Studies on the synthesis of landomycin A: Synthesis and glycosidation reactions of L-rhodinosyl acetate derivatives

Article abstract of DOI:10.1021/jo015756a

An efficient, eight-step synthesis of L-rhodinosyl acetate derivative 3 is described. The synthesis originates from methyl (S)-lactate and involves a highly stereoselective, chelate-controlled addition of allyltributylstannane to the lactaldehyde derivative 7. The β-anomeric configuration of 3 was established with high selectivity by acetylation of the pyranose precursor with Ac²O and Et₃N in CH₂Cl₂. Preliminary studies of glycosidation reactions of 3 and L-rhodinosyl acetate 10 containing a 3-O-TES ether revealed that these compounds are highly reactive glycosidating agents and that trialkylsilyl triflates are effective glycosylation promoters. The best conditions for reactions with 15 as the accepter involved use of diethyl ether as the reaction solvent and 0.2 equiv of TES-OTf at -78°C. However, the TES ether protecting group of 10 proved to be too labile under these reaction conditions, and mixtures of 16a, 17, and 18a are obtained in reactions of 10 and 15. Disaccharide 17 arises via in situ cleavage of the TES ether of disaccharide 16a, while trisaccharide 18a results from a glycosidation of in situ generated 17 (or of 16a itself) with a second equivalent of 10. These problems were largely suppressed by using 3 with a 3-O-TBS ether protecting group as the glycosyl donor and 0.2 equiv of TES-OTf as the reaction promoter. Attempts to selectively glycosylate the C(3)-OH of diol acceptors 20 or 28 gave a 70:30 mixture of 21 and 22 in the reaction of 20 and a 43:27:30 mixture of regioisomeric trisaccharides 29 and 30 and tetrasaccharide 31 from the glycosidation reaction of 28. However, excellent results were obtained in the glycosidation of differentially protected disaccharide 34 using 1.5 equiv of 3 and 0.05 equiv of TBS-OTf in CH₂-Cl₂ at -78°C. The latter step is an important transformation in the recently reported synthesis of the landomycin A hexasaccharide unit.

Full text of DOI:10.1021/jo015756a

J . Org. Chem. 2001, 66, 6389-6393
6389
Stu d ies on th e Syn th esis of La n d om ycin A: Syn th esis a n d
Glycosid a tion Rea ction s of L-Rh od in osyl Aceta te Der iva tives
William R. Roush,* Chad E. Bennett,1 and Sara E. Roberts
Department of Chemistry, University of Michigan, Ann Arbor, Michigan 48109
roush@umich.edu
Received May 15, 2001
An efficient, eight-step synthesis of L-rhodinosyl acetate derivative 3 is described. The synthesis
originates from methyl (S)-lactate and involves a highly stereoselective, chelate-controlled addition
of allyltributylstannane to the lactaldehyde derivative 7. The â-anomeric configuration of 3 was
established with high selectivity by acetylation of the pyranose precursor with Ac2O and Et3N in
CH2Cl2. Preliminary studies of glycosidation reactions of 3 and L-rhodinosyl acetate 10 containing
a 3-O-TES ether revealed that these compounds are highly reactive glycosidating agents and that
trialkylsilyl triflates are effective glycosylation promoters. The best conditions for reactions with
15 as the acceptor involved use of diethyl ether as the reaction solvent and 0.2 equiv of TES-OTf
at -78 °C. However, the TES ether protecting group of 10 proved to be too labile under these
reaction conditions, and mixtures of 16a , 17, and 18a are obtained in reactions of 10 and 15.
