Chemical synthesis and biological evaluation of cis- and trans-12,13-cyclopropyl and 12,13-cyclobutyl epothilones and related pyridine side chain analogues

Article abstract of DOI:10.1021/ja011338b

The design, chemical synthesis, and biological evaluation of a series of cyclopropyl and cyclobutyl epothilone analogues (3-12, Figure 1) are described. The synthetic strategies toward these epothilones involved a Nozaki - Hiyama - Kishi coupling to form

Full text of DOI:10.1021/ja011338b

J. Am. Chem. Soc. 2001, 123, 9313-9323
9313
Chemical Synthesis and Biological Evaluation of cis- and
trans-12,13-Cyclopropyl and 12,13-Cyclobutyl Epothilones and
Related Pyridine Side Chain Analogues
K. C. Nicolaou,*,† Kenji Namoto,† Andreas Ritze´n,† Trond Ulven,† Mitsuru Shoji,† Jim Li,†
Gina D’Amico,‡ Dennis Liotta,‡ Christopher T. French,§ Markus Wartmann,
Karl-Heinz Altmann, and Paraskevi Giannakakou§
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, Department
of Chemistry and Biochemistry, UniVersity of California, San Diego, 9500 Gilman DriVe,
La Jolla, California 92093, Chemistry Department, Emory UniVersity, 1521 Pierce DriVe,
Atlanta, Georgia 30322, Winship Cancer Institute, Emory UniVersity School of Medicine,
1365-B Clifton Rd., Atlanta, Georgia 30322, and NoVartis Pharma AG, TA Oncology Research,
CH-4002, Basel, Switzerland
ReceiVed May 31, 2001
Abstract: The design, chemical synthesis, and biological evaluation of a series of cyclopropyl and cyclobutyl
epothilone analogues (3-12, Figure 1) are described. The synthetic strategies toward these epothilones involved
a Nozaki-Hiyama-Kishi coupling to form the C15-C16 carbon-carbon bond, an aldol reaction to construct
the C6-C7 carbon-carbon bond, and a Yamaguchi macrolactonization to complete the required skeletal
framework. Biological studies with the synthesized compounds led to the identification of epothilone analogues
3, 4, 7, 8, 9, and 11 as potent tubulin polymerization promoters and cytotoxic agents with (12R,13S,15S)-
cyclopropyl 5-methylpyridine epothilone A (11) as the most powerful compound whose potencies (e.g. IC50
) 0.6 nM against the 1A9 ovarian carcinoma cell line) approach those of epothilone B. These investigations
led to a number of important structure-activity relationships, including the conclusion that neither the epoxide
nor the stereochemistry at C12 are essential, while the stereochemistry at both C13 and C15 are crucial for
biological activity. These studies also confirmed the importance of both the cyclopropyl and 5-methylpyridine
moieties in conferring potent and potentially clinically useful biological properties to the epothilone scaffold.
Introduction
icant feature in these structures for biological activity, as
opposed to the epoxide oxygen whose presence appears to
be less relevant, at least for in vitro biological activity.3,4
Remarkably, even the trans analogues 7, 8, and 11 show good
With some members in clinical trials, the epothilones com-
mand special attention as potential anticancer agents of con-
siderable promise. In addition to the several naturally occurring
substances, an impressive array of epothilone analogues have
been constructed and biologically evaluated.1,2 In a preliminary
communication,3 we reported the chemical synthesis and
biological data of 12,13-cyclopropyl and 12,13-cyclobutyl
epothilones A (3-6, Figure 1) where the epoxide moiety of
epothilone A (1, Figure 1) has been replaced by a small
cycloalkane ring. Encouraged by the biological actions of 3 and
4 we have extended these investigations to other members of
the 12,13-cycloalkane epothilone family (e.g. compounds 7-12,
Figure 1). In this article we report the details of these endeavors
including chemical synthesis and biological activities of all
analogues shown in Figure 1. Interestingly, these investigations
revealed that compounds 3 and 4 exhibit comparable potencies
to epothilone A (1) in cytotoxicity studies, supporting the notion
that the overall shape of the epothilone scaffold is a most signif-
(2) (a) Balog, A.; Meng, D.; Kamenecka, T.; Bertinato, P.; Su, D.-S.;
Sorensen, E. J.; Danishefsky, S. J. Angew. Chem., Int. Ed. Engl. 1996, 35,
2801-2803. (b) Su, D.-S.; Meng, D.; Bertinato, P.; Balog, A.; Sorensen,
E. J.; Danishefsky, S. J.; Zheng, Y.-H.; Chou, T.-C.; He, L.; Horwitz, S. B.
Angew. Chem., Int. Ed. Engl. 1997, 36, 757-759. (c) Meng, D.; Bertinato,
P.; Balog, A.; Su, D. S.; Kamenecka, T.; Sorensen, E. J.; Danishefsky, S.
J. J. Am. Chem. Soc. 1997, 119, 10073-10092. (d) Su, D.-S.; Balog, A.;
Meng, D.; Bertinato, P.; Danishefsky, S. J.; Zheng, Y.-H.; Chou, T.-C.;
He, L.; Horwitz, S. B. Angew. Chem., Int. Ed. Engl. 1997, 36, 2093-2096.
(e) Yang, Z.; He, Y.; Vourloumis, D.; Vallberg, H.; Nicolaou, K. C. Angew.
Chem., Int. Ed. Engl. 1997, 36, 166-168. (f) Nicolaou, K. C.; Sarabia, F.;
Ninkovic, S.; Yang, Z.; Angew. Chem., Int. Ed. Engl. 1997, 36, 525-527.
(g) Nicolaou, K. C.; Winssinger, N.; Pastor, J.; Ninkovic, S.; Sarabia, F.;
He, Y.; Vourloumis, D.; Yang, Z.; Li, T.; Giannakakou, P.; Hamel, E. Nature
1997, 387, 268-272. (h) Schinzer, D.; Limberg, A.; Bauer, A.; Bo¨hm, O.
