Neurochemistry Of Gaba

SYNTHESIS AND CATABOLISM OF GABA

The synthesis and metabolism of GABA is closely linked with that of glutamate and the citric acid or tricarboxylic acid (TCA) cycle (Fig. 11.1). GABA is produced by the decarboxylation of glutamate, a reaction catalysed by the enzyme glutamic acid decarboxylase (GAD). GAD is found in several non-neuronal tissues (including ovary and pancreas) but within the CNS it is a specific marker of GABAergic neurons, where it is present in the cytoplasm as both soluble and membrane-bound forms, principally in the axon terminals. Labelling with antibodies against GAD has thus proved a particularly valuable technique for the identification of these neurons and their synaptic boutons. The breakdown of GABA occurs as a transamination reaction catalysed by the mitochondrial enzyme 4-aminobutyrate aminotransferase (GABA transaminase; GABA-T). In this process the amino group from GABA is transferred onto the TCA cycle intermediate a-ketoglutarate, producing glutamate and succinic semialdehyde. The latter is in turn converted by the enzyme succinic semialdehyde dehydrogenase (SSADH) into succinate, which re-enters the TCA cycle. This synthesis and catabolism of GABA is often referred to as the 'GABA-shunt', as it acts as a shunt of the normal TCA pathway from a-ketoglutarate to succinate. Other potential routes of GABA production have been described — involving deamination and decarboxylation reactions from putrescine, spermine, spermidine and ornithine — but the vast majority of GABA is generated by means of the GABA-shunt.

GABA-T and SSADH are also present in the mitochondria of glial cells and are responsible for the degradation of GABA recovered from the extracellular space (see below). In this case the glutamate formed from the action of GABA-T is converted into glutamine by the cytosolic enzyme glutamine synthetase (GS). Glial glutamine serves as an important precursor for both neuronal glutamate and GABA. It is transported from glia into neurons where the mitochondrial enzyme phosphate-activated glutaminase (PAG) converts it back into glutamate. This neuronal glutamate can then be converted to GABA, either directly or following metabolism via the TCA cycle. The interconversion of glutamate and a-ketoglutarate is achieved by two further groups of

Brain Neurochemistry

Figure 11.1 Enzymes responsible for the synthesis and metabolism of GABA. Enzymes responsible for the syntheisis (GAD) and metabolism (GABA-T and SSADH) of GABA, and their relationship to the TCA cycle and the amino acids glutamate and glutamine. Precursor glutamate is derived from glutamine by phosphate-activated glutaminase (PAG) and from a-ketoglutarate by aminotransferases, including aspartate and alanine aminotransferases (T'ases) and GABA-T. In glia, glutamate can be converted to glutamine by glutamine synthetase (GS). Other abbreviations are given in the text. Dark-grey boxes denote enzymes present in both neurons and glia, light-grey boxes denote enzymes present only in neurons

Figure 11.1 Enzymes responsible for the synthesis and metabolism of GABA. Enzymes responsible for the syntheisis (GAD) and metabolism (GABA-T and SSADH) of GABA, and their relationship to the TCA cycle and the amino acids glutamate and glutamine. Precursor glutamate is derived from glutamine by phosphate-activated glutaminase (PAG) and from a-ketoglutarate by aminotransferases, including aspartate and alanine aminotransferases (T'ases) and GABA-T. In glia, glutamate can be converted to glutamine by glutamine synthetase (GS). Other abbreviations are given in the text. Dark-grey boxes denote enzymes present in both neurons and glia, light-grey boxes denote enzymes present only in neurons enzymes found in the mitochondria of both neurons and glia: the multi-enzyme complex glutamate dehydrogenase (GDH), and several aminotransferases (including aspartate and alanine aminotransferases) whose action is analogous to that of GABA-T. Fig. 11.2 shows the pathways of GABA metabolism in the context of a GABAergic synapse.

