondrial function was not altered between groups at 24 h post-final ACHI (Brand and Nicholls, 2011). It is possible that mitochondrial bioenergetics would be altered by single and repeated mTBI at later timepoints, considering that other groups have found reduced hippocampal mitochondrial respiration 48 h post-final injury after either 4, 7, or 10 weight-drop concussions (Severo et al., 2020), and 28 days after a single closed-head injury (Lyons et al., 2018). In line with this, another group reported that a single closed-head injury reduced hippocampal State III mitochondrial respiration 48 h post-injury, but not 6- or 24 h post-injury (Hubbard et al., 2019). They also found that mitochondrial respiration was decreased at 24- and 48 h post-injury in the cortex – suggesting that concussions may induce region-specific deleterious effects – and a second injury prolonged cortical mitochondrial impairment (Hubbard et al., 2019). Our results may also differ from other studies because of differences in injury method, impact velocity, species, anaesthetic protocols, surgery, and head restraint (Satchell et al., 2005; Picard et al., 2018; Shen et al., 2019; Salman et al., 2021). It is also important to consider the young age of the animals in this study, given that age influences response to and recovery from brain trauma (Tremblay et al., 2019), which may be tied to bioenergetic and lipidomic factors (Gilmer et al., 2010). It could be that mitochondrial functions become compromised after a threshold of cellular “wear and tear” (i.e., allostasis) has been reached – which may naturally occur with age (Allen et al., 2021) – compromising energy production, metabolism homeostasis, and mitochondrial membrane integrity, leading to mTBI neuropathology (Satchell et al., 2005; Picard et al., 2018; Shen et al., 2019; Salman et al., 2021). Notably, we observed more variable functionality in the ETC complexes and mitochondrial states of respiration in rats that were analysed 30 h post-final injury compared to rats that were analysed at 24 h post-injury. This is likely due to the limited number of rats able to be scanned per day necessitating multiple mitochondrial assays, with inter-assay variability apparent. It is also possible that varied resting times after scans before tissue collection, and subsequently, increased time between first and last tissue collection, were a factor. It would be valuable to assess mitochondrial function in the cortex after ACHIs, considering that lipid imbalances were more pronounced in the cortex than the hippocampus.

Region-specific lipid alterations induced by mTBI likely depend on the nature of the injury and its pathophysiological consequences. We found that mTBI-induced lipid alterations were more pronounced in the cortex – where repeated ACHIs elevated PE, AC, and BMP – than in the hippocampus, where no significant group differences were observed. These findings are consistent with a closed-head mTBI study that reported elevations in cortical PE levels 24 h after repeated mTBI in Tau transgenic mice, but not in the hippocampus. Notably, this study also found that PE levels had normalized 3- and 12 months post-injury (Ojo et al., 2018). However, in contrast, they also found that repeated mTBI decreased LPE and LPC levels at 24 h post-injury in the hippocampus, but not in the cortex, where LPE and LPC were elevated 3 months post-final injury (Ojo et al., 2018). Work with various TBI models suggests that PE is elevated in the acute phase (i.e., 24 h post-injury) but not in the chronic phase after an injury (Abdullah et al., 2014; Emmerich et al., 2017; Hahnefeld et al., 2022). This could also be true for LPE, LPC, and other lipids (Hahnefeld et al., 2022). Our results are in line with clinical reports that have detected higher levels of PE in cerebrospinal fluid as early as 1 day post-TBI (Pasvogel et al., 2008); however, PE levels were lower by day 6 post-injury (Pasvogel et al., 2010), further demonstrating the time-sensitive nature of lipid alterations. It is thought that heightened levels of PE may be indicative of compromised membrane integrity (Yoon et al., 2022). We also found that repeated mTBIs heightened AC levels in the cortex, which is consistent with another study that reported heightened AC levels in the developing brain after a lateral fluid-percussion injury (Chitturi et al., 2019), and a clinical study that reported higher serum AC levels is association with poorer functional outcomes 1 and 6 months post-TBI in patients with intracranial injury (Lee et al., 2023). It is possible that accumulations of AC are mediated by mitochondrial dysfunction (McGill et al., 2014) and contribute to the inflammatory response following brain trauma (Rutkowsky et al., 2014). BMP, which regulates endosomal-lysosomal trafficking and serves as a docking structure for enzymes involved in sphingolipid breakdown, is important for the integrity and function of lysosomes and has been implicated in the inflammatory response and neurodegenerative disorders (Showalter et al., 2020), which are more likely to develop after TBIs (Brett et al., 2022). BMP was increased by repeat mTBI in these conditions, but it is unclear whether alterations in BMP underlie neurobehavioral alterations or represent a biomarker of mTBI pathology. Although speculative, the accumulation of PE, BMP, and AC in the cortex but not the hippocampus may suggest that cortical mitochondria are more vulnerable after mTBI. However, cortical lipids may be altered more than hippocampal lipids due to the increased likelihood of cortical vs. hippocampal damage from the direct impact and consequent movement of the brain inside the skull. More research is also required to elucidate how imbalances in lipid metabolism, signalling, and transport influence the progression of mTBI pathology.

