neuron Glia Brain metabolism. metabolic activity

Transcription

1 BRAIN The structures within the brain are made up of about 100 billion neurons, as well as trillions of support cells called glia. Neurons in the brain are supported and nurtured by glial cells, which greatly outnumber neurons in the brain by five to ten times. The proportion of glia seems to be correlated with an animal's size: the human brain has about 90% glia cells. The defining characteristic of a neuron is its ability to transmit rapid electrical signals in the form of action potentials. Neurons are arranged in networks (circuits), and communicate with each other via specialized intercellular adhesion sites called synapses. This dense, branching network is a real "neuron forest". Neuronal signaling involves the propagation of an action potential down a neuron's axonal process to a presynaptic terminal; the depolarization of the terminal and release of neurotransmitters; binding of the released neurotransmitters to receptors on the postsynaptic membrane of another neuron; and the subsequent depolarization of this second neuron, propagating the signal further. All other neural cells that lack this property are categorized into a broad class termed glia. Glia do not fire action potentials, but instead surround and ensheath neuronal cell bodies, axons and synapses throughout the nervous system. There are a few different types of glia in the brain: oligodendrocytes, astrocytes (both classified as macroglia), and microglia. Each is needed to optimize brain function. The traditional view has been that glia look after neurons and maintain their proper functioning, having a somewhat passive role themselves. Established functions of glia include supporting neurotransmission, maintaining ionic balance in the extracellular space, and insulating axons to speed up electrical communication. However, emerging research suggests that glia, particularly astrocytes, also have an active role in brain function and information processing both during development and in adulthood. Neurons may be the more important cells in the brain that relay messages about what you are thinking, feeling, or doing. Nevertheless, they could not do it without a little help from the glial cells. Neurons and glia are not independent cellular elements of the nervous system but two strongly interconnected components. Intensive research in the past decade has revealed new, important roles for the latter. Work over the past 25 years has rendered yet another dogma about the adult central nervous system (CNS) obsolete. Ramon y Cajal had concluded that CNS structure remains essentially stable once it reaches full development, changing only in response to increasing age and degeneration. Numerous observations have shown that this is not the case. Thus, careful electron microscopic analyses of different neuronal systems described a surprising capacity of neurons and their synaptic connections to undergo significant morphological alterations. The same methodologies have revealed that it is not only neurons that can undergo remodeling but glial cells as well. Structural changes can be detected not only at the level of their somata and large processes but most importantly for neuronal function, at the level of their fine, lamellate distal processes which ensheath neuronal elements, including synapses. Brain metabolism. Although the brain represents only 2% of the body weight, it receives 15% of the cardiac output, and consumes the 20% of total body oxygen and the 25% of glucose at rest. Its stores of glycogen, creatine phosphate and ATP are minute, and would last only about 2 min at normal consumption rates. If supplies were suddenly cut off, optimal functioning would be impaired within a few seconds. The brain therefore relies on a constant supply of glucose, oxygen and blood to remove waste products. High PaCO 2, low PaO 2 and low ph, which are products of metabolic activity, tend to dilate the blood vessels and increase cerebral blood flow; changes in the opposite direction constrict the vessels and decrease blood flow. Cerebral blood flow is regulated through such mechanisms to maintain homeostasis of these chemical factors in the local tissue. The rates of production of these chemical factors depend on the rates of energy metabolism; therefore, cerebral blood flow is 1

2 adjusted to the cerebral metabolic rate. The global oxygen consumption is independent of cerebral activity and does not modify with sleep, but in several districts the neuronal metabolism can greatly change within few seconds. Carbohydrates are the substrates for oxidative metabolism. Cerebral metabolism is unique because no other tissue, except for testis, relies only on carbohydrate for energy. Except in some unusual and very special circumstances, only the aerobic utilization of glucose is capable of providing the brain with sufficient energy to maintain normal function and structure. The brain appears to have almost no flexibility in its choice of substrates in vivo. However, we will see below that the neuron metabolism is more complex than previously hypothesized. Fatty acid transport from blood into parenchymal neurons is much more difficult than other cells since the tight junctions of the blood brain barrier (BBB) severely restrict passage into the brain. The BBB (see below) does not allow the fatty acids bound to albumin to penetrate into the brain. Fatty acids must first move via transcellular transport across both the luminal (apical) and abluminal (basolateral) membranes of the endothelial cells and then across the plasma membrane of the neural cells. The mechanism of fatty acids transport into the brain remains controversial. Several studies support the notion that fatty acids can move across membranes by diffusion. Alternatively, others studies indicate that fatty acids may enter into cells via specific protein-mediated transport. It is puzzling that hydrogen-rich fatty acids are used only poorly as fuel in the brain. Accumulated non-esterified fatty acids or their activated derivatives could exert detrimental activities on mitochondria, which might trigger the mitochondrial route of apoptosis. Three particular problems have been raised: (1) ATP generation linked to β-oxidation of fatty acids demands more oxygen than glucose, thereby enhancing the risk for neurons to become hypoxic; (2) β-oxidation of fatty acids may generate superoxide, which, taken together with the poor antioxidative defense in neurons, causes severe oxidative stress; (3) the rate of ATP generation based on adipose tissue-derived fatty acids is slower than that using blood glucose as fuel, owing to the slow passage of fatty acids across the BBB. Thus, in periods of extended continuous and rapid neuronal firing, fatty acid oxidation cannot guarantee rapid ATP generation in neurons. The disadvantages connected with using fatty acids as fuel have created evolutionary pressure on lowering the expression of the β-oxidation enzyme equipment in brain mitochondria to avoid extensive fatty acid oxidation and to favor glucose oxidation in brain. It was estimated recently that about 20% of the total energy expenses of the adult brain are satisfied by oxidation of fatty acids. It is believed that fatty acid oxidation occurs almost exclusively in astrocytes, and BBB is capable to transport carnitine and fatty acids from the blood to astrocytes. During fasting however, ketone bodies may enter the brain via a specific carrier, and be used. Under normal circumstances, that is, ample glucose and few ketone bodies in the blood, the brain apparently does not oxidize ketones in any significant amounts. In prolonged starvation, the carbohydrate stores of the body are exhausted and the rate of gluconeogenesis is insufficient to provide glucose fast enough to meet the requirements of the brain; blood ketone concentrations rise as a result of the rapid fat catabolism. The brain then apparently turns to the ketone bodies as the source of its energy supply. A vast part of the brain metabolism goes into the regulation of ionic gradients in neurons and glial cells, and into the synthesis of neurotransmitter substances, such as catecholamines, glutamic acid, aspartic acid, glycine and γ-aminobutirric acid (GABA). Somatic enzyme activity was found to be related to both spontaneous and synaptically evoked activities. According to the physiological effects of the cortical cells on each other, there are two categories of neurons: excitatory and inhibitory cells. In the cortex, approximately 85-90% of neurons are excitatory, and only 10-15% are inhibitory. Excitatory cells release transmitters (mostly glutamate) at their synaptic ends that, on contact with the postsynaptic membrane, create currents that depolarize the postsynaptic cell. Inhibitory cells release transmitters (mostly GABA) that tend 2

3 to hyperpolarize the postsynaptic cell, thereby diminishing the effects of the depolarizing currents generated by the excitatory synapses. Some researchers accept that the neuromediators glutamate and GABA may undergo oxidative degradation both in neurons and astroglia: isolated brain and spinal cord mitochondria require glutamate for the maximal rates of oxidative phosphorylation. Thus, in a narrow space of spines and dendrites, several million glutamate molecules postsynaptically transported from synaptic boutons may create local cytosolic concentration of glutamate in a low mm range. Consequently, neuronal mitochondria, particularly those located at the axonal or dendritic synaptic junctions, may temporarily metabolize significant amounts of glutamate. Dendritic metabolism makes the largest single contribution to the metabolic activity of the brain: 85% of mitochondria were localized in dendrites and axon terminals. The GABAergic neurons reuptake a larger portion of the released neurotransmitter GABA compared to their glutamatergic counterparts. GABA s degradation following release from the synapse takes place in both neuronal and astrocytic cells. The GABA degradation forms succinate, and glutamate can be transformed into α-ketoglutarate. Thus, the energy metabolism of the central nervous system is much more complex and compartmentalized than it was thought. In the brain, on average, 1 mm 3 of tissue contains synapses. Because of the high density of synaptic contacts, the neuronal cells may be exposed to mediators released from hundreds of firing synapses. In a human, each neuron receives 40,000 synapses. The blood-brain barrier (BBB). The total length of capillaries in the human brain is of approximately 600 km, with a capillary surface area of about m 2 (per 1.3 kg brain). In fact, almost every neuron is perfused by its own capillary, which highlights the dimension of this interface between the blood and the brain parenchyma. Individual neurons are rarely more than 8-20 μm from a brain capillary. The central nervous system (CNS) homeostasis is essential for the proper functioning of brain cells. The blood-brain barrier (BBB), i.e. the regulated interface between the peripheral circulation and the CNS, participates in CNS homeostasis by preventing the brain from being exposed to the constant oscillations in the concentration of blood constituents and by transporting nutrients and products from brain metabolism in and out of the brain, respectively. The first experimental evidence of the BBB was described on 1885 by Paul Ehrlich, who noted that water-soluble dyes injected into the circulatory system stained all organs except the brain and spinal cord. Subsequently, on 1913 Goldman observed that injection of trypan blue directly into the cerebrospinal fluid (CSF) stained all cell types in the brain but failed to penetrate into the periphery. This suggested that there existed a barrier to the dye between the CNS and the peripheral circulation, which was referred to as the blood-brain barrier (BBB). The BBB physiologically alters the permeability of brain capillaries, so that some substances are prevented from entering brain tissue, while other substances are allowed to enter freely. The main function of the BBB is to protect the brain from changes in the blood levels of ions, amino acids, peptides, and other substances. Proper neuronal function necessitates a highly regulated extracellular environment, wherein the concentrations of ions such as Na +, K +, and Ca 2+ must be maintained within very narrow ranges. Furthermore, the metabolic demands of nervous tissue are considerable, with the CNS accounting for approximately 20% of oxygen consumption in humans. The CNS is also extremely sensitive to a wide range of chemicals; many of the substances we consume in our diet, although readily metabolized and excreted without harm to peripheral organ systems, are in fact quite neurotoxic. It is therefore essential that the interface between the CNS and the peripheral circulatory system functions as a dynamic regulator of ion balance, a facilitator of nutrient transport, and a barrier to potentially harmful molecules. In no other organ is constancy of the internal environment more important than in the brain. Elsewhere in the body, the extracellular concentrations of hormones, amino acids and potassium undergo frequent fluctuations, particularly after meals and exercise or during times of stress. In the 3

