ECHET96 Search CD [Molecules: 5] [Related articles/posters: 024 018 042 063 117 ]

New approach to 3-hydroxyisoxazole-5-carbaldehydes: key intermediates in the synthesis of new 3-hydroxyisoxazoles of pharmacological interest

Michael Schön, Regine Riess, Sabine Laschat and Volker Jäger*

Institut für Organische Chemie der Universität Stuttgart, Pfaffenwaldring 55, D-70569 Stuttgart, Germany


3-O-Protected 3-hydroxyisoxazole-5-carbaldehydes are useful building blocks in the synthesis of pharmaceutically interesting 3-hydroxyisoxazoles. We report here our results concerning the synthesis of such intermediates and their conversion, e.g. into the free vinylene muscimol, an analogue of g-vinyl-GABA (g-vinyl-g-amino butyric acid) which represents a strong inhibitor of GABA-transaminase.

Since the discovery of muscimol A1 and ibotenic acid B - two psychoactive constituents of the mushrooms Amanita muscaria and A. pantherina - much attention has been focused on the synthesis of related 3-hydroxyisoxazoles that often display remarkable central nervous system (CNS) activity2 (see Fig. 1). The most common examples are 4,5,6,7-tetrahydroisoxazolo[5,4-c]pyridin-3-ol C (THIP)1,2f,3 - a bicyclic and less toxic analogue of muscimol - and the glutamic acid analogue (S)-a-amino-3-(3-hydroxy-4-isoxazolyl)propionic acid D-[(S)-AMPA].4 THIP C was subjected to clinical trials with epileptic patients, but no significant effects were detected.3

The above compounds may be regarded as conformationally restricted bio-isosteres of neurotransmitters such as g-amino butyric acid (GABA),3 glutamic acid or N-methyl-D-aspartic acid (NMDA).4 Although the 3-hydroxyisoxazole moiety thus represents an important class of heterocycles, only a few general approaches for their synthesis are known.5

Fig. 1 Examples for naturally occurring or synthetically 3-hydroxy-isoxazoles of pharmaceutical interest

Starting from 3-hydroxyisoxazol 1, as detailed below, we have prepared the protected 3-hydroxy-isoxazole-5-carbaldehydes 2 and converted them into the unprotected 3-hydroxyisoxazole-5-carbaldehyde 3 and the 3-hydroxy-isoxazoles vinylene muscimol 4 (see Scheme 1). The aldehyde 3 is presumed to be the major first metabolite of muscimol and is supposed to cause the pronounced toxicity of this compound.1,6 Vinylene muscimol may be regarded as an analogue of the mechanism-based inhibitor of GABA transaminase g-vinyl GABA (Vigabatrin").4,7

Scheme 1 Retrosynthetic approach to 3-hydroxyisoxazole-5-carbaldehyde 3 and vinylene muscimol 4

3-Hydroxyisoxazole-5-carbaldehydes may be regarded as key intermediates in the synthesis of these compounds since the aldehyde function allows various modifications of the side chain. They were used the first time by Nakamura8 et al. and Krogsgaard-Larsen9 for the synthesis of ibotenic and homoibotenic acid respectively.

In order to reduce the ester function to the appropriate aldehyde protection of the hydroxy function of isoxazole 110 is necessary, which is usually carried out by alkylation of the 3-hydroxyisoxazole-anion. This gives rise to mixtures of the two regioisomeric O- and N-alkylated products5,8,9 which have to be separated by chromatography. Another disadvantage is the harsh conditions required for deprotection (e.g. 33% HBr/HOAc). The most common OH-protecting group for 3-hydroxyisoxazoles is the methyl group.2 There are scattered examples for the use of benzyl (Bn),11 p-methoxybenzyl (PMB),12 methoxymethyl (MOM),2b,13 phenyl-sulfonyl14 or benzoyl (Bz)15 groups in this series.

We have examined the suitability of various protecting groups for 3-hydroxyisoxazoles in order to find out whether they could selectively be introduced at oxygen and removed under mild reaction conditions. In view of O-selective alkylation of the 3-OH ester 1, we found that only isoxazoles a could be reduced to the appropriate aldehydes in good yields. The O-/N-alkylation ratios observed are summarised in Table 1.

Table 1 O-/N-Alkylation ratios from 3-hydroxyisoxazole 1

E Reagents and conditions a:ba Yields (%)b
1 Me CH2N2, A 73:27 64 (5a)
25 (5b)
2 Bn PhCH2Br, B 94:6 93 (6a)
5 (6b)
3 Bzh Ph2CHBr, B > 95:5 94 (7a)
0 (7b)
4 All H2C=CHCH2Br, B 91:9 80 (8a)
8 (8b)
5 MOM (MeO)2CH2, C < 5:95 0 (9a)
97 (9b)
6 TBDMS ButMe2SiCl, D > 95:5 39 (10a)
0 (10b)

aRatios determined from the crude products by 13C NMR. bYields of isolated analytically pure products; Reaction conditions: A Et2O, 0 oC; B K2CO3, acetone, reflux to room temp.; C P4O10, CHCl3, room temp.; D DBU, CH2Cl2, room temp.