There are several limitations in this study. First, only male rats were included, even though evidence suggests that there are sex-based differences in various aspects of the response to mTBI (Wright et al., 2017, 2021, 2022), including neurometabolite concentrations (Lyons et al., 2018) and mitochondrial dysfunction (Kalimon and Sullivan, 2021). As such, further investigation into possible sex-based changes in metabolic and lipidomic pathology in response to single and repeated mTBI is warranted, especially considering that sex hormones have neuroprotective properties (Wei and Xiao, 2013). Although the ACHI procedure features an unrestrained head, the foam platform and piston extension may limit the degree of head rotation when compared to other models, such as the lateral impact and Closed-Head Impact Model of Engineered Rotational Acceleration models. Another limitation is that we did not account for mitochondria number, which could influence overall ATP production. There is a possibility that changes in mitochondrial number may consequently affect Gln/Glc levels due to altered whole cell capacity. It should also be emphasized that whilst a single mTBI did not induce sensorimotor deficits or alter levels of metabolites, lipids, or mitochondrial bioenergetics, evidence of a concussive like injury was provided by acute neurological observations, and a consistent elevation in serum NfL in a subset of these rats presented previously (O’Brien et al., 2021). However, assessments of hippocampal function (e.g., learning and memory) as an additional measure of injury severity at each timepoint would have been beneficial, and whether alterations in neurometabolites correlated with cognitive impairment. Furthermore, our neurometabolic data is limited to one acute timepoint, although patterns of expression may change over time. Future studies could determine the persistency of concussive symptoms, including learning and memory deficits, as well as related neurometabolic and lipidomic alterations.

Alterations in metabolites Gln and Glc were observed after repeated mTBI in the hippocampus using ¹H-MRS, but these alterations were not associated with hippocampal mitochondrial dysfunction. Lipids PE, BMP, and AC were elevated in the cortex after repeated mTBI, but not the hippocampus. Our data supports the fact that repeated mTBIs in short succession can have cumulative effects, evidenced by the poorer performance in sensorimotor tests when rats received repeated vs. singular injuries, as well as greater neurometabolic and lipidomic abnormalities. More research is needed to understand the underpinning neuropathological changes responsible for these alterations.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The animal study was reviewed and approved by Alfred Research Alliance Animal Ethics Committee.

LP, SB, SS, BD, and SM contributed to the study design. LP, SB, WO’B, SS, BD, DW, and SM contributed to data collection. JA completed data analysis and drafted the manuscript. GS assisted data analysis. All authors contributed to the article and approved the submitted version.

This work was supported by the National Health and Medical Research Council to DW [grant number: 1174040] and SS [grant number: 1141643].

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.