Brain Fuel Utilization in the Developing Brain

Key Messages

•  The brain consumes up to 60% of the total energy available to
the body during development.

•  While glucose is the main source of energy for the brain in
adults, ketone bodies are essential to complement glucose to fulfil the metabolic and energy needs of the brain during its development.

•  During brain development, glucose and ketone bodies are not
only the main sources of energy but are also utilized for the biosynthesis of macro-molecules indispensable for neuronal cell proliferation, synapse formation, and myelination.

Keywords

Brain development · Brain metabolism  · Neurogenesis · Synapse · Myelin · Glucose · Ketone bodies · Aerobic glycolysis · Oxidative phosphorylation · Metabolism  · Energetic and anabolic needs

Abstract

During pregnancy and infancy, the hum an brain is growing extremely fast; the brain volume increases significantly, reaching 36, 72, and 83% of the volume of adults at 2– 4 weeks, 1 year, and 2 years of age, respectively, which is essential to establish the neuronal networks and capacity for the development of cognitive, motor, social, and emotional skills that will be continually refined throughout childhood and adulthood. Such dramatic changes in brain structure and function are associated with very large energetic demands exceeding by far those of other organs of the body. It has
been estimated that during childhood the brain may account for up to 60% of the body basal energetic requirements. While the main source of energy for the adult brain is glucose, it appears that it is not sufficient to sustain the dramatic metabolic demands of the brain during its development. Recently, it has been proposed that this energetic challenge is solved by the ability of the brain to use ketone bodies (KBs), produced from  fatty acid oxidation, as a complement source of energy. Here, we first describe the main cellular and physiological processes that drive brain development along time and how different brain metabolic pathways are engaged to support them . It has been assumed that the majority of energetic substrates are used to support neuronal activity and signal transmission. We discuss how glucose and KBs are metabolized to provide the carbon backbones used to synthesize lipids, nucleic acid, and cholesterol, which are indispensable building blocks of neuronal cell proliferation and are also used to establish and refine brain connectivity through synapse formation/elimination and myelination. We conclude that glucose and KBs are not only important to support the energy needs of the brain under development, but they are also essential substrates for the biosynthesis of macro-molecules underlying structural brain growth and reorganization. We emphasize that glucose and fatty acids supporting the production of KBs are provided in complex food matrices, such as breast milk, and understanding how their availability impacts the brain will be key to promote adequate nutrition to support brain metabolism  and, therefore, optimal brain development.

Introduction

The ability of hum ans to perform  complex mental activities, including thinking, reasoning, remembering, problem - solving, decision- making, and learning new information, depends on the ability of the brain to adapt to its environment and alter
its functional and structural organization [ 1– 4] . This is often referred to as brain, neuron, or synapse plasticity [ 5] . More- over, the brain organization is incredibly complex: it is estimated that the human brain contains more than 200 billion neurons and non- neuron cells, 1 quadrillion of connections, 100 km  of nerve fibers, and 600 km  of blood vessels [ 6, 7] . In order to maintain such dynamic abilities and sustain the functioning of this complex architecture, outstanding energy sup- ply is required. Indeed, the adult brain, accounting for a m ere 2% of body weight, is estimated to be responsible for 20% of oxygen (O  ) consumption and 20– 25% of glucose utilization [ 8, 9] . In comparison, adult vertebrate brains, with the exception of primates, use 2– 8% of total energy at resting state [ 10] .

While adult brain energy demands are astonishing, the energy requisite during early life is even higher and essential to sup- port the rapid development of the brain with a growth burst starting around the 5th gestational month and continuing postnatally, increasing the brain’s weight from  ∼ 27% of its adult weight at birth to ∼ 80% by age 2 years [ 11, 12] . In addition to its enormous demand of energy, the dramatic brain size expansion that happens during the first years of life re- quires specific nutrients, such as lipids, proteins, and micro- nutrients, which are not only the building blocks of brain structures but also support brain and cognitive functions during the rest of the lifespan [ 3, 13, 14].

In normal conditions, the main source of energy for the brain is glucose that is utilized for the generation of energy in the form  of adenosine triphosphate (ATP) from  either glycolysis or oxidative phosphorylation, the latter being 15 times m ore efficient to generate energy [ 15– 17] . Nevertheless, the particularly high energy needs of the developing hum an brain seem  not to be supported by the sole consumption of glucose. Indeed, it has recently been suggested that ketones (β- hydroxybutyrate, acetoacetate, and acetone) derived from the oxidation of newborn body fat by the liver, may provide an important additional fuel substrate for the developing brain through oxidative phosphorylation as well [ 2, 17].

The adult brain, accounting for a mere 2% of body weight, is estimated to be responsible for 20% of oxygen (O2) consumption and 20– 25% of glucose utilization.

While glucose and ketones have been traditionally considered for their canonical role in mamalian energy metabolism , recent studies showed that they play additional roles associated with brain structure development and function  [12, 17– 20] . For example, in an adult brain, 10– 12% of glucose is metabolized through glycolysis to produce lactate, despite oxygen being available for oxidative phosphorylation, a phenomenon called “aerobic” glycolysis or the “Warburg effect”  [21] . Aerobic glycolysis remains a prevalent metabolic pathway in the brain all along the lifespan and especially dur- ing brain development. It is especially crucial for the biosynthesis of cell constituents (e.g., lipids) that support key developmental processes, including synapse formation/elimination and myelination  [19, 22] . On the other hand, ketones seem  to be the substrate for the synthesis of certain macro- molecules, such as cholesterol and fatty acids, which represent around 50% of the gray matter of the brain  [17] . Recent observations also underline the importance of ketones as key signalling and epigenetic mediators, suggesting that they m ay influence gene expression involved in brain plasticity and re- organization during brain development  [23– 25] . Therefore, glucose and other fuel substrates including ketone bodies (KBs) m ay be used for other purposes than to fulfil the energy demands only and m ay play a broader role during brain development.

