# Gamma-aminobutyric acid

Post-publication activity

Curator: Eugene Roberts

Figure 1: A plaque created by Dr. C. van der Stelt, chemist and artist, in honor of Roberts’ discovery and subsequent work on GABA at a meeting honoring him in Amsterdam, 1965 (provided by Dr. Eugene Roberts).

The term GABA refers to the simple chemical substance $$\gamma$$-aminobutyric acid (NH2CH2CH2 CH2COOH). It is the major inhibitory neurotransmitter in the central nervous system. Its presence in the brain first was reported in 1950 (Roberts and Frankel, 1950a).

## Discovery of GABA and early history

The history of GABA in brain began with the discovery of the unique presence of this substance in tissue of the vertebrate central nervous system (CNS). In the course of the study of free amino acids of various normal and neoplastic tissues in several species of animals by paper chromatography, relatively large amounts of an unidentified ninhydrin-reactive material were found in extracts of fresh brains of mouse, rat, rabbit, guinea-pig, human, frog, salamander, turtle, alligator, and chick. At most, only traces of this material were found in a large number of extracts of many other normal and neoplastic tissues and in urine and blood. The unknown material was isolated from suitably prepared paper chromatograms. A study of the properties of the substance in mouse brain revealed it to be GABA. The initial identification, based on the co-migration of the unknown with GABA on paper chromatography in three different solvent systems, was followed by an absolute identification of the GABA in brain extracts by the isotope derivative method. An abstract was submitted to the Federation meetings in March of 1950 reporting the presence of GABA in brain (Roberts and Frankel, 1950a). Three papers dealing with the occurrence of GABA in brain appeared later that year in the same issue of the Journal of Biological Chemistry (Roberts and Frankel, 1950b; Udenfriend, 1950; Awapara et al., 1950). Detailed histories of the early chemical work outlined above have been published (e.g. see Roberts, 1986a).

Detailed account of the discovery of GABA here: /history.

The 3 methylene groups between the amino and carboxyl groups of GABA endow it with great structural flexibility, allowing it freedom to explore the surrounding chemical space with a continuum of structures ranging from full extension ( Figure 1, upper right) to the contiguity of the amino and carboxyl groups shown in the cyclic form ( Figure 1, lower left). Therefore, GABA has potential capacity to engage in innumerable energy-minimizing, mutually shaping interactions with molecular entities encountered in its immediate environment.

## Basic neurophysiology of GABA

For several years the presence of GABA in brain remained a biochemical curiosity and a physiological enigma. It was remarked in the first review written on GABA that “Perhaps the most difficult question to answer would be whether the presence in the gray matter of the central nervous system of uniquely high concentrations of $$\gamma$$-aminobutyric acid and the enzyme which forms it from glutamic acid has a direct or indirect connection to conduction of the nerve impulse in this tissue” (Roberts, 1956). However, later that year, the first suggestion that GABA might have an inhibitory function in the vertebrate nervous system came from studies in which it was found that topically applied solutions of GABA exerted inhibitory effects on electrical activity in the brain (Hayashi and Nagai, 1956). In 1957, the suggestion was made that indigenously occurring GABA might have an inhibitory function in the central nervous system from studies with convulsant hydrazides (Killam, 1957; Killam and Bain, 1957). Also in 1957, suggestive evidence for an inhibitory function for GABA came from studies that established GABA as the major factor in brain extracts responsible for the inhibitory action of these extracts on the crayfish stretch receptor system (Bazemore et al., 1957). Within a brief period the activity in this field increased greatly, so that the research being carried out ranged all the way from the study of the effects of GABA on ionic movements in single neurons to clinical evaluation of the role of the GABA system in epilepsy, schizophrenia, mental retardation, etc. This surge of interest warranted the convocation in 1959 of the first truly interdisciplinary neuroscience conference ever held, at which were present most of the individuals who had played a role in opening up this exciting field (Roberts et al, 1960).

