modulation of gaba a receptor function by neuroactive steroids: evidence for heterogeneity of...

Modulation of GABAA receptor function byneuroactive steroids: evidence for heterogeneityof steroid sensitivity of recombinant GABAA

receptor isoforms

Rajatavo Maitra and James N. Reynolds

Abstract: Neuroactive steroids are potent, selective allosteric modulators ofγ-aminobutyric acid type A (GABAA)receptor function in the central nervous system, and may serve as endogenous anxiolytic and analgesic agents. In orderto study the influence of subunit subtypes of the GABAA receptor on modulation of receptor function by neuroactivesteroids, we expressed human recombinant GABAA receptors inXenopusoocytes. GABA-activated membrane current,and the modulatory effects of the endogenous neurosteroid 5α-pregnan-3α-ol-20-one (allopregnanolone) and thesynthetic steroid anesthetic 5α-pregnan-3α-ol-11,20-dione (alphaxalone) were measured using two-electrodevoltage-clamp recording techniques. Allopregnanolone had similar effects to potentiate GABA-activated membranecurrent in theα1β1γ2L and α1β2γ2L receptor isoforms. In contrast, alphaxalone was much more effective as a positiveallosteric modulator on theα1β1γ2L receptor isoform. In the absence of theγ2L subunit subtype, allopregnanolone hadmuch greater efficacy, but its potency was decreased. Allopregnanolone was much more effective on theα1β1 receptorisoform compared with theα1β2 receptor isoform. The potency for alphaxalone to potentiate the GABA response wasnot altered in the absence of theγ2L subunit subtype, although its efficacy was greatly enhanced. Bothallopregnanolone and alphaxalone produced nonparallel leftward shifts in the GABA concentration–responserelationship in the absence of theγ2L subunit, decreasing the EC50 concentration of GABA and increasing the maximalresponse. Only alphaxalone increased the maximal GABA response when theγ2L subunit subtype was present. The3β-pregnane isomers epipregnanolone and isopregnanolone both inhibited the ability of allopregnanolone andalphaxalone to potentiate GABAA receptor function. However, the degree of block produced by the 3β-pregnane steroidisomers was dependent on the type of receptor isoform studied and the neuroactive steroid tested. Isopregnanolone, the3β-isomer of allopregnanolone, was significantly more effective as a blocker of potentiation caused by allopregnanolonecompared with alphaxalone in all receptor isoforms tested. Epipregnanolone had a greater efficacy as a blocker at theα1β2γ2L receptor isoform compared with theα1β1γ2L receptor isoform, and also produced a greater degree of blockof potentiation caused by allopregnanolone compared with alphaxalone. Our results support the hypothesis that theheteromeric assembly of different GABAA receptor isoforms containing different subunit subtypes results in multiplesteroid recognition sites on GABAA receptors, which in turn produces distinctly different modulatory interactionsbetween neuroactive steroids acting at the GABAA receptor. Theα and γ subunit subtypes may have the greatestinfluence on allopregnanolone modulation of GABAA receptor function, whereas theβ and γ subunit subtypes appear tobe most important for the modulatory effects of alphaxalone.

Key words: GABAA receptor, neurosteroid, allopregnanolone, alphaxalone,Xenopusoocyte.

920Résumé: Les stéroïdes neuroactifs sont des modulateurs allostériques sélectifs puissants de la fonction des récepteursde l’acideγ-amiobutyrique de type A (GABAA) dans le système nerveux central et pourraient servir d’analgésiques etd’anxiolytiques endogènes. Dans le but d’examiner l’influence de sous-types de sous-unités du récepteur GABAA sur lamodulation de la fonction des récepteurs par les stéroïdes neuroactifs, nous avons exprimé des récepteurs GABAA

recombinants humains dans des ovocytes deXenopus. Le courant membranaire activé par le GABA et les effetsmodulateurs du neurostéroïde endogène 5α-prégnane-3α-ol-20-one (alloprégnanolone) et du stéroïde synthétiqueanesthésique 5α-prégnane-3α-ol-11,20-dione (alphaxalone) ont été déterminés en utilisant des techniquesd’enregistrement de potentiel imposé à deux électrodes. L’alloprégnanolone a potentialisé de façon similaire le courant

Can. J. Physiol. Pharmacol.76: 909–920 (1998) © 1998 NRC Canada

909

Received February 17, 1998.

R. Maitra. Department of Pharmacology and Toxicology, Queen’s University, Kingston, ON K7L 3N6, Canada.J.N. Reynolds.1 Department of Pharmacology and Toxicology and Department of Anaesthesia and Critical Care, Queen’sUniversity, Kingston, ON K7L 3N6, Canada.

1Author to whom all correspondence should be addressed (e-mail: [email protected]).

membranaire activé par GABA dans les isoformes des récepteursα1β1γ2L et α1β2γ2L. À l’opposé, l’alphaxalone a étébeaucoup plus efficace comme modulateur allostérique positif sur l’isoforme du récepteurα1β1γ2L. En l’absence dusous-type de la sous-unitéγ2L, l’alloprégnanolone a été beaucoup plus efficace, mais moins puissant. Il a été beaucoupplus efficace envers l’isoforme du récepteurα1β1 qu’envers celle du récepteurα1β2. La puissance de l’alphaxalonepour ce qui est de potentialiser la réponse au GABA n’a pas été modifiée en l’absence du sous-type de la sous-unitéγ2L, bien que son efficacité ait été accrue de façon considérable. Tant l’alloprégnanolone que l’alphaxalone ontprovoqué des déplacements vers la gauche non parallèles de la relation concentration–réponse du GABA en l’absencede la sous-unitéγ2L, diminuant la concentration EC50 du GABA et augmentant la réponse maximale. Seulel’alphaxalone a augmenté la réponse maximale du GABA en présence du sous-type de la sous-unitéγ2L. Les isomères3β-prégnane, épiprégnanolone et isoprégnanolone, ont inhibé la capacité de l’alloprégnanolone et de l’alphaxalone àpotentialiser la fonction du récepteur GABAA. Toutefois, le niveau de blocage produit par les isomères stéréoïdiens3β-prégnane a été dépendant du type d’isoforme des récepteurs examinés ainsi que du stéroïde neuroactif testé.L’isoprégnanolone, le 3β-isomère de l’alloprégnanolone, a été significativement plus efficace comme bloqueur de lapotentialisation induite par l’alloprégnanolone que par l’alphaxalone dans toutes les isoformes des récepteurs examinés.L’épiprégnanolone a été beaucoup plus efficace comme bloqueur de l’isoforme du récepteurα1β2γ2L que de celle durécepteurα1β1γ2L, en plus d’induire un blocage plus important de la potentialisation provoquée par l’alloprégnanoloneque par l’alphaxalone. Nos résultats sont en accord avec l’hypothèse que l’assemblage hétéromérique de différentesisoformes des récepteurs GABAA contenant différents sous-types de sous-unités engendre de multiples sites dereconnaissance stéroïdiens sur les récepteurs GABAA, et qu’il se produit, et ce de façon indépendante, différentesinteractions modulatrices entre les stéroïdes neuroactifs agissant au niveau du récepteur GABAA. Les sous-types dessous-unitésα et γ pourraient surtout influencer la modulation par l’alloprégnanolone de la fonction du récepteurGABAA, alors que les sous-types des sous-unitésβ et γ semblent être très importants pour ce qui est des effetsmodulateurs de l’alphaxalone.

