370

Current Pharmaceutical Design, 2005, 11, 1867-1885 1867

1381-6128/05 $50.00+.00 © 2005 Bentham Science Publishers Ltd.

GABAA Receptor Channel Pharmacology

Graham A.R. Johnston*

The Adrien Albert Laboratory of Medicinal Chemistry, Department of Pharmacology, The University of Sydney, NSW2006, Australia

Abstract: GABAA receptor channels are ubiquitous in the mammalian central nervous system mediating fast inhibitoryneurotransmission by becoming permeant to chloride ions in response to GABA. The emphasis of this review is on therich chemical diversity of ligands that influence GABAA receptor function. Such diversity provides many avenues for thedesign and development of new chemical entities acting on GABAA receptors. There is also a significant diversity ofGABAA receptor subtypes composed of different protein subunits. The discovery of subtype specific agents is a majorchallenge in the continuing development of GABAA receptor pharmacology. Leads for the discovery of new chemicalentities that influence GABAA receptors come from using recombinant GABAA receptors of known subunit compositionas has been elegantly demonstrated by the refining of benzodiazepine actions with α1 subunit preferring agents showingsedative properties but not anxiolytic properties. The most recent advances in the therapeutic use of agents acting onGABAA receptors concern the promotion of sound sleep. Many herbal medicines are used to promote sleep and many oftheir active ingredients include flavonoids and terpenoids known to modulate GABAA receptor function.

INTRODUCTION

GABA (γ-aminobutyric acid), the major inhibitoryneurotransmitter in the brain, is essential for the overallbalance between neuronal excitation and inhibition that isvital to normal brain function. Too much inhibition, or toolittle excitation, can lead to coma, depression, low bloodpressure, sedation or sleep. Too much excitation, or too littleinhibition, can result in a range of conditions includingconvulsions, anxiety, high blood pressure, restlessness andinsomnia. Either imbalance in the extreme can result indeath. The exact symptoms depend on what regions of thebrain are involved and exactly what nerve cells are out ofbalance. Restoration of the balance between excitation andinhibition is a major aim of therapies that target GABA-mediated neuronal inhibition.

GABA produces neuronal inhibition by acting on anamazing diversity of membrane-bound receptors. Thesereceptors can be divided into two major types: ionotropicreceptors that are ligand-gated ion channels (GABAA andGABAC receptors), and metabotropic receptors that are G-protein coupled receptors (GABAB receptors) that act viasecond messengers [1, 2]. The ionotropic GABA receptorsbelong to the nicotinicoid superfamily of ligand-gated ionchannels as described by Le Novere and Changeux [3] thatincludes nicotinic acetylcholine, strychnine-sensitive glycineand 5HT3 receptors. The family of ionotropic GABAreceptors is divided into two subfamilies, GABAA andGABAC receptors, on the basis of their ability to formendogenous heteromeric and homomeric receptorsrespectively, and differences in their physiological andpharmacological properties [1], although GABAC receptorsare sometimes classified as subtypes of GABAA receptors

*Address correspondence to this author at the The Adrien Albert Laboratoryof Medicinal Chemistry, Department of Pharmacology, The University ofSydney, D06, Sydney, NSW 2006, Australia; Tel: 61 2 9351 6117; Fax: 612 9351 2891; E-mail: [email protected]

[4]. There is also diversity in GABAB metabotropic GABAreceptors that are heteromeric dimers [5]. There is evidencefor the existence of functional GABAA, GABAB and GABAC

receptors, as well as strychnine-sensitive glycine receptors,on a single population of retinal ganglion cells [6].

Like other members of the nicotinicoid superfamily ofligand-gated ion channels, ionotropic GABA receptors areconsidered to consist of 5 protein subunits arranged around acentral pore that constitutes the actual ion channel [1]. Eachsubunit has a large extracellular N-terminal domain whichincorporates part of the agonist/antagonist binding site,followed by three membrane spanning domains (M1-3), anintracellular loop of variable length and a fourth membranespanning domain (M4), with the C-terminal end beingextracellular. Each subunit arranges itself such that thesecond membrane-spanning domain (M2) forms the wall ofthe channel pore and the overall charge of the domaindetermines whether the channel conducts anions or cations.Both GABAA and GABAC receptors are GABA-gatedchloride ion channels causing inhibition of neuronal firing,with GABAA receptors being heteromeric, i.e. made up ofdifferent subunits (e.g. α1, β2 and γ2 subunits) and GABAC

receptors being homomeric (e.g. made up exclusively ofeither ρ1, ρ2 or ρ3 subunits; in addition ‘pseudoheteromeric’GABAC receptors made up of ρ1 and ρ2 subunits have beendescribed). The cytoplasmic loop, between the third andfourth transmembrane domains (M3 and M4), is believed tobe the target for protein kinases, required for subcellulartargeting and membrane clustering of the receptor. There are16 different subunits comprising the GABAA receptorfamily: α1-6, β1-3, γ1-3, δ, ε, π and θ [7]. In addition, thereare splice variants of many of these subunits. If all of thesesubunits could co-assemble to form functional pentamericreceptors the total number of GABAA receptors wouldbe huge. Even if the combinations were restricted tothose containing two α, two β and one other subunit, thenmore than 2000 different GABAA receptors could exist [8].

1868 Current Pharmaceutical Design, 2005, Vol. 11, No 15 Graham A.R. Johnston

In fact, studies of native GABAA receptors suggest that theremay be less than 20 widely occurring GABAA receptorsubtype combinations, with the major combinations beingα1β2/3γ2, α3β3γ2 and α2β3γ2 [7, 9].

This review is directed at some recent highlights,together with a re-evaluation of some older data, relevant toGABAA receptors as therapeutic targets with an emphasis onthe chemical diversity of ligands that influence GABAA

receptor function. Such diversity provides many avenues forthe design and development of new chemical entities actingon GABAA receptors. There are many reviews on aspectsGABAA receptors including methodological approaches tothe study of GABAA receptors [10, 11], GABAA receptorsubtypes [7, 12-15], neurosteroid modulation [16, 17], druginteractions [18], specific agonists and partial agonists [19],receptor recycling and regulation [20], novel modulators[21], medicinal chemistry [1, 22], and analysis of GABAA

receptors through mouse genetics [23]. GABAC receptors astherapeutic targets have been the subject of a recent review[24].

THERAPEUTIC USES OF AGENTS ACTING ONGABAA RECEPTORS

Agents acting on GABAA receptors have widespreadtherapeutic use as anaesthetics, anticonvulsants, anxiolyticsand sedative-hypnotics. In the main, these agents act toincrease GABA-mediated synaptic inhibition either bydirectly activating GABAA receptors or, more usually, byenhancing the action of GABA on GABAA receptors. Thislatter action is known as positive modulation [8] and isconsidered to involve agents acting on allosteric sites onGABAA receptors remote from the GABA recognition sites(orthosteric sites). Such allosteric sites are regarded as goodtargets for the development of subtype specific drugs sincethere is generally greater diversity between receptor subtypesin amino acid sequence at allosteric sites than at orthostericsites [25]. Agents that reduce the action of GABA onGABAA receptors are known as negative allosteric modula-tors (once known as ‘inverse agonists’); they have theopposite actions to those of the classical benzodiazepines.Agents that block the actions of both positive and negativeallosteric modulators are known as neutralising allostericmodulators, e.g. the classical benzodiazepine ‘antagonist’flumazenil [8]. Benzodiazepines and barbiturates areexamples of widely used therapeutic agents that act aspositive allosteric modulators at GABAA receptors.

A rich chemical diversity of agents acting GABAA recep-tors is known [1]. The discovery of subtype specific GABAA

receptor agents is a major challenge in the continuingdevelopment of therapeutic agents acting on specific wildtype and mutant GABAA receptors.

DISORDERS INVOLVING GABAA RECEPTORS

Not surprisingly, GABA as the major inhibitoryneurotransmitter is involved, directly or indirectly, in manydisorders of brain function. The major disorders for whichGABAA receptors represent important therapeutic targetsinclude anxiety disorders, cognitive disorders, epilepsies,moods disorders, schizophrenia and sleep disorders.

Heritable mutations are known to occur across thenicotinicoid superfamily of ligand-gated ion channelsincluding GABAA receptors [26]. Angelman syndrome, aneurodevelopmental disorder characterised by severe mentalretardation, epilepsy and delayed motor development hasbeen associated with deletions of GABAA receptor β3subunits [27]. GABAA receptor β3 knockout mice haveepilepsy and a phenotype with marked similarities toAngelman syndrome [28, 29].

GABA systems have been implicated in the pathogenesisof anxiety, depression and insomnia. These symptoms arepart of the core and comorbid psychiatric disturbances inpost-traumatic stress disorder (PTSD). In a study of PTSDpatients, heterozygosity of β3 GABAA receptor subunits wasassociated with higher levels of anxiety, insomnia, socialdysfunction and depression than found in homozygosity[30]. There is increasing evidence for GABA abnormalitiesin mood disorders [31] and in alcohol dependence [32, 33].Some of the neurological symptoms of guanidinoacetatemethyltransferase deficiency may be due to the partialagonist action of increased levels of guanidinoacetate onGABAA receptors [34].

The use of herbal medicines to treat depression may beassociated with actions on GABAA receptors [35]. GABAA

receptor ligands have a potential role in the treatment ofschizophrenia [36, 37], acute ischaemic stroke [38], tinnitus[39] and impaired cognition [40].

Epilepsies

Studies on the genetics of human epilepsies show thatepilepsy syndromes that have monogenic inheritance areassociated with mutations that encode subunits of voltagegated and ligand gated ion channels [41]. Heritable muta-tions in GABAA receptor subunits are strongly implicated inidiopathic generalised epilepsies [42]. Mutations in γ2GABAA receptor subunits have been described in twofamilies with generalised epilepsy syndromes [43, 44]. Onemutation associated with febrile seizures and generalisedepilepsy is in the extracellular loop connecting the TM2 andTM3 domains. Studies using recombinant receptors in frogoocytes revealed that this mutant showed smaller GABA-activated currents than did receptors lacking this mutation[43]. The other mutation was in the distal part of the Nterminus thought to make up part of the benzodiazepinebinding pocket and was associated with childhood absenceepilepsy and febrile seizures. Studies in frog oocytes showedthat this mutation had diminished sensitivity to positiveallosteric modulation by benzodiazepines [44]. Thesefindings have been questioned recently as a result of patchclamp studies using a mammalian expression system (HEKcells) that showed that the TM2-TM3 loop γ2 GABAA

receptor mutation resulted in faster deactivation rates, whilethe N terminus mutant reduced current amplitude withoutaltering benzodiazepine sensitivity [45]. Mutations inintracellular regions of γ2 GABAA receptor subunits betweenTM1 and TM2 and between TM3 and TM4 have also beenassociated with epilepsies. It appears that all of thesemutations in γ2 GABAA receptor subunits may result indiminished synaptic inhibition mediated by GABAA recep-tors and epilepsy by diverse mechanisms [46]. In addition to

GABAA Receptor Channel Pharmacology Current Pharmaceutical Design, 2005, Vol. 11, No.15 1869

these mutations, a γ2 GABAA receptor subunit splice sitemutation has been associated with childhood absenceepilepsy and febrile convulsions, but it is not known howthis influences GABAA receptor properties [47]. Anassociation analysis has shown that polymorphism in the γ2GABAA receptor gene is a susceptibility factor for febrileseizures [48]. A mutation in the TM3 region of the α1GABAA receptor subunit that influences both the efficacyand affinity of GABA has been associated with juvenilemyoclonic epilepsy [49]. Studies on recombinant receptorscontaining this TM3 α1 mutation show that it results inreduced channel open time with no change in single channelconductance [50].

Although heritable epilepsies represent only a smallfraction of epilepsies, the observed mutations associated withheritable epilepsies provide clues as to possible problemswith GABAA receptors implicated in other epilepsies. Theyalso provide targets for pharmacogenomic therapies usingdrugs acting on specific mutant GABAA receptors.

Sleep Disorders

GABA systems are known to play an important role insleep and positive allosteric modulators of GABAA receptorsare widely used to promote restful sleep [51]. Twoobservations indicate the importance of β3 GABAA receptorsubunits in sleep. Oleamide, an endogenous sleep promotingfatty acid, is inactive in β3 GABAA receptor subunitknockout mice [52]. A mutation in β3 GABAA receptorsubunits has been described in a patient with chronicinsomnia. Functional characterisation of this mutant showeda slower rate of desensitisation compared with normalGABAA receptors [53].

The treatment of insomnia is regarded as a developingmarket for agents acting on GABAA receptors. Drugscurrently used to treat insomnia include zolpidem(Ambien), zaleplon (Sonata) and zopiclone (Imovane).These drugs, Fig. (1), show some selectivity for α1 subunitcontaining GABAA receptors, acting as positive allostericmodulators. The structurally related indiplon, Fig. (1), whichis in phase III clinical trials for the treatment of insomnia,acts in a similarly selective manner [54, 55]. Also in phaseIII clinical trials is gaboxadol (THIP), a directly actingGABAA receptor partial agonist, discussed in Section 5, Fig.(3), that interacts with a GABAA receptor population that isinsensitive to benzodiazepines, zolpidem, zaleplon, zolpidemand indiplon [56].

Many herbal preparations are used to promote sleep. Forexample, chamomile tea contains the flavonoid apigenin (seeSection 7.1) which has been shown to enhance the positiveallosteric modulating effects of benzodiazepines on GABAA

receptors [24] and Valerian contains a variety of agents (seeSection 7.2) that act on GABAA receptors [57].

GABAA RECEPTOR SUBUNITS

The importance of drug interactions with receptorsubtypes made up of specific protein subunits has beenhighlighted by a number of recent findings involvinggenetically modified mice and ligands that show someselectivity for particular receptor subtypes. The finding that

different pharmacological actions of benzodiazepines resultfrom interactions with different GABAA receptor subtypeswas pivotal to these studies [14].

Fig. (1). Structures of sedative/hypnotic substances that interactpreferentially with α1 subunit containing GABAA receptors.

It must be remembered, however, that it is not only thenature of the protein subunits that influence the diversity offunction of GABAA receptor subtypes in vivo. Majorcontributors to this diversity are: presynaptic factorsincluding release probability and number of release sites;factors that determine synaptic GABA transients in the cleft,including diffusion and the actions of GABA transporters;and postsynaptic factors, including GABA receptor subtypes,their location and number, their modulation by endogenousand exogenous factors, and their interactions with postsynap-tic-anchoring proteins [58]. The transport, clustering andturnover of GABAA receptors are known to be influenced bya variety of proteins}. [20]. The phosphorylation state of thesubunits is very important with GABAA receptors as withmany other ligand-gated ion channels [59]. For example, thesensitivity of GABAA receptors to ethanol is dependent onreceptor phosphorylation; mice lacking protein kinase Cεshow increased sensitivity to ethanol while those lackingprotein kinase Cγ show decreased sensitivity [59]. Further-more, there are many endogenous ligands that influenceGABAA receptor function in vivo including metal ions suchzinc [60], steroids [17], and chemicals derived from our dietsuch as flavonoids [57].