During the aforementioned period, GABA became established as the major inhibitory neurotransmitter in the central nervous system (CNS). It was found to fulfill the “classical” requirements for neurotransmitter: proof of identity of postsynaptic action with that of the natural transmitter, presence in inhibitory nerves, releasability from terminals of identified nerves, and the presence of a rapid inactivating mechanism at synapses. Information on the GABA system, as a whole, up to 1960 has been thoroughly reviewed and extensively documented (Roberts and Eidelberg, 1960, and Roberts, et al., 1960) and major updates have appeared at intervals (Roberts, et al., 1976; Bowery, 1984; Olsen and Venter, 1986; Martin and Olsen, 2000).

Figure 2: Some metabolic relationships in nervous tissue.

## A brief synopsis of the neurochemistry of GABA

GABA is formed in the CNS of vertebrate organisms to a large extent, if not entirely, from L-glutamic acid ( Figure 2). The reaction (reaction 5) is catalyzed by L-glutamic acid decarboxylase (GAD), an enzyme found in mammalian organisms largely in neurons in the CNS, although there now are many reports of the occurrence of both GAD and GABA in neurons in the peripheral nervous system, as well as in some nonneural tissues (e.g., pancreas) and in body fluids. Brain GAD catalyzes the rapid $$\alpha$$-decarboxylation of L-glutamic acid and, of the rest of the naturally occurring amino acids, only L-aspartic acid to a very slight extent. Genes for two brain GAD isoforms have been cloned, as have families of other GABA-related proteins, such as 19 GABAA receptors and 2 to 3 GABAB receptors. It now is possible to visualize GABA, itself, and most of the proteins involved in GABA metabolism, release, and action on sections of the CNS at the light and electron microscopic levels, employing antisera to the purified components and peroxidase-labelling techniques. This has led to much more definitive data than were hitherto available through cell fractionation and lesion studies and has given detailed information of the interrelationships of GABA neurons in various nervous system regions (Roberts, 1978, 1980, 1984, 1986a).

The reversible transamination of GABA with $$\alpha$$-ketoglutarate (reaction 9) is catalyzed by a mitochondrial aminotransferase, termed GABA-transaminase (GABA-T), which in the CNS is found chiefly in the gray matter but also occurs in other tissues. The products of the transaminase reaction are succinic semialdehyde and glutamic acid. There is present an excess of a dehydrogenase that catalyzes the oxidation of succinic semialdehyde to succinic acid, which in turn can be oxidized via the reactions of the tricarboxylic acid cycle. Because succinic semialdehyde is oxidized to succinate without the intermediate formation of succinyl-coenzyme A, one consequence of the operation of the GABA shunt in brain, through which 10% to 20% of glucose metabolism may flow, is a decreased rate of phosphorylation of guanosine diphosphate (GDP) to guanosine triphosphate (GTP). The latter may be involved in activation of G proteins, formation of deoxy GTP for mitochondrial DNA synthesis, and synthesis of adenosine triphosphate (ATP). Although the exact functional significance of this GABA-dependent metabolic shunt still is not apparent, it seems certain that GABA plays a special metabolic role in brain mitochondria, which is abrogated when inhibition of GABA-T occurs. Of the keto acids normally present, only $$\alpha$$-ketoglutarate is an amino group acceptor. In addition to GABA, several other ω amino acids also are effective amino donors.

Steady-state concentrations of GABA in various brain areas normally are governed by the activity of GAD and not by GABA-T. In many inhibitory nerves, both GAD and GABA-T are present and are found throughout the neuron, GAD being more highly concentrated in the presynaptic terminals than elsewhere. The GABA-T is contained in mitochondria of all neuronal regions. GABA is a precursor of several substances found in nervous tissue and cerebrospinal fluid, among which are GABA histidine (homocarnosine), GABA-1-methylhistidine, $$\gamma$$-guanidinobutyric acid, GABA-1-cystathionine, $$\alpha$$-(GABA)-L-lysine, GABA-choline, and putreanine [(N-4-aminobutyrl)-3-aminopropionic acid]. Homocarnosine is present exclusively in brain and cerebrospinal fluid, and there are data suggesting important roles for it as an antioxidant, an optimizer of immune function, and a modifier of brain excitability.

