Structural Link between -Aminobutyric Acid Type A (GABAA

Structural Link between �-Aminobutyric Acid Type A(GABAA) Receptor Agonist Binding Site and Inner �-SheetGoverns Channel Activation and Allosteric Drug Modulation*

Received for publication, October 24, 2011, and in revised form, January 2, 2012 Published, JBC Papers in Press, January 4, 2012, DOI 10.1074/jbc.M111.316836

Srinivasan P. Venkatachalan and Cynthia Czajkowski1

From the Department of Neuroscience, University of Wisconsin-Madison, Madison, Wisconsin 53711

Background: Structural elements and protein movements underlying GABAA receptor activation are not completelyresolved.Results: Glycine insertions in the extracellular �4-�5 linker decrease GABA activation, invert antagonist efficacy, and reduceallosteric modulation.Conclusion: �4-�5 linker is critical for mediating actions of GABAA receptor orthosteric and allosteric ligands.Significance: Detailed structural-functional understanding of GABAA receptor activation is key to understanding its modula-tion in diseased and healthy states.

Rapid opening and closing of pentameric ligand-gated ionchannels (pLGICs) regulate information flow throughout thebrain. For pLGICs, it is postulated that neurotransmitter-in-duced movements in the extracellular inner �-sheet triggerchannel activation. Homology modeling reveals that the �4-�5linker physically connects the neurotransmitter binding site tothe inner �-sheet. Inserting 1, 2, 4, and 8 glycines in this regionof the GABAA receptor �-subunit progressively decreasesGABA activation and converts the competitive antagonistSR-95531 into a partial agonist, demonstrating that this linker isa key element whose length and flexibility are optimized forefficient signal propagation. Insertions in the �- and �-subunitshave little effect on GABA or SR-95531 actions, suggesting thatasymmetric motions in the extracellular domain power pLGICgating. The effects of insertions on allostericmodulator actions,pentobarbital, and benzodiazepines, have different subunitdependences, indicating that modulator-induced signaling isdistinct from agonist gating.

Electrochemical signaling in the CNS depends on ligand-gated ion channels (LGICs).2 These proteins couple the bindingof neurotransmitter to the rapid opening of an integral ion-conducting pore. The “Cys-loop” LGIC family of receptorscomprises pentameric proteins (pLGICs) that include nicotinicacetylcholine receptors (nAChRs), glycine receptors (GlyRs),GABA type A receptors (GABAARs), and serotonin type-3receptors. Although a structural picture of pLGICs is rapidlyemerging from the 4 Å resolution cryo-EM structure of the

Torpedo nAChR (1), the crystal structures of the extracellularbinding domain of the nAChR �-subunit (2) liganded and unli-ganded acetylcholine-binding proteins (AChBP), which arehomologs of the extracellular binding domain (3, 4), the crystalstructures of full-length prokaryotic pLGIC homologs fromErwinia chrysanthemi (ELIC) and Gloeobacter violaceus(GLIC) (5–7), as well as the recent crystal structure of a relatedinvertebrate pLGIC (8), our understanding of the structural ele-ments and protein movements that couple neurotransmitterbinding to channel gating is still under debate. For receptors inthis superfamily, the neurotransmitter binding site is located inthe extracellularN-terminal domain between adjacent subunitsformed by at least six noncontiguous protein regions (loopsA–F), whereas the channel gate is located 50 Å away in thetrans-membrane region (9). Neurotransmitter binding isbelieved to trigger structuralmovements at the binding site thatare propagated as a conformational wave to the channel gate(10). The secondary structure of the extracellular domain ispredominantly composed of 10 �-strands arranged in twosheets, inner and outer, that form a �-sandwich (see Fig. 1).Three flexible loops (�4-�5, �6-�7, and �8-�9 linkers) link theinner �-sheet to the outer sheet.Onemodel of receptor activation based on nAChR structural

studies suggests that agonist binding in the extracellulardomain induces a clockwise rotation of the extracellular inner�-sheet in two out of five subunits, which triggers movementsin the trans-membrane helices that result in channel gating(11). Comparison of the recent ELIC and GLIC bacterial chan-nel structures (closed and open, respectively) suggests thatchannel activation is accompanied by an anti-clockwise con-certed twist of each extracellular �-sandwich domain (6, 7, 12).In homomeric glycine receptors, receptor activation is believedto occur via a reorganization of the extracellular �-sandwichhydrophobic core and the negative subunit interface loops (13).In the GABAAR, the �-subunit �4-�5 linker links the loop A

region of theGABAbinding site to the inner�-sheet (see Fig. 1)and is in an ideal position to propagate initial binding sitemove-ments to gating movements in the channel domain. Consistent

* This work was supported, in whole or in part, by National Institutes of HealthGrant 34727 (to C. C.) from the NINDS.

1 To whom correspondence should be addressed: Dept. of Neuroscience, Uni-versity of Wisconsin-Madison, 601 Science Dr., Rm. 135, Madison, WI 53711.Tel.: 608-265-5863; Fax: 608-265-7821; E-mail: [email protected].

2 The abbreviations used are: LGIC, ligand-gated ion channel; pLGIC, penta-meric ligand-gated ion channel; ELIC, pLGIC homolog from E. chrysan-themi; GLIC, pLGIC homolog from G. violaceus; GABAAR, GABA type Areceptor; PB, pentobarbital; FZ, flurazepam; BZD, benzodiazepine; nAChR,nicotinic acetylcholine receptor; AChBP, acetylcholine-binding protein;ANOVA, analysis of variance.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 287, NO. 9, pp. 6714 –6724, February 24, 2012© 2012 by The American Society for Biochemistry and Molecular Biology, Inc. Published in the U.S.A.

6714 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287 • NUMBER 9 • FEBRUARY 24, 2012

by guest on April 4, 2018

http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

with this idea, molecular dynamic simulations of the nAChRidentifiedmotions in the�4-�5 linker that were correlatedwithmotions in loops 2, 7, andM2-M3 near the extracellular mouthof the channel (14), and mutations in the nAChR �-subunit�4-�5 linker alter channel gating (15, 16). Based on the crystalstructures of the nAChR �-subunit extracellular domain,AChBP, GLIC, ELIC, and the invertebrate glutamate-activatedchloride channel (2, 3, 5–8), this linker region spans each sub-unit and is relatively unstructured. Here, we inserted glycineresidues in the �4-�5 linkers of the GABAAR �1-, �2-, and�2-subunits to alter their length and flexibility and examinedthe effects on GABA-mediated channel activation and pento-barbital (PB)-mediated channel activation and on benzodiaz-epine (BZD) modulation of GABA responses. The absence of a�-carbon allows glycine to access energetically prohibited pro-tein dihedral angles (17).Glycine insertions in the �-subunit (�Gly)3 reduced GABA

binding and GABA-induced channel gating, and surprisingly,

converted the competitive antagonist SR-95531 into a partialagonist, demonstrating that this linker is a key structuralelement whose length and flexibility are optimized for trans-ducing GABA binding to efficient channel gating. Insertionsin the �- and �-subunits had little to no effect on GABA orSR-95531 actions, supporting the idea that asymmetric sub-unit motions in the extracellular domain help power GABA-mediated gating. Moreover, the effects of the glycine inser-tions on PB activation and BZD modulation were muchsmaller and had different subunit dependences, indicatingthat the structural mechanisms underlying GABA, PB, andBZD actions are distinct.