In this chapter, we first describe the key physiological processes underlying brain development and how brain metabolism  may be necessary to support them . We will especially focus on describing how glucose and KBs support not only the energetic but also anabolic demands of the brain during development. Of course, the brain relies on m any other nutrients for its proper development, function, and maintenance, which will be discussed by a subsequent chapter in this volume and has been previously discussed in excellent re- views [ 3, 13, 17, 26] .


Fig. 1. Timelines of neurodevelopment processes and glucose metabolism changes from conception to adulthood. The figure represents the key neurodevelopmental processes that are occurring during brain development and the changes in glucose metabolism associated with them. The top of the figure represents the major periods of human development expressed in days (d), postconceptional weeks (pcw), and years (y). The bars associated with each neurodevelopmental cellular process represent the approximative peak of the developmental period for each of them. The bottom of the figure represents the changes in glucose uptake (black line) and oxygen consumption (gray line) along time that peak around postnatal day 5. The glucose uptake is higher than the oxygen consumption, suggesting that a significant amount of glucose is metabolized via aerobic glycolysis (arrow) and parallels the rise in synapse formation and myelination. The figure has been adapted from [3, 12, 43] .

Main Physiological and Cellular Processes Underpinning Brain Development

The brain is one of the organs of the hum an body to develop the earliest, starting in utero during the 3rd gestational week, and it completes its development during the second and third decades of life, therefore being the organ with the longest development and maturation time [ 7, 22] . Once the primary organization of the brain is achieved, consisting of defining its main different regions during the embryonic period, key cellular processes emerge and proceed in developmentally overlapping waves (F ig. 1) . The generation of neurons (neuro- genesis) and their migration are initiated during the 8th gestational week, and the repertoire of neurons found in the adult neocortex is largely established before birth. Glia cell proliferation that follows neurogenesis peaks around birth, and it includes the generation of oligodendrocytes, supporting myelination and astrocytes, which have been shown to be involved in m any physiological processes in the brain, especially in the modulation of information processing, synaptic transmission, and energy dynamics [ 27– 29] . While regional variation exists, proliferation, migration, and differentiation of oligodendrocytes and astrocytes continue throughout the first 3 postnatal years, which coincides with the peak of synapse formation and neural network reorganization [ 7] . This reinforces the idea that oligodendrocytes and astrocytes play a crucial role in brain connectivity development and maturation all along childhood and adolescence.

While astrocytes and oligodendrocytes are generated and differentiated from neural progenitor cells, microglia, which are the resident macrophages of the brain involved in innate immunity, neuroprotection, synaptic pruning, and phagocytosis of cellular debris, originate from  macrophages present in the yolk sac and migrate and colonize the brain during gestational week 4.5 [ 30, 31] . Interestingly, while neural progenitors are actively dividing and generate the first neurons early during embryogenesis, astrocytes and oligodendrocytes will appear only at late embryonic time points, as mentioned earlier. Con- sequently, early microglia colonization not only precedes the peak of neurogenesis and neuronal migration but constitutes the main glial population during a large part of fetal life, suggesting that microglia are involved in early brain development. Indeed, recent studies demonstrate that microglia contribute to the regulation of neuronal numbers and migrations and actively contribute to activity- dependent synaptic reorganization during neural network establishment [ 32] .

Development of neuron arborization consisting of axon and dendrite outgrowth followed by synapse formation is the key cellular process associated with the functional maturation of the brain after neuron migration. Indeed, from  mid- gestation until the third postnatal year, immature neurons are initiating a protracted period of axon outgrowth and dendrite arborization, accompanied by the formation of synaptic junctions that ensure connectivity between neurons and lead to the formation of neuronal networks [ 33, 34] . Importantly, an overproduction of synapses will take place during the first 2 postnatal years with a burst of synapse growth between 3 and 15 months, depending on the brain areas, followed by a discrete period of synaptic pruning that typically starts during childhood and persists towards adolescence [ 35– 37] , although the visual area has been reported to undergo pruning as early as 3 months of age.

While neuronal networks are being built, oligodendrocytes generated myelin sheets are wrapped around axons, which act as insulators and lead to a dram atic increase in axonal conduction velocity and, therefore, information transmission [ 29] . Myelination starts during m id- gestation in the hum an brain, is a long process that dramatically accelerates during the first 2 postnatal years, and reaches its full maturity during the second to third decade of life [ 29, 38] . It also plays a key role in the maturation of brain networks, coordinated information processing, and ultimately cognitive performance in infants, children, and adults.

The high energy requirement of the brain is fulfilled by a constant transport of nutrients into the brain through the blood- brain barrier (BBB). The BBB tightly controls the pas- sage of selected substances in and out of the brain, provides protection against external potentially toxic agents, and is critical to maintain brain homeostasis and, thus, proper brain function [ 39, 40] . The BBB includes 3 m ajor cellular com po- nents: endothelial cells constituting the wall of blood vessels, pericytes that stabilize the BBB and are critical to maintain its integrity, and astrocytes that extend cellular processes whose endfeet ensheath the blood vessels and play a vital role in the transport of nutrients into neuronal cells [ 39] . The development and differentiation of the BBB is supposed to start in the very young embryo [ 41, 42] . Formation of blood vessels by endothelial cells is quickly accompanied by the recruitment of pericytes and astrocytes that will “seal” the BBB in order to isolate the brain from  the external environment and control transport of substances. At the tim e of birth, the pattern of brain vasculature is very similar to what it will become in the adult brain [ 40] .

The high energy requirement of the brain is fulfilled by a constant transport of nutrients into the brain through the blood-brain barrier.

Recently, significant efforts have been made to under- stand how genes orchestrate the physiological and cellular processes described above. Gene expression analysis revealed that the brain transcriptome is segregated in clusters that are spatially and temporally organized and parallel the structural and functional developmental aspects of the brain. For example, clusters of genes associated with neuronal fate specification are mainly expressed embryonically and early fetally, while genetic clusters controlling synapse formation and function are highly expressed during early childhood  [7] . Interestingly, it has also been shown that gene expression associated with mitochondria closely follows synapse density, suggesting that the proper development and maturation of brain connectivity is highly linked to energy availability  [12, 43] .