Mots clés: récepteur GABAA, neurostéroïde, alloprégnanolone, alphaxalone, ovocyte deXenopus.

[Traduit par la Rédaction]

Maitra and Reynolds

Over 55 years ago, Selye (1942) reported that metabolitesof progesterone, namely allopregnanolone (5α-pregnan-3α-ol-20-one) and pregnanolone (5β-pregnan-3α-ol-20-one),had sedative and anesthetic properties. Selye’s initial workresulted in the development and introduction of syntheticsteroid anesthetics such as alphaxalone (5α-pregnan-3α-ol-11,20-dione) into clinical medicine. Recently, there has beena resurgence of interest in the neuroactive steroids, as thesecompounds are known to be important endogenous modula-tors of neuronal activity, and may provide new therapeutictools in the treatment of anxiety, depression, and seizure dis-orders (Lambert et al. 1995; Majewska 1992; Roberts 1995).The term “neurosteroid” refers to a steroid metabolite that isformed in the central nervous system (CNS), whereas theterm “neuroactive steroid” describes all steroids, regardlessof source, that are active in the brain (Hu et al. 1987).Neurosteroids are metabolites of steroid hormones includingprogesterone and androsterone, synthesized de novo in thebrain within neurons or glial cells (Lambert et al. 1995).Neuroactive steroids can alleviate anxiety and cause hypno-sis in animals (Majewska 1992). They also exhibitanticonvulsant, analgesic, and anesthetic activity in both ani-mals and humans (Britton et al. 1991; Gilron and Coderre1996; McAuley et al. 1995). In the rat, the anti-anxiety ef-fect produced by progesterone administration was found tobe directly related to the increase in allopregnanolone levelsin blood and brain (Bitran et al. 1995).

The effects exerted by steroid hormone metabolites on theCNS are mediated by regulation of neurotransmission andmembrane excitability (McEwen et al. 1984). There iscompelling evidence that neuroactive steroids, includingthe endogenous neurosteroid allopregnanolone, are potent

and selective modulators ofγ-aminobutyric acid type A(GABAA) receptor function. Neuroactive steroids potentiatethe binding of [3H]flunitrazepam to the benzodiazepine siteof the GABAA receptor (Prince and Simmonds 1993; Slanyet al. 1995), enhance muscimol-stimulated chloride flux(Morrow et al. 1990), decrease the binding oft-[35S]butyl-bicyclophosphorothionate ([35S]TBPS) to the chloride chan-nel (Davies et al. 1997), and potentiate GABA-activatedchloride current in neuronal preparations and in recombinantreceptors expressed in HEK cells andXenopusoocytes(Davies et al. 1997; Shingai et al. 1991). The allosteric mod-ulation of GABAA receptor function by neuroactive steroidsrequires a specific binding domain on the receptor channelcomplex that is distinct from those of benzodiazepines andbarbiturates (Gee et al. 1988; Turner et al. 1989). At highconcentrations, neuroactive steroids have also demonstratedthe ability to directly activate the GABAA receptor (Lambertet al. 1995). Previous studies have shown that the brain lev-els of allopregnanolone fluctuate in response to stress, andduring the estrus cycle of rats and menstrual cycles of hu-mans (reviewed in Paul and Purdy 1992). The brain levels ofthis neuroactive steroid metabolite achieved in the brainafter swim stress in male rats (10–30 nM), during estrus(15–50 nM), and during pregnancy (≥ 100 nM) are withinthe range of concentrations known to potentiate GABAA re-ceptor function.

The GABAA receptor chloride ion channel complex is ahetero-pentameric protein composed of subunits that havebeen classified into five major families,α1–α6, β1–β3, γ1–γ3,ε, andδ (Luddens et al. 1995; Macdonald and Olsen 1994;Nayeem et al. 1994; Whiting et al. 1995). There is consid-erable evidence to suggest that the subunit composition ofGABA A receptors has a significant influence on theeffects of different modulators, including theneuroactive

© 1998 NRC Canada

910 Can. J. Physiol. Pharmacol. Vol. 76, 1998

© 1998 NRC Canada

Maitra and Reynolds 911

steroids.Subunit-dependent modulation of GABAA receptorfunction by neurosteroids could have important functionalconsequences in the whole animal. The expression of differ-ent GABAA receptor subunit subtypes varies across differentbrain regions and developmental stages (Garrett et al. 1990;MacLennan et al. 1991; Persohn et al. 1991; Zimprich et al.1991). Modulation of GABAA receptor binding by allo-pregnanolone has been reported to differ significantly acrossdifferent brain regions, being much greater in areas such asthe spinal cord and hippocampus compared with the hypo-thalamus and cerebellum (Gee and Lan 1991; Korpi andLuddens 1993; Lan et al. 1991; Sapp et al. 1992; Wilson andBiscardi 1997). Binding assays and chloride flux studies inneuronal tissue suggest the presence of multiple steroid recog-nition sites on the GABAA receptor (Prince and Simmonds1993; Hawkinson et al. 1994; Morrow et al. 1990; Sapp et al.1992). It is uncertain whether multiple steroid recognitionsites in neuronal tissue from a particular brain region repre-sent multiple steroid binding sites on individual GABAAreceptors, or reflect heterogeneity of GABAA receptorisoforms each having different sensitivities to the mod-ulatory effects of neuroactive steroids. InXenopusoocytes expressing various combinations of GABAA recep-tor subunits, the efficacy of the endogenous neurosteroidallopregnanolone to potentiate GABA-evoked membranecurrent was greatest when theα1 subunit was present in ei-ther α(x)β1 or α(x)β1γ2 receptor isoforms (Shingai et al.1991). Moreover, the presence of theγ2 subunit subtype in-creased the potency of allopregnanolone in receptor iso-forms composed ofα1β1γ2 and α2β1γ2, but decreasedsensitivity to the neurosteroid in theα3β1γ2 receptor isoform(Shingai et al. 1991).

We have been studying the influence of subunit subtypesof the GABAA receptor on modulation of receptor functionby neuroactive steroids. In preliminary experiments, it be-came apparent that allopregnanolone and the closely relatedsynthetic steroid alphaxalone, had different subunit depend-ence for positive allosteric modulation of GABAA receptorfunction. We therefore conducted a systematic study to di-rectly compare allopregnanolone and alphaxalone for modu-lation of GABAA receptor function in human recombinantreceptors expressed inXenopusoocytes. The GABAA recep-tor isoforms used in this study are among the most prevalentreceptor isoforms believed to exist in the mammalian centralnervous system (McKernan and Whiting 1996). The resultsof these studies lead us to suggest that these structurallyclosely related neuroactive steroids are differentially influ-enced by the subunit composition of the GABAA receptor. Apreliminary report of this work has previously appeared inabstract form (Maitra and Reynolds 1997).

