Total synthesis of fostriecin (CI-920)

Article abstract of DOI:10.1021/ja010195q

The first total synthesis of the potent antitumor agent fostriecin (CI-920) is described, confirming the relative and absolute stereochemistry assignments. Fostriecin is a unique phosphate monoester which exhibits weak topoisomerase II inhibition (IC

Full text of DOI:10.1021/ja010195q

J. Am. Chem. Soc. 2001, 123, 4161-4167
4161
Total Synthesis of Fostriecin (CI-920)
Dale L. Boger,* Satoshi Ichikawa, and Wenge Zhong
Contribution from the Department of Chemistry and The Skaggs Institute for Chemical Biology,
The Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, California 92037
ReceiVed January 23, 2001
Abstract: The first total synthesis of the potent antitumor agent fostriecin (CI-920) is described, confirming
the relative and absolute stereochemistry assignments. Fostriecin is a unique phosphate monoester which exhibits
weak topoisomerase II inhibition (IC50 ) 40 µM) and more potent and selective protein phosphatase 2A and
4 (PP2A and PP4) inhibition (IC50 ) 40-3 nM and 1.5 nM), resulting in mitotic entry checkpoint inhibition.
Phase I clinical trials with fostriecin, which were the first to explore the potential of this novel mechanism of
action, were halted even before therapeutic concentrations were reached or dose-limiting toxicity established
due to problems of drug stability observed during storage of naturally derived material. The synthesis of fostriecin
detailed herein is the first stage of efforts that may serve to address these limitations to the clinical examination
of this or related promising new antitumor agents.
Fostriecin (1, CI-920, Figure 1)1 is a structurally novel
phosphate ester produced by Streptomyces pulVeraceus that is
active in vitro against leukemia (L1210, IC50 0.46 µM), lung
cancer, breast cancer, and ovarian cancer, and which exhibits
efficacious in vivo antitumor activity.2 It has been investigated
in a Phase I clinical trial at the National Cancer Institute that
was halted even before dose-limiting toxicities or therapeutic
plasma levels were reached when concerns regarding drug purity
and storage stability proved problematic with the naturally
occurring material.3 Although fostriecin inhibits DNA topoi-
somerase II (IC50 ) 40 µM)4 through a novel, non-DNA-strand
cleavage mechanism, this activity is weak, and it does not induce
G2 arrest like other topoisomerase II inhibitors. Thus, it is
unlikely that this initially suggested target is responsible for
the antitumor activity of 1.5 Instead, fostriecin has been shown
to inhibit the mitotic entry checkpoint through more potent and
selective inhibition of protein phosphatases 1 (PP1), 2A (PP2A),
and 4 (PP4) (IC50 ) 45 µM, 1.5 nM, and 3.0 nM, respective-
ly).6-13 Notable in this regard is the observation that fostriecin
Figure 1.
is the most selective (PP2A/PP4 versus PP1 selectivity ) 104×)
protein phosphatase inhibitor known to date. Inhibition of the
(9) Hastie, C. J.; Cohen, P. T. W. FEBS Lett. 1998, 431, 357-361.
(10) Ho, D. T.; Roberge, M. Carcinogenesis 1996, 17, 967-972.
(11) Guo, X. W.; Th’ng, J. P. H.; Swank, R. A.; Anderson, H. J.; Tudan,
C.; Bradbury, E. M.; Roberge, M. EMBO J. 1995, 14, 976-85.
(12) For reviews of protein serine/threonine phosphatases, see: Ingre-
britsen, T. S.; Cohen, P. Eur. J. Biochem. 1983, 132, 255-261. Cohen, P.
Annu. ReV. Biochem. 1989, 58, 453-508.
(1) Tunac, J. B.; Graham, B. D.; Dobson, W. E. J. Antibiot. 1983, 36,
1595-1600. Stampwala, S. S.; Bunge, R. H.; Hurley, T. R.; Willmer, N.
E.; Brankiewicz, A. J.; Steinman, C. E.; Smitka, T. A.; French, J. C. J.
Antibiot. 1983, 36, 1601-1605.
(2) Reviews: Jackson, R. C.; Fry, D. W.; Boritzki, T. J.; Roberts, B. J.;
Hook, K. E.; Leopold, W. R. AdV. Enzyme Regul. 1985, 23, 193-215. De
Jong, R. S.; De Vries, E. G. E.; Mulder, N. H. Anti-Cancer Drugs 1997, 8,
413-418.
(3) 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.; De Vries, E. G. E. Br. J. Cancer 1999, 79, 882-887.
(4) Boritzki, T. J.; Wolfard, T. S.; Besserer, J. A.; Jackson, R. C.; Fry,
D. W. Biochem. Pharmacol. 1988, 37, 4063-4068.
(13) Other inhibitors of protein serine/threonine phosphatases are
described in the following references. (a) Okadaic acid: Tachibana, K.;
Scheuer, P. J.; Tsukitani, Y.; Kikuchi, Y.; Van Engen, D.; Clardy, J.;
Gopichand, Y.; Schmitz, F. J. J. Am. Chem. Soc. 1981, 103, 2469-2471.
(b) Microcystins: Botes, D. P.; Tuinman, A. A.; Wessels, P. L.; Viljoen,
C. C.; Kruger, H.; Williams, D. H.; Santikarn, S.; Smith, R. J.; Hammond,
S. J. J. Chem. Soc., Perkin Trans. 1 1984, 2311-2318. (c) Calyculin A:
Kato, Y.; Fusetani, N.; Matsunaga, S.; Hashimoto, K.; Fujita, S.; Furuya,
T. J. Am. Chem. Soc. 1986, 108, 2780-2781. (d) Nodularin: Botes, D. P.;
Wessels, P. L.; Kruger, H.; Runnegar, M. T. C.; Santikarn, S.; Smith, R.
