COMMENTARY

Picrotoxin-like channel blockers of GABAA receptors

Richard W. Olsen* Department of Molecular and Medical Pharmacology, Geffen School of Medicine, University of California, Los Angeles, CA 90095-1735

icrotoxin (PTX) is the prototypic vous system. Instead of an acetylcholine antagonist of GABAA receptors (ACh) target, the cage convulsants are (GABARs), the primary media- noncompetitive GABAR antagonists act- tors of inhibitory neurotransmis- ing at the PTX site: they inhibit GABAR Psion (rapid and tonic) in the nervous currents and synapses in mammalian neu- system. Picrotoxinin (Fig. 1A), the active rons and inhibit [3H]dihydropicrotoxinin ingredient in this plant convulsant, struc- binding to GABAR sites in brain mem- turally does not resemble GABA, a sim- branes (7, 9). A potent example, t-butyl ple, small amino acid, but it is a polycylic bicyclophosphorothionate, is a major re- compound with no nitrogen atom. The search tool used to assay GABARs by compound somehow prevents ion flow radio-ligand binding (10). through the chloride channel activated by This drug target appears to be the site GABA in the GABAR, a member of the of action of the experimental convulsant cys-loop, ligand-gated ion channel super- pentylenetetrazol (1, 4) and numerous family. Unlike the competitive GABAR polychlorinated hydrocarbon insecticides, antagonist bicuculline, PTX is clearly a including dieldrin, lindane, and fipronil, noncompetitive antagonist (NCA), acting compounds that have been applied in not at the GABA recognition site but per- huge amounts to the environment with haps within the ion channel. Thus PTX major agricultural economic impact (2). ͞ appears to be an excellent example of al- Some of the other potent toxicants insec- losteric modulation, which is extremely ticides were also radiolabeled and used to important in protein function in general characterize receptor action, allowing and especially for GABAR (1). Recent structure–function analysis of the chemical advances in structural modeling of family on GABAR in both insects and GABAR (Fig. 1 B and C) are consistent mammals, and to define the receptor site with action of PTX and analogous convul- (2, 4, 8). The point of ref. 2 is that this sants as NCAs. In a recent issue of PNAS, series of compounds has the same molec- Chen et al. (2) describe how numerous ular target (‘‘receptor’’) despite consider- drugs in this category, with a variety of able differences in chemical structure. pharmacological effects, can interact with Thujone, a constituent of oil of worm- the same domain of the GABAR protein wood, found in the beverage absinthe, was also shown by Casida and colleagues (11) within the ion channel. to target the GABAR channels. Absinthe PTX has an interesting history as a is the green-colored alcoholic beverage drug and research tool. Many naturally popular in findesiecleParis, now illegal occurring and synthetic convulsive agents, because of its toxicity. Absinthe was not also blockers of GABA-mediated inhibi- only the wildly popular favorite of Pari- tion, have some rather famous uses, such Fig. 1. Schematic view of the GABAA receptor– sians (in addition to wine, of course, but as stimulants, convulsants, chemical war- chloride channel and its blocker picrotoxinin. (A) there were more absinthe parlors than fare agents, and insecticides, to name a Chemical structure of picrotoxinin. (B) Structure of a cys-loop ligand-gated ion channel receptor. Sche- bread stores!), but considered then and few (3, 4). PTX is isolated from plants of matic view of a native GABAR heteropentamer, now to impart a special creative stimulus the moonseed family, Menispermaceae, with the channel in the center, formed of two to the cultural elite of the day. To para- and its close relatives tutin and coriamyr- copies of an ␣ subunit (␣1–␣6), two copies of a ␤ phrase my commentary (12), why would a ␤ ␤ ␤ tin, from the New Zealand tutu plant Co- subunit ( 1, 2, or 3), and one additional type, drug with toxic and convulsant actions riaria arborea, known as a ‘‘loco weed’’ like ␥2or␦.