Synthesis of the amino sugar from C-1027.

Article abstract of DOI:10.1021/ol010119s

[reaction: see text] The amino sugar side chain (2) from C-1027Chr was synthesized in 12 steps from mannose. The key reactions are an internal displacement by nitrogen to introduce the cis amino group at C-4 and the methylation of an enolate at C-5.

Full text of DOI:10.1021/ol010119s

ORGANIC
LETTERS
2001
Vol. 3, No. 15
2403-2406
Synthesis of the Amino Sugar from
C-1027
M. F. Semmelhack,* Yu Jiang, and Douglas Ho†
Department of Chemistry, Princeton UniVersity, Princeton, New Jersey 08544
mfshack@princeton.edu
Received June 4, 2001
ABSTRACT
The amino sugar side chain (2) from C-1027Chr was synthesized in 12 steps from mannose. The key reactions are an internal displacement
by nitrogen to introduce the cis amino group at C-4 and the methylation of an enolate at C-5.
C-10271 is one of the more intriguing of the enediyne toxins
because it is exceedingly toxic against certain cell lines2 and
shows no simple chemical trigger to initiate its biological
activity.3 It belongs to the neocarzinostatin subfamily, where
the “chromophore” (C-1027Chr, 1) is bound to a small
protein4 before being transferred to the minor groove of DNA
during the activation process.5
“strap” bearing a phenol unit and a primary amine (C), and
a reduced amino sugar unit (D). The amino sugar is expected
to be important in the binding to DNA through interaction
of the protonated amino group with the backbone phosphate
of the DNA. Modeling suggests that the sugar protrudes from
the protein in C-1027 and may be important in the transfer
from the protein to the DNA.6
The sugar 2 (Scheme 1) is structurally interesting, with
the gem-dimethyl group at C-5 and the N,N-dimethyl amino
group at C-4 in cis arrangement with the OH substituents at
C-2 and C-3. There is a variety of pyranose analogues with
two carbon substituents at C-5, including more than 10 with
the same gem-dimethyl arrangement.7,8 In general, the
syntheses of such structures proceed by creating the gem-
Scheme 1. Retrosynthesis of the C-1027 Sugar
C-1027Chr is easily dissected into four subunits, a core
structure (A), a heteroaromatic side chain (B), a lactone
* Fax: 609 258-3409.
† Director, Princeton Small Molecule X-ray Facility.
(1) For the absolute configuration and leading references, see: Iida, K.-
i., Fukuda, S., Tanaka, T., Hirama, M. Tetrahedron Lett. 1996, 37, 4997.
(2) Zhen, Y.-S.; Ming, X.-Y.; Yu, B.; Otani, T.; Saito, H.; Yamada, Y.
J. Antibiot. 1989, XLII, 1294.
(3) Hirama, M. Pure Appl. Chem. 1997, 69, 525.
(4) (a) Okuno, Y.; Otsuka, M.; Sugiura, Y. J Med. Chem. 1994, 37, 2266.
(b) Motojima, F.; Inaka, K.; Minami, Y.; Otani, T.; Miki, K. Acta
Crystallogr. Sect. D 1994, 51, 1084.
(5) (a) Xu Y. J.; Zhen Y. S.; Goldberg, I. H. Biochemistry 1994, 33,
5947. (b) Kirk, C. A.; Goodisman, J.; Beerman, T. A. Biophys. Chem. 1997,
63, 201.
10.1021/ol010119s CCC: $20.00 © 2001 American Chemical Society
Published on Web 06/26/2001

