The first total synthesis and establishment of absolute structure of luminacins C1 and C2

Article abstract of DOI:10.1016/S0040-4039(01)01657-4

Luminacins C₁ and C₂ (1 and 2), novel angiogenesis inhibitors, have been synthesized from a carbohydrate. The correlation of 1 and 2 confirms their structures to be different only at C1″ and establishes the relative and absolute conf

Full text of DOI:10.1016/S0040-4039(01)01657-4

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 7625–7628
The first total synthesis and establishment of absolute structure
of luminacins C1 and C2
Kuniaki Tatsuta,* Satoshi Nakano, Fumie Narazaki and Yusuke Nakamura
Department of Applied Chemistry, School of Science and Engineering, Waseda University, 3-4-1 Ohkubo, Shinjuku-ku,
Tokyo, 169-8555, Japan
Received 9 August 2001; revised 27 August 2001; accepted 31 August 2001
Abstract—Luminacins C1 and C2 (1 and 2), novel angiogenesis inhibitors, have been synthesized from a carbohydrate. The
correlation of 1 and 2 confirms their structures to be different only at C1%% and establishes the relative and absolute configurations.
© 2001 Elsevier Science Ltd. All rights reserved.
Luminacins C1 and C2 (1 and 2) were isolated from
Streptomyces sp. as novel angiogenesis inhibitors. Their
structures, including the relative configuration of the
carbohydrate portion, were determined mainly by
NMR studies.1 As a result, they were found to have the
same planar structure as SI-4228 and UCS15A, which
were reported to be microbial products showing antimi-
crobial, immunosuppressive,2 antitumor and bone
resorption inhibitory activities.3 However, the absolute
structure of luminacins C1 and C2 remained undeter-
mined. The significant clinical potential of luminacins
has stimulated considerable interest in the synthesis and
structure–function studies.
From both structural and retrosynthetic standpoints,
luminacins C1 and C2 (1 and 2) are reasonably expected
to be constructed from the aromatic segment 3 and the
carbohydrate segment 4 as outlined in Fig. 1. The
sensitive epoxide at C6%-8% is selectively introduced at
the latter stage of the synthesis by SN2 displacement.
The enantiopure segment 3 having the lone stereocenter
may be prepared from Grignard reaction of the benz-
aldehyde 5 with isopropylmagnesium bromide followed
by kinetic resolution. The segment 4 is prepared from
the exo-olefin 6 by ozonolysis followed by epimeriza-
tion at C5%.4 The olefin 6 can be synthesized from the
primary alcohol 7, which is derived from
L
-glucal 10
through successive Wittig olefinations.
Herein, we report the first total synthesis of luminacins
C1 and C2 (UCS15A) to disclose their absolute struc-
tures unambiguously.
Although we have synthesized all possible isomers,
especially at C1%%, C2% and the carbohydrate portion of
Figure 1.
* Corresponding author.
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01657-4

7626
K. Tatsuta et al. / Tetrahedron Letters 42 (2001) 7625–7628
luminacins C1 and C2 (1 and 2) to determine the absolute
structures, the synthesis of enantiomers (1% and 2%) of the
natural products is conveniently described from (R)- and
stereoselectively cyclized by intramolecular Michael reac-
tion to give 19 as a single product. We anticipated that
the cyclization would give the (3%S)-isomer having the
equatorial substituent at C3%, because of the presence of
the diaxial substituents at C5% and C7% due to strong
repulsion with an exo-olefin. In fact, the absolute struc-
ture was confirmed to have the 3%S-configuration by the
X-ray analysis of the derivative 29.7
(S)-3 (3R and 3S) and
10% is an inexpensive material. The natural products could
be synthesized by using -glucal 10.
D
-glucal 10% in detail. The latter
L
The synthesis of the chiral fragments required for the
assembly of 1% and 2% is outlined in Schemes 1 and 2.
