Methyllycaconitine: a Selective Probe for Neuronal A-Bungarotoxin Binding Sites

CORE Metadata, citation and similar papers at core.ac.uk

Provided by Elsevier - Publisher Connector

Volume 270, number 1,2, 45-48 FEBS 08837 September 1990

Methyllycaconitine: a selective probe for neuronal a-bungarotoxin binding sites

J.M. Ward*, V.B. Cockcroft, G.G. Lunt, F.S. Smillie and S. Wonnacott

Department of Biochemistry, University of Bath, Bath BA2 7AY, UK

Received 20 June 1990

The ability of methylly~conitine (MLA) to inhibit the binding of ezsI~-bungarotoxin to rat brain membranes, frog and human muscle extracts and the human muscle cell line TE671 has been measured. MLA showed a markedly higher affinity for the rat brain site (Ki 1.4 x IO* M) than for the muscle receptors (4 1O-s-1O-6 M). Structure modelling techniques were used to fit the structure of MLA to a nicotinic pharmacophore model. MLA is the first low molecular weight ligand to be shown to discriminate between muscle nicotinic receptors and their a-bungarotoxin- binding counterpart in the brain, and as such may be a useful structural probe for pursuing the structural and functional properties of the neuronal protein.

Methylly~~onitine; Nicotinic receptor; Brain a-bungarotoxin binding site; Nicotinic pharmacoph&z

1. INTRODUCTION site. Thus the insect and rat brain [“‘I]~-bungarotoxin binding sites share a common sensitivity to MLA. In The alkaloid N-methyllycaconitine (MLA; Fig. 1) is view of the current uncertainty about the physiological the principal toxic component of the seeds of status of [‘251]cu-bungarotoxin binding sites in the mam- Delphinium brownii. Historically, the seeds from this malian CNS (see [6,7]), a pharmacological probe show- plant have been recognized for their insecticidal proper- ing selectivity for this site would be very useful. ty, which is believed to arise from the potent an- Although the rw-bungarotoxin binding component in tagonism of insect nicotinic a~etylcholine receptors vertebrate brain shows homology with nAChR at the (nAChR) by MLA [I]. MLA is a potent competitor of protein [S] and gene [7] level, there is little direct [‘251]cY-bungarotoxin binding to nAChR in flyheads [ 11, evidence that this protein has any role in synaptic locust ganglia [2] and cockroach nerve cord [3], with Ki transmission in the brain, despite its clear nicotinic values in the nanomolar range. Complete blockade of pharmacology in binding experiments. Exceptions in- nicotinic responses from a cockroach motorneurone clude the chick optic tectum [S] and cerebellar in- was observed at 10e6 M MLA; the EDSOfor this an- terneurones [9], where ~-bungarotoxin has been shown tagonism was about lo- 7 M [3], which is comparable to to antagonize nicotinic responses. In insects, however, the sensitivity to MLA of acetylcholine responses in the the neuronal or-bungarotoxin binding protein is well nematode Ascaris [4]. This potency contrasts with the documented as a functional nAChR [lo], The potency weaker antagonism by MLA of mammalian nAChR at of MLA for neuronal a-bungarotoxin sites, both mam- the rat neuromus~ul~ junction (EDSO2 x 10e6 M) f5]. malian and invertebrate, has prompted us to further In a previous study [2] we compared the potencies of assess its specificity by comparing its potencies in com- MLA in competition binding assays for locust and rat petition binding assays for nAChR in vertebrate muscle neuronal nAChR, and reported that the [1251]~- preparations and to consider the molecular basis for its bungarotoxin binding site in mammalian brain was 3 pharmacological specificity. orders of magnitude more sensitive to MLA than was the putative nAChR labelled by E3H]nicotine in the 2. MATERIALS AND METHODS same tissue. Apart from the snake a-toxins, this is the only example of a cholinergic compound that is more 2.1. MOrerials Na ‘*‘I was purchased from Amersham International (Aylesbury, potent at the a-bungarotoxin site than the t3H]nicotine UK). cY-Bungarotoxin was purchased from Sigma (Poole, UK) and iodinated to a specific activity of 700 Ci/mmol as pieviously described [l 11. Tissue culture reagents were obtained from Flow Laboratories, Correspondence address: S. Wonnacott, Department of Irvine, Ayrshire, UK or Gibco Ltd., Uxbridge Middlesex, UK. All Biochemistry, University of Bath, Bath BA2 7AY, UK sterile plastic ware was supplied by Nunc Gibco Ltd. TE671 cells were kindly provided by Professor J. Newsom-Davis (Institute for *Present address: Department of Anesthesiology, Harvard Medical Molecular Neuroscience, John Radcliffe Hospital, Oxford, UK) and School, Shriners Burns Institute, Boston, MA, USA were cultured in monolayers in 80 cm3 flasks, essentially as described

