sign in sign up
<<
Home , GABAA receptor, Loreclezole

Proc. Nati. Acad. Sci. USA Vol. 91, pp. 4569-4573, May 1994 Neurobiology The modulatory action of at the y-aminobutyric acid type A receptor is determined by a single amino acid in the P2 and f3 subunit PETER B. WINGROVE, KEITH A. WAFFORD, CORRINA BAIN, AND PAUL J. WHITING* Merck Sharp and Dohme Research Laboratories, Terlings Park, Eastwick Road, Harlow, Essex, CM20 2QR United Kingdom Communicated by L. L. Iversen, February 7, 1994 (receivedfor review December 24, 1993)

ABSTRACT Type A yaminobutyric acid (GABAA) recep- structural determinants that contribute to the other modula- tors of the mammalian nervous system are a family of - tory sites of the GABAA receptor are currently unclear. gated channels probably formed from the coassembly of Loreclezole, {(Z)-1-[2-chloro-2-(2,4-dichlorophenyl) ethe- different subunits (a1.6, Pl39 71-3, 8) in the arrangement apy nyl]-1,2,4-triazole}, is a broad-spectrum com- or ad8. The activation of these receptors by GABA can be pound that was suggested to act via the GABAA receptor (14, modulated by a range ofcompounds acting at distinct allosteric 15). Until recently the precise mode of action of loreclezole sites. One such compound is the broad-spectrum anticonvul- was unclear. Using recombinant human GABAA receptors sant loreclezole, which we have recently shown to act via a expressed in Xenopus oocytes and transfected cells, we have specific modulatory site on the P subunit of the GABAA recently shown that loreclezole potentiates the action of receptor. The action of loreclezole depends on the type of 13 GABA by acting at a specific modulatory site on the ( subunit subunit in (16). Further, it was found that loreclezole is subunit selec- present the receptor complex; receptors containing tive, having a >300-fold higher affinity for receptors con- P2 or 13 subunits have >300-fold higher affinity for loreclezole taining P2 or (3 subunits over those containing (3 subunits. than receptors containing a PI subunit. We have used this In this report we describe the results of experiments done property to identify the amino acid residue in the Psubunit that to identify the structural determinants responsible for the determines the subunit selectivity of loreclezole. Chimeric subunit-selective action of loreclezole. By determining the P1/12 human GABAA receptor subunits were constructed and modulation by loreclezole of recombinant GABAA receptors coexpressed in Xenopus oocytes with human a, and y sub- containing a series of (-subunit chimeras and point mutants, units. The chimera 131/P2Lys237-Gly334 conferred sensitivity we have been able to demonstrate that a single amino acid, to 1 IAM loreclezole. Within this region there are four amino Asn-289 of the (32 subunit (Asn-290 of the (33 subunit) confers acids that are conserved in P2 and 13 but differ in 1. By sensitivity to this modulatory compound. mutating single amino acids of the P1 subunit to the P2/13 equivalent, only the P1 mutation of Ser-290 -* Asn conferred potentiation by loreclezole. Similarly, mutation of the homol- MATERIALS AND METHODS ogous residue in the 12 and P3 subunits to the Pi equivalent cDNA Clones. cDNAs encoding human GABAA receptor (Asn -- Ser) resulted in loss of sensitivity to loreclezole. The al, pi A(32,3,, and y2, subunits have been described (11-13, affinity for GABA and the potentiation by were 17, 18). unchanged in receptors coning the mutated 13 subunits. Construction of Chimeric 13 Subunits. The human ( subunit Thus, a single amino acid, 12 Asn-289 (P1 Asn-290), located at nucleotide sequences do not contain any conserved restric- the carboxyl-terminal end of the putative channel-lining do- tion enzyme sites that would be useful for constructing main TM2, confers sensitivity to the modulatory effects of chimeras. As such, a PCR approach was used to introduce loreclezole. into the A sequence restriction enzyme sites already present in the 2 sequence, allowing the subsequent construction of The mammalian type A )-aminobutyric acid (GABAA) re- chimeric (3i/P2 subunits. Each chimera was constructed in ceptor gene family is now known to consist of a number of the eukaryotic expression vector pCDM8 (Invitrogen), using subunit polypeptides (al-a6, (31-P3,3v-y3. 6, pi-p2) (1, 2). strategies described below. PCR was performed with Taq There is an increased body of evidence that that in polymerase (Perkin-Elmer/Cetus) as described (19), using suggests human (31 and (2 subunit cDNAs as template. Oligonucleo- these subunits coassemble in the arrangement a(yor tide primers were prepared by using an Applied Biosystems a(3S(3, 4), probably as pentamers (by analogy with the related 380B synthesizer and purified by using oligonucleotide pu- nicotinic receptor), to give a family ofreceptor subtypes that rification cartridges, according to the manufacturer's instruc- have different spatial and temporal patterns ofexpression (5). tions. PCR products were extracted with /, The activity of GABAA receptors can be allosterically -precipitated, digested with the appropriate restric- modulated by a number of agents, including ethanol and tion enzyme, purified by agarose-gel electrophoresis, and , and the clinically important and ligated as required. Constructs were verified by DNA se- (BZs) (6, 7). The BZ pharmacology of quencing. The construction ofeach ( subunit chimera was as GABAA receptor subtypes has been studied in some detail. follows: Only receptors made up of an a, 3, and 'y subunit exhibit Chimera (32A1.1: A (1 PCRproduct was generated by using high-affinity BZ binding (8), and the pharmacology is deter- the oligonucleotide primers 5'-CTAAGTTTTCGTCTTAA- mined by the type of a (9-11) and y (12), but not the (, (13), GAGAAACATTGG-3' (sense, bp 734-762 of (3 cDNA, subunit present, suggesting that this modulatory site is con- introducing an Afl II site) and 5'-TTGATACCATGGCGG- stituted by determinants on both the a and y subunits. The GATCCAGACAT-3' (antisense primer located in the