As 19 could not be converted to the desired products 1%
and 2%, the configuration at C2% was epimerized for the
total synthesis by treatment with MeONa to give a 1:1
mixture of 7% and 19. The mixture was separated by
silica-gel column chromatography to afford 7% in 45%
yield. The absolute structure was also confirmed by the
X-ray analysis.7
On one hand, the aromatic segments 3R and 3S were
synthesized from the benzaldehyde 5. Methoxymethyla-
tion of 5 was followed by Grignard reaction to give the
racemic alcohol 11.5 The kinetic resolution of 11 was
effectively carried out by acylation with 0.8 equiv. of
(−)-camphanic chloride in THF at −78°C for 30 min to
give, after recrystallization from ether–hexane, the
diastereomerically pure (R)-isomer 12R in 45% yield (the
theoretical yield is 50%). The absolute structure was
determined by the X-ray analysis.6 The unreacted 11
having the (S)-enantiomer as the major component was
acylated with enantiomeric (+)-camphanic chloride to
give 12S in 41% yield. The enantiomerically pure 11R
([h]2D6+26° (c 0.99, MeOH)) and 11S ([h]2D6−25° (c 1.1,
MeOH)) were obtained from 12R and 12S by methanol-
ysis, respectively. The O-methyl derivative 13R was
regioselectively lithiated to react with formaldehyde,
followed by O-silylation to give 14R. This was iodinated
to the desired (R)-iodide 3R.
Dihydroxylation of the olefin 7% with OsO4 was expected
to occur at the less hindered site. Acetonation of the
resulting cis-diol was followed by hydrogenolysis to give
20.
We examined the displacement of the primary alcohol of
20 with the phenylselenide group in order to obtain an
exo-olefin, but the desired displacement failed to occur.
As the C5% and C7% substituents were present in the diaxial
positions and, moreover, the O-leaving group of C7%% was
outside of the tetrahydropyran ring, the C5% hydroxyl
group could readily attack the C7%% position to form a
furan ring. After extensive experimentation, we were able
to displace the primary alcohol under more forcing
conditions by preparation of the cyclic sulfate 21, where
the nucleophile could approach from outside of the pyran
ring. Compound 20 was first treated with SOCl2 followed
by oxidation to give 21.8 The latter reacted smoothly with
the phenylselenide anion to give, after removal of the
sulfate by acidic hydrolysis,9 the desired 22 in 93% yield.
This phenylselenide was exposed to H2O2 to afford the
exo-olefin 6% by elimination.
On the other hand, the carbohydrate segment 4% was
prepared from the
D
-glucal 10%. We found that de-O-
acetylation of 10% proceeded regioselectively by KOH in
DMF at 0°C to give 15. This was successively
methoxymethylated, de-O-acetylated and benzylated to
afford 16 in high yield. Methanolysis of 16 was followed
by Swern oxidation to the ketone 17, which was submit-
ted to Wittig olefination using n-PrPPh3Br and n-BuLi.
The resulting unsaturated methyl glycoside was
hydrolyzed to the anomeric mixture 9%. Wittig reaction
of 9% with 18 gave, after purification on silica-gel column
chromatography, the a,b-unsaturated ester 8%. This was
At this stage, the configuration of the hydroxyl group in
6% was changed to the desired one by oxidation and
MOMO
R1COO
OMOM
O
MOMO
R2COO
OMOM
MOMO
HO
OMOM
HO
OH
a, b
c
O
(R)
(S)
O
R1=
H
R2=
O
O
5
11
12R
12S
CH2OTBS
OMOM
CH2OTBS
OMOM
MOMO
OMOM
MOMO
X
MOMO
d
f, g
h
X
X
12R or 12S
I
e
11R: X=
11S: X=
OH
OH
13R: X=
13S: X=
OMe
OMe
14R: X=
14S: X=
3R: X=
3S: X=
OMe
OMe
OMe
OMe
Scheme 1. (a) MOMCl, i-Pr2NEt/DMF, rt; (b) i-PrMgBr/THF, 0°C, 83% two steps; (c) (−)-camphanic chloride, Py/THF, −78°C,
45%; (d) KOH/MeOH, rt, 92%; (e) MeI, NaH/DMF, rt, 96%; (f) (HCHO)n, s-BuLi, TMEDA/THF, −78°C; (g) TBSCl,
imidazole/DMF, rt, 67% two steps; (h) NIS, s-BuLi/THF, −78°C, 80%.

