Glycosylation study on pradimicin-benanomicin antibiotics

Article abstract of DOI:10.1016/S0040-4039(00)01153-9

In relation to the synthesis of the pradimicin-benanomicin antibiotics, a model study was carried out for introducing the disaccharide moiety. The glycosyl donor 2, prepared from D-glucose and D-xylose, was allowed to react with various glycosyl acceptors in the presence of Cp₂HfCl₂ and AgClO₄. Also studied was the reactivity difference between the conformers of the trans-phenanthrendiol derivatives 12, 13 and 14. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)01153-9

Tetrahedron Letters 41 (2000) 6827±6832
Glycosylation study on pradimicin±benanomicin antibiotics
Hirohisa Kato, Ken Ohmori and Keisuke Suzuki*
Department of Chemistry, Tokyo Institute of Technology, and CREST, Japan Science and Technology (JST),
O-okayama, Meguro-ku, Tokyo 152-8551, Japan
Received 31 May 2000; revised 3 July 2000; accepted 7 July 2000
Abstract
In relation to the synthesis of the pradimicin±benanomicin antibiotics, a model study was carried out for
introducing the disaccharide moiety. The glycosyl donor 2, prepared from d-glucose and d-xylose, was
allowed to react with various glycosyl acceptors in the presence of Cp2HfCl2 and AgClO4. Also studied was
the reactivity di€erence between the conformers of the trans-phenanthrendiol derivatives 12, 13 and 14.
# 2000 Elsevier Science Ltd. All rights reserved.
Our synthetic study on the pradimicin±benanomicin antibiotics (e.g. 1) is currently focused on
two issues for introducing a sugar moiety to the C(5)-hydroxyl of the pentacyclic core.1,2
(i) Design of a suitable disaccharide donor.
(ii) E€ect of the conformation of the glycosyl acceptor (vide infra).
Described herein is the outcome of our study on these problems. Concerning problem (i), the
glycosyl ¯uoride 2 was prepared, and its reactivity as a glycosyl donor was studied. Problem (ii)
dealt with the reactivity di€erence, if any, between the conformers of the glycosyl acceptor, for
which the conformationally-restricted chiral phenanthrendiol 12 and its derivatives were used as
the model compounds.
* Corresponding author.
0040-4039/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)01153-9

6828
1. Preparation of 2 and its glycosidation study
First, the amino sugar 10 (a d-tomosamine derivative) was synthesized. By the modi®ed
procedure of Stevens,3 the known diol 34 was methanesulfonylated followed by selective
displacement of the prim-mesylate by iodide to give 5, and subsequent hydrogenolysis in the
presence of NaOAc a€orded the 6-deoxy sugar 6. Treatment of 6 with NaN3 furnished the azide
7, which was converted to the glycosyl ¯uoride 9 by acetolysis followed by treatment with
.
(HF)n pyr and saponi®cation. Benzoylation of 9 allowed the regioselective protection at 2-OH,
giving a mixture of mono-benzoates with predominance of the 2-O-benzoyl derivative (2-O-Bz:3-
O-Bz=ca. 8:1). A single recrystallization (EtOAc) gave a pure sample of 10 (mp 124±126ꢀC) in 50%
yield (Scheme 1).
Scheme 1. Reagents and conditions: (a) MsCl/pyridine, 15 h (84%); (b) NaI/2-butanone, re¯ux, 6 h (85%); (c) H2,
10% Pd±C, NaOAc/EtOH, EtOAc, 15 h (95%);ꢀ (d) NaN3, n-Bu4NCl/DMF, 110ꢀC, 10 h (87%); (e) Ac2O, conc.
ꢀ
.
H2SO4, 4 h (91%); (f) (HF)n pyridine/CH2Cl2, ^20 C!rt, 8 h; (g) K2CO3/MeOH, 0 C, 1 h (two steps, 80%); (h) BzCl/
pyridine, 0ꢀC!rt, 15 h (50%)
ꢁ1†
By exploiting the `orthogonal glycosylation' by Kanie et al.,5 the ¯uoride 10 was glycosylated
with the xylosyl sul®de 116 by using NIS and a catalytic amount of TfOH (10 mol%)7 in CH2Cl2
to give the disaccharide 2 in 93% yield (Eq. (1)). The a:b ratio was 1:5, where the requisite
b-glycoside was isolated by recrystallization (EtOH, mp 176±178ꢀC).
With the disaccharide donor 2 in hand, the glycosidation was tested with benzyl alcohol as the
glycosyl acceptor by using various promoters, and the results are summarized in Table 1. The
reaction with BF3 OEt2 or TMSOTf8b resulted in complete recovery of 2 (runs 1, 2), which
8a
.
could be ascribed to the deactivation of the promoter by the Lewis-basic functions in 2. The
8c
glycosidation was possible with SnCl2±AgClO4 (run 3), albeit in low yield. Fortunately, the
reaction proceeded nicely by employing the zirconocene- or hafnocene-based promoter (runs 4,
5).9 For example, when the ¯uoride 2 was allowed to react with benzyl alcohol in the presence of
Cp2HfCl2 and AgClO4, the glycosylation reaction occurred smoothly (^78 to ^20ꢀC), thereby
giving the corresponding b-benzyl glycoside in 95% yield.