Influence of α1 and α2 Subunits – Sedative andAnxiolytic Actions of Benzodiazepines

Benzodiazepines are considered to act on GABAA

receptors at a binding pocket at the interface between the γ2subunit and α subunits that contain a conserved histidineresidue in the benzodiazepine binding domain on the

N

N

O

N

N

N

N

O

O

N

N

Cl

O

NZolpidem Zopiclone

Zaleplon Indiplon

N

N

N

CN

N

O

N

N

N

N

O

O S

1870 Current Pharmaceutical Design, 2005, Vol. 11, No. 15 Graham A.R. Johnston

extracellular N-terminus (α1, α2, α3 and α5 subunits).Mutation of this histidine to an arginine results in GABAA

receptors that are insensitive to benzodiazepines in vitro[61]. GABAA receptors containing α4 or α6 subunits arerelatively insensitive to benzodiazepines.

Knock-in point mutations in the genes that code for α1 orα2 subunits produced mice that had different responses tobenzodiazepines. Mice with the mutant α1 subunits showedthe normal anxiolytic responses to benzodiazepines but notthe sedative effects [62, 63]. The reverse was true for micewith the mutant α2 subunits showing sedative but notanxiolytic effects in responses to benzodiazepines [64].While there is general agreement on the importance of theα1 GABAA receptor subunit in the sedative actions ofbenzodiazepines, there has been some doubt on the relativecontributions of the α2 and α3 subunits to the anxiolyticaction due to confounding effects on locomotor activity thatinfluence the assessment of anxiety [65]. Furthermore, inprofiling the influence of a range of agents onbenzodiazepine binding differences in functional activityneed to be taken into account [66]. Ultimately, validation ofGABAA receptor subtype-selective drugs needs to be carriedout using genetically modified mice [12]. Drugs selective forα2-containing GABAA receptors (found in about 15% ofdiazepam-sensitive GABAA receptors) would be expected tobe anxiolytics with greatly reduced sedative effectscompared to the non-selective benzodiazepines currently inclinical use [14]. The non-sedative anxiolytic L-838, 417,Fig. (2), is a neutralising modulator at α1 containing GABAA

receptors but is a positive modulator at α2, α3 and α5containing GABAA receptors [63]; this agent thus offers adouble pronged approach to anxiolysis without sedation byenhancing the action of GABA at α2, α3 and α5 containingGABAA receptors while diminishing the action of endo-genous benzodiazepines on α1 containing GABAA receptors.New chemical entities with functional selectivity for α2 over

α1 subtypes of GABAA receptors are being developed, e.g.3-heteroaryl-2-pyridones [67]. Recently a series of 3-phenyl-6-(2-pyridyl)methoxy-1, 2, 4-triazolo[3, 4-a]phthalazineshave been developed with Compound 62, Fig. (2), showingbinding selectivity for α2, α3 and α5 over α1 subunits andacting as an anxiolytic in the rat elevated plus maze [68]. Itshowed a good pharmacokinetic profile making it a usefultool to explore the effect of a GABAA α2/α3 selectiveagonist in vivo. The sedative effects of the α1-selective agentzolpidem are diminished in the α1 knock-in mouse,consistent with sedation being mediated via α1-containingGABAA receptors [69].

Quinolone antibiotics have long been known to interactwith receptors for GABA and other neurotransmitters [70].Modification of norfloxacin has yielded molecules such asCompound 4, Fig. (2), that positively modulates GABAA

receptors with α2 subunit selectivity and is a non-sedatinganxiolytic [71].

Importance of α5 Subunits in Spatial Memory

GABAA receptor α5 subunits account for less than 5% ofGABAA receptors in the brain. They are localised mainly tothe hippocampus where they may play a key role incognitive processes by controlling a component of synaptictransmission in the CA1 [72]. Mice lacking the α5 geneshow improved performance in the Morris water maze modelof spatial learning, whereas the performance in non-hippocampal-dependent learning and in anxiety tasks wereunaltered in comparison with wild-type controls [73].Novel selective α5 negative allosteric modulators, e.g. 6,6 - dimethyl - 3 - (2 - hydroxyethyl) thio - 1 - (thiazol - 2 - yl) - 6, 7 -dihydro-2-benzothiophen-4(5H)-one (Compound 43, Fig.(2)) have been developed that enhance spatial learning butlack the convulsant or proconvulsant activity associated withnon-selective GABAA receptor negative allosteric modula-tors [74].

Fig. (2). Structures of substances showing selectivity for GABAA receptors containing specific subunits.

NN

NN

O

N

N N

F

F

HO2C

Cl

H2NO2 S

HN

O

OH

NH

N

O

N

HN

L-838,417 Furosemide Salicylidene salicylhydrazide

OH

OO

Cl

'Compound 4'

S

SN

OS(CH2)2OH

'Compound 43'

NN

N

N

O

N

'Compound 62'

HO

GABAA Receptor Channel Pharmacology Current Pharmaceutical Design, 2005, Vol. 11, No. 15 1871

Furosemide and α6-Subunits

The loop diuretic furosemide, Fig (2), has been describedas ‘the most receptor-subtype specific’ allosteric modulatoracting on GABAA receptors [75]. Furosemide acts as anegative allosteric modulator of α6 subunit containingGABAA receptors. Such GABAA receptors are largelyrestricted to the cerebellum and show low sensitivity to theclassic GABAA antagonist bicuculline and to positivemodulation by diazepam, but can be influenced by otherbenzodiazepine receptor ligands [76, 77]. Furosemideexhibits approximately 100-fold selectivity for α6 containingreceptors over α1 containing receptors. It also acts on α4containing receptors. Mutation of a threonine to a isoleucinein the TM1 region of α1 subunits increases furosemidesensitivity by 20-fold [78]. Structure-activity studies showthat the diuretic properties of agents related to furosemideare distinct from the α6 GABAA receptor negativemodulation [79]. The positive allosteric modulator (+)-ROD188 shows selectivity for α6 GABAA receptors [80].The GABAC receptor agonist cis-4-aminocrotonic acid alsoacts as a bicuculline-sensitive agonist at recombinantα6β2γ2S GABAA receptors that are insensitive to theGABAC receptor antagonist TPMPA [81]. Niflumic acid, anonsteroidal anti-inflammatory drug, acts as an antagonist atrecombinant α6β2 receptors in a manner similar to that offurosemide, but also acts as a positive modulator at α1β2γ2receptors [82].

Gene knockout of the α6 subunit in mice resulted in anassociated inhibition of δ subunit expression withoutinfluence on exploratory activity in the open field or learningin a horizontal wire task [83]. Other studies showed the lackof effect of α6 knockout on responses to ethanol,pentobarbital and general anaesthetics [84]. In a rotating rodtest, however, α6 knockout mice were significantly moreimpaired by diazepam than were wild-type mice [85]. Thisdiazepam-induced ataxia in α6 knockout mice could bereversed by flumazenil, indicating the involvement of theremaining α1β2/3γ2 GABAA receptors on the cerebellargranule cells. This led to the conclusion that α6 subunit-dependent actions in the cerebellar cortex could becompensated by other receptor subtypes; but, if not for theα6 subunit, patients on benzodiazepine medication wouldsuffer considerably from ataxic side-effects [85]. There iselectrophysiological evidence for the coexistence of α1 andα6 subunits in a single functional GABAA receptor [86], forfurosemide-sensitive and furosemide-insensitive GABA-mediated effects on cerebellar granule cells [87] and for atonic diazepam-sensitive GABA-mediated inhibition oncultured rodent cerebellar granule cells [88].

Allelic variants in α6 GABAA subunits are associatedwith abdominal obesity and cortisol secretion [89]. In a studyof 100 patients of the effects of midazolam, a point mutation(Pro385Ser) in the α6 GABAA receptor subunit did notaffect baseline sedation, anxiety or memory, but significantlyattenuated the anxiolytic affect of low-dose midazolam [90].

Anaesthesia and Sedation Involving β2 and β3 Subunits

The intravenous general anaesthetic etomidate providesanother example of distinct actions involving differentGABAA receptor subtypes, in this case involving β2 and β3

subunits. Using genetically modified mice with etomidate-insensitive β2 subunits, it was shown that in wild-type miceetomidate produces sedation via the β2 subunit andanaesthesia via the β3 subunit [91]. Furthermore, therecovery of function in the genetically modified mice wasconsiderably improved after etomidate anaesthesia suggest-ing that β3 selective agents could be used as anaestheticswith significantly improved recovery profile [91]. The β2subunit has been shown to mediate the hypothermic effect ofetomidate [92]. Loreclezole and mefenamic acid showsimilar selectivity to etomidate with respect to β2 and β3subunits [93]. Salicylidene salicylhydrazide, Fig. (2), hasbeen shown to be a selective inhibitor of GABAA receptorsthat contain β1 subunits and thus may be a useful agent withwhich to study β subunit selectivity [94].

Splice Variants of γ2 Subunits and Sedation

The γ2 GABAA receptor subunit is generally consideredto be vital to the classical actions of benzodiazepines onGABAA receptors. Alternate splicing results in two splicevariants, a short (γ2S) and a long (γ2L) variant. Mice lackingthe γ2L variant are more sensitive to the sedative effects ofmidazolam and zolpidem, while responses to etomidate andbarbiturates are unchanged [95]. It is suggested that the lackof the γ2L variant may shift the state of α1-containingGABAA receptors from a negative allosteric modulatorpreferring conformation towards a positive allostericmodulator preferring conformation.

Ethanol and δ Subunits

The importance of the δ GABAA receptor subunit hasbeen highlighted by the discovery that ethanol at lowconcentrations known to affect humans enhances the actionof GABA on α4β3δ and α6β3δ receptor subtypes [96].Reproducible ethanol enhancement of GABA responsesoccurred at 3 mM, i.e. concentrations that are reached withmoderate ethanol consumption producing blood-ethanollevels well below the legal level for driving in mostcountries. Ethanol has been long known to influence thefunctioning of a variety of receptors usually at concentra-tions in excess of 50 mM. This had been true for recombi-nant GABAA receptors [97] until the studies on δ subunitcontaining receptors. The δ subunits appear to associatealmost exclusively with α4 and α6 subunits formingfunctional receptors that are 50 fold more sensitive to GABAand desensitise more slowly than receptor subtypes that donot contain δ subunits [83, 98]. The δ subunit protein isexpressed in brain regions expressing α4 (high in thalamus,dentate gyrus, striatum and outer cortical layers and low inhippocampus) and α6 subunit proteins (cerebellum) andappears to be associated with extrasynaptic rather thansynaptic GABAA receptors [99]. Knocking out the δ subunitgene in mice reduces their sensitivity to neurosteroids [100]and increases their susceptibility to seizures [101]. Knockingout the α6 subunit gene in mice does not alter theirsusceptibility to ethanol [84].

GABAA RECEPTOR AGONISTS

Muscimol and THIP, Fig. (3), are widely used asselective GABAA receptor agonists [1]. However, they have


potent actions on GABAC receptors which mean thatinterpretation of studies with these agents should be treatedwith some caution. No “selective” GABAA receptor agonistis known that does not have significant action on eitherGABAB and/or GABAC receptors. Muscimol, a conforma-tionally restricted analogue of GABA in which a hydroxy-isoxazole moiety replaces the carboxyl group of GABA[102], is more potent at GABAC receptors than at GABAA

receptors [103].

THIP and Ionotropic GABA Receptors

THIP (Gaboxadol, 4, 5, 6, 7-tetrahydroisoxazolo(5, 4-c)pyridin-3-ol), Fig. (3), is a conformationally restricted anal-ogue of muscimol [102].

It is a potent GABAA receptor partial agonist of highefficacy [104] that has proved to be a moderately potentGABAC receptor antagonist [103]. Unlike GABA, bothmuscimol and THIP pass the blood-brain barrier on systemicadministration [105]. Muscimol is psychoactive, while THIPis a potent analgesic. Side effects of THIP (includingsedation, dizziness, and blurred vision) meant that it had toolow a therapeutic index to be therapeutically useful as ananalgesic [106, 107]. There is renewed interest in THIP withrespect to sleep therapy [19] as it produces slow wave sleepand reduces spindling activity in non rapid eye movementsleep in humans [108].

It does appear that receptors other than classicalbenzodiazepine-sensitive, bicuculline-sensitive GABAA

receptors are involved in the effects of THIP on painperception and sleep. THIP-induced analgesia is notsensitive to bicuculline indicating that GABAA receptors arenot involved [109]. The GABAC receptor antagonist actionof THIP may contribute to its analgesic action [24]. Theanalgesic action of THIP in rats is blocked by subconvulsantdoses of picrotoxinin [110], a known GABAC receptorantagonist. Benzodiazepine-sensitive GABAA receptors donot appear to be involved in the effects of THIP on sleeppatterns [108]. The GABAC receptor antagonist TPMPA hasbeen used to probe the involvement of GABAC receptors insleep-waking behaviour [111]. The binding of THIP to ratbrain membranes, unlike that of GABA and muscimol, is notstimulated by diazepam [112]. THIP was devoid of theanticonvulsant and antiepileptogenic effects shown bydiazepam and alphaxalone in pentamethylenetetrazole-kindled mice [113].

Clinical studies with THIP have indicated that sleepquality improving effects are obtained at plasmaconcentrations of the order of 1 µM [108]. THIP showsconsiderable variation in potency on recombinant receptors:THIP acts on α1β3γ2S recombinant GABAA receptorsexpressed in oocytes as a partial agonist (EC50 350µM) andmore potently and as a full agonist on α5β3γ3 (EC50 40µM) and α5β3γ3 (EC50 29 µM) recombinant receptors[114]. On α4β3γ2 recombinant receptors THIP acts as apartial agonist (EC50 102 µM) and on α4β3δ as a‘superagonist’ (EC50 6 µM) [115]. On recombinant GABAC

receptors THIP acts as an antagonist (Kb 32 µM for ρ1 [103]and 10 µM for ρ3 receptors [116]. On this basis, α4β3δGABAA and ρ3 GABAC receptors are the most likely GABA

receptors to respond to clinically relevant 1 µM plasmaconcentrations of THIP.