Important controls in regulation of the GABA system might be exerted at points related to the availability of glutamic acid, the substrate for GABA synthesis in nerve endings by GAD (reaction 5). Glutamate carbon can originate from glucose through glycolysis and the Krebs cycle (upper right-hand corner of Figure 2), from glutamine subsequent to uptake (reaction 6), and from proline (reactions 3 and 4) and ornithine (reactions 2 and 4). Ornithine (reactions 2 and 3), but not glutamate, is an effective precursor of proline in nerve terminals, a putative inhibitory neurotransmitter. Arginine can be converted to ornithine (reaction 1), which in turn gives rise to glutamate (reactions 2 and 4), proline (reactions 2 and 3), and GABA (reactions 2, 4, and 5).

GAD requires pyridoxal phosphate (PLP), a form of vitamin B6, as a coenzyme (Roberts et al., 1964). Dietary forms of vitamin B6 are absorbed and converted efficiently in tissues to (PLP), which is synthesized in brain from ATP and pyridoxal. PLP can readily be removed from the enzyme protein of GAD causing loss of enzyme activity, and the lost enzymatic activity can be restored simply by the addition of the coenzyme. Pyridoxine-deficient animals show a decrease in the degree of saturation with the coenzyme of the enzyme protein of cerebral GAD, but no decrease is found in the content of enzyme protein in the deficient animals. Brain GAD activity is restored rapidly to normal on feeding of pyridoxine to deficient animals. Pyridoxine deficiency, however produced, results in a susceptibility to seizures in animals, including humans, probably because of decreased ability to make GABA. Seizures in an infant with a simple dietary deficiency of vitamin B6 were abolished completely almost immediately after intramuscular injection of pyridoxine. This indicates that in a normal individual there is an extremely rapid conversion of pyridoxine to pyridoxal phosphate, association of the coenzyme with the apoenzyme of GAD, and formation of GABA in nerve terminals. Hydrazides and other carbonyl-trapping agents react with the aldehyde group of PLP and decrease its availability as a coenzyme. The seizures that result when such agents are administered are partially attributable to the decreases in the amounts of releasable GABA in nerve terminals of inhibitory nerves.

## The inhibited nervous system: a global view of GABAergic function (Roberts, 1976, 1986b, 1991)

Perhaps the subject of neural inhibition had lain dormant for so many years because there was no material basis for it. Inhibitory neurons had not been identified, an inhibitory neurotransmitter had not been isolated and characterized, and postsynaptic sites for neural inhibition had not been shown. It is well to remember that it was not until 1952 (Eccles, 1982), two years after the discovery of GABA in brain, that the controversy as to whether synaptic transmission in the CNS is largely electrical or chemical in nature was settled in favor of the latter. It also was 3 years before modern molecular biology was begun by Watson and Crick (Watson and Crick, 1953).

GABA increases the permeability of membranes to specific ions in such a way as to cause the membranes to resist depolarization. For example, by acting on a particular class of receptors (GABAA), GABA produces an increase in permeability to Cl- ions that is measured as an increase in membrane conductance. GABA also produces increases in K+ conductance by action on another distinct class of receptors (GABAB) that are not colocalized with GABAA receptors. In general, GABA accelerates the rate of return of the resting potential of all depolarized membrane segments that it contacts and stabilizes undepolarized membrane segments by decreasing their sensitivity to stimulation. Thus, at many sites in the nervous system, GABA exercises inhibitory command-control of membrane potential. In this way this naturally occurring inhibitory transmitter can counteract the depolarizing action of excitatory processes to maintain the polarization of a cell at an equilibrium level near that of its resting value, acting essentially as a chemical voltage clamp. In most instances studied, GABA has been shown to exert hyperpolarizing or inhibitory effects by this mechanism. However, if high intracellular Cl- concentrations should occur, GABA can produce a decrease in membrane potential or depolarization. Data now suggest that the benzodiazepines (e.g., Valium) and barbiturates exert their pharmacologic effects largely by reacting with components of the GABAA receptor complex, thereby enhancing the efficacy of neurally released GABA.