4 central nervous system, a similar change in the composition of the interstitial fluid could lead to uncontrolled brain activity because catecholamines and certain amino acids are centrally acting neurotransmitters and potassium influences the threshold for activation of synapses. The neurovascular unit (NVU) The anatomical substrate of the BBB is the cerebral microvascular endothelium, which, together with astrocytes, pericytes, neurons, and the extracellular matrix, constitute a neurovascular unit (NVU) that is essential for the health and function of the CNS. Brain blood capillaries are unusual in two ways. Firstly, the endothelial cells which make up the walls of these vessels are sealed together at their edges by tight junctions (TJ) and adherens junctions (AJ) that form a key component of the barrier; furthermore, there are no detectable trans-endothelial pathways. These junctions prevent watersoluble substances in the blood from passing between the cells (paracellular diffusion) and therefore from freely entering the fluid environment of the brain cells. The interendothelial space of the cerebral microvasculature is characterized by the presence of a junctional complex that includes TJ, AJ, and possibly gap junctions. Whereas gap junctions (if they are present at interendothelial junctions of the BBB) mediate intercellular communication, both AJ and TJ act to restrict permeability across the endothelium. Although disruption of AJ at the BBB can lead to increased permeability, it is primarily the TJ that confers the low paracellular permeability and high electrical resistance. TJ significantly restrict the movement of small ions such as Na + and Cl -, so that the transendothelial electrical resistance (TEER), which is typically 2-20 ohm.cm 2 in peripheral capillaries, can be >1,000 ohm.cm 2 in brain endothelium. TJ are elaborate structures that span the apical region of the intercellular cleft of epithelial and endothelial barrier tissues. They function both as a zipper that affects separation of the apical and basolateral cell membranes, enabling asymmetric distribution of membrane constituents, and a fence that limits paracellular permeability. Morphologically, the TJ of the BBB resemble the tight junctions between epithelial cells rather than those between peripheral vascular endothelial cells. Much like AJ, TJ are composed of transmembrane proteins that form the primary seal linked via accessory proteins to the actin cytoskeleton. The TJ has a valuable function not only in restricting paracellular permeability (gate function), but also in segregating the apical and basal domains of the cell membrane (fence function) so that the endothelium can take on the polarized (apical-basal) properties that are more commonly found in epithelia, such as those of the gastrointestinal tract and kidney. The TJ is an intricate complex of transmembrane (junctional adhesion molecule-1, occludin, and claudins) and cytoplasmic (zonula occludens (ZO)-1 and -2, cingulin, AF-6, and 7H6) proteins linked to the actin cytoskeleton. The expression and subcellular localization of TJ proteins are modulated by several intrinsic signaling pathways, including those involving calcium, phosphorylation, and G-proteins. For instance, the C-terminal of these transmembrane proteins is linked to cytoskeletal actin through ZO-1. In response to pathological stimuli, δpkc may directly or indirectly increase phosphorylation of ZO-1, thus disrupting the association between ZO-1 and the actin cytoskeleton. But phosphorylation of distinct serine, tyrosine, and threonine residues has distinct structural and functional effects. Both extracellular Ca 2+ depletion and increased intracellular [Ca 2+ ] could lead to disruption of BBB TJ. In addition to PKC, Ca 2+ influx or release from intracellular stores can activate any number of kinase signaling cascades, leading to the activation of transcription factors (nuclear factor-κb, camp response element-binding protein, c-fos, etc.) that regulate TJ protein expression. The primary component of AJ is vascular endothelial (VE)-cadherin, a Ca 2+ -regulated protein that mediates cell-cell adhesion via homophilic interactions between the extracellular domains of proteins expressed in adjacent cells. The cytoplasmic tail of VE-cadherin binds to β-catenin 4

5 and plakoglobin, which in turn bind via α-catenin, α-actinin, and vinculin to the actin cytoskeleton, stabilizing the AJ complex. In addition, the platelet endothelial cell adhesion molecule (PECAM) mediates homophilic adhesion. Anatomically, the endothelial cells of the BBB are distinguished from those in the periphery by increased mitochondrial content, a lack of fenestrations, minimal pinocytotic activity, and the presence of TJ. The endothelial cells of the BBB are less leaky than those of the peripheral vessels. The existence of the brain barriers is in part responsible for the initial concept that the brain is an immune-privileged site, with restricted passage of immune cells into the brain. In fact, even though in a healthy person under physiological conditions, immune cell migration across the brain barriers is low, some migration exists and is required for the immune surveillance of the CNS. However, during normal aging and in several diseases of the CNS (such as stroke, trauma, brain tumor, epilepsy, multiple sclerosis, Alzheimer disease, and hypertension) changes in blood composition, brain inflammation and the facilitated entrance of immune cells through the brain barriers can potentially cause neuronal damage and cognitive dysfunction. Since the brain is contained in a rigid, bony skull, its volume has to be kept constant. The BBB plays a key role in this process, by limiting the freedom of movement of water and salts from the blood into the extracellular fluid of the brain. Whereas in other body tissues extracellular fluid is formed by leakage from capillaries, the BBB in fact secretes brain extracellular fluid at a controlled rate and is thus critical in the maintenance of normal brain volume. If the barrier is made leaky by trauma or infection, water and salts cross into the brain, causing it to swell (cerebral edema), which leads to raised intracranial pressure; this can be fatal. Acute Mountain Sickness: headache, loss of appetite, sometimes vomiting, weakness, dizziness, fatigue, and difficulty sleeping. AMS feels exactly like a bad hangover. The exact cause of AMS is still not clear. Of course, AMS is due to low oxygen, but the steps between low oxygen and AMS are fuzzy. Current thinking is that hypoxia (low oxygen) in the air and blood causes the blood vessels of the brain to dilate in an attempt to get more oxygen. Something about this vasodilation causes the headache. Because there is more blood in the brain, the brain is slightly swollen. Although everyone going to high altitude has slight brain swelling, it is worse in those who develop AMS. This swelling causes pressure on the brain as it is squeezed against the skull, and the pressure might be the cause of the symptoms. Secondly, these capillaries are enclosed by the flattened end-feet of astrocytic cells. Astrocytic end-feet almost completely surround intra-parenchymal blood vessels. These astrocyte end-feet encircling endothelial cells (covering >98% of the vascular wall), once thought to aid in the maintenance of the BBB, do not play a substantial role in this sense. However, astrocytes are strongly implicated in the induction of many BBB characteristics, leading to tighter tight junctions (physical barrier), the expression and polarized localization of transporters, including P- glycoprotein (P-gp) and GLUT1 (transport barrier), and specialized enzyme systems (metabolic barrier). From a purely structural perspective, the astrocyte is situated much like a hub in which it receives inputs from thousands of synapses and at the same time it can make contact with the local vasculature. The astrocytes indeed extend processes that surround synapses (see below) and can thereby link neuronal activity with the oxygen and nutrient supply. The perivascular endfeet of astrocytes, which are closely applied to the microvessel wall, show several specialized features characteristic of this location, including a high density of particles containing the water channel aquaporin 4 (AQP4) and the Kir4.1 K + channel, which are involved in ion and volume regulation (see below). Attached at irregular intervals to the abluminal membrane of the endothelium are pericytes. Less is known about the involvement of pericytes at the BBB. Pericytes are uniquely positioned within the 5