J.; Barna, J. C. J.; Williams, D. H. J. Chem. Soc., Perkin Trans. 1 1985,
2747-2748. Rinehart, K. L.; Harada, K.; Namikoshi, M.; Chen, C.; Harvis,
C. A.; Munro, M. H. G.; Blunt, J. W.; Mulligan, P. E.; Beasley, V. R.;
Dahlem, A.; Carmichael, W. J. Am. Chem. Soc. 1988, 110, 8557-8558.
(e) Tautomycin: Cheng, X.-C.; Ubukata, M.; Isono, K. J. Antibiot. 1990,
43, 809-819. (f) Cantharidine: Li, Y.-M.; Casida, J. E. Proc. Natl. Acad.
Sci. U.S.A. 1992, 89, 11867-11870. (g) Motuporin: Valentekovich, R. J.;
Schreiber, S. L. J.Am. Chem. Soc. 1995, 117, 9069-9070. (h) Review:
Sheppeck, J. E., II; Gauss, C.; Chamberlin, A. R. Bioorg. Med. Chem. 1997,
5, 1739-1750.
(5) Topoisomerase II inhibitors which induce G2 arrest are described in
the following references. (a) Etoposide: Chen, G. L.; Yang, L.; Rowe, T.
C.; Halligan, B. D.; Tewey, K. M.; Liu, L. F. J. Biol. Chem. 1984, 259,
13560-13566. (b) Doxorubicin: Tewey, K. M.; Rowe, T. C.; Yang, L.;
Halligan, B. D.; Liu, L. F. Science 1984, 266, 466-468. (c) Amsacrine:
Nelson, E. M.; Tewey, K. M.; Liu, L. F. Proc. Natl. Acad. Sci. U.S.A. 1984,
81, 1361-1365.
(6) Cheng, A.; Balczon, R.; Zuo, Z.; Koons, J. S.; Walsh, A. H.;
Honkanen, R. E. Cancer Res. 1998, 58, 3611-3619.
(7) Roberge, M.; Tudan, C.; Hung, S. M. F.; Harder, K. W.; Jirik, F. R.;
Anderson, H. Cancer Res. 1994, 54, 6115-6121.
(8) Walsh, A. H.; Cheng, A.; Honkanen, R. E. FEBS Lett. 1997, 416,
230-234.
10.1021/ja010195q CCC: $20.00 © 2001 American Chemical Society
Published on Web 04/11/2001

4162 J. Am. Chem. Soc., Vol. 123, No. 18, 2001
Boger et al.
mitotic entry checkpoint and selective protein phosphatase
inhibition are novel, clinically unexplored mechanisms worthy
of pursuit for introduction of a new class of antitumor agents.
Also contributing to its selective antitumor properties, fostriecin
was shown to be transported into tumor cells via the reduced
folate carrier system.14 More recently, this selective and potent
inhibition of PP2A by fostriecin has also been shown to limit
myocardial infarct size and to protect cardiomyocytes during
ischemia.15-17
Scheme 1
At the time the significance of the biological and therapeutic
effects of fostriecin was first emerging, only the drug’s two-
dimensional structure had been established,18 and its relative
and absolute stereochemistry was unknown. In initial efforts to
address these limitations to the clinical potential of fostriecin,
we first established its relative and absolute stereochemistry.19,20
This work also confirmed stereochemical assignments21 for a
family of structurally related natural products that include the
leustroducsins,22 phoslactomycins,23 phosphazomycins,23 and
phospholine (Figure 1).23 Recently, phoslactomycin F was also
found to be a selective, albeit substantially less potent, inhibitor
of PP2A (IC50 ) 4.7 µM) versus PP1 (IC50 > 1000 µM),24 and
these similarities (PP2A . PP1) and distinctions (1000-fold less
potent) with fostriecin reveal important structural features
contributing to PP2A inhibition. However, the biological
properties of these related natural products typically differ from
those of fostriecin. In addition to broad-spectrum antifungal
activities, the leustroducsins and phoslactomycins also induce
production of macrophage colony-stimulating factors by bone
marrow stromal cells contributing to the recovery of blood
leucocytes.
diastereomer of dephosphorylated fostriecin. In addition, the
route detailed herein provides access to structural analogues not
accessible by degradation or functionalization of the natural
product itself and is sufficiently general as to be adaptable to
the leustroducsins, phoslactomycins, and phospholine, for which
no synthetic efforts have yet been disclosed. Key elements of
the approach include the late-stage deprotection and sequential
introduction of the C-terminus lactone and C9 phosphate, a
Wadsworth-Horner-Emmons installation of the C6-7 trans
double bond, and stepwise assemblage of the sensitive Z,Z,E-
triene (Scheme 1). The relative and absolute stereochemistry
of the C9 center was set by utilizing an optically active starting
material (D-Glu), the C5 and C11 stereocenters were installed
via Sharpless AD reactions, and the labile C8 center was
installed enlisting an R-OSiR3 enforced Felkin-Anh versus
chelation-controlled nucleophilic addition. Throughout the
synthesis, the sensitivity of the conjugated Z,Z,E-triene and the
three allylic alcohols limited reagent selections and dictated
attentive handling during isolation or purification. By design,
the tertiary C8 allylic alcohol was introduced late in the
synthesis.