The publication costs ofthis article were defrayed in part by page charge Abbreviations: GABAA, ytaminobutyric acid type A; BZ, benzodi- payment. This article must therefore be hereby marked "advertisement" azepine. in accordance with 18 U.S.C. §1734 solely to indicate this fact. *To whom reprint requests should be addressed. 4569 Downloaded by guest on September 30, 2021 4570 Neurobiology: Wingrove et al. Proc. Natl. Acad Sci. USA 91 (1994) pCDM8 vector, bp 3348-3371). This was digested with Afl II the response was observed, normally 30 sec or less. At least and Not I (site in polylinker of pCDM8) and ligated to 3 min was allowed between each GABA application to similarly cut P2 cDNA to generate P32A1.1. prevent desensitization. Concentration-response curves Chimera 132A1.2: A Pf PGR product was generated with were calculated by using a nonlinear square-fitting program primers 5 '-CCAATGTTTCTCTTAAGACGAAAACT- to the equation fJx) = BmAx/[1+(ECso/x)n], where x is the TAG-3' (antisense, bp 734-762 of(3 cDNA introducing anAft drug concentration, EC50 is the concentration ofdrug eliciting II site) and 5'-AGTCCGAAAGAATCTGCTCCCTGCTT-3' a half-maximal response, and n is the Hill coefficient. Where (sense primer located in the pCDM8 vector, bp 1375-1400), the effect of a modulatory compound (loreclezole or fluni- digested with Afl II and HindIII (site in polylinker of trazepam) was being determined, an EC20 concentration of pCDM8), and ligated into similarly cut (2 cDNA to generate GABA was used (between 3 and 30 KLM), predetermined for (32A1.2. each individual oocyte. Chimera 2A1.3: The chimera P32A1.2 was digested with Loreclezole was a gift from Janssen, and all other com- Xho I (bp 1330 of (32 cDNA) and Xba I (site in polylinker), pounds were obtained from Sigma or Research Biochemicals removing cDNA-encoding C-terminal amino acids 373-end. (Natick, MA). This was ligated with the corresponding ( sequence gener- ated by PCR using primers 5'-CTCAGCACCCTCGAGATC- CGGAATGA-3' (sense, bp 1036-1061 of (3i cDNA, contain- RESULTS ing an Xho I site) and the antisense primer present in pCDM8 Structural Determinants Necessary for the Modulation of (bp 3348-3371) described above, also digested withXho I and GABAA Receptors by Loreclezole Are Located Between Lys- Xba I (site in polylinker of pCDM8) to create 132A1.3. 237 and Gly-334 of the P2 Subunit. We have previously Chimera (32A1.4: A HindIII (bp 929) Apa I (bp 1221) ( demonstrated that the modulatory action of loreclezole is at cDNA fragment was subcloned into pBluescript (Strata- least 300-fold selective for GABAA receptors containing a 12 gene). A Pi PCR product was generated by using oligonu- or (3 subunit over those containing a P( subunit (ref. 16 and cleotides 5'-CTTTGGGAAAGGGCCCCAGAAAAAGG-3' Fig. 1). This property was used to delineate the region of the (sense, bp 1024-1049, incorporating an Apa I site) and P2 subunit that confers sensitivity to loreclezole. A series of 5'-ACCAATTGGAGGGCCCAATCTCTTACC-3' (bp P1/P2 chimeric subunits were constructed and coexpressed in 1548-1574, incorporating an Apa I site), digested with Apa I Xenopus oocytes with a, and )o cDNAs. Large GABA-gated and ligated into the Apa I-cut (2 pBluescript DNA. This currents (0.5-2 AA) were observed with all receptors con- construct was then digested with Afl II and Xho I and ligated taining either wild-type or mutated ( subunits, with approx- to similarly cut (32A1.3 to generate j32A1.4. imately the same affinities for GABA. The ability of 1 P.M Site-Directed Mutagenesis. Oligonucleotide-directed muta- loreclezole (the EC50 concentration for modulation ofajP2Ms genesis was performed essentially as has been described (20), receptors; see ref. 16) to potentiate submaximal GABA with the modification that mutagenic primers were designed responses (=EC2o, between 3 and 30 tLM GABA, determined so as to also contain a silent base change that would result in for each individual oocyte) at GABAA receptors containing the introduction or removal of a restriction site. This result chimeric (3 subunits was then determined (Fig. 1). The allowed rapid screening of mutants by restriction digest. All determinants required for potentiation by loreclezole were mutations were confirmed by DNA sequencing. Mutagenesis delineated to the sequence encompassed by Lys-237 and of (31 subunit cDNA was performed with the subunit cDNA Gly-334 of the (2 subunit (chimera (32A1.4). As a control, all in pBluescript (Stratagene); mutants were subcloned into receptor subunit combinations were examined for their po- pCDM8 for expression studies. Mutagenesis of (2 and (33 tentiation by the BZ flunitrazepam (Fig. 1). Application of 1 subunit cDNAs was performed by using the vector pCDNA p.M flunitrazepam (which has an EC50 of 29 nM; ref. 21) I Amp (Invitrogen). The oligonucleotides used to generate potentiated the GABA responses ofall subunit combinations, each (3 subunit mutant were as follows: for (, with Thr-255 confirming their expression and coassembly, although inter- -I*le, 5'-TGTAATCAGTATCGATGGCATGTAGG-3' (in- estingly the potentiation ofreceptors containing the chimeras corporating a Cla I site); for P( with Ser-290 -- Asn, 5'- (32A1.3 and (32A1.4 was somewhat reduced. CTCCCTGAGATGGGTGTTGATGGTTG-3' (removes Sau Asn-289 of the 2 Subunit (Asn-290 of the 3 Subunit) Is I site); for (3 with Ile-308 -- Met, 5'-CCATCAGATACAT- Required for Potentiation by Loreclezole. Site-directed muta- GTCAATCGCTTTG-3' (incorporating an Afl III site); for (2 genesis was used to precisely identify the determinants within with Asn-289 -+ Ser, 5'-ACAATCAGCACCCACCTCAGG- the sequence Lys-237-Gly-334 of the P2 subunit required for GAAAC-3' (incorporating a Sau I site); for 3 with Asn-290 potentiation by loreclezole. In Fig. 2 partial amino acid -- Ser, 5'-ACCATCAGCACCCACCTTAGGGAGAC-3' (in- sequences of the human PiB, P2, and (33 subunits have been corporating a Sau I site). aligned. Because receptors containing (2 and (3 subunits, but Xenopus Oocyte Expression. Oocytes were removed from not those containing (31, are potentiated by 1 p.M loreclezole, anesthetized toads and manually defolliculated; the follicular we identified as targets for site-directed mutagenesis amino cells were removed by mild collagenase treatment (type IA, acid residues that were conserved in 12 and P3 subunits but 0.5 mg/ml for 10 min). The oocyte nuclei were directly are not conserved in (31. There are four such residues: (31 injected by using the "blind" method with 10-20 nl of Thr-255, (1 Ser-290, P3i Ile-308, and (3, Lys-334 (Fig. 2). The injection buffer (88 mM NaCl/1 mM KCI/15 mM Hepes, pH latter was disregarded, as it is located within the putative 7.0; filtered through nitrocellulose) containing GABAA re- large cytoplasmic domain between TM3 and TM4. Site- ceptor subunit cDNAs (6 ng/pl) and used 24-96 hr later. directed mutagenesis was performed on the (3, subunit, Approximately 80% of injected oocytes had responses to individually changing each ofthe other three residues to their applied GABA. For electrophysiological recording, oocytes (2/P3 equivalent. When coexpressed with a, and -tr sub- were placed in a 50-pl bath and perfused at 4-6 ml/min with units, only one mutant, (3i Ser-290 -- Asn, conferred lore- modified Barth's medium [88 mM NaCl/1 mM KCI/10 mM clezole sensitivity (Fig. 3). To confirm the critical role ofthis Hepes/0.82 mM MgSO4/0.33 mM Ca(NO3)2/0.19 mM residue, two other mutants were constructed: P2 Asn-289 -- CaCl2/2.4 mM NaHCO3, pH 7.5]. Cells were impaled with Ser and (3 Asn-290 -- Ser. In both these mutants the two 1- to 3-MQ electrodes containing 2 M KCl and voltage- asparagine residue of the (32 and (3 subunits, thought to be clamped between -40 and -70 mV. Drugs were applied in critical for loreclezole potentiation, were replaced by serine the perfusate. GABA modulators were preapplied for 30 sec found in the homologous position of the (1 subunit. When before GABA addition, which was applied until the peak of coexpressed with a, and ys,, in both cases the mutation Downloaded by guest on September 30, 2021 Neurobiology: Wingrove et al. Proc. Natl. Acad. Sci. USA 91 (1994) 4571

