Glutamate and Glutamine in Brain Disorders Lasse K. Bak, Arne Schousboe, and Helle S. Waagepetersen Abstract Diseases of the brain account for much human suffering and place a huge burden on the health care systems. Thus, research into the pathology of brain diseases and improved pharmacotherapy is of significant value. In this respect, knowledge on malfunctions of metabolic homeostasis related to the neurotransmis- sion process is still limited. As evident from this chapter, failure of the metabolic homeostasis of the two amino acids of major importance, namely glutamate and glu- tamine, is a hallmark of a wide range of both neurological and psychiatric diseases. This chapter deals with representative brain diseases as well as the methodology of research related to metabolism. In addition, future need for research and potential new targets for pharmacotherapy are discussed. Keywords Epilepsy · Glutamate–glutamine cycle · Ischemia · Metabolism · Neurodegenerative disorder · NMRS · Psychiatric disorder Abbreviations EAAT Excitatory amino acid transporter GABA γ-Aminobutyric acid GDH Glutamate dehydrogenase GS Glutamine synthetase MCAO Middle cerebral artery occlusion MS Mass spectrometry MSA Multiple system atrophy NMDA N-methyl-D-aspartate NMRS Nuclear magnetic resonance spectroscopy 3-NPA 3-Nitropropionic acid PAG Phosphate-activated glutaminase A. Schousboe (B) Faculty of Pharmaceutical Sciences, Department of Pharmacology and Pharmacotherapy, University of Copenhagen, DK-2100, Copenhagen, Denmark e-mail: as@farma.ku.dk 195 J.P. Blass (ed.), Neurochemical Mechanisms in Disease, Advances in Neurobiology 1, DOI 10.1007/978-1-4419-7104-3_7, C Springer Science+Business Media, LLC 2011 196 L.K. Bak et al. TCA Tricarboxylic acid VGLUT Vesicular glutamate transporter Contents 1 Introduction 196 2 Methodological Approaches in the Study of Glutamate and Glutamine Homeostasis in the Brain 198 3 Glutamate and Glutamine Homeostasis in Selected Brain Disorders 199 3.1 Epilepsy 199 3.2 Ischemic Conditions 201 3.3 Neurodegenerative Disorders 203 3.4 Psychiatric Disorders 205 4 Potential Drug Targets Related to Glutamate and Glutamine Homeostasis 207 5 Concluding Remarks 207 References 208 1 Introduction Together, neurological and psychiatric disorders account for the vast majority of suffering from chronic illnesses (Cowan and Kandel, 2001). Thus, such disorders are important to understand at all levels, ranging from primary care of patients to the cellular and molecular levels. The focus of this chapter is on metabolic aspects of brain disorders related to malfunction of glutamate and glutamine homeostasis. The following sections deal with representative pathologies discussing selected recent as well as future need for research and potential targets for drug treatment. Apart from being one of 20 amino acids used for protein synthesis, glutamate is the most abundant excitatory neurotransmitter in the mammalian brain. In addi- tion, glutamate is the immediate precursor for γ-aminobutyric acid (GABA), the most abundant inhibitory neurotransmitter. Moreover, glutamate serves an important function in intermediary metabolism as donor of the amino group in transaminations of α-keto acids to form other α-amino acids. The cognate keto acid of glutamate, α- ketoglutarate, is a tricarboxylic acid (TCA) cycle intermediate and glutamate can be oxidized in the TCA cycle thereby acting as an energy substrate. Glutamine, pro- duced from glutamate via the astrocytic glutamine synthetase (GS) reaction serves as a precursor for neuronal transmitter glutamate (see below). In addition, the GS reaction is considered the most important reaction for fixating ammonia in the brain playing a fundamental role for detoxifying ammonia in hyperammonemic condi- tions. Paradoxically, glutamate is a potent excitotoxin as well which means that tight homeostatic control of glutamate is of vital importance. This involves a num- ber of both cytosolic and mitochondrial enzymes as well as transporters located in the plasma, vesicular, and mitochondrial membranes (see Waagepetersen et al., 2007). Glutamate and Glutamine in Brain Disorders 197 Fig. 1 Schematic representation of the glutamate/GABA-glutamine cycle. See text for details. The present dogma dictates that all (or most) of the released glutamate (Glu) is taken up into the glial compartment, whereas released GABA is primarily accumulated into the presynaptic neuron by neuronal reuptake (as indicated by the size of the arrows); this view may change in the future. Abbreviations: GABA, γ-aminobutyric acid; GABA-T, GABA transaminase; GAD, glutamate decarboxylase; Glu, glutamate; Gln, glutamine; GS, glutamine synthetase; α-KG, α-ketoglutarate; PAG, phosphate-activated glutaminase; SSADH, succinate semialdehyde dehydrogenase; TCA, tricarboxylic acid For the