Total synthesis of landomycin A, a potent antitumor angucycline antibiotic

Article abstract of DOI:10.1021/ja205339p

The first total synthesis of landomycin A, the longest and most potent antitumor angucycline antibiotic, has been achieved in 63 steps and 0.34% overall yield starting from 2,5-dihydroxybenzoic acid, 3,5-dimethylphenol, triacetyl d-glucal, and d-xylose, with a convergent linear sequence of 21 steps.

Full text of DOI:10.1021/ja205339p

COMMUNICATION
pubs.acs.org/JACS
Total Synthesis of Landomycin A, a Potent Antitumor
Angucycline Antibiotic
Xiaoyu Yang, Boqiao Fu, and Biao Yu*
State Key Laboratory of Bio-organic and Natural Products Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of
Sciences, Shanghai 200032, China
S Supporting Information
b
ABSTRACT: The first total synthesis of landomycin A,
the longest and most potent antitumor angucycline anti-
biotic, has been achieved in 63 steps and 0.34% overall yield
starting from 2,5-dihydroxybenzoic acid, 3,5-dimethylphe-
nol, triacetyl D-glucal, and D-xylose, with a convergent linear
sequence of 21 steps.
Figure 1
having the OH groups at the 1, 6, and 11 positions protected with
he landomycins constitute a unique group of angucycline
antibiotics featuring a benz[a]anthraquinone aglycone (i.e.,
benzyl or p-nitrobenzyl groups was desired for the subsequent
sugar assembly. The synthesis of 10 was accomplished in a
convergent manner, employing Hauser annulation7 to form the
C ring and Roush’s intramolecular Michael addition5 to form the
B ring (Scheme 1). Cyanophthalide 4 and enone 5 were prepared
from 2,5-dihydroxybenzoic acid (2) and 3,5-dimethylphenol (3) in
eight and seven steps, respectively, not without complications.8
Treatment of 4 with LiHMDS in THF at ꢀ78 °C followed by
addition of 5 at 0 °C led to the desired naphthalene 6 in 48% yield.
The corresponding 1,4-adduct could transform rapidly into the
thermodynamically favored enolate having extended conjugation
with the benzene ring to escape the annulation.9 Asymmetric
T
a landomycinone) with a dearomatized B ring and deoxyoligo-
saccharide chains of various lengths attached at the C8 OH.1,2
These compounds show potent antitumor activities; the potency
varies with variation of the sugar residues and is also relevant to
the cell types.3,4 The mechanism of the antitumor action remains
uncertain, and it also seems to be dependent on the sugar
residues.3b,c The biosynthetic pathway toward these tetracyclic
decaketide glycoconjugates is intriguing, especially in the intro-
duction of the four oxygens (out of six) from air and in the
iterative introduction of the monosaccharide residues.4 The
chemical synthesis of the landomycins has challenged chemists
for many years. The major synthetic difficulties include the
following: (1) the construction of landomycinone demands
avoidance of the extremely easy process of dehydrative aroma-
tization of the B ring; (2) assembly of the deoxyoligosaccharides
requires special devices to control the stereochemistry as well as
to avoid the cleavage/anomerization of the extremely acid-labile
di- and trideoxyglycosidic linkages; and moreover, (3) attach-
ment of the sugars to the poorly nucleophilic hydrogen-bonded
C8 phenol of landomycinone is difficult. In fact, the synthesis of
landomycinone has to date been achieved only by the Roush
group;5 the synthesis of the longest hexasaccharide residue of
landomycin A has been accomplished by the Sulikowski, Roush,
Yu, and Takahashi groups, the trisaccharide fragment by the
Kirschning and O’Doherty groups, and the disaccharide frag-
ment by the McDonald group.6 However, attachment of the
sugar residues to landomycinone to accomplish the total synth-
esis of landomycins has remained an elusive task. Here we report
the total synthesis of landomycin A (Figure 1), the longest and
most potent antitumor congener of the landomycins.
reduction of ketone 6 with BH3 THF/(S)-CBS10 led to the
3
naphthodihydroquinone C6-ol derivative, which was vulnerable to
air and thus was completely converted with ceric ammonium nitrate
(CAN) at 0 °C into quinone 7 in 84% yield with 93% ee. Protection
of the C6 OH in 7 with a p-nitrobenzyl group provided 8; the
benzyl-protected counterpart was surprisingly unstable. Removal of
the 1-O-TBS protection in 8 gave the phenol, which was unstable
and thus was immediately subjected to the intramolecular Michael
addition. Thanks to the careful studies on a similar substrate by
Roush and Neitz,5 we were able to furnish the desired land-
omycinone derivative 9 in 35% yield under slightly modified
conditions [NaOEt, EtOH, 4 Å molecular sieves (MS), air,
55 °C], with isolation of the 1-methyl-3-hydroxy regioisomer
in 8% yield and recovery of the starting phenol in 27% yield.