Disaccharide 17 arises via in situ cleavage of the TES ether of disaccharide 16a , while trisaccharide
18a results from a glycosidation of in situ generated 17 (or of 16a itself) with a second equivalent
of 10. These problems were largely suppressed by using 3 with a 3-O-TBS ether protecting group
as the glycosyl donor and 0.2 equiv of TES-OTf as the reaction promoter. Attempts to selectively
glycosylate the C(3)-OH of diol acceptors 20 or 28 gave a 70:30 mixture of 21 and 22 in the reaction
of 20 and a 43:27:30 mixture of regioisomeric trisaccharides 29 and 30 and tetrasaccharide 31
from the glycosidation reaction of 28. However, excellent results were obtained in the glycosidation
of differentially protected disaccharide 34 using 1.5 equiv of 3 and 0.05 equiv of TBS-OTf in CH2-
Cl2 at -78 °C. The latter step is an important transformation in the recently reported synthesis of
the landomycin A hexasaccharide unit.
Landomycin A,2,3 a member of the angucycline antibi-
Sch em e 1
otic family,4,5 is of considerable interest as a potential
antitumor agent.3,6,7 It is known that landomycin A
inhibits DNA synthesis and G1/S cell cycle progression5,6
and that the cytostatic properties of other members of
the landomycin family depend on the length of the
oligosaccharide chain.5,6 While syntheses of the lando-
mycin A aglycone have not yet been reported, syntheses
of the hexasaccharide unit have been reported by us8 and
by Sulikowski,9 while Kirschning has reported a synthe-
sis of the repeat A-B-C trisaccharide unit (Scheme 1).10
During the course of our work on the synthesis of
hexasaccharide 1,8 we needed a convenient source of the
L-rhodinosyl acetate derivative 3 for introduction of the
C residue in the repeat trisaccharide 2. L-Rhodinose, a
2,3,6-trideoxy-L-hexose, is a constituent of several classes
of natural products including landomycin A, streptoly-
(1) A portion of this research was performed at Indiana University
and is described in detail in C.E.B.’s 2000 Ph.D. Thesis: Bennett, C.
E. Ph.D. Thesis, Indiana University, Bloomington, IN, 2000.
(2) Henkel, T.; Rohr, J .; Beale, J . M.; Schwenen, L. J . Antibiot. 1990,
XLIII, 492.
(3) Weber, S.; Zolke, C.; Rohr, J .; Beale, J . M. J . Org. Chem. 1994,
59, 4211.
(4) Rohr, J .; Thiericke, R. Nat. Prod. Rep. 1992, 9, 103.
(5) Krohn, K.; Rohr, J . Topics Curr. Chem. 1997, 188, 127.
(6) Rohr, J .; Wohlert, S.-E.; Oelkers, C.; Kirschning, A.; Ries, M.
Chem. Commun. 1997, 973.
(7) Depenbrock, H.; Bornschlegl, S.; Peter, R.; Rohr, J .; Schmid, P.;
Schweighart, P.; Block, T.; Rastetter, J .; Hanauske, A.-R. Ann. He-
matol. 1996, 73, A80.
(8) Roush, W. R.; Bennett, C. E. J . Am. Chem. Soc. 2000, 122, 6124.
(9) Guo, Y.; Sulikowski, G. A. J . Am. Chem. Soc. 1998, 120, 1392.
(10) Kirschning, A. Eur. J . Org. Chem. 1998, 2267, 7.
digin, and vineomycin B2.11-13 Numerous syntheses of
rhodinose and its derivatives are known in the litera-
10.1021/jo015756a CCC: $20.00 © 2001 American Chemical Society
Published on Web 08/24/2001

6390 J . Org. Chem., Vol. 66, No. 19, 2001
Roush et al.
Sch em e 2
Sch em e 3
silylated with triethylsilyl triflate (TES-OTf) and pyri-
dine, thereby providing TES ether 8 in 73% yield overall
from 6. Hydroboration of 8 with 9-BBN in THF followed
by careful oxidation of the organoborane intermediate
with hydrogen peroxide and sodium acetate afforded the
primary alcohol in 69-79% yield. Unfortunately, the TES
ether of this and subsequent intermediates proved to be
very sensitive to cleavage. Subsequent Swern oxidation32
of the primary alcohol then gave the corresponding
aldehyde 9 in good yield. Oxidative cleavage of the
p-methoxybenzyl ether with DDQ33 in a CH2Cl2-water
mixture then yielded a 1:1 anomeric mixture of lactols
with J 4,5 ) 1.3 Hz. These lactols were acetylated by
treatment with Ac2O and pyridine in CH2Cl2 to provide
a 1:1 anomeric mixture of rhodinosyl acetates 10 in 76%
yield over the last two steps. The stereochemistry of the
rhodinosyl acetate derivatives 10R and 10â was con-
firmed by the observation of J 4,5 coupling constants of
1.2 Hz for the R-anomer and 1.4 Hz for the â-anomer,
indicating an axial-equatorial relationship of the C(4)
and C(5) protons, and hence syn stereochemistry in 8.