Published by Elsevier Science Publishers B. Y. (Biomedical Division) ~145793/~/$3.50 0 1990 Federation of European Biochemical Societies 45 Volume 270, number 1,2 FEBSLETTERS September 1990 ” 7 3. RESULTS AND DISCUSSION N Competition binding assays with MLA were carried out on rat brain membranes, frog and human muscle “\ extracts and the human rhabdomyosarcoma cell line Q%2 TE671 (Fig. 2). Inhibition constants (the concentration Cytisine of MLA that inhibits [“‘I]cr-bungarotoxin binding by 50%) are given in Table I. In agreement with our previous findings [2], MLA was a very potent inhibitor ‘\ 0 of [1251]a-bungarotoxin binding to rat brain membranes. However, it was more than 3 orders of 0 42 =o magnitude weaker in competing for the muscle nAChR N in each of the preparations tested. Notably, human -JL4 Acetylcholine \ 2 muscle extract, prepared from amputated calf muscle [ 151, and the TE67 1 cell line displayed similar Ki values (approximately 10e5 M) for MLA. The sensitivity of muscle nAChR to MLA is very comparable to that previously derived for Torpedo NaChR and brain t3H]nicotine binding sites [2] (see Table I). At 10m4 M, MLA had no muscarinic potency, indicated by its failure to inhibit [3H]quinuclidinyl benzilate binding to rat brain membranes (data not shown). The unique discrimination by MLA in favour of

Methyllycaconitine neuronal versus muscle a-bungarotoxin binding sites Fig. 1. Structures of MLA, acetylcholine and cystisine. Hydrogen raises the issue of the structural features that underlie atoms are not shown. Unlabelled atoms are carbon. Arrows indicate the recognition of this compound; the definition of the key atoms implicated in the pharmacophore model of nicotinic such features would aid our understanding of the cholinergic ligand binding [20]. The dashed box encloses the parent molecular nature of the nicotinic receptor recognition structure, lycotonine. The N-phenyl-succinimide side chain of MLA site. The parent structure, lycoctonine [l] (see Fig. l), is is on the left. without nicotinic potency; indeed it is closely related to aconitine which interacts with sodium channels at the by Syapin et al. (121 with modifications [13]. MLA (citrate salt) was site characterized by batrachotoxin [ 191. The introduc- synthesized by Professor M.H. Benn (Department of Chemistry, University of Calgary, Alberta, Canada). tion of an ester linkage in the formation of MLA pro- vides a carbonyl oxygen (Fig. 1, arrow l), that is likely 2.2. Competition binding assays to be a key moiety in conferring cholinergic nicotinic ac- Rat brain P2 membranes, frog muscle extract and human muscle extract were prepared and assayed for [‘*‘I]cY-bungarotoxin binding G 120 as previously described [2,14,15]. TE671 cells were harvested and replated in fresh growth medium in 24-well plates at a cell density of d 5 x lo5 cells/ml, essentially as previously described [12,13]. 5 100 Triplicate wells were incubated with serial dilutions of MLA and [“‘I]~-bungarotoxin (final concentration 2.0 nM) for 90 min at 37°C. % 80 Non-speficic binding was determined in the presence of 1.O mM D- ?5 tubocurarine. After incubation, cells were washed and extracted in 0.1 M NaOH for determination of radioactivity. 1 60 .E 2.3. Structure modelling 1 40 The programs Discover and Insight from Biosym Inc. were used to H construct and view structures on an Evans and Sutherland PS 300 pic- 3 20 ture system. Energy calculations were performed using the valence w ui force-field with the potential parameters of Dauber-Osguthorpe et al. N [16]. MLA was constructed using the X-ray structure of aconitine G 0 [17], and the stereochemistry of MLA as reported by Jennings et al. -12 -10 -8 -6 -4 I2 [l]. The S- and 6-membered rings of the N-phenylsuccinimide moiety log concsntmtlon MU (W) of MLA were set orthogonal, based on the X-ray structure of N-p- bromophenyl-succinimide. The rigid nicotinic agonist, cystisine, was Fig. 2. Competition curves for MLA in various nAChR preparations. used as a template in the establishment of a nicotinic cholinergic phar- Rat brain membranes (O-O), detergent extracts of frog (o---+) macophore. Site-points used to overlay MLA and acetylcholine onto and human (&---A) muscle, and culture dishes of TE671 cells cystisine were: the nitrogen atom of the ammonium group; the carbon (ti) were incubated with serial dilutions of MLA and 2 nM atom of the carbonyl group; and the oxygen atom of the carbonyl [‘251]cY-bungarotoxin. Non-specific binding was determined in the group. Acetylcholine and MLA were template-fitted onto cystisine us- presence of unlabelled a-bungarotoxin. Values are the means of at ing a forcing constant of 25 kcal/mol [18]. least 3 separate assays with SEM indicated by the vertical bars.