EXPERIMENTAL PROCEDURES

Mutagenesis and Expression in Oocytes—Rat cDNAs encod-ing �1-, �2-, and �2S-subunits of the GABAAR, subcloned intothe pGH19 vector (18), were utilized in this study. Mutantreceptors with 1-, 2-, 4-, and 8-glycine insertions (Gly1, Gly2,Gly4, andGly8) in the�-,�-, and �-subunits were created usingQuikChange site-directedmutagenesis kit (Stratagene, La Jolla,CA). Allmutant cDNAswere confirmed using double-strandedDNA. GABAARs were expressed in Xenopus laevis oocytes as

3 Throughout this study, �Gly represents glycine insertions in the �-subunit;�Gly represents glycine insertions in the �-subunit; and �Gly representsglycine insertions in the �-subunit. In addition, Gly1, Gly2, Gly4, and Gly8indicate mutant receptors with 1-, 2-, 4-, and 8-glycine insertions.

FIGURE 1. GABAAR homology model. A, left, top down view of a pentameric GABAAR (��, counterclockwise) depicting the position of �4-�5 linker (blue)in each subunit. Right, detailed view of a single GABAAR �-subunit amino-terminal ligand binding domain depicting the system of inner (magenta) and outer(cyan) �-strands, binding-site loops (A, B, and C), and the �4-�5 linker (blue) viewed from within the extracellular channel vestibule and normal to themembrane. �K103 is shown in sticks, after which 1, 2, 4, or 8 glycines were inserted. B, sequence alignment of rat GABAAR �1-, �2, and �2-subunits with othermembers of the pLGIC family. Arrows indicate positions of the glycine insertions in the �4-�5 linker (highlighted in blue) and identified GABA (peach) and BZD(green) binding site residues in loops A and E (29 –31, 37, 52) are depicted. Torp. nAChR, Torpedo nAChR; LS-AChBP, Lymnea stagnalis-AChBP; GluCl�, glutamate-gated chloride channel alpha.

Structural Determinants of GABAAR Activation

FEBRUARY 24, 2012 • VOLUME 287 • NUMBER 9 JOURNAL OF BIOLOGICAL CHEMISTRY 6715


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

described previously (19) and recorded from 2–14 days afterinjection.Oocyte Electrophysiology—Oocytes were voltage-clamped at

�80 mV in a 200-�l chamber and continuously perfused withND96 at 10 ml/min, and data were acquired as described pre-viously (20). Stock solutions of 1 M GABA, 10 mM SR-95531(Sigma-Aldrich), and 10 mM flurazepam (FZ) (Research Bio-chemicals, Natick, CA), were prepared in ND96, stored at�20 °C, and thawed once before use. Stock solutions of 30 or 50mM PB (Research Biochemicals, Natick, CA) were preparedfresh on the day of the experiment.Concentration Response and Data Analysis—Concentration

response analyses were performed as described previously (21).For GABA concentration response experiments, each test con-centration was preceded by a low nondesensitizing concentra-tion to correct for the drift in IGABA over the course of theexperiment. Currents induced by each test concentration werenormalized to the corresponding nondesensitizing concentra-tion before curve fitting.Concentration responses for PB direct activation of wild-

type (WT) and mutant receptors were measured using twomethods. In the first method, each test PB concentration waspreceded by a low nondesensitizing PB concentration to cor-rect for any drift in PB-induced current during the experiment.Responses to each test PB concentration were then normalizedto its corresponding low nondesensitizing concentration, priorto curve fitting. In the secondmethod, the PB-induced currentswere not normalized to a low, nondesensitizing PB concentra-tion. The curve fits and calculated values obtained using the twomethods were not different, and thus, the data were pooled forstatistical analysis. At high micromolar concentrations andabove, PB blocks GABAAR current responses. The relief ofchannel block upon drug wash yields a characteristic tail cur-rent. For PB concentration-response curves, PB current ampli-tudes at high micromolar concentrations were measured usingthe tail currents.SR-95531 IC50 experiments were performed as described

previously (22). Oocytes expressingWT or mutant receptorswere challenged with EC50 GABA concentration (except forexperiments involving �Gly2, -4, and -8, which were per-formed at 30 mM GABA) followed by co-application of thesame concentration of GABA and a test concentration ofSR-95531. GABA-induced currents in the presence ofincreasing SR-95531 concentrations were then normalizedto the GABA response in the absence of SR-95531 beforecurve fitting.FZ potentiation experiments were performed at EC4–8

GABA. Potentiation is defined as ((IGABA�BZD/IGABA) � 1),where IGABA�BZD is the GABA-mediated current in the pres-ence of flurazepamand IGABA is theGABA-mediated current inthe absence of flurazepam. Because GABA EC50 values couldnot be reliably measured for �Gly2-, -4-, and -8-containingreceptors, they were not tested for flurazepam potentiation.Nonlinear regression analysis for GABA, PB, FZ, and

SR-95531 concentration response experiments was performedusing the GraphPad (San Diego, CA) Prism 4 software. GABAand PB concentration responses were fit to the following equa-tion: I � Imax/(1 � EC50/[A])n), where I is the peak response

(including the tail current for PB) to a given GABA or PBconcentration, Imax is the current amplitude (including thetail current for PB) of the maximal GABA- or PB-evokedcurrent, EC50 is the concentration of GABA or PB that pro-duces a half-maximal response, [A] is the agonist concentra-tion, and n is the Hill coefficient. For SR-95531 competitionexperiments, inhibition was calculated as IGABA�SR-95531/IGABA. Data were fit to the following equation: inhibition �1 � 1/(1 � (IC50/[Ant])n), where IC50 is the concentration ofantagonist that blocks half of IGABA; [Ant] is the concentra-

FIGURE 2. Glycine insertions in �-subunit drastically affect GABA activa-tion of GABAARs. Left panel, GABA concentration-response curves fromoocytes expressing wild-type (WT) (�, dashed line) and mutant �� GABAARscontaining 1 (Œ), 2(�), 4 (�), or 8 (●) glycine insertions in �- (top), �- (middle),or �-subunits (bottom). Data points represent the mean � S.E. from �3 inde-pendent experiments. Calculated GABA EC50 values and Hill coefficients arereported in Table 1. Right panel, representative current traces from oocytesexpressing �Gly2��, ��Gly2�, and ��Gly2 receptors elicited by increasingconcentrations of GABA (�M or mM).




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

tion of the antagonist; and n is the Hill coefficient. KI valueswere calculated using the Cheng-Prusoff/Chou equation (23,24):KI � IC50/(1� [A]/EC50), where [A] is the concentrationof GABA used, and EC50 is the concentration of GABA thatelicits a half-maximal response.

