Human physiology evolved during times of dramatic fluctuations in energy supply and demand. Coping with these changes has entrained the human body with the ability to manage energy metabolism for optimal substrate storage and use during states of either food surplus or famine, and periods of either rest or increased energy demand. The ability to efficiently adapt metabolism depending on demand or supply is known as metabolic flexibility (1). In general, the human body can aptly use moderate amounts of carbohydrates, fatty acids, and amino acids. The modern era, however, is characterized by unprecedented levels of food supply (2). This near continuous intake of calorically dense processed foods, combined with physical inactivity, reduces a predilection for, and directly impedes, metabolic flexibility (3). This is caused by substrate competition and metabolic insensitivity, characterized by distorted nutrient sensing, blunted substrate switching, and impaired energy homeostasis (4). Importantly, this metabolic inflexibility may underlie the epidemic changes in metabolic disease that affect all demographic groups and burden health-care systems (5, 6).
Defining Metabolic Flexibility
Maintaining energy homeostasis requires substrate sensing, trafficking, storage, and utilization, dependent on substrate availability (push concept) and energy requirement (pull concept). Metabolic plasticity (or adaptability) was recognized in 1983 by Saltin and Gollnick (7) when they reviewed the metabolic adaptations of skeletal muscle to exercise (see “Physical exercise”). The term metabolic flexibility was coined by Kelley et al. (8) in 1999 when they studied fuel selection in skeletal muscle in lean and obese individuals after an overnight fast. Specifically, they discovered that skeletal muscle of lean individuals showed a remarkable ability to adapt fuel preference to fasting and insulin infusions and were therefore designated as metabolically flexible (8). Insulin-resistant obese patients however manifested a lesser reliance on fatty acid oxidation compared with lean individuals and did not show increased fatty acid oxidation after fasting or reduced fatty acid oxidation after insulin infusion. Because of their inadequate responses to metabolic challenges, these patients were named “metabolically inflexible” (9). More recent work showed that, upon consumption of a high-fat diet, lean subjects with adequate metabolic flexibility were able to increase fatty acid oxidation (FAO) at the expense of glucose, whereas obese individuals were not (10). Lean individuals also showed an increased expression of genes involved in fatty acid transport and oxidation compared with little or no change in their obese counterparts (10).
The concept of metabolic flexibility was particularly linked to the capacity of mitochondria to select fuel in response to nutritional changes (4) and placed mitochondrial function at its core (11). Later, metabolic flexibility quickly expanded to encompass the ability of a given system (whole-body, organ, single cell, or organelle) to handle specific nutrients. At a molecular level, metabolic flexibility relies on the configuration of metabolic pathways that manage nutrient sensing, uptake, transport, storage, and utilization. This metabolic organization is mediated by synthesis, degradation, or activity regulation of key proteins in metabolic circuits or enzymes with a high metabolic flux (11, 12). Metabolic flexibility is not an “on-off” phenomenon, but involves tightly regulated subtle adjustments. Taken together, metabolic flexibility can be understood as an adaptive response of an organism’s metabolism to maintain energy homeostasis by matching fuel availability and demand to periodic fasting, varying meal composition, physical activity, and environmental fluctuations (13, 14).
Physiological Relevance of Metabolic Flexibility
Feeding/fasting
In healthy individuals, metabolism is distinctly different during a state of fasting compared with a state of caloric availability.
On a systemic level, maintaining metabolic homeostasis during feeding or fasting relies on multiorgan coordinated control of available fuel.
During fasting, the decrease in circulating dietary carbohydrates and lipids and decline in insulin-glucagon ratio induces a switch toward FAO (20). Glucagon stimulates hepatic glycogenolysis and ketogenesis, and the decrease in insulin suppresses hepatocyte malonyl-coenzyme A (CoA) synthesis and lipogenesis, with concomitant activation of FAO (21, 22). After an overnight fast, ketone bodies are in the micromolar range (∼30 μM for women and ∼60 μM for men), which is not detected by most assays (23). This is important because it emphasizes that an overnight fast already induces changes in metabolic flexibility in healthy humans.