The highest O-/N-alkylation ratio was obtained in the case of benzhydryl group16 (entry 3) under typical SN1 conditions. The allyl and benzyl reagents proved less regioselective. However, in contrast to the excellent O-/N-alkylation ratio for the introduction of the benzhydryl group16 (entry 3), the use of the somewhat less effective benzyl group16 (entry 2) proved preferable since the Bzh-group showed a tendency to migrate in some of the subsequent reactions applied. In the case of the TBDMS16 group, high regioselectivity was observed for the introduction step (entry 6) but accompanied low yield. An example for a protecting group which could be introduced selectively at nitrogen is the methoxymethyl moiety16 (entry 5), however the subsequent reduction to obtain the respective aldehyde failed.

With a viable method to protect the OH-function of isoxazole 1 with high regioselectivity and in good yield at hand, we explored appropriate methods for deprotection (see Scheme 2), which was successful for free variations.

  1. Acidic hydrolysis of the ester 6a using 33 % HBr/HOAc gave the free hydroxyisoxazole 1 in good yield. These conditions are similar to those applied for deprotection of 3-methoxyisoxazoles.2
  2. Catalytic hydrogenation: Although few examples are known in literature,11 hydrogenative cleavage of the O-Bn bond proved troublesome at first; however, using Rosenmundt's catalyst (Pd/BaSO4) gave no acyclic side-products and the free hydroxyisoxazole 1 was isolated in excellent yield.17
  3. Alternatively, deprotection of ester 6a with NBS using Anson's method18 was also successful, see Scheme 2.

Scheme 2 Deprotection of hydroxyisoxazole 6a

Having established selective O-benzylation and efficient deprotection conditions for 1/6a, we turned our attention to the preparation of the free aldehyde 3 (see Scheme 3). Indeed, the pure O-benzyl-protected hydroxyisoxazole 6a could readily be reduced to the corresponding aldehyde 2 by DIBAL:19

Subsequent deprotection of 2 using 33 % HBr/HOAc was successful likewise and furnished the target compound 3-hydroxyisoxazole-5-carbaldehyde 3 in 61% yield, see Scheme 3.20

Scheme 3 Synthesis of 3-hydroxyisoxazole-5-carbaldehyde (3) and vinylene muscimol 4

For the synthesis of vinylene muscimol 4, the aldehyde 2 was transformed into the allylic alcohol 11 by Grignard reaction. Subsequent Overman rearrangement21 to give the trichloroacetimidate 12 and deprotection under acidic conditions led to the hydrobromide of vinylene muscimol 4 in poor yield, probably due to the harsh rearrangement conditions. Hydrogenation of 12 using Pd/BaSO4 as a catalyst had proved troublesome since partial hydrogenation of carbon-carbon double bond occurred.


We have demonstrated a new, versatile approach towards known and new 3-hydroxyisoxazoles of pharmaceutical interest using protected 3-hydroxyisoxazole-5-carbaldehyde 2 as key intermediate. This aldehyde is now readily available in gram quantities in two steps (91% overall yield) starting from the known isoxazole 1.10 Further studies applying this strategy, e.g. for the synthesis of homoibotenic acid derivatives are in progress.


We are grateful to Fonds der Chemischen Industrie for a fellowship (R. R.) and general support and to Hoechst AG (Dr Gebert, Dr. Mildenberger) for carrying out some biological tests (no significant activity detected).