GABA is inactivated at synapses by a mechanism that involves attachment to unique membrane recognition sites, different from those for the receptor, and subsequent removal from the synaptic junction by a Na+- and Cl--dependent transport process that is similar in principle to that used for transport of many other substances. The removal of synaptically released GABA takes place by reuptake into terminals of neurons and into glial processes that invest the synapses.

 Figure 3: (A) Control section (non-immunostained) of nucleus interpositus in rat cerebellum. Neuronal soma (s). (B) Neuropil of nucleus interpositus immunostained for GAD. Soma of neuron(s), dendrite (d), reaction product (long arrows), grazed neuron soma (encircled by short arrows) with bouton-like reaction product on cell surface (b). (C) Neuron shown in Fig. 2B, photographed with Nornarski optics. Soma (s), dendrite (d), bouton-like deposits of reaction product (b). Figure 4: Electron micrographs of various types of synaptic terminals which contain GAD, the enzyme that synthesizes GABA. All specimens were obtained from the rat CNS. (a) axodentritic synapses in the substantia nigra (T1 and T2) with a dendritic shaft (D) in the pars reticulate; (b) axoaxonal synapse in cerebral cortex; (c) axosomatic synapse in dorsal horn of spinal cord; (d) axoaxonal synapse in the dorsal horn of the spinal cord; (e) dendrodentritic synapses in the glomerular layer of the olfactory bulb.

The ubiquity and extent of immunocytochemically visualized presynaptic endings of inhibitory GABAergic neurons on various structures in the vertebrate nervous system are striking. The impression is that of looking at a highly restrained nervous system ( Figure 3 and Figure 4). In coherent behavioral sequences, innate or learned, preprogrammed circuits are released to function at varying rates and in various combinations. This is accomplished largely by the disinhibition of pacemaker neurons whose activities are under the dual tonic inhibitory controls of local-circuit GABAergic neurons and of GABAergic projection neurons coming from neural command centers. According to this view, disinhibition is permissive, and excitatory input to pacemaker neurons serves mainly a modulatory role.

Disinhibition., acting in conjunction with intrinsic pacemaker activity and often with modulatory excitatory input, is one of the major organizing principles in nervous system function. For example, cortical and hippocampal pyramidal neurons are literally studded with terminals from inhibitory GABAergic neurons. Not only are the endings of the local-circuit GABAergic aspinous stellate neurons densely distributed around the somata and dendrites of the cortical pyramidal cells, but they are also located on initial axon segments, where they act as frequency filters. In addition, GABA neurons have terminals from other GABAergic neurons impinging on them. Pyramidal cells are tightly inhibited by local-circuit inhibitory neurons that may themselves be inhibited by the actions of other inhibitory neurons in such a way that disinhibition of the pyramidal neurons occurs. Local-circuit GABAergic neurons also participate in processes that result in feedforward, feedback, surround, and presynaptic inhibition and presynaptic facilitation.

Both inhibition and disinhibition play key roles in information processing in all neural regions. Normally, the principal cells in particular neural sectors may be held tightly in check by constant tonic action of inhibitory neurons. Through disinhibition, neurons in a neural sector may be released to fire at different rates and sequences and, in turn, serve to release circuits at other levels of the nervous system. Communication among neural stations and substations may take place largely by throwing of disinhibitory neural switches. This may be the way information flows from sense organ to cerebral sensory area, through associative areas to the motor cortex, and by way of the pyramidal paths to the final motor cells of the medulla and spinal cord.