M.; Cordes, M. Angew. Chem., Int. Ed. Engl. 1997, 36, 523-524. (i) May,
S.; Grieco, P. A. Chem. Commun. 1998, 1597-1598. (j) Chou, T.-C.; Zhang,
X.-G.; Balog, A.; Su, D.-S.; Meng, D.; Savin, K.; Bertino J. R.; Danishefsky,
S. J. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 9642-9647. (k) White, J. D.;
Sundermann, K. F.; Carter, R. G. Org. Lett. 1999, 1, 1431-1434. (l) Sinha,
S. C.; Sun, J.; Miller, G.; Barbas, C. F., III; Lerner, R. A. Org. Lett. 1999,
1, 1623-1626. (m) Sawada, D.; Shibasaki, M. Angew. Chem., Int. Ed. 2000,
39, 209-212. (n) Martin, H. J.; Drescher, M.; Mulzer, J. Angew. Chem.,
Int. Ed. 2000, 39, 581-583. (o) Zhu, B.; Panek, J. S. Org. Lett. 2000, 2,
2575-2578. (p) Bode, J. W.; Carreira, E. M. J. Am. Chem. Soc. 2001, 123,
3611-3612.
† The Scripps Research Institute and University of California, San Diego.
‡ Emory University.
§ Winship Cancer Institute, Emory University School of Medicine.
Novartis Pharma AG.
(1) For reviews from these laboratories, see: (a) Nicolaou, K. C.;
Roschangar, F.; Vourloumis, D. Angew. Chem., Int. Ed. Engl. 1998, 37,
2014-2045. (b) Nicolaou, K. C.; Ritze´n, A.; Namoto, K. Chem. Commun.
2001, 1523-1535.
(3) Nicolaou, K. C.; Namoto, K.; Li, J.; Ritze´n, A.; Ulven, T.; Shoji,
M.; Zaharevitz, D.; Gussio, R.; Sackett, D. L.; Ward, R. D.; Hensler, A.;
Fojo, T.; Giannakakou, P. ChemBioChem 2001, 1, 69-75.
10.1021/ja011338b CCC: $20.00 © 2001 American Chemical Society
Published on Web 08/30/2001

9314 J. Am. Chem. Soc., Vol. 123, No. 38, 2001
Nicolaou et al.
Scheme 1. Retrosynthetic Analysis and Key Fragments for
Epothilone Analogues 3-8
Scheme 1. Thus, coupling of the C7-C15 aldehyde fragment with the
heterocyclic vinyl iodide, followed by elaboration and aldol reaction
with the C1-C6 ketone segment, would lead, upon further elaboration,
to the desired seco-hydroxy acid. Cyclization according to our
Yamaguchi strategy6,7 would then furnish, upon deprotection, the desired
epothilone analogues.
The required building blocks 13-17 were synthesized as shown in
Schemes 2-5, while ketone 18 was prepared by the previously
described route.7 The first required C7-C15 aldehyde (13) was
constructed as shown in Scheme 2. Thus, Swern oxidation of optically
active alcohol 198 was followed by Wittig reaction and acid hydrolysis
to afford the homologated aldehyde 20 in 85% overall yield. A second
Wittig reaction employing commercially available phosphonium salt
21 led to a mixture of cis and trans olefins 22 (cis:trans ca. 20:1, 78%
yield), which was reduced with diimide9 to the saturated alcohol 23
(94% yield). Acetylation of the free hydroxyl group in 23 (100% yield)
yielded acetate 24, which upon hydrogenolysis of the benzyl ether
afforded alcohol 25 (78% yield). Direct hydrogenation of 22 with
palladium catalysts to simultaneously reduce the double bond and cleave
the benzyl ether proved impractical, due to significant amounts of
cyclopropyl ring-opened byproducts. Furthermore, although platinum
catalysts cleanly reduced the double bond in 22, they also reduced the
aromatic ring of the benzyl group, rather than effecting hydrogenolysis
of the C-O bond. Alcohol 25 was oxidized to the corresponding
aldehyde (89% yield) with TPAP-NMO (for abbreviations of reagents
and protecting groups, see legends in schemes), and then homologated
to the desired aldehyde 13 via enol ether 26 by the two-step procedure
described above for 20 (Wittig reaction followed by acidic hydrolysis),
in 50% overall yield.
Figure 1. Epothilones prepared in this study.
to excellent activity, while all the other analogues, where the
absolute configuration at C15 is R (5, 6, 10, and 12), are virtually
devoid of any cytotoxic activity. The most active analogue
emerging from these endeavors was the 12,13-cyclopropyl
5-methylpyridine epothilone A (11), whose impressive cyto-
toxicity rivals that of epothilone B (2), the most potent naturally
occurring epothilone.
Chemical Synthesis
Thiazole Epothilone Analogues. The chemical synthesis of the
designed 12,13-cycloalkane thiazole epothilone analogues 3-8 was
carried out according to a strategy derived from the retrosynthetic
analysis shown in Scheme 1. Thus, a Nozaki-Hiyama-Kishi coupling,5
an aldol reaction, and a Yamaguchi lactonization6,7 were employed to
disconnect the three strategic bonds as indicated, revealing building
blocks 13-16, 17, and 18. The assembly and elaboration of these
building blocks to the final targets was to follow the order shown in
Shown in Scheme 3 is the synthesis of the trans-cyclopropyl
aldehyde 14, which closely parallels that of its cis counterpart 13
described above. Thus, a Charette cyclopropanation8 of allylic alcohol
2710 yielded the enantiomerically enriched cyclopropane 29 in 98%
yield (ee >90% by Mosher ester analysis).11 Oxidation (SO3‚py)
followed by Wittig reaction afforded enol ether 30 (81% overall yield),
whose desilylation (TBAF), benzylation (NaH, BnBr), and acid
(4) Johnson, J.; Kim, S.-H.; Bifano, M.; DiMarco, J.; Fairchild, C.;
Gougoutas, J.; Lee, F.; Long, B.; Tokarski, J.; Vite, G. Org. Lett. 2000, 2,
1537-1540.
(5) (a) Takai, K.; Kitamura, T.; Kuroda, T.; Hiyama, T.; Nozaki, H.
