De novo synthesis of the trisaccharide subunit of landomycins A and E

Article abstract of DOI:10.1021/ol800697k

(Chemical Equation Presented) A highly enantio- and diastereoselective synthesis of α-L-rhodinose, β-D-olivose as well as the trisaccharide portion of landomycin A from achiral acetyl furan has been developed. The key transformations include the palladium

Full text of DOI:10.1021/ol800697k

ORGANIC
LETTERS
2008
Vol. 10, No. 11
2283-2286
De Novo Synthesis of the Trisaccharide
Subunit of Landomycins A and E
Maoquan Zhou and George A. O’Doherty*
Department of Chemistry, West Virginia UniVersity, Morgantown, West Virginia 26506
george.odoherty@mail.wVu.edu
Received March 27, 2008
ABSTRACT
A highly enantio- and diastereoselective synthesis of r-L-rhodinose, ꢀ-D-olivose as well as the trisaccharide portion of landomycin A from
achiral acetyl furan has been developed. The key transformations include the palladium-catalyzed glycosylation, Myers’ reductive rearrangement,
diastereoselective dihydroxylation, and regioselective Mitsunobu inversion. A Mitsunobu reaction on a six member ring cis-1,2-diol was found
to chemoselectively discriminate between equatorial and axial alcohols and to stereoselectively convert cis-1,2-diol into anti-1,2-diol.
The landomycins, a unique class of angucycline-type anti-
biotics, display potent antitumor and antibacterial activity.1
Landomycin A (1),2 the most complex member of the family,
was first discovered in 1990 as the principle product of
Streptomyces cyanogenus S136 bearing a deoxysugar con-
taining hexasaccharide made up of two repeating trisaccha-
rides (R-L-rhodinose-(1f3)-ꢀ-D-olivose-(1f4)-ꢀ-D-olivose)
(Figure 1). Of particular interest, landomycin A possesses
significant antitumor activity against a range of 60 cancer
cell lines.3 The structure-activity relationship (SAR) studies
suggested that the high cytotoxic activity of landomycin A
depends on the length of the glycan chain, in which
Landomycin B, J, E, D, I with shorter sugar side chain
(Figure 1), showed diminished cytotoxic activities. Although
it is known that landomycin A interacts with DNA, inhibits
DNA synthesis, and halts G1/S cell cycle progression,4 the
(1) (a) Rohr, J.; Thiericke, R. Nat. Prod. Rep. 1992, 9, 103. (b) Krohn,
K.; Rohr, J. Top. Curr. Chem. 1997, 188, 127. (c) Andriy, L.; Andreas, V.;
Andreas, B. Mol. BioSyst. 2005, 1, 117–126.
(2) (a) Henkel, T.; Rohr, J.; Beale, J. M.; Schwenen, L. J. Antibiot. 1990,
43, 492. (b) Weber, S.; Zolke, C.; Rohr, J.; Beale, J. M. J. Org. Chem.
1994, 59, 4211. The structure of the Landomycin A aglycon is still under
Figure 1. Structures of landomycins A, B, J, E, D, and I.
debate, see: (c) Roush, W. R.; Neitz, R. J. J. Org. Chem. 2004, 69, 4906.
(d) Zhu, L. I.; Luzhetskyy, A.; Luzhetska, M.; Mattingly, C.; Adams, V.;
Bechthold, A.; Rohr, J. ChemBioChem 2007, 8, 83–88.
(3) (a) Depenbrock, H.; Bornschlegl, S.; Peter, R.; Rohr, J.; Schmid,
P.; Schweighart, P.; Block, T.; Rastetter, J.; Hanauske, A.-R. Ann. Hematol.
1996, 73, A80. (b) Rohr, J.; Wohlert, S.-E.; Oelkers, C.; Kirschning, A.;
Ries, M. Chem. Commun. 1997, 973.
specific mechanism of action of landomycin A on cancer
cells has not yet been determined. As part of a program aimed
at the synthesis and biological study of carbohydrate contain-
ing natural products, we became interested in the synthesis
(4) Crow, R. T.; Rosenbaum, B.; Smith, R., III; Guo, Y.; Ramos, K. S.;
Sulikowski, G. A. Bioorg. Med. Chem. Lett. 1999, 9, 1663–1666.