following discussions, one vital component of glutamate homeostasis is particularly important to bring to mind, namely the glutamate–glutamine cycle. A more comprehensive discussion of brain glutamate and glutamine homeostasis is available elsewhere, therefore only a brief introduction is provided here (Bak et al., 2006; Waagepetersen et al., 2007). Based on discoveries of intercellular compart- mentalization of glutamine and glutamate pools, related to astrocytes and neurons, respectively, a glutamate–glutamine cycle working between neurons and astrocytes was suggested more than three decades ago (van den Berg and Garfinkel, 1971; Benjamin and Quastel, 1972; Berl and Clarke, 1983; Ottersen et al., 1992). The cycle for a glutamatergic synapse is outlined in Fig. 1 (left part) in which released neurotransmitter is t aken up into surrounding astrocytes, transformed into glutamine by the astrocyte-specific enzyme GS (Norenberg and Martinez-Hernandez, 1979) and released into the extracellular space from which it is taken up into neurons and transformed back to glutamate by phosphate-activated glutaminase (PAG; Kvamme et al., 2001). In the GABAergic synapse (Fig. 1; right part), GABA is taken up into astrocytes and catabolized to the TCA cycle intermediate succinate via the con- certed action of GABA transaminase and succinate semialdehyde dehydrogenase. 198 L.K. Bak et al. Glutamine may be synthesized from succinate via the TCA cycle including conden- sation of oxaloacetate and acetyl-CoA forming citrate and subsequent synthesis of α-ketoglutarate and conversion to glutamate. Glutamate formed by PAG activity in the GABAergic neurons is the precursor for GABA catalyzed by glutamate decar- boxylase. This fundamental neuronal–astrocytic interaction seems to be affected in many pathological conditions, as is shown in the following sections. 2 Methodological Approaches in the Study of Glutamate and Glutamine Homeostasis in the Brain Animal models of brain disorders are typically generated by pharmacological treat- ment, surgical procedures, or more recently by techniques of molecular biology. Such treatments might give rise to symptoms or pathological changes consistent with those observed in humans suffering from a given brain disease. It should be noted that many human diseases are not naturally observed in rodents, the animals of choice for many disease models. Still, these animal models constitute a vital tool for studying malfunctions and potential treatments of a wide range of brain disorders. Although primary cell cultures of neurons and astrocytes constitute a valu- able tool for studying cell-specific metabolism, their use is limited in the context of studying brain pathology as they do not constitute a native biological sys- tem. For this reason, most work done using cell cultures has been focused on mimicking pathologies such as ischemia (hypoxia and glucose-deprivation), a con- dition fairly easy to create experimentally. However, pathological conditions with a more complex pathogenesis, such as epilepsy, have typically been studied in more intact systems, for example, in vivo or ex vivo studies on animal models or acute preparations of native brain tissue. The techniques used to study metabolism in brain (pathology) include labeling of precursors with either radioactive (14C, 3H) or stable isotopes (13C, 15 N) and subsequently analyzing the incorporation of these isotopes into metabolites under different experimental conditions. The use of radio-labeled precursors is a classical technique still widely used; however, even though still a valid method, the amount of information gained using radio-labeled precursors is limited compared to employ- ing stable isotopes in combination with more advanced analytical techniques such as nuclear magnetic resonance spectroscopy (NMRS) or mass spectrometry (MS). The major difference between NMRS and MS is that the NMRS experiment reveals the location of the labeled atoms within the molecule of interest whereas MS data pro- vide information regarding only the number of labeled atoms in a given molecule. However, MS analysis is much more sensitive, faster, and inexpensive compared to NMRS. No doubt the most valuable in vivo technique for studying metabolism is NMRS. Although in vivo NMRS is a powerful technique providing real-time data, it is hampered by high cost, problems with sensitivity, low resolution, the need for anes- thetizing the animal (of course this is not necessary for human subjects), and lack Glutamate and Glutamine in Brain Disorders 199 of specific methods for differentiating between neuronal and glial compartments. Thus, in vivo NMRS relies on a number of assumptions as well as results from in vitro (cell culture) work combined with mathematical modeling to make sense of the data obtained. However, despite