Catalyst-controlled asymmetric synthesis of fostriecin and 8-epi-fostriecin

Article abstract of DOI:10.1021/ja0562043

Catalytic asymmetric synthesis of the natural antibiotic fostriecin (CI-920) and its analogue 8-epi-fostriecin and evaluation of their biological activity are described. We used four catalytic asymmetric reactions to construct all of the chiral centers of

Full text of DOI:10.1021/ja0562043

Published on Web 11/10/2005
Catalyst-Controlled Asymmetric Synthesis of Fostriecin and
8-epi-Fostriecin
Keisuke Maki,† Rie Motoki,† Kunihiko Fujii,† Motomu Kanai,*,† Takayasu Kobayashi,‡
Shinri Tamura,‡ and Masakatsu Shibasaki*,†
Contribution from the Graduate School of Pharmaceutical Sciences, The UniVersity of Tokyo,
Hongo, Bunkyo-ku, Tokyo 113-0033, Japan, and Department of Biochemistry, Institute of
DeVelopment, Aging and Cancer, Tohoku UniVersity, Seiryo-machi, Aobaku,
Sendai, 980-8570, Japan
Received September 16, 2005; E-mail: mshibasa@mol.f.u-tokyo.ac.jp
Abstract: Catalytic asymmetric synthesis of the natural antibiotic fostriecin (CI-920) and its analogue 8-epi-
fostriecin and evaluation of their biological activity are described. We used four catalytic asymmetric reactions
to construct all of the chiral centers of fostriecin and 8-epi-fostriecin; cyanosilylation of a ketone, Yamamoto
allylation, direct aldol reaction, and Noyori reduction, two of which were developed by our group. Catalytic
enantioselective cyanosilylation of ketone 13 produced the chiral tetrasubstituted carbon at C-8. Both
enantiomers of the product cyanohydrin were obtained with high enantioselectivity by switching the center
metal of the catalyst from titanium to gadolinium. Yamamoto allylation constructed the C-5 chiral carbon in
the R,â-unsaturated lactone moiety. A direct catalytic asymmetric aldol reaction of an alkynyl ketone using
LLB catalyst constructed the chirality at C-9 with the introduction of a synthetically versatile alkyne moiety,
which was later converted to cis-vinyl iodide, the substrate for the subsequent Stille coupling for the triene
synthesis. Noyori reduction produced the secondary alcohol at C-11 from the acetylene ketone 6 with
excellent selectivity. Importantly, all the stereocenters were constructed under catalyst control in this
synthesis. This strategy should be useful for rapid synthesis of stereoisomers of fostriecin.
Introduction
Fostriecin (1, CI-920) is a structurally unique antibiotic
isolated from Streptomyces pulVeraceus by a research group at
Warner Lambert-Parke Davis in 1983 (Figure 1).1 It displays
cytotoxic activity against a broad range of cancerous cell lines
such as leukemia, lung cancer, breast cancer, and ovarian cancer
in vitro, and also antitumor activity against leukemia in vivo.2
Initially, fostriecin was considered to be a topoisomerase II
poison; however, there is no relation between this mechanism
and its cancer suppression activity.3 It is proposed that the
anticancer activity is derived from the perturbation of the mitotic
entry checkpoint through potent inhibition of protein phos-
phatases (PP). Specifically, fostriecin has the most selective
Figure 1. Structures of fostriecin (1) and 8-epi-fostriecin (2).
serine/threonine phosphatase 2A (PP2A) inhibition known to
date (104 times greater selectivity for PP2A versus PP1).4 Due
to this unique property, fostriecin is expected to be a novel and
potent lead compound for antitumor drugs.5 Considering the
importance of protein phosphorylation and dephosphorylation
reactions in living organisms, it might also be a valuable
biological tool. Unfortunately, the clinical trial of fostriecin at
the National Institute of Cancer was halted in phase I due to its
instability and unpredictable purity.5 Therefore, stable analogues
of fostriecin are in high demand.
Only the two-dimensional structure of fostriecin was known6
until Boger’s pioneering synthetic and degradation studies in
1997 to determine the relative and absolute configuration of
fostriecin.7 Boger and colleagues subsequently achieved the first
total synthesis in 2001.8 Since then, several total syntheses have
† The University of Tokyo.
‡ Tohoku University.
(1) (a) Tunac, J. B.; Graham, B. D.; Dobson, W. E. J. Antibiot. 1983, 36, 1595.
(b) Stampwala, S. S.; Bunge, R. H.; Hurley, T. R.; Wilmer, N. E.;
Brankiewicz, A. J.; Steinman, C. E.; Smitka, T. A.; French, J. C. J. Antibiot.
1983, 36, 1601.
(2) Reviews: (a) Lewy, D. S.; Gauss, C.-M.; Soenen, D. R.; Boger, D. L.
Curr. Med. Chem. 2002, 9, 2005. (b) Jackson, R. C.; Fry, D. W.; Boritzki,
T. J.; Roberts, B. J.; Hook, K. E.; Leopold, W. R. AdV. Enzyme Regul.
1985, 23, 193. (c) De Jong, R. S.; De Vries, E. G. E.; Mulder, N. H. Anti-
Cancer Drugs. 1997, 8, 413. (d) Scheithauer, W.; Hoff, D. D. V.; Clark.