10.1021/ol800697k CCC: $40.75
Published on Web 05/08/2008
2008 American Chemical Society

of the trisaccharide portion of Landomycin E, which is the
repeat unit of the hexasaccharide portion of Landomycin A.
Ultimately, we hope to use this trisaccharide to study its
properties in cellular transportation and DNA binding ability
(Scheme 1).
akin to the approach by Sulikowski and co-workers (P ) Ac).6b
The trisaccharide 8 could easily be prepared from 9, which
allows us to maximize the convergency. By means of a
sequence of palladium catalyzed glycosylations and subsequent
postglycosylation modifications9 we foresaw the trisaccharide
9 as being derived from one unit of the R-L-pyranone 10 and
two units of ꢀ-D-pyranone 11. Finally, the R-L-rhodinose
precursor pyranone 10 and ꢀ-D-olivose precursor pyranone 11
could be easily prepared from achiral acylfuran 12.
Scheme 1
.
Retrosynthetic Analysis of Landomycin A
Hexasaccharide
Previously, we have shown that the R- and ꢀ-stereoisomers
of either enantiomer of pyranones 10 and 11 can be prepared
from acyl furan 12 by employing an enantioselectively Noyori
reduction (12 to 13/ent-13),10,11 an Achmatowicz oxidation,
and diastereoselective tert-butyl carbonate formation (Scheme
2).12 The ꢀ-pyranone 11 and ent-11 can be isolated in ∼50%
Scheme 2. Synthesis of D/L-R/ꢀ-Boc-pyranones
While there are no reports of the total synthesis of
landomycin A,2c,5 the synthesis of the hexasaccharide
fragment has been completed by three groups (Sulikowski,
Yu, and Roush).6 In contrast to the previous syntheses of
the landomycin A hexasaccharide, which derived their
asymmetry from carbohydrate starting material, we were
interested in a de novo asymmetric approach that would use
asymmetric catalysis to install the 22 stereocenters of the
landomycin A hexasaccharide from achiral starting materials.
While we previously have had some success using our Pd-
catalyzed glycosylation and postglycosylation reactions for
the synthesis of several naturally occurring structural motifs,
the landomycins offered the opportunity to apply this
methodology for the synthesis of new carbohydrate subunits
(e.g., R-L-rhodinose and ꢀ-D-olivose).7,8 For example, the
ꢀ-D-olivose required the development of new postglycosy-
lation transformations. In particular, we anticipated a need
for a new transformation that allowed for the net 1,2-trans-
diequatorial addition of two hydroxyl groups across a six-
membered ring alkene.
yield when the Boc-protection was preformed at elevated
temperature ((Boc)2O/NaOAc in benzene at 80 °C). Alter-
natively, R-pyranones 10 and ent-10 can be selectively
prepared (R:ꢀ ) 3:1) at low temperatures (-78 °C).
Our planned synthesis of landomycin A hexasaccharide
was built up from the previous successful approach to
2-deoxy-ꢀ-glycoside associated with digitoxin.13 To do this
would require the selective conversion of the allo-stereo-
chemistry of digitoxose into the gluco-stereochemistry of
olivose. That is to say, invert the C-3 axial alcohol to a C-3
equatorial alcohol, which we hoped to achieve without any
unnecessary protection/deprotection steps.
As part of our approach to the digitoxin trisaccharides,
we have shown the axial alcohol of the C-3/C-4 cis-diol in
allo-sugars could be regioselectively acylated via the known
stereoelectronically controlled ortho-ester hydrolysis (i.e., 14
to 15 via 16, Scheme 3).13,14 We postulated that a similar
stereoelectronic control could govern the protonation and
subsequent Mitsunobu-type opening of a dioxyphosphorane
fused to a six membered ring, as in 18 the one formed from
the reaction of 17 with PPh3 and DIAD.15 Thus, selective
protonation of the initially formed pentacoordinated
dioxyphosphorane 1816,17 should preferentially form the
oxyphosphonium betaine 19, which after subsequent SN2
displacement of Ph3PO by p-nitrobenzoate ion should afford
the C-3-OPNBz regioisomer 20.
Strategically, the two trisaccharide derivatives 8 and 9 can
be coupled together to build the final hexasaccharide in a fashion
(5) For the biosynthesis of the landomycin, see: (a) Von Mulert, U.;
Luzhetskyy, A.; Hofmann, C.; Mayer, A.; Bechthold, A. FEMS Microbiol.