Biochemical transition between feeding and fasting
Cellular fuel selection depends on the type and amount of nutrient available. Cellular responses to the changes in nutritional state are predominantly assigned to the mitochondria (see “Mitochondrial function is essential for metabolic flexibility” and “Endocrine regulation of metabolic flexibility”). As proposed in Randle’s glucose-fatty acid cycle hypothesis, in vitro studies have shown that FAO is suppressed when glucose consumption is increased and vice versa (25) (Fig. 2). In the postprandial period of a carbohydrate-rich meal when glucose and insulin are high, glucose uptake, glycolysis, and pyruvate oxidation are favored and FAO is suppressed. These glucose-dependent processes increase the concentration of malonyl-CoA, which allosterically inhibits carnitine palmitoyltransferase-1 (CPT-1), which transports fatty acids into the mitochondria for β-oxidation. Fatty acids are then alternatively used for triglyceride synthesis and stored. Additionally, the rise in pyruvate from glycolysis inhibits pyruvate dehydrogenase kinase (PDK), which reduces the phosphorylation of pyruvate dehydrogenases (PDHs), resulting in an increased glucose oxidation (4).
On the other hand, during fasting, the inhibition of FAO is released by the action of the energy stress sensor adenosine monophosphate-activated protein kinase (AMPK), which inhibits acetyl-CoA carboxylase (ACC) by phosphorylation. Inhibition of ACC lowers malonyl-CoA concentrations, resulting in an increased activity of CPT-1 and amplified transport of fatty acids into the mitochondria for β-oxidation. When FAO is preferred, acetyl-CoA and reduced nicotinamide adenine dinucleotide (NADH) levels rise, impeding PDH catalytic activity through allosteric inhibition and via activation of PDK. During fasting, PDK gene expression increases through fatty acid–dependent peroxisome proliferator–activated receptor (PPAR) signaling (26).
Additionally, an increase in fat availability elevates citrate levels, which inhibit phosphofructokinase and GLUT, impeding glucose uptake and use. Inhibition of phosphofructokinase also increases the cytosolic concentration of glucose-6-phosphate that inhibits hexokinase, making it more difficult to metabolize glucose (27). This allosteric inhibition results in a feed-forward loop, favoring FAO during times of nutrient scarcity to conserve glucose for biosynthetic processes and brain metabolism. Because glucose-derived pyruvate in the liver is no longer converted to acetyl-CoA by PDH, pyruvate can be used as a gluconeogenic precursor to avoid hypoglycemia.
Prolonged fasting and caloric/dietary restriction
The importance of energy and nutrient sensing transcription factor regulated pathways can be demonstrated during prolonged fasting. It is somewhat surprising that, until 1967, ketone bodies (KBs) were thought to have no beneficial physiological role during prolonged fasting. Currently, we know that ketone bodies are produced by the liver upon prolonged fasting. The central nervous system requires approximately 140 g of glucose per day (equivalent to almost 600 kcal), also during fasting. Plasma KB then are an important source of energy because their blood levels and oxidation rates increase (28). During fasting, KB can act as an excellent respiratory fuel: 100 g of D-3-hydroxybutyrate yields 10.5 kg of ATP (22 ATP per molecule D-3-hydroxybutyrate), whereas 100 g of glucose yields only 8.7 kg of ATP (29).
An alternative, and quantitatively less important, source of fuel during prolonged fasting is the breakdown of amino acids, in particular branched-chain amino acids (BCAAs). The mitochondrial branched-chain α-ketoacid dehydrogenase (BCKD) complex is the rate-limiting step in BCAA catabolism (33). BCKD can be allosterically inhibited when acyl-CoA concentrations and NADH levels are sufficient (34). This ensures that under fed conditions and during short intervals of fasting or light exercise, cellular proteins are conserved. With nutrient abundance, however, when BCAA are present in excess such as after a protein-rich meal, BCKD also becomes active as BCAA-derived α-ketoacids allosterically inhibit BCKDs deactivating kinase (35).
