Mitochondria play an essential role in creating energy to drive cellular function and biological processes. Dysregulation of mitochondrial function has been closely linked to numerous neurological diseases and disorders (Cabral-Costa and Kowaltowski 2020; Norat, Soldozy et al. 2020). Secondary injury occurs after initial damage of the CNS, and leads to systemic as well as tissue metabolic changes (Scholpa and Schnellmann 2017; Rabchevsky, Michael et al. 2020; Slater, Dominguez-Romero et al. 2022). The secondary injury propagates for weeks-to-months following the primary insult, and contributes to the lack of recovery in SCI patients. Over the last few years, progress has been made in studying and understanding the role and use of dietary interventions to treat the metabolic dysregulation in neurological diseases and disorders, including neurotrauma events ((Carneiro and Pellerin 2021; Chelluboina and Vemuganti 2021; Yarar-Fisher, Li et al. 2021; Field, Field et al. 2022; Yassine, Self et al. 2022). Furthermore, improving mitochondrial function after SCI has been already been shown to lead neuroprotection and functional recovery (Patel, Sullivan et al. 2012; Patel, Cox et al. 2017).
In this study, we sought to evaluate the use of KE on restoring energy metabolism following SCI. We studied mitochondrial respiratory function and oxidative phosphorylation (OXPHOS) on injured cervical spinal cord segments from rats during the acute and subacute phases of the injury. Additionally, we also wanted to investigate whether KE might potentially mimic some of the beneficial effects on mitochondrial function that we found using a ketogenic diet (KD) (Seira, Kolehmainen et al. 2021). To induce ketosis in rats that underwent C5 hemi-contusion, we delivered the KE orally (OKE), and incorporated them in the diet (KED). After SCI, animals tend to lose their appetite for the first 24H-48H. Based on previous work from our lab (unpublished data), we knew that in order to increase β-hydroxybutyrate (BHB) levels in the bloodstream of the animals during those initial hours, an oral supplementation of KE was required. Our treatment paradigm was designed with that consideration in mind, and combines the use of KED and oral administration of KE (OKE). Based on previous literature, we knew that those initial days after injury are critical in regards to cell bioenergetics dysregulation (Sullivan, Krishnamurthy et al. 2007; McEwen, Sullivan et al. 2011), so an intensive treatment regime was designed (Fig. 1A). Using this treatment paradigm, we were able to induce ketosis in the animals at all timepoints studied. We found a certain degree of variability between the different cohorts regarding the BHB levels that were reached. For example, the levels 8 hours post oral gavage (HPOG) after 2 days of treatment were singly lower in the 1 week post–SCI timepoint compared to the others. BHB levels usually peak between 3 and 7 days post-treatment, and are followed by a trend towards a decrease after that. In addition, the differences between control and treated groups becomes smaller. We believe that this time adaptation to the treatment could be associated with cell regulation of some receptors such as the Monocarboxylate transporters (MCTs) (Leino, Gerhart et al. 2001). MCTs have been identified as transporters of lactate, pyruvate, and ketone bodies (Perez-Escuredo, Van Hee et al. 2016). Some of these transporters are expressed not only in the central nervous system (MCT1, MCT2 and MCT4) (Pierre and Pellerin 2005), but in other organs and tissues such as skeletal muscle, liver, and fat tissue (Bonen, Heynen et al. 2006). MCTs can be regulated in tissues in response to changes in production or availability of their targeted metabolites (Hajduch, Heyes et al. 2000; Pierre, Parent et al. 2007). For example, diet-induced ketosis has been shown to increase MCT1 levels in rat brain (Leino, Gerhart et al. 2001), and in the injured spinal cord (Streijger, Plunet et al. 2013). Hence, we hypothesized that there may be an adaptive mechanism that explains the reduction in BHB levels in the bloodstream over time, in which an increase in circulating BHB can increase the expression of MCTs in some tissues, leading to a rapid increase in BHB uptake by the cells.
Our mitochondrial respirometry results showed that KE selectively increases mitochondrial ETC activity in the spinal cord after SCI; increases were only observed at 24H and 2 weeks post–SCI. In particular, KE led to changes of Complex I and LEAK respiration at 24H; and changes of Complex I and II, OXPHOS+ CI, and LEAK respiration at 2 weeks post-SCI. Simil arly to other studies, mitochondrial function assessment was performed in tissue homogenates. Although this approach is still very relevant to assess the overall energetic status of the cord after trauma, the spinal cord contains a variety of different cell types (i.e. astrocytes, oligodendrocytes and neurons). Determining whether KE rescues mitochondrial function in all cell types or is targeting mostly a specific cell type would be of great value in order to develop more targeted therapies. Interestingly, and in that regard, a recent work by Koppel et al. demonstrated that BHB preferentially enhances neuron over astrocyte respiration in a naïve state (Koppel, Wilkins et al. 2023).