Tetrahedron Lett. 1983, 24, 5281-5284. (b) Jin, H.; Uenishi, J.; Christ,
W. J.; Kishi, Y. J. Am. Chem. Soc. 1986, 108, 5644-5646.
(6) (a) Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M.
Bull. Chem. Soc. Jpn. 1979, 52, 1989-1993. (b) Mulzer, J.; Mareski, P.
A.; Bushmann, J.; Luger, P. Synthesis 1992, 215-228. (c) Nicolaou, K.
C.; Hepworth, D.; King, N. P.; Finlay, M. R. V.; Scarpelli, R.; Pereira, M.
M. A.; Bollbuck, B.; Bigot, A.; Werschkun, B.; Winssinger, N. Chem. Eur.
J. 2000, 6, 2783-2800.
(8) Charette, A. B.; Juteau, H.; Lebel, H.; Molinaro, C. J. Am. Chem.
Soc. 1998, 120, 11943-11952.
(9) Pasto, D. J.; Taylor, R. T. Org. React. 1991, 40, 92-155.
(10) (a) Azzena, F.; Calvani, F.; Crotti, P.; Gardelli, C.; Macchia, F.;
Pinechi, M. Tetrahedron 1995, 51, 10601-10626. (b) Trost, B. M.;
Verhoeven, T. R. J. Am. Chem. Soc. 1980, 102, 4743-4763.
(11) (a) Dale, J. A.; Dull, D. L.; Mosher, H. S. J. Org. Chem. 1969, 34,
2543-2549. (b) Dale, J. A.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95,
512-519.
(7) Nicolaou, K. C.; Ninkovic, S.; Sarabia, F.; Vourloumis, D.; He, Y.;
Vallberg, H.; Finlay, M. R. V.; Yang, Z. J. Am. Chem. Soc. 1997, 119,
7974-7991.

Synthesis of Cyclopropyl and Cyclobutyl Epothilone Analogues
J. Am. Chem. Soc., Vol. 123, No. 38, 2001 9315
Scheme 2a
Scheme 3a
a Reagents and conditions: (a) (COCl)2 (1.5 equiv) DMSO (2.0
equiv), Et3N (5.0 equiv), CH2Cl2, -78 °C; (b) MeOCH2PPh3Cl (1.5
equiv), NaHMDS (1.4 equiv), THF, -78 °C; (c) catalytic HCl, acetone:
water 9:1, 65 °C, 1 h, 85% over 3 steps; (d) 21 (1.5 equiv), n-BuLi
(3.0 equiv), THF, -78 °C, 78%; (e) (NCO2K)2 (20 equiv), HOAc (40
equiv), MeOH, py, 25 °C, 48 h, 94%; (f) Ac2O (1.1 equiv), Et3N (1.2
equiv), 4-DMAP (0.1 equiv), CH2Cl2, 25 °C, 0.5 h, 88%; (g) 20%
Pd(OH)2/C, H2 (1 atm), EtOAc:EtOH 1:1 25 °C, 2 h, 76%; (h) TPAP
(0.05 equiv), NMO (1.5 equiv), MS 4 Å, CH2Cl2, 25 °C, 1 h, 89%; (i)
MeOCH2PPh3Cl (1.2 equiv), NaHMDS (1.1 equiv), THF, 0 °C, 71%;
(j) catalytic HCl, acetone:water 9:1, 55-60 °C, 2 h, 87%. 4-DMAP )
4-(dimethylamino)pyridine, NaHMDS ) sodium hexamethyldisilazide,
NMO ) N-methylmorpholine N-oxide, py ) pyridine, TPAP ) tetra-
n-propylammonium perruthenate.
a Reagents and conditions: (a) DME (2.2 equiv), Et2Zn (2.2 equiv),
CH2I2 (4.4 equiv), 28 (1.2 equiv), CH2Cl2, 98% yield, >90% ee; (b)
Et3N (6.0 equiv), SO3‚py (3.0 equiv), CH2Cl2:DMSO 4:1, 0 °C, 2 h;
(c) MeOCH2PPh3Cl (1.5 equiv), NaHMDS (1.3 equiv), THF, -40f25
°C, 12 h, 81% over 2 steps; (d) TBAF (1.5 equiv), THF 25 °C, 2 h; (e)
NaH (1.5 equiv), BnBr (2.0 equiv), THF:DMF 5:1, 0f25 °C, 10 h; (f)
catalytic HCl, acetone:water 9:1, 50 °C, 5 h; (g) 21 (1.5 equiv),
NaHMDS (2.8 equiv), TMSCl (1.5 equiv), THF, 58% over 4 steps;
(h) (NCO2K)2 (20 equiv), HOAc (40 equiv), MeOH, py, 25 °C, 7 h;
(i) Ac2O (2.0 equiv), Et3N (5.0 equiv), 4-DMAP (0.1 equiv), CH2Cl2,
0 °C, 20 min; (j) 20% Pd(OH)2/C, H2 (1 atm), EtOAc:EtOH 1:1 25
°C, 6 h, 98% over 3 steps; (k) DMP (1.2 equiv), CH2Cl2, 0f25 °C, 40
min. 4-DMAP ) 4-(dimethylamino)pyridine, DME ) dimethoxyethane,
DMP ) Dess-Martin periodinane, NaHMDS ) sodium hexamethyl-
disilazide, py ) pyridine, TBAF ) tetrabutylammonium fluoride,
TMSCl ) chlorotrimethylsilane.
hydrolysis led to the homologated aldehyde, which reacted with the
ylide derived from phosphonium salt 21 to afford olefin 31 in 58%
yield for the four steps. Diimide reduction, acetylation, and hydro-
genolysis furnished alcohol 32 (98% overall yield). Dess-Martin
oxidation then yielded the desired aldehyde 14, which was not isolated,
but rather used immediately for the subsequent Nozaki-Hiyama-Kishi
coupling (vide infra).