Regulation of GAD

Of key importance in the synthesis of GABA is the short-term regulation of GAD activity. Increasing the availability of glutamate does not lead to an increase in the production of GABA, suggesting that GAD may normally be saturated with its substrate. Instead, the control of GAD activity is intimately linked to the enzyme's requirement for the co-factor pyridoxal-5'-phosphate (PLP; a form of vitamin B6) (Martin and Rimvall 1993). GAD exists in two states; an inactive apoenzyme (apoGAD) lacking the co-factor and active

Gaba Metabolism

Figure 11.2 Pathways for GABA metabolism. GABA is synthesised in nerve terminals by GAD. GABA produced by both GAD67 and GAD65 can be used as a neurotransmitter but GAD65 is preferentially associated with synaptic vesicles. Synaptically released GABA is recovered into neurons and glia by GABA transporters (not shown is the possible release of GABA by reversal of these transporters). In both neurons and glia, GABA is degraded in mitochondria by GABA-T. Glutamine produced in glial cells is exported to neurons and converted to glutamate (after Soghomonian and Martin 1998)

Figure 11.2 Pathways for GABA metabolism. GABA is synthesised in nerve terminals by GAD. GABA produced by both GAD67 and GAD65 can be used as a neurotransmitter but GAD65 is preferentially associated with synaptic vesicles. Synaptically released GABA is recovered into neurons and glia by GABA transporters (not shown is the possible release of GABA by reversal of these transporters). In both neurons and glia, GABA is degraded in mitochondria by GABA-T. Glutamine produced in glial cells is exported to neurons and converted to glutamate (after Soghomonian and Martin 1998)

holoenzyme (holoGAD) complexed with PLP. During the synthetic process GAD can undergo cycles of interconversion between these states. As illustrated in Fig. 11.3, in the primary reaction sequence active holoGAD combines with glutamate to form a complex, which, after decarboxylation, yields an intermediate product that is rapidly converted to GABA and free holoGAD. The intermediate product can also undergo an alternative reaction to produce succinic semialdehyde (SSA) and pyridoxamine-5'-phosphate (PMP) which, unlike PLP, dissociates from GAD to leave inactive apoenzyme, requiring additional PLP to be reactivated.

Traditionally, two processes have been considered important with respect to the regulation of GAD. First, GABA may promote conversion of GAD from its active to its inactive state, and so cause feedback inhibition of GABA synthesis. Second, ATP appears to inhibit, while inorganic phosphate promotes, the reactivation of GAD by PLP. During periods of increased neuronal activity, when the consumption of ATP increases, a rise in the level of phosphate should stimulate the conversion of inactive to active GAD, thereby increasing GABA synthesis. More recently, it has been suggested that soluble and membrane-bound forms of GAD may be differentially regulated. The soluble form of GAD is activated by a phosphatase that causes dephosphorylation while the membrane-bound form is activated following phosphorylation by a vesicular protein kinase (Hsu et al. 1999).

Glutamate Gaba Ssa

Figure 11.3 Regulation of GAD during the synthesis of GABA. Active GAD (GAD-PLP) combines with glutamate (1) to form a complex (GAD-PLP-GLU). After decarboxylation (2) this yields GABA and GAD-PLP (3). The intermediate product (GAD-INT) can undergo an alternative reaction (4) to produce succinic semialdehyde (SSA) and pyridoxamine-5'-phosphate (PMP). PMP dissociates from GAD (5) leaving inactive enzyme, which requires additional PLP to be reactivated (6), a process that is affected by ATP and inorganic phosphate

Figure 11.3 Regulation of GAD during the synthesis of GABA. Active GAD (GAD-PLP) combines with glutamate (1) to form a complex (GAD-PLP-GLU). After decarboxylation (2) this yields GABA and GAD-PLP (3). The intermediate product (GAD-INT) can undergo an alternative reaction (4) to produce succinic semialdehyde (SSA) and pyridoxamine-5'-phosphate (PMP). PMP dissociates from GAD (5) leaving inactive enzyme, which requires additional PLP to be reactivated (6), a process that is affected by ATP and inorganic phosphate

Two isoforms of GAD

In addition to the inactive and active states of GAD, there are two distinct forms of the enzyme. The two isoforms, GAD67 and GAD65, named for their respective molecular masses 67 and ~ 65kDa), are encoded by separate, independently regulated genes, GAD1 and GAD2 (Erlander et al. 1991). GAD67 and GAD65 differ substantially in their amino-acid sequence, their interaction with PLP, their kinetic properties, and their regulation (Soghomonian and Martin 1998). Individual cells contain both forms of GAD but the ratio of the two differs among different neuronal populations. GAD65 is located preferentially in nerve terminals, both in the cytosol and as a membrane-bound form closely associated with synaptic vesicles into which the newly synthesised GABA is accumulated (see below). GAD65, unlike GAD67, is not saturated with PLP and forms the majority of the apoenzyme present in brain (about half of the total GAD). This has led to the view that a proportion of GAD65 exists as a pool of inactive enzyme, ready to combine with PLP in response to cellular signals for increased GABA synthesis.