Studies on the interactions between THIP, benzodia-zepines and ethanol in the rat cortical wedge preparationprovide evidence for THIP acting on benzodiazepine-insensitive GABAA receptors in intact tissue possiblycontaining α4 subunits [56]. Rotarod studies on the effects ofTHIP on motor performance in rats showed a lack of cross-tolerance with benzodiazepines [117]. In neither study didethanol show a potentiation of the effects of THIP [56, 117].This is interesting in view of the recent findings discussed insection 4.5. of the sensitivity of α4β3δ GABAA receptors toethanol [96] and may suggest that this receptor subtype is notinvolved in some of the actions of THIP. Studies on theeffects of THIP in δ-subunit knockout mice may help sortout the role of α4β3δ GABAA receptors in THIP-inducedanalgesia and sleep.

The rat cortical wedge preparation has yielded evidenceof the importance of GABAA receptor ligands acting atextrasynaptic receptors [118]. In this preparation THIP actedas a full agonist with an EC50 of 8 µM [118]. In contrast, anunusually low activity of THIP has been reported on GABAreceptors on isolated rat dorsal roots, a tissue that certainlydoes not contain any synapses [119]. THIP was 20 timesweaker than GABA in depolarising these dorsal roots but atleast 20 times more potent than GABA is depressing overallspontaneous synaptic activity in the rat hemisected spinalcord as recorded in the ventral roots. As in the rat corticalwedge preparation, the actions of THIP in the hemisectedspinal cord may involve extrasynaptic GABA receptors, butthe results in isolated dorsal roots clearly show that not allextrasynaptic GABA receptors show high sensitivity toactivation by THIP. Interestingly, extrasynaptic GABAA

receptor channels in hippocampal slices are known to bemodulated by diazepam [120].

Fig. (3). GABAA receptor agonists and partial agonists.

HN

ON

OH

THIP

ON

OH

H2N

Muscimol

ON

HN

OH

ON

HN

OH

H2N

HN

NH

NH2

O

O

4-PIOL 4-Naphthyl-Me-4-PIOL

Compound 5b - "superagonist"


4-PIOL as a Low Efficacy Partial GABAA ReceptorAgonist

Partial agonists offer certain advantages over fullagonists [19]. Full agonists may induce receptor desensitisa-tion that can lead to tolerance and subsequent withdrawalsymptoms. The non-fused THIP analogue 4-PIOL, Fig. (3)(5-(4-piperidyl)isoxazol-3-ol), is a low efficacy GABAA

receptor partial agonist that exhibits a predominantlyantagonist profile [19]. Its activity varies with differentrecombinant GABAA receptor subtypes and in the rat corticalwedge preparation, 4-PIOL behaves as a high efficacy partialagonist [118]. Patch clamp studies on hippocampal neuronesshow that 4-PIOL is a non-desensitising partial agonistwhose action can be potentiated by benzodiazepines andbarbiturates [121]. Studies on analogues of 4-PIOLsubstituted in the 4-position of the isoxazole ring yieldedGABAA receptor antagonists of increased potency, e.g. 4-naphthyl-methyl-4-PIOL, Fig. (3), with a linear correlationbetween the lipophilicity of the 4-subsitutent and antagonistactivity, providing evidence for a hydrophobic bindingpocket at the GABA recognition site [122].

“Superagonists” at GABAA Receptors

Studies on a series of GABA amides revealed substancesthat could act as partial, full or superagonists as assessed bythe stimulation of chloride influx into mouse brainsynaptoneurosomes in a bicuculline- and picrotoxinin-sensi-tive manner [123]. Compound 5b (N, N’-1, 4-butanediylbis[4-aminobutanamide]), Fig. (3), produced a maximumresponse that was 150% that of GABA and was thus descriedas a “superagonist”. It showed similar affinity to THIP, apartial agonist with a maximum response 65% that ofGABA. The apparent ‘superagonist’ action of compound 5bcould mean that GABA is in fact a partial agonist in theseexperiments. While the exact nature of this “superagonist”action remains to be determined in functional assays usingrecombinant GABAA receptors of known subunitcomposition, the concept of ‘superagonists’ opens up anotherpossible approach to therapeutic agents acting at GABAA

receptors. THIP has been shown to display superagonistbehaviour at α4β3δ receptors with a maximum response160% that of GABA [124].

STEROIDS THAT INFLUENCE GABAA RECEPTORS

A variety of steroids are known to influence GABAA

receptors via non-genomic actions that are rapid in onset andoffset. These neuroactive steroids include neurosteroids (i.e.steroids that are synthesised in the brain), sex steroids thatoriginate in the gonads, and corticosteroids that are made inthe adrenal cortex, together with a range of synthetic steroidsand steroid analogues [16]. Most interest is centred onneuroactive steroids that act as potent positive allostericmodulators and have anxiolytic, anticonvulsant, analgesic,anaesthetic and sedative actions [17].

The CNS depressant action of steroids has been knownsince 1927 when it was shown that injection of a colloidalsuspension of cholesterol into cats caused deep anaesthesia[125]. Subsequently, cholesterol was found to potentiate theanaesthetic actions of pentobarbitone [126], but it was notuntil the extensive investigations of Seyle [127] that it

became apparent that a wide range of natural and syntheticsteroids have anaesthetic actions.

The synthetic steroid anaesthetic alphaxalone, Fig. (4),was the first steroid shown to act as a positive modulator ofGABAA receptors [128]. This was followed by the discoverythat steroid hormone metabolites, e.g. 3α-hydroxy-5α-pregnan-20-one, Fig, (4), that occur in the brain are‘barbiturate-like modulators’ of the GABAA receptor [129].This led to the concept that neurosteroids can directlymodulate GABAA receptors on the cell surface rather thanacting on receptors in the nucleus regulating geneexpression. Steroids produced outside the brain are alsoimportant modulators of GABAA receptors. For example,THDOC, Fig (4), 3α, 21-dihydroxy-5α-pregnan-20-one, is a‘neuroactive steroid’ because the sole source of this steroidappears to be the adrenals. Nonetheless, THDOC is found inthe brain where its concentration is increased during stress[130]. 3α-Hydroxy-5α-pregnan-20-one and THDOC areamong the most potent known steroid modulators of GABAA

receptors.

Fig. (4). Steroids that act on GABAA receptor function.

Studies using 3α-hydroxy-5α-pregnan-20-one indicatethat its positive modulatory actions on GABAA receptors areonly modestly influenced by the α-, β- or γ-subunits, [131].The inclusion of either an ε or δ subunit however dramati-cally alters the response with the ε subunit reducing and theδ subunit augmenting the efficacy of modulation by 3α-hydroxy-5α-pregnan-20-one [131]. The enhanced efficacy ofδ subunit containing GABAA receptors has also beenreported for THDOC [132]. As noted above, knocking out

O

HHO

5α-Pregnan-3α-ol-20-one

O

HHO

THDOC

OH

O

HHO

Alphaxalone

O

O

HHO

Ganaxolone

O

O

Cortisol

HO OHOH

OH

O

Nandrolone


the δ subunit gene in mice reduces their sensitivity to neuro-steroids [100]. These findings clearly distinguish steroidpositive modulation of GABAA receptors from flumazenil-sensitive positive modulation by benzodiazepines but dosuggest some similarities with the modulation induced byvolatile anaesthetics and ethanol. Experiments with alphax-alone on chimeric GABAA receptors indicate that the site ofaction for steroids is not the same as that for volatileanaesthetics and ethanol [133]. Studies comparing thepositive modulator effects on α6β3γ2L GABAA receptors of4 structurally distinct general anaesthetics – propofol,pentobarbitone, etomidate and 3α-hydroxy-5α-pregnan-20-one – showed that the action of all but 3α-hydroxy-5α-pregnan-20-one depended critically on a single amino acid inTM2 [134].

A novel neuroactive steroid, 6-aza-3α-hydroxy-5β-pregnan-20-one, has been used to photoaffinity label ratbrain membranes [135]. It labelled a protein identified asvoltage-dependant anion channel-1 (VDAC-1) that co-immunoprecipitated with the β2 and β3 subunits of theGABAA receptor, suggesting that neuroactive steroids maymodulate GABAA receptor function by binding to VDAC-1as an accessory protein [135]. Further studies using VDAC-1deficient mice have suggested that VDAC-1 is unlikely to beinvolved in steroid modulation of GABAA receptors [136].

In addition to alphaxalone, there are a number ofsynthetic steroids that are known to modulate GABAA

receptor function. Anabolic steroids such as nandrolone, Fig.(4), and stanazolol induce region- and subunit-specific rapidmodulation of GABAA receptor-mediated currents in the ratforebrain [137]. The antiepileptic agent, ganaxolone, Fig.(4), belongs to a novel class of neuroactive steroids calledepalons which specifically modulate GABAA receptors in thecentral nervous system (CNS). Chemically related toprogesterone but devoid of any hormonal activity, theepalons have potent antiepileptic, anxiolytic, sedative andhypnotic activities in animals [138]. Ganaxolone hasdemonstrated outstanding efficacy and better tolerability inchildren with intractable infantile spasms [139]. It has,however, been reported to exacerbate absence seizures inanimal models [140].

Cortisol, Fig. (4), is a potent bidirectional modulator ofthe action of GABA on GABAA receptors in the guinea-pigileum enhancing at low (1-10 pM) concentrations andinhibiting at higher (10-1000 nM) concentrations [141].Cortisone is a potent non-competitive inhibitor of theseGABAA receptors acting at concentrations as low as 1 pM[142]. These corticosteroids are thus the most potent agentsmodulating GABAA receptors. The actions of cortisol maybe restricted to particular GABAA receptor subtypes sincecortisol has little effect on GABAA responses in the ratcuneate nucleus [143]. Biphasic effects of corticosteroidshave been described on TBPS binding to rat brainmembranes, low (nM) concentrations enhancing binding andhigher (µM) concentrations inhibiting, the effect of nMconcentrations indicative of an antagonist action as observedat these concentrations on GABA responses in the guinea-pig ileum [144]. Cortisol (10 µM) has been shown to rapidlyincrease the spontaneous firing frequency of neurones in ratparaventricular nucleus and to inhibit whole cell potassium

currents, suggesting the cortisol may act indirectly viainactivating potassium channels [145]. High affinity bindingsites (nM) for corticosterone have been described on brainmembranes [146], and corticosterone is known to influencethe expression and activity of GABAA receptors in thehippocampus [147]. Given the risk of memory decline inpatients on corticosteroids [148], further investigations of theeffects of these steroids on GABAA receptors seemwarranted.

NATURAL PRODUCTS AND GABAA RECEPTORS

With increasing community acceptance of herbalmedicines and functional foods there is increasing interest innatural products that may influence brain function. There is aview that natural substances are inherently safer thanunnatural substances, i.e. synthetic chemicals. This view ismistaken as many of the most toxic chemicals are in factnatural products and the majority of therapeuticallybeneficial drugs are synthetic. It is the molecular structureand dose that determine the effects of substances on humanhealth, not whether they are of natural or synthetic origin[149]. There is now an impressive array of natural productsin addition to steroids that are known to influence GABAA

receptor function including substances found in beveragessuch as tea, red wine and whiskey, and in herbal preparationsincluding Ginkgo biloba and Ginseng [57]. Naturalsubstances represent a rich diversity in chemical structuresthat can lead to the development of new therapeutic agents.

Flavonoids and GABAA Receptors

Flavonoids are found in all plants in high abundance andexhibit a considerable chemical diversity with more than5,000 different flavonoids having been described. Fruits,vegetables, and beverages such as tea and red wine are majorsources of flavonoids our diet [150]. It has been estimatedthat the average daily intake of flavonoids is 1-2 g [151].Many flavonoids are polyphenolic and are thus stronglyantioxidant [152]. They have a wide variety of biologicalactivities and are being studied intensively as anticanceragents [153]. Flavonoids have a range of activities onGABAA receptors [154].

Flavonoids were first linked to GABAA receptors whenthree isoflavans isolated from bovine urine were shown toinhibit diazepam binding to brain membranes [155]. Themost potent compound was 3’, 7-dihydroxyisoflavan, Fig.(5), with an IC50 of 45 µM. Further studies searching fordiazepam-like substances using benzodiazepine bindingassays led to the discovery that the biflavonoid amento-flavone, Fig. (5), sometimes known as biapigenin, wascapable of displacing benzodiazepine binding to rat brainmembranes with a nM affinity comparable to that ofdiazepam [156]. These investigations used amentoflavoneisolated from Karmelitter Geist, an alcoholic tincture ofvarious plants used to treat anxiety and epilepsy. However itwas concluded that amentoflavone cannot be responsible forany pharmacological effects of the plant extract asamentoflavone did not influence flunitrazepam binding in thebrain in vivo following i.v. administration to mice [156]. Itwas suggested that amentoflavone was either rapidlymetabolised or did not cross the blood brain barrier, but a


recent study does indicate that amentoflavone does cross theblood brain barrier [157]. Amentoflavone occurs in a varietyof herbal preparations including St John’s wort [158] andGinkgo biloba [159]. A comprehensive battery of in vitrobinding assays has shown that amentoflavone influences avariety of G-protein coupled receptors for serotonin,dopamine and opioids at nM concentrations while having noeffect on the binding of muscimol to GABAA receptors[160]. Using recombinant α1β2γ2L GABAA receptorsexpressed in oocytes, amentoflavone has been shownrecently to be a relatively weak (4 µM) negative allostericmodulator of GABA action acting independently of classicalflumazenil-sensitive benzodiazepine modulatory sites [159].These studies on amentoflavone illustrate the difficulties ofstudying flavonoid actions – the variety of effects, the lack ofselectivity, the need for functional assays and the mismatchbetween in vitro and in vivo findings.

Apigenin, Fig. (5), a component of Matricicaria recutitaflowers (chamomile), has been characterised as a centrally-acting benzodiazepine ligand with anxiolytic effects [161].Infusions of chamomile flowers are widely used as a tea topromote sleep. Apigenin competitively inhibited (Ki 4 µM)the binding of flunitrazepam to brain membranes withoutinfluencing the binding of muscimol to GABAA receptors.Apigenin was described as having ‘a clear anxiolytic effectin mice in the elevated plus maze without evidencingsedation or muscle relaxation effects at doses similar to those

used for classical benzodiazepines’ and it was devoid ofanticonvulsant effects [161]. These finding are in contrast toa later study in rats where apigenin was shown to reduce thelatency of onset of picrotoxin-induced convulsions and toreduce locomotor activity but was devoid of anxiolytic ormuscle relaxant activities [162]. This later study showed thatapigenin could reduce GABA-activated chloride currents incultured cerebellar granule cells, an action that could beblocked by flumazenil and thus likely to involve classicalbenzodiazepine allosteric sites on GABAA receptors. Theinhibitory action of apigenin on locomotor behaviour,however, could not be blocked by flumazenil and thus couldnot ‘be ascribed to an interaction with GABAA-benzodia-zepine receptors, but to other neurotransmitter systems’[162]. Another study from the same group reported thatapigenin exerted sedative effects on locomotor activity inrats in a flumazenil-insensitive manner, whereas chrysin, astructurally related flavonoid lacking the 4’-hydroxysubstituent of apigenin, showed a clear flumazenil-sensitiveanxiolytic effect in addition to the flumazenil-insensitivesedation [163]. The apparent discrepancy between thebehavioural effects of apigenin on mice [161] and rats [162]may be due to mice having higher baseline levels of anxiety.

