Total synthesis of ravidomycin: Revision of absolute and relative stereochemistry

Article abstract of DOI:10.1016/S0040-4039(99)02235-2

First total synthesis of ravidomycin has been achieved, thereby establishing the relative and absolute stereostructure as 1. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(99)02235-2

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 1063–1067
Total synthesis of ravidomycin: revision of absolute and relative
stereochemistry
Shin Futagami, Yoriko Ohashi, Koreaki Imura, Takamitsu Hosoya,† Ken Ohmori,
Takashi Matsumoto and Keisuke Suzuki ∗
Department of Chemistry, Tokyo Institute of Technology, O-okayama, Meguro-ku, Tokyo 152-8551, Japan
Received 12 November 1999; accepted 26 November 1999
Abstract
First total synthesis of ravidomycin has been achieved, thereby establishing the relative and absolute stereostruc-
ture as 1. © 2000 Elsevier Science Ltd. All rights reserved.
We describe herein the first total synthesis of ravidomycin, an amino sugar congener of the gilvocarcin
antibiotics with an enhanced antitumor activity.1a–d This synthesis has established the relative and
absolute structure of this important compound as 1, whereas the C(50) epimer 10 was originally proposed
without specifying the absolute stereochemistry (Fig. 1).1a
Fig. 1.
Scheme 1 shows a retrosynthesis, targeting I that is epimeric to the original proposal.2 With a notion
that some of the key steps used in our previous gilvocarcin synthesis,3 e.g., the benzyne chemistry (vide
infra), might be precluded by the presence of a dimethylamino function, we opted to introduce this group
as late as possible. Thus, the neutral sugar II was envisaged as the precursor, and further disconnection
by the [4+2] cycloaddition4 led to an early synthetic intermediate III, where the problem is the selective
access to the requisite β anomer. We could reasonably assume that the α and β anomers of III would
∗
†
Corresponding author.
Current address: Department of Biomolecular Science, Gifu University, Yanagido, Gifu 501-1193, Japan.
0040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(99)02235-2
tetl 16163

1064
adopt the conformations shown below, respectively, given the strong tendency of aryl C-glycosides to
dispose the aryl group equatorial. If so, there seemed to be no obvious preference between the anomers,
because the other substituents are two equatorials and two axials, respectively. Thus, any additional
factor(s) should be included that would encourage the formation of the β-anomer.
Scheme 1.
We envisaged the stereocontrol by exploiting the effect of large silyl protection that tends to favor the
1
axial disposition (cf. the gauche interaction at the equatorial position; see Fig. 2).5 Thus, the C4 (L)
conformation would be encouraged at the stage of the glycosyl donor A,6 and if the derived oxonium
species underwent the equatorial C-glycoside formation, the desired β anomer would be obtained. Even
if a Lewis acid might induce an α/β equilibration through B, the α anomer would be unfavorable in such
a case due to the gauche interaction (Fig. 3).5a
Fig. 2.
Fig. 3.

1065
Along these lines, the silyl protected donor 67 was prepared and its C-glycosidation was attempted
(Scheme 2). With the absolute stereochemistry unknown, we deliberately chose an L-series sugar as 6,
which was prepared from the known lactone 2.8 A t-butyldiphenylsilyl group was introduced at C(30)
1
as the stereocontrolling factor. Indeed, the glycosyl donor 6 adopted the C4 (L) conformation, and,
pleasingly, the Hf-promoted reaction of 6 with the iodo phenol 7 gave β-8 as the sole product in 83%
yield.6 Neither the α anomer nor the other conformers were detected.
Scheme 2. Keys: (a) (i) piperidine, 60°C, 45 min; (ii) BnBr, NaH/THF, 20 h (two steps, 69%); (b) (i) 2 M H2SO4/MeOH, 5 h;
(ii) TsOH·H2O/benzene, 10 h (two steps, 91%); (c) (i) DIBAL/toluene, −78°C, 15 min; (ii) Ac2O, DMAP/pyr, 1 h (two steps,
91%); (d) HF·(pyr)n/CH2Cl2, 0°C, 15 min, 82%; (e) 0.2 M NaOH/MeOH, 0.5 h, 96%; (f) TBDPSCl, NaH/THF, 10 h, 97%; (g)
7, Cp2HfCl2, AgClO4/CH2Cl2, MS 4A, −78→−10°C, 1 h, 83%
With the aryl C-glycoside β-8 in hand, the remaining tasks were the construction of the aromatic
moiety and the introduction of the amino group at C(30). The phenol 8 was converted to the triflate
9 (Tf2O, i-Pr2NEt/CH2Cl2, −78°C, 20 min, 88%), which was treated with n-BuLi in the presence of
2-methoxyfuran (10), where the benzyne-furan cycloaddition4 proceeded in a regioselective manner to
give, after protection of the phenol by an MOM group, the naphthyl glycoside 11 (Scheme 3). Installation
of the azido group with inversion at C(30) was achieved by a three-step protocol to give the azide 12:
desilylation, sulfonylation with N,N0-diimidazolylsulfonate,9 and the reaction with NaN3. Reduction of
the azide with LiAlH4 gave the corresponding prim-amine, which was converted to the dimethylamine
13 by reductive dimethylation.
After removal of the MOM protection in 13, the phenol was acylated with the benzoic acid 14 to give
the ester 15 ready for the biaryl bond formation. However, the attempted Pd-catalyzed reaction of 15 gave
a disappointing result under the conditions used in our previous gilvocarcin synthesis.3,10 The major issue
was the catalyst deactivation, presumably due to the Me2N-function, and related trials were all unfruitful.
A breakthrough was provided by the Harayama conditions11 using stoichiometric Pd(OAc)2 coupled
with Bu3P and DPPP [1,3-bis(diphenylphosphino)propane] in the presence of Ag2CO3, thereby quickly
furnishing the tetracycle 16 in 70% yield. Catalytic hydrogenolysis under carefully controlled conditions
enabled the selective and stepwise removal of two of the three benzyl groups, i.e., at the phenol (<5 min)
and at the C(40) OH (30 min), and subsequent acetylation and deprotection of the C(20) OH gave the
diacetate 17. The MOM group in 17 was cleanly detached with TMSBr to give the corresponding diol,
which was selectively mesylated to give the mesylate 18. Finally, treatment of 18 with DBN effected: (1)
the removal of the acetate protection for the phenol; and (2) the elimination of a CH3SO3H, to give the
final product 19.
All the spectroscopic data of 19 were identical with those of the natural product, thereby affirming the
C(50) stereochemistry of the sugar portion (cf. the original proposal1a). Furthermore, upon acetylation
31
D
of 19, the sign of [α] of the triacetate 20, [α] −36 (c 0.045, CHCl3), was opposite to that of
D

