Facile construction of N-hydroxybenzazocine: Enantioselective total synthesis of (+)-FR900482

Article abstract of DOI:10.1002/anie.200290016

Intramolecular hydroxylamination of ω-formyl nitrobenzene 1 followed by stereoselective hydroxymethylation led to the formation of N-hydroxybenzazocine 2, a key intermediate in the enantioselective total synthesis of the antitumor antibiotic FR900482 (3).

Full text of DOI:10.1002/anie.200290016

COMMUNICATIONS
Facile Construction of N-Hydroxybenzazocine:
Enantioselective Total Synthesis of (þ)-
FR900482**
O
O
OCONH2
OH
BnO
HO
O
O
O
NH
N
OHC
N
Masashi Suzuki, Mika Kambe, Hidetoshi Tokuyama,
and Tohru Fukuyama*
MPO
1
3
OR"
O
RO
RO
The antitumor antibiotic FR900482 (1) was isolated from
Streptomyces sandaensis No. 6897 by Imanaka et al. at the
Fujisawa Pharmaceutical Co.[1] Biological studies have re-
vealed that this and the related compounds exhibit the same
level of antitumor activities as mitomycin C (2).[2] Extensive
O
O
MeO2C
MeO2C
NO2
N
O
R'O
5
4
O
O
O
O
OCONH2
OMe
OCONH2
OH
O
HO
RO
H2N
Me
RO
8
OP
X
+
OP
N
NH
O
NH
OHC
N
MeO2C
NO2
O
MeO2C
NO2
6
7
FR900482
mitomycin C 2
Scheme 1. Retrosynthesis of FR900482 (1). Bn
methoxyphenyl.
¼
benzyl, MP
¼
p-
1
investigations of the structure activity relationships and its
mode of action have revealed that this class of compounds
crosslink DNA in a fashion analogous to mitomycin C.[3] In
addition to these promising biological activities, the structure
of 1 featuring the unique hydroxylamine hemiacetal has made
it an attractive target for synthetic chemists. Although
numerous approaches[4] have been explored to construct this
densely functionalized structure, only three total syntheses[5,6]
and a formal total synthesis[7] have been reported to date.[8]
After the completion of our first total synthesis of racemic
1,[5a] we have devoted continuous efforts to establish a more
efficient route to prepare the optically active FR900482 (1).[9]
We report herein a stereocontrolled, enantioselective total
BnO
OTf
O
O
O
BnO
MeO2C
NO2
O
10
OTBS
a
OTBS
MeO2C
NO2
9
11
O
N
BnO
BnO
O
O
b
O
O
MeO2C
NO2
MeO2C
NO2
OTBS
OTBS
13
12
synthesis of
1 through a facile construction of the
OTIPS
OR1
OTIPS
N-hydroxybenzazocine intermediate.
BnO
BnO
Our synthetic plan is outlined in Scheme 1. For the
construction of the key intermediate N-hydroxybenzazocine
4, we planned to exploit intramolecular reductive hydroxyla-
mination of a fully functionalized w-formyl nitrobenzene
derivative 5. Hydroxymethylation and subsequent hydroxyl-
amine hemiacetal formation would lead to the pentacyclic
intermediate 3 in our racemic total synthesis. Cyclization
precursor 5 would be accessible from aryl acetylene 6, which
in turn would be obtained by coupling of the aromatic
fragment 7 and the terminal acetylene 8.