6 NVU to serve as vital integrators, coordinators and effectors of many neurovascular functions, including angiogenesis, BBB formation and maintenance (pericyte-derived angiopoetin can induce endothelial expression of occludin), vascular stability and angioarchitecture (the addition of pericytes to cocultures of endothelial cells and astrocytes stabilizes the formation of capillary-like structures), regulation of capillary blood flow (contractile proteins have been found in cerebral pericytes) and clearance of toxic cellular byproducts necessary for proper CNS homeostasis and neuronal function. New studies have revealed that pericyte deficiency in the CNS leads to BBB breakdown and brain hypoperfusion (due to pericapillary edema) resulting in secondary neurodegenerative changes. Pericytes have been shown to migrate away from brain microvessels in rapid response to hypoxia and traumatic brain injury; both of these conditions are associated with increased BBB permeability. Pericytes are involved in the induction and/or maintenance of barrier properties in the cerebral endothelium in a manner similar to glia. Anatomical evidence has been found for direct innervation of the microvascular endothelium and/or associated astrocytic processes by noradrenergic, serotonergic, cholinergic, and GABA-ergic neurons, as well as others. Significant loss of cholinergic innervation of cortical microvessels has been observed in Alzheimer's disease, which may lead to impaired cerebrovascular function observed in this disease (see below). Whether neurons are critical in the development of the BBB phenotype has not yet been demonstrated, but it is apparent that they can regulate critical aspects of BBB function. The extracellular matrix of the basal lamina also interacts with the cerebral microvascular endothelium. Pericytes and endothelial cells are ensheathed by the basal lamina, a membrane 30 to 40-nm thick composed of collagen type IV, heparin sulfate proteoglycans, laminin, fibronectin, and other extracellular matrix proteins. The basal lamina is contiguous with the plasma membranes of astrocyte end-feet, which ensheathe cerebral capillaries. Thus the only way for water-soluble substances to cross the BBB is by passing directly through the walls of the cerebral capillaries, and because their cell membranes are made up of a lipid/protein bilayer, they also act as a major part of the BBB. It is the tight junctions and basal lamina of the cerebral endothelial cells that play the most substantial role in maintaining the barrier. Because of its large surface area and the short diffusion distance between neurons and capillaries, the endothelium has the predominant role in regulating the brain microenvironment. Disruption of the extracellular matrix is strongly associated with increased BBB permeability in pathological states. The extracellular matrix seems to serve as an anchor for the endothelium via interaction of laminin and other matrix proteins with endothelial integrin receptors. Such cell-matrix interactions can stimulate a number of intracellular signaling pathways. Matrix proteins can influence the expression of endothelial TJ proteins, indicating that although the TJ constitute the primary impediment to paracellular diffusion, the proteins of the basal lamina are likely involved in their maintenance. Finally, components of the NVU include the circulating blood cells, such as polymorphonuclear (PMN) cells, lymphocytes and monocytes that adhere and roll along the vascular lumen and perform surveillance of neural signalling and cellular activity. The cellular anatomy of the NVU has recently been extended to include inflammatory circulating cells that participate in CNS immune surveillance by probing and interacting with the luminal surface of brain endothelium covered by thick glycocalyx, as well as perivascular macrophages and microglia, participating in innate immune responses in the neurovascular niche. BBB/NVU properties The endothelial cells of the NVU are highly polarized, with different integral membrane proteins at the luminal and abluminal surfaces. These include various receptors, enzymes and transporters that support the functions of this cellular barrier within the NVU. For example, the endothelial barrier performs vectorial transport of solutes including ions, nutrients and drugs at the blood-brain interface. It also engages in highly specialized interactions with blood cells, through specific luminal receptors, and with elements of the basal lamina and underlying cells (for example, 6

7 astrocytic endfeet and neuron terminals) at the abluminal surface, through specific abluminal plasma membrane proteins. Furthermore, astrocytes and pericytes possess their own complement of transporters, channels, receptors and signalling mechanisms with which they coordinate the role of the NVU in supporting nervous system function. The true BBB is formed by the tight junctions present between the endothelial cells of the capillaries that perfuse the brain parenchyma. Except for small lipophilic molecules such as O 2, CO 2, and NH 3, which diffuse freely across endothelial cells along their concentration gradient, access through the brain barriers of blood-born molecules requires the presence of specific transporter or receptor systems. Also other fat-soluble molecules, such as anesthetics, barbiturates, alcohol and unbound long-chain fatty acids can pass straight through the lipids in the capillary walls and so gain access to all parts of the brain. The endothelium contains transport proteins (carriers) for glucose, amino acids, purine bases, nucleosides, choline and other substances. Stereospecific, but insulin-independent, GLUT-1 glucose transporters are highly enriched in brain capillary endothelial cells and mediate the facilitated diffusion of this polar substrate through the BBB. The activity of these transporters is sufficient to transport two to three times more glucose than normally is metabolized by the brain. Monocarboxylic acids, including L-lactate, acetate, pyruvate and ketone bodies, are transported by a separate stereospecific system. Also several transporters for amino acids are present, such as excitatory amino acid transporters 1-3 (EAAT1-3) and L-system for large neutral amino acids (LAT1). Proteins such as insulin, transferrin, leptin, insulin-like growth factors and vasopressin cross the BBB by a process called receptor-mediated transcytosis. The brain capillary endothelial cell is highly enriched in receptors for these proteins, and following binding of protein to the receptor, a portion of the membrane containing the protein/receptor complex is endocytosed into the endothelial cell to form a vesicle. Although the subsequent route of passage of the protein through the endothelial cell is not known, there is eventual release of intact protein on the other side of the endothelial cell. Conversely, similar mechanisms constitute a way out of the brain for molecules produced by brain cells metabolism; both through passive diffusion and via transporters and receptors located on the brain side of the endothelial cell, such as for glutamine and Aβ peptides (see below). There are also enzymes on the lining of the cerebral capillaries that destroy unwanted peptides and other small molecules in the blood as it flows through the brain. Ecto-enzymes such as peptidases and nucleotidases are capable of metabolizing peptides and ATP, respectively, whereas intracellular enzymes such as monoamine oxidase and cytochrome P450 (1A and 2B) can inactivate many neuroactive and toxic compounds. Most neurotransmitters, such as dopamine, present in the blood do not enter the brain because of their low lipid solubility and lack of specific transport carriers in the luminal membrane of the capillary endothelial cell. In contrast, L-DOPA, the precursor for dopamine, has affinity for the large neutral amino acid-transport system and more easily enters the brain from the blood than would be predicted by its lipid solubility. This is why patients with Parkinson's disease are treated with L-DOPA rather than with dopamine; however, the penetration of L-DOPA into the brain is limited by the presence of the enzymes L-DOPA decarboxylase and monoamine oxidase within the capillary endothelial cell. This enzymatic BBB limits transendothelial passage of L-DOPA into the brain and explains the need for large doses of L-DOPA in the treatment of Parkinson's disease. Therapy is enhanced by concurrent treatment with an inhibitor of peripheral L-DOPA decarboxylase. Finally, there is another barrier process that acts against lipid-soluble molecules, which may be toxic and can diffuse straight through capillary walls into the brain. In the capillary wall there are specialized efflux pumps which bind molecules and transport them back into the blood out of the brain. P-glycoproteins (P-gp) are transmembranous, ATP-dependent pumps originally discovered for their ability to confer multidrug resistance to neoplastic cells. P-glycoproteins typically are expressed by epithelial barriers and BBB endothelial cells. 7