Preparation of cloned RNAIndividual DNAs (cDNAs) for the human GABAA receptor subunit

subtypes α1, β1, β2, and γ2L, cloned into the pCDM8 vector(Invitrogen), were a gift from Dr. P.J. Whiting, Merck Sharpe &Dohme Neuroscience Research Labs (Terlings Park, Essex, England).The vectors were linearized with the appropriate restrictionendonucleases, and RNA transcripts were prepared in vitro usingthe RiboMAX® large scale RNA production kit (Promega). Mix-tures of GABAA receptor subunit subtype RNAs (1 ng/nL of each

subunit subtype) were prepared to giveα1β1γ2L, α1β2γ2L, α1β1,and α1β2 subunit combinations.

Preparation of oocytesAnimals were housed and treated in accordance with the guide-

lines of the Canadian Council on Animal Care. All proceduresusing animals were reviewed and approved by the Queen’sUniversity Animal Care Committee. FemaleXenopus laevisfrogswere anesthetized by immersion in ice-cold water supplementedwith 2.5 mg/mL of the anesthetic agent 3-aminobenzoic acid ethylester (Sigma Chemical Co., St. Louis, Mo.). Surgical procedureswere performed on a bed of ice and 1 or 2 pieces of ovary were re-moved through a small incision made in the abdomen of the frogs.The oocytes were placed in a buffered saline containing (in mM):88 NaCl, 2 KCl, 1 CaCl2, 1MgCl2, 2.4 NaHCO3, 2 NaH2PO4,5 HEPES, 2 sodium pyruvate, 0.5 theophylline, supplemented withpenicillin (100 U/mL) and streptomycin (100µg/mL), pH 7.4.Stage V and VI oocytes were identified and the thecal, epithelial,and follicular layers were manually dissected away using fine for-ceps. RNA mixtures were injected into oocytes using a DrummondNanoject automatic injector (Drummond Scientific Company) andincubated for a period of 2–3 days at room temperature (20–22°C)prior to electrophysiological recording.

Electrophysiological recording and data analysisSingle oocytes were placed with the animal pole facing up, in a

square of nylon grid (1.5 × 1.5 mm) attached to the glass bottom ofthe flow chamber (RC recording chamber, Warner Instruments,8-mm well size). Individual oocytes were impaled with two glassmicroelectrodes filled with 3 M KCl (2–4 MΩ) and voltage-clampedat –70 mV using a Warner Instruments OC-725C oocyte clamp.The oocytes were perfused at the rate of 6–8 mL/min with bufferedsaline containing (in mM) 88 NaCl, 2 KCl, 1 CaCl2, 1 MgCl2,5 HEPES, pH 7.4. GABA or GABA plus drug(s) was applied tothe oocytes until the current response was observed to peak(30–60 s), and thecurrent responses thus generated were amplifiedand recorded on a strip chart recorder. A washout period of5–15 min was allowed in between drug applications, which wassufficient to allow recovery from desensitization and to reversedrug effects. Longer washout periods were sometimes employedafter high concentrations of steroids, but all drug effects were fullyreversible. Choice of solution flow was controlled by a series ofmanually operated valves connected to the inflow of the recordingchamber. Steroids were dissolved in dimethylsulfoxide as 10 mMstock solutions and serially diluted in the perfusion solution. Maximalconcentrations of the vehicle had no effect on or slightly reducedthe control GABA response (data not shown).Concentration–response relationships for receptor-activated currents elicited byGABA alone or potentiation of GABA-activated membrane current(IGABA) by allopregnanolone (ALP) or alphaxalone (AFX) were fitto the logistic equationI = I max [Cn/(Cn + EC50

n)] to obtain esti-mates of maximal response (Imax), EC50 values, and Hill coefficient(n). A nonlinear least-square fitting program was used to determinethe best fit to the data (Prism, Graphpad Software, Inc.). Where ap-propriate, one-samplet-test, paired or unpaired Studentt-tests, andrandomized design one-way analysis of variance (ANOVA) cou-pled with Bonferroni’s post-test for multiple comparisons wereused to determine the statistical significance of the data (Prism,Graphpad Software, Inc.). For individual experiments data werecollected from oocytes obtained from at least two different frogsand are expressed as means ± SEM.

All the neuroactive steroids used in this study were purchasedfrom Sigma Chemical Corp.

To allow comparisons of neuroactive steroid effects on

different GABAA receptor isoforms, an equi-effective con-centration of GABA was used for all experiments. From theGABA concentration–response curves for each receptorisoform, the EC10 concentrations were found to be approxi-mately 5 µM for the α1β1γ2L and α1β2γ2L receptorisoforms, and 0.5µM for the α1β1 and α1β2 receptorisoforms. The endogenous neurosteroid allopregnanolone(10 nM – 5 µM), produced a concentration-dependent en-hancement of the response to 5µM GABA with an EC50 of210 ± 21 nM and a maximal potentiation of 624 ± 11% inreceptors composed ofα1β1γ2L subunits (Fig. 1c). Simi-larly, allopregnanolone produced an equivalent increase inthe response to 5µM GABA in the α1β2γ2L receptor con-struct, with an EC50 of 231 ± 48 nM and a maximum facili-tation of 638 ± 48% (Fig. 1c). The EC50 values and themaximal enhancements of the GABA response producedby allopregnanolone in these receptor isoforms were notdifferent. In contrast, alphaxalone was found to produce a

© 1998 NRC Canada


Fig. 1. Theβ subunit subtype differentially affects potentiation ofGABA receptor-activated membrane current by allopregnanoloneand alphaxalone in oocytes expressing human recombinantGABAA receptors. (a) Current responses in oocytes expressingα1β1γ2L and α1β2γ2L receptor isoforms of the GABAA receptor.Membrane current elicited by 5µM GABA was potentiated byalphaxalone inboth of these receptor isoforms. (b) Concentration–response relationship for alphaxalone (0.1–50µM) to enhancethe response to 5µM GABA in the α1β1γ2L (r) and α1β2γ2L(e) receptor isoforms of the GABAA receptor. Alphaxaloneproduced a greater maximum potentiation of the GABA responsein receptors containing theβ1 subunit subtype (p < 0.005). TheEC50 for alphaxalone potentiation of the GABA response wasnot significantly different between these two receptor isoforms(Table 1). For this and all subsequent figures, the absence oferror bars for any data point indicates that the error bars weresmaller than the symbol used to represent the mean value.(c) Concentration–response relationship for allopregnanolone(0.01–5µM) to enhance the response to5 µM GABA in the α1β1γ2L (M) and α1β2γ2L (F) GABAA

receptor isoforms. There was no difference in either the efficacyor potency of allopregnanolone between these two receptors.

Allopregnanolone (ALP) Alphaxalone (AFX)

Receptor EC50 (µM) % IMax EC50 (µM) % IMax

α1β1γ2L 0.21±0.02a 624±11 5.2±2.0 927±49d(n=7) (n=8)

α1β2γ2L 0.23±0.05b 638±43 1.3±0.3 541±18d(n=7) (n=6)

α1β1 0.51±0.02a 3145±127c 3.4±0.9 2469±176e(n=9) (n=7)

α1β2 0.44±0.06b 1140±41c 2.6±0.4 2038±233e(n=7) (n=7)

Note: Data are mean ± SEM for the number of oocytes shown inparentheses. Values marked with the same letter represent statisticallysignificant differences:a, p < 0.001;b, amd e, p < 0.05; d, p < 0.005;c, p < 0.01.