46 Volume 270, number 1,2 FEBS LETTERS September 1990

Table I (see Fig. 3) than are the neuronal nicotinic receptors Inhibition of radioligand binding to nAChR preparations by MLA that do not recognize this snake toxin. However, com- Preparation Ligand Ki (M) for MLA parison of the chick brain cr-bungarotoxin binding pro- Frog muscle extract [rz51]c+Bgt 1.0 f 0.2 x 10-5 tein sequences with that of an a-subunit cloned from Human muscle extract [‘Z5]cu-Bgt 7.8 f 2.0 x 10-6 locust ganglion [23] reveals considerably greater TE67 1 cells [1251]a-Bgt 6.3 f 1.4 x 1O-5 homology, especially in the rather variable extracellular Rat brain P2 membranes [L251](U-Bgt 1.4 f 0.2 x 10-9 domain. In particular, there is high homology in the se- Rat brain P2 membranes [‘HInicotine 3.7 r3.7 x lo-6* Torpedo purified nAChR [1251]~-Bgt 1.1 f 0.6 x 10-6* quences flanking the vicinal cysteines (Cys 192 and 193 Locust ganglion [‘251]ru-Bgt 1.8 x 10-8* Torpedo numbering) that are implicated in the nicotinic Competition assays weer performed using 2 nM [rz51]a-Bgt. 1C50 ligand binding site [24]. Whereas the oxygen-rich bulk values were derived from linear transformations of dose-response of the lycoctonine portion of MLA (Fig. 1) might be ex- curves; Ki values were derived from ICso values (261, assuming & pected to face into the solvent around the binding site, values for [1251]cy-Bgtof 1.8 nM (frog muscle [14]), 0.5 nM (human the N-phenylsuccinimide side chain is topologically muscle [15]), 2.0 nM (TE671 cells; Ward, unpublished observations), equivalent to the reactive group in ligands that label Cys 1.5 nM (brain [2]). Values are the means f SE for at least 3 independent determinations. *Data from (21 for comparison. 192 and 193 [24]. Thus this side chain may be accom- modated in this region of the protein, the amino acid sequence of which will influence the relative binding af- finities for MLA. The homology between chick brain CY- tivity on this structure. This electronegative centre is a bungarotoxin binding proteins and locust nAChR is feature of nicotinic ligands, and is considered to be in- likely to underlie the high affinity binding of MLA volved in hydrogen bonding with key residues at the shared by these proteins (Table I). recognition site [20]. The possible relationships in receptor structure that In addition, this pharmacophore model [20] of are reflected in the pharmacological similarities are nicotinic ligands includes an electrostatic interaction outlined in Fig. 3. It should be noted however that the between the quaternary nitrogen of acetylcholine and pharmacological discrimination is not absolute; both the receptor binding site. Using the rigid agonist muscle nAChR and the high affinity nicotine binding cytisine as a definitive template, acetylcholine and site in brain show a low affinity for MLA (Table I), and MLA were superimposed to generate the conformations all the nicotinic proteins, by definition, recognize illustrated in Fig. 1 (see Materials and Methods). Thus nicotine to some extent. Indeed we have recently shown we can identify the nitrogen atom in MLA for similar that micromolar concentrations of MLA antagonize electrostatic interaction (Fig. 1; arrow 2). functional responses of neuronal nAChR that are insen- How can we account for the marked preference sitive to a-bungarotoxin [25]. shown by MLA for neuronal a-bungarotoxin binding We suggest that MLA is a potentially useful probe in sites compared with other members of the nicotinic the dissection of subtypes of nAChRs. Synthesis of receptor family of proteins? The precise spatial rela- analogues of MLA may well provide more definitive tionship between the key residues contributing to the answers to some of the proposals made here and may receptor binding site [21,22] may vary among receptor help in defining more precisely the site(s) of ligand subtypes and hence could influence the affinity of bin- recognition in this important family or neuronal recep- ding of MLA. Sequence analysis of the two (Y- tor proteins. bungarotoxin binding proteins recently cloned from chick brain [7] shows them to have lower homology Acknowledgements: This study was supported financially by R.J. with other avian nicotinic receptor subunits (less than Reynolds Tobacco Co. (SW. and G.G.L.), MRC Project Grant G8722675N (SW.) and a collaborative SERC project grant with Shell 50% in every case [7]) than there is between neuronal Research (G.G.L., V.B.C.). J.M.W. was a recipient of an SERC and muscle nicotinic receptor subunits. That is to say, Postgraduate Training Award. We are indebted to Professor Michael within the same species the enigmatic brain a- Benn for providing the MLA. bungarotoxin binding protein is less similar to the muscle nicotinic receptor, which also binds a-bungarotoxin REFERENCES

