Excitatory amino acids. Synthesis of (RS)-2-amino-3-(5-cyclopropyl-3-hydroxyisoxazol-4-yl)propionic acid, a new potent and specific AMPA receptor agonist

摘要

J. CHEM. soc. PERKIN TRANS. I 1995 221 Excitatory Amino Acids. Synthesis of (RS)-2-Amino-3-(5-cyclopropyl-3-hydroxyisoxatol-4-y1)propionic Acid, a New Potent and Specific AM PA Receptor Agonist Niels Skjaerbaek, Bjarke Ebert, Erik Falch, Lotte Brehm and Povl Krogsgaard-LarsenPharmaBiotec Research Center, Department of Medicinal Chemistry, The Royal Danish School of Pharmacy, DK-2 100 Copenhagen, Denmark The synthesis of (RS)-2-amino-3-(5-cyclopropyl-3-hydroxyisoxazol-4-yl)propionic acid 6, an analogue of the AMPA receptor agonist (RS)-2-amino-3-(3-hydroxy-5-methylisoxazol-4-yl)-propionic acid, AMPA, 3 is described. Compound 6 has been studied in vitro in radioligand binding and electrophysiological test systems and shown to be a specific AMPA receptor agonist equi- potent with AM PA.The synthesis of 6 was based on 5-cyclopropyl-3-hydroxyisoxazole 8, which was converted into the intermediate 4-bromomethyl-5-(3-bromopropyl)-2-methoxymethyl-2,3-dihydroisoxazol-3-one 9 by opening of the cyclopropane ring. Based on 'H and 13C NMR data, this conversion has been shown to proceed stepwise, the progression of the different steps being dependent on the concentration of the hydrobromic acid medium, reaction time, and temperature. An acetamidomalonate group has been regiospecifically substituted for the allylic bromine atom of 9 to give 10, and treatment of 10 with sodium hydride gave compound 11 containing a cyclopropyl group, reformed by cyclization of the 3-bromopropyl substituent of 10. Compound 11 has been fully deprotected by treatment with aqueous trifluoroacetic acid to give 6.Glutamic acid 1 is the major excitatory amino acid (EAA) neurotransmitter in the central nervous system '**and appears to play a crucial role in certain neurodegenerative disorder^.^ The physiological and pathophysiological functions of 1 are mediated by multiple ionotropic and metabotropic receptors, the former class of which comprises three receptor families 1 2 named N-methyl-D-aspartic acid (NMDA), (RS)-2-amino-3-(3-hydroxy-5-methylisoxazol-4-yl)propionicacid (AMPA), and kainic acid receptors.'-5 At the NMDA receptor complex, NMDA is a specific agonist and glycine an endogenous co- agonist, whereas (RS)-3-(2-carboxypiperazin-4-yl)propylphos-phonic acid (CPP) is a competitive antagonist and (RS)-10,l 1-dihydro-5-methyl-5H-dibenzou,dcyclohepten-5,l0-iminef (MK-801) a non-competitive antagonist.6 Whilst ibotenic acid 2' is a non-selective NMDA agonist,' AMPA 3' specifically activates AMPA receptors,' which are blocked by 6-cyano-7- nitro-l,2,3,4-tetrahydroquinoxalin-2,3-dione(CNQX).' We have previously shown that (RS)-2-amino-3-(5-ethyl-3-hydroxyisoxazol-4-y1)propionicacid 4 is slightly more potent than 3 as an AMPA receptor agonist," whereas (RS)-2-amino-3-(5-tert-butyl-3-hydroxyisoxazol-4-yl)propionicacid 5 is markedly weaker l2 (see Table 3). On the basis of structure-activity studies on 3-5 and a number of other AMPA receptor agonists we have proposed a model of the AMPA receptor binding site containing a cavity of limited size capable of accommodating lipophilic groups." In order to judge the capacity of this simple receptor model as a template for the semirational design of AMPA receptor agonists and partial agonists of potential therapeutic interest, we have now synthesized and tested pharmacologically the cyclopropyl analogue of 3, (RS)-2-amino-3-(5-cyclopropyl-3-hydroxyisoxazol-4-y1)propionicacid 6.7 This is the name used in Chemical Abstracts. According to IUPAC recommendations this compound is named: (RS)-5,10-epimino-5-methyl-10,Il -dihydro-5H-dibenzoa,dcycloheptene. HO4' HO4O 3 4 5 6 Results and Discussion Chernisrry.