Table 1. Allopregnanolone and alphaxalone potentiation ofGABA receptor-activated membrane current inXenopusoocytesexpressing human recombinant GABAA receptors.

greater enhancement of the response to 5µM GABA in theα1β1γ2L receptor isoform compared with theα1β2γ2L recep-tor isoform. Alphaxalone (10 nM – 50µM) increased cur-rent responses produced by 5µM GABA in α1β1γ2Lreceptors to a maximum of 927 ± 49% with an EC50 of 5.2 ±2 µM (Figs. 1a and 1b). The maximum calculated facilita-tion of the GABA response in theα1β2γ2L receptor isoformwas 541 ± 18% with an EC50 for alphaxalone of 1.3 ±0.3 µM (Fig. 1b). These data are summarized in Table 1.

The γ2 subunit is known to play an important role inbenzodiazepine facilitation of GABAA receptor function. Inthe absence of aγ2 subunit (receptor composed ofα andβsubunits) benzodiazepines have no effect on GABAA receptoractivation (Knoflach et al. 1992). Neurosteroids, however,have been shown to enhance GABA-activated membranecurrents in the presence and absence of theγ2 subunit(Puia et al. 1990). In the present study, both allo-pregnanolone and alphaxalone produced a robust in-crease in the response to 0.5µM GABA in receptorisoforms containingα and β subunit subtypes (Fig. 2).However, in the absence theγ2L subunit subtype, potentiationof the response to 0.5µM GABA by allopregnanolone(10 nM – 5µM) was found to be distinctly different betweenα1β1 and α1β2 receptor isoforms. Allopregnanolone pro-duced a concentration-dependent increase of the GABA re-sponse with an EC50 of 513 ± 160 nM and a maximalpotentiation of 3145 ± 127% inα1β1 GABAA receptors(Figs. 2a and 2c). In contrast, the maximal allo-pregnanolone facilitation of the GABA response inα1β2 re-ceptors was 1140 ± 41%, which was significantly lesscompared with that inα1β1 receptors (Table 1). The EC50for allopregnanolone potentiation in theα1β2 receptorisoform was 439 ± 59 nM (Table 1). Thus,allopregnanolonewas more effective in potentiating the response to 0.5µMGABA when theβ1 subunit subtype was present in receptorisoforms composed ofα1β(x) subunit combinations, with nodifference in the EC50 values for allopregnanolone. Similarly,alphaxalone modulation of the response to 0.5µM GABAwas different between theα1β1 andα1β2 receptor isoforms,although the magnitude of this difference was very small com-pared with allopregnanolone. Alphaxalone (100 nM – 20µM)potentiated the response to GABA in theα1β1 receptorisoform to a maximum of2469 ± 176% with an EC50 of

© 1998 NRC Canada


Fig. 2. Theβ subunit subtype influences potentiation of theGABA response by allopregnanolone and alphaxalone inαβreceptor isoforms of the GABAA receptor. (a) Potentiation of theresponse to 0.5µM GABA by allopregnanolone (0.5–5µM) intwo different oocytes expressingα1β1 andα1β2 GABAA

receptor isoforms. (b) Concentration–response relationship foralphaxalone (0.1–20µM) potentiation of GABA-activatedmembrane current inα1β1 (F) and α1β2 (M) receptor isoformsof the GABAA receptor. Alphaxalone was slightly more effective,with no difference in potency, on receptors containing theβ1subunit subtype (Table 1). (c) Concentration–responserelationship for allopregnanolone (0.01–5µM) potentiation ofGABA-activated membrane current inα1β1 (G) and α1β2 (O)receptor isoforms of the GABAA receptor. Allopregnanoloneexhibited a much greater maximal potentiation, but with similarpotency, of the response to 0.5µM GABA in the α1β1 comparedwith the α1β2 receptor isoform (Table 1).

3.4 ± 0.9µM (Fig. 2b). Alphaxalone was less effective inthe α1β2 receptor isoform, producing a maximal enhance-ment of the response to 0.5µM GABA of 2038 ± 233%,with an EC50 value of 2.6 ± 0.4µM (Fig. 2b). At low con-centrations of alphaxalone (<5µM), there was no difference inthe degree of potentiation of the response to 0.5µM GABA inthe α1β1 andα1β2 receptor isoforms. The calculated maximalpotentiation of GABA responses by allopregnanolone andalphaxalone with their respective EC50 values in different re-ceptor isoforms are summarized in Table 1.

© 1998 NRC Canada


Fig. 3. Effects of equivalent concentrations (the approximateEC50 values) of allopregnanolone and alphaxalone on the GABAconcentration–response relationships in theα1β2γ2L GABAA

receptor isoform. (a) Current responses obtained in two differentoocytes expressing theα1β2γ2L receptor isoform of the GABAAreceptor. Maximally effective concentrations of GABA (250–500µM)were not affected by 250 nM allopregnanolone (upper panel).However, the current response in an individual oocyte producedby 500µM GABA was still potentiated by 2µM alphaxalone(lower panel). (b) A parallel leftward shift in the GABAconcentration–response relationship is produced by 250 nMallopregnanolone in theα1β2γ2L GABAA receptor isoform. TheEC50 for the GABA activation of the receptor responses wassignificantly reduced in the presence of 250 nMallopregnanolone (Table 2). (c) A nonparallel leftward shift inthe GABA concentration–response relationship in theα1β2γ2LGABAA receptor isoform produced by 2µM alphaxalone. Themaximal effect of GABA was increased in the presence of 2µMalphaxalone. The GABA EC50 value was significantly decreasedby alphaxalone (Table 3).

GABA EC50 (µM)

Receptor Control 250 nM ALP %∆ EC50 %∆ IMax

α1β1γ2L 56±9a 22±4a 58±5 +4±6(n=8)

α1β2γ2L 66±11b 25±5b 63±3 –2±5(n=7)

α1β2 10.9±2.1c 7.0±2.4c 42±10 +38±5(n=7)Note: Data are mean ± SEM for the number of oocytes shown in

parentheses. Values marked with the same letter represent statisticallysignificant differences:a, b, and c,p < 0.005.

Table 2. GABA concentration–response data in the presence andabsence of 250 nM allopregnanolone.

GABA EC50 (µM)

Receptor Control 2µM AFX %∆ EC50 %∆ IMax

α1β1γ2L 60±9a 20±4a 69±3 +35±9(n=11)

α1β2γ2L 54±8b 15±8b 72±4 +14±5(n=10)

α1β2 6.3±0.7c 2.6±0.4c 60±4 +117±5(n=7)Note: Data are mean ± SEM for the number of oocytes shown in

parentheses. Values marked with the same letter represent statisticallysignificant differences:a, andb, p < 0.0001;c, p < 0.0005.

Table 3. GABA concentration–response data in the presence andabsence of 2µM alphaxalone.