HI Jennings, K.R., Brown, D.G. and Wright, D.P. (1986) mtbyUyc8cmitine I Experientia 42, 611-613. alpha-bungamcmin I[nicotincI PI MacAllan, D.R.E., Lunt, G.G., Wonnacott, S., Swanson, K.L., Rapoport, H. and Albuquerque, E.X. (1988) FEBS Lett. 226, nAC!hR RsaSP IlAChR IlAChR 357-363. ganglia ncumnal UUJSClC blain [31 Sattelle, D.B., Pinnock, R.D. and Lummis, S.C.R. (1989) J. [I [ venebrue Exp. Biol. 142, 215-224. [41 Colquhoun, L., Gallagher, T., Holden-dye, L. and Walker, R.J. (1990) Neurosci. Lett. S38, S113. Fig. 3. Classification of nicotinic receptors based on their high PI Nambi Aiyar, V., Benn, M.H., Hanna, T., Jacyno, J., Roth, affinity binding of MLA, or-bungarotoxin and nicotine. S.H. and Wilkens, J.L. (1979) Experientia 35, 1367-1368.

47 Volume 270, number 1,2 FEBS LETTERS September 1990

[6] Wonnacott, S. (1990) in: Nicotine Psychopharmacology: [15] Stephenson, F.A., Harrison, R. and Lunt, G.G. (1981) Eur. J. Molecular, Cellular and Behaviour Aspects, (S. Wonnacott, Biochem. 115, 91-97. M.A.H. Russel and I.P. Stolerman, eds) Oxford University [16] Dauber-Osguthorpe, P., Roberts, V.A., Osguthorpe, D.J., Press, pp. 226-277. Wolff, J., Genest, M., Hagler, A.T. (1988) Proteins: Struct. [7] Schoepfer, R., Conroy, W.G., Whiting, P., Gore, M. and Funct. Genet. 4, 31-47. Lindstrom, J. (1990) Neuron (in press). (171 Codding, P.W. (1982) Actga Crystallogr., Sect. B 38, 2519. [8] Conti-Tronconi, B.M., Dunn, S.M. J., Barnard, E.A., Dolly, [18] Struthers, R.S., Rivier, J. and Hagler, A.T. (1984) in: J.O., Lai, F.A., Ray, N. and Raferty, M.A. (1985) Proc. Natl. Conformationally Directed Drug Design: Peptides and Nucleic Acad. Sci. USA, 82, 5208-5212. Acids as Templates or Targets (J.A. Vida and M. Gordon, eds), [9] De la Garza, R., McGuire, T. J., Freedman, R. and Hoffer, B.J. American Chemical Society, Washington DC, pp. 239-261. (1987) Neuroscience 23, 887-891. [19] Catterall, W. (1980)Annu. Rev. Pharmacol. Toxicol. 20,15-43. [lo] Breer, H. and Sattelle, D.B. (1987) J. Insect Physiol. 33, [20] Sheridan, R.P., Nilakantan, R., Dixon, J.S., Venkataraghavan, 771-790. R. (1986) J. Med. Chem. 29, 899-906. [ll] Wonnacott, S., Harrison, R. and Lunt, G.G. (1982) Brain Res. [21] Dennis, M., Giraudat, J., Kotzyba-Hibert, F., Goeldner, M., 231, 365-377. Hirth, C., Chang, J.-Y., Lazure, C., Chretien, M. and [12] Syapin, P.J., Salvaterra, P.M. and Engelhardt, J.K. (1982) Changeux, J.-P. (1988) Biochemistry 27, 2346-2357. Brain Res. 231, 365-377. [22] Cockcroft, V.B., Osguthorpe, D., Barnard, E.A. and Lunt, [13] Walker, R.J., Vincent, A. and Newsom-Davis, J. (1988) J. G.G. (1990) Proteins: Struct. Funct. Genet. (submitted). Neuroimmunol. 19, 149-157. [23] Marshall, J. (1989) PhD. Thesis, University of Cambridge. [14] Rozental, R., Aracava, Y., Scoble, G.T., Swanson, K.L., [24] Kao, P.N. and Karlin, A. (1986) J. Biol. Chem. 261,8085-8088. Wonnacott, S. and Albuquerque, E.X. (1989) J. Pharmacol. [25] Wonnacott, S., Drasdo, A. and Caulfield, M. (1990) Sot. Exp. Ther. 251, 395-404. Neurosci. Abstr.