HETEROCYCLES

Communication | Regular issue | Vol. 83, No. 2, 2011, pp. 293-298
Received, 19th November, 2010, Accepted, 20th December, 2010, Published online, 28th December, 2010.
DOI: 10.3987/COM-10-12106

■ Synthetic Studies of Lepranthin, a Lichen-Produced Dimeric Macrolide. Stereoselective Synthesis of a Seco-Acid Based on Stereospecific Epoxide-Opening Reactions

Hisashi Takada, Shinji Nagumo,* Eiko Yasui, Megumi Mizukami, and Masaaki Miyashita*

Department of Applied Chemistry, Faculty of Engineering, Kogakuin University, Nakano 2665-1, Hachioji, Tokyo, 192-0015, Japan

Abstract

The stereoselective synthesis of a seco-acid derivative of lepranthin (1), a lichen-produced unique 16-membered dimeric macrolide, is described wherein all asymmetric carbon centers were constructed in a highly stereoselective manner, respectively, by using different epoxide-opening reactions of the α,β-unsaturated γ,δ-epoxy ester system and an epoxy alcohol derivative as the key steps.

Bacteria, fungi and algae produce a large number of macrolides which are classified as polyketide-derived macrolides in their biosynthetic pathways. These compounds often provided us with good opportunities discovering new drugs. Interestingly, a few macrolides have been isolated from lichens too, which may imply a symbiotic relationship between fungi and algae.1 Lepranthin (1) was isolated from the crustaceous lichen Arthonia impolita (Ehrh.) Borrer by Zopf in 1904.2 Nearly century later, a NMR investigation and X-ray analysis by Huneck et al. revealed 1 to be a 16-membered homo-macrodiolide which contains two secondary hydroxyl groups and four secondary acetates.3 Although biological properties and synthetic studies of 1 have not been reported so far, its unique diolide structure would attract attention of synthetic chemists. We report herein the stereoselective synthesis of a seco-acid derivative 28, the key monomer segment of 1, based on stereospecific epoxide-ring opening strategies.

Our synthesis started with allyl alcohol 2 which was prepared from commercially available methyl (R)-3-hydroxybutylate in five steps.4 First, 2 was converted to α,β-unsaturated γ,δ-epoxy ester 4 by a two-step reaction sequence: (1) Katsuki-Sharpless epoxidation5 with L-(+)-DET, Ti(OiPr)4, and TBHP in CH2Cl2 at -30 oC, leading to epoxy alcohol 36 (87%); (2) Dess-Martin oxidation7 followed by Wittig olefination (91% yield). Reductive cleavage of the epoxide 4 with HCO2H and Pd2(dba)3･CHCl38 smoothly occurred to give alcohol 5 in 85% yield, which was then transformed into allyl alcohol 7 through the sequence of protection of the secondary alcohol with a silyl group and subsequent DIBAH reduction. The allyl alcohol 7 thus obtained was again transformed into α,β-unsaturated γ,δ-epoxy ester 9 by the same reaction sequence as that for 2: (1) Katsuki-Sharpless epoxidation leading to 86 (86%); (2) Dess-Martin oxidation; (3) a Wittig olefination (75%, two steps). After removal of the TES group in 9 with DDQ,9 treatment of the resulting epoxy alcohol 10 with Me3Al-H2O in CH2Cl2 at -50 oC afforded the desired product 11 in 70% yield.10 Protection of syn-1,3-diol 11 with a benzylidene acetal group furnished 1211 in high yield, which was further converted to epoxy alcohol 1412 in two steps: (1) reduction with DIBAH in THF (97%); (2) Katsuki-Sharpless epoxidation with L-(+)-DET, Ti(OiPr)4, and TBHP in CH2Cl2 at -30 oC (90%). Upon treatment of 14 with Me2CuCNLi213 in Et2O at -50 to -30 oC, the regioselective methyl substitution reaction smoothly occurred to give 15 as a single product in 89% yield. Thus, the requisite five stereogenic centers in the targeted molecule were stereoselectively constructed by using different epoxide-opening reactions of the two γ,δ-epoxy unsaturated esters, 4 and 10, and the epoxy alcohol 14.