8 Thus, characteristics of the BBB are the following: tight junctions that seal the pathway between the capillary (endothelial) cells; the lipid nature of the cell membranes of the capillary wall which makes it a barrier to watersoluble molecules; carriers and ion channels; the 'enzymatic barrier' that removes molecules from the blood; the efflux pumps which extrude fat-soluble molecules that have crossed into the cells. The BBB is thus a key element in the normal functioning of the brain, and isolates it from disturbances in the composition of the fluids in the rest of the body. The successful evolution of a complex brain depends on the development of the BBB. Although the BBB is an obvious advantage in protecting the brain, it also restricts the entry from the blood of water-soluble drugs which are used to treat brain tumors or infections, such as the AIDS virus, which uses the brain as a sanctuary and hides behind the BBB from body defense mechanisms. To overcome these problems drugs are designed to cross the BBB, by making them more fat soluble. But this also means that they might enter most cells in the body and be too toxic. Alternative approaches are to make drug molecules that can ride on the natural transporter proteins in the cerebral capillaries, and so be more focused on the brain, or to use drugs that open the BBB. The blood-cerebrospinal fluid barrier (BCSFB) Besides the BBB, another barrier separates the CNS from the periphery: the blood-cerebrospinal fluid barrier (BCSFB). While the BBB has been well recognized for long, the BCSFB is more rarely mentioned, which is to regret when considering its functions, which include producing most of the cerebrospinal fluid (CSF). The BCSFB is formed by the choroid plexus (CP) epithelial cells. The CP, phylogenetically and ontogenetically conserved, develops early during embryogenesis and already constitutes a functional barrier within the first weeks of gestation. The CP is positioned within the ventricles of the brain: one in each lateral, one in the third and one in the fourth. Grossly, the CP is a lobulated structure formed by a unique and continuous line of epithelial cells originating from the ependymal wall of the ventricles, which floats in the CSF space. These epithelial cells are bound to each other by tight junctions and rest on a basal lamina and on a central core formed by connective and highly vascularized tissue. The apical side of the epithelial cells faces the CSF and contains numerous villosities, while the basolateral side faces the blood, lying in the stroma in contact with several capillaries. Of notice, the capillaries that irrigate the CP are fenestrated, i.e. at the CP there is no BBB. In addition to the fenestrated capillaries, dendritic cells, fibroblasts and macrophages populate the central core of the CP stroma. Although the passage of molecules and cells is possible in the vascularized stroma, these do not reach the CSF through paracellular transport due to the tight junctions between CP epithelial cells. Ultrastructurally, the epithelial cell contains numerous mitochondria, Golgi apparatus, smooth endoplasmic reticulum and lysosome-like vesicles, which demonstrates that it is a structure with great synthetic capacity. The main CP function is the production of CSF. The CSF is a clear, slightly viscous liquid with few cells and a protein concentration about 10 times lower than that of the blood. An adult human contains approximately 150 ml of CSF filling the ventricles, the subarachnoid space and the spinal cord, and these are renewed 3-4 times daily, testifying to the great secretory capacity of the CP epithelium. Drainage of CSF occurs at the arachnoid villus in the venous sinus. By influencing the CSF composition, which ultimately impacts on the brain parenchyma, the CP may interfere with brain homeostasis in health and in disease. Tanycytes are special ependymal cells found especially in the third ventricle of the brain and have processes extending deep into the hypothalamus. It is possible that their function is to transfer chemical signals from the CSF to the central nervous system. The term tanycyte comes from the Greek word tanus which means elongated. 8

9 Circumventricular organs Both the BBB and the BCSFB contain transporters and receptors in the basolateral and apical sides, and therefore are able to transport molecules into and out of the brain, and to convey receptormediated signals, similarly in both directions. The BBB exists in all areas of the brain, except in the so-called circumventricular organs, small sized structures lining the cavity of the third ventricle (neurohypophysis, vascular organ of the lamina terminalis, subfornical organ, pineal gland and subcommissural organ) and of the fourth ventricle (area postrema). These discrete organs have highly fenestrated capillaries that are easily penetrated by large and small molecules as well as ions. The circumventricular organs are located close to essential control centres in the hypothalamus and brain stem regions regulating respiration, blood glucose concentration, and extracellular fluid osmolality: Pineal body: secretes melatonin and neuroactive peptides. Associated with circadian rhythms. Neurohypophysis (posterior pituitary): releases neurohormones like oxytocin and vasopressin into the blood. Area postrema: "vomiting center"; when a toxic substance enters the bloodstream it will get to the area postrema and may cause the animal to throw up. In this way, the animal protects itself by eliminating the toxic substance from its stomach before more harm can be done. Subfornical organ: important for the regulation of body fluids. Vascular organ of the lamina terminalis: a chemosensory area that detects peptides and other molecules. Median eminence: regulates anterior pituitary through release of neurohormones. Astrocytes Astrocytes are star-shaped glia cells characterized by small cell somata (< 10 μm diameter), numerous highly branched fine processes that extend for distances up to 100 μm and that make contact with neuronal processes at the synapses. Astrocytes derive their name from the numerous processes that emerge from their cell-body conferring on them a star-shaped morphology. Recent studies have suggested that there may be many types of astrocytes (11 distinct phenotypes). In the mature brain, there are many more astrocytes than neurons: it is estimated that there are approximately 1.4 astrocytes for every neuron in the human brain. Our view of astrocytic morphology has been based both on Cajal's metal impregnation methods and on the intermediate filament glial fibrillary acidic protein (GFAP) staining. Until recently, the functions of astroglia were considered as secondary, supporting the neuronal functions. However, more and more evidence is accumulating, which leads us to understanding that, without astroglia, the neuronal functions would, probably, never go evolutionary higher than shellfish or worms. Neurons evolved before astrocytes and in basal taxa they must solve their energy problem unaided. However, as the nervous system became more complex, astrocytes appeared and became larger and more abundant with encephalization. Therefore, we have to consider functions of the central nervous system, as a result of concerted work of astroglia and neurons. Astrocytes are involved in almost all aspects of brain function. Some of the main functions of the astroglia are: metabolic support of neurons with nutrients such as lactate; storing glycogen, which is the glucose reserve buffer; transport of glucose and other nutrients from blood flow into the brain; neurotransmitter uptake and release and neurotransmitter synthesis de novo; regulation of ion concentration in the extracellular space; synapse formation. Furthermore, during brain development, astrocytic processes guide neurons in their migration from the wall of the ventricles to the cortex. Astrocytes induce synapse formation in several classes of 9

10 neurons, both by direct contact with neurons and by secreting factors that regulate synapse formation as well as pre- and post-synaptic functions (tripartite synapse, see below). These factors include TNF-α and thrombospondins that promote the formation and maturation of synapses. Astrocytes produce a variety of trophic factors including nerve growth factor, brain-derived neurotrophic factors, neurotrophin 3 and 4, glial-derived neurotrophic factor, and insulin-like growth factor. Astrocytes can alter how a neuron is built by directing where to make synapses or dendritic spines. Several authors suggest that the astrocyte provides spatially and temporally balanced excitation and inhibition to coordinate neuronal and synaptic networks. Astrocytes become reactive in response to a number of brain insults, including trauma, infection, epilepsy, and neurodegeneration. Certain studies suggest that reactive astrocytosis is accompanied by astrocytic proliferation, and recent cell fate mapping studies suggest that adult astrocytes are able to divide following brain injury. These studies demonstrate that astrocytic reactivity and proliferation are beneficial in the context of certain models of acute brain injury. Probably the least expected finding has been that information processing is not an exclusive property of neurons but that it is shared by astrocytes. Astrocytes thus participate in a number of interactions that are central to the development, function, and repair of the CNS. The tripartite synapse A major notion that has emerged from neuroscience over the past few years is the concept of the tripartite synapse, which has compelled neuroscientists to consider the influence of glia in synaptic function. This means that synapses do not consist of just a pre- and postsynaptic neuronal element, but that many also have an astrocytic projection that envelops the synapse. It is quite remarkable to find that all the synaptic endings in the brain are closely associated with astrocytic processes to form tripartite junctions. Each astrocyte can form 150,000 tripartite synapses. The network of astrocytic processes forms the infrastructure on which all other CNS cells and vessels are anchored. A single astrocyte creates a kind of synaptic island defined by its ensheathing processes. The combination of this structural relationship together with new-found appreciation of the presence of dynamic, activity-dependent biochemical signaling between neurons and astrocytes suggests that the astrocyte is an important integrator of neuronal activity. Recently, it has become apparent that glial cells play a much more active role in regulating neural function because they not only respond to information coming from the neurons, but also can transmit information back to the neurons. This two-way dialogue, which depends upon the neuronal-astrocyte communication and the reciprocal astrocyte-neuronal communication systems, may play a critical role in modulating neuronal activity and information processing in the brain. It is becoming evident that this active participation in brain function depends upon a process of astrocyte excitability. Astrocytes listen and talk to synapses and play roles in synaptic modulation and in mediating synaptic cross-talk. The synaptic localization of astrocytes means they are ideally placed to monitor and respond to synaptic activity. By their mere presence, astrocytes can be considered a physical barrier to restrict spillover and diffusion of locally released, potentially active molecules into the extracellular space. Moreover, by their position, they contribute to the regulation of the microenvironment in which neurons develop and function, maintaining a tight control on local ion and ph homeostasis, delivering glucose, providing metabolic substrates, and clearing away metabolic waste. Astrocytes form a syncytium Astrocytes are connected between them through gap junctions, thus giving rise to what is called the astrocytic syncytium. The connexin (Cx) family of proteins were identified as the molecular constituents of gap junction channels (GJCs). Cxs are organized as hexamers embedded in the plasma membrane (emichannels) that, when associated head-to-head between two neighbouring cells, form a GJC. Under certain conditions, Cxs can operate as hemichannels, allowing exchanges between the cytoplasm and the extracellular medium. The central pore of GJCs allows the passage of ions (ionic coupling) and small molecules (biochemical or metabolic coupling), with a cut-off selectivity of around 1 to 1.2 kda. GJCs are selective for several endogenous molecules, including 10