Dietary restriction, more commonly known as caloric restriction (CR), is the prolonged and controlled reduced intake of all dietary constituents while maintaining appropriate intake of vitamins and minerals. Recycling of cell-intrinsic macromolecules is essential to sustain metabolic processes when nutrients remain chronically scarce. This intricate salvaging process is controlled by autophagy (36). Autophagy is regulated by the deactivation of the nutrient sensors mechanistic target of rapamycin (mTOR) and v-Akt murine thymoma viral oncogene homolog 1/protein kinase B (Akt/PKB), and activation of the cellular energy status sensors AMPK and sirtuins (SIRTs) (37).
In mammals, reduced insulin/IGF-1 signaling through reduced nutrient intake inhibits Akt, which leads to activation of the forkhead box protein O (FOXO) transcription factors that upregulate the cells’ maintenance pathways: DNA repair, autophagy, and stress resistance (38). mTOR signaling stimulates growth and blocks tissue maintenance when nutrients are plentiful. However, upon CR, reduced intake of proteins, particularly of BCAA, downregulates the mTOR pathway, causing a switch toward salvage pathways such as autophagy and conserved translation (39). In response to increasing cellular AMP/ATP ratios, which rise during CR, AMPK activates catabolic pathways and represses anabolic pathways (40). In parallel, SIRT activity is increased during CR, which depends on oxidized nicotinamide adenine dinucleotide (NAD+) concentrations, and leads to protein deacetylation and improved mitochondrial function (see “SIRTs”; Fig. 3).
Caloric excess
With caloric excess, the mitochondria are overwhelmed by an excess in in substrates derived from fatty acids, glucose, and amino acids. In this setting, metabolic flexibility is altered by a push-mechanism (see “Defining Metabolic Flexibility”). Through persistent allosteric inhibition and feed-forward responses this surplus leads to mitochondrial metabolic “indecision” and ineffective substrate switching, resulting in incomplete substrate utilization for energy production and subsequent storage of substrates in (ectopic) depots (4).
With a low electron flux through the electron transport system, the NADH/NAD+ ratio in the mitochondria increases causing redox inhibition of several tricarboxylic acid (TCA) cycle enzymes and reduces the supply of carbons into the TCA cycle. This results in elevated levels of acetyl-CoA and other acyl-CoA at crucial sites where catabolism of fatty acids, glucose, and amino acids converge, leading to a state in which upstream substrate catabolic flux slows down considerably, a phenomenon coined as mitochondrial metabolic “gridlock” (4).
Physical exercise
Physical exercise
A good example of cell intrinsic metabolic programming upon physiological stimulation occurs in skeletal muscle. Skeletal muscle consists of oxidative (type I) and glycolytic (type II) fibers, which differ in their metabolic abilities. Oxidative muscle fibers have a high mitochondrial density; hence, they prefer oxidative phosphorylation for ATP production. They also contain more lipid droplets and rely on FAO. Glycolytic muscle fibers have a low mitochondrial density and rely predominantly on the breakdown of stored glycogen by glycolysis for their ATP production (45, 46). During low-intensity exercise, oxidative muscle fibers predominantly rely on FAO for their ATP production. During more intense exercise, the rising ATP utilization rate induces a metabolic switch from FAO to glucose metabolism. When maximal pyruvate production outstrips its mitochondrial import during severe intensity exercise, pyruvate is converted into lactate and NAD+ by lactate dehydrogenase in the cytosol, after which lactate is excreted from the cell. NAD+ generation helps maintain cytosolic redox potential and promotes substrate flux through glycolysis to sustain ATP generation (47).