The requisite vinyl iodide 17 was constructed from aldehyde 447
via a sequence involving (i) a modified Corey-Fuchs protocol13 with
in situ methylation of the intermediate acetylenide via intermediates
45 (88%) and 46 (97%), (ii) stereoselective hydrostannylation14 (84%),
and (iii) iodine-tin exchange (99%), as shown in Scheme 5. This
sequence represents a significant improvement, regarding both simplic-
ity and yields, over the preliminary route previously disclosed.3
The syntheses of the C12-C13-cyclobutyl aldehydes 15 and 16 were
carried out as shown in Scheme 4. As these compounds are very closely
related to the cyclopropane derivatives 13 and 14, a similar synthetic
route was again followed. Thus, starting from the monoacetate 33,
readily available through enzymatic group-selective saponification of
the corresponding diacetate,12 cis-aldehyde 34 was prepared by Dess-
Martin periodinane oxidation (95% yield), while the corresponding
trans-aldehyde 39 was conveniently available by base-catalyzed
epimerization of 34 (88% from 33). Following the route described for
the cyclopropyl derivatives, 34 and 39 were homologated to 35 and
40, respectively, and the latter compounds were coupled with the chiral
phosphorane derived from enantiomerically pure phosphonium salt 21
and NaHMDS-TMSCl to yield olefins 36 and 41, respectively.
Hydrogenation of the double bond using a platinum catalyst [olefin 36
was initially reduced with diimide because it was done in parallel with
22, where catalytic hydrogenation was not feasible, see discussion of
22 above; because the reduction was incomplete (see Supporting
Information), further catalytic hydrogenation with Pt was necessary; it
was later found that Pt hydrogenation alone worked for compound 41]
followed by standard protecting group manipulations afforded alcohols
38 and 43, which were again homologated and protected, as summarized
in Scheme 4, thus producing aldehydes 15 and 16, respectively.
With all the building blocks in hand, final assembly of epothilone
analogues 3-8 could begin. The cyclopropyl analogues 3, 5, and 7
were synthesized as shown in Scheme 6. Aldehyde 13 was coupled
with vinyl iodide 17 by the Nozaki-Hiyama-Kishi procedure employ-
ing CrCl2-NiCl2,5 furnishing a diastereomeric mixture of alcohols 48
(ca. 1:1 ratio, 56% yield, unoptimized). This mixture was taken through
the sequence until chromatographic separation of the two isomers
became feasible upon Yamaguchi macrolactonization (vide infra).6
Silylation of 48 (TBSOTf-2,6-lutidine, 100% yield) furnished silyl ether
49, which was deacetylated (DIBAL, 99% yield) to yield the advanced
intermediate alcohol 50. In a similar way, trans-aldehyde 14 was
coupled with iodide 17, but this time, oxidation (DMP) of the resulting
mixture of epimeric secondary alcohols led to ketone 56 in 75% overall
yield. Stereoselective reduction of 56 with (-)-DIPCl15 afforded alcohol
1
57 as a single stereoisomer (by H NMR spectroscopy) in 84% yield,
thus demonstrating the flexibility of this route to generate either one
(13) Grandjean, D.; Pale, P.; Chuche, J. Tetrahedron Lett. 1994, 35,
3529-3530.
(14) Betzer, J.-F.; Ardisson, J.; Lallemand, J.-Y.; Pancrazi, A. Tetrahe-
dron Lett. 1997, 38, 2279-2282.
(12) (a) Laumen, K.; Schneider, M. Tetrahedron Lett. 1985, 26, 2073-
2076. (b) Kasel, W.; Hultin, P. G.; Jones, J. B. J. Chem. Soc., Chem.
Commun. 1985, 1563-1564.
(15) (a) Brown, H. C.; Ramachandran, P. V. Acc. Chem. Res. 1992, 25,
16-24. (b) Brown, H. C.; Park, W. S.; Cho, B. T.; Ramachandran, P. V.
J. Org. Chem. 1987, 52, 5406-5412.

9316 J. Am. Chem. Soc., Vol. 123, No. 38, 2001
Nicolaou et al.
Scheme 4a
Scheme 5a
a Reagents and conditions: (a) PPh3 (4.0 equiv), CBr4 (2.0 equiv),
CH2Cl2, 0 °C, 4 h, 88%; (b) NaHMDS (1.0 equiv), MeLi (2.0 equiv),
MeI (5.0 equiv), -78f25 °C, 12 h, 97%; (c) n-BuLi (4.0 equiv), (n-
Bu3Sn)2 (4.0 equiv), CuCN (2.0 equiv), MeOH (110 equiv), THF, 87%;
(d) I2 (1.1 equiv), CH2Cl2, 0 °C, 99%. NaHMDS ) sodium hexam-
ethyldisilazide.
to our optimized protocol.6c In this manner, aldols 51 (63%) and 60
(70%) were generated and isolated with complete control of the C6-
C7 stereochemistry (as determined by 1H NMR spectroscopy). Protec-
tion of the secondary hydroxyl groups as the TBS ethers 52 and 61,
followed by a two-step oxidation of the C1 position (liberated selectively
by the action of HF‚py) and selective cleavage of the C15 TBS ether
(TBAF), afforded the hydroxy acids 53 and 62, respectively. Yamaguchi
macrolactonization6 of 53 gave a 69% combined yield of the protected
epothilone derivatives 54 and 55, which were chromatographically
separated [54 (42%); 55 (27%)]. Analogously, macrolactonization of
62 yielded bis-silyl ether 63 (53% from 61 after 5 steps). Desilylation
of 54, 55, and 63 with 20% TFA in CH2Cl2 finally afforded the desired
epothilone analogues 3, 5, and 7, respectively. The 15S configuration
of the trans analogue 7 was now further corroborated by comparison
1
of the H NMR spectrum of 7 with those of the cis isomers 3 and 5,
where the spectrum of 7 is more similar to that of 3 than to that of 5,
particularly considering the signals from the protons attached to C2
and C15 (see Supporting Information).
