Studies toward the Total Synthesis of Wiedemannic Acid - Organic

Effiette L. O. Sauer, James Hooper, Tom Woo, and Louis Barriault. Journal of ... Matthew D. Reeder, G. S. C. Srikanth, Spencer B. Jones, and Steven L...
0 downloads 0 Views 125KB Size
ORGANIC LETTERS

Studies toward the Total Synthesis of Wiedemannic Acid

2004 Vol. 6, No. 19 3329-3332

Effiette L. O. Sauer and Louis Barriault* Department of Chemistry, 10 Marie Curie, UniVersity of Ottawa, Ottawa, Canada K1N 6N5 [email protected] Received June 29, 2004

ABSTRACT

We report a novel and efficient diastereoselective synthesis of wiedemannic acid analogue 30 in 16 steps from 7 using a tandem oxy-Cope/ Claisen/ene reaction as the key step. Comparison of NMR data between wiedemannic acid (1) and analogue 30 leads us to believe that the reported stereochemistry at the ring junction of 1 is incorrect.

SalVia wiedemannii is a plant native to central Turkey (Haymana) and is the source of numerous diterpenes.1 In 1990, Ulubelen and Topcu isolated from the aerial parts of this plant a new abietane diterpene (2), wiedemannic acid (1) (Figure 1).2 A cursory glance of this natural product reveals that the major synthetic challenge of wiedemannic acid (1) resides in the stereochemical control of the five contiguous stereogenic centers at C4, C5, C8, C9, and C10, of which three carbons are quaternary centers (C4, C9, and C10).3 The retrosynthetic analysis summarized in Scheme 1 shows that the quaternary carbon center at C4 could be installed from enone 3. Tricyclic 3 could be obtained from a ringclosing metathesis reaction of 4 followed by an allylic alcohol oxidation. (1) Ulubelen, A.; Topcu, G.; Terem, B. Phytochemistry 1987, 26, 1534 and references therein. (2) Ulubelen, A.; Topcu, G. Phytochemistry 1990, 29, 2346. (3) For reviews on the synthesis of quaternary carbon centers, see: (a) Barriault, L.; Denissova, I. Tetrahedron 2003, 59, 10105. (b) Christoffers, J.; Baro, A. Angew. Chem., Int. Ed. 2003, 42, 1688. (c) Christoffers, J.; Mann, A. Angew. Chem., Int. Ed. 2001, 40, 4591. (d) Corey, E. J.; GuzmanPerez, A. Angew. Chem., Int. Ed. 1998, 110, 402. (e) Fuji, K. Chem. ReV. 1993, 93, 2037. (f) Martin, S. F. Tetrahedron 1980, 36, 419. 10.1021/ol0487635 CCC: $27.50 Published on Web 08/24/2004

© 2004 American Chemical Society

Figure 1. Structure of wiedamannic acid.

Decalin 4, bearing two contiguous quaternary centers at C9 and C10, could in turn be obtained from a tandem oxyCope/Claisen/ene reaction of 1,2-vinylcyclohexanol allyl ether 5.4 Finally, the tandem precursor 5 can be readily prepared from isopulgenone derivative 6. To develop a highly diastereoselective synthetic route to create the A and B rings of wiedemannic acid (1), a model study was envisaged where the isopropyl group occupying the equatorial position at C13 was replaced by an equatorial methyl at C12.

Scheme 1

The synthesis of our model study began by following our recently reported reaction sequence4 for the rapid conversion of isopulgenone 75 to intermediate 10 (Scheme 2). A