a) su)subulnit chimeras b) Loreclezole

120 *: :. . FIG. 1. Modulation by lore- clezole and flunitrazepam of hu- (_ 60| man GABAA receptors containing chimeric 131/132 subunits. (a) Dia- grammatic representation of 31/B32 chimeras. f1 sequences are repre- sented by clear boxes, and 132 se- i -'2 F quences are represented by CZ c) Flunitrazepam shaded boxes. The hatched areas 160 represent the putative signal pep- a) tide and transmembrane domains 0.l.- A 8 (TM1-4). Potentiation of human a113yrs subunit combinations by loreclezole (b) and by flunitra-

.. zepam (c). The 13 subunit present I~~~~~~~li'-J:~~~~~ :11 , in each combination is indicated beneath each bar. Bars represent the percent potentiation of a sub-

I maximal GABA response (EC20, i:-...... _i __LI. ..t -fir- -TY T - of predetermined for each individual oocyte). Data are the mean -~? " SEM of 2 four oocytes. resulted in loss of loreclezole sensitivity, confirming the DISCUSSION importance of this asparagine residue. Typical responses In this study we investigated the structural determinants that demonstrating the effect of loreclezole on GABAA receptor confer subunit-selective modulation of GABAA receptor subunit combinations containing wild-type or mutant sub- P function by the specific anticonvulsant compound lore- units are shown in Fig. 4. For all subunit combinations, the clezole. A single amino acid residue, /32 Asn-289 ((3 Asn-290) affinities for GABA were not affected, and flunitrazepam was responsible for confering sensitivity to loreclezole: its could potentiate the response to GABA (Fig. 3b), demon- strating that the receptor subunits had been expressed and removal from g32 and 13 resulted in loss of sensitivity (Fig. 3), correctly assembled. whereas Ser -* Asn substitution in the homologous position of /1 resulted in the gain of loreclezole sensitivity (Figs. 3-5). Concentration-effect curves were generated to determine the EC50 for the potentiation by loreclezole of the response a) Loreclezole to GABA (Fig. 5). Receptors containing wild-type 12 or mutant /31Ser-290 -- Asn had essentially the same EC50 value 120- (1.2 + 0.4 gM and 2.0 ± 0.15 ,gM, respectively), Hill slope (0.7 0.11 and 0.9 ± 0.05, respectively), and maximum potentiation (214 ± 9% and 181 ± 29%, respectively). These < data demonstrate that Asn-289 of the (32 subunit (Asn-290 of 60Lo

the f33 subunit) is the critical determinant for potentiation of CD GABAA receptors by loreclezole.