Lett. 2004, 230, 91–97. (b) Baig, I.; Kharel, M.; Kobylyanskyy, A.; Zhu,
L.; Rebets, Y.; Ostash, B.; Luzhetskyy, A.; Bechthold, A.; Fedorenko, V. A.;
Rohr, J. Angew. Chem., Int. Ed. 2006, 45, 7842–7846
.
2284
Org. Lett., Vol. 10, No. 11, 2008

Scheme 3
.
Synthesis of 2-Deoxy-ꢀ-glucose from
2-Deoxy-ꢀ-allose
Scheme 4. Synthesis of Monosaccharide ꢀ-D-Olivose
4-step sequence: palladium catalyzed glycosylation with
BnOH, Luche reduction20 of the keto group, Myers’ reduc-
tive 1,3-allylic transposition21 and dihydroxylation,22 gave
exclusively the diol 22 in 56% overall yield.13 Under
Mitsunobu reaction conditions, the digitoxose 22 was cleanly
converted to protected ꢀ-D-olivose 23 in 71% yield while
the benzoate group also served as a temporary protecting
group. A simple hydrolysis of benzoate group with K2CO3
fashioned the monosaccharide ꢀ-D-olivose 24 in 98% yield.
With a viable route to the ꢀ-D-oliVo-sugar 24 in hand, our
efforts turned to the construction of the di- and trisaccharide
(Schemes 5 and 6). Thus, the oliVo-sugar 23 and ꢀ-D-
pyranone 11 were subjected to the typical palladium-
catalyzed glycosylation conditions, which afforded the C-4
glycosylated disaccharide 25 in 85% yield with complete
To our delight, the direct Mitsunobu reaction18 on the diol
17 gave only the C-3 equatorial nitro-benzoate product 20
in 64% yield along with 30% recovered 17 (Scheme 3). Upon
basic hydrolysis, ester 20 was converted to the 2-deoxy-ꢀ-
glucose 21, which has the desired ꢀ-olivose stereochemistry
(2,6-dideoxy-ꢀ-glucose). The stereochemistry was confirmed
by 1H NMR coupling constant analysis. The coupling
(10) Fujii, A.; Hashiguchi, S.; Uematsu, N.; Ikariya, T.; Noyori, R. J. Am.
Chem. Soc. 1996, 118, 2521–2522.
(11) (a) Li, M.; Scott, J. G.; O’Doherty, G. A. Tetrahedron Lett. 2004,
45, 1005–1009. (b) Li, M.; O’Doherty, G. A. Tetrahedron Lett. 2004, 45,
6407–6411.
3
constant JH3-H4 in diol 17 is 3.0 ppm, which is the typical
axial-equatorial proton coupling constant; while in diol 21,
(12) For the preparation of Boc-protected pyranones, see: (a) Babu, R. S.;
O’Doherty, G. A. J. Carb. Chem. 2005, 24, 169–177. (b) Guo, H.;
O’Doherty, G. A. Org. Lett. 2005, 7, 3921–3924.
the axial-axial JH3-H4 equal to 8.4 ppm.19
3
With this highly chemoselective discrimination and ste-
reoselective inversion of the 1,2-diol established, we then
applied this transformation to the synthesis of landomycin
A monosaccharide ꢀ-D-olivose (Scheme 4). As the starting
point, both the enantiopure pyranones R-L-10 and ꢀ-D-11
were diastereoselectively prepared as the major products
(Scheme 2).9a,13 Subjecting pyranone 11 to the following
(13) Zhou, M.; O’Doherty, G. A. J. Org. Chem. 2007, 72, 2485–2493.
(14) (a) King, J. F.; Allbutt, A. D. Can. J. Chem. 1970, 48, 1754–1769.
(b) Lowary, T. L.; Hindsgaul, O. Carbohydr. Res. 1994, 251, 33–67.
(15) (a) Garcia-Delgado, N.; Riera, A.; Verdaguer, X. Org. Lett. 2007,
9, 635–638. (b) Mathieu-Pelta, I.; Evans, S. A., Jr. J. Org. Chem. 1992,
57, 3409–3413.
(16) (c) Robinson, P. L.; Barry, C. N.; Kelley, J. W.; Evans, S. A., Jr.