In summary, the hum an brain undergoes a rapid growth from  the 4th gestational week to the 3rd postnatal year. Subsequently, the rate of growth slows down and the brain is subjected to a significant reorganization, which is dominated by synaptic pruning and myelination that extends throughout the third decade of life. It is important to keep in m ind that brain development is not structurally and functionally homogeneous with age. Indeed, associative regions of the neocortex and especially prefrontal cortex are slower to mature than motor and sensory cortices, for example. Therefore, while being less important than during the prenatal or early postnatal periods, brain growth and especially brain reorganization at cellular and molecular levels continue beyond childhood to early adulthood. Refinement of neuronal networks is indeed thought to be critical for the functional specification of brain regions and crucial for the development of higher cognitive functions and behavior.

Energetic and Anabolic Demands during Human Brain Development

As briefly mentioned earlier, energy demands for vertebrate species correspond to 2– 8% of the total energy provided by basal metabolism , while the hum an adult brain requires as much as 20– 25% of it  [10] . The energy demand during brain development is even m ore striking; it has been estimated that the newborn hum an brain, which represents about 13% of lean body weight, is consuming around 60% of the body’s daily requirement [ 12, 15, 17, 19, 20, 44] . This dramatic energetic demand persists and is even increased during child- hood; while a child brain at age 10 years accounts for 5– 10% of the body m ass, it approximately consumes 50% of the total basal metabolic rate of the hum an body [ 3, 12] .

Why are the energetic costs associated with brain function so high in humans and especially during brain development? In order to understand this, it is first necessary to identify the brain components and processes that cost energy. From an evolutionary point of view, a comparison of glucose and oxygen metabolic rates of the adult brain in awake mammals (rodents, macaque, baboon) suggests that the total metabolic cost is a simple linear function of the number of neurons present in the brain  [45] . This is in agreement with a recent approximation of neural cellular energy demands which estimates that neurons consume 75– 80% of the energy produced, whereas the rest is used for glia- based processes  [15, 46, 47] . Two main reasons may explain why neurons have high energetic demands: first, the generation of action potentials along the axons and synaptic transmission from  neuron to neuron are based on electrochemical and cellular processes, such as ion fluxes, neurotransmitter release and re-uptake, and vesicle cycling, which are energetically costly  [15,44, 46] . A signalling mechanism  at the synapse has been suggested to be especially energy consuming; for example, it has been estimated that 80% of the energy in m hippo- cam pal axons is expended by postsynaptic potentials  [48] . Second, the ability of the brain to change and adapt continuously along the lifespan is due to the constant remodeling of its architecture that culminates by the addition or the elimination of synapses to strengthen or weaken neuronal net- work activities accordingly. Constant synthesis of proteins, lipids, and amino acids is necessary to support the molecular modifications that underlie brain plasticity, which contributes to increasing brain energy expenditure  [19, 46, 49, 50] . Rapid turnover of proteins and lipids is crucial to support dendritic spines and synapse modification, which are essential for learning and memory processes  [51] . Nevertheless, it has been widely recognized that the majority of the energy in the adult brain is used to maintain its physiological baseline activity, including neuron and synaptic resting membrane potential, while changes in brain activity required to sustain specific cognitive tasks linked to synaptic plasticity result in an increase of energy demands by only 5%  [18, 46, 52] . Importantly, brain development has significant additional energetic needs that are essential to support the constant and sustained synthesis of the molecular building blocks (proteins, lipids, and nucleic acids) underlying the rapid development and maturation of neuronal networks. In particular, at birth, the brain is about 25% of the weight of an adult brain, by age 2 years it is about 75% of its adult size, and around the age of 7 years, the hum an brain has reached its maximal size. Therefore, postnatal growth, especially during early childhood, happens rapidly and is not the result of the addition of new neurons, since neurogenesis mainly happens prenatally. In- stead, it is the development and the maturation of neurons already present at birth that account for the increase in brain biomass and energy demands, including axon growth, dendritic arborization elaboration, synaptic formation/elimination, and axon myelination  [7, 18, 36, 37] . Interestingly, brain energy metabolism  requirement follows the development and maturation of the brain, reaching its peak in energy demand per gram  of tissue during postnatal year 2 and 3, especially when the rates of synapse formation and myelination are reaching their maximal intensity  [3, 53] . Indeed, metabolic and especially anabolic demands are expected to in- crease with the addition of new synaptic connections and myelin wraps around axons. Therefore, there is a very close temporal and spatial relationship between the brain metabolic and anabolic needs and the cellular and physiological changes of the neural tissue through development and adulthood. 


Fig. 2. Key biochemical pathway involved in glucose metabolism. Blood glucose is crossing the BBB in order to enter the brain. Glucose enters cells through glucose transporters (GLUTs) and is immediately phosphorylated to generate glucose- 6- phosphate (glucose- 6P). Glucose- 6P is used as the metabolic substrate for different biochemical pathways. First, it is converted into 2 molecules of pyruvate through glycolysis that generate ATP and NADH. Pyruvate is then either reduced in lactate, consuming one molecule of NADH, or is metabolized in acetyl- CoA. Lactate can be released in the extracellular space through monocarboxylate transporters and is used as a source of energy or as a biosynthetic substrate by neurons and oligodendrocytes. Second, acetyl- CoA is metabolized through the tricarboxylic acid (TCA) cycle and oxidative phosphorylation that produce ATP and CO 2 while consuming oxygen. The complete oxidation of glucose produces larger amounts of energy in the form of ATP in the mitochondria (30– 36 ATPs) compared to glycolysis (2 ATPs). Alternatively, acetyl- CoA is used for the synthesis of fatty acid and amino acids. Glucose- 6P can also be metabolized along the pentose phosphate pathway (PPP) and leads to the generation of ribulose- 5- phosphate (ribulose- 5- P, use in the synthesis of nucleic acid) and NADPH. NADPH is important to support fatty acid synthesis but also for the regulation of glutathione metabolism. Glutathione exists in the reduced form (GSH) or as a disulfide from (GSSG). The reduced form GSH is a source of reducing equivalent that can neutralize reactive oxygen species (ROS), such as hydroxyperoxides (ROOH). GSH is converted into GSSG, which is then recycled back to GSH by using NADPH as an electron donor. Glutathione is, therefore, a key antioxidant that protects cells against oxidative stress and is also critically involved in the control of cell redox homeostasis. Ultimately, nucleic acid, fatty acid, amino acid synthesis, and the control of redox homeostasis are providing the necessary energy and source of macromolecules that support neurodevelopment processes in brain development. The figure has been modified from [15] .