(C) Cytoplasmic end of the TM2͞ channel ␣-helix for two ␤3 subunits (black dots in B; possibly be considered pleasant or at least that caused occasional poisonings in cows desirable? Thujone, like PTX, is excita- and even in people. PTX was shown to be the other three subunits are not shown) in GABAR, showing the PTX site residues near 2Ј–9Ј (residues tory on the brain (analeptic) [not a de- a stimulant and could cause tonic-clonic are numbered 1Ј–23Ј from the N-terminal bottom pressant, like marijuana]. Such an agent convulsions at fairly low doses, but was of the helix to the C-terminal extracellular end of may produce mood elevation and antide- listed in the Merck Index as late as the the helix). pressant, anxiety-generating, and alerting 1970s as an ‘‘antidote for barbiturate over- effects, as opposed to the anxiolytic, seda- dose’’ (5). It was shown to antagonize tive, and amnestic effects of GABA-en- inhibitory pathways in the nervous sys- Casida and Palmer (8) have synthesized hancing drugs like benzodiazepines and tem activated by GABA, and GABA- and studied a series of synthetic potent ethanol (1). Do not forget, however, that enhancing drugs like barbiturates and neurotoxic chemicals, the ‘‘cage convul- in absinthe one is balancing the stimula- benzodiazepines reverse its action. In vitro, sants.’’ These compounds were developed PTX inhibited GABA-activated inhibitory as insecticides, aimed at the common tar- currents in neurons and crayfish muscle get, acetylcholineresterase, like the well Conflict of interest statement: No conflicts declared. 3 (4, 6). Radioactive [ H]dihydropicrotoxi- known malathion. These cage convulsants See companion article on page 5185 in issue 13 of volume nin bound to specific sites in crayfish mus- were not inhibitors of this enzyme, but 103. cle and mammalian brain that could be they were equally toxic, like nerve gas, but *E-mail: [email protected]. identified as GABARs (7). were acting on another target in the ner- © 2006 by The National Academy of Sciences of the USA

www.pnas.org͞cgi͞doi͞10.1073͞pnas.0601121103 PNAS ͉ April 18, 2006 ͉ vol. 103 ͉ no. 16 ͉ 6081–6082 Downloaded by guest on September 30, 2021 tory effect of thujone with the intoxicating phy. Crystallography data on the snail this domain to the functional activation and depressant effects of ethanol. ACh binding protein (19), which adapts mechanism. All of the data on nAChR Another important development in a homopentameric structure that binds appear to be translatable to the other establishing the connection between ACh, yielded a structure that is not only members of the receptor superfamily, in- PTX-like NCA and GABAR was the homologous to the nicotinic ACh recep- cluding GABAR. Thus the TM2 channel independent identification of the site of tor (nAChR), but verified the subunit domain is established for GABAR, and action of insecticides like dieldrin. In arrangements and interfaces and agonist the action of PTX-like NCA is accepted addition to the critical demonstration and modulator binding pocket placement: by most workers as direct channel block- by Casida and Palmer (8) and others amino acids involved in agonist site bind- ing. The residues involved in NCA action that many insecticides inhibited ing have been identified in the extracellu- are all in TM2 and are consistent with the t-butylbicyclophosphoro[35S]thionate lar domain at subunit interfaces by a site of binding of these ligands. (TBPS) binding (and GABAR chan- combination of affinity labeling and mu- The transmembrane domain of nels), a GABAR homolog gene was tagenesis (20–22). Homologous polypep- GABAR and related members of the cloned from insects by using dieldrin tide loops are involved in GABAR ligand receptor superfamily like the glycine resistance in houseflies as a screen. binding pockets. The GABARs have