G. M.; Shillis, J. L.; Elslager, E. F. Eur. J. Clin. Oncol. 1986, 22, 921.
(3) (a) Roberge, M.; Tudan, C.; Hung, S. M. F.; Harder, K. W.; Jirik, F. R.;
Anderson, H. Cancer Res. 1994, 54, 6115. (b) Guo, X. W.; Th’ng, J. P.
H.; Swank, R. A.; Anderson, H. J.; Tudan, C.; Brandbury, E. M.; Roberge,
M. EMBO J. 1995, 14, 976. (c) Ho, D. T.; Roberge, M. Carcinogenesis
1996, 17, 967. (d) Walsh, A. H.; Cheng, A.; Honkanen, R. E. FEBS Lett.
1997, 416, 230. (e) Hastie, C. J.; Cohen, P. T. W. FEBS Lett. 1998, 431,
357. (f) Cheng, A.; Balczon, R.; Zuo, Z.; Koons, J. S.; Walsh, A. H.;
Honkanen, R. E. Cancer Res. 1998, 58, 3611.
(4) Boritzki, T. J.; Wolfard, T. S.; Besserer, J. A.; Jackson, R. C.; Fry, D. W.;
Biochem. Pharmacol. 1998, 37, 4063.
(5) De Jong, R. S.; Mulder, N. H.; Uges, D. R. A.; Sleijfer, D. Th.; Hoppener,
F. J. P.; Groen, H. J. M.; Willemse, P. H. B.; Van der Graaf, W. T. A.; Dc
Vries, E. G. E. Br. J. Cancer 1999, 79, 882.
(6) Honkanson, G. C.; French, J. C. J. Org. Chem. 1985, 50, 462.
(7) Boger, D. L.; Hikota, M.; Lewis, B. M. J. Org. Chem. 1997, 62, 1748.
9
10.1021/ja0562043 CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 17111-17117
17111

A R T I C L E S
Maki et al.
Scheme 1. Retrosynthetic Analysis of Fostriecin (1)
Figure 2. Summary of the stereocontrol strategies in the total syntheses
of fostriecin.
been reported,9,10 including one from our group.11 Stereoselective
construction of fostriecin’s four chiral carbons is the key for
success of the total synthesis. The strategies utilized in the
reported total syntheses are summarized in Figure 2. Although
various methods are available for construction of the chiral
trisubstituted carbons at C-5, 9, and 11, there are few methods
available for the chiral tertiary alcohol at C-8. Our synthesis is
unique in that all the stereogenic centers were constructed using
external asymmetric catalysts. Specifically, the tertiary alcohol
at C-8 was constructed through catalytic enantioselective
cyanosilylation of a ketone developed in our group.12 An
advantage of our synthesis over that of other groups is its
applicability to the synthesis of stereoisomers of fostriecin. For
example, it is possible to synthesize 8-epi-fostriecin 2 just by
switching the enantioselectivity of the cyanosilylation. It is not
very straightforward to synthesize 2 using the synthetic methods
of the other groups. Investigation of the stereochemistry and
biological activity relationship is the starting point for under-
standing the protein-drug interactions. In this sense, we are
especially interested in the effect of the configuration of the
C-8 tertiary alcohol of fostriecin and, more generally, in how
the configuration of the chiral tetrasubstituted carbon is distin-
guished in nature. In this paper, we report catalytic asymmetric
synthesis of fostriecin and its epimer 8-epi-fostriecin and the
effect of the configuration of the tertiary alcohol (C-8) on their
biological activity.
Results and Discussion
Synthetic Plan. One of our focal points in this project was
to assess and demonstrate the power of catalytic asymmetric
reactions in the practical synthesis of a biologically active
compound with a considerably complex structure. Thus, we
planned to construct all of the stereogenic centers using four
catalytic asymmetric reactions. This strategy will be advanta-
geous for rapid synthesis of stereoisomer analogues of fostriecin,
if each reaction proceeds with catalyst control.
Our retrosynthesis is shown in Scheme 1. Because of its
instability, the triene moiety should be introduced at a late stage
through Stille coupling of vinyl tin 3 and vinyl iodide 4.13 The
stereogenic center at C-11 of 4 should be produced through
Noyori transfer hydrogenation14 of alkynyl ketone 6. The
secondary alcohol at C-9 can be constructed through a direct
catalytic asymmetric aldol reaction between aldehyde 8 and
alkynyl methyl ketone 9 promoted by a heterobimetallic LLB
complex 7.15 R,â-Unsaturated lactone can be constructed through
Yamamoto allylation16 of aldehyde 10, acryloylation, and ring
closing metathesis. Finally, we planned to apply catalytic
(8) Boger, D. L.; Ichikawa, S.; Zhong, W. J. Am. Chem. Soc. 2001, 123, 4161.
(9) (a) Chavez, D. E.; Jacobsen, E. N. Angew. Chem., Int. Ed. 2001, 40, 3667.
(b) Reddy, Y. K.; Falck, J. R. Org. Lett. 2002, 4, 969. (c) Miyashita, K.;
Ikejiri, M.; Kawasaki, H.; Maemura, S.; Imanishi, T. Chem. Commun. 2002,
9, 742. (d) Wang, Y.-G.; Kobayashi, Y.; Org. Lett. 2002, 4, 4615. (e) Esumi,
T.; Okamoto, N.; Hatakeyama, S. Chem. Commun. 2002, 9, 3042. (f)
Miyashita, K.; Ikejiri, M.; Kawasaki, H.; Maemura, S.; Imanishi, T. J. Am.