As such, regular physical exercise is a classic example of how metabolic flexibility is regulated by transcription factors. During acute exercise, an increased AMP/ATP ratio, sensed by AMPK, increases transcription, translation, and activity of the transcriptional coactivator PPAR gamma coactivator 1-alpha (PGC1α) (40, 48). PGC1α is a regulator of exercise-induced adaptations in the capacity of oxidative phosphorylation (OXPHOS) in skeletal muscle (49, 50). In particular, PGC1α interacts and coactivates many transcription factors and nuclear receptors that are involved in mitochondrial energy homeostasis and metabolic adaptations, such as nuclear respiratory factors (NRFs) and PPARs. NRFs regulate the expression of nuclear genes encoding OXPHOS proteins, and PPARs regulate the transcription of genes that encode enzymes involved in lipid transport and catabolism (51). The increase of mitochondrial biogenesis and FAO improves insulin sensitivity. The role of PGC1α in metabolic flexibility is underlined by observations that basal PGC1α skeletal muscle expression is reduced in sedentary subjects (49).
Hibernation and cold exposure
Research in migratory birds (53), killifish (54), and in animals that enter states of torpor or hibernation (55) has revealed that temperature can influence metabolic reprogramming and metabolic flexibility.
(…)
Master regulators of browning in humans are PPARγ and PGC1α (59). PPAR agonists in white adipocytes caused conversion to a brite/beige molecular phenotype in vitro, defined by increased mitochondrial oxygen consumption and amplified expression of the thermogenic gene program, including uncoupling protein 1, which is located in the inner mitochondrial membrane and uncouples mitochondrial respiration, releasing it as heat (60).
Obese subjects showed lower BAT activity than lean subjects, and low BAT activity is associated with metabolic dysfunctions such as T2DM and aging (62). Although noradrenalin can induce browning, adrenergic therapy does not activate BAT to the same extent as cold exposure, and incurs adverse cardiovascular effects (59). It seems that increasing BAT activity and induction of browning by short-term cold exposure in humans can increase insulin sensitivity (63), but more work needs to be performed before it can be regarded as a suitable treatment of obesity and T2DM.
Interplay between circadian rhythm and metabolic flexibility
The circadian clock enables organisms to anticipate the diurnal variation in metabolic substrates (64). The circadian clock can be divided into the central and peripheral clocks. The central, or master clock, is located in the hypothalamic suprachiasmatic nucleus and can function autonomously without any external input, but can be entrained by temporally relevant external input such as, temperature, light, and feeding/fasting cycles (65). The central circadian pacemaker mainly acts through its powerful influence over the endocrine system (66). The peripheral clocks are synchronized by the central clock and are present in almost all mammalian tissues, where they regulate tissue specific gene expression (65).
Indeed, the circadian clock can have major effects on metabolic flexibility and is even able to coordinate temporal and spatial organization of lipids and circadian rhythmicity of mitochondrial function (67, 68). The peripheral circadian clock has several mechanisms to influence metabolism including regulation of metabolite levels, interaction with nutrient sensors, control of rate-limiting metabolic enzymes, and modulation of nuclear receptors (65). Circadian proteins directly coordinate with key regulators of nutrient homeostasis, including AMPK, cAMP response element-binding protein, and Akt/PKB, which are themselves driven by daily rhythms of nutrient supply (69, 70). Gene expression of key metabolic enzymes show diurnal variations (71). Expression of nuclear receptors such as members of the PPAR family and estrogen-related receptor family are also under circadian clock control (65). Moreover, diurnal variations in human skeletal muscle oxidative capacity were recently observed and may be linked to the circadian clock in muscle (72).
The circadian clock is under influence from, and can synchronize to, external stimuli such as food intake and diet (64). For instance, the liver synchronizes its peripheral circadian rhythm based on the availability of circulating metabolites (71, 73). Peripheral tissues communicate dietary signals to the brain via endocrine cues such as ghrelin, leptin, and insulin, meaning feeding rhythm strongly contributes to the reciprocal relationship of the circadian clock and metabolism (74, 75). The nutrient sensors AMPK and SIRT1 that modulate circadian gene expression and circadian oscillating metabolites such as cAMP and NAD+, are known regulators of the circadian clock (64, 65, 76).