1066
Scheme 3. Keys: (a) (i) n-BuLi, 10/THF, −78°C, 0.5 h; (ii) MOMCl, NaH/DMF, 1 h (two steps, 77%); (b) CsF/DMF, 100°C, 3
h, 90%; (c) (Imd)2SO2, NaH/DMF, −40→0°C, 0.5 h, 93%; (d) NaN3, BnMe3NCl/DMF, 60°C, 3 h, 57%; (e) (i) LiAlH4/THF,
0°C, 0.5 h; (ii) aq. HCHO, NaBH3(CN)/CH3CN, 0.5 h (two steps, 91%); (f) (i) 1 M HCl/acetone, reflux, 3 h; (ii) 14, EDCI,
DMAP/CH2Cl2, 2 h (two steps, 98%); (g) Pd(OAc)2, DPPP, Bu3P, Ag2CO3/DMF, 130°C, 15 min, 70%; (h) (i) H2, Pd black,
2 M HCl/MeOH, 0.5 h; (ii) Ac2O, DMAP/pyr, 45 min (two steps, 76%); (i) H2, Pd black, 2 M HCl/MeOH, 3 h, 86%; (j)
TMSBr/CH2Cl2, −78→−15°C, 45 min, 59%; (k) MsCl /pyr, −20°C, 0.5 h, quant.; (l) DBN/CH3CN, 80°C, 0.5 h, 64%
26
the corresponding naturally derived triacetate, lit.1b [α] +33 (c 0.072, CHCl3). Thus, the absolute
D
stereochemistry of ravidomycin was proven to be the enantiomer of 19, i.e., 1.
Acknowledgements
1
We thank Prof. J. A. Findlay, University of New Brunswick, for the copies of the H NMR and IR
spectra of ravidomycin, and also to Dr. Y. Lefebvre, Wyeth-Ayerst, for providing an authentic specimen of
the natural product. We thank Prof. Harayama, Okayama University, for the kind information on the Pd-
promoted coupling. This work was financially supported by the Ministry of Education, Science, Culture
and Sports of Japanese Government [grant-in-aid for Scientific Research on Priority Area (08245103)].

1067
References
1. (a) Findlay, J. A.; Lin, J.-S.; Radics, L.; Rachit, S. Can. J. Chem. 1981, 59, 3018; (b) Findlay, J. A.; Lin, J.-S.; Radics, L.
ibid. 1983, 61, 323; (c) Sehgel, S. N.; Czerkawski, H.; Kudelski, A.; Pander, K.; Saucier, R.; Vezina, C. J. Antibiot. 1983, 36,
355; (d) Narita, T.; Matsumoto, M.; Mogi, K.; Kukita, K.; Kawahara, R.; Nakashima, T. ibid. 1989, 42, 347. For isolation of
related substances, see: Yamashita, N.; Shin-ya, K.; Furihata, K.; Hayakawa, Y.; Seto, H. ibid. 1998, 51, 1105.
2. An NOE experiment gave a clue to the correct stereochemistry, where the NOE observed between C(30) and C(50) proton
was conclusive. Also, personal communication from Dr. Y. Saito, Kyowa Hakko Co., is acknowledged.
3. (a) Hosoya, T.; Takashiro, E.; Matsumoto, T.; Suzuki, K. J. Am. Chem. Soc. 1994, 116, 1004; (b) Matsumoto, T.; Hosoya,
T.; Suzuki, K. ibid. 1992, 114, 3568.
4. Matsumoto, T.; Hosoya, T.; Katsuki, M.; Suzuki, K. Tetrahedron Lett. 1991, 32, 6735.
5. (a) Hosoya, T.; Ohashi, Y.; Matsumoto, T.; Suzuki, K. Tetrahedron Lett. 1996, 37, 663; (b) Eliel, E. L.; Satici, H. J. Org.
Chem. 1994, 59, 688.
1
6. H NMR showed that each of the compounds, α-6, β-6, and 8, adopts mostly the 1C4 conformation.
7. All compounds were fully characterized by spectroscopic means.
8. Ireland, R. E.; Wilcox, C. S. J. Org. Chem. 1980, 45, 197.
9. Hanessian, S.; Vatele, J.-M. Tetrahedron Lett. 1981, 22, 3579.
10. Deshpande, P. P.; Martin, O. R. Tetrahedron Lett. 1990, 31, 6313.
11. Harayama, T.; Akiyama, T.; Nakano, A. Chem. Pharm. Bull. 1997, 45, 1723.