J. Am. Chem. Soc. 1985, 107, 5210–5219. (d) Robinson, P. L.; Barry, C. N.;
Bass, S. W.; Jarvis, S. E.; Evans, S. A., Jr. J. Org. Chem. 1983, 48, 5396–
5398. Luckenbach, R. Dynamic Stereochemistry of Pentacoordinated
Phosphorus and Related Elements; George Thieme: Stuttgart, FRG, 1973.
(17) A pentacoordinated dioxyphosphorane like 18 was suggested to
have a trigonal-bipyramidal configuration with a the two P-O bonds to
occupying an equatorial-axial position on the phosphorous, see: (a) Holmes,
R. R. J. Am. Chem. Soc. 1978, 100, 433. (b) Holmes, R. R. Pentaroordinated
Phosphorus; American Chemical Society: Washington, DC, 1980; Vols. I,
11; ACS Monograph No. 175, p 176.
(6) For chemical synthesis of the trisaccharide, see: (a) Kirschning, A.
Eur. J. Org. Chem. 1998, 2267. For chemical synthesis of hexasaccharide,
see: (b) Guo, Y.; Sulikowski, G. A. J. Am. Chem. Soc. 1998, 120, 1392.
(c) Yu, B.; Wang, P. Org. Lett. 2002, 4, 1919. (d) Roush, W. R.; Bennett,
C. E. J. Am. Chem. Soc. 2000, 122, 6124. For chemical synthesis of the
aglycon, see ref 2c.
(7) For the stereocontrolled synthesis of rare deoxysugars, see: (a) Sherry,
B. D.; Loy, R. N.; Toste, F. D. J. Am. Chem. Soc. 2004, 126, 4510–4511.
(b) Tanaka, H.; Yoshizawa, A.; Takahashi, T. Angew. Chem., Int. Ed. 2007,
14, 2505–2507. (c) Davidson, M. H.; McDonald, F. E. Org. Lett. 2004, 6,
1601–1603. (d) McDonald, F. E.; Reddy, K. S.; Diaz, Y. J. Am. Chem.
Soc. 2000, 122, 4304–4309.
(18) For reviews on Mitsunobu reaction, see: (a) Mitsunobu, O. Synthesis
1981128. (b) Hughes, D. L. Org. Prep. 1996, 28, 127–164. For examples
of the Mitsunobu reaction on acyclic 1,2-diols, see: (c) He, L.; Wanunu,
M.; Byun, H.-S.; Bittman, R. J. Org. Chem. 1999, 64, 6049–6055. (d) Ko,
S. Y. J. Org. Chem. 2002, 67, 2689–2691. (e) Pautard-Cooper, A.; Evans,
S. A., Jr. J. Org. Chem. 1989, 54, 2485–2488, 4974.
(8) For a review of deoxysugar syntheses, see: Kirschning, A.; Jesberger,
M.; Scho¨ning, K.-U. Synthesis 2001, 507–540.
(19) To the best of our knowledge, this is the first example of this type
of 1,2-cis-diol Mitsunobu in a six-membered ring system, which when used
in conjunction with a highly diastereoselective dihydroxylation reaction
constitutes a nice solution to the problem associated with the net 1,2-
diequatorial addition to a cyclohexene.
(9) (a) Babu, R. S.; O’Doherty, G. A. J. Am. Chem. Soc. 2003, 125,
12406–12407. (b) Comely, A. C.; Eelkema, R.; Minnaard, A. J.; Feringa,
B. L. J. Am. Chem. Soc. 2003, 125, 8714–8715. (c) Kim, H.; Men, H.;
Lee, C. J. Am. Chem. Soc. 2004, 126, 1336–1337. For its application in the
de novo synthesis of oligosaccharides, see: (d) Babu, R. S.; Zhou, M.;
O’Doherty, G. A. J. Am. Chem. Soc. 2004, 126, 3428–3429. (e) Zhou, M.;
O’Doherty, G. A. Org. Lett. 2006, 8, 4339–4342. (f) Guppi, R. S.; Zhou,
M.; O’Doherty, G. A. Org. Lett. 2006, 8, 293–296. (g) Guo, H.; O’Doherty,
G. A. Angew. Chem., Int. Ed. 2007, 46, 5206–5208. (h) Shan, M.;
O’Doherty, G. A. Org. Lett. 2006, 8, 5149–5152.