c–e
h
O
OR2
MeO2C
NO2
MeO2C
i
NO2
OR3
OR
14: R1, R2, R3 = H
17: R = TBS
18: R = H
f
15: R1, R2 = H, R3 = TBS
g
16: R1 = H, R2 = Ts, R3 = TBS
Scheme 2. Synthesis of epoxy alcohol 18. a) [Pd(OAc)2] (0.1 equiv), PPh3
(0.2 equiv), THF/NEt3 (2:1 v/v), 658C, 1 h, then room temperature, 12 h,
75%; b) pyrrolidine (2 equiv), benzene, room temperature, 1 h, then
aqueous AcOH (50%), room temperature, 2 h; c) Zn(BH4)2 (1.2 equiv),
Et2O, À308C, 3 h, 94% (2 steps), 9:1 diastereoselectivity; d) TIPSOTf
(3 equiv), 2,6-lutidine (6 equiv), CH2Cl2, room temperature, 7 h; e) AcOH/
H2O (5:1 v/v), 1008C, 4 h, 61% (2 steps); f) TBSCl (1.2 equiv), NEt3
(2.4 equiv), DMAP (0.1 equiv), CH2Cl2, room temperature, 13 h; g) TsCl
(1.2 equiv), DABCO (2 equiv), CH2Cl2, room temperature, 1.5 h; h) NaH
(1.5 equiv), DMF, 08C!RT, 0.5 h, 76% (3 steps); i) CSA (0.1 equiv),
Preparation of the epoxy alcohol precursor 18 commenced
with Sonogashira-coupling of acetylene 9[10] and aryl triflate
10[5b] to provide aryl acetylene 11 (Scheme 2).[11,12] At this
juncture, it was necessary to devise a regioselective trans-
MeOH, room temperature, 1 h. TIPS
¼
triisopropylsilyl, OTf
¼
[*] Prof. Dr. T. Fukuyama, M. Suzuki, M. Kambe, Dr. H. Tokuyama
Graduate School of Pharmaceutical Sciences, The University of Tokyo
7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033 (Japan)
Fax : (þ 81)3-5802-8694
trifluoromethanesulfonate, TBS ¼ tert-butyldimethylsilyl; DMAP ¼ 4-
dimethylaminopyridine, Ts ¼ p-toluenesulfonyl, DABCO ¼ 1,4-diazabi-
cyclo[2.2.2]octane, DMF
phorsulfonic acid.
¼
N,N-dimethylformamide, CSA
¼
10-cam-
E-mail: fukuyama@mol.f.u-tokyo.ac.jp
[**] This work was supported by CREST, the Japan Science and
Technology Corporation (JST), and by the Grant-in-Aid for Scientific
Research on Priority Areas (A) ™Exploitation of Multi-Element
Cyclic Molecules∫ from the Ministry of Education, Culture, Sports,
Science and Technology, Japan.
formation of the acetylene into the required ketone under
mild conditions. To this end, we developed a novel conjugate
addition of secondary amines to ortho-nitroaryl acetylenes.
4686
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4124-4686 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2002, 41, No. 24

COMMUNICATIONS
Thus, addition of pyrrolidine proceeded smoothly at room
temperature to furnish intermediate enamine 12, which upon
treatment with AcOH/H2O (50%) in a one-pot procedure
gave the desired ketone 13 in excellent yield. After stereo-
selective reduction of the ketone[13] and protection of the
resultant alcohol, both the acetonide and the TBS group were
removed by heating in aqueous acetic acid to give triol 14. The
desired epoxide was then obtained by a three-step sequence:
TBS protection of the primary alcohol, tosylation of the
sterically less-hindered secondary alcohol,[14] and treatment
with NaH. Finally, selective deprotection of the TBS group
afforded epoxy alcohol 18.
For the ensuing hydroxymethylation and hemiacetal for-
mation, we developed a sequential one-pot procedure.
Hydroxymethylation was best effected by treatment of ketone
20 with formalin in the presence of a catalytic amount of
LiOH to furnish the desired 21 with high diastereoselectivity
(94:6).[15] Acidification of the reaction mixture with HCl (1n)
afforded hemiacetal 22,[16] which was subjected to acetonide-
formation conditions to give the pentacyclic compound 23 in
56% yield from ketone 20.[17] Acetonide 23 was then reduced
with DIBAL, and the resultant benzyl alcohol 24 was
protected as the p-methoxyphenyl ether to give the pentacy-
clic compound 3.
Having synthesized epoxy alcohol 18 in a straightforward
manner, we then focused on the facile construction of the
N-hydroxybenzazocine 19. Alcohol 18 was oxidized with
Dess Martin periodinane to the corresponding aldehyde,
which was then subjected to a variety of reductive conditions
for construction of the desired N-hydroxybenzazocine. After
numerous attempts, we found that catalytic hydrogenation
over Pt/C (5%) in MeOH cleanly afforded N-hydroxybenza-
zocine 19 as the sole product (89% overall yield from 17). No
product of further reduction was observed during the hydro-
genation. Protection of the hydroxylamine of 19 as the
1-methoxy-1-methylethyl ether followed by deprotection of
the TIPS group, and Swern oxidation furnished ketone 20
(Scheme 3).