The cis-cyclobutyl thiazole epothilones 4 and 6 were assembled in
an analogous fashion, as summarized in Scheme 7. A Nozaki-
Hiyama-Kishi coupling between cis-aldehyde 15 and the side chain
vinyl iodide 17 afforded the secondary alcohol 64 (89% yield) as a 1:1
mixture of C15 epimers. Protective group manipulations and oxidation
yielded, via intermediates 65 and 66, aldehyde 67, which smoothly
underwent the stereoselective aldol coupling reaction with ketone 18,
thus producing alcohol 68. Further manipulation of protecting groups
and oxidation of the C1 position yielded hydroxy acid 72, which was
cyclized by applying the Yamaguchi protocol to afford the two lactones
73 and 74. At this point, the C15 epimers 73 and 74 were chromato-
graphically separated and deprotected to yield the desired cis-cyclobutyl
epothilones 4 and 6, respectively, and in good overall yields.
The trans-cyclobutyl thiazole epothilone 8 was prepared by a similar
sequence, as detailed in Scheme 8. Thus, after the Nozaki-Hiyama-
Kishi coupling between aldehyde 16 and iodide 17, the resulting alcohol
was oxidized to the corresponding enone 75, which was then stereo-
selectively reduced with (-)-DIPCl15 to afford only the (15S)-epimer
76. The remaining steps followed the same sequence described for the
cis compounds (see Scheme 7), and proceeded smoothly and in similar
yields, affording the targeted trans-cyclobutyl epothilone 8.
Pyridine Epothilone Analogues. Some of the most active epothilone
analogues prepared to date include within their structures a pyridine
side chain as a replacement for the thiazole moiety of the naturally
occurring substances.16 Given the very promising preliminary results
with cyclopropane epothilone analogue 3, we reasoned that combining
these two structural modifications might result in highly active
compounds despite the absence of the epoxide oxygen. Such compounds
(e.g. 9-12, Figure 1) may be metabolically more stable leading to
longer in vivo lifetime and lower toxicity. In an effort to improve the
a Reagents and conditions: (a) DMP (1.2 equiv), NaHCO3 (5.0
equiv), CH2Cl2, 25 °C, 3 h, 95%; (b) starting with alcohol 33: (COCl)2
(1.1 equiv), DMSO (2.2 equiv), Et3N (5.0 equiv), CH2Cl2, -78 °C;
then Et3N, 25 °C, 5 days, 88% over 2 steps; (c) MeOCH2PPh3Cl (1.15
equiv), NaHMDS (1.10 equiv), THF, -78f25 °C, 89%; (d) 0.12 N
HCl (aq):acetone (1:9), reflux, 1 h, 98% (35), 94% (40); (e) 21 (2.0
equiv), NaHMDS (3.8 equiv), THF, 0 °C, 2 h; then TMSCl (2.0 equiv),
25 °C, 20 min; then 35 (or 40), THF, -78f25 °C, 20 h, 59% (36),
83% (41); (f) (NCO2K)2 (20 equiv), AcOH (40 equiv), py:MeOH (5:
1), 25 °C, 48 h; then PtO2 (0.05 equiv), H2 (1 atm), MeOH, 25 °C, 20
min, 82%; (g) 10 wt % Pt on carbon (0.02 equiv), H2 (1 atm), EtOAc,
25 °C, 8 h, 96%; (h) TBSOTf (1.0 equiv), 2,6-lutidine (2.5 equiv),
CH2Cl2, -78f0 °C, 20 min; (i) DIBAL (2.0 equiv), CH2Cl2, -78 °C,
5 min, 99% (38), 90% (43) for 2 steps; (j) DMP (1.2 equiv), NaHCO3
(5.1 equiv), CH2Cl2, 25 °C, 3 h, 94% (38); (k) (COCl)2 (1.1 equiv),
DMSO (2.2 equiv), Et3N (5.0 equiv), CH2Cl2, -78f25 °C, 97% (43);
(l) MeOCH2PPh3Cl (1.15 equiv), NaHMDS (1.10 equiv), THF,
-78f25 °C; (m) 0.12 N HCl (aq):acetone (1:9), reflux, 1 h; (n) Ac2O
(1.1 equiv), Et3N (2.5 equiv), 4-DMAP (0.02 equiv), CH2Cl2, 0 °C, 20
min, 60% (15), 62% (16) for 3 steps. DIBAL ) diisobutylaluminum
hydride, 4-DMAP ) 4-(dimethylamino)pyridine, DMP ) Dess-Martin
periodinane, NaHMDS ) sodium hexamethyldisilazide, py ) pyridine,
TMSCl ) chlorotrimethylsilane.
or both C15 epimers. The 15S stereochemistry was assumed based on
the chirality of the reducing agent. Compound 57 was protected as a
TBS ether (58, TBSOTf, 2,6-lutidine, 91% yield), which was then
deacetylated with DIBAL to provide the desired alcohol 59 in 93%
yield. The stage was now set for the stereoselective aldol coupling which
was employed to simultaneously create the C6-C7 bond and set the
stereochemistry at these stereocenters. To this end, DMP oxidation of
50 and 59, respectively, was immediately followed by aldol addition
of the previously described C1-C6 ketone 187 using LDA according
(16) Nicolaou, K. C.; Scarpelli, R.; Bollbuck, B.; Werschkun, B.; Pereira,
M. M. A.; Wartmann, M.; Altmann, K.-H.; Zaharevitz, D.; Gussio, R.;
Giannakakou, P. Chem. Biol. 2000, 7, 593-599.

Synthesis of Cyclopropyl and Cyclobutyl Epothilone Analogues
J. Am. Chem. Soc., Vol. 123, No. 38, 2001 9317
Footnote to Scheme 6 (continued)
Scheme 6a
THF, 25 °C, 16-26 h, 46% over 4 steps (53); (m) 2,4,6-trichlorobenzoyl
chloride (2.4 equiv), Et3N (6.0 equiv), THF, 0 °C, 1 h, then 4-DMAP
(2.2 equiv), toluene, 75 °C, 3-11 h, 42% (54), 27% (55), 53% over 5
steps (63); (n) 20% TFA in CH2Cl2, 0 °C, 2 h, 78% (3), 65% (5); (o)
25% TFA in CH2Cl2, 25 °C, 7 h, 73%. DIBAL ) diisobutylaluminum
hydride, DIPCl ) diisopinocampheyl chloroborane, 4-DMAP )
4-(dimethylamino)pyridine, DMP ) Dess-Martin periodinane, LDA
) lithium diisopropylamide, NaHMDS ) sodium hexamethyldisilazide,
py ) pyridine, TBAF ) tetrabutylammonium fluoride.