Glucose Requirements to Support Energy and Anabolic Demands during Brain Development

Measurements of cerebral metabolic rate for glucose (CMRGlc) and for oxygen (CMRO  ), which is a measure of glucose and oxygen utilization in the brain, show a constant in- crease in their values during the first 2 years of life, reach approximately 2 and 1.5 times the adult value around 3– 5 years of life, respectively, and then gradually decrease to the aver- age adult value during the second decade of life (F ig. 1)  [ 3, 54].

These observations suggest that energy demand increases during brain development, presumably due to an increase in synaptic transmission. Intriguingly, glucose utilization is in- creased to a greater extent com pared to brain oxygen utilization. Indeed, at birth, CMRGlc and CMRO  measurement showed that the glucose consumption rate exceeds oxidative phosphorylation by around 34% [ 54, 55] . Moreover, Goyal et al. [ 20] reported that oxygen utilization during childhood ac- counts for approximately 70% of the total glucose consumption in a child brain. Since oxygen in the brain is utilized almost entirely for the oxidation of carbohydrate through oxidative phosphorylation to generate ATP, these results suggest that glucose m ay play additional functions to being an energetic substrate (F ig. 2)  [ 3, 18, 56] . The preference of converting the glucose metabolite pyruvate produced through glycolysis into lactate or using it as a source of carbon for biosynthetic processes instead of converting it to ATP despite the availability of oxygen has been named aerobic glycolysis or the Warburg effect [ 21] . Aerobic glycolysis has been well described in tum or tissues that metabolize approximately 10- fold m ore glucose to lactate than norm al tissues, to provide substrates for biosynthesis of cell constituents and support cancer cell proliferation [ 57] . Nevertheless, since neurogenesis at birth is limited and restricted to specific brain areas, metabolic and anabolic demands supported by glucose are presumably due to the maturation of preexisting neurons, refinement of synaptic connectivity, glia proliferation, and rapid rise in axon myelination.

Glycolytic byproducts are a crucial source of carbons to produce glutathione, NADPH, and riboses along the pentose phosphate pathway (PPP), which are them selves essential for the synthesis of fatty acids and nucleotide sensitive, respectively, and to maintain oxidative stress homeostasis (F ig. 2) [ 58] . Biosynthesis of macro-molecules from  glucose metabolites is critical to support key physiological processes behind
proper brain growth and maturation; it has been shown, for exam ple, that axon growth, synapse formation, and myelination rely critically on aerobic glycolysis [ 20, 59, 60] . Interestingly, aerobic glycolysis is predominant in the white matter com pared to the gray matter, and it has been shown that glycolytic byproducts, such as lactate, are especially important for myelin production by oligodendrocytes [ 61, 62] . While it has been assumed that most of the glucose is used for ion pumping to maintain synaptic activity, these findings highlight that glucose is critically involved in anabolic requirem ents be- yond energetic dem ands during neurodevelopm ent [ 18, 63, 64].

As discussed, aerobic glycolysis varies through the lifespan depending on regional and temporal metabolic and anabolic demands and seem s to be critical as well during early fetal brain development; measurement of glucose uptake in 12-  to 21- week previable human fetuses demonstrated that around one- third of the total body glucose was consumed by the brain and only half of it was presumably oxidized [ 65] . Studies perform ed in preterm  infants, when neurogenesis is still active, demonstrated a very low rate of oxygen consumption and suggested that 90% of glucose is dedicated to aerobic glycolysis [ 66, 67] . While studies during early stages of brain development are limited, these data indicate that the fetal brain is highly dependent on aerobic glycolysis, possibly due to the large requirement of denovo biosynthesis of lipids, amino and nucleic acids that are associated with neuron generation and proliferation (Fig. 2) .

The peak of CMRGlc happens around postnatal year 5 and stays elevated until 10 years of age [ 3, 54, 55] . While aging, glucose consumption slightly declines before reaching its adult life level during the second decade of life, and aerobic glycolysis decreases to one- third of its value in adulthood, representing 8– 10% of glucose utilization [ 55] . Nevertheless, aerobic glycolysis in some areas of the brain, such as the m e- dial and lateral parietal and prefrontal cortices, can contribute to as much as 20– 25% of glucose utilization [ 68, 69] . These brain areas integrate multi-modal sensory information and participate in complex cognitive functions, such as executive function and self- awareness, that necessitate a high level of synaptic plasticity and, therefore, a significantly high biomoecular turnover. It is, therefore, possible that during brain development glucose utilization is the key to not only provide energy through oxidative phosphorylation but also to support increase in biomass and macr-omolecule biosynthesis through aerobic glycolysis. Importantly, aerobic glycolysis seem s to be especially crucial prenatally to support neurogenesis and then postnatally to support mainly neuronal growth, synapse formation, and myelination and eventually the proper develop- m ent of neuronal networks underlying cognitive function. During brain m aturation, a gradual m etabolic switch between aerobic glycolysis and oxidative phosphorylation is happening. As neuronal networks mature, oxidative phosphorylation is predominant and maintains basal synaptic activity, while aerobic glycolysis is prevalent in areas where brain plasticity is engaged to sustain experience- dependent structural and functional changes that accompany higher cognitive functions (Fig. 2 ).