The extent to which metabolites have control over the circadian clock is currently unknown. It is likely, however, that an interplay of all metabolites (the metabolome) acts as a cue for the circadian clock because many metabolites affect its phase, amplitude, and/or period of oscillations (75).
Disruption of the circadian rhythm, or circadian misalignment, in human subjects can result in insulin resistance (78). As a result, (night)shift workers are at a greater risk to develop obesity, T2DM, cardiovascular disease, and metabolic syndrome (79, 80).
Aging
The metabolic influence on aging and lifespan has gained increased attention over the past decade; as such, major regulators of metabolic flexibility play dominant roles in aging (81, 82). Indeed, metabolic flexibility is negatively correlated with aging (83) and targeting metabolic flexibility as a cause for aging and related comorbidities may provide cues to delay the onset of age-related diseases and prolong health span. Mitochondrial dysfunction and a cellular shift toward a glycolytic phenotype is intimately linked to senescence and the “senescence-associated secretory phenotype” that entails the secretion of multiple factors such as proinflammatory cytokines, proteases, and growth factors that have potent local and systemic effects such as inflammation and metastasis (84, 85). Perturbation of mitochondrial function and nutrient-sensing pathways, particularly related to glucose homeostasis, is a hallmark of aging (86). As such, the nutrient and energy sensing pathways (insulin/IGF, mTOR, AMPK, and SIRTs) are causally involved in fitness and longevity (39, 87). Consistent with this, CR increases health span and/or lifespan in model organisms and improves several markers of health in humans (38) (see “Prolonged fasting and caloric/dietary restriction”).
It is currently unclear how metabolic flexibility is perturbed in the elderly, because few metabolic flexibility studies have been conducted in late middle-aged and aged populations.
A study in middle-aged postmenopausal women showed that endurance training improved work-related ability to mobilize and oxidize free fatty acids, suggesting that in the elderly metabolic flexibility can still be trained (88). Importantly, exercise is a principal preventive strategy to improve metabolic flexibility and prolong healthy aging (89) (see “Exercise training”).
In summary, on a cellular level, acute metabolic flexibility is a universal property of healthy cells (1). Metabolic flexibility, and therefore substrate flux, is principally determined by the reciprocity of metabolic circuitries, of which the presence is dependent on the cell’s gene and protein expression, nutrient availability and/or demand. On a systemic level, metabolically active organs such as the liver, muscle, heart, and adipose tissue, communicate to best organize the utilization of available fuel. This holistic and vis-à-vis orchestration of available nutrients to sustain whole-body energy homeostasis has ensured organism survival and is therefore interwoven with both healthy and diseased states of metabolism.
Regulation of Metabolic Flexibility
The intrinsic qualitative and quantitative capacity of cells to oxidize or store energy is dependent on the molecular organization of their metabolic pathways.
Mitochondrial function is essential for metabolic flexibility
Mitochondria are pliable organelles, and they adapt their morphology to nutrient availability and in doing so regulate OXPHOS activity and substrate preference (90, 91).
Many studies support the idea that deregulation of mitochondrial function underlies the onset of metabolic inflexibility [reviewed in Muoio (4)], although a causal link between the two still remains to be fully established (100, 101).
Endocrine regulation of metabolic flexibility
Gut endocrine regulation of metabolic flexibility
For example, glucagonlike peptide-1 (GLP-1) is released by enteral L cells and agonizes pancreatic insulin secretion. However, GLP-1 exerts direct inhibitory effects on hepatic glucose production via direct hepatic or neuronal inhibition (108). Also, GLP1 may contribute to reduced intestinal lipoprotein production. Likewise, bile acids also facilitate nutrient trafficking in a hormone-like fashion (109). Bile acids induce insulin secretion directly via the transmembrane bile acid receptor Takeda G protein–coupled receptor 5 (on β cells and indirectly via stimulation of L-cell derived GLP-1 and subsequent insulin release (110–112). Additionally, the postprandial increase in bile acids also increases insulin sensitivity and energy expenditure (109). In contrast to these two examples, the peptide ghrelin is produced in stomach X/A cells and actually decreases during food intake. The postprandial decrease in ghrelin lowers hepatic glucose production while increasing peripheral glucose-uptake in both skeletal muscle and adipose tissue (113). All in all, the combined effects of different factors during food-intake and fasting facilitate the organisms’ fuel availability and metabolic flexibility.