With the key intermediate 3 in hand, we completed the total
synthesis of optically active FR900482 (1) by modifying the
protocol established during our racemic synthesis.[5a] Thus,
regioselective opening of the epoxide 3 with LiN3 and
mesylation of the resultant alcohol gave acetonide 25
(Scheme 4). Conversion of 25 into hydroxy carbonate 26
was effected by a three-step sequence involving acidic
hydrolysis of the acetonide, treatment with triphosgene, and
deprotection of the p-methoxyphenyl group with ceric
ammonium nitrate.[18] The resultant alcohol 26 was oxidized
to the aldehyde, which was protected as the dimethyl acetal.[19]
After formation of the aziridine by heating with PPh3 in the
presence of iPr2NEt, hydrogenolysis of the benzyl ether
followed by treatment with HClO4 in THF/H2O afforded
aldehyde 28. Finally, ammonolysis of the cyclic carbonate
provided exclusively the desired FR900482 (1), whose spec-
OTIPS
O
BnO
OTIPS
O
BnO
a, b
c–e
MeO2C
MeO2C
NO2
N
O
O
O
O
OH
BnO
BnO
BnO
HO
OH
19
18
O
O
c–e
a, b
OMs
N3
OH
O
O
O
BnO
BnO
N
N
7
MPO
MPO
9
3
25
f
O
O
MeO2C
MeO2C
N
O
N
O
O
O
O
O
O
OMe
OMe
BnO
20
21
O
O
i, j
f–h
OMs
N3
O
MeO
NH
N
N
O
OH
OH
BnO
BnO
HO
MeO
O
26
27
g
O
O
O
O
R
N
MeO2C
N
O
O
O
OCONH2
OH
HO
23: R = CO2Me
24: R = CH2OH
3: R = CH2OMP
22
h
i
O
k
NH
O
NH
OHC
N
OHC
N
Scheme 3. Construction of the pentacyclic compound 3. a) Dess Martin
periodinane (1.4 equiv), CH2Cl2, 08C!RT, 0.5 h; b) H2 (1 atm), Pt/C (5%;
15 wt%), MeOH, room temperature, 2 h, 89% (from 17); c) 2-methoxy-
propene (22 equiv), TsOH¥H2O (0.1 equiv), CH2Cl2, room temperature,
10 min; d) TBAF (3.5 equiv), THF, room temperature, 12 h, 85% (2 steps);
e) (COCl)2 (2 equiv), DMSO (4 equiv), CH2Cl2, À788C, 0.5 h, then NEt3
(6 equiv), À788C!RT, 0.5 h, 82%; f) aqueous HCHO (37%; 115 equiv),
LiOH (0.4 equiv), THF/H2O (20:3 v/v), 08C, 5 h, then HCl (1n; 2 equiv),
08C!RT, 14 h; g) 2-methoxypropene (5 equiv), PPTS (0.1 equiv), 2,2-
dimethoxypropane/acetone (1:1 v/v), room temperature, 3 h; separation of
the isomers, 56% (from 20); h) DIBAL (3 equiv), CH2Cl2, À788C, 1 h,
99%; i) 4-methoxyphenol (2 equiv), PPh3 (2 equiv), DEAD (2 equiv),
benzene, room temperature, 15 min, 96%. TBAF ¼ tetrabutylammonium
fluoride, DMSO ¼ dimethyl sulfoxide, PPTS ¼ pyridinium p-toluenesul-
28
1
Scheme 4. Completion of the total synthesis of 1. a) LiN3 (27 equiv), DMF/
H2O (10:1 v/v), 1208C, 3.5 h, 83%; b) MsCl (2 equiv), NEt3 (3 equiv),
CH2Cl2, room temperature, 2.5 h, 80%; c) TFA (8 equiv), CH2Cl2, room
temperature, 3 h; d) (Cl3CO)2C ¼ O (5 equiv), pyridine (6 equiv), CH2Cl2,
08C, 30 min, 92% (2 steps); e) (NH4)2Ce(NO3)6 (2.5 equiv), MeCN/H2O
(4:1 v/v), room temperature, 10 min, 84%; f) PCC (2 equiv), MgSO4
(4 equiv), CH2Cl2, room temperature, 1.5 h; g) CSA (0.08 equiv),
CH(OMe)3/MeOH (1:4 v/v), room temperature, 1 h, 81% (2 steps);
h) PPh3 (2 equiv), iPr2NEt (1.2 equiv), THF/H2O (10:1 v/v), 608C, 1.5 h,
85%; i) H2 (1 atm), Pd/C (10%; 15 wt%), EtOH, room temperature, 2.5 h;
j) HClO4 (1%; 0.2 equiv), THF/H2O (10:1 v/v), room temperature, 5 h;
k) NH3 (gas), THF, room temperature, 3 h, 89% (3 steps). Ms
methanesulfonyl, TFA ¼ trifluoroacetic acid, PCC ¼ pyridinium chloro-
¼
fonate, DIBAL
azodicarboxylate.