As we examined glycosidations of 10, it became ap-
parent that the TES ether protecting group was too labile
under the glycosidation conditions (vide infra). Conse-
quently, the rhodinose derivative 3 containing a much
more robust tert-butyldimethylsilyl (TBS) protecting
group was also synthesized. Intermediate 11 was pre-
pared in 74% overall yield from 6 by substituting tert-
butyldimethylsilyl chloride, imidazole and DMF for the
TES ether protection step in the sequence employed for
synthesis of 8. Hydroboration of 11 using catecholborane
in the presence of Wilkinson’s catalyst ((Ph3P)3RhCl)34
produced a boronic ester that was oxidized by treatment
with aqueous NaOH and H2O2, thereby giving primary
alcohol 12 in 88-90% yield (Scheme 3). Oxidation of this
alcohol under Swern conditions,32 followed by deprotec-
tion of the PMB ether by treatment with DDQ in wet
CH2Cl2, provided a mixture of hemiacetals. Acylation of
this mixture with Ac2O and Et3N afforded a mixture of
â-rhodinosyl acetate 3 and anisaldehyde. None of the
R-anomer of 3 was observed under these conditions.35 The
contaminating anisaldehyde was removed by extraction
of this mixture with aqueous NaHSO3 to form a water-
soluble bisulfite adduct. Rhodinosyl acetate 3 was thus
obtained in 72% yield over the last three steps. In this
way, rhodinose derivative 3 was prepared on multigram
scale in eight steps from commercially available methyl
(S)-lactate in 38% overall yield.
ture,14-23 including several that originate from more
readily available sugars.24-28 We were most attracted to
the routes that originated from methyl (S)-lactate14,19,20
and adopted modifications of Schlessinger’s route for
synthesis of 3.20,29
We initially targeted the triethylsilyl (TES)-protected
10 as the rhodinosyl acetate donor for the landomycin A
synthesis. Accordingly, commercially available methyl
(S)-lactate (4) was protected as a p-methoxybenzyl (PMB)
ether by treatment with p-methoxybenzyl trichloroace-
timidate (5) and 0.3 mol % of TfOH in cyclohexane, giving
PMB ether 6 in 82% yield (Scheme 2).30 Controlled
reduction of 6 with 1.2 equiv of DIBAL in a 1:1 mixture
of CH2Cl2-hexane solvent mixture at -78 °C provided
lactaldehyde 7 in excellent yield. Treatment of 7 with
MgBr2‚OEt2 and allyltributylstannane in CH2Cl2 at -23
°C provided the expected homoallylic alcohol in 84% yield
with excellent stereoselectivity.20,31 This alcohol was then
(11) Rinehart, K. L., J r.; Beck, J . R.; Borders, D. B.; Kinstle, T. H.;
Krauss, D. J . Am. Chem. Soc. 1963, 85, 4038.
(12) Imamura, N.; Kakinuma, K.; Ikekawa, N.; Omura, S. J .
Antibiot. 1981, 34, 1517.
(13) Kirschning, A.; Bechthold, A. F.-W.; Rohr, J . Top. Curr. Chem.
1997, 188, 1.
(14) Kelly, T. R.; Kaul, P. N. J . Org. Chem. 1983, 48, 2775.
(15) DeShong, P.; Leginus, J . M. Tetrahedron Lett. 1984, 25, 5355.
(16) Itoh, T.; Yoshinaka, A.; Sato, T.; Fujisawa, T. Chem. Lett. 1985,
1679.
