The brain–liver axis in exercise: ketone production as a driver of cognitive function?

Physical exercise (EX) is widely recognized as an effective non-pharmacological strategy to improve brain health, through the stimulation of neuroprotective mechanisms. EX increases long-term potentiation (LTP), increases hippocampal neurogenesis, reduces oxidative stress and improves mitochondrial function in the brain. Currently, there is a widespread consensus that the neurotrophin brain-derived neurotrophic factor (BDNF) plays a key role in mediating the positive effects of EX on the brain. Among the mechanisms underlying EX-induced cognitive improvements, there is a growing attention focused on inter-organ communication. In this regard, emerging evidence has underlined the importance of the liver–brain axis in improving exercise-induced cognition (Cefis et al. 2023). Over the past few decades, studies have identified β-hydroxybutyrate (BHB), the most abundant ketone body in mammals, as a critical mediator in this axis. During fasting and EX, the liver synthesizes ketone bodies as an alternative energy source for the brain. These ketone bodies also support mitochondrial function, reduce oxidative stress and enhance cognitive processes (Jensen et al. 2020). Despite these positive effects, it is unclear whether ketone bodies directly contribute the pro-cognitive effects of EX, thereby underscoring the need for further investigation. In this context, the recent study of Kelty et al. (2025), published in the Journal of Physiology, explores an innovative and under-researched connection between hepatic ketogenesis and cognitive function, offering valuable insights into liver–brain crosstalk. Using an adeno-associated virus (AAV) approach, the authors investigated the role of hepatic ketone production in exercise-induced neuroprotection by knocking down (KD) the 3-hydroxymethylglutaryl-CoA synthase 2 (HMGCS2) – a key mitochondrial enzyme catalysing the first step of ketogenesis – in the livers of healthy young female rats. To specifically target the liver, they used AAV8 and injected it directly into the portal vein to maximize liver distribution. To investigate the impact of ketogenesis on the pro-cognitive effects of EX, animals underwent two types of EX on a treadmill: an acute bout of exercise (30 min) or a chronic endurance exercise (4 weeks). The authors then used a cutting-edge multi-omic approach to deeply characterize the impact of blocking ketone body production following acute or chronic exercise. They combined proteomic and phospho-proteomic analyses in the frontal cortex using data-independant acquisition parallel accumulation-serial fragmentation (DIA-PASEF) with a metabolomic analysis utilizing ultrahigh-performance liquid chromatography MS/MS in plasma. Of note, the data collected with these promising approaches are part of the Molecular Transducer of Physical Activity Consortium (MoTrPAC), an ambitious initiative aimed at mapping the complex biological responses to physical exercise (MoTr PAC Study Group et al. 2024). In addition, the MoTrPAC initiative aims also to create a comprehensive database accessible to all, and encourages the standardization of exercise protocols to ensure consistency between studies, an improvement in reproducibility and translational impact of the outcomes. This study stands out by being conducted exclusively on female rats, addressing a significant gap in the literature, as most similar studies have predominantly focused on males. Although, according to the human protein atlas there is no sex difference in HMGCS protein expression, it would be relevant to verify whether EX response differs between sexes. The authors also integrate both in vitro and in vivo approaches, employing widely recognized and well-established models, allowing a comprehensive analysis of the mechanisms underlying exercise-induced neuroprotection. Furthermore, the study incorporates two distinct exercise protocols, facilitating a nuanced investigation of both immediate and long-term effects of physical activity on hepatic ketogenesis and neuroprotection. This dual approach provides an exhaustive understanding of how different durations of exercise influence physiological and molecular mechanisms. The present study revealed that liver-specific HMGCS2 KD significantly reduced frontal cortex mitochondrial respiration and cytosolic calcium signalling. In addition, key markers of synaptic plasticity, involved in LTP, long-term depression and synaptogenesis, were significantly diminished in the frontal cortex of HMGCS2 KD rats. These alterations were associated with impaired spatial memory, as assessed by the Y-maze test, in HMGCS2 KD rats compared to the scramble group. Interestingly, proteomic analysis in the frontal cortex revealed that HMGCS2 KD rats exhibited increased markers of mitochondrial dysfunction and reduced markers of oxidative phosphorylation and respiratory electron transport, following an acute bout of EX (1 h), compared to scramble-EX rats. While an acute bout of EX does not seem to be sufficient to reverse the negative effects in the frontal cortex caused by the HMGCS2 KD, endurance EX (4 weeks) induced positive effects. Indeed, the Y-maze performance and the levels of synaptic plasticity markers were reversed in the frontal cortex of the HMGCS2 KD-EX rats compared to the HMGCS2 KD-SED group. The use of the Y-maze provided a focused and robust assessment of spatial and short-term memory, but incorporating other cognitive tests such as a spontaneous alternation task for working memory or the Morris water maze for long-term retention could provide a more comprehensive understanding between different memory