(20) (a) Luche, J. L. J. Am. Chem. Soc. 1978, 100, 2226–2227. (b)
Haukaas, M. H.; O’Doherty, G. A. Org. Lett. 2001, 3, 401–404.
(21) (a) Myers, A. G.; Zheng, B. J. Am. Chem. Soc. 1996, 118, 4492–
4493. (b) Haukaas, M. H.; O’Doherty, G. A. Org. Lett. 2002, 4, 1771–
1774.
(22) VanRheenen, V.; Kelly, R. C.; Cha, D. Y. Tetrahedron Lett. 1976,
17, 1973–1976.
Org. Lett., Vol. 10, No. 11, 2008
2285

Scheme 5
.
Synthesis of Disaccharide Benzyl
Scheme 6. Synthesis of Landomycin A and E Trisaccharide
ꢀ-D-Olivose-(1f4)-ꢀ-D-olivose
stereocontrol at the anomeric center (Scheme 5). Luche
reduction of the keto-group in pyranone 25 provided a
mixture of allylic alcohols, which when exposed to the
Myers’ reductive 1,3-allylic transposition conditions provided
olefin 26 in 83% yield for two steps. At this stage, the
replacement of p-nitrobenzoyl group (PNBz) is necessary
because another Mitsunobu reaction will take place and these
two hydroxyl groups will need to be differentiated. Thus,
the p-nitrobenzoate group was easily hydrolyzed with K2CO3
and MeOH to give alcohol 27 in 98% yield. Protection of
the alcohol with TBSCl provided the olefin 28 in 72% yield.
The dihydroxylation of 28 gave exclusively the diol 29 in
95% yield.
When applying the regioselective Mitsunobu reaction to
diol 29, the benzoate 30 was afforded in 85% yield. Because
the C-3 free hydroxy group will be the glycosyl acceptor
for next glycosylation, the protection of the C-4 alcohol and
deprotection of C-3 benzoate are needed. Thus, treating the
alcohol 30 with TBSCl provided the silyl ether 31 in 82%
yield. The initial effort toward the hydrolysis of benzoate in
31 with K2CO3 and MeOH proved to be troublesome, giving
the desired product 32 along with the C-3-OTBS product in
a ratio of 1:1.4. Presumably the C-3-OTBS product is formed
by a base promoted TBS group migration from C-4 hydroxy
group to C-3 hydroxy group. A much cleaner reaction
resulted from the reduction of the benzoate 31 with DIBAL-
H, which gave alcohol 32 in 99% yield.
Luche reduction to give exclusively the allylic alcohol 34
in 93% yield. Because the stereochemistry of the C-4
hydroxy group was not identical to the target compounds, it
had to be reversed. Gratifyingly, this was achieved by
Mitsunobu reaction, which gave the desired product 35 in
97% yield. Upon hydrolysis, the benzoate in 35 was removed
to give the alcohol 36 in 98% yield. The diimide reduction
of the olefin in 36 provided the trisaccharide 37. Finally,
the landomycin A and E trisaccharide 38 was achieved by
removing the two TBS groups with TBAF (97%). Overall,
the trisaccharide 38 was prepared from the achiral acylfuran
12 in 22 steps and 4.5% overall yield.
In conclusion, a highly enantioselective route to landomycin
E trisaccharide 38 has been developed. The key transformations
include: the palladium catalyzed glycosylation reaction, Myers’
reductive rearrangement, diastereoselective dihydroxylation, and
the regioselective Mitsunobu reaction. This is the first example
of the chemoselective discrimination and stereoselective inver-
sion of the 1,2-diol on a sugar system using a direct Mitsunobu
reaction on diol. The preparation of other analogues as well as
landomycin A and E are ongoing.
Acknowledgment. We are grateful to the NSF (CHE-
0749451) and the ACS-PRF (47094-AC1) for the support
of our research program.
Supporting Information Available: Complete experimen-
tal procedures and spectral data for all new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
The last sugar of the trisaccharide was R-L-rhodinose,
which was derived from the R-L-pyranone 10 (Scheme 6).
The glycosylation of alcohol 32 and R-L-10 provided the
trisaccharide pyranone 33 in 95% yield, which underwent
OL800697K
2286
Org. Lett., Vol. 10, No. 11, 2008