Current Rise Time and Maximal Current Amplitude RatioAnalysis—10�90% apparent rise times forGABA-induced cur-rents at saturating GABA concentrations (10 mM for WT and100 mM for �Gly1) were calculated using a built-in feature inthe WinWCP data acquisition software (provided by J. Demp-ster, Univ. of Strathclyde, Glasgow, UK). Data from n � 3oocytes from at least two different batches were pooled for sta-tistical analysis.For maximum PB versus maximum GABA-induced current

ratio determinations, each oocyte (WTor�Gly1) was first chal-lenged with maximum GABA concentration (10 mM for WTand 100mM for �Gly1) and then allowed to recover completelyby washing sufficiently to remove any trace remnants of GABAfrom previous exposure, which was followed by application ofmaximum PB concentration (10 mM for WT and 30 mM for�Gly1). GABA- and PB-induced currents at saturating concen-trations were then measured, and ratios (maximum PB/maxi-mum GABA) were determined by dividing the maximum PBcurrent amplitude by maximum GABA current amplitude foreach experiment. Ratiometric data from n � 3 oocytes from atleast two different batches were pooled for statistical analysis.PB current amplitudes at saturating concentrations were mea-sured using tail currents.Statistical Analysis—LogEC50 values for GABA, PB, FZ,

potentiation, and LogKI values for SR-95531 concentrationresponses were analyzed using one-way ANOVA followed by apost hoc Dunnett’s test to determine the level of significancebetween wild-type (WT) and mutant receptors at an �-level of0.05. The Dunnett’s test compares group means and is used toidentify samples whose means are significantly different fromthe mean of a reference group, in our case the WT sample.10–90% rise times in response to maximal GABA concentra-tion and ratios of maximum PB versus maximum GABA cur-rent amplitudes for oocytes expressing WT and mutantGABAARs were analyzed using unpaired two-tailed Student’s ttest. All data reported are mean � S.E. unless noted otherwise.Structural Modeling—Homology modeling was performed

as described previously (20). Briefly, we modeled the GABAARextracellular domain after the AChBP (3) structure and mod-eled its trans-membrane domain after the structure of thenAChR trans-membrane domain (at a resolution of 4 Å) solvedbyMiyazawa et al. (11) (Protein Data Bank (PDB) code 1OED).

FIGURE 3. �Gly1 reduces GABA efficacy and slows GABA rise times.A, representative current traces from WT- or �Gly1-expressing oocytes inresponse to sequential applications of saturating concentrations of GABAand PB are depicted. Concentrations of GABA and PB (in mM) used were 10and 10 for WT and 100 and 30 for �Gly1 receptors. Bars represent mean � S.E.of maximum PB/maximum GABA current ratios from (n) experiments for WTand ��Gly1� receptors. B, representative saturating GABA current tracesfrom WT- and �Gly1-expressing oocytes are peak normalized to highlight theslowing in GABA 10 –90% rise time for �Gly1 mutant receptors. Bars representmean � S.E. of 10 –90% rise times (in seconds) from (n) experiments for WTand ��Gly1� receptors. *, values are significantly different from WT, p � 0.05(one-way ANOVA).

TABLE 1Summary of GABA and SR-95531 concentration response dataData are mean � S.E. for n experiments from at least two different batches of oocytes. Mutant/WT (mut/WT) EC50 or IC50 ratios were calculated. SR-95531 experimentswere done at EC50 GABA for WT, �Gly, �Gly1, and �Gly insertions and at 30 mM GABA for �Gly2, 4, and 8 insertions. *, values are significantly different fromWT, p �0.05 (one-way ANOVA).

Receptor GABA (EC50, �M) nH mut/WT n SR-95531 (IC50, nM) SR-95531 (KI) nH mut/WT IC50 n

WT 14.1 � 2.2 1.4 � 0.2 1 6 158 � 22 76 � 11 �1.1 � 0.1 1 3�Gly1 28.8 � 3.5 1.2 � 0.1 2 3 272 � 27 133 � 13* �1.1 � 0.1 2 3�Gly2 27.2 � 5.4 1.3 � 0.2 2 3 142 � 17 70 � 9 �1.1 � 0.1 1 5�Gly4 155 � 70* 0.9 � 0.1 11 3 338 � 23 187 � 13* �1.2 � 0.1 2 3�Gly8 31.0 � 2.1* 1.2 � 0.1 2 4 270 � 45 116 � 19 �1.0 � 0.1 2 3�Gly1 370 � 90* 0.6 � 0.1* 26 8 1390 � 350 767 � 192* �1.0 � 0.1 9 3�Gly2 �150 mM �10,000 10 7500 � 1400 �1.0 � 0.1 47 3�Gly4 �150 mM �10,000 3 31,000 � 11,300 �1.0 � 0.1 200 3�Gly8 � 1 M �50,000 3 13,300 � 3100 �1.1 � 0.2 85 4�Gly1 41.5 � 3.3* 1.2 � 0.1 3 3 195 � 7 100 � 3 �1.3 � 0.1 1 3�Gly2 43.7 � 12.6* 1.2 � 0.2 3 3 252 � 18 118 � 8 �1.2 � 0.1 2 3�Gly4 27.1 � 10 1.4 � 0.2 2 4 180 � 13 76 � 2.5 �1.2 � 0.1 1 3�Gly8 43.0 � 4* 1.1 � 0.1 3 3 234 � 18 117 � 9 �1.3 � 0.1 2 3




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Subsequently, the two structures were docked, and globalenergy minimizations were undertaken followed by examiningthe protein for gross structural distortions. The GABAARmodel images were developed using PyMOL (Schrödinger,LLC, New York).

RESULTS

Effects of Glycine Insertions on GABA Actions—Gly1, Gly2,Gly4, or Gly8 were inserted after position Lys-103 in the �-sub-unit (�K103) and after aligned positions in the �- (�K105) and�- (�K118) (Fig. 1,A andB) subunits to evaluate how increasing

the length and flexibility of the linker regions in the �-, �-, and�-subunits would affect GABAAR function. Oocytes wereinjected with mutant and wild-type �-, �-, and �-subunitcRNAs to form �� receptors and functionally characterizedusing two-electrode voltage clamp. All of the mutant subunitsassembled into functional receptors that responded to GABA.In general, the�Gly and �Gly insertions hadminimal effects onGABAEC50 values (�4-fold) as compared withWT (14.1� 2.2�M), except for �Gly4 that had an 11-fold increase in GABAEC50 (Fig. 2, Table 1). The �Gly and �Gly insertions had noeffects on the Hill slopes for GABA activation. The maximalGABA-activated currents elicited from receptors containingthe �Gly or �Gly insertions ranged from 3.5 to 11 �A and didnot significantly differ fromWT (9.8 � 1 �A; n � 6).

In contrast, �Gly1 increased GABA EC50 26-fold and signif-icantly reduced the Hill slope, whereas �Gly2, -4, or -8increased GABA EC50 more than 10,000-fold (Fig. 2, Table 1).

FIGURE 4. Glycine insertions in only �-subunit affect SR-95531 inhibitionof GABA-activated current. Left panel, SR-95531 inhibition curves of GABA(EC50) currents from oocytes expressing wild-type (�, dashed line) andmutant �� GABAARs containing 1 (Œ), 2 (�), 4 (�), or 8 (●) glycine insertionsin �- (top), �- (middle), or �-subunits (bottom). Data points represent themean � S.E. from �3 independent experiments. Calculated SR-95531 KI val-ues and Hill coefficients are reported in Table 1. Right panel, representativecurrent traces from oocytes expressing �Gly2�� (top), ��Gly2� (middle), and��Gly2 (bottom) receptors. Traces from �Gly2�� and ��Gly2 receptors areGABA (EC50) currents in the absence and presence of SR-95531. Traces from��Gly2� receptors are from applications of 100 mM GABA or 100 �M SR-95531to illustrate that SR-95531 alone elicits currents and is a partial agonist for thismutant receptor.