Other endocrine factors affecting metabolic flexibility
Many other circulating factors are involved in metabolic flexibility, including cytokines and other peptides that are expressed, produced, and released by adipocytes (adipokines), muscle (myokines), and liver (hepatokines). Although the precise function of many of these factors remains elusive, some exert autocrine, paracrine, or endocrine effects that are fundamental for organ cross-talk in the regulation of energy homeostasis (114).
Fully functional adipocytes reduce lipotoxicity in tissues such as the heart and liver and they maintain a healthy balance of adipokines, which exert paracrine effects on adipocytes in their direct vicinity and endocrine effects on the central nervous system, immune system, and peripheral tissues. Adiponectin suppresses glucose production in the liver and enhances FAO in skeletal muscle (117, 118). Indeed, high levels of the insulin-sensitizing, antiapoptotic, and anti-inflammatory hormone adiponectin has been proposed to improve metabolic flexibility of adipose tissue, enhancing its function under metabolic challenges (119). Leptin is a central feedback indicator for the brain on the amount of stored energy in the body and rises in concert with the amount of adipose tissue (120). Leptin also exhibits diurnal expression patterns that are dampened in obese subjects (see “Interplay between circadian rhythm and metabolic flexibility”) (121).
Myokines provide skeletal muscle with the ability to mediate whole-body metabolism via endocrine signaling to adipose tissue, liver, pancreas, heart, and brain (122). In particular, myokines play important roles in mediating the positive effects of exercise on whole-body metabolism [reviewed in Oh et al. (114)]. Production of myokines is predominantly influenced by skeletal muscle contraction and can alter glucose disposal, FAO, and lipolysis.
Hepatokines can also regulate whole-body metabolic flexibility and some are even considered as potential targets for the treatment of cardiovascular disease [reviewed in Jung et al. (116)]. For example, fetuin-A has a major role in the regulation of insulin sensitivity as fetuin-A–deficient mice showed improved insulin sensitivity (116).
Interestingly, some of these paracrine and endocrine factors are secreted by multiple tissues, and their local function and impact on metabolic flexibility can depend on their origin and local plasma concentrations. IL-6, for instance, when secreted by skeletal muscle, stimulates AMPK activity and in this way increases glucose uptake and β-oxidation in muscle and adipose tissue (114). IL-6, is, however, also secreted by adipocytes from obese patients and negatively affects metabolic flexibility by decreasing insulin signaling and glucose uptake because of its proinflammatory properties (114, 123). Another example is fibroblast growth factor 21 (FGF21), which, as a myokine, increases GLUT1 expression in skeletal muscle, boosting glucose uptake (114). FGF21 from muscle also exerts endocrine-like effects on WAT, increasing lipolysis and β-oxidation and inducing browning (126).
Epigenetic regulation of metabolic flexibility
Evidence that epigenetic changes drive metabolic inflexibility in humans is emerging (127).
In this way, histones can act as metabolic sensors, converting changes in metabolism into stable patterns of gene expression, a concept named “metabolic memory” (130, 131).
Metabolic sensory epigenetic programming enzymes include: histone acetyltransferases that acetylate histones using acetyl-CoA (133), SIRTs that deacetylase histones using NAD+ as a cofactor (42), histone methyl-transferases that methylate or remove methyl marks using S-adenosylmethionine as a methyl donor, or lysine-specific demethylases that use flavin adenine dinucleotide as a cofactor (134). How these epigenetic regulators are targeted to specific sites, such as promotor regions, how transient their epigenetic markers are, and how these changes are inherited, is still under active investigation (135).