Herein, we report the first total synthesis of fostriecin that
not only serves to confirm the structural and stereochemical
assignments (5R,8R,9R,11R) of 1 but also possesses the potential
of addressing the issues of chemical purity and storage stability
limiting clinical trials on the natural product. In this regard, the
efforts are complementary to the early and only other synthetic
efforts disclosed to date on 1 by Just and O’Connor,20 which
culminated in the preparation of an unnatural 5R,8R,9S,11R
(14) Fry, D. W.; Besserer, J. A.; Boritzki, T. J. Cancer Res. 1984, 44,
3366-3370.
(15) Armstrong, S. C.; Kao, R.; Gao, W.; Shivell, L. C.; Downey, J.
M.; Honkanen, R. E.; Ganote, C. E. J. Mol. Cell. Cardiol. 1997, 29, 3009-
3024.
(16) Armstrong, S. C.; Gao, W.; Lane, J. R.; Ganote, C. E. J. Mol. Cell.
Cardiol. 1998, 30, 61-73.
(17) Weinbrenner, C.; Baines, C. P.; Liu, G.-S.; Armstrong, S. C.; Ganote,
C. E.; Walsh, A. H.; Honkanen, R. E.; Cohen, M. V.; Downey, J. M.
Circulation 1998, 98, 899-905.
(18) Early work assigned the absolute configuration at C5 as R, see:
Hokanson, G. C.; French, J. C. J. Org. Chem. 1985, 50, 462-466.
(19) Boger, D. L.; Hikota, M.; Lewis, B. M. J. Org. Chem. 1997, 62,
1748-1753.
(20) Just, G.; O’Connor, B. Tetrahedron Lett. 1988, 29, 753-756.
(21) The absolute stereochemistry of the leustroducsins, phoslactomycins,
and phospholine was tentatively assigned on the basis of the chiroptical
properties of Mosher esters of the alcohols. Shibata, T.; Kurihara, S.; Yoda,
K.; Haruyama, H. Tetrahedron 1995, 51, 11999-12012.
(22) Leustroducsins: Kohama, T.; Enokita, R.; Okazaki, T.; Miyaoka,
H.; Torikata, A.; Inukai, M.; Kaneko, I.; Kagasaki, T.; Sakaida, Y.; Satoh,
A.; Shiraishi, A. J. Antibiot. 1993, 46, 1503-1511. Kohama, T.; Nakamura,
T.; Kinoshita, T.; Kaneko, I.; Shiraishi, A. J. Antibiot. 1993, 46, 1512-
1519.
Synthesis of the C1-C6 Subunit of Fostriecin. The
precursor to the fostriecin lactone was derived from 2, which
was prepared in four steps from p-methoxybenzyl 5-hexenoate
enlisting a Sharpless AD reaction as described in our earlier
structural assignment work on 1.19 Thus, oxidation of 2 to the
corresponding selenoxide (H2O2) and in situ elimination (85%),
Dibal-H reduction of the resulting unsaturated lactone 3,
followed by protection of the lactol provided 425 (90%, two
steps), Scheme 2. Deprotection of the TBDPS ether (Bu4NF,
91%) and oxidation of the resulting alcohol 525 with TPAP/
NMO provided the aldehyde 625 (90%). Embodied in this key
(23) Phoslactomycins: Fushimi, S.; Nishikawa, S.; Shimazu, A.; Seto,
H. J. Antibiot. 1989, 42, 1019-1025. Fushimi, S.; Furihata, K.; Seto, H. J.
Antibiot. 1989, 42, 1026-1036. Phosphazomycins: Tomiya, T.; Uramoto,
M.; Isono, K. J. Antibiot. 1990, 43, 118-121. Phospholine: Ozasa, T.;
Tanaka, K.; Sasamata, M.; Kaniwa, H.; Shimizu, M.; Matsumoto, H.;
Iwanami, M. J. Antibiot. 1989, 42, 1339-1343. Ozasa, T.; Suzuki, K.;
Sasamata, M.; Tanaka, K.; Kobori, M.; Kadota, S.; Nagai, K.; Saito, T.;
Watanabe, S.; Iwanami, M. J. Antibiot. 1989, 42, 1331-1338.
(24) Usui, T.; Marriott, G.; Inagaki, M.; Swarup, G.; Osada, H. J.
Biochem. (Tokyo) 1999, 125, 960-965.
(25) The intermediates 4-6, 30-33, and 38 were isolated as a single
diastereomer of undefined stereochemistry at the anomeric center. We have
assigned this as trans-(2R,6R), in agreement with an independent recent
report of 5 and 6; see: Christmann, M.; Bhatt, U.; Quitschalle, M.; Claus,
E.; Kalesse, M. Angew. Chem., Int. Ed. 2000, 39, 4364-4366; Bhatt, U.;
Christmann, M.; Quitschille, M.; Claus, E.; Kalesse, M. J. Org. Chem. 2001,
66, 1885-1893. See also: Murakami, N.; Wang, W.; Aoki, M.; Tsutsui,
Y.; Higuchi, K.; Aoki, S.; Kobayashi, M. Tetrahedron Lett. 1997, 38, 5533-
5536.

Total Synthesis of Fostriecin
J. Am. Chem. Soc., Vol. 123, No. 18, 2001 4163
Scheme 2
Scheme 3
C-terminus subunit of 1 is the C5 stereocenter whose R absolute
stereochemistry was set through a Sharpless AD reaction.