This transformation was reproducible in a gram-scale synthesis.
Protection of the C1 phenol with a benzyl group (84%) and
removal of the 8-O-MOM group [MgBr2 Et2O, THF, room
3
temperature (rt), 81%]5 provided the desired landomycinone 10.
The phenolic C8 OH in landomycinone 10 is situated on an
electron-deficient naphthoquinone skeleton and hydrogen-
bonded, which makes its glycosidation extremely challenging.
In addition, the vulnerability of the substrate toward aromatization
To realize the total synthesis of a landomycin, any of the
preceding synthetic approaches toward landomycinone and the
deoxyoligosaccharides must be adjusted to enable selective
removal of protecting groups and the subsequent coupling
of the aglycone and sugars. Thus, landomycinone derivative 10
Received: June 9, 2011
Published: July 25, 2011
r
2011 American Chemical Society
12433
dx.doi.org/10.1021/ja205339p J. Am. Chem. Soc. 2011, 133, 12433–12435
|

Journal of the American Chemical Society
COMMUNICATION
Scheme 1
Scheme 3
11 is crucial to the reaction; the similar reaction with the 6-deoxy
counterpart led only to the corresponding glycal. Selective
removal of the 40-O-TBS group on 13 provided 14 (85%), which
was ready for subsequent sugar elongation.
Scheme 2
To assemble landomycin A in a convergent manner, penta-
saccharide trifluoroacetimidate 27 was prepared (Scheme 3).15
Glycosylation of 1511b with L-rhodinosyl acetate 168 under
the catalysis of TBSOTf (0.2 equiv) at ꢀ78 °C afforded the
thermodynamically favored R-disaccharide 17 (94%). Selective
removal of the 40-O-TBS group in 17 was surprisingly proble-
matic; with TBAF/HOAc at 80 °C, the desired 18 was obtained
in moderate yield (49%) with 28% recovery of 17. Coupling of
1911b with 2,6-dideoxy-2-iodoglucopyranosyl trifluoroacetimi-
date 208 under the action of TBSOTf (0.2 equiv) at ꢀ78 °C
afforded β-disaccharide 21 in excellent yield (91%) along with
the corresponding R-disaccharide in ∼5% yield. Removal of the
of the B ring would rule out the applicability of most of the powerful
glycosylation conditions. From the viewpoint of the sugar part,
formation of the β-D-olivopyranoside linkage is also a difficult task
that requires special devices.6 Thus, not surprisingly, numerous
attempts at glycosylation of 10 failed, including adoption of the
previously well-established methods with 2-deoxy-2-iodoglycosyl
trichloroacetimidates,11 2-deoxy-2-selenophenyl-R-D-glucopyanoses
(under Mitsunobu conditions),12 and 2,3-O-thionocarbonyl-1-
thioglycosides as donors.6c,13 Finally came a fortuitous try with
glycosyl iodide under anionic conditions, a protocol developed
by the Gervay-Hague group.14 Thus, treatment of 3-O-acetyl-6-
bromo-4-O-tert-butyldimethylsilyl-2-deoxy-D-glucopyranosyl
acetate (11, six steps from triacetyl D-glucal)8 with TMSI
(CH2Cl2, 0 °C) led to the clean formation of R-iodide 12.
SN2-type substitution of iodide 12 using the naphthoate anion
derived from landomycinone 10 (KHMDS, 18-crown-6, 4 Å MS,
THF) at ꢀ20 °C furnished β-glycoside 13 in a satisfactory 63%
yield, along with a 14% yield of the undesired product formed by
elimination to provide the corresponding Δ5,6 derivative
(Scheme 2). It is worth noting that the 6-bromo substituent in
3-O0-TBS group in 21 was effected nicely with HF Et3N
3
(MeCN, 60 °C), affording 22 in 90% yield. Glycosylation of
22 with L-rhodinosyl acetate 238 promoted by TBSOTf (0.2
equiv) at a higher temperature (ꢀ50 °C) gave the desired
trisaccharide 24 (89%). Selective cleavage of the anomeric
acetate in 24 (N2H4 H2O, CH2Cl2, MeOH, rt)11 followed by
3
trifluoroacetimidate formation16 afforded trisaccharide imidate
25 nearly quantitatively. Coupling disaccharide acceptor 18 with
trisaccharide imidate 25 under the catalysis of TBSOTf (0.05
equiv) in the presence of 5 Å MS at ꢀ78 °C in CH2Cl2 afforded
the desired pentasaccharide 26 in a satisfactory 88% yield.