Pathophysiology of Metabolic Inflexibility
Obesity
Obesity is predominantly associated with elevated levels of plasma free fatty acids (141). High circulating levels of free fatty acids inhibit glycogen synthase activity and PDH activity, which leads to reduced disposal and oxidation of glucose. Besides adipocyte metabolic dysfunction, skeletal muscle mitochondrial capacity and β-oxidation are reduced. Specifically, upregulation of PPARα and its downstream targets in response to high-fat feeding are defective (142).
Ectopic fat deposition is related to metabolic abnormalities and defects in insulin sensitivity, T2DM, cardiovascular disease, and cancer (143). Finally, obesity is associated with a state of chronic low-grade inflammation because ectopic fat depots release more inflammatory mediators than peripheral fat depots and infiltration of macrophages (24). Metabolic inflexibility and fat deposition therefore likely reinforce one another in a vicious cycle.
Metabolic syndrome and T2DM
The best example of compromised metabolic flexibility in metabolic syndrome is a deteriorated insulin-mediated substrate switching. As such, metabolic inflexibility is at the core of the pathophysiology of insulin resistance (146). After a high-fat meal, patients with metabolic syndrome have higher levels of glycaemia and lower skeletal muscle free fatty acid uptake compared with healthy individuals. In response to fasting, skeletal muscle from patients with insulin resistance are less able to switch to FAO compared with healthy individuals (146). An increased dependency on glucose oxidation and decreased reliance on FAO in offspring from patients with T2DM suggests that impaired FAO may precede insulin resistance (146–148).
Interestingly, studies have demonstrated that BCAA and associated metabolites are strongly associated with insulin resistance and T2DM (153). Based on the theory of mitochondrial metabolic gridlock and anaplerosis, excessive BCAA metabolites are proposed to clog the β-oxidation machinery, particularly in skeletal muscle and liver, and thus contribute to accumulation of incompletely oxidized intermediates of fatty acids, particularly in the presence of a high-fat diet.
Systemic low-grade inflammation
In the case of obesity and insulin resistance, systemic inflammation can trigger and propagate metabolic inflexibility. Systemic inflammation and metabolic inflexibility can cause a vicious circle because metabolic inflexibility can also trigger systemic inflammation. How this is regulated at the cellular and molecular level is currently unknown, but hyperglycemia-induced mitochondrial ROS production can stimulate inflammation by signaling factors (157), such as protein kinase C, p38 MAPK, and c-Jun-N-terminal kinase (158).
Although the mechanism and specific mediators in lipid-induced inflammation are not completely understood, the endoplasmic reticulum (ER) is central to these responses because this is where both lipid biosynthesis and esterification processes as well as inflammatory pathways converge. Disrupted lipid synthesis in the ER can change ER membrane composition, leading to ER stress, dysfunction, and ultimately cell death, triggering inflammation (160).
Metabolic flexibility in immune responses
Metabolic flexibility and the accompanied rerouting of metabolic flux are essential for immune function. Following immune stimulation, naive lymphocytes that rely on β-oxidation of fatty acids and pyruvate oxidation via the TCA cycle become active and engage in glycolysis and glutaminolysis (161).
Memory T cells also use glucose and other fuels to synthesize triglycerides, which are then used in FAO (163).
Contrary to the dogma that innate immunity is nonspecific and lacks memory, classic innate immune cells such as macrophages, natural killer cells, and monocytes can become epigenetically reprogrammed by infection or vaccination, which confers nonspecific protection from secondary infection, a phenomenon called trained immunity (164).
Adipose tissue macrophages that have been activated and rely on glucose are proinflammatory (type M1) and contribute to adipose inflammation and insulin resistance. Conversely, macrophages that rely on fatty acid metabolism secrete anti-inflammatory cytokines and thus preserve insulin sensitivity of liver and adipose tissue (type M2) (167, 168).