allo- receptor also have been shown to be Comparing resistant and sensitive ani- steric ligand binding pockets in the extra- sites of action of general anesthetics like Rdl the volatile agent isoflurane and high- mals led to the identification of the cellular domain involving modified agonist Ն gene in Drosophila, which turned out to sites at subunit interfaces for BZ ligands dose ( 100 mM) ethanol (26) and i.v. be the insect orthologue of the mamma- and possibly ethanol (1, 14, 15, 23). anesthetics like etomidate and propofol lian GABAR ␤ subunit (13). The muta- The ␣-helical transmembrane domains (27). The anesthetics appear to bind in a tion conferring resistance to dieldrin of this receptor superfamily remain to be water-filled pocket formed by the mem- (which acts like PTX to noncompeti- solved structurally at the atomic level, but brane-spanning helices, not within, but tively antagonize GABAR) corresponds computer-assisted analysis of cryoelectron behind, the channel pore (26). The model of Chen et al. (2) for to A2ЈS, a residue in the TM2 domain͞ microscopic images of Torpedo nAChR the NCA binding site in the channel of ion channel of the Rdl subunit of the (24) have afforded a structure at 4-Å res- GABAR is based on the ␤3 homopen- GABAR protein, and is needed for olution. This approximation is consistent tameric GABAR, which has a high [35S]TBPS binding (14–18). As described with biochemical evidence: first, noncom- affinity for the NCAs, resembling the by Chen et al. (2), this residue and oth- petitive channel blockers of nAChR were insect channel, and the symmetry of the ers lining the channel at the cytoplasmic found to attach to residues in TM2 (20– protein facilitates modeling and ligand (N-terminal) end (Fig. 1C) participate in 22). Second, cysteine scanning accessibility positioning. The model leads to the novel the binding site for dieldrin, PTX, and mutagenesis showed that the residues pro- conclusion that several NCA molecules of the rest of the toxicants and insecticides posed to line the channel on the lumen ␣ diverse chemical structure all fit into the in the NCA family. An important con- side of the TM2 -helix were accessible to same ligand site in the 8.5-Å pore and tribution showing that this domain solvent, whereas those residues on the block the channel, interacting with the (amino acids in the channel wall in buried parts of the helix were not (21, 25). same TM2 amino acid residues A2Ј,T6Ј, TM2) actually provides a contact Channel blocking functional data were and L9Ј (Fig. 1C), a valuable insight. Nev- point for NCA ligands, rather than just consistent with the pore being formed by ertheless, native mammalian GABARs conformational coupling, was the dem- the five TM2 domains of the heteropen- are not homopentamers and differ from onstration that an affinity-labeled sulf- tameric protein. These residues (2Ј,6Ј, the ␤3 homomer in that the presence of hydryl reagent based on the structure of 9Ј, etc.; Fig. 1C) include the pore size- additional subunits alters the structure– fipronil bound covalently to the cysteine determining residues, the impediment to activity profile for NCAs, consistent with mutation at the 2Ј residue (V257) in ion flow in the middle of the pore (L9Ј), differences in heteromeric GABAR, and TM2 of the GABAR ␣1 subunit (17). and site of NCA action, including binding, channel gating and allosteric interactions The type of structural modeling and thus probably actual NCA binding sites between sites for other modulators and analysis in Chen et al. (2) is especially (2, 17, 18, 23). Third, residues in TM2 are NCAs. The PTX-type NCA drugs for valuable for the ligand-gated ion channel able to affect the agonist sensitivity and GABAR remain interesting today. receptor superfamily, whose structures channel opening probability (14, 20–23), have not been solved by x-ray crystallogra- consistent with an allosteric coupling of I thank Jacob Hanchar for helpful discussions.