K. Tatsuta et al. / Tetrahedron Letters 42 (2001) 7625–7628
7627
Scheme 2. (a) KOH/DMF, 0°C, 69%; (b) MOMCl, i-Pr2NEt/PhMe, rt; (c) NH3/MeOH, 40°C; (d) BnCl, t-BuOK/DMF, rt, 95%
three steps; (e) Ph3P·HBr, MeOH/PhMe, rt, 89%; (f) (COCl)2, DMSO, Et3N/CH2Cl2, −78 to 0°C, 96%; (g) n-PrPPh3Br,
n-BuLi/THF, −78°C to rt, 84%; (h) aq. AcOH, 50°C, 80%; (i) PhMe, 100°C, 76%; (j) MeONa (cat.)/PhMe, 70°C, 80%; (k)
MeONa/MeOH, 60°C, 45%; (l) OsO4, NMO/aq. MeCN, rt, quant.; (m) 2-methoxypropene, PPTS/CHCl3, rt, quant.; (n) H2,
Pd(OH)2-C/EtOH, rt, quant.; (o) SOCl2, Et3N/CH2Cl2, −78°C; (p) RuCl3·nH2O, NaIO4/CCl4–MeCN–H2O, rt; (q) PhSeNa/THF,
0°C; (r) aq. H2SO4/THF, rt; (s) aq. H2O2, NaHCO3/THF, rt to 50°C, 67% five steps; (t) TPAP, NMO, MS4A/CH2Cl2, rt; (u)
LiAlH4/THF, 0°C to rt; (v) TBDPSCl, imidazole/DMF, rt, 77% three steps; (w) BnBr, NaH/DMF, rt, 95%; (x) O3/CH2Cl2, −78°C
then Ph3P, 80%; (y) TFA/CH2Cl2, rt, 80%; (z) Tf2O, Py/CH2Cl2, −78°C; (aa) DIBAL-H/PhMe, −78°C; (bb) BnBr, NaH/DMF,
rt; (cc) TBAF/THF, rt; (dd) (COCl)2, DMSO, Et3N/CH2Cl2, −78°C to 0°C, 88% five steps.
hydride reduction. Simultaneously, the ester group of 6%
was reduced to a primary alcohol, which was selectively
silylated to 23. O-Benzylation of 23 was followed by
ozonolysis and de-acetonation to give the lactone 24.
This was converted by SN2 displacement to the epoxide
and reduced with DIBAL to the lactol 25. O-Benzyla-
Scheme 3. (a) n-BuLi/THF, −78 to 0°C; (b) TPAP, NMO, MS4A/CH2Cl2, rt, 70% two steps; (c) H2, Pd(OH)2-C/EtOH, rt, 45%;
(d) THF–AcOH–H2O, 60°C, 65%; (e) MnO2/CHCl3, rt; (f) THF–AcOH–H2O, 60°C, 61% two steps.

7628
K. Tatsuta et al. / Tetrahedron Letters 42 (2001) 7625–7628
tion of 25 to give exclusively the benzyl b-glycoside 26
was followed by de-O-silylation and oxidation of the
resulting alcohol to the aldehyde 4%.
(1H, s, CHO-1), 12.97 (1H, s, OH-6), 14.16 (1H, s, OH-2).