While aging, glucose consumption slightly declines before reaching its adult life level during the second decade of life, and aerobic glycolysis decreases to one- third of its value in adulthood.

Ketones as a Source of Energy for the Brain during Development

Glucose is the primary metabolic substrate used by the brain to generate ATP in the central nervous system  of adult mammals [ 8] . However, during brain development and maturation, the demand of energy and the high rate of macro-molecular biosynthesis exceed the availability of blood glucose. The most relevant additional fuels to support extra energetic needs of brain development are KBs, especially
β- hydroxybutyrate and acetoacetate [ 17, 70– 72] . KBs are short- chain fatty acids (SCFAs) derived mainly from  liver β- oxidation of fatty acids that are available for the brain in direct proportion to their concentration in the blood [ 17, 73] . KBs are then oxidized through oxidative phosphorylation in the mitochondria of neuronal cells to generate ATP [ 17] . While the use of KBs for brain development is starting in utero, post- natal brain development also highly depends on KBs [ 65] . An interruption of continuous transplacental nutrient and energy supplies at birth necessitates the newborn to adapt quickly to the new metabolic environment and, especially, to move from continuous feeding to alternate periods of feeding and fasting [71] . This leads to a rapid metabolization of available substrates to produce energy, first, acutely from  newborn reserve and, second, from alimentation in the form  of milk from  the m other [71, 74] . At birth, brain glucose reserves are very limited and could support brain needs for a few hours only. Metabolization and utilization of glucose from  other organs, such as muscle, or tissue breakdown are not a viable long- term  solution to support growth development. Interestingly, at birth, newborns are in a state of perm anent mild ketosis (0.1– 0.5 m M β- hydroxybutyrate), which is independent of feeding status or hypoglycem ia [ 75] . Moreover, the brain up- take of KBs is 4– 5 times faster in infants and children than in adults, which means that infant and child metabolism  is pro- gram m ed to actively produce KBs from  liver β- oxidation and that the brain is dependent on KBs to support its metabolic and anabolic needs. Indeed, it has been estimated that KBs may be able to replace for up to two- third of brain energy demands when glucose availability is low [ 17] .

It has now been recognized that fatty acids, especially medium - chain fatty acids (MCFAs) that constitute up to 10–20% of fatty acids contained in breast m ilk, are one of the m ain substrates used to produce SCFAs and maintain sustained ketosis in infants  [72] . MCFAs are either directly converted into KBs by β- oxidation in the liver that will be taken up by the brain or they can be stored in adipose tissues and can be used later to support energy demands during a fasting period ( Fig. 3,  4  ). Moreover, it is estimated that hum an m ilk contains around 15– 17% SCFAs, which are highly ketogenic and may support brain energy and anabolic needs immediately  [72, 76, 77].

Interestingly, body fat deposition during development is unique in hum ans; the hum an fetus starts to accumulate fat in subcutaneous adipose tissues during m id- gestation and has 500– 600 g of subcutaneous fat at birth, while most mammals have a very limited amount of adipose tissues and, therefore, inability to store either MCFAs or SCFAs [ 17].

What are the main advantages for the brain to use KBs during development? First, SCFAs are either immediately avail- able through the milk, or MCFAs stored within subcutaneous adipose tissues can be mobilized and metabolized to KBs. Second, the use of KBs as an alternative source of energy preserves glucose utilization for additional key metabolic path- ways, such as the PPP, as described earlier. Third, KBs are not only a high energetic substrate but are also used as anabolic metabolites. For example, cholesterol is the main carbon source for the synthesis of cholesterol, which represents 20% of total brain lipids. Cholesterol is not only indispensable to properly build the cell membrane, but it is also crucial to build axon  yelination [ 78, 79].

Finally, the generation of KBs from MCFAs and SCFAs is very fast com pared to other fatty acids, since they directly reach the liver through the portal vein, by- passing the lym phatic system , and are β- oxidized into the m i-tochondria without the usual activation of the enzym e carni- tine palm itoyltransferase (Fig. 3).

Therefore, contrary to the adult brain, it appears that KBs are essential for brain development and maturation as they are not only an essential source of energy to complement glucose to entirely fulfil the brain metabolic needs but also, sim ilar to glucose, to support the anabolic demands associated with cell proliferation, growth, and maturation.

Conclusion

From  conception to the third year of life, brain size increases dram atically, leading to the formation and expansion of neuronal networks that will eventually be reorganized and re- shaped according to a variety of genetic and environmental factors [ 3] . Amongst the latter, nutrition is important for optimal brain development, since it provides glucose, KBs, and ketogenic fatty acids, which are the main metabolic substrates of the brain, during its development and maturation. Aerobic glycolysis and biosynthesis of macro-molecules from  glucose seem s to be particularly important to support the establishment and maintenance of synaptic plasticity associated with higher cognitive functions. Utilization of KBs may, therefore, be used to “free” glucose for aerobic glycolysis, while sup- porting energy demands for synaptic transmission. In addition, KBs are also necessary for the synthesis of specific macro-molecules, such as cholesterol, as discussed earlier. Nevertheless, m any questions remain: first, it is still unknown what are the intrinsic and extrinsic factors that trigger aerobic glycolysis and oxidative phosphorylation in the brain, depending on its stage of development; second, it is not well understood how the balance between the use of glucose and KBs as energetic versus anabolic substrates is regulated. It will, there- fore, be crucial to elucidate the genetic, metabolic, and physiological processes during brain development that dictate brain metabolism  changes. Such nutrients are normally provided in complex food matrices, such as breast m ilk. Under- standing fully how specific nutrients, especially in the context of food intake, interact together and affect brain metabolism (please see the chapter in this issue focusing on essential nutrients for early brain development: “Nutritional Factors in Fetal and Infant Brain Development” by Cheatham ) will help us to better define the minimal requirements to support it and promote brain development.