FIGURE 5. Glycine insertions in both �-subunits and �-subunits affect PBactivation of GABAARs. Left panel, PB concentration-response curves fromoocytes expressing wild-type (�, dashed line) and mutant �� GABAARs con-taining 1 (Œ), 2 (�), 4 (�), or 8 (●) glycine insertions in �- (top), �- (middle), or�-subunits (bottom). Data points represent the mean � S.E. from �3 inde-pendent experiments. Calculated PB EC50 values and Hill coefficients arereported in Table 2. Right panel, representative current responses fromoocytes expressing �Gly2��, ��Gly2�, and ��Gly2 receptors elicited byincreasing concentrations of PB (�M or mM). Peak PB-activated currents weremeasured after wash-out (tail current) and used for curve fitting.




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

GABA EC50 values for �Gly2, -4, or -8 could only be estimatedbecause the high concentrations of GABA (�300 mM) neededto reach maximal current responses changed the extracellularsolution osmolarity and could not be used. Themaximal GABAcurrent amplitudes for �Gly1 receptors were significantlysmaller (1.28� 0.1�A;n� 14) as comparedwithWTreceptors(9.8� 1 �A; n� 6). To determine whether �Gly1 was affectingGABA efficacy and/or receptor expression, we measured andcompared currents induced by a saturating GABA concentra-tion with those induced by a saturating PB concentration in thesame oocyte. For WT receptors, saturating concentrations ofPB and GABA elicited currents similar in magnitude (IPB max/IGABAmax ratio � 1.03 � 0.07, Fig. 3A). In oocytes expressing�Gly1 receptors, the currents elicited by saturating concentra-tions of GABA were 4-fold smaller than currents induced bysaturating concentrations of PB, indicating a reduction inGABA efficacy (Fig. 3A). We also measured apparent 10–90%current rise times at saturating GABA concentrations for�Gly1 andWT receptors. Although current onset is limited bythe slow solution-exchange timeswhen recording fromoocytes(300 ms), differences in apparent current rise times betweenWT andmutant receptors would imply changes in channel gat-ing. GABA apparent rise times for �Gly1 receptors were signif-icantly slower thanWT receptors (0.35 � 0.02 s for WT versus1.5 � 0.2 s for �Gly1, Fig. 3B), suggesting that insertion of asingle glycine in the �4-�5 linker of the �-subunit reducedchannel opening.Effects of Glycine Insertions on Gabazine (SR-95531) Actions—

We also examined the effect the glycine insertions had on theability of the competitive antagonist SR-95531 to inhibitGABA-activated currents.�Gly and�Gly insertions caused lessthan 3-fold changes in SR-95531 KI as compared with WT(76 � 11 nM) (Fig. 4, Table 1). Larger increases in SR-95531 KIwere observed for the �Gly insertions. Insertion of a single gly-cine resulted in an �10-fold increase in SR-95531 KI (Table 1).Because the GABA EC50 values were too right-shifted to beprecisely determined for �Gly2, -4, and -8 receptors (Table 1),we approximated the -fold changes in SR-95531 KI using theCheng-Prusoff equation (KI � IC50/(1 � ([A]/EC50)), where Ais the concentration of GABA used in the experiment, EC50 isthe GABA concentration that elicits half-maximal response,and IC50 is the concentration of SR-95531 that inhibits 50% of

the GABA-induced current. The equation predicts that at [A]�� EC50, KI approaches IC50, and at [A] � EC50, KI � (IC50)/2.The SR-95531 inhibition experiments for�Gly2, -4, and -8 used30 mM GABA, which from curve-fitting estimates is below theGABA EC50 (Table 1). Thus, for �Gly2 receptors, we estimatethat the SR-95531 KI is between 3750 nM (using 30 mM as anupper limit forGABAEC50) and 7500 nM (assuming that 30mM

is �� EC50), an �50–100-fold increase in KI as compared withWT. The estimated -fold increases in SR-95531 KI for �Gly4and -8 receptors were even larger (100–400-fold).

FIGURE 6. Glycine insertions in �-, �-, and �-subunits decrease FZ poten-tiation of GABA-mediated currents. Left panel, FZ concentration-responsecurves from oocytes expressing wild-type (�, dashed line) and mutant ��GABAARs containing 1 (Œ), 2 (�), 4 (�), or 8 (●) glycine insertions in �- (top), �-(middle), or �-subunits (bottom). Potentiation is defined as {((IGABA � FZ)/IGABA) � 1}, where IGABA � FZ is the GABA current in the presence of FZ andIGABA is the current elicited from GABA (EC4 – 8 concentration). Data pointsrepresent the mean � S.E. from �3 independent experiments. Calculated FZEC50 values, maximal potentiation values, and Hill coefficients are reported inTable 3. Right panel, representative GABA (EC4 – 8) current responses fromoocytes expressing �Gly2��, ��Gly1� and ��Gly2 receptors elicited in theabsence and presence of increasing concentrations of FZ (�M). Note that FZpotentiation of GABA responses was not measured for �Gly2-, -4-, and -8-con-taining receptors (see “Results”).

TABLE 2Summary of Pentobarbital concentration response dataData are mean � S.E. for n experiments from at least two different batches ofoocytes. Mutant/WT (mut/WT) EC50 ratios were calculated. *, values are signifi-cantly different fromWT, p � 0.05 (one-way ANOVA).

Receptor PB (EC50, �M) nH mut/WT n

WT 231 � 10 2.4 � 0.3 1 3�Gly1 288 � 84 2.5 � 0.1 1 3�Gly2 385 � 56 2.6 � 0.2 2 4�Gly4 720 � 74* 1.3 � 0.1* 3 3�Gly8 484 � 100* 1.5 � 0.2* 2 4�Gly1 1910 � 320* 1.2 � 0.1* 8 8�Gly2 3020 � 510* 2.1 � 0.2 13 5�Gly4 1840 � 320* 2.1 � 0.2 8 5�Gly8 2500 � 200* 2.5 � 0.1 11 3�Gly1 1280 � 60* 2.9 � 0.2 6 5�Gly2 2230 � 310* 2.4 � 0.2 10 3�Gly4 1390 � 50* 2.5 � 0.1 6 3�Gly8 1570 � 150* 2.0 � 0.3 7 4