(17) Servi, S. J . Org. Chem. 1985, 50, 5865.
(18) Hatakeyama, S.; Sakurai, K.; Takano, S. Heterocycles 1986, 24,
633.
(19) Boeckman, R. K., J r.; Potenza, J . C.; Enholm, E. J . J . Org.
Chem. 1987, 52, 469.
(20) Schlessinger, R. H.; Graves, D. D. Tetrahedron Lett. 1987, 28,
4381.
(21) Dondoni, A.; Fantin, G.; Fogagnolo, M.; Pedrini, P. Tetrahedron
1989, 45, 5141.
(22) Herczegh, P.; Kovacs, I.; Laszlo, A.; Dinya, Z.; Sztaricskai, F.
J . Liebigs Ann. Chem. 1991, 599.
(23) Sobti, A.; Sulikowski, G. A. Tetrahedron Lett. 1995, 36, 4193.
(24) Laland, S.; Overend, W. G.; Stacey, M. J . Chem. Soc. 1950, 738.
(25) Haines, A. H. Carbohydr. Res. 1972, 21, 99.
(26) Knollmann, R.; Dyong, I. Chem. Ber. 1975, 108, 2021.
(27) El Khadem, H. S.; Cermak, R. C. Carbohydr. Res. 1979, 75,
335.
(28) Cardillo, G.; Orena, M.; Sandri, S.; Tomasini, C. Gazz. Chim.
Ital. 1984, 114, 325.
(29) Schlessinger used O-benzyl lactate as the starting material. We
elected to use a PMB ether protecting group in the work described
here.
(30) Nakajima, N.; Horita, K.; Abe, R.; Yonemitsu, O. Tetrahedron
Lett. 1988, 29, 4139.
(31) Keck, G. E.; Boden, E. P. Tetrahedron Lett. 1984, 25, 265.
(32) Tidwell, T. T. Org. React. 1990, 39, 297.
(33) Horita, K.; Yoshioka, T.; Tanaka, T.; Oikawa, Y.; Yonemitsu,
O. Tetrahedron 1986, 42, 3021.
(34) Evans, D. A.; Fu, G. C.; Hoveyda, A. H. J . Am. Chem. Soc. 1992,
114, 6671.
(35) Bols, M.; Hansen, H. C. Acta Chem. Scand. 1993, 47, 818.

Synthesis of Landomycin A
Sch em e 4
J . Org. Chem., Vol. 66, No. 19, 2001 6391
McDonald activated the glycal with p-TsOH in the
presence of an acceptor to afford R-glycosides in good
yields.
To determine if the rhodinosyl acetate derivatives 3
and 10 would undergo highly stereoselective R-glycosi-
dation reactions with a relatively hindered secondary
hydroxyl acceptor,38-40 we explored the reactions of 3 and
10 with glucopyranose 1547 (see Table 1). These reactions
were performed at -78 °C typically in Et2O using a slight
excess of 15 (1.2-1.6 equiv). We initially used TES ether
10 as the glycosyl donor and TMS-OTf as the catalyst.
However, as shown in the first entry of Table 1, a 52:36:
12 mixture of three products was obtained in 41% yield
when the reaction was performed in CH2Cl2/disaccharide
16a , disaccharide 17 in which the C(4′)-hydroxyl had
been deprotected under the reaction conditions, and
trisaccharide 18a resulting from glycosidation of the
intermediate 17 (or 16a ) with a second equivalent of 10.
All three products had R-configurations at the new
glycosidic centers. The ratio of products 16a :17:18a
improved to 77:11:12 when the reaction was performed
in Et2O, which presumably attenuates the Lewis acidity
of the TMS-OTf catalyst (Table 1, entry 2). An improved
product ratio of 86:10:4 was obtained for products 16a ,
17, and 18a when 10 was used as the donor with TES-
OTf as the glycosidation catalyst (Table 1, entry 3).
Unfortunately, these products were isolated in a lower
combined yield (58%) under from this reaction.