Chem. Soc. 2003, 125, 8238. (g) Trost, B. M.; Frederiksen, M. U.; Papillon,
J. P. N.; Harrington, P. E.; Shin, S.; Shireman, B. T. J. Am. Chem. Soc.
2005, 127, 3666.
(10) Synthetic studies: (a) Just, G.; O’Connor, B. Tetrahedron Lett. 1988, 29,
753. (b) Liu, S. U.; Huang, D. F.; Huang, H. H.; Huang, L. Chin. Chem.
Lett. 2000, 11, 957. (c) Cossy, J.; Pradaux, F.; BouzBouz, S. Org. Lett.
2001, 3, 2233. (d) Kiyotsuka, Y.; Igarashi, J.; Kobayashi, Y. Tetrahedron
Lett. 2002, 43, 2725. (e) Marshall, J. A.; Bourbeau, M. P. Org. Lett. 2003,
5, 3197. (f) Ramachandran, P. V.; Liu, H. P.; Reddy, M. V. R.; Brown, H.
C. Org. Lett. 2003, 5, 3755.
(11) Fujii, K.; Maki, K.; Kanai, M.; Shibasaki, M. Org. Lett. 2003, 5, 733.
(12) (R)-selective: (a) Hamashima, Y.; Kanai, M.; Shibasaki, M. J. Am. Chem.
Soc. 2000, 122, 7412. (b) Hamashima, Y.; Kanai, M.; Shibasaki, M.
Tetrahedron Lett 2001, 42, 691. (S)-selective: (c) Yabu, K.; Masumoto,
S.; Yamasaki, S.; Hamashima, Y.; Kanai, M.; Du, W.; Curran, D. P.;
Shibasaki, M. J. Am. Chem. Soc. 2001, 123, 9908. (d) For a recent review,
see: Kanai, M.; Kato, N.; Ichikawa, E.; Shibasaki, M. Synlett 2005, 1491.
(13) Jacobsen, Imanishi, Hatakeyama, and Falck utilized Stille or Suzuki cross-
coupling for the triene construction. Vinyl iodide 4 is the common
intermediate of Jacobsen, Imanishi, Hatakeyama, and Kobayashi.
(14) Matsumura, K.; Hashiguchi, S.; Ikariya, T.; Noyori, R. J. Am. Chem. Soc.
1997, 119, 8738.
(15) (a) Yoshikawa, N.; Yamada, Y. M. A.; Das, J.; Sasai, H.; Shibasaki, M. J.
Am. Chem. Soc. 1999, 121, 4168. (b) Yamada, Y. M. A.; Yoshikawa, N.;
Sasai, H.; Shibasaki, M. Angew. Chem., Int. Ed. Engl. 1997, 36, 1871.
(16) Yanagisawa, A.; Kageyama, H.; Nakatsuka, Y.; Asakawa, K.; Matsumoto,
Y.; Yamamoto, H. Angew. Chem., Int. Ed. 1999, 38, 3701.
9
17112 J. AM. CHEM. SOC. VOL. 127, NO. 48, 2005

Catalytic Asymmetric Synthesis of Fostriecin
A R T I C L E S
Scheme 2. Catalytic Enantioselective Cyanosilylation of Ketones
Scheme 3 a
a Reagents and conditions: (a) Ti(OiPr)4 (5 mol %), ligand 12 (5 mol
%), TMSCN, THF, -25 °C, 93%, 85% ee; (b) 6 N HCl/EtOH, 60 °C,
83%; (c) NaBH4, MeOH, 88%; (d) p-NO2-C6H4COCl, py, CH2Cl2, 100%;
recryst. from CH2Cl2/hexane, 78% (>99% ee); (e) K2CO3, MeOH-CH2Cl2,
i
100%; (f) TIPSCl, imidazole, DMF, 96%; (g) MOMCl, Pr2NEt, CH2Cl2,
enantioselective cyanosilylation of ketone 13 developed in our
group12 to control the C-8 stereocenter.
100%; (h) Na, NH3, -78 °C, 93%; (i) MnO2, CH2Cl2, 93%; (j) AgF (20
mol %), (R)-p-tol-BINAP (20 mol %), allyltrimethoxysilane, MeOH, -20
°C, 80% (d.r. ) 28:1); (k) acryloyl chloride, Et3N, CH2Cl2, 76%; (l) first
generation Grubbs catalyst (15 mol %), CH2Cl2, 94%; (m) 3HF-Et3N, THF,
90%; (n) DMP, CH2Cl2, 92%.
Synthesis of Fostriecin (1). Our synthesis of fostriecin began
with the catalytic asymmetric cyanosilylation of ketone 13. A
titanium complex derived from ligand 11 or 12 and Ti(OiPr)4
in a 1:1 ratio generally gives (R)-ketone cyanohydrins with
excellent enantioselectivity, while a gadolinium complex derived
from 11 and Gd(OiPr)3 in a 2:1 ratio generally gives (S)-ketone
cyanohydrins (Scheme 2).12 In natural fostriecin synthesis, the
former reaction was used. Both the reactivity and selectivity of
this reaction heavily depended on the protective group of the
allylic alcohol of the substrate. Examination of several sub-
strates17 indicated that the benzyl group was the protective group
of choice. Thus, using 5 mol % of the catalyst derived from
ligand 11, product 16 was obtained in 94% yield with 71% ee
(-20 °C, 36 h). Enantioselectivity was improved using ligand
12 containing a benzoyl group at the catechol,12b and the product
was obtained in 86% yield and 87% ee (-25 °C, 48 h). The
higher enantioselectivity obtained when using 12 compared to
11 might be partly due to the enhanced Lewis acidity of titanium
by the electron-withdrawing group at the catechol, which would
more efficiently stabilize the hypothetical dual activation
transition state 14. This reaction was performed on a 50 g scale
without a significant change in efficiency (93%, 85% ee). The
chiral ligand 12 was recovered in 95% yield after silica gel
column chromatography, which could be reused at least several
times without any loss of catalyst activity.