-Ethyl cyclopropylpropiolate 7 was converted into 5-cyclopropyl-3-hydroxyisoxazole8 by treatment with hydroxylamine under basic conditions (70 yield) (Scheme 1).Treatment of8 with a solution of 1,3,5-trioxane in hydrobromic acid (62) and subsequent treatment of the intermediate, containing bromomethyl groups at positions 2 and 4 of the ring, with methanol under previously described reaction conditions gave 9 in 99 yield. As expected,l4 these strongly acidic reaction conditions caused opening of the cyclopropane ring to give the 3-bromopropyl group in 9. A Sorensen reaction regiospecifically converted compound 9 into 10 (6673, and 222 J. CHEM. SOC. PERKIN TRANS. i 1995 Table 1 "C NMR spectroscopic data for 12-17 6,Solvent Compound (aq. HBr, ) C-3' C-4 C-5 C-I' C-2' C-3' 2-CH,-X 4-CH,-Br 12 47 164.5 92.2 181.5 9.3 11.6 11.6 69.2b -12 62 165.0 92.4 181.2 9.4 11.7 11.7 69.2' -13 62 166.1 92.8 183.0 9.7 12.2 12.2 38.5' -14 47 163.5 105.9 176.3 8.3 10.3 10.3 68.6* 19.2 14 62 163.7 105.8 176.4 8.6 10.5 10.5 68.4' 18.9 15 62 166.9 105.4 177.0 8.2 10.4 10.4 40.0' 18.7 16 62 163.7 107.8 173.6 28.2 25.0 33.1 68.9' 18.6 17 62 166.3 107.4 174.1 28.2 24.8 33.2 39.6' 18.1 'The assignment of the 6 values for C-3 as well as for C-4, C-5, C-l', C-2' and C-3', in the mixture of compounds 12 and 13 are tentative.The same applies for the mixtures of compounds 14 and 15, and of compounds 16 and 17. X = OH. X = Br. Table 2 Ratios of intermediates in the conversion of 8 into 9 based on I3C NMR integrals Solvent Reaction time (aq. HBr, ) (h) 47 0.2 62 0.2 47 48 62 48 62 0.2 62 6 0 cEC-6: i OEt 7 81ii Br-Ldo Br Br 101iv Q -HoH$; V 11 S Product ratios Temperature ("C) 12:13 14:15 16:17 20 5 20 5 60 60 1oo:o --50:50 ---1oo:o --50:50 --60:40 --55:45 isoxazole amino acid 6 in 58 yield using aqueous tri-fluoroacetic acid (1 mol dm3).The reactions involved in the conversion of 8 into 9 were studied in detail by I3C NMR spectroscopy (Table 1). Under carefully controlled reaction conditions it could be demon- strated that bromomethylation/hydroxymethylation reactions and the subsequent opening of the cyclopropane ring proceeded stepwise (Scheme 2). Compound 8 was initially converted into an equilibrium mixture of 12 and 13 by treatment with trioxane and 62 hydrobromic acid at 5 "C for 0.2 h (Table 2), with no detectable formation of other reaction products.After 48 h under these reaction conditions, the further conversion of 12 and 13 into an equilibrium mixture of 14 and 15 was complete. This virtually quantitative bromomethylation at C-4 was shown to proceed without detectable opening of the cyclopropane ring. Using 47 hydrobromic acid at 20"C, compound 8 was converted exclusively into the hydroxymethyl compound 12 after 0.2 h reaction time (Table 2). Analogously, only the hydroxymethyl compound 14 could be detected after 48 h under the same reaction conditions. At 60deg;C and by using 62 hydrobromic acid for 0.2 h, only a (60:40)mixture of 14 and 15 could be detected, but after 6 h under these conditions a complete opening of the cyclopropane ring has taken place, with formation of an equilibrium mixture of the 5-(3-bromopropyl) compounds 16 and 17.Pharmacdogy,--Compound 6 was shown to bind to the AMPA receptor site with an affinity (IC,o = 0.035 pmol dm-3) comparable with those of AMPA 3 (IC,o = 0.040 pmol dm-3) and 4 (IC50 = 0.030 pmol dm-3) but much higher than that observed for 5 (ICs0 = 2.1 pmol dm3) (Table 3). Compound 6 was shown electrophysiologically in vitro to be an AMPA receptor agonist (EC,, = 5.5 pmol dm-3) slightly weaker than AMPA 3 (ECSo= 3.5 pmol dm-3) and 4 (EC50= 2.3 pol dm-3). The pharmacological profile of 6is, however, markedly different from that of AMPA 3.Thus, whereas AMPA 3 is a much more potent inhibitor of the binding of the AMPA antagonist, C3HCNQX, in the presence of potassium thio- cyanate (IC,o = 0.4 pmol dm-3) than in the absence of this Scheme 1 i, NH,OH, NaOH; ii, (CH,0)3, 62 aq. HBr, MeOH; iii, AcNHCH(CO,Me),; iv, NaH; v, CF,CO,H (1 mol dm-3) treatment of 10 with sodium hydride gave the cyclopropyl analogue 11in 74 yield. This cyclization reaction proceeds via deprotonation of the highly activated methylene group LX to the isoxazole ring.' Under the reaction conditions used, intra- molecular N-alkylation, with formation of a hexahydroazocine ring, or a Claisen type reaction between the acetyl group and one of the methoxycarbonyl groups of 10 to form a five-membered P-dicarbonyl ring structure,I6 were not observed. Compound 11 was fully deprotected to give the desired acidic J.CHEM. SOC. PERKIN TRANS. I 1995 Table 3 In uitro radioligand binding and electrophysiological data ICs0(pmol dm-3) ElectrophysiologyCompound C3HAMPA C3HCNQX C3HCNQX + KSCN (EC,,, pmol dm-') AMPA3 0.040 f 0.005 18 f 5 0.4 f 0.1 3.5 f 0.2 4 0.030 f 0.015 n.d." 0.4 _+ 0.1 2.3 f 0.2 5 2.1 f 0.2 100 100 34 f 2 6 0.035f 0.015 12 f 4 4.7 f 2.0 5.5 f 1.0 Values are mean values f standard error of means from at least three separate experiments. Not determined. 3' 8 I aq. HBr (47 or 62)O"0 7OH t -F 4 B r *"-J - Br Br 1s 17 Scheme 2 chaotropic agent (ICs0 = 18 pmol dm-'), such a marked difference is not observed for 6 (Table 3).These receptor affinity data strongly suggest that whereas AMPA 3 very selectively interacts with the high-affinity agonist conformation of the AMPA receptor, 6 binds with almost equal affinity to the high- affinity (agonist) and low-affinity (antagonist) conformation of the receptor in spite of its agonist character." We have previously shown that adding potassium thiocyanate to the C3HCNQX binding assay increases the affinity of agonists and decreases the affinity of antagonists. Thus, this result may reflect that the two enantiomers of 6 have different pharmacological profiles. These aspects make 6 an interesting tool for studies of AMPA receptor mechanisms.The selectivity of 6 as an AMPA receptor agonist is emphasized by the observation that 6 did not significantly affect the binding of 'Hlkainic acid (kainic acid receptor) or C3HCPP, E3HMK-801, and C3Hglycine (NMDA receptor complex) (ICs0 100 pmol drn-,). Experimental General.-Dichloromethane was dried over sodium hydride. Dimethylformamide was dried as described. Unless otherwise stated, solutions were dried using magnesium sulfate. Solvents were removed under reduced pressure by rotatory evaporation. Column chromatography was performed as de~cribed.'~ All new compounds were colourless, unless otherwise stated. Structure and homogeneity of new compounds were confirmed by m.p. determinations (capillary tubes, uncorrected), elemen- tal analyses (determined by Mr.M. G. Cornali, Micro- analytical Laboratory, Leo Pharmaceutical Products, Ballerup, Denmark), TLC (silica FL5., plates, Merck), and NMR spectroscopy. 