These results established that the TES ether of 10 was
too labile under the reaction conditions. Accordingly, we
decided to employ the TBS-protected rhodinosyl acetate
3 instead. The TMS-OTf-promoted glycosidation of 3 and
15 still afforded three products: 16b, 17, and 18b; but
they were isolated in a combined yield of 65% and in a
94:3:3 ratio (Table 1, entry 4). We have previously used
TBS-OTf as the promoter of glycosidation reactions of
sensitive substrates,48-50 in all cases taking advantage
of the diminished Lewis acidity of this reagent compared
to TMS-OTf to minimize production of unwanted side
products (including trans-silylation of silyl ether protect-
ing groups). We were surprised, therefore, that the ratio
of products from the TBS-OTf-catalyzed glycosidation of
3 and 15 was reduced to 83:7:10 (Table 1, entry 6).
However, use of TES-OTf as the Lewis acid provided the
products in a 96:3:1 ratio and a combined yield of 68%
(Table 1, entry 5).
A second route to 3 was also explored in which
lactaldehyde derivative 7 was subjected to a chelate
controlled reaction with 3-butenylmagnesium bromide.36,37
Best results were obtained when this reaction was
performed in Et2O at 0 °C, and the diastereoselectivity
of the reaction under these conditions was ca. 20:1.
Protection of the major diastereomer 13 as a TBS ether
and deprotection of the C(5)-PMB ether then gave 14 in
high yield (Scheme 4). Finally, ozonolysis of 14 and
acylation of the resulting mixture of hemiacetals provided
L-rhodinosyl acetate 3 in 58% yield for the final two steps.
Although this synthesis is one step shorter than the route
that proceeds by way of 11, the diastereoselectivity of the
key carbonyl addition step is lower and the two isomers
are difficult to separate especially on large scale. Con-
sequently, this route was never scaled up to provide
significant quantities of 3 for use in glycosylation studies.
In contrast to the numerous syntheses of rhodinose
derivatives that have been developed, there are
fewer reports of glycosidation reactions of rhodinosyl
donors.13,38-40 In an early example, Boeckman treated a
protected rhodinosyl acetate with BF3‚OEt2 in pyrrolidine
as solvent at 23 °C to afford a â-N-rhodinosyl pyrrolidine
in 90% yield.19 Schlessinger has reported the conversion
of a rhodinose lactol to a â-hemiaminal in quantitative
yield simply by stirring the lactol and amine in MeOH.41
In his syntheses of the landomycin A hexasaccharide and
a trisaccharide fragment of PI-080, Sulikowski used
rhodinosyl tetrazoles to construct 2-deoxy-R-glycosidic
linkages.9,42 Rhodinose glycals have also been used as
donors by Kirschning43 and McDonald.44 Kirschning
activated the glycal with NIS in the presence of an
acceptor to provide R-glycosides in good yields,45,46 while
In contemplating the synthesis of the landomycin A
hexasaccharide, we initially considered the possibility
that disaccharide 1951 could serve as the A-B (and A′B′)
disaccharide unit. We hoped that the derived diol 20
could be selectively glycosylated at the C(3)-hydroxyl,
since the electron withdrawing properties of the pyran
ring oxygen should make the C(4)-hydroxyl less nucleo-
philic than the C(3)-hydroxyl. Accordingly, deprotection
of the TBS ethers of disaccharide 19 using Et3N‚HF in
CH3CN at 65 °C gave diol 20 in 85% yield (Scheme 5).
However, treatment of a mixture of diol 20 and TBS-
protected rhodinosyl acetate 3 with TES-OTf (0.1 equiv)
in Et2O at -78 °C afforded a 70:30 mixture of the desired
(36) Nemoto, H.; Shiraki, M.; Fukumoto, K. J . Org. Chem. 1996,
61, 1347.
(37) Reetz, M. T. Angew. Chem., Int. Ed. Engl. 1984, 23, 556.
(38) Toshima, K.; Tatsuta, K. Chem. Rev. 1993, 93, 1503.
(39) Schmidt, R. R. Angew. Chem., Int. Ed. Engl. 1986, 25, 212.