Synthesis of the C7-C18 Subunit of Fostriecin. The C7-
C18 subunit of 1 was prepared by first assembling C8-C12
with introduction of the C9 and C11 stereocenters followed by
stepwise introduction of the sensitive triene and finally conver-
sion to the â-ketophosphonate 27 for linkage to C1-C6. The
stepwise order to the triene introduction permitted the assessment
of the relative stability of the C12-C13 Z-alkene, the C12-
C15 Z,Z-diene, and the C12-C17 Z,Z,E-triene and offered the
potential of postponing the introduction of any of its sensitive
segments to a late stage of the synthesis. PMB protection (95%)
and Dibal-H reduction (96%) of optically active lactone 7,
available in two steps from D-glutamic acid,26 followed by
subsequent mesylation and in situ elimination27 provided the
dihydrofuran 10 (73% from 8), incorporating the C9 stereocenter
(Scheme 3). Extensive dimerization of the lactol 9 was observed
unless the mesylation was conducted under dilute reaction
conditions (ca. 0.02-0.03 M) in the presence of excess Et3N
(15 equiv) at temperatures below -20 °C. Sharpless asymmetric
dihydroxylation28 on 10, employing (DHQD)2AQN,29 gave a
>10:1 ratio of diol 11 (100%) as an anomeric mixture with
clean delivery of the C2 alcohol cis to the C4 hydroxymethyl
substituent installing the fostriecin C11 center (Scheme 3).
After empirical experimentation, selective protection of the
C2 secondary alcohol over the anomeric C1 alcohol was
achieved by low-temperature (-78 °C) treatment of 11 with
TBSOTf (1.0 equiv, CH2Cl2) in the presence of Et3N (3.0 equiv),
providing 12 in superb yield (90%, >10:1 anomeric mixture).
Analogous, but less effective, conversions were achieved with
TBSOTf-lutidine (1/1.5 equiv, CH2Cl2, -78 °C, 1.5 h, 69%
based on recovered 11) or through conversion of diol 11 to the
dibutylstannylene (1.1 equiv of Bu2SnO, CH3OH, reflux, 1.5
h),30,31 followed by silylation with CF3CONMeTBS/Et3N (3-6
equiv, DMF, 0-5 °C, 3 h, 30-54%). In the latter case, the
intermediate silylation of the anomeric alcohol was detected as
the predominant product (0.5 h, 0-5 °C) and observed to
undergo silyl transfer to the C2 alcohol (3 h, 25 °C), Scheme
4.
Scheme 4
Lactol 12 was condensed with the Still-Gennari phospho-
nate32 ((CF3CH2O)2P(O)CH2CO2Me, KHMDS, 18-crown-6,
THF, -78 °C) to afford 13 (88%, Z:E ) 29:1), and this step
constitutes introduction of the first cis olefin of the sensitive
Z,Z,E-triene. Notably, 13 cyclized to the corresponding tetrahy-
drofuran32 by Michael addition of the liberated alkoxide onto
the newly generated Z unsaturated ester if excess KHMDS was
used as the base, and this could be avoided by careful use of a
stoichiometric amount of base. An alternative approach involv-
(26) Herdeis, C. Synthesis 1986, 232-233. Lee, Y. S.; Kang, S. S.; Choi,
J. H.; Park, H. Tetrahedron 1997, 53, 3045-3056.
(27) Takle, A.; Kocienski, P. Tetrahedron 1990, 46, 4503-4516.
(28) Kolb, H. C.; VanNieuwenhze, M. S.; Sharpless, K. B. Chem. ReV.
1994, 94, 2483-2547.
(29) Becker, H.; Sharpless, K. B. Angew. Chem., Int. Ed. Engl. 1996,
35, 448-451.
(32) Still, W. C.; Gennari, C. Tetrahedron Lett. 1983, 24, 4405-4408.
For the cyclized tetrahydrofuran: colorless oil, [R]22D -22 (c. 1.0, CHCl3);
1H NMR (CDCl3, 500 MHz) δ 7.26 (2H, d, J ) 8.8 Hz), 6.86 (2H, d, J )
8.8 Hz), 4.52 (1H, d, J ) 11.8 Hz), 4.48 (1H, d, J ) 11.8 Hz), 4.25 (1H,
m), 4.13 (2H, m), 3.80 (3H, s), 3.69 (3H, s), 3.58 (1H, dd, J ) 6.6, 9.9
Hz), 3.43 (1H, dd, J ) 5.1, 9.9 Hz), 2.50 (1H, dd, J ) 5.5, 15.0 Hz), 2.47
(30) David, S.; Hanessian, S. Tetrahedron 1985, 41, 643-663. De
Bernardo, S.; Tengi, J. P.; Sasso, G. J.; Weigele, M. J. Org. Chem. 1985,
50, 3457-3462.
(31) Nicolaou, K. C.; van Delft, F. L.; Conley, S. R.; Mitchell, H. J.;
Jin, Z.; Rodriguez, R. M. J. Am. Chem. Soc. 1997, 119, 9057-9058.