0 TM1 TM2

31 238 -IF, NGI GYF I71QTYMPStL1TILSVT5/SF;ZJINYDASAARlAL-,GI TTVLTMT C: b) Flunitrazepam 0 6- 160- 32 237 VIY LQT-MPSiLIT:L2iF'TNYDIk Al li 1SMT CZ

0.4 33 238 rLKRIJTGYFIIQTYMPSiLITILSWvSFWINYDASAARVALGITTVLTMT CD 0' 0_ TM3 80) 288 TIIsI'HLRETLPKIPYVKAIDiYLMGCFVJFViF' ALLEYA fillY YIFFGkG

287 TInFHLRETLPKIPYVKAIDmYLMGCFlVFV'F'mALLEYA7 VNY] FEGrG

0~~~~~~~~ 288 TILLnHLRE1TLPKI PYVKAIDmYLMGCFVF'' 1ALI1FYAfVNYLL GrG ?15 1Q 9

FIG. 2. Identification of the amino acid residue in the 13 subunit required for potentiation of GABAA receptor function by the anti- convulsant loreclezole; alignment of the predicted amino acid se- FIG. 3. Modulation by loreclezole and flunitrazepam of human quences of the human GABAA receptor Pi, 132, and 13 subunit GABAA receptor combinations containing mutated subunits. Po- sequences. Only the region of the subunit sequences corresponding tentiation of human GABAA receptor subunit combinations by to Lys-237-Gly-334 of 132 are shown. Numbering indicates the amino loreclezole (a) and flunitrazepam (b). The subunit present in each acid number, assigning the initiating methionine as 1. The putative subunit combination is indicated beneath each bar. Bars represent transmembrane domains are indicated by TM1-3. Residues in bold- the percent potentiation of a submaximal GABA response (EC20, face type are those that are conserved only in 232 and 133. The boxed predetermined for each individual oocyte). Data are the mean + residue indicates the position of 132 Asn-289 and 133 Asn-290. SEM of - four oocytes. Downloaded by guest on September 30, 2021 4572 Neurobiology: Wingrove et al. Proc. Natl. Acad. Sci. USA 91 (1994)

6pM GABA +1 pM Loreclezole 6pM GABA

p1

50 nA

2 min

6pM GABA +1pM Loreclezole 6pM GABA

50 nA

2 min

1 OpM GABA +1pM Loreclezole 10pM GABA

P1Ser29OAsn

250 nA

2 min

2OpM GABA +1 pM Loreclezole 2OpM GABA

P2Asn289Ser

50 nA

2 min

FIG. 4. Effect of 1 pM loreclezole on GABA-induced currents inXenopus oocytes expressing human GABAA receptors composed of alpi-3, ai(32y2, al(3(Ser29OAsn)V2, and alP2(Asn289Ser)y2. Loreclezole was applied for 30 sec before coapplication of GABA (an EC20 concentration, predetermined for each oocyte) as indicated by the bars. Recovery was observed after a 3-min wash period.

Our data demonstrate that a single-amino acid residue tunately there is no radioligand for the loreclezole-binding determines the sensitivity of GABAA receptors to lore- site; such a radioligand would be a very useful toolfor further clezole. Further investigation is required to definitively de- characterization of the mechanism of interaction of lore- termine whether this asparagine residue is required for bind- clezole with the GABAA receptor. ing ofloreclezole (i.e., located at the binding site) or whether It is only the P2 and (3 subunits that have an asparagine it is required for the transduction of the allosteric changes residue at position 289/290. The (,3 subunit (17) and all a (11, that result from the binding of loreclezole to the receptor. It 23) and y (24) subunits have a serine, and the 8 subunit has is possible to speculate that the molecular interaction is via a methionine (25). Considering other members of the ligand- a hydrogen bond(s) between the triazole moiety ofloreclezole gated gene superfamily, the al and the amide group of asparagine. However, a single hy- subunit has a serine at the homologous position (26), whereas drogen bond contributes 0.5-1.8 kcal mol-1 of binding en- the (31 subunit has a cysteine (27). The cation-gating 5-hy- ergy, equivalent to perhaps a factor of2-20 in the dissociation droxytryptamine-type 3 receptor (28) and Torpedo nicotinic