construct the chiral lactone part, we investigated several catalytic
asymmetric allylations.19 Yamamoto’s AgF-catalyzed reaction
gave the best results in the current system. Diastereoselectivity
was as high as 28:1 in favor of the desired isomer 20.20 R,â-
Unsaturated lactone 21 was constructed from 20 through
treatment with acryloyl chloride and ring closing olefin me-
tathesis using the first generation Grubbs’ catalyst.21 Deprotec-
tion of the primary alcohol and oxidation with Dess-Martin
periodinane (DMP) produced aldehyde 8, the substrate for the
catalytic asymmetric aldol reaction.
Before performing the catalytic asymmetric aldol reaction
with an alkynyl ketone in the actual system, its feasibility was
tested using a model aldehyde 22, because this type of reaction
had not been explored.22 Neither the lithium enolate nor zinc
enolate of 9 provided the target molecule 23 in reasonable yield
(Table 1, entries 1 and 2). On the other hand, Lewis acid-
Bro¨nsted base two-center asymmetric catalyst23 LLB (10 mol
%) promoted the reaction in good yield with high enantiose-
lectivity (entry 3). Less satisfactory results were obtained when
the more basic LLB/KOH system15a was utilized (entry 4),
probably because the highly basic catalyst facilitated the
retroaldol reaction. Application of the best reaction conditions
to the actual system (8 + 9) afforded the desired aldol adduct
with 3.6:1 diastereomeric ratio in 65% yield.24 All attempts to
The cyanohydrin 16 was transformed to the primary alcohol
17 via ethanolysis and NaBH4 reduction (Scheme 3). Enantio-
merically pure 17 was obtained after converting to the p-
nitrobenzoyl ester 18, recrystallization, and hydrolysis (78%
yield in three operations). Subsequent protection of the primary
and tertiary alcohols with TIPS and MOM groups, respectively,
followed by debenzylation led to the allyl alcohol, which was
oxidized by MnO2 to give the allylation precursor 10.18 To
(19) Keck allylation (Keck, G. E.; Krishnamurthy, D.; Grier, M. C. J. Org. Chem.
1993, 58, 6543) gave unsatisfactory results, perhaps due to the coordination
of the MOM-oxy group to the titanium of the catalyst. See SI for details.
(20) The C-5 configuration was determined after conversion to 4 by comparing
the 1H and 13C NMR with Imanishi’s NMR chart. The stereoselectivity
was consistent with Yamamoto’s transition state model. AgF-catalyzed
reaction in the absence of the chiral ligand gave a 1:1 diastereomixture.
(21) Grubbs, R. H.; Chang, S. Tetrahedron 1998, 54, 4413.
(22) Trost recently reported a catalytic enantioselective aldol reaction using an
alkynyl ketone as a donor: Trost, B. M.; Fetes, A.; Shireman, B. T.; J.
Am. Chem. Soc. 2004, 126, 2660.
(23) For a recent review on two-center asymmetric catalysis, see: Shibasaki,
M.; Kanai, M.; Funabashi, K. Chem. Commun. 2002, 1985.
(17) TBS, TBDPS, MOM, Bz, Ac, and PMB groups were tested. See Supporting
Information (SI) for details.
(18) In the previous communication, debenzylation and oxidation were conducted
with LiDBB and TPAP, respectively. The current reaction conditions
produced higher chemical yield and reproducibility.
9
J. AM. CHEM. SOC. VOL. 127, NO. 48, 2005 17113

A R T I C L E S
Maki et al.
p-nitrobenzenesulfonylhydrazide (NBSH).28 By modifying Iman-
ishi’s procedure, properly protected vinyl iodide 4 was afforded
in three steps. Stille coupling between vinyl tin 329 and 4 under
ligand-free conditions30 proceeded uneventfully to form 26, the
common intermediate of the fostriecin synthesis developed by
Jacobsen, Imanishi, and Hatakeyama. Thus, the formal total
synthesis of fostriecin was accomplished.