Recent studies on recombinant receptors in oocytes haveshown that µM apigenin inhibited the activation of α1β1γ2SGABAA receptors in a flumazenil-insensitive manner andhad a similar effect on ρ1 GABAC receptors [164]. Other

Fig. (5). Flavonoids that act on GABAA receptors.

O

OH

HO

3',7-Dihydroxyisoflavan

OHO

OOHOH

Genistein

OHO

OOH

OOH

HO OH

O

OH

Amentoflavone

OHO

OOH

R1

R2

Q

O

O

R1

NO2

R1=H, R2=OH ApigeninR1=MeO, R2=OH HispidulinR1=MeO, R2=H Oroxylin AR1=Me, R2=OH 6-Methylapigenin

R=NO2 3',6-DinitroflavoneR=Cl 6-Chloro-3'-nitroflavone

Graham Johnston

Placed Image


studies on recombinant α1β2γ2L GABAA receptors describean inhibitory effect of apigenin on GABA responses and, inaddition, describe an enhancement of the diazepam-inducedpositive allosteric modulation of GABA responses byapigenin [57, 165]. Such a second order modulation byapigenin of benzodiazepine modulation of the activation byGABA of GABAA receptors may indicate that apigeninneeds to work through an endogenous benzodiazepinesystem to influence behaviour in a flumazenil-sensitivemanner. Overall, it seems that the effects of apigenin onGABAA receptors are complex and involve both flumazenil-sensitive and flumazenil–insensitive components, and thatother receptors could be involved in the behavioural effectsof apigenin. Genistein, Fig. (5), the isoflavone equivalent ofapigenin, is a phytoestrogen with a wide variety ofpharmacological effects on animal cells [166]. It is widelyused as a tyrosine kinase inhibitor but its action as a negativemodulator of the action of GABA on recombinant GABAA

receptors is the result of a direct action on the receptors andis independent of tyrosine kinase [167, 168].

Hispidulin, Fig. (5), 4’, 5, 7-trihydroxy-6-methoxyflav-one, i.e. the 6-methoxy derivative of apigenin), was isolatedtogether with apigenin from Salvia officinalis (Sage) recentlyusing a benzodiazepine binding assay-guided fractionation[169]. Hispidulin was some 30 times more potent thatapigenin in displacing flumazenil binding. Preparations ofsage have been used in herbal medicine to assist memory[170, 171] and an extract of Salvia lavandulaefolia (Spanishsage) has been shown to enhance memory in healthy youngvolunteers [172]. Hispidulin has been shown to act as apositive allosteric modulator of α1, 3, 5, 6β2γ2S GABAA

receptor subtypes showing little subtype selectivity being alittle more potent at α1, 2, 5β2γ2S subtypes than at α3,6β2γ2S subtypes [173]. The positive modulatory action of 10µM hispidulin at α1β2γ2S receptors was reduced from 47%to 17% by flumazenil, indicating that sites other thanclassical flumazenil-sensitive benzodiazepine sites wereinvolved in the action of hispidulin. As hispidulin did notinfluence the action of GABA on α1β2 GABAA receptors,hispidulin does not interact with low affinity flumazenil-insensitive benzodiazepine sites [174] in contrast to otherflavonoids such as 6-methylflavone [175]. Of significance isthe ability of hispidulin to act as a positive modulator atα6β2γ2L GABAA receptors unlike diazepam; 10 µMhispidulin enhanced the action of GABA at these receptorsby 65%, this action being reduced by 1 µM flumazenil to37% [173]. Hispidulin was shown to have an anticonvulsantaction in seizure prone Mongolian gerbils and to pass theblood brain barrier [173]. Flavonoids structurally related tohispidulin, and that influence benzodiazepine binding, havebeen isolated from Scutellaria baicalensis, an important herbin traditional Chinese medicine [176]. Oroxylin A, Fig. (5),5, 7-dihydroxy-6-methoxyflavone, i.e. hispidulin lacking the4’-hydroxy group), inhibits flunitrazepam binding at 1 µMand on oral administration as a neutralising allostericmodulator blocking the anxiolytic, myorelaxant and motorincoordination effects, but not the sedative and anticon-vulsant effects elicited by diazepam [177]. 6-Methyla-pigenin, Fig. (5), 4’,5,7-dihydroxy-6-methylflavone) isolatedfrom Valeriana wallichii, a known sedative herb, influencesbenzodiazepine binding at 0.5 µM in manner suggesting it

may be a positive modulator of GABAA receptors [178].Thus, flavones substituted in the 6-position with a methoxyor methyl substituent have interesting effects on GABAA

receptor function and may contribute to the properties ofsome herbal preparations. Natural and synthetic 2’-hydroxy-substituted flavones are also of interest [179]. Severalflavonoid glycosides including goodyerin [180], linarin andhesperidin [181] are also being studied as sedative andanticonvulsant agents likely to interact with GABAA

receptors.

Using a combinatorial chemistry approach, a range ofrelatively simple flavones have been synthesised andevaluated initially for activity in a benzodiazepine bindingassay [182]. This approach led to some very interestingcompounds that were further evaluated in a variety ofpharmacological tests as GABAA receptor ligands [154]. Themost active anxiolytic flavone was 3’, 6-dinitroflavone, Fig.(5), which was 30 times more potent than diazepam. It wasorally active and had minimal sedative action at anxiolyticdoses. In contrast, 6-chloro-3’-nitroflavone, Fig. (5), had noanxiolytic properties and abolished the anxiolytic,anticonvulsant and amnesic effects of diazepam [154].

There have been extensive structure-activity studiesaimed as developing models of flavonoid pharmacophoresfor their interaction with GABAA receptors [183-187]. Aproblem common to most of these studies is that the activitydata is based on ligand binding studies to what is now knownto be a mixture of benzodiazepine binding sites. Given ourincreased knowledge of the diversity of benzodiazepine andflavonoid actions on cloned receptors of defined subunitcomposition, future structure-activity studies need to bebased on data from functional studies on GABA receptors ofknown subunit composition [175].

Terpenoids and GABAA Receptors

Terpenoids are widespread in plants, especially in whatare known as essential oils that can be extracted from plantsand have a wide range of uses from perfume constituents topaint thinners. Terpenoids are oxygenated products formallyderived from C5 isoprene units and are classified by thenumber of C5 units in their structure. Thus monoterpenoidshave 2xC5 units, sesquiterpenoids 3xC5 units, diterpenoids4xC5 units and triterpenoids 6xC5 units. The most widelyused terpenoid in studies on GABAA receptors is thesesquiterpenoid lactone picrotoxinin, Fig. (6), a non-competitive antagonist at GABAA receptors [1]. A number ofother terpenoids, however, are of interest for their actions onGABAA receptors.

Bilobalide, Fig. (6), a sesquiterpenoid lactone fromGinkgo biloba that bears some structural similarities topicrotoxinin, including a lipophilic side chain and ahydrophilic cage, is also a non-competitive antagonist atGABAA receptors [188]. Both bilobalide and picrotoxininappear to act at sites in the chloride channel of GABAA

receptors and are thus negative allosteric modulators. Thecognition-enhancing effects of Ginkgo extracts may be partlymediated by bilobalide acting to enhance hippocampalpyramidal neuronal excitability [189]. While picrotoxinin isa convulsant, bilobalide is an anticonvulsant [189, 190]. Aswith the α5 subunit preferring negative allosteric modulator


mentioned in Section 4.2, the lack of convulsant action in anagent that reduces GABA action may be important forenhancement of cognition. The lack of convulsant action ofbilobalide may result from subunit selectivity but this has yetto be established. The structurally-related ginkgolides,especially ginkgolide B, also act as negative modulators atGABAA receptors [191]. They also inhibit strychnine-sensitive glycine receptors and platelet activating factor[191, 192]. Bilobalide and the ginkgolides reducebarbiturate-induced sleeping time in mice, an effect perhapsrelevant to the clinically observed ‘vigilance-enhancing’ andantidepressant-like actions of Ginkgo extracts [193].

The monoterpenoid α-thujone, Fig. (6), is a psychoactivecomponent of absinthe, a liqueur popular in France in the19th and early 20th centuries. It is found in extracts ofwoodworm and some other herbal medicines and beveragessince ancient Egyptian times [194]. α-Thujone is a convul-sant that acts as a negative allosteric modulator of GABAA

receptors [195]. It also acts as an antagonist of 5HT3 recep-tors by influencing agonist-induced desensitisation [194].

The structurally-related substance thymol, Fig. (6), aconstituent of thyme essential oil, is a flumazenil-insensitivepositive allosteric modulator of GABAA receptors [196]. Athigher concentrations, thymol had a direct action on GABAA

receptors similar to that of the anaesthetic propofol and otherphenols [197]. The anticonvulsant effects of thymoquinone,the major constituent of Nigella sativa seeds, may be due topositive modulation of GABAA receptors [198].

(+)-Borneol, Fig. (6), a monoterpenoid found in manyessential oils, is a flumazenil-insensitive positive allostericmodulator of recombinant GABAA receptors of low affinitybut very high efficacy producing 12 fold enhancement of theaction of 10 µM GABA at a concentration of 450 µM [57,199]. (+)-Borneol is found in high concentrations in extractsof Valerian officinalis that are widely used to reduce thelatency of sleep onset, the depth of sleep and the perceptionof well-being. Extracts of Valerian are known to contain alarge number of constituents including flavonoids andterpenoids, many of which are considered to be active atGABAA receptors. The sesquiterpenoid valerenic acid, Fig.(6), has a direct partial agonist action on GABAA receptors[200]. Isocurcumenol, Fig. (6), a sesquiterpenoid fromCyperus rotundus, was found to inhibit [H-3]Ro15-1788binding and enhance [H-3]flunitrazepam binding in thepresence of GABA in a manner consistent with it acting as apositive allosteric modulator [201].

Ginsenosides, triterpenoid glycosides that are the majoractive constituents of Panax ginseng, are known tonegatively modulate nicotinic and NMDA receptor activity.Of a series of ginsenosides, ginsenoside Rc was the mostpotent (EC50 53 µM) in enhancing the action of GABA onrecombinant α1β1γ2S GABAA receptors expressed inoocytes [202].

Sage and GABAA Receptors

Sage has been used widely to treat memory deficits andextracts of Salvia lavandulaefolia (Spanish Sage) have been

Fig. (6). Terpenoids that act on GABAA receptors.

O

α-Thujone

HO

Thymol

O

Thymoquinone

O

HO

H

OHO

lsocurcumenol (+)-Borneol Valerenic acid

O

O

O

OH O

O

Picortoxinin Bilobalide

OO H

OH

O

O

HO

H

O

H

O

COOH


shown to enhance memory in healthy young volunteers[172].

In addition to the flavonoids apigenin, hispidulin andlinarin (see section 7.1.), a number of terpenoids have beenextracted from varieties of Salvia (sage) that influencebenzodiazepine binding [169]. The diterpenoid lactonegaldosol, Fig. (7) from the common sage Saliva officinalis,inhibited flumazenil binding at 0.8 µM [169].

The diterpenoid quinone miltirone, Fig. (7), from theChinese medicinal herb Salvia miltriorrhiza, inhibitedflunitrazepam binding at 0.3 µM and was orally active inanimal models as a tranquilliser without muscle relaxantproperties [203]. Structure-activity studies on miltirone ledto the development of a synthetic compound that was muchmore potent than miltirone on flunitrazepam binding (IC500.05 µM) [204].

The structurally-related diterpenoids carnosic acid andcarnosol, Fig. (7), extracted from Salvia officinalis, while notinfluencing diazepam or muscimol binding, did inhibit TBPSbinding [205]. This suggests that, like flavonoids,diterpenoids can influence GABAA receptors in a mannerindependent of classical benzodiazepine sites and could bemissed in benzodiazepine binding assays. The structures ofgaldosol, carnosic acid and carnosol, Fig. (7), contain the o-isopropylphenolic moiety that is present in thymol, Fig. (6),and the anaesthetic agent propofol.

Fig. (7). Diterpenoids from Salvia that influence GABAA receptors.

Sage also contains α-thujone, Fig. (6), a known GABAA

receptor antagonist as noted above, which may influence theGABA enhancing effects of hispidulin, galdosol, miltirone,carnosic acid, carnosol and related compounds in sageextracts. The levels of α-thujone in individual sage plants areknown to vary considerably [206].

Dietary and Environmental Chemicals that mayInfluence GABAA Receptors

There are many chemicals in food and beverages,together with other chemicals in our environment, that areknown to be capable of influencing the function of GABAA

receptors in addition to the flavonoids, terpenoids andethanol discussed in previous sections. GABA itself occurswidely in plants being involved in pH regulation, nitrogenstorage, plant development and defence [207]. High levels ofGABA in plants extracts may be a confusing factor inevaluating the effects of such extracts on GABAA receptorsin binding or functional assays, but such GABA will notnormally influence GABAA receptors in the brain oningestion due to the blood brain barrier.