Disclosure Statement

P.S. is an employee of Société des Produits Nestlé SA and the writing of this article was supported by Nestlé Nutrition Institute. The author declares no other conflicts of interest.


Fig. 3. Comparison of the absorption of medium- chain fatty acids (MCFAs) and long- chain fatty acid (LCFAs). MCFAs are rapidly absorbed from the gut and directly reach the liver through the portal vein. LCFAs are first integrated into chylomicrons and are primarily absorbed through the lymphatic system before reaching the target organs, including the liver, from the peripheral circulation. In the liver, MCFAs and LCFAs are converted into ketone bodies (KBs) (see Fig. 4) that are then released in the peripheral circulation before reaching the brain through the BBB. KBs are then converted to acetyl- CoA that is metabolized to produce either cholesterol in the smooth endoplasmicreticulum or energy in the form of ATP through the TCA cycle. The figure has been modified from [75] .


Fig. 4. Ketone body synthesis. Fatty acids are converted to acetyl- CoA through β- oxidation. Acetyl- CoA can then either enter the TCA cycle to generate ATP, or 2 molecules of acetyl- CoA are condensed in acetoacetyl- CoA. Acetoacetyl- CoA is then used to produce acetoacetate that can be used to produce either acetone or D- β-
hydroxybutyrate. The figure has been adapted from [17] .