'H and 13C NMR spectra were recorded at 200 and 50.32 MHz, respectively, on a Bruker AC-200 instrument. Signal positions in the 'H NMR spectra are given as 6 values relative to tetramethylsilane (TMS), when CDCI, was used as a solvent, and relative to dioxane (6 3.70), when water was used. Coupling constants (J)are given in Hz. 'C NMR signals were assigned through their multiplicity in the coupled spectra or through DEPT spectra, and using the CDC1, peak (8 76.93) or the dioxane peak (6 67.40), when CDCI, or water, respectively, were used as solvents.NMR Kinetic Experiments.-Compound 8 (100 mg, 0.80 mmol) and 1,3,5-trioxane (108 mg, 1.20 mmol) were dissolved in hydrobromic acid of different concentrations (0.5 cm') in an NMR tube (Tables 1 and 2). Temperature experiments were carried out in a refrigerator (5 "C), at room temperature (20 "C), or by using an oil bath (60 "C). 5-Cyclopropyl-3-hydroxyisoxazole-A solution of ethyl cyclopropylpropiolate 72o in methanol (100 cm') was added dropwise to an ice-cold solution of hydroxylamine hydrochlor- ide (5.21 g, 75 mmol) and sodium hydroxide (6.0 g, 150 mmol) in water (75 cm3). The reaction mixture was stirred at 0 "C for 3 h and then at room temperature for 20 h. The reaction mix- ture was acidified with concentrated hydrochloric acid and evaporated. To the oily residue was added water (50 cm3), and the precipitate was collected and washed with water (20 cm').After drying, the solid was extracted with diethyl ether (2 x 30 cm3), and the combined extracts were evaporated. Recrystal- lization of the residue from diethyl ether-light petroleum gave the title compound 8 (2.61 g, 70), m.p. 103-105 "C (lit.,21 m.p. 106-107 "C); SH(CDC13) 9.5 (1 H, S, OH), 5.35 (1 H, S, 4-H), 1.85 (1 H, m) and 0.95 (4 H, m); Gc(CDC13) 10.3 (CH), 23.6 (2 x CH,), 105.2 (C-4), 168.2 (C-3) and 173.5 (C-5). 4-Bromomethyl-5-(3-bromopropyl)-2-methox~methyl-2,3-di-hydroisoxazol-3-one 9.-5-Cyclopropyl-3-hydroxyisoxazole 8 (2.17 g, 17.3 mmol) and 1,3,5-trioxane (2.34 g, 26.0 mmol) were treated with hydrobromic acid (40 cm', 62) for 18 h at 60 OC.This mixture was extracted with dichloromethane (3 x 50 cm3) and then methanol (50cm') was added to the combined organic extracts and the mixture was stirred for 2 h at 20 OC. Addition of dichloromethane (35 cm3), washing with water (2 x 50 cm3), drying and removal of the organic solvents, produced the title compound 9 (5.90 g, 99) as a yellow oil; 6,(CDCl,) 2.28 (2 H, quintet, J7.7), 2.90 (2 H, t, J7.7), 3.40 (3 H, s, OCHJ, 3.49 (2 H, t, J 7.7), 4.23 (2 H, s, CH,Br) and 5.17 (2 H, s, N-CH,O); Gc(CDC1,) 18.5 (BrCH,C-4), 24.7 (BrCH,CH,), 28.4 (CH,CH,C-5), 31.7 (BrCH,CH,), 57.0 (OCH,), 75.2 (NCH,O), 107.7 (C-4), 165.5 (C-3) and 169.9 (C-5) (Found: C, 31.35;H, 3.75;N, 4.15.C,H,,Br,NO, requiresC, 31.51; H, 3.82; N, 4.08). Methyl 2-Acetamido-3-5-(3-bromopropyl)-2-methoxy-methyl-3-oxo-2,3-dihydroisoxazol-4-yl-2-methoxycarbonyl-propionate 10.-A 60 suspension of sodium hydride in mineral oil (384 mg, 9.6 mmol) was added over a period of 10 min to a solution of dimethyl acetamidomalonate (1.82 g, 9.6 mmol) in dimethyl formamide (I5 cm'). After stirring for 15 min, a solution of compound 9 (3.00 g, 8.75 mmol) in dimethylformamide (7.5 cm3) was added during 10 min. Stirring was continued for 10 h, and then the reaction mixture was evaporated to dryness, dissolved in dichloromethane (15 cm3), and washed with sodium hydroxide (1 mol dm-3; 15 cm3, OOC) and water (2 x 15 cm3, OOC). The organic phase was dried, evaporated and subjected to column chromatography silica