Further insights into the role of GAD isoforms in the synaptic release of GABA have been provided by the techniques of gene manipulation. Mice lacking the GAD67 gene have a greatly reduced level of brain GABA (Asada et al. 1997). The neurological significance of this reduction is difficult to ascertain: GABA appears essential for the normal development of the palate and one consequence of the reduced production in GABA in these mice is a cleft palate that is responsible for their death soon after birth. In contrast, mice lacking the GAD65 gene show only a modest reduction in total brain GABA but exhibit spontaneous seizures and a greater susceptibility to chemical convulsants (Asada et al. 1996; Kash et al. 1997). In these mice basal GABAergic transmission is normal but GABA release during sustained synaptic activation is reduced (Obata et al. 1999; Tian et al. 1999). Together these results suggest that GAD67 is responsible for the synthesis of most brain GABA, but GAD65 is intimately involved in synthesis of GABA required for the refilling of the releasable pool of synaptic vesicles.

Inhibitors of GAD

Several drugs are known to inhibit GAD, either directly or through interaction with the co-factor PLP. The largest group of inhibitors are the hydrazides, such as isoniazid, semicarbazide and thiosemicarbazide. These are carbonyl-trapping agents that react with the aldehyde group of PLP; as many other enzymes use PLP as a co-factor, these agents are not specific for GAD. Two other agents, allylglycine and 3-mercaptopropionic acid, are competitive inhibitors of GAD. In general, GAD inhibitors reduce the level of GABA in the brain and cause seizures in experimental animals that, in the case of the hydrazides, can be overcome by application of vitamin B6, the precursor of PLP. Similarly, in humans an inherited defect in pyridoxine metabolism is characterised by a low concentration of GABA in the cerebrospinal fluid, and intrauterine or neonatal seizures that also respond to treatment with vitamin B6. These findings support the notion that maintained synthesis of GABA is an important factor in the control of overall brain excitability.

STORAGE OF GABA

Within nerve terminals, GABA, like other classical non-peptide neurotransmitters, is stored in synaptic vesicles into which it is accumulated by active transport. The uptake of GABA from the cytosol (where it is present at a concentration of a few millimolar) into the vesicle lumen (where it may reach several 100 millimolar) is dependent on a vesicular protein that transports cytosolic GABA in exchange for lumenal protons. The proton electrochemical gradient that drives this uptake is generated by a H+-ATPase located in the vesicle membrane. Like vesicular transporters for monoamines and acetylcholine, the 'GABA transporter' recognises more than one substrate, and can also transport glycine (see below).

A gene (unc-47) encoding a vesicular GABA transporter was first identified from experiments on the simple nervous system of the nematode worm C. elegans. Mammalian homologues were subsequently cloned from rat and mouse; these were named VGAT (for vesicular GABA transporter; Mclntire et al. 1997) or VIAAT (for vesicular inhibitory amino acid transporter; Sagne et al. 1997), respectively. These essentially identical clones have sequences predicting proteins of approximately 520 amino acids with ten transmembrane domains and, when expressed in mammalian cell lines, confer vesicular GABA and glycine transport. Immunohistochemical studies showed that VGAT/VIAAT is concentrated not only in the terminals of GABAergic neurons but also in those of neurons known to use glycine as a neurotransmitter (Gasnier 2000). As yet, no specific blockers or modulators of VGAT/VIAAT activity have been identified.

UPTAKE OF GABA

Once released from a vesicle, GABA molecules are able to activate receptors located on the pre- or postsynaptic membrane before rapidly diffusing out of the synaptic cleft. The ultimate removal of GABA from the extracellular space, and the maintenance of a low extracellular GABA concentration (low micromolar), is achieved by the high-affinity Na+- and Cl~-dependent uptake of GABA into both GABAergic neurons and glial cells. Like the accumulation of GABA into vesicles, this is a secondary active transport mechanism, but in this case GABA uptake is coupled to the movement of Na+ down its electrochemical gradient into the cell.