References

1 Knickmeyer RC, Gouttard S, Kang C, Evans D, Wilber K, Sm ith JK,
et al. A structural MRI study of hum an brain development from birth to 2 years. J  Neurosci.  2008 Nov;  28(47):  12176– 82.
2 Cunnane SC, Harbige LS, Crawford MA. The importance of energy and nutrient supply in hum an brain evolution.  Nutr Health . 1993;
9(3):  219– 35.
3 Goyal MS, Iannotti LL, Raichle ME. Brain Nutrition: A Life Span Approach. Annu Rev Nutr.  2018 Aug;  38(1):  381– 99.
4 Laughlin SB, Sejnowski TJ. Com m unication in neuronal networks. Science.  2003 Sep;  301(5641):  1870– 4.
5 Pascual- Leone A, Am edi A, Fregni F, Merabet LB. The plastic hu-
m an brain cortex. Annu Rev Neurosci.  2005;  28(1):  377– 401.
6 Wong AD, Ye M, Levy AF, Rothstein JD, Bergles DE, Searson PC. The blood- brain barrier: an engineering perspective. Front Neuroeng.  2013 Aug;   6:  7.
7 Silbereis JC, Pochareddy S, Zhu Y, Li M, Sestan N. The Cellular and
Molecular Landscapes of the Developing Hum an Central Nervous System . Neuron . 2016 Jan;  89(2):   248– 68.
8 Sokoloff L. Energetics of functional activation in neural tissues. Neurochem  Res.  1999 Feb;  24(2):   321– 9.
9 Raichle ME, Mintun MA. Brain work and brain imaging. Annu Rev Neurosci.  2006;   29(1):  449– 76.
10 Mink JW, Blumenschine RJ, Adam s DB. Ratio of central nervous system  to body metabolism  in vertebrates: its constancy and functional basis. A m  J Physiol.  1981 Sep;  241(3):R203– 12.
11 Dobbing J, Sands J. Quantitative growth and development of human brain. Arch Dis Child.  1973 Oct;  48(10):  757– 67.
12 Goyal MS, Venkatesh S, Milbrandt J, Gordon JI, Raichle ME. Feeding the brain and nurturing the mind: linking nutrition and the gut microbiota to brain development. Proc Natl Acad Sci USA.  2015 Nov;  112(46):  14105– 12.
13 Mattei D, Pietrobelli A. Micronutrients and Brain Development. Curr Nutr Rep.  2019 Jun;  8(2):  99– 107.
14 Zheng L, Fleith M, Giuffrida F, O’Neill BV, Schneider N. Dietary Polar Lipids and Cognitive Development: A Narrative Review. A dv Nutr . 2019 Nov;  10(6):  1163– 76.
15 Magistretti PJ, Allaman I. A cellular perspective on brain energy metabolism  and functional imaging. Neuron . 2015 May;  86(4):883– 901.
16 Cunnane S, Nugent S, Roy M, Courchesne- Loyer A, Croteau E, Tremblay S, et al. Brain fuel metabolism , aging, and Alzheimer’s disease. Nutrition.  2011 Jan;   27(1):  3– 20.
17 Cunnane SC, Crawford MA. Energetic and nutritional constraints on infant brain development: implications for brain expansion during hum an evolution. J  Hum  Evol.  2014 Dec;  77:   88– 98.
18 Bauernfeind AL, Babbitt CC. The appropriation of glucose through prim ate neurodevelopment. J  Hum  Evol.  2014 Dec;  77:  132– 40.
19 Bauernfeind AL, Barks SK, Duka T, Grossman LI, Hof PR, Sherwood CC. Aerobic glycolysis in the prim ate brain: reconsidering the implications for growth and maintenance.  Brain Struct Funct.  2014 Jul;  219(4):  1149– 67.
20 Goyal MS, Hawrylycz M, Miller JA, Snyder AZ, Raichle ME. Aerobic glycolysis in the human brain is associated with development and neotenous gene expression. C ell Metab. 2014 Jan;  19(1):  49– 57.
21 Warburg O, Wind F, Negelein E. The Metabolism  of Tumors in the
Body. J  Gen Physiol.  1927 Mar;  8(6):  519– 30.
22 Stiles J, Jernigan TL. The basics of brain development. Neuropsychol Rev.  2010 Dec;  20(4):  327– 48.
23 Puchalska P, Crawford PA. Multi- dimensional roles of ketone bodies in fuel metabolism , signaling, and therapeutics. Cell Metab. 2017 Feb;  25(2):   262– 84.
24 Newman JC, Verdin E. β- hydroxybutyrate: much m ore than a metabolite. Diabetes Res Clin Pract.  2014 Nov;  106(2):  173– 81.
25 Newman JC, Verdin E. Ketone bodies as signaling metabolites.Trends Endocrinol Metab.  2014 Jan;  25(1):   42– 52.
26 Cusick SE, Georgieff MK. The Role of Nutrition in Brain Development: The Golden Opportunity of the “First 1000 Days”. J  Pediatr. 2016 Aug;  175:  16– 21.
27 Elsayed M, Magistretti PJ. A New Outlook on Mental Illnesses: Glial Involvement Beyond the Glue. Front Cell Neurosci.  2015 Dec; 9: 468.
28 Allam an I, Bélanger M, Magistretti PJ. Astrocyte- neuron metabolic relationships: for better and for worse.  Trends Neurosci.  2011 Feb;  34(2):76– 87.
29 Purger D, Gibson EM, Monje M. Myelin plasticity in the central nervous system . N euro-pharmacology.  2016 Nov;  110(Pt B):  563– 573.
30 Thion MS, Ginhoux F, Garel S. Microglia and early brain development: an intimate journey. Science.  2018 Oct;  362(6411):  185– 9.
31 Menassa DA, Gomez- Nicola D. Microglial Dynamics During Human Brain Development. Front Immunol.  2018 May;  9:1014.
32 Lenz KM, Nelson LH. Microglia and Beyond: Innate Immune Cells As Regulators of Brain Development and Behavioral Function. Front Immunol.  2018 Apr;  9:   698.
33 Zecevic N. Synaptogenesis in layer I of the hum an cerebral cortex in the first half of gestation. Cereb Cortex . 1998 Apr- May; 8(3):245– 52.
34 Huttenlocher PR, Dabholkar AS. Regional differences in synapto- genesis in hum an cerebral cortex.  J Com p Neurol.  1997 Oct; 7(2):  167– 78.
35 Huttenlocher J, Levine S, Vevea J. Environmental input and cognitive growth: a study using time- period comparisons. Child Dev. 1998 Aug;  69(4):1012– 29.
36 Rakic P, Bourgeois JP, Eckenhoff MF, Zecevic N, Goldm an- Rakic PS. Concurrent overproduction of synapses in diverse regions of the prim ate cerebral cortex. Science.  1986 Apr;  232(4747):  232– 5.
37 Hensch TK. Critical period regulation. A nnu Rev Neurosci.  2004; 27(1):  549– 79.
38 Bruchhage MM, Ngo GC, Schneider N, D’Sa V, Deoni SC. Functional connectivity correlates of infant and early childhood cognitive development. B rain Struct Funct.  2020 Mar; 225(2): 669– 81.
39 Siegenthaler JA, Sohet F, Danem an R. ‘Sealing off the CNS’: cellular and molecular regulation of blood- brain barriergenesis. Curr Opin Neurobiol.  2013 Dec;  23(6):   1057– 64.
40 Haddad- Tóvolli R, Dragano NR, Ramalho AF, Velloso LA. Development and Function of the Blood- Brain Barrier in the Context of Metabolic Control. Front Neurosci. 2017 Apr; 11: 224.
41 Fossan G, Cavanagh ME, Evans CA, Malinowska DH, Møllgård K,
Reynolds ML, et al. CSF- brain permeability in the immature sheep fetus: a CSF- brain barrier. B rain Res.  1985 Feb;  350(1- 2):  113– 24.
42 Marín- Padilla M, Knopm an DS. Developmental aspects of the intracerebral microvasculature and perivascular spaces: insights into brain response to late- life diseases. J  Neuropathol Exp Neurol.  2011 Dec;  70(12):  1060– 9.