11 second messengers (camp, IP 3 and Ca 2+ ), amino acids (glutamate, aspartate and taurine), nucleotides (ADP, ATP, CTP and NAD), energy metabolites (glucose, glucose-6-phosphate and lactate), small peptides (glutathione) and RNA (24mer), but not for large molecules, such as nucleic acids, proteins and lipids. The transjunctional and transmembrane voltage dependence of GJC conductance can also act as a selectivity filter to limit and compartmentalize the diffusion of biological signalling molecules. To coordinate information transfer in a reliable and efficient manner, astrocytes need a strong modality of intercellular communication: Cx-mediated pathways can certainly fulfil this function. Indeed, injections of intercellular tracers into one astrocyte revealed that hundreds of cells can contribute to astroglial networks in the hippocampus and other regions of the brain. In the adult brain, Cx43 (also known as GJA1) and Cx30 (also known as GJB6) are the main Cxs in astrocytes. K + spatial buffer mechanism The concentration of extracellular K + needs to be tightly regulated since its accumulation in the extracellular space can alter neuronal excitability dramatically. In normal brain activity, neurons release neurotransmitters and K +, and take up Na +, while glucose metabolism generates water. The neurotransmitters and ions are generally recycled, whereas water must be removed from the brain and excreted. Astrocytes contribute to ionic, amino acid, neurotransmitter and water homeostasis of the brain in several ways, and astrocytes that form perivascular endfeet at the BBB have a particular role in these processes. Astrocytes take up K + via the glial isoform of the Na + /K + pump and Na +,K +,2Cl - co-transporters (NKCC) or through specific ion channels, like inward-rectifying K + (Kir4.1) channels and Ca 2+ - activated potassium channels (BK channels). They then transfer it from sites of accumulation to areas with lower concentrations, to finally extrude it in the extracellular space, a kind of K + spatial buffer mechanism that may depend on gap junction coupling. For both channel- and transportermediated K + uptake, the net ion gain results in osmotic water uptake and slight cell swelling; the high density of aquaporin 4 (AQP4) water channels in perivascular astrocytic endfeet facilitates redistribution of this water. As the brain endothelium has low water permeability, it is likely that the excess metabolic water joins the interstitial fluid being secreted into the pericapillary space by the endothelium. Interstitial fluid outflow involves perivascular spaces around large vessels, and clearance routes either through the CSF or following alternative pathways to neck lymphatics. In this way the astrocytic membrane potential is maintained close to the equilibrium potential of K + : when extracellular K + rises, K + ions flow into astrocytes through these inward rectifiers. Astrocytespecific knockout of Kir4.1, the major Kir channel in these glia, results in seizure activity and premature death. Generally, astrocytes have a high resting K + conductance and are coupled by gap junctions (see below): for this reason, the first major function assigned to these cells is the clearance of extracellular K + following elevated periods of neuronal activity. The K + released during periods of neuronal activity is buffered by a K + siphoning process whereby the K + is taken up by astrocytic process near the neurons and is then transferred throughout the astrocytic syncytial system finally to be released from the end foot processes on the blood vessels. The presence in astrocytes of Ca 2+ -activated potassium channels (BK channels) allows for coupling of astrocytic Ca 2+ signaling (discussed below) to the release of K + from astrocytic endfeet onto blood vessels. Astrocytes, by expressing different transporters, are also involved in the clearance of synaptically released neurotransmitters, like glutamate and GABA (see below). Taken together, such passive and active properties of astrocytes serve to limit interneuronal communication mediated by volume transmission and prevent spillover of transmitters originating in synapses, thus preserving point-topoint neuronal communication. 11

12 Astrocyte-neuron lactate shuttle. Processing of information in the brain is metabolically expensive. It has been said previously that the brain accounts for only 2% of our body mass but 20% of our oxygen and glucose consumption. Neurons contribute at most 50% of cerebral cortical volume, and the astrocyte-to-neuron ratio increases with increasing brain size; this is an important consideration when approaching the study of the cellular bases of brain energy metabolism in humans. It is therefore clear that glucose reaching the brain parenchyma through the circulation should provide energy substrates to a variety of cell types, only a portion of which are neurons. Neurons do not have direct access to glucose, from which they are separated by four plasma membranes: two endothelial and two astrocytic. Astrocyte processes are wrapped around synaptic contacts (the tripartite synapse), whereas particular astrocytic profiles, the end-feet, surround intraparenchymal capillaries, and provide a cellular zone interposed between the bloodstream and other elements of the brain parenchyma. Astrocytic end-feet cover 99.7% of the brain capillary, giving astrocytes control over glucose flux into the tissue. This latter structural feature has long been suggested as evidence indicating a role of astrocytes in the transit of glucose and oxygen from blood to neurons. Importantly, both endothelial and astrocytic cells express the glucose transporter GLUT1 at the surface of these structures to allow for the efficient transfer of glucose into the CNS. Astrocytes express GLUT1 which is responsible for the uptake of glucose from associated capillaries. In contrast, neurons primarily express GLUT3 transporters. While a defect of GLUT3 in mice is not important for neuronal metabolism, a defect of GLUT1 results in decreased glucose uptake and severe neurological defects, suggesting that essential glucose uptake in the CNS primarily occurs in a GLUT1-dependent manner in astrocytes. Thus, unlike most other cells in the body, neurons are not able to control their own fuel supply, at least not directly. Intracellular distance is a second handicap for neurons, because most axons and dendrites are too far from their soma to receive timely metabolic support. For example, ATP would take over 1 h to diffuse between soma and a needy dendrite located 1 mm away, far too long compared with the ATP turnover, which occurs in seconds. The energy dissipated by brain is high, but is even higher in heart and kidney. What makes the brain stand out in terms of energy demand is the nature of individual cells and particularly of specific regions within these cells. For a neuron jumping from quiescence to a firing rate of 10 Hz, ATP consumption increases approximately sevenfold; such an energy demand without a matching increase in the generation of energy would deplete the ATP pool within seconds. Paradoxically, the neuron is the only highly energetic cell devoid of energy storage. Liver cells store fat and glycogen, and muscle and kidney cells store glycogen, but neurons posses neither of the two. Owing to their electrical signaling, neurons have higher energy requirements than astrocytes but they are slower at metabolizing glucose. Transcriptome analysis of brain cells revealed weaker expression of glycolytic enzymes in neurons than in astrocytes, but the most dramatic divergence has been found at the post-translational level. Under resting conditions, glucose can be actively used through the PPP in neurons due to the low activity of the glycolytic-promoting enzyme, PFKFB3, an inducible isozyme of 6-phosphofructo-2-kinase (PFK-2), which is continuously degraded by the E3 ubiquitin ligase. Neurons have a functional block at phosphofructokinase that secures a sustained flux through the pentose phosphate pathway (PPP), the anabolic branch that keeps the cytosol reduced and protected against oxidative stress. Within the brain, neurons are the cells most vulnerable to excess reactive oxygen and nitrogen species (ROS and RNS). During their long lifespan, neurons are subjected to considerable free radical-mediated DNA damage. Protection against such damage is likely to require an efficient antioxidant system. Neurons are intrinsically equipped with a biochemical mechanism that links glucose metabolism to antioxidant defence. Neurons actively metabolize glucose through the PPP, which maintains the antioxidant glutathione in its reduced state, hence exerting neuroprotection. As neurons have very low concentrations of glutathione, and low activity of γ-glutamyl cysteine synthetase necessary for its synthesis, it can be speculated that the diversion of glucose towards the generation of reduced glutathione has evolved 12

13 as a defence mechanism against oxidative stress. However, their survival relies also on the antioxidant protection promoted by neighbouring astrocytes. The down side of antioxidation and protection against apoptosis is that neurons cannot activate glycolysis efficiently on ATP demand. Both astrocytes and neurons have the capacity to fully oxidize glucose and/or lactate. But the available results suggest that astrocytes primarily rely on high rates of glycolysis in the cell cytosol and respond to ATP demand with robust glycolytic stimulation, whereas nerve cells primarily use mitochondrial oxidative phosphorylation to support their energy demands. Post-translational divergence between neurons and astrocytes is also found for pyruvate dehydrogenase (PDH), the gatekeeper of mitochondria. Neurons maintain an active PDH, funneling pyruvate into the Krebs cycle, whereas PDH in astrocytes is tonically inhibited by phosphorylation, shunting pyruvate towards lactate production and export. Thus glucose, taken up by astrocytic processes from cerebral blood vessels, via GLUT1, is metabolized at a high capacity by glycolysis to produce lactate which is released into the extracellular space via monocarboxylate transporters (MCTs). Extracellular lactate is then taken up by neurons and converted to pyruvate, likely by lactate dehydrogenase (LDH). The isoforms of LDH are differentially expressed in neurons and in astrocytes; whereas neurons express preferentially LDH1 mainly pyruvate-producing astrocytes express LDH5 associated with high lactate-producing tissues. Furthermore, astrocytes express the monocarboxylate transporters-1 and -4 (MCT1 and MCT4), which are responsible for lactate efflux, whereas neurons express MCT2, specialized in lactate influx. Glucose as the sole energy substrate cannot support neuronal survival in vitro, whereas lactate and pyruvate are adequate substrates for brain tissue in vitro. In fact, synaptic activity can be maintained in cerebral cortical slices with only lactate or pyruvate as a substrate. Lactate transport from astrocytes to neurons is essential for long-term memory. Genetic knockdown of the lactate transporters of astrocytes (MCT4), oligodendrocytes (MCT1), and neurons (MCT2) induced amnesia and disrupted long-term memory formation. Significantly, exogenous lactate (but not glucose) rescued memory formation in MCT4 and MCT1 knockdowns but not in MCT2 knockdown, emphasizing the relevance of lactate transfer between glial cells and neurons. Taken together, these results suggest that the astrocytic lactate export by MCT-1 and/or MCT-4, and subsequent import into neurons through MCT2, is essential for long-term memory. In brain white matter, an extra layer of complexity comes from oligodendrocytes and their intimate relationship with axons. In the optic nerve, a white matter tract that can be studied in vitro, lactate sustained firing as effectively as glucose. Further supporting the importance of the intercellular lactate shuttle, knockdown of the oligodendrocyte lactate transporter MCT1 induced axonal damage that was rescued by exogenous lactate. The high consumption of locally produced lactate in the presence of physiological glucose is shown by the fact that lactate displaces glucose as a brain energy substrate when arterial lactate is increased either by vigorous exercise or by venous infusion. Thus lactate and pyruvate can sustain synaptic activity in vitro, but because of their limited permeability across the brain capillaries, they cannot substitute for plasma glucose to maintain brain function. Pyruvate is subsequently converted in neurons to acetyl CoA, through the activity of the pyruvate dehydrogenase complex and then enters the tricarboxylic acid (TCA) cycle followed by the generation of NADH to fuel oxidative phosphorylation via the mitochondrial electron transport chain. In this way neurons oxidize lactate to CO 2 and H 2 O. A further biochemical difference is seen for glycogen, the sole energy reserve for the brain, which is stored only in astrocytes. The glycogen turnover rate is very rapid in brain. Glycogen in the brain is worth min of the resting metabolic rate, and is actively degraded during hypoglycemia, ischemia, and strenuous exercise. The physiological activation of specific neuronal circuits results in the mobilization of glial glycogen stores. What is the metabolic fate of the glycosyl units mobilized from glycogen? As noted earlier, lactate, an efficient fuel for neuronal activity, is the main metabolic intermediate released by astrocytes. No glucose is released from astrocyte cultures. 13