Table 1. Direct Catalytic Enantioselective Aldol Reaction of
Alkynyl Ketone in a Model System
entry
promoter
additive
temp. (
°
C)
time (h)
yield (%)
ee (%)
1a
LDA
LDA
(R)-LLB
(R)-LLB
none
ZnCl2
none
KOH, H2O
-78
0
-20
-40
1
12
22
64
28
0
76
63
Synthesis of 8-epi-Fostriecin. Despite intensive studies, the
structure-activity relationship of fostriecin, its binding feature
to PP, and the molecular mechanism of PP inhibition are not
yet clarified.2 Specifically, the structure-activity relationship
of 1 is limited to information obtained by either molecular
modeling calculations or compounds that could be derivatized
from 1 and its natural analogues. Three important pieces of
information confirmed so far are as follows: (1) the R,â-
unsaturated lactone is a fundamental motif for the potent activity
and PP2A selectivity of 1; (2) the phosphate ester at C-9 is
essential; (3) the terminal allylic alcohol is not necessary.31 To
obtain further insight into the binding mode of 1 to PP, we were
interested in the effect of the configuration of chiral carbons
on fostriecin’s biological activity, especially the effect of chiral
tertiary alcohol at C-8. The methyl group at C-8 is proposed to
be a common pharmacophore in naturally occurring PP1 and
PP2A inhibitors, and it mimics the methyl group of phospho-
threonine, a PP substrate.32 Currently, the role of the tertiary
alcohol is not well understood, but it might mimic the hydroxy
group of the substrate threonine or displace the enzyme’s metal-
bound water nucleophile in the active site.2a Because our
synthetic strategy is advantageous for preparing other isomers
of fostriecin, we extended our strategy to the synthesis of 8-epi-
fostriecin (2). A main interest in this synthetic endeavor is
whether catalyst control dominates over substrate control in this
linear system.
2a,b
3c
87
60
4c,d
a 2 equiv of 9 and 1 equiv of LDA were used. b 1 equiv of ZnCl2 to 9
was used as an additive. c 6 equiv of 9 and 10 mol % of (R)-LLB were
used. d 9 mol % of KOH and 10 mol % of H2O were used as additives.
Scheme 4 a
The first reaction was (S)-selective cyanosilylation of 13 using
a gadolinium complex (2 mol %) derived from ligand 11
(Scheme 5). Product 27, which is the enantiomer of the previous
16, was obtained in 95% yield with 86% ee. This reaction was
performed on a 120 g scale. 27 was effectively converted to
the allylation precursor 28 following the procedure for the
fostriecin synthesis. Yamamoto allylation of this substrate
produced the desired isomer 29 with excellent selectivity (d.r.
) 16:1). Thus, stereochemistry at C-8 did not significantly affect
stereoselectivity in this allylation, and the catalyst chirality
a Reagents and conditions: (a) (S)-LLB (10 mol %), 9, THF, -20 °C,
65% (d.r. ) 3.6:1); (b) 2,2-dimethoxypropane, PPTS, acetone, 80%; (c)
i
Noyori’s catalyst 5 (10 mol %)-KOH (10 mol %), PrOH, 49%; (other
isomers: 18%); (d) TBSOTf, 2,6-lutidine, CH2Cl2, -78 °C, 73%; (e) NIS,
AgNO3, acetone, 88%; (f) NBSH, Et3N, THF-iPrOH, 40% (6% recovery);
(g) 1 M HCl aq. MeOH, 47%; (h) TBSOTf, 2,6-lutidine, CH2Cl2, -78 °C;
TESOTf, -78 °C to -10 °C, 52%; (i) 1 M HCl aq.-THF-CH3CN (1:3:
6), -10 °C, 52%; (j) PdCl2(CH3CN)2, 3, DMF, 85%.
improve the moderate selectivity, including catalyst tuning by
modifying the BINOL structure, were unsuccessful.25
(26) Other diastereomers (9S,11S and 9S,11R) were obtained in 14% and 4%
isolated yield. The relative configuration of C-9 and C-11 was determined
by NOE measurement. See ref 11 for details.
With the desired aldol in hand, construction of the remaining
stereogenic center at C-11 was performed using Noyori hydro-
genation. Thus, after acetonide formation to generate 6, Noyori’s
transfer hydrogenation effectively produced the secondary
alcohol 24 in a highly stereoselective manner (d.r. ) >15:1,
Scheme 4). Diastereomerically pure 24 was obtained by
preparative TLC purification.26 At this stage, all the chiral
centers of fostriecin were introduced. 24 was transformed to
cis-vinyl iodide 25 via three steps: TBS protection of the
secondary alcohol at C-11, iododesilylation of the TMS
acetylene,27 and diimide reduction of the iodoalkyne using
(27) Nishikawa, T.; Shibuya, S.; Hosokawa, S.; Isobe, M. Synlett 1994, 485.
(28) Myers, A. G.; Zheng, B.; Morassaghi, M. J. Org. Chem. 1997, 62, 7507.
(29) Mapp, A. K.; Heathcook, C. H. J. Org. Chem. 1999, 64, 23.
(30) Stille, J. K.; Groh, B. L. J Am. Chem. Soc. 1987, 109, 813.
(31) (a) Buck, S. B.; Hardouin, C.; Ichikawa, S.; Soenen, D. R.; Gauss, G.-M.;
Hwang, I.; Swingle, M. R.; Bonness, K. M.; Honkanen, R. E.; Boger, D.
L. J. Am. Chem. Soc. 2003, 125, 15694. (b) Leopold, W. R.; Shillis, J. L.;
Mertus, A. E.; Nelson, J. M.; Roberts, B. J.; Jackson, R. C. Cancer Res.
1984, 44, 1928. (c) Fry, D. W.; Besserer, J. A.; Boritzki, T. J. Cancer Res.
1984, 44, 3366. (d) Mamber, S. W.; Okasinski, W. G.; Pinter, C. D.; Tunac,
J. B. J. Antibiot. 1986, 39, 1467.
(24) The diastereomeric ratio was determined based on the 1H NMR analysis
of the crude mixture of the aldol reaction. The diastereomers were not
separable at this stage. The relative configuration of C-8 and C-9 was
determined by NOE measurement on acetonide 6.