  1. For reviews on muscimol and related compounds see (a) Krogsgaard-Larsen, P.; Brehm, L.; Schaumburg, K. Acta Chem. Scand. B 1981, 35, 311. (b) De Feudis, F. V. Rev. Pure Appl. Pharmacol. Sci. 1982, 3, 319.
  2. Recent articles included: (a) Madsen, U.; Frydenvang, K.; Ebert, B.; Johansen, T. N.; Brehm, L.; Krogsgaard-Larsen, P. J. Med. Chem. 1996, 39,183. (b) Skjoerboek, N.; Ebert, B.; Falch, E.; Brehm, L.; Krogsgaard-Larsen, P. J. Chem. Soc. Perkin Trans. I, 1995, 221. (c) Ebert, B.; Lenz, S.; Brehm, L.; Bregnedal, P.; Hansen, J. J.; Frederiksen, K.; Bøgesø, K. P.; Krogsgaard-Larsen, P. J. Med. Chem. 1994, 37, 878. (d) Johansen, T. N.; Frydenvang, K.; Ebert, B.; Krogsgaard-Larsen, P.; Madsen, U. J. Med. Chem. 1994, 37, 3252. (e) Madsen, U.; Wong, E. H. F. J. Med. Chem. 1992, 35, 107. (f) Hjeds, H.; Christensen, I. T.; Cornett, C.; Frølund, B.; Falch, E.; Pedersen, J. B.; Krogsgaard-Larsen, P Acta Chem. Scand. 1992, 46, 772. (g) Brehm, L.; Johansen, J. S.; Krogsgaard-Larsen, P. J. Chem. Soc. Perkin Trans. I 1992, 2059. (h) Krogsgaard-Larsen, P.; Ferkany, J. W.; Nielsen, E. Ø.; Madsen, U.; Ebert, B.; Johansen. J. S.; Diemer, N. H.; Bruhn, T.; Beattie, D. T.; Curtis, D. R. J. Med. Chem. 1991, 34, 123.
  3. Krogsgaard-Larsen, P.; Frølund, B.; Jørgensen, F. S.; Schousboe, A. J. Med. Chem. 1994, 37, 2489.
  4. (a) Krogsgaard-Larsen, P. Comprehensive Med. Chem. 1990, 3, 493. (b) Johnson, R. L.; Koerner, J. F. J. Med. Chem. 1988, 31, 2057.
  5. Christensen, S. B.; Krogsgaard-Larsen, P. Acta Chem. Scand. B 1974, 38, 625. For synthesis of 3-hydroxyisoxazoles in general see e.g. Lang Jr., S.A.; Lin, Y.-I. in "Comprehensive Heterocyclic Chemistry", Vol. 6; Katritzky, A.R.; Rees, C.W., Ed.; Pergamon; London, New York 1984.
  6. Fowler, L. J.; Lowell, D. H.; John, R. A. J. Neurochem. 1983, 41, 1751.
  7. (a) Metcalf, B. W.; Jung, M. J.; Lippert, B.; Casara, P.; Böhlen, P.; Schechter, P. J. in GABA Neurotransmitters; Krogsgaard-Larsen, P.; Scheel-Krüger, J.; Kofod, H., Ed.; Munksgaard: Copenhagen, 1990. (b) Allan, R. D.; Johnston, G. A. R. Med. Res. Rev. 1983, 3, 91.
  8. Kishida, Y.; Hiraoka, T.; Ide, J.; Terada, A.; Nakamura, N. Chem. Pharm. Bull. 1966, 14, 92. (9) (a) Hansen, J. J.; Krogsgaard-Larsen, P. J. Chem. Soc. Perkin Trans. I 1980, 1826. (b) Hansen, J. J.; Krogsgaard-Larsen, P. J. Chem. Soc. Chem. Comm. 1979, 87. (c) Krogsgaard-Larsen, P.; Christensen, S. B. Acta Chem. Scand. B 1976, 30, 281.
  9. (a) Bennouna, C.; Petrus, F.; Verducci, J. Bull. Soc. Chim. Fr. 1980, 478. (b) Jäger, V.; Frey, M. Liebigs Ann. Chem. 1982, 817. (c) Frey, M.; Jäger, V. Synthesis 1985, 478.
  10. Madsen, U.; Brehm, L.; Krogsgaard-Larsen, P. J. Chem. Soc. Perkin Trans. I 1988, 359.
  11. Frølund, B.; Kristiansen, U.; Brehm, L.; Hansen, A. B.; Krogsgaard-Larsen, P.; Falch, E. J. Med. Chem. 1995, 38, 3287.
  12. Begtrup, M.; Sløk, F. A. Synthesis 1993, 861.
  13. Nakamura, N.; Tajima, Y.; Sakai, S. Heterocycles 1982, 17, 235.
  14. Schlewer, G.; Krogsgaard-Larsen, P. Acta Chem. Scand. B 1984, 38, 815.
  15. Loewenthal, H. J. Q. in Protective Groups in Organic Chemistry; McOmie, J. F. W. Ed.; Plenum Press: New York, 1981.
  16. It is noteworthy that the identical reaction conditions applied to the N-benzyl product 6b furnished the open chain N-benzyl amide: Schön, M. Dissertation, Universität Stuttgart (planned).
  17. Anson, M. S.; Montana, J. G. Synlett 1994, 219.
  18. Winterfeldt, E. Synthesis 1975, 617.
  19. Deprotection using H2, Pd/BaSO4 furnished the appropriate benzyl alcohol in 98% yield: Schön, M. Dissertation, Universität Stuttgart (planned).
  20. Overman, L. E. J. Am. Chem. Soc. 1976, 98, 2901.