Scheme 2

lithium-halogen exchange of 8 in diethyl ether at -100 °C followed by the addition of 7 afforded the corresponding tertiary alcohol in 50% yield. The latter was treated with KH and allyl bromide in DME to provide allyl ether 9 in 95% yield. Allyl ether 9 was converted into Decalin 10 in 90% yield (dr > 25:1) by a microwave-assisted tandem oxyCope/Claisen/ene reaction.6 The tandem pericyclic reaction is initiated by an oxy-Cope rearrangement to generate in situ macrocycle 11, which spontaneously rearranges via a Claisen [3,3] shift reaction to the macrocyclic ketone 12 (Scheme 3). The latter is poised (4) Barriault, L.; Sauer, E. L. O. J. Am. Chem. Soc. 2004, 126, 8569. (5) Corey, E. J.; Ensley, H. E.; Suggs, J. W. J. Org. Chem. 1976, 41, 380. (6) For review on microwaves in organic synthesis, see: (a) Loupy, A.; Petit, A.; Hamelin, J.; Texier-Boullet, F.; Jacquault, P.; Mathe´, D. Synthesis 1998, 1213. (b) Majetich, G.; Hichs, R. J. J. MicrowaVe Power Electromagnetic Energy 1995, 30, 27. (c) Loupy, A.; Perreux, L. Tetrahedron 2001, 57, 9199. (d) Lidstro¨m, P.; Tierny, J.; Wathey, B.; Westman, J. Tetrahedron 2001, 57, 9225. 3330

Scheme 3

to cyclize via a transannular carbonyl ene reaction to give Decalin 10. Recently, we demonstrated that the stereochemical outcome at C5, C8, C9, and C10 is governed by the preferential conformation of macrocycles 11 and 12 at the transition state for the Claisen and ene reactions.4 Removal of the tert-butyldimethylsilyl ether group was achieved with TBAF in THF to give alcohol 13 in 91% yield. The latter was then oxidized with TPAP and NMO to give the corresponding aldehyde 14 in 95% yield.7 Installation of the vinyl moiety, required for the subsequent ring-closing metathesis reaction (RCM), was accomplished by the addition of vinylmagnesium bromide to 14 in THF. This alkylation, however, yielded an inseparable mixture of allylic alcohol 15 (Felkin-Anh product) and lactol 16 (epimer of 15 at C5) in 67% combined yield. This mixture was then treated with Grubbs’s catalyst8 to give alcohol 17, which was easily separated from lactol 16. To ensure high yields for the RCM step, however, the exclusive formation of alkylation product 15 would be required. Suspecting that the free tertiary alcohol at C9 in 14 was responsible for the epimerization at C5 and subsequent lactol formation, its conversion to a silyl ether was considered. Accordingly, diol 13 was treated with TMSCl and KHMDS in THF at -78 °C followed by a mild basic hydrolysis of the bis-silylated material in methanol to give the monoprotected alcohol 18. The crude material was then oxidized with TPAP and NMO in DCM to give aldehyde 19 in 88% yield for the three steps (Scheme 4). With the tertiary alcohol now protected, the addition of vinylmagnesium bromide to aldehyde 19 was attempted. Much to our delight, the alkylation now gave exclusively the Felkin-Ahn alkylation product 20 in 97% yield (dr > 25:1). RCM reaction of 20 with Grubbs’s catalyst gave the corresponding allylic alcohol 21, which was directly oxidized to give enone 22 in 91% yield for the two steps. The transformation of enone 22 to wiedemannic acid analogue 30 is summarized in Scheme 5. 1,2-Alkylation of 22 with methyllithium in THF at -78 °C gave alcohol 23 (98%), which then underwent a 2,3-rearrangement in the presence of PCC to afford 24 in 91% yield (Scheme 5). The (7) Ley, S. V.; Griffith, W. P. Aldrichim. Acta 1990, 23, 1. (8) (a) Scholl, S.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999, 1, 953. (b) Sanford, M. S.; Love, J. A.; Grubbs, R. H. J. Am. Chem. Soc. 2001, 123, 6543. Org. Lett., Vol. 6, No. 19, 2004

Scheme 4

Figure 2. ORTEP view of 31.

stereoselective 1,4-addition of vinylcuprate, readily generated from vinylmagnesium bromide and CuI in THF, provided ketone 25 in 80% yield as a single detectable isomer (dr > 25:1). With the final quaternary center now installed, our attention was turned to the deoxygenation of the carbonyl