The remaining task for the synthesis of a seco-acid was discrimination of the five hydroxyl groups in 15. To this end, sequential oxidations of 15 with TEMPO14 and then with NaClO215 followed by esterification with CH2N2 produced ester 17 in 67% overall yield. Next, the hydroxyl group in 17 was protected with a MOM group by treatment with MOMCl, DIPEA, and NaI in 1,2-DME, giving rise to 18 in 85% yield. Among discrimination of the five hydroxyl groups, the most difficult task was that between C5 and C7 hydroxyl groups protected by benzylidene acetal. All attempts aiming at a regioselective reductive cleavage of the benzylidene acetal moiety in 18 failed unfortunately. Eventually, distinction between these hydroxyl groups was performed as follows. Removal of the silyl group in 18 with TBAF/AcOH in DMF (90%) followed by treatment of the resulting alcohol with BzCl and pyridine in CH2Cl2 furnished 20 (96%), which was converted to diol 21 by catalytic hydrogenolysis with PtO2 in EtOH. Unexpectedly, the benzene ring in 20 was smoothly hydrogenated concomitantly to produce 21. Further treatment of 21 with PPTS16 in refluxing 1,2-dichloroethane in the presence of pyridine resulted in facile lactonization to give lactone 22, whose hydroxyl group was then protected with ethyl vinyl ether and PPTS16 in CH2Cl2 to afford ethoxyethyl ether 23 quantitatively. Unfortunately, however, subsequent hydrolysis of 23 under alkaline conditions underwent elimination of the MOM group to give unsaturated lactone exclusively. To overcome this difficulty, the lactone 23 was reduced with LiAlH4 in THF and subsequent regioselective protection of the primary alcohol with a TBDPS group produced diol 25 (87%). After protection of the diol with acetyl groups (84%), removal of the TBDPS group with TBAF in THF gave the primary alcohol (90%), which was successfully converted to seco-acid 2817 for the total synthesis of lepranthin (1), in two steps: (1) TEMPO oxidation; (2) sodium chlorite oxidation (82%).

In summary, we completed the asymmetric synthesis of the seco-acid 28, the key monomer of lepranthin (1), based on stereospecific epoxide-opening reactions including the stereospecific methylation reaction of the γ,δ-epoxy unsaturated ester 10 with Me3Al-H2O system. Further studies of the crucial macrolactonization of the seco-acid 28 toward total synthesis of lepranthin (1) are now in progress in our laboratory.

ACKNOWLEDGEMENTS
Financial support from the Ministry of Education, Culture, Sports, Science and Technology, Japan (a Grant-in-Aid for Scientific Research (B) (No. 19350027) and Advanced Promotion Research Program for Education of Graduate School) is gratefully acknowledged.