dialkyl quaternary center at an early stage followed by ring
closure. The previous synthesis of 2 began with the gem-
dimethyl group in place using the epoxide (80% ee) from
asymmetric epoxidation of 3,3-dimethylallyl alcohol.9 That
synthesis involved about 17 steps and 1-2% overall yield,
with good stereoselectivity except in the step that introduced
the vicinal cis-diol grouping (C-2/C-3).
We viewed the target molecule as an interesting problem
in alkylation of a pyranose ring at C-5, an approach made
difficult by the presence of a potential leaving group at C-4
and the deactivating effect of the ring oxygen toward
generation of an enolate anion at C-5.10 Success in this
alkylation step would allow for direct variation of the
substitution pattern at C-5 in this and related sugars. Our
approach also takes advantage of the availability of common
sugars as single enantiomers. The retrosynthesis is sum-
marized in Scheme 1.
A key transform is the displacement of the C-4 hydroxyl
group of mannose (3) by an amine function, using an
intramolecular delivery and resulting in 4. Oxidation of C-6
sets up the molecule (5) for enolate generation and methyl-
ation to give 6.
The synthesis began with R-methyl-D-mannopyranoside
(3) of >99% enantiopurity.11 According to well-established
procedures,12 double acetonide formation followed by selec-
tive hydrolysis of one acetonide produced 7, Scheme 2.
which was quite sensitive toward handling at room temper-
ature. The internal SN2 reaction was initiated with KH in
THF at -30 °C and gave the bicyclic compound 4 as
colorless crystals of mp 153-154 °C, in 78% yield.14
In a direct approach, LiAlH4 reduction of 4 produced the
dimethyl amino alcohol 10 in quantitative yield. The oxida-
tion of 10 to 11 or 12 was difficult, and over 16 different
oxidation conditions were evaluated. Conventional conditions
for producing the carboxylic acid oxidation level (e.g., 11)
such as KMnO4 or K2Cr2O7 gave complex mixtures of the
products from oxidation and elimination. PDC,15 PCC,16 the
Dess-Martin reagent,17 Pd(II) under various conditions,18
Pt/C with O2,19 and photo-oxidation,20 also gave oxidation
products with elimination (13/14). Swern,21 Moffatt,21
TEMPO,22 and NCS/Me2S23 produced the aldehyde, but
always with elimination (14) and significant byproducts. For
example, Swern oxidation of 10 (oxalyl chloride, DMSO,
TEA, THF, -60 °C) gave 14 in 67% yield. Finally, the use
of catalytic Ru(II) with NMO as the stoichiometric oxidant24
under mildly basic conditions gave acceptable results,
producing the aldehyde 12 in 81% yield. Unfortunately, the
amino aldehyde 12 was quite unstable on handling and could
not be oxidized efficiently to the carboxylic acid or ester
under either basic [Ag(I)]25 or acidic oxidizing conditions.
To minimize loss of the amino group at C-4 during