## GABA and diseases of the CNS

Defects in coordination between the GABA system and other neurotransmitter and modulator systems may involve a local brain region, several brain regions, or the entire CNS. Enhanced synchrony of neuronal firing (e.g., in seizures) may arise in several ways: increased rate of release of synaptic excitatory transmitters, blockade of inhibitory transmitter receptor mechanisms, desensitization of receptors to inhibitory transmitters, decreased availability of inhibitory transmitter, decreased activity of inhibitory neurons, and increased formation or activation of electrotonic (gap) junctions. Immunocytochemical studies of the sensorimotor cortex in experimental epilepsy in monkeys showed highly significant reductions in numbers of GABAergic terminals of electrographically proved epileptogenic sites of alumina gel application. Electronmicroscopic observations showed a marked loss of axosomatic synapses on the pyramidal cells and a replacement of synaptic appositions with astrocytic processes in the alumina cream-treated animals. However, the symmetric, presumably excitatory synapses on the dendrites of these pyramidal cells appeared to be largely intact. Comprehensive biochemical studies complementary to the morphologic ones showed a significant correlation with seizure frequency only with losses in GABAergic receptor-related binding and decreased GAD activity. Current data support the notion that actual destruction or inactivation of inhibitory interneurons is one of the major cerebral defects predisposing to seizures, at least in the case of focal epilepsy (Roberts, 1986b). Mutations in GABAA receptor now have been shown to predispose individuals to various types of seizures (Macdonald, et al., 2004). GABA neurons play important roles in control mechanisms in various hypothalamic and brain stem centers. If their activity within these structures is compromised, abnormally enhanced responses may be observed, for example, in emotional reactivity, cardiac and respiratory functions, blood pressure, food and water intake, sweating , insulin secretion, liberation of gastric acid, and motility of the colon.

The roles of GABA neurons in information processing in various regions of the nervous system are so varied and complex that it appears doubtful that many useful drug therapies will come from approaches that are aimed at affecting one or another aspect of GABAergic function at all GABA synapses. Currently there are no drugs that are process and site specific. In this regard, the detailed molecular characterization that is being carried out of the enzymes of GABA metabolism, GABA receptors and transporters, the components of GABA receptor-associated anion channels, and the relationships among these structures and the lipidic membrane components in which they are imbedded should give rise to many opportunities for devising specific therapeutic modalities (e.g., see Roberts, 2006).

## GABA, The quintessential neurotransmitter: electroneutrality, fidelity, and specificity (Roberts, 1993)

Isoelectric Points (PI) of Major Naturally-Occurring Amino Acids and Peptides in Animal Tissues (From Greenstein, J.P., Winitz, M. Chemistry of the Amino Acids, Vol. 1. New York: John Wiley & Sons, 1961, pp. 486-489).
Amino Acid pI
Aspartic acid2.77
Glutamic acid3.22
Cystine5.03
Taurine5.12
Asparagine5.41
Phenylalanine5.48
Homocystine5.53
Threonine5.64
Glutamine5.65
Tyrosine5.66
Serine5.68
Methionine5.74
Hydroxyproline5.74
Tryptophan5.89
Citrulline5.92
Isoleucine5.94
Valine5.96
Glycine5.97
Leucine5.98
Alanine6.00
Sarcosine6.12
Proline6.30
β-Alanine6.90
Cysteine6.94
Homocysteine7.05
$$\gamma$$-Aminobutyric acid7.30
Histidine7.47
$$\delta$$-Amino-n-valeric acid7.52
$$\epsilon$$-Amino-n-caproic acid7.60
l-Methylhistidine7.67
Carnosine8.17
Anserine8.27
Lysine9.59
Ornithine9.70
Arginine 11.15

Nature’s choice of GABA as the major inhibitory neurotransmitter is an example of evolutionary optimization. Alone of the known neurotransmitters, GABA is an electroneutral zwitterion (isoelectric point, 7.3) at physiologic pH, the ionization constants for both its amino and carboxyl groups being sufficiently far removed from neutrality so that shifts of pH in the physiologic range produce little change in net charge (Table 1). This endows GABA with a capacity for higher fidelity of information transmittal than that of other known major neurotransmitters, enabling it, in “stealth” fashion, to escape the charged minefields encountered in passage through the dense extracellular environment lying between presynaptic sites of release and postsynaptic sites of action. Coordinate enhancement with progressive acidification occurs in GABAergic inhibitory function because GABA formation and its anion channel-opening efficacy are increased while its metabolic destruction by transamination and removal by transport are decreased. Diminution of GABAergic inhibitory function occurs on alkalinization. Contrariwise, acidification decreases postsynaptic efficacy of glutamate, the major excitatory neurotransmitter, and alkalinization increases it.