4164 J. Am. Chem. Soc., Vol. 123, No. 18, 2001
Boger et al.
ing reaction of lactol 12 with iodomethylenetriphenylphos-
phorane33 to provide the corresponding Z-iodoalkene was not
successful. After protection of 13 with Et3SiOTf (95%), the
resulting ester 14 was reduced with Dibal-H to give alcohol 15
(90%), which was subsequently oxidized with Dess-Martin
periodinane34 to provide aldehyde 16 (100%). Attempts to
homologate 16 to the corresponding Z,Z-iododiene enlisting
iodomethylenetriphenylphosphorane33 provided the desired io-
dide35 (78% from 15, Z:E ) 4.2:1) but with modest stereose-
lection. Not only were the two isomers not separable by
chromatography, but the product proved labile to isomerization
during attempts at purification. Consequently, we employed a
two-step procedure enlisting the Corey and Fuchs reaction36 of
aldehyde 16 to first obtain dibromide 17 (94% from 15),
followed by selective reduction of 17 with Bu3SnH-
Pd(PPh3)4,37 to cleanly yield the substantially more stable cis-
vinyl bromide 18.35 In contrast to the Z,Z-iododiene, 18 proved
stable to handling during purification. In initial efforts to
elaborate 18 to the requisite Z,Z,E-triene unit found in the natural
product, the removal of PMB group following Stille coupling
with 22 proved problematic. Consequently, the PMB group was
exchanged for an acetate by treatment of 17 with DDQ (95%)
and acetylation of 19 to provide 20 (93%). Selective reduction
of 20 with Bu3SnH-Pd(PPh3)4 furnished cis-vinyl bromide 21
(84%) with only occasional and trace generation of the over-
reduced product (5-10%). Stille coupling38 of 21 with vinyl-
stannane 2239 was found to be low-yielding under typical
conditions. Enlisting conditions first established with a model
substrate (Scheme 5), the Stille coupling of 21 with 22
conducted in i-Pr2NEt provided the key intermediate, Z,Z,E-
triene 23, in excellent yield (82%) with nearly perfect stereo-
chemical intergrity (g98:2). While 21 and 23 are only margin-
ally stable at room temperature when stored as oils, no
decomposition or isomerization was observed over several
months when they were stored as solutions in nonpolar solvents
at 0-5 °C. Although deacetylation of 23 using standard
Scheme 5
hydrolytic methods (K2CO3/MeOH, LiOH/THF/H2O) led to
acetate removal followed by migration and/or desilylation of
the neighboring triethylsilyl group, reduction with Dibal-H gave
the desired primary alcohol 24 (98%), which was converted to
aldehyde 25 by Dess-Martin oxidation34 (91%).
The remaining conversion of 25 to the ketophosphonate 27
set the stage for a Wadsworth-Horner-Emmons reaction to
unite the C1-C6 and C7-C18 subunits. Initial attempts to add
the anion derived from diethyl methylphosphonate to aldehyde
25 in THF met with limited success. However, when the reaction
was performed in the nonpolar solvent toluene, superb conver-
sions (1:1 mixture of diastereomers) were observed, and the
â-ketophosphonate 27 was obtained after Dess-Martin oxida-
tion of 26 (90% for two steps).40
Synthesis of Dephosphoryl Fostriecin: Correlation and
Confirmation of Structure. Prior to completing the total
synthesis of 1, we first wanted to correlate our synthetic
intermediates with those of the natural product to ensure they
possessed the C8 natural stereochemistry. This was accom-
plished by linking the C1-C6 and C7-C18 subunits, the
diastereocontrolled introduction of the C8 methyl group,
elaboration of the lactone, and correlation with fostriecin
derivatives that lack the phosphate. Initial efforts first condensed
a â-ketophosphonate related to 27 with the sensitive lactone
aldehyde 28,19 which provided 2941 in modest conversions (10-
36%). The best conversions were obtained by using the Roush-
Masamune procedure42 with aldehyde 28 generated in situ by
Swern oxidation of the corresponding alcohol as described19
and appeared to be limited by the sensitivity of the aldehyde to
the basic reaction conditions (eq 1). This, coupled with our
inability to cleanly add a methyl nucleophile to the C8 carbonyl
versus the lactone, led to the use of the lactone precursor 6.
Wadsworth-Horner-Emmons reaction of â-ketophosphonate
(1H, dd, J ) 7.0, 15.0 Hz), 2.22 (1H, ddd, J ) 6.2, 7.3, 12.8 Hz), 1.71
(1H, ddd, J ) 5.9, 11.8 Hz) 0.87 (9H, s), 0.05 (3H, s), 0.04 (3H, s); IR
(neat) νmax 2942, 1734, 1515, 1247, 1093, 837 cm-1; MALDIFTMS (DHB)
m/z 447.2175 (M + Na+ C22H36O6Si requires 447.2173).
(33) Stork, G.; Zhao, K. Tetrahedron Lett. 1989, 30, 2173-2174.
(34) Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113, 7277-
7287.