¼
diisobutylaluminum hydride, DEAD
¼
diethyl
chromate.
Angew. Chem. Int. Ed. 2002, 41, No. 24
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4124-4687 $ 20.00+.50/0
4687

COMMUNICATIONS
unambiguously confirmed by observation of NOE interactions
between 7-H and 9-H.
[16] In addition to 22, formation of a side product, which was tentatively
assigned as hydroxylamine hemiacetal diastereomer 22’, was ob-
served. This mixture was subjected to the next acetonide formation
without separation.
tral data were completely identical with those reported in
literature.[1b]
In conclusion, we have completed a highly efficient total
synthesis of FR900482 (1). The present synthesis features a
facile formation of N-hydroxybenzazocine by intramolecular
reductive hydroxylamination and an ensuing facile construc-
tion of the hydroxylamine hemiacetal. The synthetic strategy
described above should be applicable to the synthesis of
analogues of FR900482 as well as of other benzazocine
derivatives. Application of this approach to the synthesis of
mitomycin C is currently under investigation in our laborato-
ries.
[17] The ratio of 22/22’ was almost the same as that of 23 and a side
product, which was tentatively assigned as 23’. Furthermore, depro-
tection of the acetonide of 23 and 23’ under acidic conditions (HCl
(1n) in THF, room temperature) gave only 22 and 22’, respectively.
These observations would indicate that neither epimerization of C7
nor interconversion of the hemiacetal diastereomers via the eight-
membered ring ketone occurred during the acetonide formation.
Received: August 6, 2002 [Z19901]
[1] a) M. Iwami, S. Kiyoto, H. Terano, M. Kohsaka, H. Aoki, H. Imanaka,
J. Antibiot. 1987, 40, 589; b) S. Kiyoto, T. Shibata, M. Yamashita, T.
Komori, M. Okuhara, H. Terano, M. Kohsaka, H. Aoki, H. Imanaka,
J. Antibiot. 1987, 40, 594; c) I. Uchida, S. Takase, H. Kayakiri, S.
Kiyoto, M. Hashimoto, J. Am. Chem. Soc. 1987, 109, 4108.
[2] K. Shimomura, O. Hirai, T. Mizota, S. Matsumoto, J. Mori, F.
Shibayama, H. Kikuchi, J. Antibiot. 1987, 40, 600.
[3] a) R. M. Williams, S. R. Rajski, S. B. Rollins, Chem. Biol. 1997, 4, 127;
b) M. M. Paz, P. B. Hopkins, J. Am. Chem. Soc. 1997, 119, 5999;
c) R. M. Williams, S. B. Rollins, S. R. Rajski, J. Am. Chem. Soc. 1998,
120, 2192.
[18] T. Fukuyama, A. A. Laird, L. M. Hotchkiss, Tetrahedron Lett. 1985,
26, 6291.
[19] The aldehyde was protected as the dimethyl acetal to prevent
reduction during hydrogenolysis of the phenolic benzyl ether.
[4] For representative examples, see: a) N. Yasuda, R. M. Williams,
Tetrahedron Lett. 1989, 30, 3397; b) R. J. Jones, H. Rapoport, J. Org.
Chem. 1990, 55, 1144; c) S. J. Miller, S. H. Kim, Z. R. Chen, R. H.
Grubbs, J. Am. Chem. Soc. 1995, 117, 2108; d) H. J. Lim, G. A.
Sulikowski, Tetrahedron Lett. 1996, 37, 5243; e) S. Mithani, D. M.
Drew, E. H. Rydberg, N. J. Taylor, S. Mooibroek, G. I. Dmitrienko, J.
Am. Chem. Soc. 1997, 119, 1159.
[5] a) T. Fukuyama, L. Xu, S. Goto, J. Am. Chem. Soc. 1992, 114, 383;
b) J. M. Schkeryantz, S. J. Danishefsky, J. Am. Chem. Soc. 1995, 117,
4722.