these shortcomings, refinement of this tech- nique promises to bring about an increasing amount of valuable data in the coming decades. To date, the best way to probe glial and neuronal metabolism is the use of labeled glucose and acetate. The underlying principle is that acetate is specifically taken up and metabolized by astrocytes because of specific uptake into this cellular compart- ment (Waniewski and Martin, 1998). It is then presumed that oxidative metabolism of glucose primarily takes place in neurons (Taylor et al., 1996; Qu et al., 2000). A number of studies take advantage of this by employing differentially labeled glu- cose and acetate. Thus, combined injection of [1-13C]glucose and [1,2-13C]acetate into an experimental animal produces different labeling patterns in metabolites, that is, mono-labeled from [1-13C]glucose (neuronal compartment) and double-labeled from [1,2-13C]acetate (astrocytic compartment; see Fig. 2 for detailed explanation). 3 Glutamate and Glutamine Homeostasis in Selected Brain Disorders 3.1 Epilepsy More than 13 decades ago, an epileptic seizure was defined as an occasional, sud- den, and excessive discharge of grey matter (Jackson, 1873). However, seizures are merely symptoms of underlying brain pathologies although in most cases no causative brain disorder can be identified and a purely descriptive diagnosis is made. Current drugs for treating epilepsy target either sodium channels or neurotransmitter metabolic enzymes, transporters, or receptors; although the mechanisms of action are not always clear, most drugs seem to either inhibit the glutamatergic (excitatory) system or potentiate the GABAergic (inhibitory) system. The concept of regarding epilepsy as an imbalance between excitation and inhibition seems to implicate that glutamatergic or GABAergic systems may be involved in the pathology. Indeed, many studies have revealed that the levels of several amino acids including gluta- mate, GABA, and glutamine are altered in epilepsy in animal models as well as in humans. Biopsies taken from human subjects suffering from temporal lobe epilepsy revealed decreased glutamate–glutamine cycling in sclerotic hippocampal tissue, as evidenced by labeling in glutamate and glutamine from [2-13C]glucose infused prior to resection of the hippocampal tissue (Petroff et al., 2002). In fact, glutamate– glutamine cycling seems to be a mechanism commonly affected in epilepsy. Accordingly, inhibiting astrocytic GS (and thus glutamine transfer to neurons for transmitter glutamate synthesis) in hippocampal neuronal/astrocytic cocultures or hippocampal slices (in which epileptiform activity was induced by GABAA 200 L.K. Bak et al. Fig. 2 Schematic representation of synapses showing how differentially labeled [1-13C]glucose ([1-13C]Glc) and [1,2-13C]acetate produces different labeling patterns in major metabolites. [1,2-13C]Acetate is specifically taken up into astrocytes (see text for references) and metab- olized in the TCA cycle to α-[4,5-13C]ketoglutarate (α-[4,5-13C]KG) which is converted to [4,5-13C]glutamate ([4,5-13C]Glu) and [4,5-13C]glutamine ([4,5-13C]Gln) which is subsequently transferred to neurons. In neurons, [4,5-13C]glutamine is converted to [4,5-13C]glutamate and subsequently to [1,2-13C]GABA in GABAergic neurons. [1-13C]glucose is primarily metabolized in the neuronal compartment (see text for references). Glycolytic processing of [1-13C]glucose leads to [3-13C]pyruvate ([3-13C]Pyr) which is metabolized to α-[4-13C]ketoglutarate and sub- sequently to [4-13C]glutamate and [2-13C]GABA in GABAergic neurons. The depicted scheme is the simplest possible labeling patterns produced when α-[4,5-13C]ketoglutarate leaves the TCA cycle in the first turn. In addition, [4,5-13C]glutamate formed from [4,5-13C]glutamine in the neuronal compartment may to a large degree be converted to α-[4,5-13C]ketoglutarate and metab- olized in the TCA cycle thus forming additional labeling patterns in glutamate and GABA. Such cycling of astrocyte-derived glutamine has been suggested to be substantial in GABAergic neurons (Waagepetersen et al., 1999) receptor block) reduced spontaneous epileptiform spiking activity, indicating that the reduced flow of glutamine to the neurons reduced neuronal activity (Bacci et al., 2002). The same effect on spike activity was observed in cultured hip- pocampal neurons in which glutamine transport had been blocked (Bacci et al., 2002). Furthermore, intraperitoneal injection of leucine or its cognate keto acid, α-ketoisocaproate, augmented the occurrence of absence seizures in genetic absence epilepsy rats from Strasburg (GAERS; Dufour et al., 2001a, b). This was argued to be mediated by a decrease in the amount of glutamate available for neurotransmis- sion, which may correlate to seizure activity in this animal model