Tea and coffee contain a range of chemicals in additionto GABA that have been shown to influence recombinantbovine α1β1 GABAA receptors. Extracts of green, oolong orblack tea contained catechins, especially (-)-epicatechingallate and (-)-epigallocatechin gallate, that inhibited GABAresponses and alcohols, such as leaf alcohol and linalool,Fig. (8), that enhanced GABA responses at concentrations of1 mM [208]. Coffee extracts contained theophylline, whichinhibited GABA responses in a non-competitive mechanism(Ki 0.55 mM), and theobromine, which inhibited in acompetitive manner (Ki 3.8 mM), while a number ofcompounds including 1-octen-3-ol and sotolone, Fig. (8),enhanced GABA responses [209]. When 1-octen-3-ol (100mg/kg) was orally administered to mice prior tointraperitoneal administration of pentobarbitone, the sleepingtime of mice induced by pentobarbital increased significantly[209]. Sotolone is a key component in the “nutty” and“spicy-like” aroma of oxidative aged port wine [210]. Manycomponents in the fragrance of whiskey, in particular ethyl3-phenylpropanoate, Fig. (8), strongly enhanced GABAA

responses [211]. When applied to mice through respiration,ethyl 3-phenylpropanoate delayed the onset of convulsionsinduced by pentylenetetrazole. The extract of other alcoholicdrinks such as wine, sake, brandy, and shochu alsopotentiated GABA responses to varying degrees [211].Although these fragrant components are present in alcoholicdrinks at low concentrations (extremely small quantitiescompared with ethanol), they may also modulate the moodor consciousness through the potentiation of GABAA

responses after absorption into the brain because thesehydrophobic fragrant compounds are easily absorbed into thebrain through the blood-brain barrier and are severalthousands times as potent as ethanol in the potentiation ofGABAA receptor-mediated responses [211]. The aging ofwhiskey results in enhanced potency of the fragrance inpotentiating GABAA responses and in prolongingpentobarbitone-induced sleeping time in mice [212]. As allof the above tests on chemicals from tea, coffee and whiskeywere carried out on recombinant bovine α1β1 GABAA

receptors, the observed effects are independent of classicalbenzodiazepine-sensitive sites. Several perfume constituentshave been shown to act as positive modulators of GABAA

receptors including the terpenoids eugenol, citronellol andhinokitol, Fig. (8) [213]. The tea flavonoid (-)-epigallo-catechin gallate has been shown to be some 10 times morepotent than apigenin as a second order modulator of thepositive modulation by diazepam of α1β2γ2L humanrecombinant GABAA receptors expressed in oocytes [165].

Simple disaccharides are able to enhance TBOB bindingto GABAA receptors [214]. Lactose (EC50 1.5 µM) was 100-600 fold more potent than maltose or sucrose. Lactose didnot influence flunitrazepam binding and its effects on TBOB

O

O

HO

OH

OO

HO

OH

HO2 C

Miltirone Galdosol

Carnosic acid

HO

OH

Garnosol

O

O


binding could be blocked by GABA. Regional differences inthe potency of lactose enhancement of TBOB bindingsuggest that the effect might be GABAA receptor subtypeselective [214].

As noted in section 6, cholesterol has been long known toproduce deep anaesthesia in cats following injection of acolloidal suspension [125] and to potentiate pentobarbitone-induced anaesthesia [126]. Dietary cholesterol and agentsthat alter cholesterol levels may influence GABAA receptorfunction in the brain. Alterations in membrane cholesterol indissociated hippocampal neurones alters GABAA receptorproperties [215]. Cholesterol enrichment increased thepositive modulatory effects of the nonsteroidal agentspropofol, flunitrazepam and pentobarbitone but reduced thepositive modulatory effects of the steroids pregnanolone andalphaxalone. Depletion of membrane cholesterol increasedthe effects of pregnanolone and alphaxalone withoutinfluencing the effects of the nonsteroidal modulators.Increases in dietary cholesterol in rats has been shown todepress brain waves as measured by EEG [216], an effectthat may be attributable to changes in GABAA receptorfunction. There has been speculation about an associationbetween brain cholesterol and Alzheimer’s disease and thesuggestion that cholesterol-lowering strategies influence theprogression of this disease [217]. There is some evidencethat statins can reduce cholesterol turnover in the brain thusenabling statins to reduce the incidence of Alzheimer’sdisease [218]. Such treatments might influence GABAA

receptor function through alteration of cholesterol levels inthe brain.

Inhaled drugs of abuse such as the solvents toluene, 1, 1,1-trichloroethane and trichlorethylene act as positivemodulators of GABAA receptors, acting in a mannersuggesting that their sites of action may overlap with thoseof ethanol and volatile anaesthetics [219]. Another study,however, found toluene to be an antagonist of the activationof GABAA receptors [220].

CONCLUSIONS

When I reviewed GABAA receptor pharmacology in1996 [8] the then literature prompted a conclusion that thereappeared to be at least 11 distinct sites on GABAA receptorsfor interactions with specific ligands. The likely sites were:

(1) agonist/partial agonist/competitive antagonist recognitionsites; (2) picrotoxinin sites; (3) sedative-hypnotic barbituratesites; (4) neuroactive steroid sites; (5) benzodiazepine sites;(6) ethanol sites; (7) sites for inhalation anaesthetics; (8) sitesfor furosemide associated with α6 subunits; (9) sites forZn2+; (10) sites for a variety of divalent cations, such asCa2+, Sr2+, Ba2+, Cd2+, Mn2+, and Mg2+; and (11) sites forLa3+. I noted that it was likely that there were subtypes ofneuroactive steroid sites and that there were certainlysubtypes of benzodiazepine sites. In addition, I noted thatthere were possibly sites associated with (a) phospholipidsinteracting with GABAA receptor protein subunits, (b)cyclic nucleotide protein kinase activity involvedphosphorylation of the intracellular loop of some GABAAreceptor protein subunits, and (c) the interaction of GABAAreceptors and microtubules that may anchor receptor clustersat postsynaptic membranes.

The situation has become even more complex since 1996.Thanks to the use of genetically modified mice we nowknow the importance of the different types of GABAA

receptor subunits for the actions of particular agents – e.g.α1 and α2 subunits for the sedative and anxiolytic actions ofbenzodiazepines respectively, α5 subunits for agentsinfluencing spatial memory, and δ subunits for the potentaction of ethanol. Thanks to the use of recombinant receptortechnology, expressing GABAA receptors of known subunitcomposition, we now have detailed knowledge of the actionsof the increasingly chemically diverse range of naturalproducts on GABAA receptor function. We know offlavonoids that influence GABAA receptor function aspositive and negative allosteric modulators, some of theseactions of flavonoids being sensitive to the classicalbenzodiazepine antagonist flumazenil and other actionsbeing insensitive. It is a similar story with terpenoids. Thus itis likely that there are at least two distinct sites, flumazenil-sensitive and flumazenil-insensitive, on GABAA receptorsfor flavonoids and terpenoids of which the flumazenil-sensitive sites may overlap with classical benzodiazepinesites.

It appears likely that the various proposed sites onGABAA receptors overlap and interact making it difficult toput a meaningful figure on the number of distinct sitesthough my 1996 figure of 11 now seems very conservative.We await detailed three-dimensional structural information

Fig. (8). Volatile substances in beverages and perfumes that influence GABAA receptor.

HO

HO

OH OH

OMe

HOEtO

O

O

OH

OO

HO

Leaf alcohol Citronellol Linalool 1-Octen-3-ol

Eugenol Ethyl 3-phenylpropanoate Hinekitol Sotolone


on the various subtypes of GABAA receptors in order towork out exactly where all of these chemically diverseligands interact. Such information is starting to emergethrough homology modelling based on a 4Å resolutionstructure of a nicotinic receptor [221], e.g. the structure ofthe proposed propofol binding site involving the M2 and M3regions of GABAA receptors [222].

The diversity of sites on GABAA receptors representtargets for the further development of specific agents actingon particular GABAA receptor subtypes. The structures ofthe various ligands described in this and other reviews serveas leads for the discovery of new chemical entities for thetreatment of disorders involving specific GABAA receptors.

ACKNOWLEDGEMENTS

The author is grateful to his many colleagues who havecontributed to his studies on GABAA receptors includingRobin Allan, Erica Campbell, Mary Chebib, Rujee Duke,Renee Granger, Belinda Hall, Jane Hanrahan, ShelleyHuang, Povl Krogsgaard-Larsen, Ken Mewett, Hue Tran andPaul Whiting, and to the Australian National Health andMedical Research Council and Polychip Pharmaceuticals forfinancial support.

REFERENCES

[1] Chebib M, Johnston GAR, GABA-activated ligand gated ionchannels: Medicinal chemistry and molecular biology. J Med Chem2000; 43: 1427-1447.

[2] Bormann J. The 'ABC' of GABA receptors, Trends Pharmacol Sci2000; 21: 16-19.

[3] Le Novere N, Changeux JP. The Ligand Gated Ion Channeldatabase: an example of a sequence database in neuroscience, Phil.Trans. Roy Soc London - Series B: Biol Sci 2001; 356: 1121-1130.

[4] Barnard EA, Skolnick P, Olsen RW, Möhler H, Sieghart W, BiggioG, et al . International Union of Pharmacology - XV - Subtypes ofγ-aminobutyric acidA receptors - classification on the basis ofsubunit structure and receptor function. Pharmacol Rev 1998; 50:291-313.

[5] Vacher CM, Bettler B, GABAB receptors as potential therapeutictargets, Curr. Drug Targ. CNS Neurol Disord 2003; 2: 248-259.

[6] Rotolo TC, Dacheux RF. Two neuropharmacological types ofrabbit ON-alpha ganglion cells express GABAC receptors. VisNeurosci 2003; 20: 373-84.

[7] Whiting PJ. GABAA receptor subtypes in the brain: a paradigm forCNS drug discovery? Drug Discovery Today 2003; 8: 445-450.

[8] Johnston GAR. GABAA receptor pharmacology. Pharmacol Therap1996; 69: 173-198.

[9] McKernan RM, Whiting PJ. Which GABAA-receptor subtypesreally occur in the brain?. Trends Neurosci 1996; 19: 139-143.

[10] Burt DR. Reducing GABA receptors. Life Sci 2003: 73: 1741-1758.

[11] Smith AJ, Simpson PB. Methodological approaches for the studyof GABAA receptor pharmacology and functional responses. AnalBioanal Chem 2003; 377: 843-851.

[12] Rosahl TW. Validation of GABAA receptor subtypes as potentialdrug targets by using genetically modified mice. Curr Drug TargCNS Neurol Disord 2003; 2: 207-212.

[13] Möhler H, Fritschy JM, Rudolph U. A new benzodiazepinepharmacology. J. Pharmacol Exp Therap 2002; 300: 2-8.

[14] Rudolph U, Crestani F, Möhler H. GABAA receptor subtypes:dissecting their pharmacological functions. Trends Pharmacol Sci2001; 22: 188-194.

[15] Sieghart W. Unravelling the function of GABAA receptor subtypes.Trends Pharmacol Sci 2000; 21: 411-413.

[16] Hamilton NM. Interaction of steroids with the GABAA receptor.Curr Top Med Chem 2002; 2: 887-992.

[17] Lambert JJ, Belelli D, Peden DR, Vardy AW, Peters JA.Neurosteroid modulation of GABAA receptors. Prog. Neurobiol2003; 71: 67-80.

[18] Korpi ER, Grunder G, Luddens H. Drug interactions at GABAA

receptors, Prog Neurobiol 2002; 67: 113-159.[19] Krogsgaard-Larsen P, Frølund B, Liljefors T, Specific GABA-A

agonists and partial agonists. The Chemical Record 2002; 2: 419-430.

[20] Kneussel M, Dynamic regulation of GABAA receptors at synapticsites. Brain Res. - Brain Res Rev 2002; 39: 74-83.

[21] Sigel E, Dodd RH. Novel positive allosteric modulators of GABAA

receptors. Drugs of the Future 2001; 26: 1191-1197.[22] Frølund B, Ebert B, Kristiansen U, Liljefors T, Krogsgaard-Larsen

P. GABAA receptor ligands and their therapeutic potentials. CurrTop Med Chem 2002; 2: 817-832.

[23] Rudolph U, Möhler H. Analysis of GABAA receptor function anddissection of the pharmacology of benzodiazepines and generalanesthetics through mouse genetics. Ann Rev Pharmacol Toxicol2004; 44: 475-498.

[24] Johnston GAR, Chebib M, Hanrahan JR, Mewett KN. GABAC

receptors as drug targets. Curr Drug Targ CNS Neurol Disord2003; 2: 260-268.

[25] Christopoulos A. Allosteric binding sites on cell-surface receptors:Novel targets for drug discovery. Nature Rev Drug Disc 2002; 1:198-210.

[26] Vafa B, Schofield PR. Heritable mutations in the glycine, GABAA,and nicotinic acetylcholine receptors provide new insights into theligand-gated ion channel superfamily. Int Rev Neurobiol 1998; 42:285-332.

[27] Holopainen IE, Metsahonkala EL, Kokkonen H, Parkkola RK,Manner TE, Nagren K, et al. Decreased binding of [11C]flumazenilin Angelman syndrome patients with GABAA receptor β3 subunitdeletions. Ann Neurol 2001; 49: 110-113.

[28] Homanics GE, Delorey TM, Firestone LL, Quinlan JJ, HandforthA, Harrison NL, et al. Mice devoid of γ-aminobutyrate type Areceptor β3 subunit have epilepsy, cleft palate, and hypersensitivebehavior. Proc Natl Acad Sci USA 1997; 94: 4143-4148.

[29] Sinkkonen ST, Homanics GE, Korpi ER. Mouse models ofAngelman syndrome, a neurodevelopmental disorder, displaydifferent brain regional GABAA receptor alterations. Neurosci Lett2003; 340: 205-208.

[30] Feusner J, Ritchie T, Lawford B, Young RM, Kann B, Noble EP.GABAA receptor beta 3 subunit gene and psychiatric morbidity in apost-traumatic stress disorder population. Psych Res 2001; 104:109-117.

[31] Taylor M, Bhagwagar Z, Cowen PJ, Sharp T. GABA and mooddisorders. Psych Med 2003; 33: 3873-93.

[32] Loh EW, Ball D. Role of the GABA-Abeta 2, GABA-Aalpha 6,GABA-Aalpha 1 and GABA-Agamma 2 receptor subunit genescluster in drug responses and the development of alcoholdependence. Neurochem Int 2000; 37: 413-423.

[33] Parsian A, Zhang ZH. Human chromosomes 11p15 and 4p12 andalcohol dependence: Possible association with the GABRB1 gene.Am J Med Genet 1999; 88: 533-538.

[34] Neu A, Neuhoff H, Trube G, Fehr S, Ullrich K, Roeper J, et al.Activation of GABAA receptors by guanidinoacetate: A novelpathophysiological mechanism. Neurobiol Dis 2002; 11: 298-307.

[35] Leung JW, Xue H. GABAergic functions and depression: fromclassical therapies to herbal medicine. Curr Drug Targ CNS NeurolDisord 2003; 2: 363-374.

[36] Wassef AA, Dott SG, Harris A, Brown A, O'Boyle M, Meyer WJ,et al. Critical review of GABA-ergic drugs in the treatment ofschizophrenia. J Clin Psychopharmacol 1999; 19: 222-232.

[37] Wassef A, Baker J, Kochan LD. GABA and schizophrenia: areview of basic science and clinical studies. J Clin Psycho-pharmacol 2003; 23: 601-640.

[38] Green AR, Hainsworth AH, Jackson DM. GABA potentiation: alogical pharmacological approach for the treatment of acuteischaemic stroke. Neuropharmacology 2000; 39: 1483-1494.