(25) For a beneficial effect of BINOL substitution on a catalytic asymmetric
nitroaldol reaction, see: Shibasaki, M.; Sasai, H.; Arai, T. Angew. Chem.,
Int. Ed. Engl. 1997, 36, 1236.
(32) Gauss, C.-M.; Sheppeck, J. E., II; Narin, A. C.; Chamberlin, R. Bioorg.
Med. Chem. 1997, 5, 1751.
9
17114 J. AM. CHEM. SOC. VOL. 127, NO. 48, 2005

Catalytic Asymmetric Synthesis of Fostriecin
A R T I C L E S
Scheme 5 a
Scheme 6 a
a Reagents and conditions: (a) Gd(OiPr)3 (2 mol %), ligand 11 (4 mol
%), TMSCN, EtCN, -65 °C, 95%, 86% ee; (b) AgF (5 mol %), (R)-p-
tol-BINAP (5 mol %), allyltrimethoxysilane, MeOH, -20 °C, 88% (d.r. )
16:1); (c) acryloyl chloride, Et3N, CH2Cl2, 0 °C, 87%; (d) second generation
Grubbs catalyst (2 mol %), CH2Cl2, reflux, 93%; (e) 3HF-Et3N, THF, 50
°C, 82%; (f) DMP, H2O (0.5 equiv), CH2Cl2, 94%.
Table 2. Additive Effect in Direct Catalytic Asymmetric Aldol
Reaction of 31
entry
additive (mol %)
time (h)
d.r. (anti/syn)
1
2
3
4
5
6
7
8
none
25
24
4
10
15
24
34
34
2:1
3:1
4:1
2.5:1
1.5:1
2:1
a Reagents and conditions: (a) (S)-LLB (10 mol %), LiOTf (20 mol %),
9, THF, -20 °C; (b) 2,2-dimethoxypropane, 50 °C, 65% (for two steps,
LiOTf(10)
LiOTf(20)
LiOTf(30)
NaOTf(10)
KOTf(10)
LiClO4(10)
LiPF6(10)
i
d.r. ) 4:1); (c) Noyori’s catalyst 5 (3.3 mol %), KOH (3 mol %), PrOH,
81%; (d) NIS, AgNO3 (0.6 equiv), EtOH (10 equiv), acetone, 0 °C, 87%;
(e) NBSH, NaHCO3, MeOH, 58% (31% recovery); (f) 1 M HCl aq. MeOH,
73%; (g) TBSOTf, 2,6-lutidine, CH2Cl2, -78 °C; TESOTf, -78 °C to -40
°C, 89%; (h) 1 M HCl aq.-THF-CH3CN (1:3:6), -10 °C, 53% (20%
recovery); (i) PdCl2(MeCN)2 (50 mol %), 3, DMF, 40 °C, 85%; (j)
iPr2NP(Oallyl)2 (40), 1H-tetrazole, CH2Cl2, 0 °C; I2, py-H2O-THF (1:2:
7), -20 °C (k) Pd(PPh3)4 (100 mol %), PPh3, formamide, THF; 48% HF
aq.-H2O-MeCN (1:2:20), py, 50% (4 steps).
2:1
2:1
determined the stereochemical outcome. This reaction was
followed by ester formation with acryloyl chloride and ring
closing metathesis catalyzed by the second generation Grubbs’
ruthenium complex33 to form R,â-unsaturated lactone 30. The
TIPS group was then removed, and the resultant primary alcohol
was oxidized to the aldol precursor 31 by DMP.34 This
oxidation, however, was tricky, and the yield of 31 was variable.
Further investigations clarified that a small amount of water
was essential for reproducible conversion. Thus, aldehyde 31
was obtained in 94% yield in the presence of 0.5 equiv of H2O.
A similar beneficial effect of water in DMP oxidation was
reported by Schreiber et al.35
The (S)-LLB-catalyzed asymmetric aldol reaction of aldehyde
31 and alkynyl ketone 9 was then performed under the optimized
conditions in natural fostriecin synthesis. Although the reaction
proceeded under catalyst control in favor of the desired isomer
32, the diastereoselectivity was only 2:1 (Table 2, entry 1).36
In contrast to the case of natural fostriecin synthesis, the
stereoselectivity of the catalyst did not match the 1,2-chirality
transfer raised by the C-8 chiral tertiary alcohol in this case.
To improve the selectivity, we intensively investigated the effect
of additives (Table 2). Among the additive salts screened, 20
mol % LiOTf gave the best results;37 the diastereomeric ratio
was improved to 4:1 (entry 3). Other alkali metal triflates and
other lithium counterions other than triflate gave almost the same
diastereoselectivity as that in the absence of additives.
With the desired aldol 32 in hand, the next task was the
construction of the C-11 stereocenter. Due to the instability of
32, it was directly transformed to acetonide 33 without column
chromatographic purification. As might be expected, Noyori
reduction of 33 proceeded under catalyst-control irrespective
of the configuration of C-8, giving product 34 with excellent
selectivity (d.r. ) >10:1, Scheme 6).36 When the resulting
propargyl alcohol 34 was subjected to the same iododesilylation
conditions as the natural fostriecin synthesis (NIS and AgNO3
in acetone), the desired iodoalkyne 35 was obtained in moderate
yield (62%) with concomitant generation of an epimerized
(35) Meyer, S. D.; Schreiber, S. L. J. Org. Chem. 1994, 59, 7549.