1. Martin, D. L. & Olsen, R. W., eds. (2000) GABA in the 10. Squires, R. F., Casida, J. E., Richardson, M. & Sae- 19. Brejc, K., van Dijk, W. J., Klaassen, R. V., Schuur- Nervous System (Lippincott Williams & Wilkins, Phil- derup, E. (1983) Mol. Pharmacol. 23, 326–336. mans, M., van der Oost, J., Smit, A. B. & Sixma, T. adelphia). 11. Hold, K. M., Sirisoma, S. I., Ikeda, T., Narahashi, T. K. (2001) Nature 411, 269–276. 2. Chen, L., Durkin, K. A. & Casida, J. E. (2006) Proc. & Casida, J. E. (2000) Proc. Natl. Acad. Sci. USA 97, 20. Corringer, P.-J. LeNovere, N. & Changeux, J.-P. (2000) Natl. Acad. Sci. USA 103, 5185–5190. 3286–3291. Annu. Rev. Pharmacol. Toxicol. 40, 431–458. 3. Porter, L. A. (1967) Chem. Rev. (Washington, DC), 441–464. 12. Olsen, R. W. (2000) Proc. Natl. Acad. Sci. USA 97, 21. Karlin, A. (2002) Nat. Rev. Neurosci. 3, 102–114. 4. Olsen, R. W. & Gordey, M. (2000) in Handbook of 4417–4418. 22. Chiara, D. C., Trinidad, J. C., Wang, D., Ziebell, Experimental Pharmacology, Pharmacology of Ionic 13. ffrench-Constant, R. H., Steichen, J. C., Rocheleau, M. R., Sullivan, D. & Cohen, J. B. (2003) Biochemistry Channel Function: Activators and Inhibitors, eds. T. A., Aronstein, K. & Roush, R. T. (1993) Proc. Natl. 42, 271–283. Endo, M., Kurachi, Y. & Mishina, M. (Springer, Acad. Sci. USA 90, 1957–1961. 23. Chang, Y. & Weiss, D. S. (2000) in GABA in the Heidelberg), Vol. 147, pp. 497–515. 14. Olsen, R. W., Chang, C.-S. S., Li, G., Hanchar, H. J. Nervous System, eds. Martin, D. L. & Olsen, R. W. 5. Windholz, M., Budavari, S., Stroumtsos, L. Y. & & Wallner, M. (2004) Biochem. Pharmacol. 68, 1675– (Lippincott Williams & Wilkins, Philadelphia), pp. Fertig, M. N., eds. (1976) The Merck Index (Merck, Rahway, NJ), 9th Ed. 1684. 127–139. 6. Takeuchi, A. & Takeuchi, N. (1969) J. Physiol. (Lon- 15. Sigel, E. (2002) Curr. Top. Med. Chem. 2, 833–840. 24. Miyazawa, A., Fujiyoshi, Y. & Unwin, N. (2003) don) 205, 377–391. 16. Xu, M., Covey, D. F. & Akabas, M. H. (1995) Biophys. Nature 423, 949–958. 7. Ticku, M. K. & Olsen, R. W. (1979) Neuropharma- J. 69, 1858–1867. 25. Horenstein, J., Wagner, D. A., Czajkowski, C. & Aka- cology 18, 315–318. 17. Perret, P., Sarda, X., Wolff, M., Wu, T. T., Bushey, D. bas, M. H. (2001) Nat. Neurosci. 4, 477–485. 8. Casida, J. E. & Palmer, C. J. (1988) Adv. Biochem. & Goeldner, M. (1999) J. Biol. Chem. 274, 25350– 26. Yamakura, T., Bertaccini, E., Trudell, J. R. & Harris, Psychopharm. 45, 109–123. 25354. R. A. (2001) Annu. Rev. Pharmacol. Toxicol. 41, 23–51. 9. Bowery, N. G., Collins, J. F. & Hill, R. G. (1976) 18. Jursky, F., Fuchs, K., Buhr, A., Tretter, V., Sigel, E. & 27. Rudolph, U. & Mohler, H. (2004) Annu. Rev. Phar- Nature 261, 601–603. Sieghart, W. (2000) J. Neurochem. 74, 1310–1316. macol. Toxicol. 44, 475–498.

6082 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0601121103 Olsen Downloaded by guest on September 30, 2021