14 A specific role for glycogen-derived lactate has been suggested from the observation that pharmacological inhibition of glycogen phosphorylase in rat hippocampus impaired the formation of long-term memory, an effect that could be rescued by exogenous lactate and that confirmed the amnesic effect of phosphorylase inhibition. Glycogen is toxic for neurons. Neurons have glycogen synthase but keep it inactive, and do not express glycogen phosphorylase. Mutations of the inactivation mechanism lead to accumulation of aberrant glycogen (excessive phosphate content and branches at abnormally long intervals) in neurons and the development of Lafora disease, a fatal genetic condition characterized by progressive ataxia and dementia. From this and other evidences, it can be inferred that most of the activation-induced glucose uptake in the brain parenchyma, as visualized by the PET, occurs in astrocytes. In vivo as well as in vitro studies indicate that the physiological stimulation of a given brain region triggers a rapid activation of glycogenolysis (proven to be exclusively astrocytic) and glycolysis, which in turn result in the release of lactate. The energy load is further distributed by glucose/lactate-permeable gap junctional communication between astrocytes, between astrocytes and oligodendrocytes, and between oligodendrocytes themselves. Given the low expression of glycolytic enzymes in oligodendrocytes, it seems plausible that oligodendrocytes may act as lactate conveyors, supplying astrocytic lactate to axons on demand. The outsourcing of the energetic function of glycolysis into astrocytes is advantageous for neurons in several ways. 1) It allows neurons to generate energy on demand without compromising their antioxidant status. Because the volume of the astrocyte that envelops a given neuronal process is much larger than the process itself, outsourcing multiplies the amount of lactate readily available at any lactate concentration. 2) Another important consequence of outsourcing glycolysis to glial cells is the spatial separation of lactate production from lactate clearance. In an isolated cell, the glycolytic rate is controlled by the ATP/ADP ratio, which is in turn determined by mitochondrial activity and energy consumption. Because this is a negative feedback mechanism, pyruvate and lactate levels may only decrease during activity. The physical and therefore regulatory separation of glial glycolysis from the energy demands of neurons introduces a degree of freedom for glycolysis, allowing lactate production to vary independently of neuronal energy status. 3) Neuronal activity is indicative of an increase in energy demand. Astrocytes have a key role in coupling neuronal activity to the use of glucose in the brain (neurovascular coupling, i.e. the link between neuronal activity and energy supply from blood flow, in which astrocytes participate). Astrocytes take up glutamate, the primary excitatory neurotransmitter in the CNS, and convert it to glutamine as part of a recycling mechanism (see also below). In doing so, glutamate triggers a cascade of molecular events leading to an enhancement of glucose utilization by astrocytes. Collectively, the astrocyte-neuron lactate shuttle model posits that: (1) neuronal activity increases extracellular glutamate which is taken up by astrocytes via Na + /glutamate co-transporters leading to activation of the Na + /K + -ATPase; (2) within seconds the need of ATP triggers the glucose uptake by GLUT1 and glycolysis in astrocytes which leads to a large increase in the production of lactate followed by its release into the extracellular space and by the generation of new ATP; (3) the transport of lactate into neurons is used as an energy substrate for oxidative and nonoxidative derived ATP production; (4) meanwhile the glycolytic rate and activation of the Na + /K + -ATPase is acutely stimulated by a rise in extracellular K + (see above). Recently glutamate released from neurons has been shown to generate metabolic waves in cultured astrocytes, resulting in coordinated uptake of glucose by gap junction-connected astrocytes, thereby amplifying the metabolic response. Therefore, glutamatergic synaptic activity enhances both glucose uptake and glucose trafficking in astroglial networks and might serve to efficiently supply 14

15 energy metabolites to remote sites of high neuronal demand. The diffusion of glucose among hippocampal astrocytes is enhanced by neuronal activity. This process is inhibited by astrocytespecific genetic ablation of connexins 43 and 30. Following its uptake by astrocytes, glucose is either stored as glycogen or metabolized to lactate. Astrocytic glucose and glycogen metabolism has been shown to be regulated by neuronal mechanisms. The astrocyte-neuron lactate shuttle model is not completely convincing, and at least needs the intervention of further metabolic pathways. For instance, up to 20% of the brain s energy metabolism is supplied by the β-oxidation of fatty acids in astrocytic mitochondria. This means that upon neuronal activation, astrocytes simultaneously supply neurons with lactate as a substrate for neuronal mitochondria, whereas astrocytic mitochondria supply high levels of ATP to maintain the glutamate/glutamine cycle (see below) and other ATP-consuming functions. In this way, both the aerobic glycolysis and oxidative phosphorylation will proceed simultaneously in the astrocyte as quasi-irreversible pathways. The idea that β-oxidation of fatty acids may be the preferable substrate for production of acetyl- CoA in astrocytes is supported by the properties of the PDH in astrocytic mitochondria. All components of PDH are expressed in both neurons and astrocytes in culture. However, in astrocytes, the PDH activity is kept strongly inhibited through phosphorylation of the PDHα subunit. In contrast, neuronal PDH operates close to maximum with much lower levels of phosphorylated PDHα. Dephosphorylation of astrocytic PDHα restores the PDH activity and lowers lactate production. This intrinsic property of the astrocytic mitochondrial PDH will favor glucose entering the glycolytic pathway and production of lactate, rather than entering the TCA cycle. Together, these properties of key enzymes in astrocytes direct more pyruvate to formation of lactate. This will also promote β-oxidation of fatty acids that will supply the TCA cycle with acetyl- CoA. Oxaloacetate for the citrate synthase reaction of TCA is provided by the pyruvate carboxylase, which is mainly an astrocytic enzyme, whereas its activity is low in the neuron. In astrocytes, fatty acids may be the major source for acetyl-coa. When the neuronal synaptic terminals become quiet, astrocytes will also become metabolically inactivated because the accumulation of lactate will inhibit glycolysis. 15

