Past, present and future of research on brain energy metabolism in bipolar disorder

Nearly a hundred years ago, Looney and Childs reported that the level of lactate in the blood of patients with schizophrenia was 38.8% higher than in healthy individuals1, making bioenergetic dysfunction one of the oldest biomarkers proposed in psychiatry. Blood lactate concentration is classically understood to be an indicator of inefficient metabolism, and this finding is buttressed by related neglected putative biomarkers in major psychotic disorders, including reduced pH and abnormal carbohydrate metabolism. These historic findings suggest a shift away from efficient aerobic energy generation towards the less efficient glycolytic pathway, which has been confirmed by contemporary methodologies. Indeed, the brain is the most energy dependent tissue, and the phenomenology of many psychiatric conditions features energy dysregulation. Bipolar disorder is likely to be a disorder of brain energy metabolism. It can be conceptualized as the outcome of a state dependent biphasic dysregulation of energy generation, with depression characterized by inability to meet metabolic demands and mania by excessive energy generation2. Neuroimaging evidence supporting this hypothesis has been available for almost half a century, with Baxter et al3 showing that people in bipolar depressed or mixed states had significantly lower supratentorial whole brain glucose metabolic rates than those in unipolar depression or bipolar mania and healthy controls. In contrast, resting energy expenditure is increased in mania, as is oxygen consumption (VO2). Mitochondria are at the heart of energy generation. They are the source of adenosine triphosphate (ATP), hence energy through oxidative phosphorylation in the electron transport chain. In addition, mitochondria have other critical roles such as calcium homeostasis, reactive oxygen species production, and modulation of growth factors such as brain-derived neurotrophic factor. These aggregate roles regulate synaptic plasticity and the activation of neurotransmitters, which are ultimately related to higher-order functions such as learning and memory. Critically, dysregulation of all these elements are documented in bipolar disorder4. Mitochondrial dysfunction can instigate a vicious cycle of progressive cellular damage that aggravates the disease, drives neuroprogression and premature aging, and compromises clinical and functional outcomes. Progress in mitochondria research has been largely driven by post-mortem human brain analyses and animal studies. For example, Andreazza et al5, using post-mortem methodology, explored the components of the mitochondrial electron transport chain in the prefrontal cortex. Bipolar disorder patients had significantly reduced complex I activity compared to people with major depressive disorder or schizophrenia, adjusted for antipsychotic or antidepressant medication, age, sex, post-mortem interval, and brain pH. However, phase-specific effects are poorly captured by post-mortem methodology4. Many known medications for bipolar disorder have extensive effects on mitochondrial function6. Lithium regulates the expression of numerous mitochondrial enzymes. In human brain tissue, it activates several respiratory chain enzyme complexes, including complexes I + III, II + III and specific complex I subunits such as NDUFB9, NDUFAB1 and NDUFS7, as well as succinate dehydrogenase. It also inhibits GSK-3B, which in its turn inhibits the conversion of pyruvate into acetyl-coenzyme A (acetyl-CoA). Both lithium and valproate have antioxidant effects; they can attenuate cell death induced by peroxide or rotenone, and they both stabilize intracellular calcium dynamics and thereby inhibit glutamate-induced increases in intracellular calcium levels, protein oxidation, lipid peroxidation, DNA fragmentation, and cell death. Human magnetic resonance spectroscopy data show that lithium increases levels of N-acetylaspartate (NAA), a putative neuroprotective marker. Lithium also increases NAA/creatine ratio, also regarded as a marker of neuronal health. Importantly, a rodent model of bipolar disorder with face, predictive and construct validity was developed by Kasahara et al7. They generated transgenic mice with a neuronally specific mutation of the mitochondrial DNA polymerase (POLG) gene. These mice had abnormal monoamine function and mitochondrial DNA mutations in the forebrain. They showed distorted day-night wheel-running activity, and switched to mania-like high activity with antidepressant administration and to standard "normal" activity with lithium treatment, which supports the model's predictive and face validity. The mitochondrial DNA abnormalities were consistent with post-mortem studies in bipolar disorder patients. We do not have robust biomarkers reliably indexing mitochondrial function in stored samples. While metabolomic markers such as lactate and the lactate-to-pyruvate ratio are used, levels are affected by many medical and lifestyle variables, and can change after venipuncture through ongoing metabolism. Several exercise physiology measures can provide insights into mitochondrial function, including maximal oxygen consumption, VO2max (the maximum amount of oxygen used during intense exercise) and respiratory exchange ratio, RER (the ratio of carbon dioxide produced to oxygen consumed during metabolism). Neuroimaging technologies – such as 18F-fluorodeoxyglucose positron emission tomography, blood oxygen level-dependent functional magnetic resonance imaging, near-infrared spectroscopy, and magnetic resonance spectroscopy – can index brain energy metabolism, but carry cost as well as feasibility and hence translational barriers. Presently, the flux biogenesis assay (Seahorse) is the most advanced methodology to assess mitochondrial function. It can measure oxygen flux, oxygen consumption rate, proton flux, as well as the extracellular acidification rate, a marker of glycolysis. However, it carries methodological complexities, usually requiring live cells such as lymphocytes that do not necessarily reflect central nervous system activity. So, what does the future hold? If there is indeed a regulatory failure in brain energy metabolism underpinning the state dependent oscillations between mania and depression, then detecting the responsible regulatory target is a top order candidate to unpick the cause of the disorder and identify mechanistic targets for intervention. Several master regulators of mitochondrial function have been explored, including peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC-1α) and calcium/calmodulin-dependent protein kinase 28. Several plausible therapeutic options targeting mitochondrial biogenesis and function are already available. There is active interest in the ketogenic diet as a potential therapeutic strategy. Ketones are an alternative energy source by feeding acetyl-CoA into the Krebs cycle, by-passing the traditional pathway through glycolysis. Exercise, which is well established to have benefits in mood and related disorders, impacts mitochondrial biogenesis amongst other pathways. Mitochondria are also a potentially druggable target. Several strategies have been explored in recent trials, including N-acetylcysteine, coenzyme Q10, alpha lipoic acid, acetyl-L-carnitine, and choline. Very preliminary findings suggest that augmenting pharmacological mitochondrial strategies with lifestyle strategies such as exercise might provide synergistic benefits. Trimetazidine is a selective inhibitor of 3-ketoacyl-CoA thiolase, a key enzyme in fatty acid oxidation. By selectively inhibiting beta-oxidation of free fatty acids, this drug promotes glucose oxidation and increases ATP production per molecule of oxygen in the brain, as measured by Seahorse assay9. It has been in use for 50 years as an anti-angina agent to increase metabolic efficiency in the heart when metabolic processes are compromised, and may have potential in bipolar disorder. In summary, research on the mitochondrial foundation of bipolar disorder is rapidly progressing, and has the potential to unlock a variety of new therapeutic prospects.