Scheme 7a
a Reagents and conditions: (a) 17 (1.5 equiv), CrCl2 (12.6 equiv),
NiCl2 (0.13 equiv), DMSO, 25 °C, 6 h, (89%, 2:3 mixture of C15
epimers); (b) TBSOTf (1.0 equiv), 2,6-lutidine (2.5 equiv), CH2Cl2,
-78f0 °C, 20 min; (c) DIBAL (2.0 equiv), CH2Cl2, -78 °C, 5 min,
99% for 2 steps; (d) DMP (1.2 equiv), CH2Cl2, 25 °C, 1.5 h; (e) LDA
(3.1 equiv), 18 (3.0 equiv), THF, -78 °C, 4 min, 67% for 2 steps; (f)
TBSOTf (2.0 equiv), 2,6-lutidine (2.5 equiv), CH2Cl2, -20f25 °C,
12 h, 96%; (g) HF‚py, py, THF, 0f25 °C, 3 h, 91% (h) DMP (1.2
equiv), NaHCO3 (1.5 equiv), CH2Cl2, 25 °C, 2 h; (i) NaClO2 (5.0 equiv),
2-methyl-2-butene (7.5 equiv), NaH2PO4 (2.5 equiv), t-BuOH:H2O
4:1, 25 °C, 10 min, 93% for 2 steps; (j) TBAF (12 equiv), THF, 25
°C, 26 h, 54%; (k) 2,4,6-trichlorobenzoyl chloride (2.4 equiv), Et3N
(6.0 equiv), THF, 0 °C, 1 h, then 4-DMAP (2.2 equiv), toluene, 75 °C,
3 h, 21% (73), 38% (74); (l) 20 v/v% TFA in CH2Cl2, 0 °C, 2 h, 61%
(4), 60% (6). DIBAL ) diisobutylaluminum hydride, 4-DMAP )
4-(dimethylamino)pyridine, DMP ) Dess-Martin periodinane, LDA
) lithium diisopropylamide, py-pyridine, TBAF ) tetrabutylammonium
fluoride.
a Reagents and conditions: (a) 17 (1.5-2.0 equiv), CrCl2 (10-13
equiv), NiCl2 (0.02-0.13 equiv), DMSO, 25 °C, 6-12 h, 56% (48),
91% from 32; (b) DMP (1.2 equiv), CH2Cl2, 0f25 °C, 0.5 h, 83%; (c)
(-)-DIPCl (3.0 equiv), Et2O, -15f25 °C, 18 h, 84%; (d) TBSOTf
(1.1-2.0 equiv), 2,6-lutidine (2.5 equiv), CH2Cl2, -78 °C, 0.5-1 h,
91-100%; (e) DIBAL (2.0-3.1 equiv), CH2Cl2, -78 °C, 15 min to 1
h, 93-96%; (f) DMP (1.2 equiv), CH2Cl2, 25 °C, 1.5 h; (g) LDA (3.1
equiv), 18 (3.0 equiv), THF, -78 °C, 4 min, 63% (51), 70% (60); (h)
TBSOTf (2.0 equiv), 2,6-lutidine (2.5 equiv), CH2Cl2, -20f25 °C,
1.5-12 h, 86% (52), 94% (61); (i) HF‚py, py, 0-25 °C, 3-4 h; (j)
DMP (1.2-1.5 equiv), NaHCO3 (1.5 equiv), CH2Cl2, 25 °C, 15 min to
2 h; (k) NaClO2 (5.0 equiv), 2-methyl-2-butene (7.5 equiv), NaH2PO4
(2.5 equiv), t-BuOH:H2O 4:1, 25 °C, 10-20 min; (l) TBAF (12 equiv),
overall synthesis of these compounds, and to accommodate future
preparation of other side chain-modified analogues via a convergent

9318 J. Am. Chem. Soc., Vol. 123, No. 38, 2001
Nicolaou et al.
Scheme 8a
Scheme 9. Retrosynthetic Analysis and Key Fragments for
Epothilone Analogs 9-12
protection of the resulting alcohol 88 as its TBDPS ether (TBDPSCl-
imid.), afforded alkene 89 in 89% yield. Hydrogenation of the double
bond in 89 with concomitant cleavage of the benzyl ether gave primary
alcohol 90 in 75% yield. This compound (90) was then converted into
the corresponding iodide (91) in 93% yield by exposure to I2/PPh3.
Coupling of 91 with alkyne 92 (n-BuLi, 72% yield), followed by
removal of the TBS group (BF3‚OEt2) from the resulting alkyne 93,
produced the propargylic alcohol 94 (89% yield). This compound was
used as a common precursor to prepare both the cis- and the trans-
cyclopropyl pyridine epothilone analogues (9-12). The synthesis of
the cis series of compounds commenced with a nickel boride reduction17
of alkyne 94 to furnish cis olefin 95 in 95% yield (Scheme 10), while
the corresponding trans alkene (97) was prepared from the same
intermediate (94) via reduction with LiAlH3(OMe)18 (83% yield).
Charette cyclopropanation8 of 95 and 97 smoothly afforded the
cyclopropanes 96 (99% yield) and 98 (93%) in >95% de, as judged
by 1H NMR spectroscopic analysis of the corresponding Mosher
esters.11 Subsequent benzylation of the primary hydroxyl group,
followed by removal of the silyl group at the other end of the molecule,
led to the desired primary alcohols 84 and 85, respectively.