2%: 3.27 (1H, dd, J=7.0 and 6.0, H-8%), 3.65 (1H, ddd,
J=10.0, 7.5 and 4.0, H-2%), 4.18 (1H, br dd, 12.0 and 4.5,
H-5%), 4.40 (1H, ddd, J=11.5, 7.5 and 1.5, H-3%), 4.95 (1H,
s, H-7%), 8.04 (1H, s, H-4), 10.42 (1H, s, CHO-1), 12.97
(1H, s, OH-6), 14.18 (1H, s, OH-2). 3R and 3S: 4.33 (1H,
d, J=6.5, H-1%%), 4.69 (1H, d, J=10.5, CH2-1), 4.73 (1H,
d, J=10.5, CH2-1), 7.76 (1H, s, H-4). 4%: 3.30 (1H, t,
J=6.5, H-8%), 4.12 (1H, dd, J=11.5 and 4.5, H-5%), 4.29
(1H, ddd, J=12.0, 5.0 and 2.0, H-3%), 4.55 (1H, s, H-7%),
9.76 (1H, d, J=2.0, H-1%). 6% [in (CD3)2CO]: 3.64 (3H, s,
CO2Me), 4.02 (1H, ddd, J=4.0, 4.0 and 2.0, H-5%), 4.14
(1H, ddd, J=11.0, 8.5 and 3.0, H-3%), 4.22 (1H, dd,
J=10.5 and 3.0, H-8%), 4.59 (1H, d, J=1.0, H-7%%), 4.61
(1H, d, J=1.0, H-7%%). 7%: 3.67 (3H, s, CO2Me), 4.15 (1H,
ddd, J=11.5, 7.5 and 2.0, H-3%), 4.43 (1H, dd, J=3.5 and
3.0, H-5%), 5.60 (1H, t, J=7.5, H-8%). 12R and 12S: 0.90,
1.05, 1.10 (3H, each s, camphanyl Me), 6.06 (1H, d,
J=7.0, H-1%%), 6.66 (1H, dd, J=8.0 and 2.0, H-3), 6.79
(1H, d, J=2.0, H-1), 7.16 (1H, d, J=8.0, H-4). 14R and
14S: 4.35 (1H, d, J=6.5, H-1%%), 4.68 (1H, d, J=10.5,
CH2-1), 4.76 (1H, d, J=10.5, CH2-1), 6.95 (1H, d, J=9.0,
H-3), 7.24 (1H, d, J=9.0, H-4). 21: 3.70 (3H, s, CO2Me),
4.27 (1H, dd, J=12.0 and 2.5, H-8%), 4.58 (1H, ddd,
J=12.0, 6.5 and 3.5, H-3%), 4.63 (1H, dd, J=13.5 and 2.5,
H-7%%), 4.70 (1H, dd, J=13.5 and 1.5, H-7%%), 4.80 (1H, dd,
J=3.5 and 2.5, H-5%). 24: 3.70 (1H, dd, J=10.5 and 4.5,
H-1%), 3.73 (1H, dd, J=10.5 and 6.5, H-1%), 3.83 (1H, dd,
J=10.0 and 2.5, H-8%), 3.92 (1H, dd, J=6.5 and 6.5, H-5%),
4.69 (1H, ddd, J=8.0, 8.0 and 4.5, H-3%). (1%%R,2%R)-27:
3.29 (1H, t, J=6.5, H-8%), 3.62 (1H, ddd, J=8.5, 8.0 and
4.0, H-2%), 4.12 (1H, dd, J=12.0 and 4.5, H-5%), 4.29 (1H,
ddd, J=11.5, 8.0 and 1.5, H-3%), 4.54 (1H, s, H-7%), 4.69
(1H, d, J=10.5, CH2-1), 4.76 (1H, d, J=10.5, CH2-1),
7.61 (1H, s, H-4). (1%%S,2%R)-27: 3.28 (1H, t, J=6.5, H-8%),
3.62 (1H, ddd, J=8.0, 8.0 and 4.0, H-2%), 4.12 (1H, dd,
J=12.0 and 4.5, H-5%), 4.28 (1H, ddd, J=11.5, 8.0 and
1.5, H-3%), 4.54 (1H, s, H-7%), 4.71 (1H, d, J=10.5, CH2-1),
4.78 (1H, d, J=10.5, CH2-1), 7.62 (1H, s, H-4).