An example of metabolic inflexibility and disrupted inflammatory assuagement is sepsis. During sepsis, a profound change in acute leukocyte metabolism occurs. Metabolic inflexibility drives sepsis-related innate immunoparalysis as the metabolism through glycolysis, β-oxidation, and OXPHOS pathways in leukocytes is downregulated, resulting in their inability to mount any response whatsoever (171). A sudden mitochondrial complex I dysfunction in sepsis (172), possibly linked to the overproduction of NO and ROS, may be one of the causes of an upstream mitochondrial gridlock, and has been observed to relate to organ dysfunction (172).
Cardiovascular disease
Cardiac performance is sustained by fatty acid and glucose oxidation, although fatty acids are the preferred substrate in the heart because of the higher energy yield compared with glucose. This flux is mediated by a high expression of PPARα-regulated genes encoding key proteins in fatty acid uptake, esterification, and oxidation (175). Under energetically demanding conditions such as exercise, the heart switches to the oxidation of glucose and lactate (176). An increase in heart rate increases mitochondrial calcium concentration (177), allowing higher mitochondrial ATP production rates to sustain the increased energetic load of the heart. Upon exercise-induced sympathetic nervous system stimulation, β-adrenergic signaling increases glycolytic flux via cAMP activation of cAMP-dependent protein kinase A, increasing pyruvate production and glucose metabolism. Protein kinase A also activates phosphofructokinase-1 and PDH, stimulating the heart to rapidly oxidize glucose even in the presence of fatty acids (178). As a consequence, triglyceride accumulation in cardiomyocytes likely leads to abnormal lipid signaling, increased ROS production, ER stress, and mitochondrial dysfunction (179). Glucose metabolism is enhanced in a similar manner through insulin and nutrient stress signaling via Akt and AMPK, respectively (178).
Obesity can cause metabolic inflexibility of the heart and alter substrate selection (179).
Cancer
Metabolic inflexibility is correlated to an increased risk of certain types of cancers (183) and metabolic syndrome is associated with a 33% elevated total cancer mortality (184).
Diet composition is also correlated to development of certain cancers [reviewed in Potter et al.187)]. High-fat diets for instance have particularly been related to increased risk of colorectal (188), pancreatic (189), breast (190), lung (191), and prostate cancer (192).
Tumor growth and metabolic flexibility
Our understanding of cancer metabolism has rapidly advanced in recent years. Most cancer cells show a remarkable metabolic flexibility, which allows a survival advantage in the face of their energetic demand and the environmental supply of nutrients. Mitochondrial-mediated flexibility is central in this process [reviewed in Vyas et al. (199)].
Most tumors display one, if not several, cancer associated metabolic hallmarks: (1) deregulated uptake of glucose and amino acids, (2) use of opportunistic modes of nutrient acquisition, (3) use of glycolysis/TCA cycle intermediates for biosynthesis and reduced nicotinamide adenine dinucleotide phosphate (NADPH) production, (4) increased demand for nitrogen, (5) alterations in metabolite-driven gene regulation, and (6) specialized metabolic interactions with the microenvironment (206). As a prominent feature of cell activation and proliferation, tumor cells chiefly require increased amounts of glucose and glutamine to survive (207).
The metabolic reprogramming that underlies increased glucose consumption for use in glycolysis, as opposed to OXPHOS, is known as the Warburg effect. In 1924, Otto Warburg discovered that cancer cells metabolize glucose differently than cells of normal tissues: that even in conditions of sufficient oxygen availability, cancer cells convert glucose into lactate instead of using glucose for OXPHOS (208). Warburg hypothesized that cancer cells have mitochondrial defects and impaired aerobic respiration that forces them to rely on glycolysis. Today, we understand that mitochondrial respiration is not impaired but that cancer cells place emphasis on acquisition and generation of building blocks necessary for cell division. They do so by enhancing biosynthetic metabolism using glycolytic intermediates (209).
(…)
Referência:
Endocr Rev. 2018 Aug 1;39(4):489-517.