Total Synthesis of (ꢀ )-FR66979**
Richard Ducray and Marco A. Ciufolini*
In the late 1980s, scientists at the Fujisawa Co. (Japan)
unveiled a new class of antitumor agents with general
structure 1 (Scheme 1).[1] These substances, denoted FR-
66979 (1a) and FR-900482 (1b), are structurally related to the
mitomycins (see mitomycin C (2)).[2] Indeed, the two families
of anticancer agents possess comparable bioactivity[3] and are
believed to act by a similar mechanism, yet FR-type
compounds are less toxic than mitomycins, probably as a
result of the absence of a quinoid nucleus.[4] Derivatives of 1b
are currently undergoing clinical trials.[5]
[6] a) T. Katoh, E. Itoh, T. Yoshino, S. Terashima, Tetrahedron 1997, 53,
10229; b) T. Yoshino, Y. Nagata, E. Itoh, M. Hashimoto, T. Katoh, S.
Terashima, Tetrahedron 1997, 53, 10239; c) T. Katoh, Y. Nagata, T.
Yoshino, S. Nakatani, S. Terashima, Tetrahedron 1997, 53, 10253.
[7] I. M. Fellows, D. E. Kaelin, Jr., S. F. Martin, J. Am. Chem. Soc. 2000,
122, 10781.
[8] After submission of this manuscript, we learned that two groups have
independently completed enantioselective total syntheses; see the
previous communication: a) T. C. Jude, R. M. Williams, Angew. Chem.
2002, 114, 4877; Angew. Chem. Int. Ed. 2002, 41, 4683; and the
following communication: b) R. Ducray, M. A. Ciufolini, Angew.
Chem. 2002, 114, 4882; Angew. Chem. Int. Ed. 2002, 41, 4688.
[9] For an approach utilizing intramolecular [3þ2] cycloaddition of nitrile
oxide, see: M. Kambe, E. Arai, M. Suzuki, H. Tokuyama, T.
Fukuyama, Org. Lett. 2001, 3, 2575.
The biomedical potential and unusual architecture of
compounds 1 have stimulated substantial interest at a
[10] G. Pandey, M. Kapur, Tetrahedron Lett. 2000, 41, 8821. Alternatively,
we prepared 9 by a modified procedure: 2,3-di-O-isopropylidene-l-
threitol, NaH, TBSCl, THF; cat. TEMPO (2,2,6,6-tetramethyl-1-
piperidinyloxy, free radical), PhI(OAc)2, CH2Cl2; dimethyl-1-diazo-2-
oxopropylphosphonate, K2CO3, MeOH (S. M¸ller, B. Liepold, G. J.
Roth, H. J. Bestmann, Synlett 1996, 521).
[11] K. Sonogashira, Y. Tohda, N. Hagihara, Tetrahedron Lett. 1975, 16,
4467.
[12] The coupling reaction under typical conditions ([PdCl2(PPh3)2] and
CuI in Et3N) gave a considerable amount of the homocoupling by-
product of the acetylene.
[13] T. Nakata, Y. Tani, M. Hatozaki, T. Oishi, Chem. Pharm. Bull. 1984,
32, 1411.
[14] J. Hartung, S. H¸nig, R. Kneuer, M. Schwarz, H. Wenner, Synthesis
1997, 1433.
[15] Ketone 21 could be isolated when the reaction was quenched with
NH4Cl (1n) instead of HCl (1n), and its diastereomeric ratio was
estimated by means of 1H NMR spectroscopy. The structure of 21 was
[*] Prof. Dr. M. A. Ciufolini, R. Ducray
Laboratoire de Synthõse et Mÿthodologie Organiques
CNRS UMR 5078
Universitÿ Claude Bernard Lyon 1
and
…cÙle Supÿrieure de Chimie, Physique, Electronique de Lyon
43, Bd. du 11 Novembre 1918, 69622 Villeurbanne cedex (France)
Fax : (þ 33)4-7243-2963
E-mail: ciufi@cpe.fr
[**] We thank the MENRT (doctoral fellowship to R.D.), the CNRS, and
the Rÿgion RhÙne-Alpes for support of our research. We are grateful
to Ms. Laurence Rousset and Dr. Denis Bouchu for the mass spectral
data, to Dr. Bernard Fenet for the NMR spectroscopic data. Finally,
we thank the Fujisawa Co. for a gift of natural FR-66979. M.A.C. is the
recipient of a Merck & Co. Academic Development Award.
Supporting information for this article is available on the WWW under
http://www.angewandte.org or from the author.
4688
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4124-4688 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2002, 41, No. 24