constant (22). Because the change in EC50 for loreclezole at acetylcholine receptor a, (3, y, 8 subunits (see ref. 29) all have al,81y2 versus ai(31(Ser290Asn)y2 is at least 300-fold (Fig. 5), leucine at the homologous position. Thus, at least for the then other molecular interactions would be required. Unfor- GABAA and glycine receptor subunits, there appears to be Downloaded by guest on September 30, 2021 Neurobiology: Wingrove et al. Proc. Natl. Acad. Sci. USA 91 (1994) 4573 2. Wisden, W. & Seeburg, P. H. (1992) Curr. Opin. Neurobiol. 2, 263-369. 3. Duggan, M. J., Pollard, S. & Stephenson, F. A. (1992) J. Neurochem. 58, 72-76. -~200- 4. Quirk, K., Gillard, N. P., Ragan, C. I., Whiting, P. J. &

0 McKernan, R. M. (1994) J. Biol. Chem., in press. 5. Laurie, D. J., Wisden, W. & Seeburg, P. H. (1992)J. Neurosci. 0 150 12, 4151-4172. 0 6. Olsen, R. W. & Venter, J. C., eds. (1986) / GABA-A Receptors and Channels: Structural and 100 Functional Properties (Liss, New York). C 0 7. Doble, A. & Martin, I. L. (1992) Trends Pharmacol. Sci. 13, CL 76-81. 50- 8. Pritchett, D. B., Sontheimer, B. D., Shivers, B. D., Ymer, S., Kettenmann, P. R., Scofield, P. R. & Seeburg, P. H. (1989) Nature (London) 338, 582-585. 9. Pritchett, D. B., Luddens, H. & Seeburg, P. H. (1989) Science 0.0001 0.001 0.01 0.1 1 10 100 245, 1389-1392. 10. McKernan, R. M., Quirk, K., Prince, R., Cox, P. A., Gillard, [Loreclezole] (pM) N. P., Ragan, C. I. & Whiting, P. J. (1991) 9, 667-676. 11. Hadingham, K. L., Wingrove, P., Le Bourdelles, B., Palmer, FIG. 5. Concentration-effect curves for the potentiation by lore- K. J., Ragan, C. I. & Whiting, P. J. (1993) Mol. Pharmacol. 43, clezole of human GABAA receptors expressed in Xenopus oocytes. 970-975. 12. K. P. 0, ai(nt2; *, ai02n; o, a1131(Ser289Asn)y2. Data for the first two Wafford, A., Bain, C. J., Whiting, J. & Kemp, J. K. combinations are taken from Wafford et al. (16) and is shown for (1993) Mol. Pharmacol. 44, 437-442. comparison. An EC20 concentration of GABA was used, predeter- 13. Hadingham, K. L., Wingrove, P. B., Wafford, K. A., Bain, C., mined for each individual oocyte. Data were fitted as described in the Kemp, J. A., Palmer, K. J., Wilson, A. W., Wilcox, A. S., methods. Sikela, J. M., Ragan, C. I. & Whiting, P. J. (1993) Mol. Phar- macol. 44, 1211-1218. 14. Waquier, A., Fransen, J., Melis, W., Ashton, D., Gillardin, some flexibility regarding the amino acid found in this posi- J. M. & Lewi, P. J. (1990) Drug Dev. Res. 19, 375-392. tion. 15. Ashton, D., Fransen, J. F., Heeres, J., Clinke, G. H. C. & Perhaps the most intriguing observation is the location of Janssen, P. A. J. (1992) Epilepsy Res. 11, 27-36. this asparagine residue is its location in the TM2 helical 16. Wafford, K. A., Bain, C. J., Quirk, K., McKernan, R. M., domain, now generally agreed to line the pore of the ion Wingrove, P. B., Whiting, P. J. & Kemp, J. A. (1994) Neuron, the in press. channel. The precise location of this residue is toward 17. Schofield, P. R., Pritchett, D. B., Sontheimer, H., Ketten- carboxyl-terminal end of TM2, i.e., on the extracellular side, mann, H. & Seeburg, P. H. (1989) FEBS Lett. 244, 361-364. six residues away from the conserved leucine residue (P2 18. Wagstaff, J., Chaillet, J. R. & Lalande, M. (1991) Genomics 11, Leu-283), which in Torpedo nicotinic acetylcholine receptor 1071-1078. (Leu-251) is thought to be at the narrowest point ofthe channel 19. Whiting, P. J., McKernan, R. M. & Iversen, L. L. (1990) Proc. many in TM2 have been mutated and Natl. Acad. Sci. USA 87, 9966-9970. (30). Although residues 20. Wafford, K. A. & Whiting, P. J. (1992) FEBS Lett. 313, 113- changes in channel function have been demonstrated (31), the 117. residue at the position homologous to 32 Asn-289 does not 21. Wafford, K. A., Whiting, P. J. & Kemp, J. A. (1992) Mol. appear to have been subjected to site-directed mutagenesis. Pharmacol. 