of nonconvulsive absence epilepsy (Danober et al., 1998). However, a more direct approach to study the metabolic disturbances involved in cortical and thalamic brain regions in the GAERS model was performed by combined injection of [1-13C]glucose and Glutamate and Glutamine in Brain Disorders 201 [1,2-13C]acetate, showing increased glutamate–glutamine cycling in the cortex but not in the thalamus in conjunction with a decreased amount of cortical GABA (Melø et al., 2006). Hence, increased glutamatergic input from the cortex to the thalamus may affect thalamic filter function, thus playing a role in inducing absence seizures in the GAERS model. In general, epilepsy animal models show metabolic disturbances of both neu- rons and astrocytes. However, it has been argued that the initial or primary change might take place in only one of these cell types (e.g., Sonnewald and Kondziella, 2003; Kondziella et al., 2003). A recent study in a rat model of lithium–pilocarpine-induced temporal lobe epilepsy showed the same extent of [1,2-13C]acetate metabolism in controls as in epileptic animals, implying that astro- cytic metabolism is not compromised in these animals (Melø et al., 2005). However, glutamate labeling from [1-13C]glucose was reduced, suggesting that the metabolic malfunction in this epileptic model is in the neuronal compartment. Although glutamate and glutamine metabolism is certainly affected in epileptic conditions, it does not seem clear whether this is what is causing the seizure activ- ity or vice versa; basically a chicken and egg dilemma that may be best resolved by mapping the pathogenesis taking all initial and persistent changes into account. However, knowledge of metabolic malfunctions is very useful in the context of developing novel drugs for symptomatic treatment of epileptic disorders that target metabolic changes (see Section 4 for a further discussion on this matter). 3.2 Ischemic Conditions Cerebral ischemia, (i.e., absolute or relative shortage of blood supply to a part of the brain such as following an ischemic stroke) is a serious condition that can lead to physical impairment or death. Stroke is a leading cause of death and adult disability in the industrialized part of the world (Thom et al., 2006). The outcome of a cerebral ischemic episode is greatly influenced by the duration, that is, the time from onset of ischemia until reperfusion; treatment is directed at the cause of the impaired blood supply, that is, thrombolysis or surgical intervention (e.g., see the review by Juttler et al., 2006). The cellular and molecular events taking place during an ischemic episode have been studied extensively in animal models as well as in tissue preparations and cell cultures. The lack of oxygen and glucose causes a dramatic chain of events ultimately leading to cell death and necrosis of the affected tissue. It is gener- ally thought that excitotoxic insults are key elements in the pathology, as neurons might succumb to excitotoxic mechanisms rather than energy deprivation per se, especially in the penumbral zone (Huang et al., 1997). One of the initial events in ischemia may be impairment or reversal of astrocytic glutamate uptake ( Rossi et al., 2000), resulting in increased extracellular levels of glutamate and concomitant dysfunction of the glutamate–glutamine cycle. Interestingly, the eventual demise of astrocytes in ischemia might be due to lack of intracellular glutamate (caused by a reversal of the glutamate transporter) for synthesis of glutathione and associated 202 L.K. Bak et al. oxidative damage, a mechanism possibly involving the lipoxygenase pathway (Re et al., 2006). However, it is clear from cell culture studies that astrocytes are much more resistant to deprivation of oxygen and glucose than neurons (e.g., Almeida et al., 2002). A widely employed rodent model of cerebral ischemia is middle cere- bral artery occlusion (MCAO), a surgical procedure in which the middle cerebral artery is temporarily or permanently occluded resulting in infarction of only one hemisphere (Longa et al., 1989). As discussed below, the MCAO model or variations thereof have been employed in a number of studies. These studies indicate mal- functions in cellular metabolism and impairment of neuronal–astrocytic interactions both during the ischemic episode and following reperfusion. After 1 h of reperfusion following 2 or 3 h of MCAO in conscious rats, regional as well as cellular differences were found with regard to incorporation of 14C into glutamate and glutamine from [14C]glucose or [14C]acetate (Thoren et al., 2005, 2006). It was found that labeling of glutamine from [14C]acetate was reduced com- pared to the nonaffected hemisphere in striatum, but not in focal or perifocal cortical tissue (Thoren et al., 2005). This signifies the regional diversity