(36) The relative configurations between C-8, C-9, and C-11 were determined
by NOE and/or NOESY measurements on the 33 and 34 derivative and
later by X-ray crystallographic analysis of 37. See SI for details.
(37) It is proposed that the LLB-LiOTf oligomeric species is the actual
asymmetric catalyst under these conditions. See: (a) Horiuchi, Y.;
Gnanadesikan, V.; Ohshima, T.; Masu, H.; Katagiri, K.; Sei, Y.; Yamaguchi,
K.; Shibasaki, M. Chem.sEur. J. 2005, 11, 5195. (b) Gnanadesikan, V.;
Horiuchi, Y.; Ohshima, T.; Shibasaki, M. J. Am. Chem. Soc. 2004, 126,
7782.
(33) Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999, 1, 953.
(34) Other oxidants such as PDC, TPAP, and IBX did not promote the reaction.
Swern oxidation gave the product in 67% yield.
9
J. AM. CHEM. SOC. VOL. 127, NO. 48, 2005 17115

A R T I C L E S
Maki et al.
Table 3. IC50 Values (in µM) in Inhibition of PP2A and PP1a
assay
inhibitor
PP2A
PP1
A
okadaic acid
fostriecin (1)
8-epi-fostriecin (2)
okadaic acid
fostriecin (1)
8-epi-fostriecin (2)
okadaic acid
fostriecin (1)
8-epi-fostriecin (2)
okadaic acid
fostriecin (1)
0.0001
0.002
4
0.006
5
40
0.004
6
0.1
>100
>100
B
C
D
1
4
>1000
0.2
9
>4000
2
4
770
0.001
40
Figure 3. ORTEP drawing of 37.
8-epi-fostriecin (2)
6
510
dioxane 36 (20%). Generation of Lewis acidic silyl species
(TMSNO2 or TMS-succinimide) during this reaction might
cause the formation of 36. Thus, Lewis acid mediated equilibra-
tion between the dioxolane and the dioxane, silicon trap of the
tertiary alcohol to terminate this equilibrium, and epimerization
at the propargylic position to the more stable isomer with the
two substituents at the equatorial positions should produce 36.
To avoid the adverse effects of the Lewis acidic silyl species,
we added EtOH as a scavenger. The yield of 35 then dramati-
cally increased to 87%, with minimum generation of undesired
36 (2%).
cis-Reduction of the iodoalkyne 35 with NBSH, followed by
deprotection in acidic media, gave triol 37 (Figure 3). X-ray
crystallographic analysis of 37 unequivocally determined the
stereochemistry of all the stereogenic centers. The two-step
protection-deprotection sequence worked reasonably well as
in the case of natural fostriecin synthesis, giving the coupling
precursor 38. The Stille coupling between 38 and 3 proceeded
at 40 °C under ligandless conditions, and the triene was obtained
in 85% yield. Careful temperature control was necessary for
this Stille coupling reaction, because isomerization of the triene
occurred gradually at 60 °C, while at room temperature the
reaction was rather slow.
Finally, the phosphate moiety was added to the C-9 alcohol
to complete the synthesis of 2. We used allyl-protected
phosphoramidite 40 for this phosphorylation, according to
Hatakeyama’s fostriecin synthesis. The application of Hatakeya-
ma’s conditions using TBHP as an oxidizing reagent of the
initially formed phosphite, however, produced a complex
mixture of products. TLC analysis of the reaction revealed that
decomposition occurred at the oxidation step. Thus, changing
the oxidizing reagent to I2 and performing the reaction at -20
°C under careful TLC monitoring, the phosphorylation product
was obtained in reasonable yield (80% based on TLC analysis).
Deprotection of the phosphate and global desilylation were
conducted under slightly modified conditions of the reported
procedure.9e These reactions should also be quenched in 4 h,
otherwise significant decomposition occurs. After purification
through a short pad reversed-phase silica gel column chroma-
tography,38 the synthesis of 8-epi-fostriecin was achieved.
Biological Assay. Inhibitory activities of okadaic acid
(control), fostriecin (1), and 8-epi-fostriecin (2) against PP1 and
PP2A were evaluated. Four kinds of assays were conducted in
which the enzymatic activity was determined by hydrolysis of
phosphorylated histone (assay A),39 phosphorylated casein (assay
a Substrates of PPs are phosphorylated histone in assay A, phosphorylated
casein in assay B, p-nitrophenyl phosphate in assay C, or commercial
phosphorylated peptide or ProFluor Ser/Thr Phosphatase Assay Kit in assay
D. Assays were performed at least three times.
B), p-nitrophenyl phosphate (assay C),40 or commercial phos-
phorylated peptide (assay D)41 (Table 3). Although the IC50
values of okadaic acid were similar in the four assay systems
(IC50 ) 0.1-6 nM for PP2A and 0.1-2 µM for PP1), those of
1 and 2 were highly dependent on the phosphorylated substrates.
1 exhibited excellent PP2A selective inhibitory activity in assay
A, as previously reported (2 nM for PP2A and >100 µM for
PP1),3f but almost no selectivity or even slight PP1-selectivity
in assays B-D. 2 was a weaker PP inhibitor than 1 in assays
A-C. Specifically, 2 did not demonstrate inhibitory activity in
assay C using p-nitrophenyl phosphate as the substrate, which
is more precisely a substrate analogue of tyrosine phosphatase.
On the other hand, 2 inhibited PP2A more selectively and with
higher potency than 1 in assay D.