With sufficient quantities of enantio-pure segments 3R,
3S and 4% in hand, we turned to their combination as
shown in Scheme 3. Lithiation of 3R followed by
addition of 4% yielded the alcohol, which was oxidized to
the ketone (1%%R,2%R)-27. Deprotection by hydrogenoly-
sis and hydrolysis gave the mono-O-MOM derivative
28. The primary alcohol was oxidized to the aldehyde,
followed by removal of the O-MOM group to give
(+)-luminacin C1 (1%) ([h]2D6+97° (c 0.47, CHCl3)). When
treated with the enantiomer 3S, compound 4% gave
similarly (1%%S,2%R)-27. This was converted into (−)-lumi-
nacin C2 (2%) ([h]2D6−54° (c 0.17, CHCl3)) by similar
procedures through (1%%S,2%R)-28 as described above.
Both synthetic products 1% and 2% were identical in
NMR, IR and mass spectral analyses with natural
luminacins C1 and C2 (1 and 2), although the signs of
their optical rotations were completely opposite to those
of the natural products.10 Finally, luminacins C1 and C2
(1 and 2) could be synthesized from -glucal 10 with 3S
L
and 3R, respectively, by the same procedures described
above. Thus, the correlation of 1 and 2 confirmed their
structures to be different only at C1%% and established the
relative and absolute configurations to be as shown in
Fig. 1.
Acknowledgements
We are grateful to the financial support by Grant-in-Aid
for Specially Promoted Research from the Ministry of
Education, Culture, Sports, Science and Technology.
6. The X-ray data of 12R has been deposited with CCDC
(CCDC 165798).
References
7. The derivative 29 was prepared from 7% as shown in
Scheme 4. The X-ray data of 29 has been deposited with
CCDC (CCDC 167727).
1. Naruse, N.; Kageyama-Kawase, R.; Funahashi, Y.; Wak-
abayashi, T.; Watanabe, Y.; Sameshima, T.; Dobashi, K.
J. Antibiot. 2000, 53, 579.
2. Suzuki, M.; Kobayashi, I.; Mitsutake, K. Jpn. Kokai
O
OBn
O
O
S
Tokkyo Koho 1983, 116, 686.
O
3. (a) Tamaoki, T.; Sugawara, K.; Hamada, M.; Nakano,
H.; Mizukami, T.; Yamashita, Y.; Kosaka, N.; Sugawara,
T. Jpn. Kokai Tokkyo Koho 1996, 268, 888; (b) Sharma, S.
V.; Oneyama, C.; Yamashita, Y.; Nakano, H.; Sugawara,
K.; Hamada, M.; Kosaka, N.; Tamaoki, T. Oncogene
2001, 20, 2068.
CO2Me
H
CO2Me
(R)
a - d
O
O
O
H
(S)
BnO
29
7'
Scheme 4. (a) mCPBA/CHCl3, rt, 95%; (b) H2, Pd(OH)2–C/
EtOH, rt, quant.; (c) SOCl2, Et3N/CH2Cl2, rt; (d)
RuCl3·nH2O, NaIO4/CCl4–MeCN–H2O, rt, 80% two steps.
4. The carbon-numbering protocol parallels conveniently
that of the natural product 1.
5. All compounds were purified by silica-gel column chro-
matography and/or recrystallization, and were fully char-
acterized by spectroscopic means. 1H NMR (600 MHz: l,
ppm from TMS, and J in Hz) spectra were in CDCl3
solution, unless otherwise stated. Significant 1H NMR
spectral data are the following. 1%: 3.27 (1H, dd, J=7.0
and 5.5, H-8%), 3.60 (1H, m, H-2%), 4.18 (1H, ddd, J=12.0,
12.0 and 5.0, H-5%), 4.39 (1H, ddd, J=11.5, 7.0 and 1.5,
H-3%), 4.96 (1H, d, J=2.5, H-7%), 8.08 (1H, s, H-4), 10.41
8. Gao, Y.; Sharpless, K. B. J. Am. Chem. Soc. 1988, 110,
7538.
9. Kim, B. M.; Sharpless, K. B. Tetrahedron Lett. 1989, 30,
655.
10. Authentic samples of luminacins C1 (1: ([h]2D6−98° (c 0.28,
CHCl3)) and C2 (2: UCS15A; ([h]2D6+54° (c 0.13, CHCl3))
were kindly provided by K. Dobashi, Mercian Corp. and
Y. Kanda, Kyowa Hakko Kogyo Co., Ltd.