gel, Woelm, 0.063-0.200 mm; toluenexthyl acetate- acetic acid (24: 75: l), which afforded the title compound 10 (2.62 g, 66) as an oil; G,(CDCl,) 2.04 (3 H, s, CH,CON), 2.16 (2 H, quintet, J 7.9, 2.72 (2 H, t, J 74, 3.32 (2 H, s, 4-CH2), 3.37(3H,s,OCH3),3.43(2H,t,J7.5),3.8l(6H,s,2x CH,), 5.12 (2 H, s, NCH,O) and 7.24 (1 H, s, NH); G,(CDCl,) 22.7 (CH,CO), 24.3 (BrCH,CH,), 25.7 (CH,CH,C-5), 28.6 (BrCH,CH2), 31.4 (CCH,C-4), 53.4 (2 x CO,CH,), 56.9 (OCH,), 65.0 (NHC), 74.9 (NCH,O), 103.9 (C-4), 167.3 (C-3), 167.7 (2 x CO,CH,), 169.4 (C-5) and 170.1 (CON) (Found: C, 42.6; H, 5.3; N, 6.5.C16H23BrN208 requires C, 42.66; H, 5.15; N, 6.22). Methyl 2-Acetamido-3-(5-cyclopropyl-2-methoxymethyl-3-0~0-2,3-dihydroisoxazol-4-yI)-2-methoxycarbonylpropionate 11.-At -11 'C a solution of 10 (848 mg, 1.88 mmol) in acetonitrile (5 cm3) was added during 2 min to a suspension of sodium hydride (60, 301 mg, 7.5 mmol) in acetonitrile (25 cm3).After being stirred at 20deg;C for 2 h, the solvent was evaporated and water (10 cm') was added to the residue. The mixture was extracted with dichloromethane (3 x 20 cm3), and the combined extracts were dried and evaporated and then the residue was purified by column chromatography silica gel, Woelm, 0.063-0.200 mm; toluene-ethyl acetate-acetic acid (24: 75: I) to give the title compound 11 (515 mg, 74), m.p. 130-1 30.5 OC (ethyl acetate-light petroleum); G,(CDCl,) 1.05 (4H, m), 1.94 (1 H, m), 2.04 (3 H, s, CH,CON), 3.34 (3 H, s, OCH,), 3.41 (2 H, s, 4-CH2), 3.83 (6 H, s, 2 x CO,CH,), 5.04 (2 H, s, NCH,O) and 7.10 (1 H, s, NH); G,(CDCl,) 7.5 (CHC-5),20.7(2 x CO,CH3),22.9(CH,CO),25.6(2x CH,), 26.7 (CCH2C4), 57.0 (OCH3), 65.3 (NHC), 75.3 (NCH,O), 102.5 (C-4), 167.5 (C-3), 168.3 (2 x CO,CH,), 169.6 ((2-5) and 172.6 (CON) (Found: C, 52.05; H, 6.1; N, 7.40.Cl,H,,N,08 requires C, 51 37; H, 5.99; N, 7.57). (RS)-2-Amino-3-(5-cyclopropyl-3-hydroxyisoxazol-4-yl)prop-ionic Acid 6.-Compound 11(275 mg, 0.74 mmol) was refluxed in aqueous trifluoroacetic acid (1 mol drn-,; 5 cm3) for 12 h and then the mixture was evaporated to dryness. The residue was twice dissolved in water (2 x 5 cm3) and re-evaporated and then twice dissolved in toluene (2 x 5 cm3) and re-evaporated.Preparative TLC acetonitrile-water-acetic acid (8 :1 : 1); R,= 0.331 gave the title compound 6(91 mg, 58), m.p. 217-218 "C (decomp.) (water); 6,(D,O) 1.00 (4 H, m), 1.89 (1 H, m), 2.82 (2 H, d, 4-CH2, J 7) and 3.87 (1 H, t, CH,CH, J 7); S,(D,O) 13.2(CHC-5),25.5(2 x CHJ,30.3(CH,C-4), 58.9(CH), 103.5 (C-4), 169.9 (C-5), 179.1 (C-3) and 181.0 (C0,H) (Found: C, 50.15; H, 5.65; N, 13.2. C9H1,N204, 25 mol H,O requires C, 49.87; H, 5.82; N, 12.93). J. CHEM. SOC. PERKIN TRANS. 1 1995 Radioligand Binding Assays.-The membrane preparations used in the C3HAMPA, C3Hkainic acid, C3HCPP, C3HMK- 801, C3Hglycine and C3HCNQX binding assays were prepared as described.22 3HAMPA,23 C3Hkainic acid,24 C3HCPPZ5 and C3HCNQX 7,26 binding was performed following pub- lished procedures.C3HMK-801 binding to fully stimulated membranes was performed essentially as described earlier. 27 C3Hglycine binding was carried out by a modified version of the method described,28 using filtration through Whatman GF/B filters instead of centrifugation to isolate bound ligand. CNQX was a gift from Dr. Poul Jacobsen, Novo-Nordisk, M810v. Denmark. Electrophysio1ogy.