Temporally, mitochondrial dysfunction has been shown to start as early as 2h after SCI and to continue until 24H after SCI (Sullivan, Krishnamurthy et al. 2007). Our measurements obtained beyond the 24H mark show that, in fact, mitochondrial dysregulation continues for at least 2 weeks after SCI. This suggests that a feasible therapeutic window not only needs to start early, as previously suggested (Sullivan, Krishnamurthy et al. 2007), but may also need to be extended for at least 2 weeks following injury. Contrary to what we observed using a KD (Seira, Kolehmainen et al. 2021), the treatment with KE did not show significant benefits at 1 week after SCI, however similar effects were seen at 2 weeks after SCI. While no differences between the injury parameters were found between groups (see Supplementary Fig. 1), inter-individual variability amongst animals, cohorts (van der Goot, Kooij et al. 2021), and most likely differences in the biochemical composition of the treatments might be some of the factors leading to those differences at that specific timepoint. Indeed, treatment with KE only supplies one ketone body, β-hydroxybutyrate (BHB), whereas when using a KD, the fats from the diet are broken down in the liver to the three ketone bodies: β-hydroxybutyrate (BHB), acetoacetate (AcAC), and acetone. AcAc can be further broken down into BHB (Dhillon and Gupta 2023). Additionally, AcAc has also been shown to act as signaling metabolite to promote muscle cell growth (Rahman, Muhammad et al. 2014; Zou, Meng et al. 2016; Zhong, Miao et al. 2021), increase mitochondrial function in kidney cells in vitro (Denoon, Sunilkumar et al. 2020), protect against glutamate toxicity in neurons (D'Agostino, Pilla et al. 2013), and improve motor coordination and cognition in mice with Angelman syndrome (Ciarlone, Grieco et al. 2016). Whether AcAc might activate signaling pathways in the CNS that could contribute to the early beneficial effects of the KDs after neurotrauma still needs to be further investigated.
Differences in glycemic control between KD and KE treatments might also be relevant to explain the differences in mitochondrial function rescue observed in this study when compared to our previous KD study. For example, KD has been proven to have a therapeutic effect on glucose levels after TBI (Ritter, Robertson et al. 1996). Indeed, high-glucose levels have been associated with induction of mitochondrial dysfunction in cardiac models, retina, and neurons (Russell, Golovoy et al. 2002; Dassanayaka, Readnower et al. 2015; Fiorello, Treweeke et al. 2020; Lam, Cheung et al. 2022). Contrary to KD, in our experience, KE treatment alone does not change glucose levels to the same extent that KD does (data not shown). Although glucose levels decrease initially, it quickly recovers to baseline after KE. Moreover, in a comparative study, Modica et al., described that the supplementation with ketone diester had no effect on glucose when compared to KD (Modica, Flores-Felix et al. 2021). This may suggest that in our study, the inability of KE to decrease glucose might be accountable for the less positive effects on mitochondrial bioenergetics observed at 1 week post SCI. Nonetheless, the 2 weeks post-SCI data indicate that KE can still confer relevant bioenergetic benefits, but that these may be slightly different and delayed.
Similar to our previous KD study, we investigated the protein expression levels of all the ETC complexes. Interestingly, general changes in protein expression were found at 48H, 1 and 2 weeks. Specifically KE seemed to significantly reduce the protein expression of some of the subunits of the complexes when compared to the control treated group. Only the 48H timepoint was an exception, in which we saw an increase in protein expression for Complex II after treatment (Fig. 3). In fact, at this timepoint, we also saw an increase in expression in the control treated group for Complex II and Complex V. The enhanced expression of some of the complexes after injury may reflect a compensatory mechanism to sustain cellular basal oxygen consumption after an increase in cellular energy demand in the injured spinal cord. Furthermore, although not in a time dependent manner, variability in the protein expression of the ETC complexes (up- or down-regulation) in different regions of the brain have been previously seen in a model of repeated stress in mild traumatic brain injury (mTBI) (Xing, Barry et al. 2013). Thus, it would seem reasonable to think that variability in protein expression of the ETC complexes in the spinal cord exists at different timepoints as well.
Even though KE led to changes in protein expression, the increase in activity through Complexes I and II was not correlated with an increase or change in protein expression of the targeted subunits (NDUFB8 and SDH respectively). This lack of correlation between activity and protein expression levels has been previously described in cardiac tissue; in a pressure-overload hypertrophy model in rabbits, in human atrial fibrillation (Griffiths, Friehs et al. 2010; Emelyanova, Ashary et al. 2016), and in response to exercise training (Jacobs and Lundby 2013; Montero, Cathomen et al. 2015). Proposed mechanisms for the lack of synergy between subunit protein expression and respiratory function are post-translational modifications of subunits as a result of changes in oxidation. Unfortunately, we did not measure reactive oxygen production, carbonylation, nor mutations of mitochondrial DNA in this study (Ryan, Backos et al. 2012; Alexeyev, Shokolenko et al. 2013; Emelyanova, Ashary et al. 2016; Ngo, Sverdlov et al. 2019). Moreover, the formation of enzymatic supercomplexes between some of the mitochondrial complexes might also be contributing to the lack of correlation between individual complex activity and protein content. In fact, it’s been described that these supercomplexes might compensate complex dysfunction by stabilizing individual complexes such as CI to changes in the environment. Interestingly, the regulation of these supercomplexes has been also been linked to neurological disorders (Nesci, Trombetti et al. 2021). Whether some of these mechanisms occur in the CNS, or are involved in traumatic events is still unknown. Lastly, the understanding of the mechanisms underlying the different outcomes in regards of mitochondrial activity between KE and KD treatments at 1 week after SCI, will require further research.
In summary, our study provides evidence that KE partially mitigate mitochondrial dysfunction in the acute and subacute phases after SCI. The rescue of mitochondrial function predominantly affects Complexes I and II. Furthermore, our data show that this improvement might not be fully correlated with changes in protein expression of the different ETC complexes. Overall, the work presented here provides support for the beneficial use of KE as an alternative to KD to treat acute metabolic dysfunction that occurs after SCI.