In this manner the delicate balance between excitation and inhibition in the brain is maintained within the adaptive range in response to local or global activity that acidifies the environment in which it occurs. Accelerated metabolism after nerve activity results in accelerated formation of carbon dioxide and lactic acid; the accompanying acidification applies physiologic “brakes,” so to speak, preventing structural and functional damage from taking place. When GABAergic-glutamatergic relations are unbalanced by glutamatergic overactivity, seizures may occur. For example, the excitement experienced at an athletic event with the attendant hyperventilation and consequent alkalinization not infrequently causes seizures in susceptible individuals. Overbalancing in favor of the GABA system can lead to maladaptive decrement in neural activity and even to coma.

The properties of the simple GABA molecule itself, and of the machinery built to support its function, make it eminently suitable to guide the brain in a “civilized” manner. The yin-yang relationship between the glutamatergic excitatory and GABAergic inhibitory systems is played out on the tightrope of a delicate balance, and imbalances between them lead to serious disorders.

No $$\alpha$$-, $$\beta$$-, or $$\omega$$- amino acid known to occur in any abundance in animal tissues approaches GABA in molar efficacy at the GABAA receptor. Therefore, the noise level created by nonspecific effects at the GABAA receptor are minimal, ensuring quantitative fidelity of the neural messages delivered by GABA.

The “charm” of GABA lies in nature’s choice of this simple molecule, made from the common metabolic soil of glutamic acid, for the all-important role as major controller of the infinitely complex machinery of the brain, allowing it to operate in the manner best described as freedom without license. Try as one might, one cannot come up with a better choice for the job (Roberts, 1991, 1993).

## References

Awapara, J., Landua, A.J., Fuerst, R., and Seale, B. Free gamma-aminobutyric acid in brain. Journal of Biological Chemistry 187:35-9, 1950.

Bazemore, A.W., Elliott, K.A.C., Florey, E. Isolation of Factor I. Journal of Neurochemistry 1:334-339, 1957.

Bowery, N.G., ed. Actions and Interactions of GABA and Benzodiazepines. New York: Raven Press, 1984.

Eccles, J.C. The synapse: from electrical to chemical transmission. Annual Review of Neuroscience, 5:325-339, 1982.

Hayashi, T., Nagai, K. Action of ω-amino acids on the motor cortex of higher animals, especially γ-amino-β-oxybutyric acid as the real inhibitory principle in brain. In: Abstracts of reviews: Abstracts of communications. Brussels: Twentieth International Physiological Congress, p. 410, 1956.

Killam, K.F. Convulsant hydrazides. II. Comparison of electrical changes and enzyme inhibition induced by the administration of thiosemicarbazide. Journal of Pharmacological and Experimental Therapeutics. 119:263-271, 1957.

Killam, K.F., Bain J.A. Convulsant hydrazides. I. In vitro and in vivo inhibition of vitamin B6 enzymes by convulsant hydrazides. Journal of Pharmacological and Experimental Therapeutics. 119:255-262, 1957.

Macdonald, R.L., Gallagher, M.J., Feng, H.-J., Kang, J. GABAA receptor epilepsy mutations. Biochemical Pharmacology. 68:1497-1506, 2004.

Martin, D. L. & Olsen, R. W., eds. GABA in the Nervous System. Philadelphia: Lippincott Williams & Wilkins, 2000.

Olsen, R.W. and Venter J.C., editors Benzodiazepine/GABA Receptors and Chloride Channels: Structural and Functional Properties. New York: Alan R. Liss, Inc., 1986.

Roberts, E. Formation and utilization of γ-aminobutyric acid in brain. Progress in Neurobiology. I. Neurochemistry. Korey, S.R., and Burnberger, J.I. eds. New York: Hoeber-Harper, pp. 11-25, 1956.