-A rat cortical slice preparation for testing the depolarizing activities of EAAs described by Harrison and SimmondsZ9 was used in a modified version. Wedges (500 pm thick) of rat brain containing cerebral cortex and corpus callosum were placed with the cortex part between two layers of absorbent fibre and the corpus callosum part between two other layers of absorbent fibre.The two halves were electrically insulated from each other with a grease gap. The cortical part was constantly perfused with a magnesium- free, oxygenated Krebs buffer to which the compounds tested were added, whereas the corpus callosum part was perfused with a magnesium- and calcium-free Krebs buffer. The two parts were each in contact with an Ag/AgCI electrode through which DC potentials were measured and plotted on a chart recorder. Acknowledgements This work has been supported by a grant from The Lundbeck Foundation. The NMR spectrometer is a gift from the Velux Foundation of 198 1. Dr. Per Veds~, The Royal Danish School of Pharmacy, is gratefully acknowledged for fruitful discussions and helpful assistance during this work.The secretarial assistance of Mrs. Anne Nordly is gratefully acknowledged. References 1 Excitatory Amino Acids and Synaptic Transmission eds. H. V. Wheal and A. M. Thomson, Academic Press, London, 1991. 2 Excitatory Amino Acid Receptors: Design of Agonists and Antagonists, eds. P. Krogsgaard-Larsen and J. J. Hansen, Ellis Horwood, Chichester, 1992. 3 Excitatory Amino Acids in Health and Disease, ed. D. Lodge, Wiley, Chichester, 1988. 4 J. C. Watkins, P. Krogsgaard-Larsen and T. Honore, Trends Pharmacol. Sci., 1990, 11,25. 5 Excitatory Amino Acids, ed. R. P. Simon, Thieme Medical Publishers, New York, 1992. 6 The NMDA Receptor, eds. J. C. Watkins and G. Collingridge, Oxford University Press, Oxford, 1989.7 U. Madsen, L. Brehm and P. Krogsgaard-Larsen, J. Chem. Soc., Perkin Trans. I, 1988,359. 8 P. Krogsgaard-Larsen, T. Honort, J. J. Hansen, D. R. Curtis and D. Lodge, Nature, 1980, 284,64. 9 J. J. Hansen and P. Krogsgaard-Larsen, J. Chem. Soc., Perkin Trans. I, 1980, 1826. 10 T. Honoramp;,Med. Res. Rev., 1989, 9, I. 11 U. Madsen, B. Frelund, T. M. Lund, B. Ebert and P. Krogsgaard- Larsen, Eur. J. Med. Chem., 1993,28, 791. 12 J. Lauridsen, T. Honort and P. Krogsgaard-Larsen, J. Med. Chem., 1985,28,668. 13 M. Begtrup and F. A. Slek, Synthesis, 1993,9, 861. 14 A, J. Gordon, J. Chem. Educ., 1967,44,481. 15 P. Krogsgaard-Larsen, A. L. N. Larsen and K. Thyssen, Acta Chem. Scand., Ser. B, 1978,32, 469.16 L. Brehm, J. S. Johansen and P. Krogsgaard-Larsen, Acta Chem. Scand., Ser. E, 1985,39, 807. 17 B. Ebert. U. Madsen, T. M. Lund, S. M. Lenz and P. Krogsgaard-Larsen, Neurochem. Int., 1994,24, 507. J. CHEM. SOC. PERKIN TRANS. I 1995 18 D. R. Burfield and R. H. Smithers, J. Org. Chem., 1978,43,3966. 19 W. R. Still, M. Kahn and A. Mitra, J. Org. Chem., 1978,43,2923. 20 0.Hormi, Org. Synth., 1988,66, 173. 21 K. Sato, S. Sugai and K. Tomita, Agric. Biol. Chem., 1988,50, 1831. 22 R. W. Ransom and N. L. Stec, J. Neurochem., 1988,51,830. 23 T. Honore and M. Nielsen, Neurosci. Lett., 1985,54, 27. 24 D. J. Braitman and J. T. Coyle, Neuropharmacology, 1987, 26, 1247. 25 D. E. Murphy, J. Schneider, C. Boehm, J. Lehmann and K. Williams, J. Pharmacol. Exp. Ther., 1987,240,778. 26 T. Honore, J. Drejer, E. 0. Nielsen and M. Nielsen, Biochem. Pharmacol., 1989,38,3207. 27 A. C. Foster and E. H. F. Wong, Br. J. Pharmacol., 1987,91,403. 28 J. A. Kemp, A. C. Foster, P. D. Leeson, T. Priestly, R. Tridgett, L. L. Eversen and G. N. Woodruff, Proc. Natl. Acad. Sci. USA, 1988,85,6547. 29 N. L. Harrison and M. A.Simmonds, Br. J. Pharmacol.,1985,84,38 1. Paper 4j05492.J Received 9th Sepiernber 1994 Accepted 13th October 1994