Differences in the inhibitory activity and subtype selectivity
between these three compounds are interesting from the
viewpoint of protein-drug interactions and their inhibitory
mechanisms. Although both okadaic acid and fostriecin are
protein phosphatase inhibitors, the cellular responses caused by
these compounds are different: okadaic acid has been reported
to promote or suppress tumor formation depending on the
duration of the drug application or cell type, while fostriecin is
an antitumor agent. This contrasting outcome might be related
to our observation that okadaic acid inhibited dephosphorylation
of all four substrates examined with the IC50 values in the
nanomolar range, while fostriecin’s IC50 values varied over 4
orders of magnitude depending on the substrate. These results
also suggest that fostriecin might selectively interfere with
specific intracellular signaling pathways depending on the
phosphorylated substrate proteins. Based on the results in assays
B, C, and D, 8-epi-fostriecin (2) appears to have an even higher
substrate specificity with a higher PP2A selectivity than natural
fostriecin (1). The difference between 1 and 2 should reflect
the difference in their three-dimensional structures,42 and
therefore interactions with the phosphatases. The stereochemistry
at C-8 has a great influence on their biological activity, and
this finding supports the proposed hypothesis that the methyl
and hydroxyl groups at C-8 are important pharmacophores that
(39) (a) Honkanen, R. E.; Zwillwe, J.; Moore, R. E.; Daily, S. L.; Khatra, B.
S.; Dukelow, M.; Boynton, A. L. J. Biol. Chem. 1990, 265, 19401. (b)
Huang, X.; Swingle, M. R.; Honkanen, R. E. Methods Enzymol. 2000, 315,
579.
(40) Zwiller, J.; Ogasawara, E. M.; Nakamoto, S. S.; Boynton, A. L. Biochem.
Biophys. Res. Commun. 1988, 155, 767.
(41) The assay was performed using the commercially available ProFluor Ser/
Thr Phosphatase Assay Kit.
(38) 8-epi-Fostriecin 2 is not very stable. It decomposes partially on reversed-
phase preparative TLC at 4 °C, and significantly on reversed-phase HPLC
under buffered conditions at rt (eluent, MeCN/pH 6.8 phosphate buffer )
12:88).
9
17116 J. AM. CHEM. SOC. VOL. 127, NO. 48, 2005

Catalytic Asymmetric Synthesis of Fostriecin
A R T I C L E S
mimic the substrate threonine. Considering its high selectivity,
both in terms of the substrates and the protein phosphatases
subtypes, 8-epi-fostriecin is a unique biological tool. Further
evaluation of the anticancer activity of 8-epi-fostriecin is
ongoing.
facilitate the synthesis of the other stereosiomers of fostriecin.
Four kinds of biological assays to evaluate the phosphatase
inhibitory activity were performed. The results clarified that the
inhibitory activities of fostriecin and 8-epi-fostriecin were
substrate-dependent, and the stereochemistry at C-8 greatly
affected the biological activity. Furthermore, in one assay, 8-epi-
fostriecin was a more selective and more potent PP2A inhibitor
than natural fostriecin. Therefore, 8-epi-fostriecin might be
useful as a unique biological tool.
Conclusions
We accomplished a catalytic enantioselective synthesis of
fostriecin and 8-epi-fostriecin. All of the stereogenic centers of
these compounds were constructed by combining catalytic
enantioselective cyanosilylation of a ketone, catalytic asym-
metric allylation, a direct catalytic asymmetric aldol reaction
using an alkynyl ketone, and Noyori reduction, two of which
were developed in our laboratories. Importantly, these key
reactions proceeded under catalyst control. This strategy should
Acknowledgment. Financial support was provided by a
Grant-in-Aid for Specially Promoted Research of MEXT. We
gratefully acknowledge Professor Imanishi for providing us the
NMR charts of the synthetic intermediates, Mr. M. Waki for
technical support and useful advice about biological assay,
Professor S. Kobayashi and Mr. H. Kiyohara for X-ray crys-
tallographic analysis, and Dr. T. Ohshima and Mr. S. Nakamura
for kind advice about theoretical calculations.
(42) Preliminary conformation search by molecular mechanics calculations using
CONFLEX5 [(a) Goto, H.; Osawa, E. J. Am. Chem. Soc. 1989, 111, 8950.
(b) Goto, H.; Osawa, E. J. Chem. Soc., Perkin Trans. 2 1993, 187. (c)
Goto, H.; Osawa, E. THEOCHEM 1993, 285, 157. (d) Goto, H.;
Kawashima, Y.; Kashimura, M.; Morimoto, S.; Osawa, E. J. Chem. Soc.,
Perkin Trans. 2 1993, 1647. (e) Osawa, E.; Goto, H.; Hata, T.; Deretey, E.
THEOCHEM 1997, 399, 229.] indicated that natural fostriecin is confor-
mationally more flexible than 8-epi-fostriecin. Among about 8000 conform-
ers generated in the MMFF94S force field using this program, more than
90% of them were distributed in the 2 kcal/mol range from the most stable
conformer in the case of 8-epi-fostriecin. In the case of natural fostriecin,
however, ca. 75% of conformers existed in the same range.
Supporting Information Available: Full experimental details
including structure assignments and results of biological assays
are provided. This material is available free of charge via the
Internet at http://pubs.acs.org.
JA0562043
9
J. AM. CHEM. SOC. VOL. 127, NO. 48, 2005 17117