(35) Characterization for iodide (4.2:1 Z,Z-iododiene versus E,Z-iodo-
diene): 1H NMR (Z,Z-iododiene) δ 7.68 (1H, m), 7.19 (2H, m), 6.84 (2H,
m), 6.85 (1H, m), 6.25 (1H, t, J ) 10.9 Hz), 6.08 (1H, d, J ) 7.6 Hz), 5.70
(1H, t, J ) 10.1 Hz), 4.97 (1H, m), 4.32 (2H, s), 4.23 (1H, m), 3.36 (2H,
m), 3.30 (3H, s), 2.00 (1H, m), 1.82 (1H, m), 1.18 (9H, m), 1.10 (9H, s),
1
0.08 (6H, m), 0.21 (3H, s), 0.17 (3H, s); H NMR (E,Z-iododiene) δ 7.64
(1H, m), 7.21 (2H, m), 6.86 (2H, m), 6.00 (1H, d, J ) 14.0 Hz), 5.56 (1H,
d, J ) 11.1 Hz), 5.44 (1H, t, J ) 9.7 Hz), 5.15 (1H, m), 4.37 (2H, s), 4.25
(1H, m), 3.37 (2H, m), 3.31 (3H, s), 2.00 (1H, m), 1.82 (1H, m), 1.18 (9H,
m), 1.10 (9H, s), 1.08 (6H, m), 0.22 (3H, s), 0.18 (3H, s); IR (neat) νmax
2936, 2864, 1241, 1092 cm-1; FABHRMS (NBA-CsI) m/z 765.1248 (M
+ Cs+, C28H49IO4Si2 requires 765.1269). For 18: 1H NMR (C6D6, 400
MHz) δ 7.21-7.23 (2H, m), 6.91 (1H, dd, J ) 7.5, 11.0 Hz), 6.82-6.85
(2H, m), 6.38 (1H, dd, J ) 10.5, 11.0 Hz), 5.88 (1H, d, J ) 7.5 Hz), 5.65
(1H, dd, J ) 9.5, 10.5 Hz), 4.94-4.99 (1H, m), 4.31-4.38 (2H, m), 4.19-
4.23 (1H, m), 3.41 (1H, dd, J ) 6.0, 9.5 Hz), 3.36 (1H, dd, J ) 4.9, 9.5
Hz), 3.33 (3H, s), 1.97 (1H, ddd, J ) 4.4, 8.2, 13.8 Hz), 1.73 (1H, ddd, J
) 4.6, 7.3, 13.8 Hz), 1.10 (9H, t, J ) 7.9 Hz), 1.02 (9H, s), 0.76 (6H, q,
J ) 7.9 Hz), 0.17 (3H, s); FABHRMS (NBA-CsI) m/z 717.1428 (M +
Cs+, C28H49BrO4Si2 requires 717.1407).
(40) Initial studies conducted with the analogous TBS-protected vinnyl-
stannane related to 22 provided the corresponding C18 OTBS derivatives
of 23-27, 30, and 31. Selective deprotection of the C9 OTES group was
not achieved, leading to the adoption of reagent 22. Characterization data
for these intermediates are provided in the Supporting Information.
(41) Characterization for 29: 1H NMR (CD3CN, 500 MHz) δ 7.02-
6.97 (1H, m), 6.92-6.84 (1H, m), 6.80-6.75 (1H, m), 6.60-6.56 (1H, m),
6.19-6.13 (1H, m), 5.97-5.86 (1H, m), 5.82-5.66 (3H, m), 5.59-5.64
(1H, m), 5.23-5.09 (2H, m), 4.70-4.67 (1H, m), 4.13-4.12 (2H, m), 2.28-
2.12 (2H, m), 2.08-1.92 (1H, m), 1.68-1.58 (1H, m), 1.13-0.90 (27H,
m), 0.80-0.72 (6H, m), 0.24-0.18 (6H, m), 0.09-0.08 (6H, m); FAB-
HRMS (NBA-CsI) m/z 825.3367 (M + Cs+, C37H68O6Si3 requires
825.3378).
(36) Corey, E. J.; Fuchs, P. L. Tetrahedron Lett. 1972, 3769-3772.
(37) Uenishi, J.; Kawahama, R.; Yonemitsu, O.; Tsuji, J. J. Org. Chem.
1998, 63, 8965-8975. Uenishi, J.; Kawahama, R.; Shiga, Y.; Yonemitsu,
O.; Tsuji, J. Tetrahedron Lett. 1996, 37, 6759-6762. Uenishi, J.; Kawahama,
R.; Yonemitsu, O.; Tsuji, J. J. Org. Chem. 1996, 61, 5716-5717.
(38) Evans, D. A.; Black, W. C. J. Am. Chem. Soc. 1993, 115, 4497-
4513. Gibbs, R. A.; Krishnan, U. Tetrahedron Lett. 1994, 35, 2509-2512.
(39) Jung, M. E.; Light, L. A. Tetrahedron Lett. 1982, 23, 3851-3854.
Bansal, R.; Cooper, G. F.; Corey, E. J. J. Org. Chem. 1991, 56, 1329-
1332.

Total Synthesis of Fostriecin
J. Am. Chem. Soc., Vol. 123, No. 18, 2001 4165
Scheme 6
(THF) gave predominantly the undesired 1,4-adduct43 (90%,
>10:1). Alternative conditions including MeLi/18-crown-6,44
MeLi/LiCl, and MeTi(iOPr)3 either did not substantially
45
improve the selectivity or led to no reaction. Even the combina-
tion of MeLi-CeCl3,46 which gave the most encouraging results,
provided a mixture of 1,2-adduct 31 versus 1,4-adduct in THF.
Ultimately, we established that when ketone 30 was introduced
as a toluene solution into the MeLi/CeCl3 slurry in THF, the
1,2-adduct 3125 was formed preferentially (>20:1) in 96% yield
as a mixture of two diastereomers (3:1, 1H NMR). Correlation
of the major isomer with natural product derivatives as detailed
below established its stereochemistry as 8R as expected and
derived from a Felkin-Anh addition. We briefly explored
whether this diastereoselection could be improved by increasing
the size of the C9 O-silyl protecting group. However, an
analogous ratio of diastereomers was obtained with the corre-
sponding C9 OTBS derivative.