a Reagents and conditions: (a) 17 (1.5 equiv), CrCl2 (12.6 equiv),
NiCl2 (0.13 equiv), DMSO, 25 °C, 6 h, 91%; (b) DMP (1.2 equiv),
NaHCO3 (5.0 equiv), CH2Cl2, 25 °C, 3 h; (c) (-)-DIPCl (3.0 equiv),
Et2O, -15f25 °C, 18 h, 47%, for 2 steps; (d) TBSOTf (1.0 equiv),
2,6-lutidine (2.5 equiv), CH2Cl2, -78f0 °C, 20 min; (e) DIBAL (2.0
equiv), CH2Cl2, -78 °C, 5 min, 84% for 2 steps; (f) DMP (1.2 equiv),
CH2Cl2, 25 °C, 1.5 h; (g) LDA (3.1 equiv), 18 (3.0 equiv), THF, -78
°C, 4 min, 75% for 2 steps; (h) TBSOTf (2.0 equiv), 2,6-lutidine (2.5
equiv), CH2Cl2, -20f25 °C, 12 h, 96%; (i) HF‚py, py THF, 0f25
°C, 3 h, 81%; (j) DMP (1.2 equiv), NaHCO3 (1.5 equiv), CH2Cl2,
25 °C, 2 h; (k) NaClO2 (5.0 equiv), 2-methyl-2-butene (7.5 equiv),
NaH2PO4 (2.5 equiv), t-BuOH:H2O 4:1, 25 °C, 10 min; (l) TBAF (12
equiv), THF, 25 °C, 26 h, 47% for 3 steps; (m) 2,4,6-trichlorobenzoyl
chloride (2.4 equiv), Et3N (6.0 equiv), THF, 0 °C, 1 h; then 4-DMAP
(2.2 equiv), toluene, 75 °C, 3 h, 50%; (n) 20% TFA in CH2Cl2, 0 °C,
2 h, 79%. DIBAL ) diisobutylaluminum hydride, DIPCl ) diisopi-
nocampheylchloroborane, 4-DMAP ) 4-(dimethylamino)pyridine, DMP
) Dess-Martin periodinane, LDA ) lithium diisopropylamide, py )
pyridine, TBAF ) tetrabutylammonium fluoride.
The requisite side chain vinyl iodide 86 was synthesized as shown
in Scheme 11. A Sonogashira coupling of 5-methyl-2-bromopyridine
99 with propyne19 yielded alkyne 100 in 98% yield. This was then
hydrostannylated, and the tin was exchanged for iodine (86% for two
steps) by the same method as that employed to prepare the thiazole
side chain precursor 17 (Scheme 5), thus yielding iodide 86 via stannane
101 (100% yield).
The final stages of the synthesis of the targeted pyridine analogues
are depicted in Schemes 12 and 13. Oxidation of alcohols 84 and 85
with Dess-Martin periodinane was followed by the stereoselective aldol
coupling with ketone 187 previously employed (vide supra). This
coupling was performed according to our general procedure,6c yielding
aldols 102 (75% yield) and 108 (89% yield) with a dr of ca. 10:1 (by
1H NMR spectroscopy) in both cases. Further elaboration of these
compounds (102 and 108) involved TBS protection of their secondary
alcohols, selective removal of the primary TBS group (HF‚py),
oxidation of the resulting primary alcohol (DMP; NaClO2), and
methylation of the so obtained carboxylic acid, leading to compounds
strategy, a slightly different scheme for their total synthesis was
designed based on the retrosynthetic analysis shown in Scheme 9. The
devised strategy for the construction of the pyridine cycloalkane
epothilones (9-12) is similar to that utilized for the total synthesis of
their thiazole counterparts except for the reversal of the coupling order
of the fragments. Thus, the aldol reaction of building blocks 84 and 85
with ketone 18 will now precede the Nozaki-Hiyama-Kishi coupling
with vinyl iodide 86.
The required building blocks 84 and 85 were prepared as shown in
Scheme 10. A Wittig reaction between the ylide derived from the
enantiomerically pure phosphonium salt 21 and NaHMDS-TMSCl, and
the commercially available aldehyde 87 (68% yield), followed by
(17) Taber, D. F.; Herr, R. J.; Gleave, D. M. J. Org. Chem. 1997, 62,
194-198.
(18) Ashby, E. C.; Dobbs, F. R.; Hopkins, H. P., Jr. J. Am. Chem. Soc.
1975, 97, 3158-3162.
(19) Arcadi, A.; Cacchi, S.; Carnicelli, V.; Marinelli, F. Tetrahedron
1994, 50, 437-452.

Synthesis of Cyclopropyl and Cyclobutyl Epothilone Analogues
J. Am. Chem. Soc., Vol. 123, No. 38, 2001 9319
Scheme 12a
Scheme 10a
a Reagents and conditions: (a) NaHMDS (2.1 equiv), TMSCl (1.1
equiv), THF, -78f25 °C, 6 h, 68%; (b) TBDPSCl (1.1 equiv),
imidazole (2.0 equiv), DMF, 25 °C, 1 h, 89%; (c) 10% Pd/C, H2 (1
atm), MeOH:THF 5:1, 50 °C, 10 h, 75%; (d) PPh3 (1.4 equiv), 4-DMAP
(0.01 equiv), I2 (1.5 equiv), imidazole (2.0 equiv), MeCN/Et2O, 5:3,
25 °C, 1 h, 93%; (e) n-BuLi (3.3 equiv), 3-(tert-butyldimethylsilylox-
y)propyne (92), (3.5 equiv), THF/HMPA, 6:1, -78 to -30 °C, 2.5 h,
72%; (f) BF3‚OEt2 (2.0 equiv), CH2Cl2, 25 °C, 1.5 h, 89%; (g) NiCl2
(1.0 equiv), NaBH4 (1.0 equiv), EDA (3.0 equiv), H2 (1 atm), EtOH,
0 °C, 1 h, 95%; (h) LiAlH4 (1.0 equiv), MeOH (1.0 equiv), THF, 50
°C, 0.5 h, 83%; (i) DME (2.2 equiv), Et2Zn (2.2 equiv), CH2I2 (4.4
equiv), ent-28 (1.2 equiv), CH2Cl2, -15f25 °C, 6 h, 99% (96), 93%
(98); (j) Ag2O (3.0 equiv), BnBr (2.6 equiv), TBAI (0.1 equiv), toluene,
24 h, 25f50 °C; (k) TBAF (5.0 equiv), THF, 25 °C, 4 h, 83% (84),
85% (85) over 2 steps. 4-DMAP ) 4-(dimethylamino)pyridine, DME
) dimethoxyethane, DMP ) Dess-Martin periodinane, EDA )
ethylenediamine, HMPA ) hexamethylphosphoramide, NaHMDS )
sodium hexamethyldisilazide, TBAF ) tetrabutylammonium fluoride,
TBAI ) tetrabutylammonium iodide.