oxidation, the cyclic carbamate 4 was cleaved with NaOH
to give the N-methyl amino alcohol 15 (99% yield) and the
amino group was modified with a Cbz group (16; Scheme
3). This protected version behaved well during Swern
Scheme 2. Internal Substitutiona
Scheme 3. Oxidation of Amino Alcohol 10
a (a) MeNCO, Et3N, 23 °C, 7 days, 98%; (b) Tf2O, CH2Cl2, 0
°C, 0.5 h, 90%; (c) KH, THF, 18-crown-6, -30 °C, 3 h, 78%.
Selective reaction with methyl isocyanate gave the carbamate
8, poised for internal displacement. The best yields (98%)
were obtained with long reaction times (7 days) at 23 °C;
attempts to accelerate the process with, for example, CuCl13
gave lower yields. The hydroxyl group at C-4 was activated
as the triflate (9; pale yellow needles with mp 67-71 °C),
oxidation to the aldehyde (17, 97%). Further oxidation to
the ester was best achieved with iodine under basic condi-
tions,26 giving ester 5 in about 70% yield overall for the two-
(6) Okuno, Y. Otsuka, M.; Sugiura, Y. J. Med. Chem. 1994, 37, 2266.
(7) For a 5,5-dimethyl sugar in a gyrase B inhibitor (novobiocin), see:
Laurin, P.; Ferroud, D.; Klich, M.; Dupuis-Hamelin, C.; Mauvais, P.;
Lassaigne, P.; Bonnefoy, A.; Musicki, B. Bioorg. Med. Chem. Lett. 1999,
2079. For a 5,5-dimethyl sugar in an antibiotic (lipiarmycin), see: Arnome,
A.; Nasini, G. J. Chem. Soc., Perkin Trans. 1 1987, 1353.
(8) One specific solution is the introduction of a hydroxymethyl group
(HCHO, NaOH, H2O) R to a CHO group at C-5: Mazur, A. W.; Hiler, G.
D. J. Org. Chem. 1997, 62, 4471.
(9) Lida, K.-I.; Ishii, T.; Hirama, M. Tetrahedron Lett. 1993, 34, 4079
(10) This difficulty has been discussed explicitly, e.g.: Werschkun, B.;
Theim, J. Synthesis 1999, 121. Also described in this paper is a Claisen
rearrangement approach to introduction of a quaternary center at C-4 of a
furanose derivative.
(14) This reaction was performed 15 times on a scale of 0.2-5 g. The
yields ranged from 95% to 65%, with an average of 78%.
(15) Corey, E. J.; Schmidt, G. Tetrahedron Lett. 1979, 32, 399.
(16) Corey, E. J.; Ensley, H. E. J. Am. Chem. Soc. 1975, 97, 6908.
(17) Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113, 7277.
(18) Tamaru, Y.; Yamamoto, Y.; Yamada, Y.; Yoshida, Z-i. Tetrahe-
dron Lett. 1979, 32, 1401.
(19) Pravdic, N.; Keglevic, C. Tetrahedron 1965, 21, 1897.
(20) Ottenheim, H. C. J.; DeMan, J. H. M. Synthesis 1975, 163.
(21) Mancuso, A. J.; Swern, D. Synthesis 1981, 165.
(22) TEMPO was used in stoichiometric amounts according to an early
procedure: Semmelhack, M. F.; Chou, C. S.; Cortes, D. A. J. Am. Chem.
Soc. 1983, 105, 4492.
(11) Purchased as 99% ee from the Aldrich Chemical Co.
(12) Evans, M.; Parrish, F. W. Carbohydr. Res. 1977, 54, 105.
(13) Duggan, M. E.; Imagire, J. S. Synthesis 1989, 131.
(23) Corey, E. J.; Kim, C. V. J. Am. Chem. Soc. 1972, 94, 7586.
2404
Org. Lett., Vol. 3, No. 15, 2001