The concentration–response relationship for GABA inthe α1β2γ2L receptor isoform was shifted to the left by250 nM allopregnanolone (approximate EC50 value)(Fig. 3b) with no change in the maximal GABA response(Fig. 3a) or the Hill slope. The EC50 for GABA wasdecreased from 66 ± 11µM to 25 ± 5 µM by 250 nMallopregnanolone. A similar shift in the GABAconcentration–response relationship without any change in the maximumwas obtained in theα1β1γ2L receptor isoform, where theEC50 for GABA was reduced to 22 ± 4µM from 56 ± 9µMby 250 nM allopregnanolone (Table 2). In contrast, 2µMalphaxalone (concentration approximately equivalent toEC50 value) produced a nonparallel leftward shift of theGABA concentration responses in bothα1β1γ2L andα1β2γ2L receptor isoforms. The EC50 for GABA was de-creased from 60 ± 9µM to 20 ± 4 µM in α1β1γ2L recep-tors, with a 35 ± 9% (p < 0.01) increase in the maximalGABA response (Table 3). Similarly, 2µM alphaxalonereduced the EC50 value for GABA from 54 ± 8µM to 15 ±8 µM in GABAA receptors composed ofα1β2γ2L subunitswith a 14 ± 5% (p < 0.05) enhancement of the maximumGABA response (Figs. 3a, 3c, and Table 3). At these con-centrations (250 nM and 2µM, respectively), allo-pregnanolone and alphaxalone did not elicit any directcurrent response in any GABAA receptor isoform tested(data not shown). Thus, the effects produced by theneuroactive steroids represent allosteric modulation of theGABA site on the receptor.

Anesthetic agents have consistently demonstrated theability to lower the EC50 for GABA and simultaneously in-crease the maximal GABA-activated receptor responses inrecombinant GABAA receptor isoforms composed ofα andβ subunits (Lam and Reynolds 1998; Horne et al. 1993).Both allopregnanolone and alphaxalone produced a nonparal-lel leftward shift in the concentration–response relationshipfor GABA in the α1β2 GABAA receptor isoform. Allo-pregnanolone (250 nM) caused a significant increase in thevalues for the maximal response to GABA by 38 ± 5% (p <0.001) and lowered the GABA EC50 value from 10.9 ±2.1 µM to 7.0 ± 2.4 µM (Figs. 4a, 4b, and Table 2).Alphaxalone (2µM) also produced a similar effect on theGABA concentration–response relationship. However, thechange in the maximal GABA response was robust and morepronounced compared with that found with allo-pregnanolone in theα1β2 GABAA receptor isoform. The

© 1998 NRC Canada


Fig. 4. Allopregnanolone and alphaxalone both produce anonparallel leftward shift in the GABA concentration–responserelationship in theα1β2 GABAA receptor isoform. (a) Currentresponses obtained from two different oocytes expressing theα1β2 GABAA receptor isoform show an increase in the responseto a maximally effective concentration of GABA (100µM) byboth250 nM allopregnanolone and 2µM alphaxalone. (b) TheGABA concentration–response relationship was shifted to the leftby 250 nM allopregnanolone, and the maximal GABA responsewas increased. The EC50 value for GABA in α1β2 receptors wassignificantly decreased by 250 nM allopregnanolone (Table 2).(c) Effect of 2µM alphaxalone on GABA concentrationresponses in theα1β2 receptor isoform. Alphaxalone caused arobust increase in the maximal response to GABA andsignificantly decreased the EC50 value for GABA (Table 3).

© 1998 NRC Canada


Fig. 5. Potentiation ofGABA-activated currents by allopregnanolone (250 nM) and alphaxalone (2µM) is differentially blocked by epipregnan-olone in human recombinantα1β1γ2L andα1β2γ2L GABAA receptor isoforms. (a andc) Current traces obtained from oocytes expressing theα1β1γ2L (a) or theα1β2γ2L (c) receptor isoforms. Traces are the membrane current response to 5µM GABA in the presence of either 250 nMallopregnanolone or 2µM alphaxalone, which were subsequently blocked by 3 and 10µM epipregnanolone. (b) Cumulative data (n = 5–14)showing the percent inhibition of the potentiation of the response to 5µM GABA by 250 nM allopregnanolone or 2µM alphaxalone caused by3 and 10µM epipregnanolone in theα1β1γ2L receptor isoform. Epipregnanolone was equally effective in inhibiting potentiation of the GABAresponse caused by equivalent concentrations of either allopregnanolone or alphaxalone.There was no change in the degree of inhibition between3 and 10µM epipregnanolone. (d) Cumulative data (n = 7 or 8) showing the percent inhibition of the potentiation of the response to 5µM GABA by250 nM allopregnanolone or 2µM alphaxalone caused by 3 and 10µM epipregnanolone in theα1β2γ2L receptor isoform. A greater inhibition wasachieved with 10µM epipregnanolone. Epipregnanolone was significantly more effective in blocking potentiation of the GABA response caused byallopregnanolone compared with alphaxalone.

© 1998 NRC Canada


maximal response was potentiated by 117 ± 5% (p < 0.0001)and the EC50 for GABA was reduced from 6.3 ± 0.7µM to2.6 ± 0.4µM with 2 µM alphaxalone (Fig. 4c, Table 3).

Epipregnanolone (5β-pregnan-3β-ol-20-one) and iso-pregnanolone (5α-pregnan-3β-ol-20-one) are inactive isomersof 3α-pregnane steroids, and themselves do not produce anyalteration in GABA-activated membrane currents. Both iso-pregnanolone and epipregnanolone were found to partiallyinhibit allopregnanolone and alphaxalone facilitation ofGABA currents. However, the degree of block was differentfor the two neuroactive steroids and in different receptorisoforms. Increasing concentrations of epipregnanolone(0.1–30µM) produced a progressive block of the potent-iation of the response to 5µM GABA caused by 250 nMallopregnanolone in theα1β2γ2L receptor isoform. Inhibi-tion caused by epipregnanolone reached a maximum at aconcentration of 10–20µM, and no further reduction wasobtained at any higher concentrations (data not shown).Similarly, epipregnanolone (0.5–10µM) decreased en-hancement of the response to 5µM GABA produced by250 nM allopregnanolone in theα1β1γ2L receptor isoform(Fig. 5a). However, in theα1β1γ2L receptor isoform, blockby epipregnanolone of the potentiation caused by allo-pregnanolone was small (approximately 25% reduction),exhibited weak concentration dependence, and was maxi-mal at 3µM epipregnanolone (Fig. 5b). In contrast, in theα1β2γ2L receptor isoform, block by epipregnanolone of thepotentiation caused by allopregnanolone was more pro-nounced (approximately 50% reduction), concentrationdependent, and maximal at approximately 10µM epi-pregnanolone (Figs. 5c and 5d). A similar profile ofepipregnanolone effects were obtained when alphaxalonepotentiation of the response to 5µM GABA was tested inboth theα1β1γ2L and α1β2γ2L GABAA receptor isoforms.That is, a greater degree of block by epipregnanolone ofthe potentiation caused by 2µM alphaxalone was observedin the α1β2γ2L receptor isoform compared with theα1β1γ2L receptor isoform (Figs. 5b and 5d). Moreover,whereas the magnitude of the block by epipregnanolone ofthe potentiation caused by allopregnanolone and alphaxalonein the α1β1γ2L receptor construct was identical (Fig. 5b),both 3 and 10µM epipregnanolone produced a greaterdegree of block of allopregnanolone compared with alphaxalonein the α1β2γ2L GABAA receptor isoform (Fig. 5d).