6829
Table 1
Glycosidation of 2 with benzyl alcohol by using various promoters
Under the hafnocene-promoted conditions, thus found, the glycosidation of 2 with various
acceptors was tested, where all the reactions listed in Table 2 were complete within 30 min at
^20ꢀC.10 A sec-alcohol as well as a tert-alcohol can be smoothly glycosylated in high yields (runs
1, 2). The stereoselectivity showed an interesting dependence on the acceptors. A benzyl alcohol
with a methoxycarbonyl group was glycosylated with good b-selectivity (run 3), whereas a multi-
functionalized orsellinate derivative showed rather poor selectivity (run 4). Interestingly, phenol
showed a complete a-selectivity (run 5).11 It is notable that the reaction stopped at the O-glycosylated
stage in this case, whereas such reaction conditions usually facilitates the O!C glycoside
rearrangement to give an aryl C-glycoside,12 which could again be ascribed to the diminution of
the Lewis acidity by many Lewis basic functionalities.
Table 2
Glycosidation of 2 with various acceptorsa)

6830
2. Reactivities of the trans-phenanthrendiol derivatives
With these data in hand, we turned our attention to the conformer dependence of the diol
acceptor. In order to address this issue, the enantiomerically pure phenanthrendiol (S,S)-1213 was
chosen as the model substrate, because the interconversion of the conformers, 12eq and 12ax, is
slow enough at ambient temperature, thereby allowing us to examine the reactivity of the
conformers as separate entities.14,15 In addition to 12, its mono-protected derivatives, 13 and 14,
were tested for this study.
Table 3 shows the results of the glycosylation of these compounds with 2 in the presence of
Cp2HfCl2 and AgClO4. It turned out that the diols 12eq and 12ax gave only poor results (runs 1,
2). The diequatorial diol 12eq showed a poor reactivity, presumably due to the internal hydrogen
bonding that makes the hydroxy groups less reactive. On the other hand, the diaxial counterpart
12ax showed higher reactivity, giving the mono-glycosylated product along with the bis-glycosylated
product. Interestingly, all the anomeric centers were b in the mono- as well as the bis-glycosylated
product.
Table 3
Glycosylation of various phenanthrendiol derivatives with 2a)
We then examined the glycosylation of mono-protected derivatives, i.e. the acetates 13eq and
13ax, and the benzyl ethers 14eq and 14ax. Although these compounds were fairly reactive, the
stereoselectivity depended on the protecting group. The stereoselectivity was poor for the acetates
13eq and 13ax (runs 3, 4), whereas better for the benzyl ethers 14eq and 14ax (runs 5, 6). It should
be noted that the diaxial conformer consistently showed higher b-selectivity than the diequatorial
counterparts.
In summary, we have described a model study for the introduction of a disaccharide moiety in
the pradimicin±benanomicin antibiotics, which would hopefully contribute to the completion of
the total synthesis. Further work is now in progress in our laboratories.