in susceptibility to ischemic insults apparently based on differences in properties of glial cells. In contrast, when [14C]glucose was employed, labeling in both glutamate and glu- tamine decreased in both striatum and cortex (Thoren et al., 2006). This indicates that neuronal glucose metabolism in both regions is significantly affected which is consistent with a reported decrease in 2-deoxy-glucose utilization in a similar model (Belayev et al., 1997). Surprisingly, for both brain regions the ATP: ADP ratio and phosphocreatine level were found to be maintained following reperfusion; however, increases in [14C]lactate was manifest throughout both regions (Thoren et al., 2006). The cellular origin of this lactate may be both neuronal and astro- cytic, probably reflecting increased anaerobic glycolysis secondary to lack of O 2 for oxidative metabolism. It should be noted that astrocytes contain the only supply of glycogen in the brain, although it is of limited size (McKenna et al., 2006). Glycogen may support anaer- obic glycolysis in astrocytes during ischemia producing lactate which is released into the extracellular space; however, the functional role of brain glycogen during normal or ischemic conditions is not known in detail (McKenna et al., 2006). In this respect, lactate has been suggested to function as an (obligatory) energy substrate for neurons recovering from hypoxia or aglycemia (Fowler, 1993; Schurr et al., 1997a, b, c; Cater et al., 2001), although these observations have been argued to be an artifact of the preparation procedures of the experimental models employed (Okada and Lipton, 2007). In line with this, exogenous pyruvate may serve a neuro- protective role in the postischemic brain (e.g., Desagher et al., 1997) being primarily a neuronal substrate as evidenced in mice employing injections of [3-13C]pyruvate (Gonzalez et al., 2005); however, as noted by these authors, the clinical use may be hampered by the risk of seizures induced by high doses of exogenous pyruvate. In a series of studies in rats by Håberg et al. (1998, 2001, 2006), different time periods of MCAO (ranging from 30 to 240 min) were studied ex vivo employ- ing injection of [1-13C]glucose and [1,2-13C]acetate and subsequent analysis by NMRS. Already in the early stage of ischemia (30 min of MCAO) astrocytic Glutamate and Glutamine in Brain Disorders 203 metabolism was compromised in the ischemic core (lateral caudoputamen and lower parietal cortex) as evidenced by decreased synthesis of [4,5-13C]glutamine from [1,2-13C]acetate (Håberg et al., 2001). In addition, decreased neuronal formation of [4,5-13C]glutamate from the [4,5-13C]glutamine formed in astrocytes was evident, suggesting impaired transfer of glutamine from astrocytes to neurons (Håberg et al., 2001). Even though neurons can maintain glutamate synthesis throughout 240 min of MCAO, the total amount of glutamate in the affected tissue is decreased after an ischemic episode (Håberg et al., 2001), underlining the serious metabolic deficien- cies occurring under such conditions. Interestingly, [1,2-13C]GABA was already absent from ischemic brain regions after 30 min of MCAO (Håberg et al., 2001). This may indicate that (i) at the GABAergic synapse, astrocyte-to-neuron trafficking of [4,5-13C]glutamine was completely abolished or (ii) that the [4,5-13C]glutamate derived from [4,5-13C]glutamine is metabolized in the neuronal TCA cycle rather than being used directly for synthesis of GABA. As indicated above, primary astrocytic malfunction may cause subsequent harm to the neurons but certainly astrocytes play an important role for neuronal sur- vival following reperfusion as well. In rats, after 120 min of MCAO followed by 120 min of reperfusion interesting metabolic differences were found between the ischemic core (lateral caudoputamen and lower parietal cortex) and the penumbral zone (frontoparietal cortex; Håberg et al., 2006). Following reperfusion, astro- cytic metabolism was significantly improved in the penumbral zone, as evidenced by increased metabolism of [1,2-13C]acetate whereas in the ischemic core the situation was the reverse, showing decreased [1,2-13C]acetate metabolism com- pared to 120 min of MCAO alone (Håberg et al., 2006); this was most likely caused by decreased activity of acetyl-CoA synthetase, catalyzing the conversion of [1,2-13C]acetate to [1,2-13C]acetyl-CoA. In the penumbra, label from [1- 13C]glucose into [4-13C]glutamate was not affected by 120 min reperfusion compared to after 120 min of MCAO alone, which was in contrast to the ischemic core where almost no [4-13C]glutamate was present at this point (Håberg et al., 2006). This suggests that after 120 min of MCAO, neuronal metabolism in the ischemic core is not rescued by 120 min of reperfusion whereas neuronal integrity is preserved in the penumbral zone. The MCAO animal model seems to be a robust system in which to investigate the metabolic changes occurring in ischemia. Not surprising, rather severe dysfunc- tional homeostasis of glutamate and glutamine metabolism seem to be present in ischemia, including interference with neuronal–astrocytic interactions. However, the severity of these dysfunctions seemed to vary among brain regions, possibly due to differential properties of the local glial cells. 