43 Goyal MS, Raichle ME. Gene expression- based modeling of human cortical synaptic density. Proc Natl Acad Sci USA . 2013 Apr;
110(16):  6571– 6.
44 Bordone MP, Salm an MM, Titus HE, Am ini E, Andersen JV, Chakraborti B, et al. The energetic brain -  A review from  students to students. J  Neurochem .  2019 Oct;  151(2):  139– 65.
45 Herculano- Houzel S. Scaling of brain metabolism  with a fixed energy budget per neuron: implications for neuronal activity, plasticity and evolution. P LoS One. 2011 Mar; 6(3):e17514.
46 Harris JJ, Jolivet R, Attwell D. Synaptic energy use and supply. Neuron . 2012 Sep;  75(5): 762– 77.
47 Hyder F, Herm an P, Bailey CJ, Møller A, Globinsky R, Fulbright RK, et al. Uniform  distributions of glucose oxidation and oxygen extraction in gray matter of normal human brain: no evidence of regional differences of aerobic glycolysis. J Cereb Blood Flow Metab . 2016 May; 36(5): 903– 16.
48 Alle H, Roth A, Geiger JR. Energy- efficient action potentials in hippocam pal mossy fibers. Science. 2009 Sep; 325(5946): 1405– 8.
49 Magistretti PJ. Neuron- glia metabolic coupling and plasticity. Exp Physiol.  2011 Apr; 96(4): 407– 10.
50 Sheng M, Kim  E. The postsynaptic organization of synapses. Cold Spring Harb Perspect Biol.  2011 Dec; 3(12): a005678.
51 Holtmaat A, Svoboda K. Experience- dependent structural synaptic plasticity in the mammalian brain. N at Rev Neuroscii. 2009 Sep;10(9): 647– 58.
52 Magistretti PJ, Pellerin L; Astrocytes Couple Synaptic Activity to Glucose Utilization in the Brain. Astrocytes Couple Synaptic Activity to Glucose Utilization in the Brain. News Physiol Sci. 1999 Oct; 14(5):177– 82.
53 Deoni S, Dean D 3rd, Joelson S, O’Regan J, Schneider N. Early nutrition influences developmental myelination and cognition in infants and young children. Neuroim age.  2018 Sep;178: 649– 59.
54 Chugani HT. A critical period of brain development: studies of cerebral glucose utilization with PET. P rev Med.1998 Mar- Apr; 27(2): 4– 8.
55 Chugani HT. Positron emission tomography: principles and applications in pediatrics. Mead Johnson Symp Perinat Dev Med. 87;(25):15– 8.
56 Goyal MS, Raichle ME. Glucose Requirements of the Developing Hum an Brain. J  Pediatr Gastroenterol Nutr. 2018 Jun; 66 Suppl 3:S46– 9.
57 Koppenol WH, Bounds PL, Dang CV. Otto Warburg’s contributions to current concepts of cancer m etabolism . N at Rev Cancer. 2011 May;11(5): 325– 37.
58 Magistretti PJ. Synaptic plasticity and the Warburg effect. C ell Metab. 2014 Jan;19(1):   4– 5.
59 Segarra- Mondejar M, Casellas- Díaz S, Ram iro- Pareta M, Müller- Sánchez C, Martorell- Riera A, Hermelo I, et al. Synaptic activity-induced glycolysis facilitates membrane lipid provision and neurite outgrowth. E MBO J . 2018 May;  37(9):e97368.
60 Della- Flora Nunes G, Mueller L, Silvestri N, Patel MS, Wrabetz L, Feltri ML, et al. Acetyl- CoA production from  pyruvate is not necessary for preservation of myelin. Glia.  2017 Oct; 65(10): 1626– 39.
61 Sánchez- Abarca LI, Tabernero A, Medina JM. Oligodendrocytes use lactate as a source of energy and as a precursor of lipids. Glia .01 Dec; 36(3): 321– 9.
62 Rinholm  JE, Ham ilton NB, Kessaris N, Richardson WD, Bergersen LH, Attwell D. Regulation of oligodendrocyte development and myelination by glucose and lactate. J  Neurosci.  2011 Jan; 31(2): 538– 48.
63 Magistretti PJ, Pellerin L, Rothm an DL, Shulm an RG. Energy ondem and. Science. 1999 Jan; 283(5401): 496– 7.
64 Attwell D, Laughlin SB. An energy budget for signaling in the grey matter of the brain.  J Cereb Blood Flow Metab . 2001 Oct; 21(10):1133– 45.
65 Adam  PA, Räihä N, Rahiala EL, Kekom äki M. Oxidation of glucose
and D- B- OH- butyrate by the early hum an fetal brain. Acta Paediatr Scand. 1975 Jan; 64(1): 17– 24.
66 Altm an DI, Perlm an JM, Volpe JJ, Powers WJ. Cerebral oxygen metabolism  in newborns. Pediatrics. 1993 Jul; 92(1): 99– 104.
67 Yoxall CW, Weindling AM. Measurement of cerebral oxygen consumption in the hum an neonate using near infrared spectroscopy: cerebral oxygen consumption increases with advancing gestational age. Pediatr Res. 1998 Sep; 44(3):  283– 90.
68 Vaishnavi SN, Vlassenko AG, Rundle MM, Snyder AZ, Mintun MA, Raichle ME. Regional aerobic glycolysis in the hum an brain. Proc Natl Acad Sci USA. 2010 Oct;  107(41): 17757– 62.
69 Blazey T, Snyder AZ, Su Y, Goyal MS, Lee JJ, Vlassenko AG, et al. Quantitative positron emission tomography reveals regional differences in aerobic glycolysis within the hum an brain. J  Cereb Blood Flow Metab. 2019 Oct; 39(10):  2096– 102.
70 Achanta LB, Rae CD. β- Hydroxybutyrate in the Brain: One Molecule, Multiple Mechanisms. Neurochem Res.  2017 Jan;  42(1): 35– 49.
71 Nehlig A, Pereira de Vasconcelos A. Glucose and ketone body utilization by the brain of neonatal rats. Prog Neurobiol. 1993 Feb; 40(2):163– 221.
72 Cunnane SC, Courchesne- Loyer A, Vandenberghe C, St- Pierre V, Fortier M, Hennebelle M, et al. Can Ketones Help Rescue Brain Fuel Supply in Later Life? Implications for Cognitive Health during Aging and the Treatment of Alzheimer’s Disease. Front Mol Neurosci. 2016 Jul; 9: 53.
73 Cunnane SC, Schneider JA, Tangney C, Trem blay- Mercier J, For-
tier M, Bennett DA, et al. Plasma and brain fatty acid profiles in mild cognitive impairment and Alzheimer’s disease.  J Alzheimers Dis. 2012; 29(3): 691– 7.
74 Erecinska M, Cherian S, Silver IA. Energy m etabolism  in m am m a- lian brain during development.  Prog Neurobiol.  2004 Aug;  73(6): 7– 445.
75 Cunnane SC, Courchesne- Loyer A, St- Pierre V, Vandenberghe C,
Pierotti T, Fortier M, et al. Can ketones compensate for deteriorating brain glucose uptake during aging? Implications for the risk and treatment of Alzheimer’s disease. Ann N Y Acad Sci.  2016 Mar; 1367(1): 12– 20.
76 Sarda P, Lepage G, Roy CC, Chessex P. Storage of medium - chain triglycerides in adipose tissue of orally fed infants. Am  J Clin Nutr.
1987 Feb; 45(2): 399– 405.
77 Koletzko B, Demmelmair H, Socha P. Nutritional support of infants
and children: supply and metabolism  of lipids. Baillieres Clin Gastroenterol. 1998 Dec;  12(4): 671– 96.
78 Piom elli D, Astarita G, Rapaka R. A neuroscientist’s guide to lipidomics. N at Rev Neurosci.  2007 Oct;  8(10):  743– 54.
79 Koper JW, Lopes- Cardozo M, Van Golde LM. Preferential utilization of ketone bodies for the synthesis of myelin cholesterol in vivo. Biochim  Biophys Acta.  1981 Dec; 666(3): 411– 7.