Isopregnanolone (0.1–10µM), the other 3β-pregnane iso-mer used for this study, also partially decreased thepotentiation of the response to 5µM GABA caused by250 nM allopregnanolone or 2µM alphaxalone in theα1β1γ2L receptor isoform (Fig. 6a). However, in contrast toepipregnanolone, the blocking effect of isopregnanolonewas significantly greater against the potentiation of theGABA response caused by allopregnanolone comparedwith alphaxalone (compare Figs. 5b and 6b). The interac-tion between isopregnanolone (10µM) and allopreg-nanolone (250 nM) or alphaxalone (2µM) to modulate theresponse to GABA was also tested on theα1β2γ2L andα1β1 receptorisoforms. Isopregnanolone was found toproduce asignificantly greater block of the potentiation ofthe GABA response caused by allopregnanolone com-pared withalphaxalone in both theα1β2γ2L and α1β1 re-ceptor isoforms (Fig. 6c).

The endogenous neurosteroid allopregnanolone showedno difference in potency or efficacy to potentiate GABAAreceptor-activated membrane current when theβ subunit sub-type was changed in receptors composed ofα1β(x)γ2L sub-units. In contrast, alphaxalone was found to be significantlymore effective to potentiate the GABA response in theα1β1γ2L receptor isoform compared with theα1β2γ2L recep-tor isoform. Previous studies have also suggested that theβsubunit subtype has a significant influence on modulation ofGABAA receptor function by alphaxalone. In theα1γ2GABAA receptor isoform expressed in human embryonickidney (HEK) cells, alphaxalone has no effect (Harris et al.1995) or slightly reduces (Slany et al. 1995) high affinity[3H]flunitrazepam binding. Inclusion of aβ subunit subtypewith α1γ2 subunits produced receptors in which alphaxalonewas able to potentiate [3H]flunitrazepam binding (Harris etal. 1995). Moreover, alphaxalone was more potent as a mod-ulator of [3H]flunitrazepam binding in theα1β2γ2 receptorisoform compared with theα1β1γ2 and α1β3γ2 receptorisoforms (Harris et al. 1995). Similarly, inXenopusoocytesexpressing theα1β1γ2S, α1β2γ2S, or α1β3γ2S receptorisoforms, alphaxalone was 2–3 times more potent to potenti-ate GABA receptor-activated membrane current in receptorscontaining theβ2 subunit subtype, but was significantlymore potent to directly activate the receptor when theβ1subunit subtype was present in the receptor construct (Sannaet al. 1997). We also found that alphaxalone was 2–3 timesmore potent in theα1β2γ2L receptor isoform compared wihthe α1β1γ2L receptor isoform (Table 1), although the dif-ference did not achieve statistical significance. Taken together,these data suggest that theβ subunit subtype has a significantlygreater influence on the modulation of GABAA receptor func-tion by alphaxalone compared with allopregnanolone.

Removal of theγ2L subunit subtype had a dramatic in-fluence on the modulatory effects of alphaxalone andallopregnanolone to potentiate GABAA receptor function.In receptor isoforms composed ofα1β1 andα1β2 subunitcombinations, both alphaxalone and allopregnanolonewere significantly more effective in potentiating the re-sponse to 0.5µM GABA, the approximate EC10 concen-tration of GABA in these receptor isoforms (Table 1).However, whereas omitting theγ2L subunit subtype hadno effect on the potency of alphaxalone to modulate re-ceptor activation, it significantly decreased the potency ofallopregnanolone at both theα1β1 and α1β2 receptorisoforms (Table 1). This is inagreement with the work ofShingai et al. (1991), who reported that the presence of theγ2 subunit increased the potency of allopregnanolone to po-tentiate GABAA receptor-activated membrane current inXenopusoocytes expressing theα1β1γ2 and α2β1γ2 receptorisoforms. Interestingly, the presence of theγ2 subunit decreasedsensitivity to allopregnanolone in theα3β1γ2 receptor isoform(Shingai et al. 1991). In the present study, removal of theγ2L subunit subtype revealed a very significant differencein the efficacy of allopregnanolone to modulate the GABAresponse in theα1β1 receptor isoform compared with theα1β2 receptor isoform (Fig. 2c). In contrast, there was lessinfluence of theβ subunit subtype on alphaxalone modula-tion of receptor function in the absence of theγ2L subunitsubtype. Taken together, our data, along with the reports by

Shingai et al. (1991) and Harris et al. (1995), suggest thatthe α and γ subunit subtypes have the greatest influenceon allopregnanolone modulation of GABAA receptor func-tion, and that theβ and γ subunit subtypes are most impor-tant for the modulatory effects of alphaxalone.

Isopregnanolone and epipregnanolone are the inactive ormarginally active isomers of the 3α-pregnane steroid metab-olites allopregnanolone and pregnanolone, respectively. Bothisopregnanolone and epipregnanolone have been reported tobe competitive antagonists of the 3α-pregnane neurosteroidbinding site on the GABAA receptor (Prince and Simmonds1992, 1993). However, there is also evidence that thesestereoisomers may act as partial agonists at the GABAA re-ceptor, resulting in weak facilitation of receptor function(Kokate et al. 1994) and partial block of the effects ofallopregnanolone (Pignataro and Fiszer de Plazas 1997;Viapiano and Fiszer de Plazas 1997). In our study, neitherisopregnanolone nor epipregnanolone had any effect on theresponse to GABA alone at concentrations up to 30µM.Higher concentrations could not be tested because of thepoor water solubility of the pregnane steroids. At least oneother study (Poisbeau et al. 1997) supports the finding thatepipregnanolone has little or no effect as a modulator ofGABAA receptor function.

In the present study, both isopregnanolone andepipregnanolone were found to at least partially blockpotentiation of the response to GABA caused by eitherallopregnanolone or alphaxalone. However, the degree ofblock produced by the 3β-pregnane steroid isomers wasdependent on the type of receptor isoform examined and theneuroactive steroid tested. In theα1β1γ2L receptor isoform,epipregnanolone was a relatively weak blocker of po-tentiation of the GABA response caused by both allo-pregnanolone and alphaxalone, with no difference in effecton the two neuroactive steroids (Fig. 5b). In contrast, in theα1β2γ2L receptor isoform, epipregnanolone was a more ef-fective blocker, and produced a significantly greater effect

© 1998 NRC Canada


Fig. 6. Isopregnanolone produces a greater inhibition ofallopregnanolone potentiation of the GABA response comparedwith alphaxalone in oocytes expressing theα1β1γ2L, α1β2γ2L,

and α1β2 GABAA receptor isoforms. (a) Current responses to5 µM GABA with either allopregnanolone or alphaxalone in thepresence or absence of isopregnanolone in oocytes expressingthe α1β1γ2L isoform are shown. (b) Potentiation of the GABAresponse by 250 nM allopregnanolone is subject to aconcentration-dependent reduction by isopregnanolone (0.1–10µM),in the α1β1γ2L receptor isoform. A maximum inhibition of thealphaxalone potentiation of the GABA response was obtainedwith 1 µM isopregnanolone. No further reduction was achievedwith a higher concentration (10µM) of the of the antagonist.However, 10µM isopregnanolone produced a greater degree ofblock of the potentiation of the GABA response produced byallopregnanolone compared with alphaxalone (n = 5–14).(c) Cumulative data showing the blocking effect of 10µMisopregnanolone on potentiation of the GABA response producedby allopregnanolone or alphaxalone in theα1β1γ2L, α1β2γ2L,and α1β1 GABAA receptor isoforms. In all receptor isoformstested, 10µM isopregnanolone was more effective in reducingpotentiation caused by allopregnanolone compared withalphaxalone.