3.3 Neurodegenerative Disorders Neurodegeneration is a progressive, fatal deterioration of neuronal function, playing a part in many brain pathologies including Alzheimer’s, Huntington’s, and Parkinson’s diseases and olivopontocerebellar atrophy, which are commonly 204 L.K. Bak et al. referred to as neurodegenerative diseases. Olivopontocerebellar atrophy and Huntington’s disease are part of the group of trinucleotide repeat or polyglutamine disorders, featuring excessive repeats of genomic CAG sequences on diff erent chromosomes affecting gene function. Generally, only few studies have been con- ducted on glutamate and glutamine metabolism in neurodegenerative disorders. Some studies have determined levels of amino acids in blood and brain or cere- brospinal fluid (CSF) and found some differences; however, levels of metabolites only offer limited information on dysfunctional metabolism. Interestingly, there seems to be an increased interest in employing NMRS in studying these pathologies and representative studies are discussed below. Glutamate metabolism and glutamate–glutamine cycling might be compromised in Alzheimer’s disease, inasmuch as aberrant expression of glutamate metaboliz- ing enzymes including glutamate dehydrogenase (GDH), GS, and PAG has been reported (Robinson, 2000, 2001; Burbaeva et al., 2005). In addition, a study employing in vivo NMR spectroscopy suggested decreased glutamatergic neuro- transmission activity and TCA cycling rate in patients suffering from Alzheimer’s disease, as suggested by labeling patterns in glutamate and glutamine from infused [1-13C]glucose (Lin et al., 2003). Neurodegeneration related to GDH malfunction and subsequent glutamate toxic- ity may constitute part of the pathology of some forms of multiple system atrophies (MSAs) known as olivopontocerebellar atrophy or spinocerebellar ataxia affecting the cerebellum, pons, and inferior olives (review, Plaitakis et al., 1993); symptoms include Parkinsonism, ataxia, and autonomic dysfunction. GDH deficiencies were found in patients suffering from cerebellar degeneration (Plaitakis et al., 1979, 1980) and abnormal levels of brain glutamate have been observed in patients suffering from MSA (Perry et al., 1981). In addition, brain TCA cycle metabolism may also be compromised as the activity of α-ketoglutarate dehydrogenase, the rate-limiting enzyme in the TCA cycle, seems to be decreased in postmortem biopsies from MSA patients (Mastrogiacomo and Kish, 1994). Apparently, no studies employing stable isotope-labeled substrates and NMRS or MS seem to have been performed in either MSA patients or any animal models of cerebellar degeneration. One problem may be that no valid animal model is available that displays the right combination of pathological changes compared to patients suffering from GDH-deficient forms of MSA (Miret-Duvaux et al., 1990), although systemic administration to rodents of 3-acetylpyridine has been argued to produce lesions corresponding to those observed in patients (see the review by Plaitakis et al., 1993). The elucidation of metabolic defects in MSAs would be of significant value for understanding these debilitating neurodegenerative diseases. The levels of glutamine and glutamate were found to increase and decrease, respectively, in CSF of patients suffering from Parkinson’s disease (Mally et al., 1997), however, other studies employing in vivo NMRS found that the [gluta- mate+glutamine]/[creatine] ratio in the basal ganglia and striatum of Parkinson’s disease patients was no different from healthy control s ubjects (Clarke et al., 1997; Taylor-Robinson et al., 1999). On the other hand, a study employing 1-methyl-4- phenyl-1,2,3,6-tetrahydropyridine (MPTP; a drug found to induce striatal lesions . regions in the GAERS model was performed by combined injection of [1-13C]glucose and Glutamate and Glutamine in Brain Disorders 201 [1,2-13C]acetate, showing increased glutamate–glutamine cycling in. glutamine are altered in epilepsy in animal models as well as in humans. Biopsies taken from human subjects suffering from temporal lobe epilepsy revealed decreased glutamate–glutamine cycling in. transporter Contents 1 Introduction 196 2 Methodological Approaches in the Study of Glutamate and Glutamine Homeostasis in the Brain 198 3 Glutamate and Glutamine Homeostasis in Selected Brain Disorders