[39] Shulman A, Strashun AM, Goldstein BA. GABAA-benzodiazepine-chloride receptor-targeted therapy for tinnitus control: preliminaryreport. Int Tinnitus J 2002; 8: 30-6.

[40] Maubach K. GABAA receptor subtype selective cognitionenhancers. Curr Drug Targ CNS Neurol Disord 2003; 2: 233-239.

[41] Scheffer IE, Berkovic SF. The genetics of human epilepsy. TrendsPharmacol Sci 2003; 24: 428-433.


[42] Jones-Davis DM, Macdonald RL. GABAA receptor function andpharmacology in epilepsy and status epilepticus. Curr OpinPharmacol 2003; 3: 12-18.

[43] Baulac S, Huberfeld G, Gourfinkel-An I, Mitropoulou G, BerangerA, Prud'homme JF, et al. First genetic evidence of GABAA receptordysfunction in epilepsy: a mutation in the gamma 2-subunit gene.Nature Genetics 2001; 28: 46-48.

[44] Wallace RH, Marini C, Petrou S, Harkin LA, Bowser DN, PanchalRG, et al . Mutant GABAA receptor gamma 2-subunit in childhoodabsence epilepsy and febrile seizures. Nature Genetics 2001; 28:49-52.

[45] Bianchi MT, Song LY, Zhang H, Macdonald RL. Two differentmechanisms of disinhibition produced by GABAA receptormutations linked to epilepsy in humans. J Neurosci 2002; 22: 5321-5327.

[46] Macdonald RL, Bianch MT, Feng H. Mutations linked togeneralized epilepsy in humans reduce GABAA receptor current.Exp Neurol 2003; 184: S58-67.

[47] Kananura C, Haug K, Sander T, Runge U, Gu WL, Hallmann K, etal. A splice-site mutation in GABRG2 associated with childhoodabsence epilepsy and febrile convulsions. Arch Neurol 2002; 59:1137-1141.

[48] Chou IC, Peng CT, Huang CC, Tsai JJ, Tsai FJ, Tsai CH.Association analysis of gamma 2 subunit of gamma-aminobutyricacid type A receptor polymorphisms with febrile seizures. PediatRes 2003; 54: 26-29.

[49] Cossette P, Liu LD, Brisebois K, Dong HH, Lortie A, Vanasse M,et al. Mutation of GABRA1 in an autosomal dominant form ofjuvenile myoclonic epilepsy. Nature Genetics 2002; 31: 184-189.

[50] Fisher JL. A mutation in the GABAA receptor α1 subunit linked tohuman epilepsy affects channel gating properties. Neuropharma-cology 2004; 46: 629-637.

[51] Gottesmann C. GABA mechanisms and sleep. Neuroscience 2002;111: 231-239.

[52] Laposky AD, Homanics GE, Basile A, Mendelson WB. Deletion ofthe GABAA receptor beta 3 subunit eliminates the hypnotic actionsof oleamide in mice. NeuroReport 2001; 12: 4143-4147.

[53] Buhr A, Bianchi MT, Baur R, Courtet P, Pignay V, Boulenger JP,et al. Functional characterization of the new human GABAA

receptor mutation beta 3(R192H). Hum Genet 2002; 111: 154-160.[54] Smith AJ, Alder L, Silk J, Adkins C, Fletcher AE, Scales T, et al.

Effect of alpha subunit on allosteric modulation of ion channelfunction in stably expressed human recombinant gamma-aminobutyric acid(A) receptors determined using Cl-36 ion flux.Mol Pharmacol 2001; 59: 1108-1118.

[55] Sanna E, Busonero F, Talani G, Carta M, Massa F, Peis M, et al.Comparison of the effects of zaleplon, zolpidem, and triazolam atvarious GABAA receptor subtypes. Eur J Pharmacol 2002; 451:103-110.

[56] Storustovu S, Ebert B. Gaboxadol: in vitro interaction studies withbenzodiazepines and ethanol suggest functional selectivity. Eur JPharmacol 2003; 467: 49-56.

[57] Johnston GAR. Dietary chemicals and brain function. J Proc RoySoc NSW 2003; 135: 57-71.

[58] Cherubini E, Conti F. Generating diversity at GABAergicsynapses. Trends Neurosci 2001; 24: 155-162.

[59] Smart TG. Regulation of excitatory and inhibitoryneurotransmitter-gated ion channels by protein phosphorylation.Curr Opin Neurobiol 1997; 7: 358-367.

[60] Hosie AM, Dunne EL, Harvey RJ, Smart TG. Zinc-mediatedinhibition of GABAA receptors: discrete binding sites underliesubtype specificity. Nature Neurosci 2003; 6: 362-369.

[61] Benson JA, Low K, Keist R, Möhler H, Rudolph U. Pharmacologyof recombinant γ-aminobutyric acidA receptors rendered diazepam-insensitive by point-mutated α-subunits. FEBS Letters 1998; 431:400-404.

[62] Rudolph U, Crestani F, Benke D, Brunig I, Benson JA, FritschyJM, et al. Benzodiazepine actions mediated by specific gamma-aminobutyric acid(A) receptor subtypes. Nature 1999; 401: 796-800.

[63] McKernan RM, Rosahl TW, Reynolds DS, Sur C, Wafford KA,Atack JR, et al. Sedative but not anxiolytic properties ofbenzodiazepines are mediated by the GABAA receptor alpha(1)subtype. Nature Neurosci 2000; 3: 587-592.

[64] Low K, Crestani F, Keist R, Benke D, Brunig I, Benson JA, et al.Molecular and neuronal substrate for the selective attenuation ofanxiety. Science 2000; 290: 131-134.

[65] Reynolds DS, McKernan RM, Dawson GR. Anxiolytic-like actionof diazepam: which GABAA receptor subtype is involved? TrendsPharmacol Sci 2001; 22: 402-403.

[66] Atack JR. Anxioselective compounds acting at the GABAA

receptor benzodiazepine binding site. Curr Drug Targ CNS NeurolDisord 2003; 2: 213-232.

[67] Collins I, Moyes C, Davey WB, Rowley M, Bromidge FA, QuirkK, et al. 3-Heteroaryl-2-pyridones: Benzodiazepine site ligandswith functional selectivity for α2/α3-subtypes of human GABAA

receptor-ion channels. J Med Chem 2002; 45: 1887-1900.[68] Carling RW, Moore KW, Street LJ, Wild D, Isted C, Leeson PD, et

al. 3-Phenyl-6-(2-pyridyl)methoxy-1, 2, 4-triazolo[3, 4-a]phthalazines and analogues: High-affinity γ-aminobutyric acid-Abenzodiazepine receptor ligands with α2, α3 and α5-subtypebinding selectivity over α1. J Med Chem 2004; 47: 1807-1822.

[69] Crestani F, Martin JR, Möhler H, Rudolph U. Mechanism of actionof the hypnotic zolpidem in vivo. Br J Pharmacol 2000; 131: 1251-1254.

[70] Dodd PR, Davies LP, Watson WE, Nielsen B, Dyer JA, Wong LS,et al. Neurochemical studies on quinolone antibiotics: effects onglutamate, GABA and adenosine systems in mammalian CNS.Pharmacol Toxicol 1989; 64: 404-411.

[71] Johnstone TB, Hogenkamp DJ, Coyne L, Su J, Halliwell RF, TranMB, et al. Modifying quinolone antibiotics yields new anxiolytics.Nat Med 2004; 10: 31-32.

[72] Caraiscos VB, Elliott EM, You-Ten KE, Cheng VY, Belelli D,Newell JG, et al. Tonic inhibition in mouse hippocampal CA1pyramidal neurons is mediated by α5 subunit-containing gamma-aminobutyric acid type A receptors. Proc Natl Acad Sci USA 2004;101: 3662-3667.

[73] Collinson N, Kuenzi FM, Jarolimek W, Maubach KA, Cothliff R,Sur C, et al. Enhanced learning and memory and alteredGABAergic synaptic transmission in mice lacking the alpha 5subunit of the GABAA receptor. J Neurosci 2002; 22: 5572-5580.

[74] Chambers MS, Atack JR, Broughton HB, Collinson N, Cook S,Dawson GR, et al. Identification of a novel, selective GABAA α5receptor inverse agonist which enhances cognition. J Med Chem2003; 46: 2227-2240.

[75] Jackel C, Kleinz R, Makela R, Hevers W, Jezequel S, Korpi ER, etal. The main determinant of furosemide inhibition on GABA(a)receptors is located close to the first transmembrane domain. Eur JPharmacol 1998; 357: 251-256.

[76] Knoflach F, Benke D, Wang Y, Scheurer L, Luddens H, HamiltonBJ, et al. Pharmacological modulation of the diazepam-insensitiverecombinant gamma-aminobutyric acid(a) receptors alpha-4-alpha-2-gamma-2 and alpha-6-beta-2-gamma-2. Mol Pharmacol 1996;50: 1253-1261.

[77] Thompson SA, Whiting PJ, Wafford KA. Barbiturate inrteractionsat the human GABAA receptor: dependnece on receptor subunitcombination. Br J Pharmacol 1996; 117: 521-527.

[78] Thompson SA, Arden SA, Marshall G, Wingrove PB, Whiting PJ,Wafford KA. Residues in transmembrane domains I and IIdetermine gamma-aminobutyric acid type A receptor subtype-selective antagonism by furosemide. Mol Pharmacol 1999; 55:993-999.

[79] Luddens H, Lang HJ, Korpi ER. Structure-activity relationship offurosemide-derived compounds as antagonists of cerebellum-specific GABAA receptors. Eur J Pharmacol 1998; 344: 269-277.

[80] Thomet U, Baur R, Razet R, Dodd RH, Furtmuller R, Sieghart W,et al. A novel positive allosteric modulator of the GABAA receptor:the action of (+)-ROD188. Br J Pharmacol 2000; 131: 843-850.

[81] Wall MJ. Cis-4-amino-crotonic acid activates alpha 6 subunit-containing GABAA but not GABAC receptors in granule cells ofadult rat cerebellar slices. Neurosci Lett 2001; 316: 37-40.

[82] Sinkkonen ST, Mansikkamaki S, Moykkynen T, Luddens H, Uusi-Oukari M, Korpi ER. Receptor subtype-dependent positive andnegative modulation of GABAA receptor function by niflumic acid,a nonsteroidal anti-inflammatory drug. Mol Pharmacol 2003; 64:753-63.

[83] Jones A, Korpi ER, Mckernan RM, Pelz R, Nusser Z, Makela R, etal. Ligand-gated ion channel subunit partnerships - GABA(a)receptor alpha(6) subunit gene inactivation inhibits delta subunitexpression. J Neurosci 1997; 17: 1350-1362.


[84] Homanics GE, Ferguson C, Quinlan JJ, Daggett J, Snyder K,Lagenaur C, et al. Gene knockout of the α6 subunit of the γ-aminobutyric acid type A receptor - lack of effect on responses toethanol, pentobarbital, and general anesthetics. Mol Pharmacol1997; 51: 588-596.

[85] Korpi ER, Koikkalainen P, Vekovischeva OY, Makela R, Kleinz R,Uusi-Oukari M, et al. Cerebellar granule-cell-specific GABAA

receptors attenuate benzodiazepine-induced ataxia: evidence fromα6-deficient mice. Eur J Neurosci 1999; 11: 233-240.

[86] Sigel E, Baur R. Electrophysiological evidence for the coexistenceof alpha 1 and alpha 6 subunits in a single functional GABAA

receptor. J Neurochem 2000; 74: 2590-2596.[87] Wall MJ. Furosemide reveals heterogeneous GABAA receptor

expression at adult rat Golgi cell to granule cell synapses,Neuropharmacology 2002; 43: 737-749.

[88] Leao RM, Mellor JR, Randall AD. Tonic benzodiazepine-sensitiveGABAergic inhibition in cultured rodent cerebellar granule cells.Neuropharmacology 2000; 39: 990-1003.

[89] Rosmond R, Bouchard C, Bjorntorp P. Allelic variants in theGABAA alpha 6 receptor subunit gene (GABRA6) is associatedwith abdominal obesity and cortisol secretion. Int J Obes RelMetabol Disord 2002; 26: 938-941.

[90] Hoffman WE, Balyasnikova IV, Mahay H, Danilov SM, BaughmanVL. GABA alpha 6 receptors mediate midazolam-inducedanxiolysis. J Clin Anes 2002; 14: 206-209.

[91] Reynolds DS, Rosahl TW, Cirone J, O'Meara GF, HaythornthwaiteA, Newman RJ, et al. Sedation and anesthesia mediated by distinctGABAA receptor isoforms. J Neurosci 2003; 23: 8608-8617.

[92] Cirone J, Rosahl TW, Reynolds DS, Newman RJ, O'Meara GF,Hutson PH, et al. Gamma-aminobutyric acid type A receptor beta 2subunit mediates the hypothermic effect of etomidate in mice.Anesthesiology 2004; 100: 1438-1445.

[93] Halliwell RF, Thomas P, Patten D, James CH, Martinez TA, MilediR, et al. Subunit-selective modulation of GABAA receptors by thenon-steroidal anti-inflammatory agent, mefenamic acid. Eur JNeurosci 1999; 11: 2897-2905.

[94] Thompson SA, Wheat L, Brown NA, Wingrove PB, Pillai GV, etal. Salicylidene salicylhydrazide, a selective inhibitor of β1-containing GABAA receptors. Br J Pharmacol 2004; 142: 97-106.

[95] Quinlan JJ, Firestone LL, Homanics GE. Mice lacking the longsplice variant of the gamma 2 subunit of the GABAA receptor aremore sensitive to benzodiazepines. Pharmacol Biochem Behav2000; 66: 371-374.

[96] Wallner M, Hanchar HJ, Olsen RW. Ethanol enhances alpha 4 beta3 delta and alpha 6 beta 3 delta gamma-aminobutyric acid type Areceptors at low concentrations known to affect humans. Proc NatlAcad Sci USA 2003; 100: 15218-15223.

[97] Harris RA. Transfected cells for study of alcohol actions onGABA(a) receptors. Addict Biol 1996; 1: 157-163.

[98] Sur C, Farrar SJ, Kerby J, Whiting PJ, Atack JR, McKernan RM.Preferential coassembly of alpha 4 and delta subunits of thegamma-aminobutyric acid(A) receptor in rat thalamus. MolPharmacol 1999; 56: 110-115.

[99] Peng Z, Hauer B, Mihalek RM, Homanics GE, Sieghart W, OlsenRW, et al. GABAA receptor changes in delta subunit-deficientmice: Altered expression of alpha 4 and gamma 2 subunits in theforebrain. J Comp Neurol 2002; 446: 179-197.