on potentiation caused by allopregnanolone compared withalphaxalone (Fig. 5d). Isopregnanolone, the 3β-isomer ofallopregnanolone, was also more effective as a blocker ofpotentiation caused by neuroactive steroids at theα1β2γ2Lreceptor isoform (Fig. 6c). In addition, isopregnanolone wassignificantly more effective as a blocker of potentiation ofthe GABA response caused by allopregnanolone comparedwith alphaxalone in all receptor isoforms tested (Fig. 6c).There have been conflicting reports concerning the interac-tion between neuroactive steroids acting at the GABAA re-ceptor and the blocking actions of the 3β isomers. Pignataroand Fiszer de Plazas (1997) tested the effects of allo-pregnanolone, alphaxalone, and epipregnanolone on[3H]flunitrazepam binding in a chick optic lobe membranepreparation. They concluded that allopregnanolone and alphax-alone acted through a common site on the GABAA receptor,since a maximal concentration of alphaxalone failed to en-hance the binding of [3H]flunitrazepam produced byallopregnanolone alone. A very different conclusion wasreached by Prince and Simmonds (1993), who tested theinteractions between allopregnanolone, pregnanolone, alphax-alone, andtheir 3β isomers on [3H]flunitrazepam binding to amembrane preparation obtained from whole rat brain. Theyconcluded that allopregnanolone and alphaxalone had bothcommon and distinct binding sites on the GABAA receptor,and that the modulatory effects of the neuroactive steroidswere antagonized by epipregnanolone through differentmechanisms. They postulated the presence of three differentsites of action for the neurosteroids: a population ofepipregnanolone-insensitive receptors selective for alphax-alone; apopulation of epipregnanolone-sensitive recep-tors selective for pregnanolone; and a set of receptors thatare antagonized by epipregnanolone and that do not distin-guish between the neuroactive steroids. In the rat mem-brane preparation, alphaxalone did increase the bindingof [3H]flunitrazepam in the presence of a maximally effec-tive concentration of allopregnanolone (Prince andSimmonds 1993). Our results clearly support the notion thatthe heteromeric assembly of different GABAA receptorisoforms containing different subunit subtypes results inmultiple steroid recognition sites on GABAA receptors,which in turn produces distinctly different modulatory inter-actions between neuroactive steroids acting at the GABAAreceptor. While the exact nature of these binding sites re-quires further study, the interface between adjacent receptorsubunits seems a likely candidate. This would allow for mul-tiple, distinct recognition sites, the nature of which would bedetermined by the subunit composition and the arrangementof subunits within the complex. This would also account forthe cooperative interaction observed between differentclasses of subunits (e.g.,α andγ subunit subtypes) in deter-mining the pharmacological properties of neuroactive ste-roids acting at the GABAA receptor.

Bitran, D., Shiekh, M., and McLeod, M. 1995. Anxiolytic effect ofprogesterone is mediated by the neurosteroid allopregnanoloneat brain GABAA receptors. J. Neuroendocrinol.7: 171–177.

Britton, K.T., Page, M., Baldwin, H., and Koob, G.F. 1991. Anxio-lytic activity of steroid anesthetic alphaxalone. J. Pharmacol.Exp. Ther.258: 124–129.

Davies, P.A., Kirkness, E.F., and Hales, T.G. 1997. Modulation bygeneral anesthetics of rat GABAA receptors comprised ofα1β3and β3 subunits expressed in human embryonic kidney 293cells. Br. J. Pharmacol.120: 899–909.

Garrett, K.M., Saito, N., Duman, R.S., Abel, M.S., Ashton, R.A.,Fujimori, S., Beer, B., Tallman, J.F., Vitek, M.P., and Blume,A.J. 1990. Differential expression of gamma-aminobutyricacidA receptor subunits. Mol. Pharmacol.37: 652–657.

Gee, K.W., and Lan, N.C. 1991.γ-Aminobutyric acidA receptorcomplexes in rat frontal cortex and spinal cord show differentialresponses to steroid modulation. Mol. Pharmacol40: 995–999.

Gee, K.W., Bolger, M.B., Brinton, R.E., Coirini, H., and McEwen,B.S. 1988. Steroid modulation of the chloride ionophore in ratbrain: structure-activity requirements, regional dependence, andmechanism of action. J. Pharmacol. Exp. Ther.246: 803–812.

Gilron, I., and Coderre, T.J. 1996. Preemptive analgesic effects ofsteroid anesthesia with alphaxalone in the rat formalin test. An-esthesiology,84: 572–579.

Harris, B.D., Wong, G., Moody, E.J., and Skolnick, P. 1995.Diferent subunit requirements for volatile and nonvolatile anes-thetics atγ-aminobutyric acid type A receptors. Mol. Pharmacol.47: 363–367.

Hawkinson, J.E., Kimbrough, C.L., McCauley, L.D., Michael,M.B., Lan, N.C., and Gee, K.W. 1994. The neuroactive steroid 3α-hydroxy-5β-pregnan-20-one is a two-component modulator ofligand binding to the GABAA receptor. Eur. J. Pharmacol.269:157–163.

Horne, A.L., Harkness, P.C., Hadingham, K.L., Whiting, P., andKemp, J.A 1993. The influence of theγ2L subunit on the modu-lation of responses to GABAA receptor activation. Br. J.Pharmacol.108: 711–716.

Hu, Z.Y., Bourreau, E., Jung-Testas, I., Robel, P., and Baulieu,E.-E. 1987. Neurosteroids: oligodendrocyte mitochondria con-vert cholesterol to pregnenolone. Proc. Natl. Acad. Sci. U.S.A.84: 8215–8219.

Knoflach, F., Backus, K.H., Giller, T., Malherbe, P., Pflimlin, P.,Mohler, H., and Trube, G. 1992. Pharmacological andelectrophysiological properties of recombinant GABAA recep-tors comprising of theα3, β1 andγ2 subunits. Eur. J. Neurosci.4: 1–9.

Kokate, T.G., Svensson, B.E., and Rogawski, M.A. 1994. Anti-convulsant activity of neurosteroids: correlation withγ-aminobutyric acid-evoked chloride current potentiation. J.Pharmacol. Exp. Ther.270: 1223–1229.

Korpi, E.R., and Luddens, H. 1993. Regionalγ-aminobutyric acidsensitivity of t-butylbicyclophosphoro[35S]thionate binding de-pends on γ-aminobutyric acidA receptor α-subunit. Mol.Pharmacol44: 87–92.

Lam, D.W., and Reynolds, J.N. 1998. Modulatory and direct ef-fects of propofol on recombinant GABAA receptors expressed inXenopusoocytes: influence ofα- and γ2-subunits. Brain Res.784: 179–187.