Scheme 5

at C2. Subjecting 25 to Wolff-Kishner conditions effectively reduced the carbonyl and removed the tertiary silyl ether at C9 to give intermediate 26 in yields ranging from 8 to 78%. Despite considerable experimental efforts to improve the reliability of this reaction, however, the chemical yields remained highly variable, making its use in the synthesis less than viable. As an alternative means of reducing the carbonyl, a Barton-McCombie radical deoxygenation was considered. The preparation of 26 commenced with a carbonyl reduction by DIBAl-H in THF to provide the corresponding axial alcohol. The latter was treated with thiocarbonyldiimidazole and DMAP in dichloromethane to afford 27 in 84% yield for the two steps. Treatment of 27 in the presence of a catalytic amount of AIBN and Ph3SnH in refluxing benzene afforded the deoxygenated product 28 in 74% yield.9 Following removal of the silyl ether with TBAF in THF, the masked carbonyl and aldehyde moieties, at C7 and C4, respectively, were revealed by treatment with OsO4 and NMO followed by NaIO4 cleavage of the resulting diols to afford 29 in 83% yield. Finally, aldehyde 29 was transformed to the corresponding carboxylic acid 30 in 78% yield by a mild NaClO2 oxidation.10 The relative stereochemistry at C4, C5, C8, C9, and C10 of compounds 29 and 30 was established by COSY and NOESY 1H NMR experiments and later confirmed by singlecrystal X-ray analysis of the methyl ester derivative 31 (CH2N2, Et2O, 94%) (Figure 2). At this stage, spectroscopic data of 30 was compared to that of wiedamannic acid (1). Major differences were found in both the chemical shifts and coupling constants of their 1H and 13C NMR spectra.11 In particular, significant variation at C5, C6, and C8 was noticed (Table 1). On the assumption that the equatorial methyl at C12 should not influence the chemical shift of remote carbons and protons in the A and B rings (C1-C10, C19, and C20), the observed differences suggest that the proposed structure of wiedemannic acid (1) is incorrect. (9) Ogilvie, W. W.; Yoakim, C.; Doˆ, F.; Hache´, B.; Lagace´, L.; Naud, J.; O’Meara, J. A.; De´ziel, R. Bioorg. Med. Chem. 1999, 7, 1521. (10) Kraus, G. A.; Taschner, M. J. J. Org. Chem. 1980, 45, 1175. (11) See NMR spectra in Supporting Information.

Org. Lett., Vol. 6, No. 19, 2004

3331

Table 1. CDCl3

1H

and

13C

13C

NMR Data (δ, ppm) of 1 and 30 in

13C

atom

NMR 1 (ppm)

NMR 30 (ppm)

5 6

52.2 37.0

41.8 41.1

8

52.6

52.1

1H

1H

NMR 1 (ppm)

NMR 30 (ppm)

1.08a 2.27 axc 1.90 eqd 3.49g

2.92b 2.33 axe 2.09 eqf 2.34h

a Dd, 8 and 4 Hz. b Dd, 13.8 and 3.0 Hz. c Dd, 14 and 8 Hz. d Dd, 14 and 4 Hz. e Dd, 14.4 and 13.4. f Dd, 14.4 and 3.2 Hz. g Dd, 12 and 6 Hz. h Dd, 11.9 and 3.8 Hz.

In conclusion, we have reported a novel and efficient diastereoselective synthesis of wiedemannic acid analogue

3332

30 in 16 steps from 7. This synthesis took advantage of the tandem oxy-Cope/Claisen/ene reaction, recently developed in our laboratory, to construct a Decalin system having the proper stereochemistry at C5, C8, C9, and C10. Comparison of NMR data between 1 and 30 leads us to believe that the structure of 1 may have been misassigned. Acknowledgment. We thank NSERC, Merck-Frosst, Boehringer Ingelheim, Bristol Myers Squibb, AstraZeneca Canada, Canada Foundation for Innovation, Ontario Innovation Trust, and the University of Ottawa for generous funding. E.L.O. Sauer thanks NSERC for a PGS-A JuliePayette postgraduate scholarship. We thank Prof. Ulubelen for providing copies of 1H and 13C NMR spectra of 1. Supporting Information Available: Experimental procedures and spectroscopic data and copies of 1H and 13C NMR for compounds 1 (authentic sample) and 13-31. This material is available free of charge via the Internet at http://pubs.acs.org. OL0487635

Org. Lett., Vol. 6, No. 19, 2004