process. Consistent with this, while the authors focused on the frontal cortex, providing a valuable and precise examination of a brain region critical for spatial memory and short-term memory, extending the investigation to the hippocampus – the main region involved in both short- and long-term memory as well as spatial navigation – could offer complementary insights and a wider understanding of how hepatic ketogenesis affects different brain cognitive areas. Despite the inhibition of hepatic HMGCS2 via the AAV short hairpin RNA (shRNA) approach, circulating plasma ketone bodies remain detectable. Interestingly, an acute bout of EX is sufficient to increase ketone levels in HMGCS2 KD rats. These findings raise several important considerations, particularly regarding the efficacy of the sh-AAV-mediated KD of liver HMGCS2. Furthermore, since HMGCS2 enzyme is also expressed in the kidney and gastrointestinal tract, this suggests a potential metabolic compensation mechanism that may partially offset the loss of hepatic ketogenesis. These findings emphasize the importance of inter-organ crosstalk in regulation of energy metabolism and suggest that exploring extra-hepatic ketogenesis could provide further insights into the role of ketone in exercise-induced neuroprotection. Moreover, ketone bodies have been shown to exert effects beyond their primary role as energy substrates, particularly by inducing an anti-inflammatory response. This effect may involve activation of the GPR41-dependent pathway in microglia and/or afferent neurons, as suggested by the authors in their discussion. In the context of future translational studies, it would also be pertinent to evaluate the impact of impaired ketogenesis on neurodegenerative diseases. Given the crucial role of ketone bodies in maintaining brain energy metabolism, mitochondrial function and inflammation control, chronic deficits in ketogenesis could contribute to cognitive decline and neurodegeneration. In this regard, investigating whether there are human models with genetic mutations that impair ketone metabolism, such as mutations in the HMGCS2 gene, could provide better understanding of the long-term neurological consequences of disrupted ketone metabolism. Such studies could help bridge the gap between preclinical findings and potential therapeutic strategies targeting metabolic pathways in neurodegenerative diseases. While previous research has linked EX to improved cognition, these findings reinforce the well-established neuroprotective effects of EX by identifying hepatic ketogenesis as a potential mediator of these benefits. From an ageing and neurodegenerative perspective, this study is particularly relevant since cognitive decline is associated with metabolic dysfunction, mitochondrial impairment and reduced neurotrophic support. Conversely, cognitive impairment is also commonly observed in metabolic diseases, highlighting the bidirectional relationship between metabolic health and brain function. This study investigated the impact of two types of endurance EX and revealed that practising chronically was more effective in reversing the effects of HMGCS2 KD. In light of their findings, it would be relevant to explore other modalities of EX, such as high-intensity interval training, which may produce similar effects and increase the possibility of proposing this modality to a large population. Indeed, exploring different exercise modalities would provide a broader understanding of how various physical activities influence these mechanisms and their potential therapeutic applications. Finally, investigating the effect of different types of EX could both supply the MoTrPAC and, in the future, it will help clinicians to adapt exercise prescription based on the specific needs of different individuals. In summary, the present study provides critical insights into the relationship between ketogenesis and brain function, showing that endurance exercise can mitigate cognitive deficits associated with impaired hepatic ketone production. These findings underscore the essential role of ketone bodies, particularly BHB, as a neuroprotective metabolite that supports mitochondrial function, neuroplasticity and cognitive resilience. This research emphasizes the necessity of investigating inter-organ crosstalk in mediating the benefits of EX. In the context of EX, the liver–brain axis appears to be a driver of cognitive function and understanding how peripheral metabolites influence neural outcomes could open new therapeutic ways for preventing cognitive decline. Future studies should continue exploring the interplay between exercise, ketone metabolism and brain health to optimize strategies that mitigate neurodegenerative disease and age-related cognitive impairments. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. No competing interests declared. J.W.: Conception or design of the work; Drafting the work or revising it critically for important intellectual content; Final approval of the version to be published; Agreement to be accountable for all aspects of the work. P.G.: Conception or design of the work; Drafting the work or revising it critically for important intellectual content; Final approval of the version to be published; Agreement to be accountable for all aspects of the work. M.C.: Conception or design of the work; Drafting the work or revising it critically for important intellectual content; Final approval of the version to be published; Agreement to be accountable for all aspects of the work. All authors have approved the final version of the manuscript and agreed to be accountable for all aspects of the work. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed. None.