Drugs which block the uptake of GABA may be beneficial in conditions of reduced GABA function, as they are likely to prolong the action of synaptically released GABA (Thompson and Gahwiler 1992). The uptake of GABA is inhibited by a variety of simple GABA analogues, including nipecotic acid, ^-alanine, 2,4-diaminobutyric acid (DABA), c«-3-aminocyclohexane-carboxylic acid (ACHC), 4,5,6,7-tetrahydroisoxazolo [4,5-c]pyridin-3-ol (THPO) and guvacine (Fig. 11.4), but as most of these do not cross the blood-brain barrier they have been of experimental interest only. In early studies, a number of compounds were suggested to preferentially inhibit GABA uptake into neurons (DABA and ACHC) or glia (^-alanine and THPO), while others were clearly non-selective (nipecotic acid and guvacine).

Cloned GABA transporters

This simple distinction between glial and neuronal uptake has required revision following the molecular cloning of a family of four Na+- and CL-dependent GABA transporters, each encoded by a different gene: GAT-1, GAT-2, GAT-3 and BGT-1 (reviewed by Palacin et al. 1998). The nucleotide sequence of GAT-1 predicts a protein of 599 amino acids with a presumed structure containing twelve membrane-spanning regions. The transport of each GABA molecule into the cell is coupled to the movement of 2 Na+ and 1 Cl". All the GABA transporters share a similar structure, with approximately 50% amino acid identity. GAT-1 appears to be mainly neuronal in origin as its mRNA is found in neurons and it is inhibited more effectively by neuronal than by glial uptake inhibitors. Nevertheless, immunohistochemical studies suggest some expression in glial cells. GAT-2 is found in cells of the ependyma and arachnoid membrane surrounding the brain and may play a role in the regulation of GABA in cerebrospinal fluid (it is also found in other tissues such as liver). GAT-3 is present in brain, principally in glia, but also in some neurons. BGT-1 was isolated from kidney and transports the osmolyte betaine as well as GABA (hence betaine/GABA transporter). It is present in the brain but its precise location and role are unclear.

In parallel with the identification of distinct transporters for GABA there has been continued interest in the development of selective blockers of these transporters and the therapeutic potential that could result from prolonging the action of synaptically released GABA. It has been known for a long time that certain pro-drugs of nipecotic acid (e.g. nipecotic acid ethyl ester) are able to cross the blood-brain barrier and are effective anticonvulsants in experimental models of epilepsy. More recently, several different systemically active lipophillic compounds have been described that act selectively on GAT-1, GAT-2 or GAT-3 (Fig. 11.4). Of these, tiagabine (gabitril), a derivative of nipecotic acid that acts preferentially on GAT-1, has proved clinically useful in cases of refractory epilepsy.

METABOLISM OF GABA

Once recovered into GABAergic nerve terminals or glia, GABA is metabolised to succinic semialdehyde and then to succinate. As detailed above, these reactions are catalysed by GABA-T and SSADH, respectively. The actions of these two enzymes are closely linked. Aminotransferase reactions are reversible but GABA-T breaks down GABA, rather than producing it, because the irreversible action of SSADH rapidly oxidises the product SSA to succinate (Fig. 11.1). SSA may also be reduced by the enzyme succinic semialdehyde reductase (SSAR) to form y-hydroxybutyric acid (GHB).

Succinic Acid Gabt

Tiagabine

NMC-711

SKF-99976-A

COO H

COOH

COOH

GAT-1

Brain Transmitter Transportation
MeO-

SlMAP-5114

COOH

COOH

GAT-3

GAT-3

Figure 11.4 Blockers of GABA transport.The upper panel shows the structure of several GABA analogues that inhibit GABA transport into neurons or glia (see text). The lower panels show more recently developed compounds that exhibit selectivity for various cloned GABA transporters

Inhibitors of GABA-T

Inhibition of GABA-T leads to an elevation of brain GABA and, presumably because of an enhanced presynaptic availability of the transmitter, this has an anticonvulsant effect. Inhibitors of GABA-T include aminooxyacetic acid, 5-amino-1,3-cyclohexadi-enenecarboxylic acid (gabaculine), y-vinyl GABA (vigabatrin) and 2-propylpenatanoic acid (valproate). The first two are PLP antagonists and are of experimental interest only. Vigabatrin is an irreversible inhibitor of GABA-T and has been used clinically as an anticonvulsant. Valproate is a widely used anticonvulsant but it is not clear to what extent inhibition of GABA-T contributes to its therapeutic properties, as it also inhibits SSADH and SSAR, and inhibits Na+ currents, thus limiting neuronal firing.

Was this article helpful?

+1 0

Post a comment