Pentasaccharide acetate 26 was then transformed to trifluoro-
acetimidate 27 in a manner similar to the 24 f 25 transforma-
tion in an excellent yield of 88%.
Coupling of landomycinone monoglycoside 14 with penta-
saccharide trifluoroacetimidate 27 was achieved under the cata-
lysis of TBSOTf (0.03 equiv) in the presence of 5 Å MS in
CH2Cl2 at ꢀ78 °C, affording the desired hexasaccharide 28 in a
12434
dx.doi.org/10.1021/ja205339p |J. Am. Chem. Soc. 2011, 133, 12433–12435

Journal of the American Chemical Society
COMMUNICATION
Scheme 4
(2) (a) Henkel, T.; Rohr, J.; Beale, J. M.; Schwene, L. J. Antibiot.
1990, 43, 492. (b) Weber, S.; Zolke, C.; Rohr, J.; Beale, J. M. J. Org.
Chem. 1994, 59, 4211. (c) Zhu, L.; Ostash, B.; Rix, U.; Nur-e-Alam, M.;
Mayers, A.; Luzhetskyy, A.; Mendez, C.; Salas, J. A.; Bechthold, A.;
Fedorenko, V.; Rohr, J. J. Org. Chem. 2005, 70, 631.
(3) (a) Depenbrock, H.; Bornschlegl, S.; Peter, R.; Rohr, J.; Schmid,
P.; Schweighart, P.; Block, T.; Rastetter, J.; Kanauske, A. R. Ann.
Hematol. 1996, 73, A80/316. (b) Crow, R. T.; Rosenbaum, B.; Smith,
R., III; Guo, Y.; Ramos, K. S.; Sulikowski, G. A. Bioorg. Med. Chem. Lett.
1999, 9, 1663. (c) Korynevska, A.; Heffeter, P.; Matselyukh, B.; Elbling,
L.; Micksche, M.; Stoika, R.; Berger, W. Biochem. Pharmacol. 2007,
74, 1713. (d) Shaaban, K. A.; Srinivasan, S.; Kumar, R.; Damodaran, C.;
Rohr, J. J. Nat. Prod. 2011, 74, 2.
(4) (a) Ostash, B.; Rix, U.; Rix, L. L. R.; Liu, T.; Lombo, F.;
Luzhetskyy, A.; Gromyko, O.; Wang, C.; Bra~na, A. F.; Mꢀendez, C.;
Salas, J. A.; Fedorenko, V.; Rohr, J. Chem. Biol. 2004, 11, 547.
(b) Luzhetskyy, A.; Zhu, L.; Gibson, M.; Fedoryshyn, M.; D€urr, C.;
Hofmann, C.; Hoffmeister, D.; Ostash, B.; Mattingly, C.; Adams, V.;
Fedorenko, V.; Rohr, J.; Bechthold, A. ChemBioChem 2005, 6, 675.
(c) Zhu, L.; Luzhetskyy, A.; Luzhetska, M.; Mattingly, C.; Adams, V.;
Bechthold, A.; Rohr, J. ChemBioChem 2007, 8, 83.
good 78% yield (Scheme 4). Stronger conditions (e.g., a bit more
TBSOTf or higher temperature) led to cleavage of the 2,3,6-
trideoxy-R-glycosidic linkages and the C8ꢀO-20-deoxy-β-glyco-
sidic linkage. Removal of the bromide and iodide and the benzylic
protecting groups on the fragile landomycin A precursor 28
proved to be a big challenge. We first screened conditions on a
landomycinone disaccharide15 and then applied the optimized
conditions to hexasaccharide 28. Thus, hydrogenation (10 atm)
of 28 with Raney Ni followed by oxidation of the resulting
hydroquinone with DDQ afforded the desired deoxyhexasac-
charide 29 in 44% yield, with the bromide, three iodide, and three
benzylic groups being cleaved. Finally, the five acetyl groups
remaining on the sugar residue were removed with NaOMe,
providing landomycin A (1) in 67% yield. The analytical data of 1
were in good agreement with those reported for the natural
product.2a,c,8
In conclusion, the first total synthesis of landomycin A (1), the
longest and most potent antiumor congener of the angucycline
antibiotics, has been achieved in 63 steps and 0.34% overall yield
starting from 2,5-dihydroxybenzoic acid, 3,5-dimethylphenol,
triacetyl D-glucal, and D-xylose, with a convergent linear sequence
of 21 steps. The synthesis of other landomycin members and
their derivatives has become a feasible task that would facilitate
in-depth studies of their unusual spectrum of antitumor activities.