6831
Acknowledgements
We thank Professor Kazuhiko Saigo, the University of Tokyo, for helpful discussion and Ms.
Miou Kobayashi for her early contribution. Technical support from the Daicel Corporation for
the optical resolution of 12 is gratefully acknowledged.
References
1. For synthetic studies on this class of natural products, see: Kitamura, M.; Ohmori, K.; Kawase, T.; Suzuki, K.
Angew. Chem., Int. Ed. Engl. 1999, 38, 1229±1232, and references cited therein.
2. For isolation and structure elucidation, see: (a) Oki, T.; Konishi, M.; Tomatsu, K.; Tomita, K.; Saitoh, K.;
Tsunakawa, M.; Nishio, M.; Miyaki, T.; Kawaguchi, H. J. Antibiot. 1988, 41, 1701±1704; (b) Takeuchi, T.; Hara,
T.; Naganawa, H.; Okada, M.; Hamada, M.; Umezawa, H.; Gomi, S.; Sezaki, M.; Kondo, S. J. Antibiot. 1988, 41,
807±811.
3. Stevens, C. L.; Blumbergs, P.; Otterbach, D. H.; Strominger, J. L.; Matsuhashi, M.; Dietzler, D. N. J. Am. Chem.
Soc. 1964, 86, 2937±2938.
4. Richtmyer, N. K. Methods in Carbohydr. Chem. 1976, 2, 107±113.
5. Kanie, O.; Ito, Y.; Ogawa, T. J. Am. Chem. Soc. 1994, 116, 12073±12074.
6. Ferrier, R. J.; Furneaux, R. H. Carbohydr. Res. 1976, 52, 63±68.
7. (a) Veeneman, G. H.; van Leeuwen, S. H.; van Boom, J. H. Tetrahedron Lett. 1990, 31, 1331±1334; (b)
Konradsson, P.; Udodong, U. E.; Fraser-Reid, B. Tetrahedron Lett. 1990, 31, 4313±4316.
8. (a) Nicolaou, K. C.; Chucholowski, A.; Dolle, R. E.; Randall, J. L. J. Chem. Soc., Chem. Commun. 1984, 1155±
1156; (b) Hashimoto, S.; Hayashi, M.; Noyori, R. Tetrahedron Lett. 1984, 25, 1379±1382; (c) Mukaiyama, T.;
Murai, Y.; Shoda, S. Chem. Lett. 1981, 431±432.
9. (a) Suzuki, K.; Maeta, H.; Matsumoto, T.; Tsuchihashi, G. Tetrahedron Lett. 1988, 29, 3571±3574; (b) Suzuki, K.;
Maeta, H.; Matsumoto, T. Tetrahedron Lett. 1989, 30, 4853±4856.
10. In the presence of Cp2HfCl2±AgClO4, the b-xylosyl glycoside undergoes anomerization in the reactions at higher
temperature, typically 0ꢀC or above.
11. This result could be rationalized as follows: The b-phenyl glycoside, formed if any, would be more reactive,
thereby undergoing a neighboring group-assisted departure of the phenoxide, possibly in a reversible manner, and
the less reactive a-glycoside was accumulated.
12. (a) Matsumoto, T.; Katsuki, M.; Suzuki, K. Tetrahedron Lett. 1988, 29, 6935±6938; (b) Matsumoto, T.; Hosoya,
T.; Suzuki, K. Tetrahedron Lett. 1990, 31, 4629±4632.
13. Enantiomerically pure (S,S)-12eq14 was prepared by the SmI2-promoted pinacol coupling of the corresponding
dialdehyde (þ)-1516,17 followed by the enantiomer separation by using preparative chiral HPLC column
chromatography (Daicel Chiralcel OJ, 20 mmÂ250 mm, eluted with hexanes:i-PrOH=90:10, ¯ow rate 2.5 mL/
min). The CD spectrum showed that the latter fraction was the (S,S)-enantiomer from the literature data.15 The
corresponding diaxial conformer, (S,S)-12ax, was obtained via the following steps: (1) Ac2O/pyridine; (2) Á/
toluene; (3) K2CO3/MeOH. Partial protection of the hydroxy groups in 12 a€orded the mono-protected
compounds 13 and 14.

6832
14. Kato, H.; Ohmori, K.; Suzuki, K. Chirality, 2000, 12, 548±550.
15. Armstrong, R. N.; Lewis, D. A. J. Org. Chem. 1985, 50, 907±908; (b) Armstrong, R. N.; Lewis, D. A.; Ammon,
H. L.; Prasad, S. M. J. Am. Chem. Soc. 1985, 107, 1057±1058.
16. Ohmori, K.; Kitamura, M.; Suzuki, K. Angew. Chem., Int. Ed. Engl. 1999, 38, 1226±1229.
17. Karnes, H. A.; Kybett, B. D.; Wilson, M. H.; Margrave, J. L.; Newman, M. S. J. Am. Chem. Soc. 1965, 87, 5554±
5558.