16 Glutamate-glutamine cycle. A crucial requirement for the action of glutamate and other neurotransmitters in the CNS is that their extracellular concentrations be kept low. The level of glutamate in plasma is high compared with that in cerebrospinal fluid. The BBB essentially prevents the entry of glutamate into the CNS, and most of the glutamate that is present in the brain is synthesized de novo in brain. In young animals, the BBB has a higher permeability to glutamate, and excessive dietary uptake of glutamate can be excitotoxic. The BBB becomes impermeable to glutamate at seven days after birth. Neurons contain about 5-10 mm glutamate in their cytoplasm, whereas astrocytic concentrations are somewhat lower about 2-3 mm. The lower concentration of glutamate in astrocytes is attributed to the presence of glutamine synthetase, an enzyme that converts glutamate to glutamine as part of the recycling of synaptically released glutamate. The concentration of glutamate in the cerebrospinal fluid is about 1 μm, but can increase to 20 μm under pathological conditions that are characterized by defects in the BBB and/or by cellular damage, such as stroke, trauma, multiple sclerosis and meningitis. The exact level of extracellular glutamate is not known, owing to the limitations of the available techniques (maybe in the range of 1-2 μm, but this is probably an overestimation). In fact, the glutamate transporters that are expressed by both neurons and astrocytes (see below) have the capacity to reduce glutamate concentrations to the nanomolar range. Astrocyte functions include modulation of synaptic function via glutamate transporters, which convey glutamate from the synaptic cleft into the cell in a process known as the glutamateglutamine cycle. Contrary to common sense, glutamatergic neurons are incapable of de novo synthesis of glutamate from glucose via α-ketoglutarate, and there is ample evidence that most of the released neurotransmitter glutamate is lost to surrounding astrocytes. As glutamate is formed from the tricarboxylic acid (TCA) cycle metabolite α-ketoglutarate, this faces the glutamatergic neurons with the paradox of how to replenish the glutamate pool without draining their TCA cycle of intermediates. The solution to this paradox lies in the discovery of intercellular compartmentation of glutamine and glutamate pools, related to astrocytes and neurons, respectively, which led to the suggestion of a glutamate-glutamine cycle working between glutamatergic neurons and astrocytes. Glutamate is removed from the synaptic cleft by a group of glutamate transporters (EAAT: excitatory amino acid transporters). EAAT1 and EAAT2 in humans are primarily expressed on astrocyte membranes, but are widely distributed in the CNS also in neurons and microglia. Human EAAT3 is also found outside of the CNS. Human EAAT4 is expressed by cerebellar neurons. The glutamate transporters are capable of concentrating intracellular glutamate more than 10,000- fold compared with the extracellular environment. Glutamate uptake into astrocytes occurs through a cotransport with Na + and results in the activation of Na + /K + -ATPase, probably by an increase in the intracellular concentration of Na +. The cellular uptake of glutamate is coupled also to the transport of protons and potassium. The stoichiometry of this process is such that the inward movement of one glutamate, three Na + ions and one H + are coupled to the outward transport of one K + ion (the transporter is electrogenic, as there is a net gain of two positive charges for each glutamate that is transported). In astrocytes, glutamate may also derive from endogenous sources, mainly from α-ketoglutarate, a product of the TCA cycle. In glial cells, glutamate is converted to glutamine (Gln) by glutamine synthetase before it is released into the extracellular space. By being the sole supplier of Gln to synapses, astrocytes have the potential to shape neuronal activity and information processing by controlling neuronal inhibition. Gln is taken up by glutamatergic neurons and converted to glutamate by glutaminase before being packaged into synaptic vesicles. The levels of glutamate are maintained in vesicles in the 16

17 presynaptic neurons by two vesicular glutamate transporters, VGLUT1 and 2. This process is driven by an electrochemical gradient of H + established by V-ATPase, which, together with the VGLUT activity, affect vesicular glutamate content and subsequently the glutamatergic signaling. VGLUT is believed to exchange one proton for one glutamate molecule during the transport process. When the synapse is activated, glutamate is released into the synaptic cleft and activates the ionotropic and metabotropic glutamate receptors (see below). Glutamate receptor activation stimulates further downstream cellular pathways. For instance, glutamate release stimulates glycolysis and lactate accumulation (see above). Indeed, the energy costs of the Na + /K + -ATPase pump and glutamine synthesis result in decrease of [ATP], and increase of [ADP], [Pi], and [AMP], which largely stimulate glycolysis in the astrocytes with resultant lactate production. For each molecule of glutamate recycled, one molecule of ammonia will be produced in the neurons. This ammonia will have to be translocated out of the neurons and back to the astrocytes for detoxification (resynthesis of glutamine). The conventional assumption is that unionized ammonia (NH 3 ) can pass the BBB by diffusion, whereas translocation of the ionized species (NH 4 + ) can be neglected. Since 99% of ammonia exists in ionic form at physiologic ph, it might enter the brain mainly by a transcellular route, using an array of K + channels and transporters or by substituting other cations with similar hydrated radius. This is because an elevated ammonia concentration has detrimental effects on a number of cellular functions, including inhibition of pyruvate dehydrogenase and α-ketoglutarate dehydrogenase. Also Gln may function as a carrier of excess ammonium, and is transported across the BBB to be disposed off by the circulation. The enzymes necessary for sustaining glutamate homeostasis are heterogeneously distributed among neurons and astrocytes. The glutamine synthesizing enzyme, glutamine synthetase, is selectively localized in astrocytes and the enzyme transforming glutamine to glutamate, glutaminase, is preferentially expressed in neurons. Finally, pyruvate carboxylase, the quantitatively most active anaplerotic enzyme in the brain (producing oxaloacetate) is mainly confined to the astrocytic compartment. These are the factors that force neurons to rely on astrocytes for providing precursors for anaplerosis of their TCA cycle constituents. Glutamate transporter-mediated uptake is critical for preventing the sustained activation of ionotropic glutamate receptors that would lead to excitotoxic cell death. Excitotoxicity involves sustained elevations of intracellular Ca 2+, which initiates a complex cascade of intracellular events that lead to destruction of the cell (see below). Astrocytes modulate the cerebral blood flow. The increased energy demand of active neurons is also met by local increases in blood flow in the area of elevated neuronal activity. This phenomenon, which was first described in the late 1800s, is a fundamental event in brain function. Local increases in blood flow result from the rapid dilation of arterioles and capillaries of a restricted area in response to an episode of high neuronal activity. As a consequence, blood flow increases in that region within a few seconds, thereby ensuring that most active neurons receive an adequate supply of oxygen and metabolic substrates for energy consumption. Local accumulation of metabolic products has been initially proposed to directly control blood flow. Traditionally, it was thought that active neurons generate a metabolic signal (a fall in O 2 or glucose concentration, or a rise in CO 2 concentration), which triggers an increase in blood flow. Although under particular circumstances, such as brain hypoxia or ischemia, this process may indeed affect blood vessels, results obtained over the last few years provide conclusive support for the view that blood flow is directly coupled to neuronal activity rather than to local energy needs. For instance, the fractional increase in blood flow induced by sustained neuronal activity is at least 4-fold greater than the increase in ATP consumption by the neurons. It has been discovered that neurotransmitter-mediated signalling, particularly by glutamate, has a major role in regulating cerebral blood flow. Much of the regulation of blood flow is mediated by 17

18 astrocytes, which are an important integrator also of the local control of cerebrovasculature. The present knowledge on the multiple signaling pathways that during activation lead to the production of vasoactive factors suggests that the molecular mechanism at the basis of functional hyperemia is highly complex and may not necessarily be the same in all brain regions. Although various aspects remain to be elucidated, most recent studies highlight a central role of neuron-to-astrocyte signaling in the local control of microcirculation. Because of their polarized anatomical structure and of the vicinity of their endfeet with contractile elements of blood vessels, such as smooth muscle cells in arterioles and pericytes in capillaries, astrocytes have been long proposed to contribute to the regulation of the blood flow during neuronal activity. The ability of astrocytes to remove from the extracellular space around active synapses K + ions increasingly concentrated there following high neuronal activity and to redistribute them, through the syncytium, to distal regions, was originally considered a plausible mechanism to couple neuronal activity with dilation of vessels. Astrocytes can increase blood flow in response to neuronal activity by releasing K + ions from their endfeet apposed to arterioles, because modest increases in extracellular K + concentration (up to ~10 mm [K + ] o ) hyperpolarize smooth muscle cells. But more recent evidence suggests that the control of microcirculation by astrocytes is not based simply on the spatial buffering of K + hypothesis, but rather involves a more complex mechanism and a number of different molecules. In hippocampal and cortical slices glutamate released at active synapses triggers Ca 2+ oscillations in astrocytes that increase in frequency according to increasing levels of neuronal activity. While this observation demonstrates that astrocytes are sophisticated sensors of neuronal activity, it also represents a clue to the possibility that astrocytes transfer to blood vessels information on the level of neuronal activity. Indeed, neuronal activity-dependent Ca 2+ elevations in astrocytes were observed to propagate to perivascular endfeet. Such a signal provides a mechanistic basis for the graded response of the blood flow to different levels of neuronal activity, thereby strengthening the idea of a distinct astrocytic role in neurovascular coupling. Importantly, high-frequency stimulation of neuronal afferents was found to trigger both Ca 2+ elevations in astrocyte endfeet and dilation of cerebral arterioles. When neuronal activity releases glutamate at synapses, some of the released glutamate escapes the synaptic cleft and activates metabotropic glutamate receptors (mglurs) (see below) on astrocytes, thus increasing [Ca 2+ ] i in astrocytes. Furthermore, Ca 2+ elevations triggered in astrocytes by either a metabotropic glutamate receptor (mglur) agonist, or direct mechanical stimulation of individual astrocytes by a patch pipette, also evoked dilation of cortical arterioles, while inhibition by mglur antagonists of Ca 2+ oscillations evoked in astrocytes by synaptic glutamate, or the incubation with cyclooxygenase (COX) inhibitors that block prostaglandin synthesis, reduced neuronal activity-dependent dilation of cerebral arterioles. However, the situation is more complicated. In hippocampal slices, Ca 2+ elevations in astrocyte endfeet have been observed to evoke also arteriole constriction. Thus, emerging evidence implicates an increase in intracellular Ca 2+ concentrations ([Ca 2+ ]i) in astrocytes in either vasodilation or vasoconstriction, depending on the nature of the signal that triggered the [Ca 2+ ]i increase. According to these results, a model is proposed in which astrocytes can encode different levels of neuronal activity into defined Ca 2+ oscillation frequencies that, at the level of perivascular endfeet, mediate the release of dilating agents as well as constrictive agents. There is strong evidence that astrocytes can control blood flow through the production and release of metabolites of arachidonic acid (AA). When glutamate released from neurons activates astrocyte mglurs, the resultant rise in [Ca 2+ ] i activates phospholipase A2 (PLA2), evoking the production of AA from membrane phospholipids. AA can cause vasodilatory and vasoconstrictive actions, since the stimulation of astrocytes in culture with glutamate receptor agonists triggers the formation of various AA metabolites with different vasoactive properties. The build-up of AA leads, for instance, to the production of several cyclooxygenase (COX)-dependent prostaglandins (PGs), such as PGD 2, PGE 2, and PGI 2, which dilate nearby arterioles, and other PGs having vasoconstrictive activity (PGF 2α, TxA 2 ). 18