stage oxidation; during large scale procedures, small amounts
of the epimer of 5 at C-5 were observed.
Scheme 5. Appearance of a â-Lactam During Dimethylation
Direct methylation of 5 to give 6 could be achieved in up
to 80% yield by enolate generation under carefully controlled
conditions; elimination of the N at C-4 was the competing
process. For example, by using lithium (tetramethylpiperi-
dide) (LiTMP), in THF at -78 °C, the elimination product
13 was obtained in 96% yield. With KN(SiMe3)2 added to a
mixture of 5 and excess MeI in THF/HMPA at -78 °C, the
desired methylation produced a mixture 6a and 6b (3:2) in
76% yield (averaged over 13 runs on various scales)
accompanied by minor amounts of the elimination product.
The mixture could not be separated. The final stage involved
reduction to the epimeric alcohols 18a and 18b, activation
as the mesylate (19a and 19b), and a second reduction to
produce 20. Removal of the acetonide unit completed the
synthesis of C-1027 sugar 2 (Scheme 4).27
Scheme 4. Completon of the Synthesisa
enolate with excess KN(SiMe3)2 followed by addition of
excess MeI led to double methylation and a mixture of
epimers, 22a and 22b, in 62% yield. The dimethylation is a
delicate reaction. For example, reaction of 5 with excess of
LiTMP at -78 °C followed by addition of MeI produced
three major products, the expected methylated products 22a
and 22b (ratio ca. 3:2, 42% yield) along with the â-lactam
23 (35% yield). The structure of the unexpected product 23
was confirmed by X-ray crystallography.28 We assume that
deprotonation of the amino group occurs first and that the
potassium base gives more rapid enolate formation, removing
the possibility of â-lactam formation. With slower proton
abstraction (LiTMP), deprotonation at nitrogen occurs first,
but ring closure to the lactam competes with the second
deprotonation at C-5; a final methylation gives 22a/b and
23.
a (d) NaOH/H2O, ∆, 4 h, 99%; (e) Cbz-Cl, K2CO3, dioxane-
water, 23 °C, 8 h, 95%; (f) DMSO, oxalyl chloride, TEA, THF,
-60 °C, 97%; (g) KOH, I2, MeOH, 45 min, 0 °C, 71%; (h)
KN(SiMe3)2, MeI, THF/HMPA, -78 °C, 45 min, 76% as mixture
of diastereomers; (i) LiAlH4, ether, 23 °C, 4 h, 96%; (j) MsCl,
NaH, ether, 0 °C, 2 h, 85%; (k) LiAlH4, ether, 23 °C, 4 h, 59%; (l)
HF, MeCN, 23 °C, 40 min, 79%.
A double methylation approach was also explored, by
using the unprotected amino group in place to inhibit
elimination during enolate generation. The Cbz group was
removed from 5 to give the amino-ester 21. Simultaneous
deprotonation of the amino group and generation of the
Partly to confirm the structures and also to recycle
material, 23 was treated with MeONa to give the methyl-
amine 24, and then N-methylation gave 22a as a single
isomer. The mixture 22a/22b underwent LiAlH4 reduction
to 18a/b, which are on the path to 2 in Scheme 5.
(24) Sharpless, K. B.; Akashi, K.; Oshima, K. Tetrahedron Lett. 1976,
29, 2503.
The most efficient pathway follows 12 steps (Schemes 2
and 3; steps a through l) in 10% overall yield with high
(25) Based on the procedure of Campaigne, E. and LeSuer, W. M. in
Organic Syntheses; Wiley: New York, 1963; Collect. Vol. IV, p 919.
(26) Yamada, S.; Morizone, D.; Yamamoto, K. Tetrahedron Lett. 1992,
45, 2503.
(28) A few other examples of a â-lactam fused to a pyranose have been
examined by X-ray crystallography, but none with the orientation of
substitutents (4-amino-5-acyl) of 23. Preparation by a Mitsunobu reaction
produced a 4-amino-3-acyl lactam: Zegrodka, O.; Abramski, W.; Urbanc-
zyk-Lipowska, Z.; Chmielewski, M. Carbohydr. Res. 1998, 307, 33.
Preparation by a cycloaddition to a glycal produced a 2-amino-3-acyl
lactam: Chmielewski, M.; Kaluza, Z.; Suwinska, K.; Rosenbaum, D.;
Duddek, H.; Magnus, P. D.; Huffman, J. C. Carbohydr. Res. 1990, 203,
183. Assembly of the pyranose ring onto a â-lactam gave a 3-amino-4-
acyl derivative: Hart, D. J.; Leroy, V.; Merriman, G. H.; Young, D. G. J.
J. Org. Chem. 1992, 57, 5670.
1
(27) Comparison with the natural material9 was primarily via H NMR
spectral data comparison. Synthetic: 1H NMR (400 MHz, CDCl3) δ 4.61
(1H, d, J ) 7.84, H1), 4.45 (1H, dd, J ) 2.95, 3.3 Hz, H3), 3.49 (3H, s,
OCH3), 3.25 (1H, dd, J ) 3.1, 7.5 Hz, H2), 2.45 (6H, s, NMe2), 2.30 (1H,
d, J ) 2.95 Hz, H4), 1.55 (3H, s, Me at C5), 1.30 (3H, s, Me at C5).
Natural: 1H NMR (400 MHz, CDCl3) δ 4.65 (1H, d, J ) 7.8, H1), 4.42
(1H, dd, J ) 2.8, 3.1 Hz, H3), 3.52 (3H, s, OCH3), 3.35 (1H, dd, J ) 3.1,
7.5 Hz, H2), 2.45 (6H, s, NMe2), 2.30 (1H, d, J ) 2.95 Hz, H4), 1.55 (3H,
s, Me at C5), 1.30 (3H, s, Me at C5).
Org. Lett., Vol. 3, No. 15, 2001
2405

diastereoselectivity following from a convenient enantio-
merically pure starting material. It demonstrates that, while
delicate, the alkylation of a C-5 enolate in this series is a
viable approach to creating the quaternary center.
Supporting Information Available: Experimental pro-
cedures and characterizations for compounds 2, 4, 5, 6a/b,
8, 10, 15, 16, 17, 18a, 20, 21, 22a, and 23 and the X-ray
data for 23. This material is available free of charge via the
Internet at http://pubs.acs.org.
Acknowledgment. We thank the National Institutes of
Health for financial support through research grant CA
54819.
OL010119S
2406
Org. Lett., Vol. 3, No. 15, 2001