(5) Roush, W. R.; Neitz, R. J. J. Org. Chem. 2004, 69, 4906. In this
paper, Roush and Neitz described the synthesis of the structure originally
assigned to the aglycone of landomycin A (i.e., landomycinone) and
confirmed the structure of the synthetic sample by X-ray analysis. However,
comparison of spectroscopic data for synthetic landomycinone with
published data for the natural aglycone as well as with copies of 1H NMR
spectra provided by Prof. Rohr to Prof. Roush led these authors to conclude
that their synthetic material was not identical to naturally occurring land-
omycinone. However, a paper subsequently appeared from Rohr’s labora-
tory that provided a new set of 1H NMR data for natural landomycinone
that exactly matched the data obtained by Roush and Neitz for their
synthetic landomycinone (ref 2c). The latter information, together with our
successful total synthesis of landomycinone A described here, leaves no
doubt that Roush and Neitz indeed synthesized landomycinone.
(6) (a) Guo, Y.; Sulikowski, G. A. J. Am. Chem. Soc. 1998, 120, 1392.
(b) Roush, W. R.; Bennett, C. E. J. Am. Chem. Soc. 2000, 122, 6124.
(c) Yu, B.; Wang, P. Org. Lett. 2002, 4, 1919. (d) Tanaka, H.; Yamaguchi,
S.; Yoshizawa, A.; Takagi, M.; Shin-ya, K.; Takahashi, T. Chem.—Asian J.
2010, 5, 1407. (e) Kirschning, A. Eur. J. Org. Chem. 1998, 2267. (f) Zhou,
M.; O’Doherty, G. A. Org. Lett. 2008, 10, 2283. (g) McDonald, F. E.;
Reddy, K. S. J. Organomet. Chem. 2001, 617ꢀ618, 444.
(7) Mal, D.; Pahari, P. Chem. Rev. 2007, 107, 1892.
(8) See the Supporting Information for details.
(9) Couladouros, E. A.; Strongilos, A. T.; Papageorgiou, V. P.; Plyta,
Z. F. Chem.—Eur. J. 2002, 8, 1795.
(10) Corey, E. J.; Bakshi, R. K.; Shibata, S. J. Am. Chem. Soc. 1987,
109, 5551.
(11) (a) Roush, W. R.; Gung, B. W.; Bennett, C. E. Org. Lett. 1999,
1, 891. (b) Durham, T. B.; Roush, W. R. Org. Lett. 2003, 5, 1875.
(12) Roush, W. R.; Lin, X. F. J. Am. Chem. Soc. 1995, 117, 2236.
(13) Yu, B.; Yang, Z. Org. Lett. 2001, 3, 377.
(14) Lam, S. N.; Gervay-Hague, J. Org. Lett. 2003, 5, 4219.
(15) We had examined the glycosylation of landomycinone mono-
glycoside 14 with a relevant monosaccharide donor and found 3,4-di-O-
acetyl-2,6-dideoxy-2-iodo-D-glucopyranosyl trifluoroacetimidate to be
superior to afford the corresponding β-disaccharide in good yield (data
not shown).
’ ASSOCIATED CONTENT
S
Supporting Information. Experimental details, charac-
b
terization data, and NMR spectra for new compounds. This
material is available free of charge via the Internet at http://pubs.
acs.org.
’ AUTHOR INFORMATION
Corresponding Author
byu@mail.sioc.ac.cn
(16) (a) Yu, B.; Tao, H. Tetrahedron Lett. 2001, 42, 2405. (b) Yu, B.;
Sun, J. Chem. Commun. 2010, 46, 4668.
’ ACKNOWLEDGMENT
This work was supported by the NSFC (20932009 and
20921091) and the MOST of China (2010CB529706).
’ REFERENCES
(1) (a) Rohr, J.; Thiericke, R. Nat. Prod. Rep 1992, 9, 103. (b) Krohn,
K.; Rohr, J. Top. Curr. Chem. 1997, 188, 127.
12435
dx.doi.org/10.1021/ja205339p |J. Am. Chem. Soc. 2011, 133, 12433–12435