[100] Mihalek RM, Banerjee PK, Korpi ER, Quinlan JJ, Firestone LL,Mi ZP, et al. Attenuated sensitivity to neuroactive steroids ingamma-aminobutyrate type A receptor delta subunit knockoutmice. Proc Natl Acad Sci USA 1999; 96: 12905-12910.

[101] Spigelman I, Li Z, Liang J, Cagetti E, Samzadeh S, Mihalek RM, etal. Reduced inhibition and sensitivity to neurosteroids inhippocampus of mice lacking the GABAA receptor delta subunit. JNeurophysiol 2003; 90: 903-910.

[102] Johnston GAR. GABA chemistry: analogs of GABA as therapeuticand investigational agents. In GABA In The Nervous System: TheView At Fifty Years, DL. Martin R. Olsen, eds.; LippincottWilliams and Wilkins: Philadelphia 2000; pp. 65-80.

[103] Woodward RM, Polenzani L, Miledi R. Characterization ofbicuculline/baclofen-insensitive (rho-like) gamma-aminobutyricacid receptors expressed in Xenopus oocytes. II. Pharmacology ofgamma-aminobutyric acidA and gamma-aminobutyric acidB

receptor agonists and antagonists. Mol Pharmacol 1993; 43: 609-625.

[104] Krogsgaard-Larsen P, Johnston GAR, Lodge D, Curtis DR. A newclass of GABA agonist. Nature 1977; 268: 53-55.

[105] Krogsgaard-Larsen P, Frølund B, Kristiansen U, Frydenvang K,Ebert B. GABAA and GABAB receptor agonists, partial agonists,antagonists and modulators - design and therapeutic prospects. EurJ Pharm Sci 1997; 5: 355-384.

[106] Kjaer M, Neilson H. The analgesic effect of the GABA agonistTHIP in patients with chronic pain of malignant origin. Br J ClinPharmacol 1983; 16: 477-485.

[107] Vaught JL, Pelley K, Costa LG, Setler P, Enna SJA. A comparisonof the antinociceptive responses to the GABA-receptor agonistsTHIP and baclofen. Neuropharmacology 1985; 24: 211-216.

[108] Faulhaber J, Steiger A, Lancel M. The GABAA agonist THIPproduces slow wave sleep and reduces spindling activity in NREMsleep in humans. Psychopharmacology 1997; 130: 285-291.

[109] Zorn SH, Enna SJ. The GABA agonist THIP attentuatesantinociception in the mouse by modifying central cholinergictransmission. Neuropharmacology 1987; 26: 433-437.

[110] Tatsuo M, Yokoro CM, Salgado JV, Pesquero S, Santana M,Francischi JN. Hyperalgesic effect induced by barbiturates,midazolam and ethanol - pharmacological evidence for GABA-areceptor involvement. Brazil J Med Biol Res 1997; 30: 251-256.

[111] Arnaud C, Gauthier P, Gottesmann C. Study of a GABAC receptorantagonist on sleep-waking behavior in rats. Psychopharmacology2001; 154: 415-419.

[112] Skerritt JH, Johnston GAR. Diazepam stimulates the binding ofGABA and muscimol but not THIP to rat brain membranes.Neurosci Lett 1983; 38: 315-320.

[113] Hansen SL, Sperling BB, Sanchez C. Anticonvulsant andantiepileptogenic effects of GABAA receptor ligands inpentylenetetrazole-kindled mice. Prog Neuropsychopharm BiolPsych 2004; 28: 105-13.

[114] Ebert B, Whiting PJ, Krogsgaard-Larsen P, Kemp JA. Molecularpharmacology of γ-aminobutyric acid type A receptor agonists andpartial agonists in oocytes injected with different α, β, and γreceptor subunit combinations. Mol Pharmacol 1994; 46: 957-963.

[115] Brown N, Kerby J, Bonnert TP, Whiting PJ, Wafford KA.Pharmacological characterization of a novel cell line expressinghuman alpha(4)beta(3)delta GABAA receptors. Br J Pharmacol2002; 136: 965-974.

[116] Vien J, Duke RK, Mewett KN, Johnston GAR, Shingai R, ChebibM. trans-4-Amino-2-methylbut-2-enoic acid (2-MeTACA) and (+/-)-trans-2-aminomethylcyclopropanecarboxyic acid (+/-)-TAMP)can differentiate rat rho 3 from human rho 1 and rho 2 recombinantGABAC receptors. Br J Pharmacol 2002; 135: 883-890.

[117] Voss J, Sanchez C, Michaelsen S, Ebert B. Rotarod studies in therat of the GABAA receptor agonist gaboxadol: lack of ethanolpotentiation and benzodiazepine cross-tolerance. Eur J Pharmacol2003; 482: 215-222.

[118] Ebert B, Storustovu SI, Mortensen M, Frølund B. Characterizationof GABAA receptor ligands in the rat cortical wedge preparation:evidence for action at extrasynaptic receptors?. Br J Pharmacol2002; 137: 1-8.

[119] Allan RD, Evans RH, Johnston GAR. γ-Aminobutyric acidagonists: an in vitro comparison between depression of spinalsynaptic activity and depolarization of spinal root fibres in the rat.Br J Pharmacol 1980; 70: 609-615.

[120] Lindquist CE, Ebert B, Birnir B. Extrasynaptic GABAA channelsactivated by THIP are modulated by diazepam in CA1 pyramidalneurons in the rat brain hippocampal slice. Mol Cell Neurosci2003: 24: 250-257.

[121] Kristiansen U, Lambert J. Benzodiazepine and barbiturate ligandsmodulate responses of cultured hippocampal neurones to theGABAA receptor partial agonist, 4-piol. Neuropharmacology 1996;35: 1181-1191.

[122] Mortensen M, Frølund B, Jorgensen AT, Liljefors T, Krogsgaard-Larsen P, Ebert B. Activity of novel 4-PIOL analogues at humanalpha(1)beta(2)gamma(2S) GABAA receptors - correlation withhydrophobicity. Eur J Pharmacol 2002; 451: 125-132.

[123] Carlier PR, Chow ESH, Barlow RL, Bloomquist JR. Discovery ofnon-zwitterionic GABAA receptor full agonists and a superagonist.Bioorg Med Chem Lett 2002; 12: 1985-1988.

[124] Adkins CE, Pillai GV, Kerby J, Bonnert TP, Haldon C, McKernanRM, et al. α4β3δ GABA A receptors characterized by fluorescenceresonance energy transfer-derived measurements of membranepotential. J Biol Chem 2001; 276: 38934-38939.


[125] Cashin MF, Moravek V. The physiological action of cholesterol.Am J Physiol 1927; 82: 294-298.

[126] Starkenstein E, Weden H. The physiology and pharmacology ofsterols. IV. Influence of cholesterol in the action of hypnotics andnarcotics. Arch exp Pathol Pharmak 1936; 182: 700-714.

[127] Seyle H. Correlations between the chemical structure and thepharmacological actions of the steroids. Endocrinology 1942; 30:437-453.

[128] Harrison NL, Simmonds MA. Modulation of the GABA receptorcomplex by a steroid anaesthetic. Brain Res 1984; 323: 287-292.

[129] Majewska MD, Harrison NL, Schwartz RD, Barker JL, Paul SM.Steroid hormone metabolites are barbiturate-like modulators of theGABA receptor. Science 1986; 232: 1004-1007.

[130] Purdy RH, Morrow AL, Moore PH, Paul SM. Stress-inducedelevations of γ-aminobutyric acid type A receptor-active steroids inthe rat brain. Proc Natl Acad Sci USA 1991; 88: 4553-4557.

[131] Belelli D, Casula A, Ling A, Lambert JJ. The influence of subunitcomposition on the interaction of neurosteroids with GABAA

receptors, Neuropharmacology 2002; 43: 651-661.[132] Wohlfarth KM, Bianchi MT, Macdonald RL. Enhanced

neurosteroid potentiation of ternary GABAA receptors containingthe delta subunit. J Neurosci 2002; 22: 1541-1549.

[133] Rick CE, Ye Q, Finn SE, Harrison NL. Neurosteroids act on theGABAA receptor at sites on the N-terminal side of the middle ofTM2. NeuroReport 1998; 9: 379-383.

[134] Pistis M, Belelli D, McGurk K, Peters JA, Lambert JJ.Complementary regulation of anaesthetic activation of human(alpha(6)beta(3)gamma(2L)) and Drosophila (RDL) GABAreceptors by a single amino acid residue. J Physiol 1999; 515: 3-18.

[135] Darbandi-Tonkabon R, Hastings WR, Zeng CM, Akk G, ManionBD, Bracamontes JR, et al. Photoaffinity labeling with aneuroactive steroid analogue. 6-azi-pregnanolone labels voltage-dependent anion channel-1 in rat brain. J Biol Chem 2003; 278:13196-13206.

[136] Darbandi-Tonkabon R, Manion BD, Hastings WR, Craigen WJ,Akk G, Bracamontes JR, et al. Neuroactive steroid interactionswith voltage-dependent anion channels: lack of relationship toGABAA receptor modulation and anesthesia. J Pharmacol ExpTherapeut 2004; 308: 502-511.

[137] Jorge-Rivera JC, McIntyre KL, Henderson LP. Anabolic steroidsinduce region- and subunit-specific rapid modulation of GABAA

receptor-mediated currents in the rat forebrain. J Neurophysiol2000; 83: 3299-3309.

[138] Monaghan EP, Navalta LA, Shum L, Ashbrook DW, Lee DA.Initial human experience with ganaxolone, a neuroactive steroidwith antiepileptic activity. Epilepsia 1997; 38: 1026-1031.

[139] Reddy DS, Newer gabaergic agents for pharmacotherapy ofinfantile spasms. Drugs Today 2002; 38: 657-675.

[140] Snead OC, Ganaxolone. a selective, high-affinity steroid modulatorof the γ-aminobutyric acid-A receptor, exacerbates seizures inanimal models of absence. Ann Neurol 1998; 44: 688-691.

[141] Ong J, Kerr DI, Johnston GAR. Cortisol: a potent biphasicmodulator at GABAA-receptor complexes in the guinea pig isolatedileum. Neurosci Lett 1987; 82: 101-106.

[142] Ong J, Kerr DI, Capper HR, Johnston GAR. Cortisone: a potentGABAA antagonist in the guinea-pig isolated ileum. J PharmPharmacol 1990; 42: 662-664.

[143] Andres-Trelles F, Bibby V, Lustman S, Simmonds MA. Effects ofcortisol on GABAA receptor-mediated responses compared inguinea-pig ileum and rat cuneate nucleus. Neuropharmacology1989; 28: 705-708.

[144] Majewska MD. Antagonist-type interaction of glucocorticoids withthe GABA receptor-coupled chloride channel. Brain Res 1987;418: 377-382.

[145] Zaki A, Barrett-Jolley R. Rapid neuromodulation by cortisol in therat paraventricular nucleus: an in vitro study. Br J Pharmacol 2002;137: 87-97.

[146] Orchinik M, Hastings N, Witt D, McEwen BS. High affinitybinding of corticosterone to mammalian neuronal membranes -possible role of corticosteroid binding globulin. J Steroid BiochemMol Biol 1997; 60: 229-236.

[147] Orchinik M, Carroll SS, Li YH, McEwen BS, Weiland NG.Heterogeneity of hippocampal GABAA receptors: Regulation bycorticosterone. J Neurosci 2001; 21: 330-339.

[148] Keenan PA, Kuhn TW. Do glucocorticoids have adverse effects onbrain function? CNS Drugs 1999; 11: 245-251.

[149] Topliss JG, Clark AM, Ernst E, Hufford CD, Johnston GAR,Rimoldi JM, et al. Natural and synthetic substances related tohuman health - (IUPAC Technical Report). Pure Appl Chem 2002;74: 1957-1985.

[150] Aherne SA, O'Brien NM. Dietary flavonols: Chemistry, foodcontent, and metabolism. Nutrition 2002; 18: 75-81.

[151] Havsteen BH. The biochemistry and medical significance of theflavonoids. Pharmacol Therap 2002; 96: 67-202.

[152] Heim KE, Tagliaferro AR, Bobilya DJ. Flavonoid antioxidants:chemistry, metabolism and structure-activity relationships. J NutritBiochem 2002; 13: 572-584.

[153] Le Marchand L. Cancer preventive effects of flavonoids - a review.Biomed Pharmacotherap 2002; 56: 296-301.

[154] Marder M, Paladini AC. GABAA receptor ligands of flavonoidstructure. Curr Top Med Chem 2002; 2: 853-867.

[155] Luk KC, Stern L, Weigele M, O'Brien RA, Spirst N. Isolation andidentification of "diazepam-like" compounds in bovine brain. J NatProd 1983; 46: 852-861.

[156] Nielsen M, Frokjaer S, Braestrup C. High affinity of the naturally-occurring biflavonoid, amentoflavone, to brain benzodiazepinereceptors in vitro. Biochem Pharmacol 1988; 37: 3285-3287.

[157] Gutmann H, Bruggisser R, Schaffner W, Bogman K, Botomino A,Drewe J. Transport of amentoflavone across the blood-brain barrierin vitro. Plant Med 2002; 68: 804-807.

[158] Baureithel KH, Büter KB, Engesser A, Burkard W, Schaffer W.Inhibition of benzodiazepine binding in vitro by amentoflavone, aconstituent of various species of Hypericum. Pharm Acta Helv1997; 72: 153-157.

[159] Hanrahan JR, Chebib M, Davucheron NM, Hall BJ, Johnston GAR.Semisynthetic preparation of amentoflavone: a negative modulatorat GABAA receptors. Bioorg Med Chem Lett 2003; 13: 2281-2284.

[160] Butterweck V, Nahrstedt A, Evans J, Hufeisen S, Rauser L, SavageJ, Poedak B, et al. In vitro receptor screening of pure constitutentsof St. John's wort reveals novel interactions with a number ofGPCRs. Psychopharmacology 2002; 162: 193-202.

[161] Viola H, Wasowski C, Levi de Stein M, Wolfman C, Silvera R,Medina AE, et al. a component of Matricaria recutita flowers, is acentral benzodiazepine receptors-ligand with anxiolytic effects.Plant Med 1995; 61: 213-216.

[162] Avallone R, Zanoli P, Puia G, Kleinschnitz M, Schreier P, BaraldiM. Pharmacological profile of apigenin, a flavonoid isolated fromMatricaria chamomilla. Biochem Pharmacol 2000; 59: 1387-1394.

[163] Zanoli P, Avallone R, Baraldi M. Behavioural characterisation ofthe flavonoids apigenin and chrysin. Fitoterapia 2000; 71: S117-S123.