Deprotection of the secondary C9 TES ether along with the
exchange of the isopropyl acetal to an ethyl acetal was effected
by treatment of 31 with PPTS-EtOH (25 °C, 3.5 h, 93%) and
was required to permit eventual introduction of the C9 phos-
phate.47 However, prior to this completion of the natural product
synthesis, correlation of synthetic and naturally derived 36 was
carried out to ensure our intermediates bore the correct relative
and absolute configurations. Without optimization, reprotection
of the secondary alcohol as its OTBS ether 3325 (7.2 equiv of
TBSOTf, Et3N, CH2Cl2, -78 °C, 4 h, 77%) followed by aqueous
acid hydrolysis of the acetal (0.5 N aqueous HCl-acetone 1:4,
0 °C, 1 h, 50%) provided the lactol 34, which was cleanly
oxidized to the lactone 35 (30 equiv of 50% Ag2CO3-Celite,
benzene, 80 °C, 2 h, 80%). Deprotection of 35 (HF‚pyr,
pyridine-THF, 25 °C, 4 d) furnished a 3:1 mixture of C8
diastereomers, of which the major diastereomer correlated with
the tetraol 36 derived from dephosphorylation of the natural
product18,19 (HPLC: Porasil column, 3.9 mm × 300 mm,
0-50% EtOAc-hexane gradient elution, 1.0 mL/min, Rt ) 9.45
min, minor peak at Rt ) 8.72 min; NovaPak C18 column, 3.9
mm × 300 mm, 0-50% CH3CN-H2O gradient elution, 1.0
mL/min, Rt ) 18.43 min, minor peak at Rt ) 18.86 min).
In addition, the 9,11,18-triacetate 37 was prepared from
naturally derived 36 by acetylation (Ac2O, pyr, 5 °C, 5 h, 70%)18
and correlated with the major isomer of a sample prepared from
synthetic 33 (i. Bu4NF, THF, 30 min, 25 °C; ii. Ac2O, Et3N,
DMAP, CH2Cl2, 30 min, 25 °C, 75% for two steps; iii. 0.5 N
aqueous HCl-acetone 1:4, 0 °C, 45 min; iv. 50% Ag2CO3-
Celite, benzene, 80 °C, 1.5 h, 41% for two steps) by HPLC
(Chiralcel OD column, 4.6 mm × 250 mm, 10% i-PrOH-
hexane, 1.0 mL/min, Rt ) 48.3 min, minor peak at Rt ) 41.3
min) and 1H NMR (C8-CH3 at δ 1.25, minor isomer C8-CH3
at δ 1.28).
27 with aldehyde 6 was carried out smoothly in toluene,
furnishing the desired trans R,â-unsaturated ketone 3025 (91%),
Scheme 6.
These studies permitted spectroscopic and HPLC correlation
of the synthetic samples with the naturally derived samples of
36 and 37. However, none of the intermediate C8 diastereomers
were chromatographically separable to the extent that they were
(43) Leonard, J.; Mohialdin, S.; Reed, D.; Ryan, G.; Swain, P. A.
Tetrahedron 1995, 51, 12843-12858.
(44) Yamamoto, Y.; Maruyama, K. J. Am. Chem. Soc. 1985, 107, 6411-
6413.
(45) Weidmann, B.; Seebach, D. HelV. Chim. Acta 1980, 63, 2451-
2454.
Direct addition of MeLi to 30 via Felkin-Anh attack was
anticipated to provide 31. However, treatment of 30 with MeLi
(46) Imamoto, T. Pure Appl. Chem. 1990, 62, 747-752.
(47) When this reaction was conducted on the corresponding C18 OTBS
derivative, isolation of the C8,C9,C18-triol was observed. Treatment of this
triol with TBDPSCl (7 equiv, Et3N, DMAP, CH2Cl2, 1 h, 86%) also
provided 32. Characterization data for C8,C9,C18-triol are provided in the
Supporting Information.
(42) Blanchette, M. A.; Choy, W.; Davis, J. T.; Essenfeld, A. P.;
Masamune, S.; Roush, W. R.; Sakai, T. Tetrahedron Lett. 1984, 25, 2183-
2186.

4166 J. Am. Chem. Soc., Vol. 123, No. 18, 2001
Boger et al.
Scheme 7
Scheme 8
extensively investigated, direct methods of phosphate introduc-
tion (POCl3, proton sponge; Bu4N‚H2PO4, Cl3CCN) were not
successful and simply provided recovered starting material. The
global deprotection of 41 was carried out in two stages.
Prolonged treatment with unbuffered HF (5% H2O-CH3CN, 3
d) required for acid-catalyzed PMB ester deprotection led to
extensive degradation and olefin isomerization, which precluded
the ability to isolate 1 from the reaction mixture. The alternative
treatment with HF‚pyr (pyr-THF, 4 d) removed the silyl ether
protecting groups but failed to cleave the PMB esters cleanly.
Thus, we employed a two-stage deprotection, enlisting an initial,
brief treatment with HF (5% H2O-CH3CN, 15 min) to remove
the PMB esters, followed by addition of pyridine (25% pyr-
CH3CN/H2O) to conduct the slow silyl ether deprotections under
buffered conditions. Under these conditions, the C18 and C11
O-silyl protecting groups were removed relatively rapidly (1-2
d), whereas the hindered tertiary C8 OTBS required prolonged
reactions times.
Degradation and Functionalization of Natural Fostriecin.
Degradation and functionalization of natural fostriecin provided
authentic samples of our correlation intermediates (Scheme 8).