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Interestingly, even at high concentrations, SR-95531 onlyinhibited 70% of the GABA-induced current for �Gly2-, -4-,and -8-containing receptors as compared with 100% inhibitionseen for WT and �Gly1 receptors (Fig. 4). To investigate themechanism underlying the partial inhibition, we appliedSR-95531 in the absence of GABA. At concentrations �30 �M,SR-95531 directly gated �Gly2-, -4-, and -8-containing recep-tors (Fig. 4, middle inset, data not shown for �Gly4 and -8).These concentrations of SR-95531 never elicited currents fromWT-or�Gly1-containing receptors. Thus, for�Gly2, -4, and -8receptors, SR-95531 behaved as a weak partial agonist.Effects of Glycine Insertions on PBActions—PB is an allosteric

modulator of the GABAAR that binds at a site distinct fromGABA (25). At high concentrations, PB can directly open thechannel. The single channel conductances of GABAARs acti-vated by PB and GABA are similar (26, 27), suggesting that thereceptor open-state channel structures induced by their bind-ing are similar (28). To test whether disrupting the linkerregions in the �-, �-, and �-subunits affected PB activation, wemeasured PB concentration responses from wild-type andmutant receptors. Glycine insertions in both the �-subunitsand the �-subunits significantly increased PB EC50 values by�10-fold as compared with WT receptors (231 � 10 �M) (Fig.5, Table 2). The�Gly insertions alteredPBEC50� 3-fold (Fig. 5,Table 2). The mean maximal currents elicited by PB for all themutant receptors ranged from 3.3 to 12.4 �A and did not sig-nificantly differ fromWT (9.1� 1.3�A;n� 12), indicating thatnone of the glycine insertions in any of the subunits alteredGABAAR surface expression.Effects of Glycine Insertions on FZ Actions—BZDs modulate

GABA responses by binding at a site formed at the interfacebetween the extracellular N-terminal regions of the �- and�-subunits (29–31) (Fig. 1, A and B). At subsaturating concen-trations of GABA, positive BZD modulators increase GABA-induced current. To examine whether the linker regions wereinvolved in mediating BZD-positive allosteric modulation, wetested the effects the Gly insertions had on the ability of theBZD-positive modulator FZ to potentiate GABA currents. ForWT GABAARs, FZ maximally potentiated EC4–8 GABA-in-duced current with a potentiation value of 2.49 � 0.18 and anEC50 of 468 � 94 nM (Fig. 6, Table 3). All of the �Gly and the�Gly1 insertions reduced FZ potentiation of EC4–8GABA-me-diated current responses (50–60%) without changing the FZEC50 (Fig. 6, Table 3), suggesting that these regions are impor-tant for mediating BZD efficacy. Because the magnitude of the

maximal BZD potentiation of IGABA measured is highlydependent upon the effective GABA concentration beingapplied (32), FZ potentiation of GABA currents for �Gly2-, -4-,and -8-containing receptors were not measured because theirGABA EC50 values could not be precisely determined. The�Gly insertions not only reduced FZpotentiation ofGABAcur-rents (40–50% for �Gly4 and �Gly8) but also increased FZEC50values �4-fold (Fig. 6, Table 3).

DISCUSSION

Here, by altering the length and flexibility of the �-subunit�4-�5 linker, we demonstrate that this linker plays a criticalrole in mediating agonist-induced GABAAR functionalresponses. Moreover, agonists and allosteric modulators reactuniquely toward insertions in the linker in the�-,�-, and�-sub-units of the GABAAR, suggesting that the allosteric trajectoriesunderlying their actions are distinct. Overall, our data identifythe �4-�5 linker as a key structural element in the GABAARthat shapes the energetic landscape associated with channelactivation and drug modulation.

�-Subunit Linker Region Couples GABA Binding and Gating—Glycine insertions in the �-subunit linker region, which physi-cally connects the GABA binding site (loop A) to the extracel-lular domain inner �-sheet, increased GABA EC50 more than10,000-fold, whereas insertions in the �- and �-subunits hadlittle to no effect on GABA EC50 (Fig. 7A, Table 1). These dataare consistent with structural and molecular dynamic studies(1, 3, 4, 33–35) in the pLGIC family that indicate asymmetricsubunit motions in the extracellular domain help power ago-nist-mediated gating. Recently, photochemical cleavage of the�-subunit GABAAR linker was shown to disrupt GABA activa-tion (36). In this study, the cleavage site was located near the�/� interface close to loop E of the GABA binding site (at�M113, see Fig. 1). We speculate that proteolysis at this sitelikely altered the structure of the GABA binding site, whichresulted in the loss of GABA-mediated functional responsesobserved. Here, the �-subunit glycine insertions are locatednear the �/� and �/� interfaces (at non-GABA binding siteinterfaces). Surprisingly, inserting even up to 8 glycine residuesin the �- and �-subunits was tolerated, indicating that thelength and/or flexibility of the linker in these subunits is notcritical for GABA-mediated current responses. In the crystalstructures of the nAChR �-subunit extracellular domain,AChBP, GLIC, ELIC, and the invertebrate glutamate-activatedchloride channel (2, 3, 5–8), the linker region is relatively

TABLE 3Summary of flurazepam concentration response dataData are mean � S.E. for n experiments from at least two different batches of oocytes. Mutant/WT (mut/WT) FZ EC50 ratios and maximal potentiation values{((IGABA � FZ)/IGABA) � 1} were calculated. FZ experiments were done at EC4–8 GABA. *, ¶, values are significantly different fromWT, p � 0.05 (one-way ANOVA).

Receptor FZ (EC50, nM) Potentiation (P) nH mut/WT EC50 mut/WT (P) n

WT 468 � 94 2.56 � 0.16 1.6 � 0.1 1 1 12�Gly1 523 � 113 0.97 � 0.1¶ 1.2 � 0.1 1 0.4 6�Gly2 567 � 87 1.23 � 0.08¶ 1.8 � 0.2 1 0.5 4�Gly4 532 � 61 0.86 � 0.05¶ 1.2 � 0.2 1 0.3 4�Gly8 690 � 246 1.17 � 0.13¶ 1.7 � 0.2 1.5 0.5 4�Gly1 518 � 87 1.08 � 0.1¶ 1.5 � 0.1 1 0.4 4�Gly1 820 � 105 2.27 � 0.1 1.5 � 0.1 1.8 0.9 5�Gly2 1310 � 144* 2.53 � 0.14 1.2 � 0.1 2.8 1 4�Gly4 2190 � 284* 1.53 � 0.07¶ 1.4 � 0.1 4.7 0.6 4�Gly8 1540 � 191* 1.47 � 0.06¶ 1.3 � 0.1 3.3 0.6 3




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

unstructured and located on the surface of each subunit facingthe extracellular channel vestibule, which likely allows the gly-cine insertions to be accommodated without large perturba-tions in the overall folding and structure of the �-sandwichcores of the subunits. Unstructured loop regions may impartflexibility that is essential for protein function.If the linker in the �-subunit is involved in propagating