References

1. S. Omura, Macrolide Antibiotics; Academic Press: 2002.
2. W. Zopf, Liebigs Ann. Chem., 1904, 336, 46. CrossRef
3. K. Polborn, W. Steglich, J. D. Connolly, and S. Z. Huneck, Naturforsch. B: Chem. Sci., 1995, 50, 1111.
4. R. M. Garbaccio, S. J. Stachel, D. K. Baeschlin, and S. J. Danishefsky, J. Am. Chem. Soc., 2001, 123, 10903. CrossRef
5. T. Katsuki and K. B. Sharpless, J. Am. Chem. Soc., 1980, 102, 597. CrossRef
6. Diastereo ratio of compounds 3 and 8 was determined by 1H NMR spectrum.
7. D. B. Dess and J. C. Martin, J. Org. Chem., 1983, 48, 4155. CrossRef
8. M. Oshima, H. Yamazaki, I. Shimizu, M. Nisar, and J. Tsuji, J. Am. Chem. Soc., 1989, 111, 6280. CrossRef
9. K. Tanemura, T. Suzuki, and T. Horaguchi, J. Chem. Soc., Perkin Trans. 1, 1992, 2997. CrossRef
10. M. Miyashita, M. Hoshino, and A. Yoshikoshi, J. Org. Chem., 1991, 56, 6483. CrossRef
11. Benzyl proton peak in 1H NMR spectrum of compound 12 showed the NOE correlations with peaks at the C5 and C7 positions (each ca. 10%).
12. The peaks for a diastereomer were not observed in the 1H NMR spectrum of compound 14.
13. B. H. Lipshutz, J. Kozlowski, and R. S. Wilhelm, J. Am. Chem. Soc., 1982, 104, 2305. CrossRef
14. A. D. Mico, R. Margarita, L. Parlanti, A. Vescovi, and G. Piancatelli, J. Org. Chem., 1997, 62, 6974. CrossRef
15. (a) B. O. Lindgren and T. Nilsson, Acta Chem. Scand., 1973, 27, 888; CrossRef (b) G. A. Kraus and M. Taschner, J. Org. Chem., 1980, 45, 1175; CrossRef (c) G. A. Kraus and B. Roth, J. Org. Chem., 1980, 45, 4825; CrossRef (d) B. S. Bal, W. E. Childers, Jr., and H. W. Pinnick, Tetrahedron, 1981, 37, 2091. CrossRef
16. M. Miyashita, A. Yoshikoshi, and P. A. Grieco, J. Org. Chem., 1977, 42, 3772. CrossRef
17. Seco-acid 28 was obtained as a diastereomeric mixture (ca. 0.55 : 0.45) concerning the ethoxyethyl group. [α]D27 -19.21 (c 1.38, CHCl3); FAB-MS (POSI) m/z 465 (MH+), 433, 386; HR-FABMS m/z 465.2722 (calcd for C22H41O10: 465.2700); IR (ATR) 2978, 1732, 1456 cm-1; 1H NMR (400 MHz, CDCl3) δ 5.16-5.11 (1H, m), 5.09-5.00 (1H, m), 4.72 (1H×0.55, q, J = 5.6 Hz), 4.69-4.65 (2H, m), 4.60 (1H×0.45, q, J = 5.2 Hz), 3.70-3.43 (5H, m), 3.39 (3H, s), 2.98-2.90 (1H, m), 2.15-1.98 (2H, m), 2.06 (3H×0.55, s), 2.05 (3H×0.45, s), 2.02 (3H×0.45, s), 2.01 (3H×0.55, s), 1.95-1.73 (3H, m), 1.60-1.49 (1H, m), 1.284 (3H×0.45, d, J = 5.2 Hz), 1.283 (3H×0.55, d, J = 5.6 Hz), 1.26-1.22 (6 H, m), 1.21 (3H×0.45, t, J = 6.8 Hz), 1.16 (3H×0.55, t, J = 7.2 Hz), 0.97 (3H×0.55, d, J = 7.6 Hz), 0.96 (3H×0.45, d, J = 7.2 Hz); 13C NMR (100 MHz, CDCl3) δ 179.39 (C), 179.14 (C), 170.69 (C), 170.67 (C), 170.22 (C), 170.19 (C), 101.23 (CH), 98.26 (CH2), 98.23 (CH2), 97.55 (CH), 83.25 (CH), 83.10 (CH), 72.25 (CH), 71.27 (CH), 71.07 (CH), 68.84 (CH), 68.32 (CH), 67.84 (CH), 60.49 (CH2), 60.07 (CH2), 56.28 (CH3), 56.26 (CH3), 43.05 (CH), 43.01 (CH), 41.33 (CH2), 41.05 (CH2), 38.51 (CH for both isomers), 35.83 (CH2), 34.42 (CH2), 21.27 (CH3), 21.25 (CH3), 21.13 (CH3), 21.10 (CH3), 20.75 (CH3), 20.63 (CH3), 20.36 (CH3), 20.34 (CH3), 15.27 (CH3), 15.25 (CH3), 12.83 (CH3), 12.77 (CH3), 11.36 (CH3), 11.22 (CH3).