19 Through the astrocyte cytochrome P-450 epoxygenase CYP2C, AA can form also epoxyeicosatrienoic acids (EETs). EETs hyperpolarizes smooth muscle cells and triggers dilation of cerebral vessels; the pharmacological inhibition of P-450 epoxygenase results in reduction of the basal blood flow in the cerebral cortex. EETs vasodilate by activating the smooth muscle cell largeconductance Ca 2+ -activated and voltage-dependent K + (BK Ca ) channel. On the other hand, the diffusion of AA to the smooth muscle may lead to the accumulation of 20- hydroxyeicosatetraenoic acid (20-HETE) that causes vasoconstriction. 20-HETE, that derives from ω-hydroxylation of AA by CYP4A, a cytochrome P-450 enzyme subtype, depolarizes smooth muscle cells by inhibiting the opening of the smooth muscle BK Ca channel, and also enhances Ca 2+ influx through voltage-dependent Ca 2+ channels. Under normal physiological conditions 20-HETE has been proposed to play a crucial role in the maintenance of myogenic tone in cerebral blood vessels. The constrictive action of 20-HETE may thus control, together with that of dilating agents, the extent of neuronal activity-dependent increases in blood flow. This action of 20-HETE can also account for the constriction of cerebral blood vessels associated with ischemia, since Ca 2+ elevations and Ca 2+ waves are known to occur in the astrocytes during this pathological brain condition, increasing the BK Ca channels opening. PGE 2 can relax vascular smooth muscle by binding to prostaglandin receptors which increase the activation of protein kinase A (PKA) by camp, whereas other prostaglandins acting on different receptors may cause vasoconstriction (such as PGF 2α and TxA 2 ) by increasing Ca 2+. In smooth muscle (and pericytes) Ca 2+ binds to calmodulin and activates the myosin light chain kinase (MLCK), leading to phosphorylation of the myosin light chain (MLC). Phosphorylation of MLC allows the myosin ATPase to be activated by actin and the muscle to contract. On the other hand, relaxation mediated by an increase in intracellular camp or cgmp is often associated with a decrease in Ca 2+ sensitivity. PKA phosphorylates MLCK at two sites, site A and B. Phosphorylation of site A but not site B decreases the affinity of MLCK for the Ca 2+ /calmodulin complex. Thus, via PKA, camp increases the phosphorylation of MLCK (which in this way is inactivated) and thus decreases the phosphorylation of MLC (inhibiting the smooth muscle contraction). It is possible that other mechanisms are involved. The hyperpolarization caused by K + syphoning (see above) could reduce the influx of Ca 2+ in vascular smooth muscle cells through voltage-gated channels and could dilate the vessels. However, astrocytes may dilate vessels through a different K + -based mechanism. The increase in [Ca 2+ ] i caused by mglurs activation has been reported to lead to the opening of large-conductance Ca 2+ - activated K + (BK Ca ) channels in astrocyte endfeet, releasing K + onto vessels. Astrocytes produce other vasoactive substances, including also nitric oxide (NO), carbon monoxide (CO) and ATP. Synaptic release of glutamate activates neuronal N-methyl-D-aspartate receptors (NMDARs, see below), resulting in Ca 2+ entry into neurons and activation of neuronal nitric oxide synthase (nnos). This releases NO, which dilates vessels. In the cortex, inhibition of nnos reduces the increase in blood flow that is associated with neural activity, suggesting a role for NO in neurovascular coupling. The action of NO is complex, since NO produced by both neurons and glia has been reported to downregulate the formation of the cytochrome P-450 subtypes CYP2 that produce EETs as well as to inhibit 20-HETE formation. Furthermore, endothelial nitric oxide synthase (enos) can be activated by flow-induced shear stress or by acetylcholine (ACh). NO stimulates the guanylyl cyclase, increasing cyclic GMP, which stimulates the cyclic GMPdependent protein kinase (PKG). PKG causes smooth-muscle relaxation by mechanisms that are still being defined and that might include a reduction in cytosolic Ca 2+ (by enhanced Ca 2+ export and/or by reduced inositol trisphosphate receptor-mediated Ca 2+ mobilization) and 19

20 dephosphorylation of myosin light chains (by activation of MLC phosphatase and/or by sequestration of MLCK in a phosphorylated form that is not readily activated by Ca 2+ /CaM). CO is produced by heme oxygenase (HO)-catalyzed heme degradation. HO has two active isoforms, HO-1 (inducible) and HO-2 (constitutive). Both HO-1 and HO-2 are membrane-bound proteins, anchored to the endoplasmic reticulum membrane. HO-2 is highly expressed in endothelial and smooth muscle cells and in adjacent astrocytes in the brain. Overall, HO activity of the brain tissue exceeds that of systemic organs, including even the liver and spleen. In response to various physiological/pathophysiological stimulations, including glutamate, seizures, hypoxia, hypotension, and ADP, HO-2 activity in vivo is rapidly increased by Ca 2+ without altering HO-2 expression. CO dilates cerebral arterioles: the inhibition of HO attenuates the dilation induced by these stimuli. CO dilates by activating smooth muscle cell BK Ca channels, which are functionally a heme protein. CO, by binding to the channel-bound reduced heme and other channel locations, changes the association of heme with the channel, leading to channel activation. Thus BK Ca channel-bound heme is a receptor for CO, and CO binding increases BK Ca channel Ca 2+ sensitivity. The resulting smooth muscle cell hyperpolarization decreases [Ca 2+ ] i, leading to vasodilation. The predominant cell type that causes CO production and cerebral arteriole dilation to topical glutamate is the astrocyte rather than the endothelial cell. CO-induced cell signaling has been proposed to occur via guanylyl cyclase and PKG activation, but CO is far less effective at activating guanylyl cyclase than is NO. Thus, cgmp as a direct mediator of CO-induced vasodilation under physiological conditions appears unlikely. In cerebral arterioles in vivo, in contrast to dilation to acute CO, a prolonged exposure of cerebral arterioles to elevated CO produces progressive constriction by inhibiting nitric oxide synthase. All together, these observations hint at the possibility that astrocytes release both dilating and constrictive agents. While these two opposing actions seem in conflict, since both have been seen to occur in vivo, the challenge is to identify the conditions that select for the respective actions. Whether a rise in astrocyte [Ca 2+ ] i causes dilation or constriction may in part be determined by the pre-existing tone of the vessel, but the O 2 concentration also has a key role in determining this. The concentration of O 2 in the extracellular space is μm. Thus, at in vivo levels of O 2, we expect NO and 20-HETE synthesis to be significantly limited by the amount of O 2 available, whereas the enzymes synthesizing EETs and prostaglandins are allowed to operate. Two other factors were also shown to contribute significantly to the effects of lowering [O 2 ]. As O 2 concentrations decrease, the lack of energy for ATP synthesis causes an increase in the level of extracellular adenosine, which binds to adenosine A 2A receptors on vascular smooth muscle cells to depress vessel constriction, by increasing camp. In addition, a decrease in the rate of oxidative phosphorylation relative to the rate of glycolysis results in lactate production. Monocarboxylate transporters release the lactate into the extracellular space, where it reduces the clearance of extracellular PGE 2 by the prostaglandin transporter PGT. Thus, when PGE 2 is released from astrocytes, extracellular PGE 2 increases to a greater degree, resulting in larger arteriole dilations. This effect of lactate may partly explain why, in humans and rats in vivo, cerebral blood flow is regulated by the lactate/pyruvate concentration ratio. Interestingly, lactate is released into the extracellular space during synaptic activity (see above), which should promote vasodilation. These observations confirm that the mechanism that governs the blood flow response to neuronal activity is complex and relies probably on different vasoactive agents in different brain regions, and they underline the central role of astrocytes in functional hyperemia. The important action of astrocytes in the control of microvasculature raises also the possibility that an astrocyte dysfunction could be implicated in the dysregulation of cerebral circulation in brain pathologies, for example, in the defective neurovascular coupling that is associated with Alzheimer s disease, as well as in the vascular abnormal responses during stroke, trauma, and spreading depression. 20