GABA-triggered ligand binding sitemovements to the channel,one would predict that mutations in this region would affectligand binding and channel gating. To evaluate whether theglycine insertions in the �-subunit altered the GABA bindingsite structure, we examined the ability of the competitive antag-onist SR-95531 to inhibit GABA-gated current.�Gly insertionssignificantly increased SR-95531 IC50 (Fig. 7B, Table 1), sug-gesting that the orthosteric binding pocket was altered and thatthe shifts in GABA EC50 observed with the glycine insertionsare in part due to a change in GABAmicroscopic binding affin-ity. The proximity of the�Gly insertions to loopA of theGABAbinding site (37) and the �10,000-fold changes in GABA EC50measured also suggest that, at least in part, the insertions alterGABA binding (25). In support of this idea, in the relatednAChR, deletion of residues in the �4-�5 linker in the �-sub-unit altered microscopic acetylcholine binding as well as inhib-ited channel gating (15).The �Gly insertions also affected channel gating. When the

currents elicited from maximum PB versus maximum GABAwere compared in the same oocyte, the GABA currents elicitedfrom�Gly1-containing receptors were�4–5-fold smaller thanthe PB currents (Fig. 3A), indicating that �Gly1 transformedGABA into a partial agonist and decreased GABA efficacy.Moreover, �Gly1 increased the 10–90% apparent current risetimes for maximum GABA (Fig. 3B), suggesting that �Gly1alters channel gating by decreasing the channel opening rate. IntheGABAAR, nAChR, and serotonin type-3 receptor, mutatingresidues in the �4-�5 linker near loop A increase unligandedchannel opening (15, 16, 37, 38), indicating that this regioninfluences channel gating. Moreover, agonist-mediated move-ments in the nAChR �-subunit �4-�5 linker have beenobserved (39). Thus, although a detailed understanding of theeffects of the glycine insertions on GABA binding and channelgating will require higher resolution kinetic studies, takentogether, the data provide strong evidence that the �-subunit�4-�5 linker is an important structural element involved intransducing GABA binding to channel gating.A surprising observation seen with the �Gly2–8 insertions

was that the competitive antagonist SR-95531 was convertedinto a weak partial agonist and elicited ionic currents in theabsence of GABA (Fig. 4). Because glycine does not have a�-carbon, it can adopt a large variety of conformations andimpart a regional flexibility that may lead to a widening of theGABA binding pocket. This widening may allow the relativelybulkier SR-95531 to promote local movements in the GABA

FIGURE 7. Summary of effects of glycine insertions on GABA (A), SR-95531(B), PB (C), and FZ (D and E) actions. Bars represent mean � S.D. ofmutant/WT (mut/WT) ratios of EC50, IC50, and FZ potentiation values for 1, 2, 4,

and 8 glycine insertions in the �-, �-, and �-subunits. Dashed lines represent aratio of 1 (no difference in mutant as compared with WT). For �Gly2, -4, and -8mutant receptors, errors for GABA EC50 and SR-95531 IC50 mutant/WT ratioscould not be calculated because GABA EC50 for these mutant receptors couldnot be precisely determined (see “Results”). *, values are significantly differentfrom WT, p � 0.05 (one-way ANOVA).




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

binding pocket that trigger and stabilize an open channel pro-tein conformation.Moreover, the binding site expansionwouldhinder GABA, a smaller molecule, from optimally establishingthe necessary structural contacts and triggering the localmove-ments required (i.e. binding site contraction) for receptor acti-vation and would thus convert it into a partial agonist. Agonistaffinity for theGABAbinding site is linearly correlated to ligandlength (40) consistentwith the idea that altering the binding sitesize would alter GABA-site ligand actions.In the scenario that GABAAR activation involves a 10° clock-

wise rotation of the inner �-sheet, as is suggested for nAChR(1), inserting glycine(s) in the �-subunit �4-�5 linker could actmuch like a shock absorber and dampen the strength of thesignal being transmitted to the inner �-strands. Alternatively,the glycine insertions may affect local positioning of criticalresidue(s) important for GABAAR activation. In GABAARhomology models, a potential salt bridge between residues�K102 (loop A), which is nearby the insertion point of the gly-cines, and�D56 in the�-subunit (loop 2) is observed.Mutationof either of these residues alters GABA-mediated currentresponses, suggesting that this salt bridge is important for link-ing the binding site to loop 2 in the gating interface (41). Con-sistent with this idea, recent work fromCadugan andAuerbach(16) has shown that �A96 (loop A) and �I49 (loop 2) in themouse nAChR, which are aligned to �K102 and �D56 of theGABAAR, respectively, are energetically coupled.GABAAR Activation by PB Has Different Structural Require-

ments than GABA—Previous studies have shown that thestructural elements underlying GABAAR activation by GABAand PB, which bind at different sites (25), are different (36, 42,43). The differential effects of the glycine insertions on GABAand PB activation provide additional support for this idea. Gly-cine insertions in both the �-subunit and the �-subunit linkersincreased PB EC50 (Fig. 7C, Table 2), whereas insertions in the�-subunit were well tolerated. Consistent with the �Gly inser-tions beingwell tolerated, photochemical cleavage of the�4-�5linker in the GABAAR �-subunit does not affect PB EC50 or itsmaximumcurrent amplitudes (36). Regardless of the number ofglycine insertions in either the�-subunits or the �-subunits, PBEC50 increased about 10-fold (Table 2), suggesting thatGABAAR activation by PB does not depend on length and/orflexibility of the�- or�-subunit linkers. The differences inmag-nitude of the effects of the glycine insertions on GABA and PBactivation as well as differences in the subunit specificity of theeffects indicate that the allosteric structural pathways linkingGABA binding and pentobarbital binding to GABAAR activa-tion are different.We also examined the effects of introducing 2glycine residues (Gly2) simultaneously in both the �-subunitand the �-subunit linkers (��Gly2�Gly2) on PB EC50 (data notshown). We hypothesized that if the �- and �-subunit linkerscontributed independently to the PB activation cascade, thenthe dual subunit insertions would have additive effects on PBEC50 and increase PB EC50 by �100-fold. The changes in PBEC50 for ��Gly2�Gly2 were not additive and increased onlyabout 17-fold (EC50 � 3994 � 269 �M, n � 5 versus 231 � 10�M, n � 3 for WT), suggesting that the mutations in theselinkers may affect a shared downstream element in the trans-duction pathway mediating PB activation of the receptor. Pre-

vious studies (44–46) have also identified the �-subunit as akey participant in PB activation.GABAAR Modulation by Flurazepam Has Different Struc-

tural Requirements than GABA—Although spatially distinct,the GABA and BZD binding sites allosterically communicatewith each other (47). Glycine insertions in each of the GABAARsubunits decreased BZD modulation of GABA responses (Fig.7, D and E), suggesting that allosteric coupling of BZD bindingto modulation of GABA responses involves multiple subunitsand transduction pathways. Insertions in the �-subunit linkerdecreased FZ maximal potentiation of IGABA without affectingFZ apparent affinity (EC50) (Fig. 7,D and E, Table 3), suggestingthat the �-subunit linker is critical for mediating BZD agonistefficacy. The �-subunit �4-�5 linker spans the �-subunit andphysically connects the BZD and GABA binding sites (Fig. 1, Aand B). We speculate that BZD-initiated movements in theBZDbinding site are propagated directly through the�-subunitlinker to the GABA binding site and that the �Gly insertionshinder this propagation. BZDmodulation of GABAARwas alsoaltered by �Gly insertions (Fig. 7, D and E), which increasedBZD EC50 as well as decreased BZD efficacy (Fig. 7, D and E,Table 3). The �Gly insertions are positioned near the �/� inter-face (a nonbinding site interface). Interestingly, mutations ofArg-43 at this interface alter BZD unbinding (48) consistentwith structural perturbations at this interface having long rangeeffects on BZD actions. �Gly1 also decreased BZD efficacy(Table 3, Fig. 7D). Some studies suggest that BZD agonistsincrease unliganded GABAAR channel opening (49–51).Because �Gly1 appears to stabilize a closed state of theGABAAR by slowing channel opening (Fig. 3B), one mightexpect that �Gly1 would decrease BZD efficacy. Overall, theglycine insertions in the �4-�5 linkers of each of the subunitsaltered BZD actions, suggesting that BZD agonist-inducedstructural changes are transmitted via multiple pathways.Linear free energy relationship analysis in the nAChR indi-

cates that the�4-�5 linkermoves early in the activation process(15) and likely transfers the energy imparted by the agonist toloop 2 of the coupling interface that then leads to receptor acti-vation. Thus, we envision that increasing the length and flexi-bility of this linker in the GABAAR �-subunit decreases GABAactivation presumably by accruing an energetic barrier, indicat-ing that the structural dynamics of this linker are optimized forefficient signal propagation within the native protein.