Roberts, E. Disinhibition as an organizing principle in the nervous system: the role of the GABA system. Application to neurologic and psychiatric disorders. In: Roberts, E., Chase, T.N., and Tower, D.B., eds., GABA in Nervous System Function, New York, Raven Press, pp. 515-539, 1976.

Roberts, E. Roles of GABA neurons in information processing in the vertebrate CNS. In: Karlin, A., Tennyson, V.M., Vogel, H.J., eds. Neuronal Information Transfer. New York: Academic Press, pp. 213-239, 1978.

Roberts, E. γ-Aminobutyric acid (GABA): a major inhibitory transmitter in the vertebrate nervous system. In: Levi-Montalcini, R., ed. Nerve Cells, Transmitters and Behaviour. Rome: Pontifical Academy of Sciences, 163-213, 1980.

Roberts, E. γ-Aminobutyric acid (GABA): from discovery to visualization of GABAergic neurons in the vertebrate nervous system. In: Actions and Interactions of GABA and Benzodiazepines, Bowery, N.G., ed. New York: Raven Press, pp. 1-25, 1984.

Roberts, E. GABA: The road to neurotransmitter status. In: Benzodiazepine/GABA Receptors and Chloride Channels: Structural and Functional Properties, R.W. Olsen and J.C. Venter, editors, pp. 1-39. New York: Alan R. Liss, Inc., 1986a.

Roberts, E. Failure of GABAergic inhibition: a key to local and global seizures. Advances in Neurology,44:319-341, 1986b.

Roberts, E. Living systems are tonically inhibited, autonomous optimizers, and disinhibition coupled to variability generation is their major organizing principle: inhibitory command-control at levels of membrane, genome, metabolism, brain, and society. Neurochemical Research 16:409-421, 1991.

Roberts, E. Adventures with GABA: Fifty Years On. In: GABA in the Nervous System: The View at Fifty Years, D.L. Martin and Richard W. Olsen, editors, pp. 1-24, Lippincott, Williams & Wilkins, Philadelphia, 2000.

Roberts, E. GABAergic malfunction in the limbic system resulting from an aboriginal genetic defect in voltage-gated Na+-channel SCN5A is proposed to give rise to susceptibility to schizophrenia. Advances in Pharmacology 54:119-145, 2006.

Roberts, E., Baxter, C.F., Van Harreveld, A., Wiersma, C.A.G., Adey, W.R., and Killam, K.F., eds. Inhibition in the Nervous System and Gamma-aminobutyric Acid. Oxford: Pergamon Press, 1960.

Roberts, E., Chase, T.N., and Tower, D.B., eds., GABA in Nervous System Function, New York, Raven Press, 1976.

Roberts, E. and Eidelberg, E. Metabolic and neurophysiological roles of γ-Aminobutyric acid. International Review of Neurobiology 2:279-332, 1960.

Roberts, E. and Frankel, S. γ-Aminobutyric acid in brain. Federation Proceedings 9:219, 1950.

Roberts, E. and Frankel, S. γ-Aminobutyric acid in brain: Its formation from glutamic acid. Journal of Biological Chemistry 187:55-63, 1950.

Roberts, E. and Sherman, M.A. GABA—the quintessential neurotransmitter: electroneutrality, fidelity, specificity, and a model for the ligand binding site of GABAA receptors. Neurochemical Research 18:365-376, 1993.

Roberts, E., Wein, J., Simonsen, D.G. γ-Aminobutyric acid (GABA), vitamin B6 and neuronal function: a speculative synthesis. Vitamins and Hormones 22:503-559, 1964.

Udenfriend, S. Identification of gamma-aminobutyric acid in brain by the isotope derivative method. Journal of Biological Chemistry 187:65-9, 1950.

Watson, J.D. and Crick, F.H.C. Molecular structure of nucleic acids: a structure for deoxyribose nucleic acid. Nature 171:737-8, 1953.

Internal references