43, 240-244. Site-directed mutagenesis has also been used to begin to 22. Fersht, A. R. (1987) Trends Biochem. Sci. 12, 301-304. identify the structural determinants that contribute to the 23. Seeburg, P. H., Wisden, W., Verdoorn, T. A., Pritchett, D. B., Werner, P., Herb, A., Luddens, H., Sprengel, R. & BZ-binding site. Studies have concentrated on the a subunit. Sakmann, B. (1990) Cold Spring Harbor Symp. Quant. Biol. 55, A single amino acid substitution (al Gly-228 -+ Glu) increases 29-40. the affinity of several selective BZ compounds without affect- 24. Herb, A., Wisden, W., Luddens, H., Puia, G., Vicini, S. & ing the affinity of nonselective compounds (32). Similarly, Seeburg, P. H. (1992) Proc. Natl. Acad. Sci. USA 89, 1433- conversion of a6 Arg-119 -> His conferred high-affinity bind- 1437. ing ofdiazepam to receptors (33). It is also clear that the 25. Shivers, B. H., Killisch, I., Sprengel, R., Sontheimer, H., a6p32y Kohler, M., Schofield, P. R. & Seeburg, P. H. (1989) Neuron y subunit contributes to the BZ-binding site and influences the 3, 327-337. pharmacology (12, 34). Indeed, some BZ compounds will bind 26. Grenningloh, G., Schmeiden, V., Schofield, P. R., Seeburg, to recombinant GABAA receptors that lack an a subunit (ref. P. H., Siddique, T., Mohandas, T. K., Becker, C.-M. & Betz, 35 and K.A.W., unpublished observations). It is thus likely H. (1990) EMBO J. 9, 771-776. that the BZ-binding site will be made up ofcontributions from 27. Grenningloh, G., Pribila, I., Prior, P., Multhaup, G., Bey- both the a and 'y subunits. Together, these data suggest that reuther, K., Taleb, 0. & Betz, H. (1990) Neuron 4, 963-970. although a single amino acid residue can be a critical deter- 28. Maricq, A. V., Peterson, A. S., Brake, A. J., Myers, R. M. & minant for the binding ofcertain compounds to the BZ-binding Julius, D. (1991) Science 254, 432-437. 29. Noda, M., Takahashi, H., Tanabe, T., Toyosato, M., Kikyo- site, ligand-binding sites are complex structures made up of tani, S., Furutani, Y., Hirose, T., Takashima, H., Inayama, S., numerous determinants that are not necessarily closely lo- Miyata, T. & Numa, S. (1983) Nature (London) 302, 528-532. cated in the primary sequence. This is certainly the case for the 30. Unwin, N. (1993) J. Mol. Biol. 229, 1101-1124. acetylcholine-binding site ofthe nicotinic acetylcholine recep- 31. Changeux, J.-P., Galzi, J.-L., Devillers-Theiry, A. & Bertrand, tor (31). In this report we have demonstrated that a single D. (1992) Q. Rev. Biophys. 25, 395-432. amino acid residue determines to a very large degree the 32. Pritchett, D. B. & Seeburg, P. H. (1991) Proc. Natl. Acad. Sci. modulatory effect of loreclezole. Further studies will be re- USA 88, 1421-1425. 33. Weiland, H. A., Luddens, H. & Seeburg, P. H. (1992) J. Biol. quired to determine definitively if P2 Asn-289 is, indeed, Chem. 267, 1426-1429. located at the loreclezole-binding site and what, if any, con- 34. Stephenson, F. A., Duggan, M. J. & Pollard, S. (1990) J. Biol. tributions are made by other amino acid residues. Chem. 265, 21160-21165. 35. Im, H. K., Im, W. B., Hamilton, B. J., Carter, D. B. & Von- 1. Burt, D. R. & Kamatchi, G. L. (1991) FASEB J. 5, 2916-2923. voigtlander, P. F. (1993) Mol. Pharmacol. 44, 866-870. Downloaded by guest on September 30, 2021