Acknowledgments—We thank Dr. Andrew J. Boileau for helpful dis-cussions and Dr. Lisa M. Sharkey for constructing �Gly2, -4, and -8mutant cDNAs and measuring GABA concentration responses for�Gly2�� and �Gly4�� receptors.We also thank Dr. Ken Satyshur forhelp with homology modeling and Say Thao for assistance with con-struction of Gly1 mutants in �-, �-, and �-subunits.

REFERENCES1. Unwin, N. (2005) Refined structure of the nicotinic acetylcholine receptor

at 4 Å resolution. J. Mol. Biol. 346, 967–9892. Dellisanti, C. D., Yao, Y., Stroud, J. C., Wang, Z. Z., and Chen, L. (2007)

Crystal structure of the extracellular domain of nAChR �1 bound to�-bungarotoxin at 1.94 Å resolution. Nat. Neurosci. 10, 953–962

3. Brejc, K., vanDijk,W. J., Klaassen, R. V., Schuurmans,M., vanDerOost, J.,




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Smit, A. B., and Sixma, T. K. (2001) Crystal structure of an ACh-bindingprotein reveals the ligand binding domain of nicotinic receptors. Nature411, 269–276

4. Hansen, S. B., Sulzenbacher, G., Huxford, T., Marchot, P., Taylor, P., andBourne, Y. (2005) Structures of Aplysia AChBP complexes with nicotinicagonists and antagonists reveal distinctive binding interfaces and confor-mations. EMBO J. 24, 3635–3646

5. Bocquet, N., Prado de Carvalho, L., Cartaud, J., Neyton, J., Le Poupon, C.,Taly, A., Grutter, T., Changeux, J. P., and Corringer, P. J. (2007) A pro-karyotic proton-gated ion channel from the nicotinic acetylcholine recep-tor family. Nature 445, 116–119

6. Hilf, R. J., and Dutzler, R. (2008) X-ray structure of a prokaryotic pentam-eric ligand-gated ion channel. Nature 452, 375–379

7. Hilf, R. J., and Dutzler, R. (2009) Structure of a potentially open state of aproton-activated pentameric ligand-gated ion channel. Nature 457,115–118

8. Hibbs, R. E., and Gouaux, E. (2011) Principles of activation and perme-ation in an anion-selective Cys-loop receptor. Nature 474, 54–60

9. Akabas, M. H. (2004) GABAA receptor structure-function studies: a reex-amination in light of new acetylcholine receptor structures. Int. Rev. Neu-robiol. 62, 1–43

10. Purohit, P., Mitra, A., and Auerbach, A. (2007) A stepwise mechanism foracetylcholine receptor channel gating. Nature 446, 930–933

11. Miyazawa, A., Fujiyoshi, Y., and Unwin, N. (2003) Structure and gatingmechanism of the acetylcholine receptor pore. Nature 423, 949–955

12. Bocquet, N., Nury, H., Baaden, M., Le Poupon, C., Changeux, J. P., De-larue, M., and Corringer, P. J. (2009) X-ray structure of a pentamericligand-gated ion channel in an apparently open conformation. Nature457, 111–114

13. Miller, P. S., Topf, M., and Smart, T. G. (2008) Mapping a molecular linkbetween allosteric inhibition and activation of the glycine receptor. Nat.Struct. Mol. Biol. 15, 1084–1093

14. Szarecka, A., Xu, Y., and Tang, P. (2007) Dynamics of heteropentamericnicotinic acetylcholine receptor: implications of the gating mechanism.Proteins 68, 948–960

15. Chakrapani, S., Bailey, T. D., and Auerbach, A. (2003) The role of loop 5 inacetylcholine receptor channel gating. J. Gen. Physiol. 122, 521–539

16. Cadugan, D. J., and Auerbach, A. (2010) Linking the acetylcholine recep-tor channel agonist binding sites with the gate. Biophys. J. 99, 798–807

17. Robinson, C. R., and Sauer, R. T. (1998) Optimizing the stability of single-chain proteins by linker length and composition mutagenesis. Proc. Natl.Acad. Sci. U.S.A. 95, 5929–5934

18. Liman, E. R., Tytgat, J., and Hess, P. (1992) Subunit stoichiometry of amammalian K� channel determined by construction of multimericcDNAs. Neuron 9, 861–871

19. Boileau, A. J., Kucken, A. M., Evers, A. R., and Czajkowski, C. (1998)Molecular dissection of benzodiazepine binding and allosteric couplingusing chimeric �-aminobutyric acidA receptor subunits.Mol. Pharmacol.53, 295–303

20. Mercado, J., and Czajkowski, C. (2006) Charged residues in the �1 and �2pre-M1 regions involved in GABAA receptor activation. J. Neurosci. 26,2031–2040

21. Newell, J. G., and Czajkowski, C. (2003) The GABAA receptor �1-subunitPro-174–Asp-191 segment is involved inGABA binding and channel gat-ing. J. Biol. Chem. 278, 13166–13172

22. Holden, J. H., and Czajkowski, C. (2002) Different residues in the GABAA

receptor �1T60-�1K70 region mediate GABA and SR-95531 actions.J. Biol. Chem. 277, 18785–18792

23. Chou, T. (1974) Relationships between inhibition constants and fractionalinhibition in enzyme-catalyzed reactions with different numbers of reac-tants, different reactionmechanisms, and different types andmechanismsof inhibition.Molecular Pharmacology 10, 235–247

24. Cheng, Y., and Prusoff, W. H. (1973) Relationship between the inhibitionconstant (K1) and the concentration of inhibitor which causes 50% inhi-bition (I50) of an enzymatic reaction. Biochem. Pharmacol. 22, 3099–3108

25. Amin, J., andWeiss, D. S. (1993) GABAA receptor needs two homologousdomains of the �-subunit for activation by GABA but not by pentobarbi-tal. Nature 366, 565–569

26. Jackson, M. B., Lecar, H., Mathers, D. A., and Barker, J. L. (1982) Singlechannel currents activated by �-aminobutyric acid, muscimol, and (�)-pentobarbital in cultured mouse spinal neurons. J. Neurosci. 2, 889–894

27. Akk, G., and Steinbach, J. H. (2000) Activation and block of recombinantGABAA receptors by pentobarbitone: a single-channel study. Br. J. Phar-macol. 130, 249–258