The tetraol 36 was obtained from 1 by alkaline phosphatase
hydrolysis (100%) as previously described.18,19 Selective protec-
tion of the primary alcohol was accomplished by treatment of
1 with TBDPSCl (91%), providing 42 as described.19 Conver-
sion of 42 to the correlation sample 40 could be accomplished
by sequential or concurrent C11 and C8 OTBS ether formation.
Thus, treatment of 42 with TBSCl (1.5 equiv, 3.0 equiv of
imidazole, DMF, 25 °C, 1 h) cleanly provided 43 (62%) derived
from selective C11 OTBS ether formation. Without optimiza-
tion, treatment of 43 with TBSOTf (3.0 equiv) in the presence
of 2,6-lutidine (6.0 equiv) at -20 °C (3 h, CH2Cl2) provided
40 (45%), derived from protection of the tertiary C8 versus
secondary C9 alcohol, and an equivalent amount of 35 derived
from secondary C9 alcohol protection. Like the silylation
reaction of 32, the generation of 40 presumably proceeds by
C9 O-silyl ether formation and subsequent migration to provide
the more stable C8 OTBS ether and is facilitated by conducting
the reaction at -20 versus -78 °C. Compound 40 derived from
natural fostriecin proved to be identical in all respects (1H NMR,
13C NMR, [R]25D, IR, MS) with synthetic material. The
correlation sample 40 could be prepared even more conveniently
by direct TBS silylation of 42 (4.0 equiv of TBSOTf, 8.0 equiv
of lutidine, CH2C12, -20 °C, 3 h) in higher conversion (54%).
preparatively useful. Thus, this preliminary correlation was
conducted with the 3:1 mixture of C8 diastereomers.
Completion of the Total Synthesis of Fostriecin. In the
course of handling 31 and its related C9 OTBS derivative 34,
a base-catalyzed migration of the C9 O-silyl group to the
adjacent C8 tertiary alcohol was observed. We took advantage
of this observation and developed conditions whereby treatment
of 32 with TBSOTf (2.0 equiv, 5 equiv of 2,6-lutidine, -20
°C, 45 min) provided predominantly 3825 (70%) (Scheme 7).
Thus, silylation at -78 °C cleanly provided 33 (77%) as detailed
in Scheme 6, but when this reaction was conducted at the higher
temperature of -20 °C, a subsequent C9-to-C8 O-silyl migration
presumably occurred to provide primarily 38. Hydrolysis to the
lactol 39 and oxidation to the lactone cleanly provided 40, at
which point the problematic separation of the C8 diastereomers
was straightforward. Thus, only four intermediates in the total
synthesis (31, 32, 38, and 39) were prepared and characterized
as the 3:1 mixture of C8 diastereomers. Synthetic 40 was found
to be identical in all respects with a sample of 40 prepared by
degradation and functionalization of 1 (1H NMR, 13C NMR,
IR, MS, [R]25D). Global silyl ether deprotection of 40 was
accomplished by treatment with HF‚pyr in pyridine-THF (25
°C, 4 d) with a solid NaHCO3 workup, providing the tetraol
36, which also proved to be spectroscopically and chromato-
graphically indistinguishable from naturally derived 36.18,19
Conversion of 40 to the bis-PMB-protected48 C9 phosphate upon
treatment with PCl3 (4 equiv, pyridine, 25 °C) followed by
PMBOH (9 equiv) and subsequent phosphite oxidation H2O2-
H2O) provided 41 in superb conversion (91%), which upon
global silyl ether deprotection (HF; HF-pyridine) provided
fostriecin that was identical in all aspects with authentic material.
Notably, oxidation of the intermediate phosphite (31P NMR δ
141.0) to the phosphate (31P NMR δ -0.28) was slow and could
be monitored by 31P NMR, and the bis-PMB phosphate ester
was stable to purification by chromatography. Although not
Conclusions
The first total synthesis of fostriecin (CI-920) is disclosed
which serves to confirm its relative and absolute stereochemical
(48) Evans, D. A.; Gage, J. R.; Leighton, J. L. J. Am. Chem. Soc. 1992,
114, 9434-9453.

Total Synthesis of Fostriecin
J. Am. Chem. Soc., Vol. 123, No. 18, 2001 4167
assignments. Improvements in this approach and its extension
to the preparation of key analogues and partial structures of the
natural product are in progress in efforts to define the structural
features of 1 contributing to the potent and selective protein
phosphatase inhibition, and these efforts will be disclosed in
due course.
preliminary studies leading to 17, and initial studies on formation
of the Z,Z,E-triene. We thank Dr. Robert J. Schultz of the Drug
Synthesis and Chemistry Branch, Developmental Therapeutics
Program, Division of Cancer Treatment and Diagnosis, National
Cancer Institute, for the generous supply of natural fostriecin
(NSC 339638).
Supporting Information Available: Full experimental
details of the total synthesis of 1, details of the coupling defined
in Table 1, full experimental details and characterization of the
corresponding C18 OTBS route (Supporting Information Scheme
1), the correlation conversions of 32 to 36 and 37, and the
preparations of 43 and 40 from 42 (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
Experimental Section
Full experimental details are provided as Supporting Informa-
tion.
Acknowledgment. We gratefully acknowledge financial
support of the National Institute of Health (CA42056) and The
Skaggs Institute for Chemical Biology. We thank Drs. B. Lewis
and Y.-S. Jung for development of the synthesis of 11,
JA010195Q