(40) Marazabadi, C. H.; Franck, R. W. Tetrahedron 2000, 56, 8385.
(41) Schlessinger, R. H.; Graves, D. D. Tetrahedron Lett. 1987, 28,
4385.
(42) Sobti, A.; Kim, K.; Sulikowski, G. A. J . Org. Chem. 1996, 61, 6.
(43) Drager, G.; Garming, A.; Maul, C.; Noltemeyer, M.; Thiericke,
R.; Zerlin, M.; Kirschning, A. Chem. Eur. J . 1998, 4, 1324.
(44) McDonald, F. E.; Zhu, H. Y. H. J . Am. Chem. Soc. 1998, 120,
4246.
(47) Garegg, P. J .; Hultberg, H.; Wallin, S. Carbohydr. Res. 1982,
108, 97.
(48) Roush, W. R.; Hartz, R. A.; Gustin, D. J . J . Am. Chem. Soc.
1999, 121, 1990.
(49) Roush, W. R.; Gung, B. W.; Bennett, C. E. Org. Lett. 1999, 1,
891.
(50) Roush, W. R.; Narayan, S. Org. Lett. 1999, 1, 899.
(51) Roush, W. R.; Bennett, C. E. J . Am. Chem. Soc. 1999, 121, 3541.
(45) Thiem, J . In Trends in Synthetic Carbohydrate Chemistry;
Horton, D., Hawkins, L. D., McGarvey, G. J ., Eds.; ACS Symposium
Series; American Chemical Society: Washington, DC, 1989; Vol. 386;
p 131 and literature cited therein.
(46) Thiem, J .; Klaffke, W. Top. Curr. Chem. 1990, 154, 285.

6392 J . Org. Chem., Vol. 66, No. 19, 2001
Roush et al.
Ta ble 1. Glycosid a tion s of L-Rh od in osyl Aceta tes 3 a n d 10 w ith Accep tor 15
product
(ratio)b
16:17:18
15
(equiv)
yielda
(%)
donor
Lewis acid
solvent
10
10
10
3
3
3
1.3
1.6
1.3
1.5
1.5
1.2
TMS-OTf (0.1 equiv)
TMS-OTf (0.1 equiv)
TES-OTf (0.2 equiv)
TMS-OTf (0.25 equiv)
TES-OTf (0.2 equiv)
TBS-OTf (0.2 equiv)
CH2Cl2
Et2O
Et2O
Et2O
Et2O
Et2O
41
71
58
65
68
66
52:36:12
77:11:12
86:10:4
94:3:3
96:3:1
83:7:10
a
b
Yield of products isolated by chromatography. Glycosides 16a and 18a derive from reactions with 10 as the donor, whereas 16b and
18b were obtained from reactions in which 3 was the glycosyl donor.
Sch em e 5
Sch em e 6
R-1,3-trisaccharide 21 and the undesired R-1,4-trisac-
charide 22 in a combined yield of 60%. Use of TBS-OTf
as the Lewis acid under the same glycosidation conditions
did not change the reaction regioselectivity. Lowering the
reaction temperature to -100 °C with TES-OTf as
catalyst also did not improve the selectivity of this
experiment. The inter-glycosidic connectivities of 21 (δ
3.42 for H3, δ 3.09 for H4) and 22 (δ 3.61 for H3, δ 3.04
for H4) were established by acetylation of the free
hydroxyl in each with Ac2O in pyridine to afford the
acetylated derivatives 23 (δ 3.77 for H3, δ 4.71 for H4)
1
and 24 (δ 5.18 for H3, δ 3.33 for H4). The H NMR data
summarized here provide unequivocal confirmation of the
connectivity of the new glycosidic bonds in 21 and 22.