[164] Goutman JD, Waxemberg MD, Donate-Oliver F, Pomata PE,Calvo DJ. Flavonoid modulation of ionic currents mediated byGABAA and GABAC receptors. Eur J Pharmacol 2003; 461: 79-87.

[165] Campbell EL, Chebib M, Johnston GAR. The dietary flavonoidsapigenin and (-)-epigallocatechin gallate enhance the positivemodulation by diazepam of the activation by GABA ofrecombinant GABAA receptors. Biochem Pharmacol 2004; 68:1631-1638.

[166] Dixon RA, Ferreira D. Genistein Phytochemistry 2002; 60: 205-211.

[167] Dunne EL, Moss SJ, Smart TG. Inhibition of GABAA receptorfunction by tyrosine kinase inhibitors and their inactive analogues.Mol Cell Neurosci 1998; 12: 300-310.

[168] Huang RQ, Fang MJ, Dillon GH. The tyrosine kinase inhibitorgenistein directly inhibits GABAA receptors. Mol Brain Res 1999;67: 177-183.

[169] Kavvadias D, Monschein V, Sand P, Riederer P, Schreier P.Constituents of sage (Salvia officinalis L.) with in vitro affinity tohuman brain benzodiazepine receptor. Plant Med 2003; 69: 113-117.

[170] Perry EK, Pickering AT, Wang WW, Houghton PJ, Perry NS.Medicinal plants and Alzheimer's disease: from ethnobotany tophytotherapy. J Pharm Pharmacol 1999; 51: 527-534.

[171] Perry NS, Howes M-J, Houghton P, Perry E. Why sage may be awise memory remedy: Effects of Salvia on the nervous system.Med Arom Plants Indust Profiles 2000; 14: 207-223.

[172] Tildesley NT, Kennedy DO, Perry EK, Ballard CG, Savelev S,Wesnes KA, et al. Salvia lavandulaefolia (Spanish Sage) enhancesmemory in healthy young volunteers. Phamaco Biochem Behav2003; 75: 669-674.

1884 Current Pharmaceutical Design, 2005, Vol. 11, No.15 Graham A.R. Johnston

[173] Kavvadias D, Sand P, Youdim KA, Qaiser MZ, Rice-Evans C,Baur R, et al. The flavone hispidulin, a benzodiazepine receptorligand with positive allosteeric properties, traverses the blood-brainbarrier and exhibits anti-convulsive effects. Br J Pharmacol 2004,142: 811-820

[174] Walters RJ, Hadley SH, Morris KDW, Amin J. Benzodiazepinesact on GABAA receptors via two distinct and separablemechanisms. Nature Neurosci 2000; 3: 1274-1281.

[175] Hall BJ, Chebib M, Hanrahan JR, Johnston GAR. Flumazenil-independent positive modulation of γ-aminobutyric acid by 6-methylflavone at human recombinant α1β2γ2L and α1β2 GABA A

receptors. Eur J Pharmacol 2004; 491: 1-8.[176] Wang HY, Hui KM, Chen YJ, Xu SX, Wong JTF, Xue H.

Structure-activity relationships of flavonoids, isolated fromScutellaria baicalensis , binding to benzodiazepine site of GABAA

receptor complex. Plant Med 2002; 68: 1059-1062.[177] Huen MS, Leung JW, Ng W, Lui WS, Chan MN, Wong JT, et al.

5, 7-Dihydroxy-6-methoxyflavone, a benzodiazepine site ligandisolated from Scutellaria baicalensis Georgi, with selectiveantagonistic properties. Biochem Pharmacol 2003; 66: 125-32.

[178] Wasowski C, Marder M, Viola H, Medina JH, Paladini AC.Isolation and identification of 6-methylapigenin, a competitiveligand for the brain GABAA receptors, from Valeriana wallichii.Plant Med 2002; 68: 934-936.

[179] Huen MS, Hui KM, Leung JW, Sigel E, Baur R, Wong JT, et al.Naturally occurring 2'-hydroxyl-substituted flavonoids as high-affinity benzodiazepine site ligands. Biochem Pharmacol 2003; 66:2397-2407.

[180] Du XM, Sun NY, Takizawa N, Guo YT, Shoyama Y. Sedative andanticonvulsant activities of goodyerin, a flavonol glycoside fromGoodyera schlechtendaliana. Phytother Res 2002; 16: 261-263.

[181] Fernandez S, Wasowski C, Paladini AC, Marder M. Sedative andsleep-enhancing properties of linarin, a flavonoid-isolated fromValeriana officinalis. Pharmacol Biochem Behav 2004; 77: 399-404.

[182] Marder M, Viola H, Bacigaluppo JA, Colombo MI, Wasowski C,Wolfman C, et al. Detection of benzodiazepine receptor ligands insmall libraries of flavone derivatives synthesized by solution phasecombinatorial chemistry. Biochem Biophys Res Comm 1998; 249:481-485.

[183] Dekermendjian K, Kahnberg P, Witt MR, Sterner O, Nielsen M,Liljefors T. Structure-activity relationships and molecular modelinganalysis of flavonoids binding to the benzodiazepine site of the ratbrain GABAA receptor complex. J Med Chem 1999; 42: 4343-4350.

[184] Hong X, Hopfinger AJ. 3D-pharmacophores of flavonoid bindingat the benzodiazepine GABAA receptor site using 4D-QSARanalysis. J Chem Inf Comput Sci 2003; 43: 324-336.

[185] Huang XQ, Liu T, Gu JD, Luo XM, Ji RY, Cao Y, et al. 3D-QSARmodel of flavonoids binding at benzodiazepine site in GABAA

receptors. J Med Chem 2001; 44: 1883-1891.[186] Kahnberg P, Lager E, Rosenberg C, Schougaard J, Camet L,

Sterner O, et al. Refinement and evaluation of a pharmacophoremodel for flavone derivatives binding to the benzodiazepine site ofthe GABAA receptor. J Med Chem 2002; 45: 4188-4201.

[187] Marder M, Estiu G, Blanch LB, Viola H, Wasowski C, Medina JH,et al. Molecular modelling and QSAR analysis of the interaction offlavone derivatives with the benzodiazepine site of GABAA

receptor complex. Bioorg Med Chem 2001; 9: 323-335.[188] Huang SH, Duke RK, Chebib M, Sasaki K, Wada K, Johnston

GAR. Bilobalide, a sesquiterpene trilactone from Ginkgo biloba, isan antagonist at recombinant alpha(1) beta(2)gamma(2L) GABAA

receptors. Eur J Pharmacol 2003; 464: 1-8.[189] Sasaki K, Oota I, Wada K, Inomata K, Ohshika H, Haga M. Effects

of bilobalide, a sesquiterpene in Ginkgo biloba leaves, onpopulation spikes in rat hippocampal slices. Comp BiochemPhysiol C-Toxicol Pharmacol 1999; 124: 315-321.

[190] Sasaki K, Hatta S, Haga M, Ohshika H. Effects of bilobalide ongamma-aminobutyric acid levels and glutamic acid decarboxylasein mouse brain. Eur J Pharmacol 1999; 367: 165-173.

[191] Ivic L, Sands TT, Fishkin N, Nakanishi K, Kriegstein AR,Stromgaard K. Terpene trilactones from Ginkgo biloba areantagonists of cortical glycine and GABAA receptors. J Biol Chem2003; 278: 49279-85.

[192] Chatterjee SS, Kondratskaya EL, Krishtal OA. Structure-activitystudies with Ginkgo biloba extract constituents as receptor-gated

chloride channel blockers and modulators. Pharmacopsychiatry2003; 36: S68-77.

[193] Brochet D, Chermat R, DeFeudis FV, Drieu K. Effects of singleintraperitoneal injections of an extract of Ginkgo biloba (EGb 761)and its terpene trilactone constituents on barbital-induced narcosisin the mouse. Gen Pharmacol 1999; 33: 249-256.

[194] Deiml T, Haseneder R, Zieglgansberger W, Rammes G,Eisensamer B, Rupprecht R, et al. α-Thujone reduces 5-HT3receptor activity by an effect on the agonist-induceddesensitization. Neuropharmacology 2004; 46: 192-201.

[195] Hold KM, Sirisoma NS, Ikeda T, Narahashi T, Casida JE. a-thujone (the active component of absinthe): γ-aminobutyric acidtype A receptor modulation and metabolic detoxification. Proc NatlAcad Sci USA 2000; 97: 3826-3831.

[196] Priestley CM, Williamson EM, Wafford K, Sattelle DB. Thymol, aconstituent of thyme essential oil, is a positive allosteric modulatorof human GABAA receptors and a homo-oligomeric GABAreceptor from Drosophila melanogaster. Br J Pharmacol 2003;140: 1363-1372.

[197] Mohammadi B, Haeseler G, Leuwer M, Dengler R, Krampfl K,Bufler J. Structural requirements of phenol derivatives for directactivation of chloride currents via GABAA receptors. Eur JPharmacol 2001; 421: 85-91.

[198] Hosseinzadeh H, Parvardeh S. Anticonvulsant effects ofthymoquinone, the major constituent of Nigella sativa seeds, inmice. Phytomedicine 2004; 11: 56-64.

[199] Granger RE, Campbell EL, Johnston GAR. Valerian monoterpenesproduce unprecedented enhancememnt of GABA action at GABAA

receptors. J Neurochem 2004; 88: (Suppl. 1) 99.[200] Yuan CS, Mehendale S, Xiao Y, Aung HH, Xie JT, Ang-Lee MK.

The gamma-aminobutyric acidergic effects of valerian andvalerenic acid on rat brainstem neuronal activity. Anesth Analg2004; 98: 353-358.

[201] Ha JH, Lee KY, Choi HC, Cho J, Kang LS, Lim JC, et al.Modulation of radioligand binding to the GABAA-benzodiazepinereceptor complex by a new component from Cyperus rotundus.Biol Pharmaceut Bull 2002; 25: 128-130.

[202] Choi SE, Choi S, Lee JH, Whiting PJ, Lee SM, Nah SY. Effects ofginsenosides on GABAA receptor channels expressed in Xenopusoocytes. Arch Pharmac Res 2003; 26: 28-33.

[203] Lee CM, Wong HN, Chui KY, Choang TF, Hon PM, Chang HM.Miltirone, a central benzodiazepine receptor partial agonist from aChinese medicinal herb Salvia miltiorrhiza. Neurosci Lett 1991;127: 237-241.

[204] Chang HM, Chui KY, Tan FW, Zhong ZP, Lee CM, Sham HL, etal. Structure-activity relationship of miltirone, an active centralbenzodiazepine ligand isolated from Salvia miltiorrhiza Bunge(Danshen). J Med Chem 1991; 34: 1675-1692.

[205] Rutherford DM, Nielsen MPC, Hansen SK, Witt MR, BergendorffO, Sterner O. Isolation from salvia officinalis and indentification oftwo diterpenes which inhibit t-butylbicyclophosphoro[35S]thionatebinding to chloride channel of rat cerebrocortical membranes invitro. Neurosci Lett 1992; 135: 224-226.

[206] Perry NB, Anderson RE, Brennan NJ, Douglas MH, Heaney AJ,McGimpsey JA, et al. Essential oils from Dalmation sage (Salviaofficinalis l.): vaiations among individuals, plant parts, seasons, andsites. J Ag Food Chem 1999; 47: 2048-2054.

[207] Shelp BJ, Bown AW, McLean MD. Metabolism and functions of γ-aminobutyric acid. Trends Plant Sci 1999; 4: 446-452.

[208] Hossain SJ, Hamamoto K, Aoshima H, Hara Y. Effects of teacomponents on the response of GABAA receptors expressed inXenopus oocytes. J Ag Food Chem 2002; 50: 3954-3960.

[209] Hossain SJ, Aoshima H, Koda H, Kiso Y. Effects of coffeecomponents on the response of GABAA receptors expressed inXenopus oocytes. J Ag Food Chem 2003; 51: 7568-7575.

[210] Ferreira ACS, Barbe J-C, Bertrand A. 3-Hydroxy-4, 5-dimethyl-2(5H)-furanone: a key odorant of the typical aroma of oxidativeaged port wine. J Ag Food Chem 2003; 51: 4356-4363.

[211] Hossain SJ, Aoshima H, Koda H, Kiso Y. Potentiation of theionotropic GABA receptor response by whiskey fragrance. J AgFood Chem 2002; 50: 6828-6834.

[212] Koda H, Hossain SJ, Kiso Y, Aoshima H. Aging of whiskeyincreases the potentiation of GABAA receptor response. J Ag FoodChem 2003; 51: 5238-44.


[213] Aoshima H, Hamamoto K. Potentiation of GABAA receptorsexpressed in Xenopus oocytes by perfume and phytoncid. BiosciBiotech Biochem 1999; 63: 743-748.

[214] Rezai N, Duggan C, Cairns D, Lees G, Chazot PL. Modulation of[3H] TBOB binding to the rodent GABAA receptor by simpledisaccharides. Biochem Pharmacol 2003; 65: 619-23.

[215] Sooksawate T, Simmonds MA. Influence of membrane cholesterolon modulation of the GABAA receptor by neuroactive steroids andother potentiators. Br J Pharmacol 2001; 134: 1303-1311.

[216] Agar A, Yargicoglu P, Senturk KU, Oner G. The role of dietcholesterol changes on EEG. Int J Neurosci 1994; 75: 103-109.

[217] Simons M, Keller P, Dichgans J, Schulz JB. Cholesterol andAlzheimer's disease: is there a link?. Neurology 2002; 57: 1089-1093.

[218] Locatelli S, Lutjohann D, Schmidt HH, Otto C, Beisiegel U, vonBergmann K. Reduction of plasma 24S-hydroxycholesterol

(cerebrosterol) levels using high-dosage simvastatin in patientswith hypercholesterolemia: evidence that simvastatin affectscholesterol metabolism in the human brain. Arch Neurol 2002; 59:213-216.

[219] Beckstead MJ, Weiner JL, Eger EI, Gong DH, Mihic SJ. Glycineand gamma-aminobutyric acid(A) receptor function is enhanced byinhaled drugs of abuse. Mol Pharmacol 2000; 57: 1199-1205.

[220] Meulenberg CJ, Vijverberg HP. Selective inhibition of γ-aminobutyric acid type A receptors in human IMR-32 cells by lowconcentrations of toluene. Toxicology 2003; 190: 243-248.

[221] Miyazawa A, Fujiyoshi Y, Unwin N. Structure and gatingmechanism of the acetylcholine receptor pore. Nature 2003; 423:949-955.

[222] Bali M, Akabas MH. Defining the propofol binding site location onthe GABAA receptor. Mol Pharmacol 2004; 65: 68-76.

370

Documents