28. Rosen, A., Bali, M., Horenstein, J., and Akabas, M. H. (2007) Channelopening by anesthetics and GABA induces similar changes in the GABAA

receptor M2 segment. Biophys. J. 92, 3130–313929. Buhr, A., and Sigel, E. (1997) A pointmutation in the �2-subunit of �-ami-

nobutyric acid type A receptors results in altered benzodiazepine bindingsite specificity. Proc. Natl. Acad. Sci. U.S.A. 94, 8824–8829

30. Duncalfe, L. L., Carpenter, M. R., Smillie, L. B., Martin, I. L., and Dunn,S.M. (1996) Themajor site of photoaffinity labeling of the �-aminobutyricacid type A receptor by [3H]flunitrazepam is histidine 102 of the �-sub-unit. J. Biol. Chem. 271, 9209–9214

31. Mihic, S. J., Whiting, P. J., Klein, R. L., Wafford, K. A., and Harris, R. A.(1994) A single amino acid of the human �-aminobutyric acid type Areceptor �2-subunit determines benzodiazepine efficacy. J. Biol. Chem.269, 32768–32773

32. Walters, R. J., Hadley, S. H., Morris, K. D., and Amin, J. (2000) Benzodiaz-epines act on GABAA receptors via two distinct and separable mecha-nisms. Nat. Neurosci. 3, 1274–1281

33. Gao, F., Bren,N., Burghardt, T. P., Hansen, S., Henchman, R.H., Taylor, P.,McCammon, J. A., and Sine, S. M. (2005) Agonist-mediated conforma-tional changes in acetylcholine-binding protein revealed by simulationand intrinsic tryptophan fluorescence. J. Biol. Chem. 280, 8443–8451

34. Henchman, R. H., Wang, H. L., Sine, S. M., Taylor, P., and McCammon,J. A. (2003)Asymmetric structuralmotions of the homomeric�7 nicotinicreceptor ligand binding domain revealed by molecular dynamics simula-tion. Biophys. J. 85, 3007–3018

35. Henchman, R. H., Wang, H. L., Sine, S. M., Taylor, P., and McCammon,J. A. (2005) Ligand-induced conformational change in the �7 nicotinicreceptor ligand binding domain. Biophys. J. 88, 2564–2576

36. Hanek, A. P., Lester, H. A., and Dougherty, D. A. (2010) Photochemicalproteolysis of an unstructured linker of theGABAAR extracellular domainprevents GABA but not pentobarbital activation. Mol. Pharmacol. 78,29–35

37. Boileau, A. J., Newell, J. G., and Czajkowski, C. (2002) GABAA receptor �2Tyr-97 and Leu-99 line the GABA binding site: insights into mechanismsof agonist and antagonist actions. J. Biol. Chem. 277, 2931–2937

38. Price, K. L., Bower, K. S., Thompson, A. J., Lester, H. A., Dougherty, D. A.,and Lummis, S. C. (2008) A hydrogen bond in loop A is critical for thebinding and function of the 5-HT3 receptor. Biochemistry 47, 6370–6377

39. Mourot, A., Bamberg, E., andRettinger, J. (2008)Agonist- and competitiveantagonist-inducedmovement of loop 5 on the �-subunit of the neuronal�4�4 nicotinic acetylcholine receptor. J. Neurochem. 105, 413–424

40. Jones, M. V., Sahara, Y., Dzubay, J. A., andWestbrook, G. L. (1998) Defin-ing affinity with the GABAA receptor. J. Neurosci. 18, 8590–8604

41. Venkatachalan, S. P., Thao, S., and Czajkowski, C. (2008) An intrasubunitsalt bridge linking loop A and loop 2 is critical for GABAA receptor acti-vation. FASEB J. 22 (Meeting Abstract Supplement, SanDiego, CA, 2008),1127.13

42. Mercado, J., and Czajkowski, C. (2008) �-Aminobutyric acid (GABA) andpentobarbital induce different conformational rearrangements in theGABAA receptor �1 and �2 pre-M1 regions. J. Biol. Chem. 283,15250–15257

43. Sancar, F., and Czajkowski, C. (2011) Allosteric modulators induce dis-tinct movements at the GABA binding site interface of the GABAA recep-tor. Neuropharmacology 60, 520–528

44. Serafini, R., Bracamontes, J., and Steinbach, J. H. (2000) Structural do-mains of the human GABAA receptor 3 subunit involved in the actions ofpentobarbital. J. Physiol. 524, 649–676

45. Pistis, M., Belelli, D., McGurk, K., Peters, J. A., and Lambert, J. J. (1999)Complementary regulation of anesthetic activation of human (�6�3�2L)and Drosophila (RDL) GABA receptors by a single amino acid residue.J. Physiol. 515, 3–18

46. Amin, J. (1999) A single hydrophobic residue confers barbiturate sen-




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

sitivity to �-aminobutyric acid type C receptor. Mol. Pharmacol. 55,411–423

47. Sharkey, L. M., and Czajkowski, C. (2008) Individually monitoring ligand-induced changes in the structure of the GABAA receptor at benzodiaz-epine binding site and nonbinding site interfaces. Mol. Pharmacol. 74,203–212

48. Goldschen-Ohm,M. P.,Wagner, D. A., Petrou, S., and Jones, M. V. (2010)An epilepsy-related region in the GABAA receptormediates long distanceeffects on GABA and benzodiazepine binding sites. Mol. Pharmacol. 77,35–45

49. Campo-Soria, C., Chang, Y., andWeiss, D. S. (2006)Mechanism of action

of benzodiazepines on GABAA receptors. Br. J. Pharmacol. 148, 984–99050. Downing, S. S., Lee, Y. T., Farb, D. H., and Gibbs, T. T. (2005) Benzodiaz-

epine modulation of partial agonist efficacy and spontaneously activeGABAA receptors supports an allosteric model of modulation. Br. J. Phar-macol. 145, 894–906

51. Rüsch, D., and Forman, S. A. (2005) Classic benzodiazepinesmodulate theopen-close equilibrium in �1�2�2L �-aminobutyric acid type A recep-tors. Anesthesiology 102, 783–792

52. Kloda, J. H., and Czajkowski, C. (2007) Agonist-, antagonist-, and benzo-diazepine-induced structural changes in the �1Met-113–Leu-132 regionof the GABAA receptor.Mol. Pharmacol. 71, 483–493




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Srinivasan P. Venkatachalan and Cynthia CzajkowskiModulation

-Sheet Governs Channel Activation and Allosteric DrugβBinding Site and Inner ) Receptor AgonistA-Aminobutyric Acid Type A (GABAγStructural Link between

doi: 10.1074/jbc.M111.316836 originally published online January 4, 20122012, 287:6714-6724.J. Biol. Chem.

10.1074/jbc.M111.316836Access the most updated version of this article at doi:

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here

http://www.jbc.org/content/287/9/6714.full.html#ref-list-1

This article cites 51 references, 21 of which can be accessed free at


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/lookup/doi/10.1074/jbc.M111.316836

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;287/9/6714&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/287/9/6714

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=287/9/6714&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/287/9/6714

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/287/9/6714.full.html#ref-list-1

http://www.jbc.org/

Documents

Structural Link between -Aminobutyric Acid Type A (GABAA