of diol 28 and rhodinosyl acetate 3 in Et2O at -78 °C
afforded three products: the desired R-1,3-trisaccharide
29 in 28% yield, the regioisomeric R-1,4-trisaccharide 30
in 18% yield, and tetrasaccharide 31 in 20% yield. Once
again, the connectivity in 29 and 30 was determined by
1H NMR analysis of the acetate derivatives prepared by
acylation of the free hydroxyl groups. The structures of
acetate derivative 32 [high-resolution FAB, calcd for
Because the regioselectivity of the glycosidation of 20
was poor, we decided to remove the C(2′)-iodide substitu-
ent so as to decrease the steric crowding of the C(3′)-
hydroxyl group. This strategy seemed appropriate, since
Thiem had demonstrated in his synthesis of the C-D-E
trisaccharide unit of olivomycin A that glycosidation of
diol 26 with glycal 25 provided trisaccharide 27 with
glycosidation occurring exclusively on C(3)-OH (Scheme
6).52 Accordingly, diol 28 was prepared in 52% yield by
reduction of 20 with Bu3SnH and AIBN in refluxing
benzene.53,54 However, TES-OTf-catalyzed glycosidation
(54) The Bu3SnH reduction of 20 was performed under standard
conditions, and provided 28 in 52% yield. In addition, enone i was
obtained in 30% yield.
(52) Thiem, J .; Gerken, M. J . Org. Chem. 1985, 50, 954, and
references therein.
(53) Neumann, W. P. Synthesis 1987, 665.

Synthesis of Landomycin A
Sch em e 7
J . Org. Chem., Vol. 66, No. 19, 2001 6393
to the rhodinose unit was exclusively R, and the 3:1
anomeric mixture in the A ring was unchanged. Disac-
charide 34 was poorly soluble in Et2O at -78 °C;
consequently, we used CH2Cl2 for this reaction even
though the preliminary studies (Table 1) suggested that
these conditions would be less than optimal. It is also
interesting to note that use of TBS-OTf (0.05 equiv) as
the glycosidation promoter gave excellent results in this
case, again in contrast to the results summarized in Table
1. Perhaps the excellent result in this case is due to the
low catalyst loading and relatively short reaction time.
In summary, the examples presented here establish
that the TBS-protected rhodinosyl acetate 3 is a highly
reactive and highly stereoselective donor for synthesis
of R-L-rhodinosyl glycosides. The best results have been
obtained when TES-OTf or TBS-OTf are used as the
glycosidation catalysts at -78 °C. Although it was not
possible to achieve regioselective glycosidations of diols
20 and 28, excellent results were obtained in the gly-
cosidation of disaccharide 34. Additional progress toward
the synthesis of landomycin A and full details of our
synthesis of hexasaccharide 1 will be reported separately.
C40H58O12SSiNa (M + Na)+ 813.3316, found 813.3318
m/z], acetate derivative 33 [high-resolution FAB, calcd
for C40H58O12SSiNa, (M + Na)+ 813.3316, found 813.3282
m/z], and 31 [low resolution FAB spectrum for C50H80O13-
SSi2Na, (M + Na)+ calcd 999, found 999 m/z] were also
confirmed by mass spectral data.
It was apparent that the development of an efficient
synthesis of the landomycin A repeat trisaccharide would
require use of a fully differentially protected B (or B′)
ring monosaccharide unit, such that the regioselectivity
issues encountered in glycosidations of 20 and 28 would
be avoided. Ultimately, this was accomplished by using
disaccharide 34,8 a 3:1 mixture of anomers at the A ring
glycosyl acetate, as the A-B disaccharide acceptor. Thus,
treatment of 34 with 1.5 equiv of 3 and 0.05 equiv of TBS-
OTf in CH2Cl2 at -78 °C in the presence of 4 Å molecular
sieves provided the A-B-C repeat trisaccharide 2 in 91%
yield (Scheme 7). The newly formed B-C glycosyl linkage
Ack n ow led gm en t. We thank the National Insti-
tutes of Health (GM 38907) for financial support.
Su p p or tin g In for m a tion Ava ila ble: Experimental de-
tails for synthesis of 3, representative procedures for the
glycosidation reactions, and 1H NMR spectra of 2R, 2â, 3, 6,
10R, 10â, 36, and 38. This material is available free of charge
via the Internet at http://pubs.acs.org.
J O015756A