Lambert, J.J., Belelli, D., Hill-Venning, C., and Peters, J.A. 1995.Neurosteroids and GABAA receptor function. Trends Pharmacol.Sci. 16: 295–303.

Lan, N.C., Gee, K.W., Bolger, M.B., and Chen, J.S. 1991. Differen-tial responses of expressed recombinant humanγ-aminobutyricacidA receptors to neurosteroids. J. Neurochem.57: 1818–1821.

Luddens, H., Korpi, E.R., and Seeburg, P.H. 1995. GABAA/benzo-diazepine receptor heterogeneity: neurophysiological implica-tions. Neuropharmacology,34: 245–254.

© 1998 NRC Canada


© 1998 NRC Canada


Macdonald, R.L., and Olsen, R.W. 1994. GABAA receptor chan-nels. Annu. Rev. Neurosci.17: 569-602.

MacLennan, A.J., Brecha, N., Khrestchatisky, M., Sternini, C.,Tillakaratane, N.J., Chiang, M.Y., Anderson, K., Lai, M., andTobin, A.J. 1991. Independent cellular and ontogenic expressionof mRNAs encoding three alpha polypeptides of the rat GABAAreceptor. Neuroscience,43:369-380.

Maitra, R., and Reynolds, J.N. 1997. Modulation of GABAA re-ceptor function by alphaxalone and allopregnanolone is differen-tially influenced by the subunit composition of the receptor.Society for Neuroscience 27th Annual Meeting. October 25–30,1997, New Orleans, LA. Society for Neuroscience, Washington,D.C. p. 960.

Majewska, M.D. 1992. Neurosteroids: endogenous bimodal modu-lators of the GABAA receptor. Mechanism of action and physio-logical significance. Prog. Neurobiol.38: 379–395.

McAuley, J.W., Reynolds, I.J., Kroboth, F.J., Smith, R.B., andKroboth, P.D. 1995. Orally administered progesterone enhancessensitivity to triazolam in postmenopausal women. J. Clin.Psychopharmacol.15: 3–11.

McEwen, B.S., Biegon, A., Fischette, C., Luine, V., Parsons, B,and Rainbow, T. 1984. Toward a neurochemical basis of steroidaction. In Frontiers in neuroendocrinology. Vol. 8.Edited byL. Martini and W.F. Ganong. Raven Press, New York. pp. 153–176.

McKernan, R.M., and Whiting, P.J. 1996.Which GABAA-receptorsubtypes really occur in the brain? Trends Neurosci.19: 139–143.

Morrow, A.L., Pace, J.R., Purdy, R.H., and Paul, S.M. 1990. Char-acterization of steroid interactions withγ-aminobutyric acid re-ceptor-gated chloride ion channels: Evidence for multiplesteroid recognition sites. Mol. Pharmacol.37: 263–270.

Nayeem, N., Green, T.P., Martin, I.L., and Barnard, E.A. 1994.Quarternary structure of the native GABAA receptor determinedby electron microscopic image analysis. J. Neurochem.62: 815–818.

Paul, S.M., and Purdy, R.H. 1992. Neuroactive steroids. FASEB J.6: 2311–2322.

Persohn, E., Malherbe, P., and Richards, J.G. 1991. In situ hybrid-ization histochemistry reveals a diversity of GABAA receptorsubunit mRNAs in neurons of the rat spinal cord and dorsal rootganglia. Neuroscience,42: 497–507.

Pignataro, L., and Fiszer de Plazas S. 1997. Epipregnanolone actsas a partial agonist on a common neurosteroid modulatory siteof the GABAA receptor complex in avian CNS. Neurochem.Res.22: 221–225.

Poisbeau, P., Feltz, P., and Schlichter, R. 1997. Modulation ofGABAA receptor-mediated IPSCs by neuroactive steroids in arat hypothalamo-hypophyseal coculture model. J. Physiol.500:475–485.

Prince, R.J., and Simmonds, M.A. 1992. 5β-Pregnan-3β-ol-20-one,a specific antagonist at the neurosteroid site of the GABAA re-ceptor complex. Neurosci. Lett.135: 273–275.

Prince, R.J., and Simmonds, M.A. 1993. Differential antagonismby epipregnanolone of alphaxalone and pregnanolonepotentiation of [3H]flunitrazepam binding suggests more thanone class of binding site for steroids at GABAA receptors.Neuropharmacology,32: 59–63.

Puia, G., Santi, M.R., Vicini, S., Pritchett, D.B., Purdy, R.H., Paul,S.M., Seeburg, P.H., and Costa, E. 1990. Neurosteroids act onrecombinant GABAA receptors. Neuron,4: 759–765.

Roberts, E. 1995. Pregnanolone—from Selye to Alzheimer and amodel of the pregnenolone sulfate binding site on the GABAAreceptor. Biochem. Pharmacol.49: 1–16.

Sanna, E., Murgia, A., Casula, A., and Biggio, G. 1997. Differen-tial subunit dependence of the actions of the general anestheticsalphaxalone and etomidate atγ-aminobutyric acid type A recep-tors expressed inXenopus laevisoocytes. Mol. Pharmacol.51:484–490.

Sapp, D.W., Witte, U., Turner, D.M., Longoni, B., Kokka, N., andOlsen, R.W. 1992. Regional variation in steroid anesthetic mod-ulation of [35S]TBPS binding toγ-aminobutyric acidA receptorsin rat brain. J. Pharmacol. Exp. Ther.262: 801–808.

Selye, H. 1942. Correlations between the chemical structure andthe pharmacological actions of the steroids. Endocrinology,30:437–453.

Shingai, R., Sutherland, M.L., and Barnard, E.A. 1991. Effects ofsubunit types of the cloned GABAA receptor on the response toa neurosteroid. Eur. J. Pharmacol.206: 77–80.

Slany, A., Zezula, J., Fuchs, K., and Sieghart, W. 1995. Allostericmodulation of [3H]flunitrazepam binding to recombinantGABAA receptors. Eur. J. Pharmacol.291: 99–105.

Turner, D.M., Ransom, R.W., Yang, J.S.-J., and Olsen, E.W. 1989.Steroid anesthetics and naturally occuring analogs modulate theγ-aminobutyric acid receptor complex at a site distinct from bar-biturates. J. Pharmacol. Exp. Ther.248: 960–966.

Viapiano, M.S., and Fiszer de Plazas, S. 1997. 3 Beta-OH-5beta-pregnan-20-one enhances [3H]GABA binding in develop-ing chick optic lobe. J. Recept. Signal Transduct. Res.17: 585–597.

Whiting, P.J., McKernan, R.M., and Wafford, K.A. 1995. Structureand pharmacology of vertebrate GABAA receptor subtypes. Int.Rev. Neurobiol.38: 95–139.

Wilson, M.A., and Biscardi, R. 1997. Influence of gender and brainregion on neurosteroid modulation of GABA responses in rats.Life Sci. 60: 1679–1691.

Zimprich, F., Zezula, J., Seighart, W., and Lassman, H. 1991.Immunohistochemical localization of theα1, α2 andα3 subunitof the GABAA receptor in the rat brain. Neurosci. Lett.127:125–128.

modulation of gaba a receptor function by neuroactive steroids: evidence for heterogeneity of...

Documents