a Reagents and conditions: (a) DMP (1.2 equiv), CH2Cl2, 25 °C,
1.5 h; (b) LDA (2.5 equiv), 18 (2.4 equiv), THF, -78 °C, 4 min, 75%
(102), 89% (108) over 2 steps, (c) TBSOTf (4.0 equiv), 2,6-lutidine
(5.0 equiv), CH2Cl2, -20f25 °C, 1 h, 93% (103), 100% (109); (d)
HF‚py, py, 25 °C, 2 h; (e) DMP (1.2 equiv), NaHCO3 (1.5 equiv),
CH2Cl2, 25 °C, 6 h; (f) NaClO2 (5.0 equiv), 2-methyl-2-butene (7.5
equiv), NaH2PO4 (2.5 equiv), t-BuOH:H2O 4:1, 25 °C, 10 min; (g)
TMSCHN2 (2.0 equiv), MeOH:benzene 1:1, 92% (104), 88% (110)
over 4 steps; (h) 20% Pd(OH)2/C, H2 (1 atm), EtOAc:EtOH 1:1 25 °C,
6 h, 93% (105), 80% (111); (i) DMP (1.2 equiv), NaHCO3 (1.5 equiv),
CH2Cl2, 25 °C, 1.5 h; (j) MeOCH2PPh3Cl (1.5 equiv), NaHMDS (1.3
equiv), THF, -40f25 °C, 70% (106), 79% (112); (k) TsOH (20 equiv),
dioxane:H2O 10:1, 50 °C, 5 h; then silylation as in (c), 61% (107),
76% (113). DMP ) Dess-Martin periodinane, LDA ) lithium
diisopropylamide, NaHMDS ) sodium hexamethyldisilazide, py )
pyridine.
Scheme 11a
a Reagents and conditions: (a) Pd(PPh3)2Cl2 (0.01 equiv), CuI (0.02
equiv), propyne (1 atm), DMF/i-Pr2NH, 6:5, 25 °C, 3 h, 98%; (b)
n-BuLi (4.0 equiv), (n-Bu3Sn)2 (4.0 equiv), CuCN (2.0 equiv), MeOH
(110 equiv), THF, -10 °C, 15 h, 86%; (c) I2 (1.05 equiv), CH2Cl2, 25
°C, 5 min, 100%.
and homologation to install the C15 carbon atom, thus yielding
aldehydes 107 and 113 via enol ethers 106 and 112, respectively.
The cis-aldehyde 107 was then subjected to the Nozaki-Hiyama-
Kishi coupling with vinyl iodide 86 to yield methyl ester 114 (43%,
unoptimized), which was hydrolyzed to the corresponding acid (115)
in 76% yield (Scheme 13). The ester hydrolysis (114f115) was
extremely slow, requiring 4 days for completion. When the same
104 and 110, as shown in Scheme 12. Hydrogenolysis of the benzyl
ether from 104 and 110 was followed by oxidation of the resulting
primary alcohols (105 and 111) to the corresponding aldehydes (DMP)

9320 J. Am. Chem. Soc., Vol. 123, No. 38, 2001
Nicolaou et al.
Scheme 13a
a Reagents and conditions: (a) NaBH4 (1.1 equiv), CH2Cl2/EtOH,
7:3, -78 °C, 1 h, 72%; (b) LiOH, H2O/t-BuOH, 1:1, 40 °C, 48 h,
98%; (c) EDC (2.0 equiv), 4-DMAP (0.5 equiv), TMSE-OH:CH2Cl2
2:1, 25 °C, 2 h, 83%; (d) DMP (2.5 equiv), py (10 equiv), CH2Cl2, 0
°C, 2.5 h, 93%; (e) 86 (2.0 equiv), CrCl2 (10 equiv), NiCl2 (0.02 equiv),
DMSO, 25 °C, 12-36 h, 43% (114), 71% (121); (f) LiOH, H2O:t-
BuOH 2:3, 25 °C, 4 days, 76%; (g) TBAF (18 equiv), THF, 0 °C, 2 h;
(h) 2,4,6-trichlorobenzoyl chloride (9.0 equiv), Et3N (22 equiv), THF,
0 °C, 1 h, then 4-DMAP (3.0 equiv), toluene, 75 °C, 3 h, 38% (116),
32% (117), 31% (123), 39% (124); (i) 20% TFA in CH2Cl2, 25 °C,
2-22 h, 60% (9), 59% (10), 89% (11), 74% (12). 4-DMAP )
4-(dimethylamino)pyridine, DMP ) Dess-Martin periodinane, EDC
) 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride, py
) pyridine, TBAF ) tetrabutylammonium fluoride, TMSE ) 2-(tri-
methylsilyl)ethyl.
Figure 3. In vitro tubulin polymerization. Comparison of the abilities
of epothilone A (1), cis-(15S)-CP-py-epo (9), trans-(15S)-CP-py-epo
(11), and trans-(15R)-CP-py-epo (12) to induce tubulin polymerization
in the absence of microtubule-associated proteins (MAPs). Polymeri-
zation reactions with epothilone B (2) and paclitaxel (Taxol) are
included for comparison. In each assay 10 µM (1 mg/mL) purified rat
brain tubulin in G-PEM buffer (80 mM PIPES, 1 mM GTP, 1 mM
EGTA, 1 mM MgCl2, pH 6.9) was mixed with 3 µM (curve 1) or 10
µM (curve 2) compound and polymerization at 25 °C was followed
for 30 min. The optical density at 350 nm (absorbance 350) was then
recorded at 15 s intervals for 30 min. Curves 0 are included as controls,
showing tubulin polymerization reactions under similar conditions in
the absence of the respective compound.
Figure 2. Cyclopropyl epothilones prepared in a previous study.23
hydroxy acid and protected (TMSE-OH, EDC, 4-DMAP), affording
the TMSE ester 119 in 81% yield. Direct hydrolysis of aldehyde 113
was unsuccessful, which dictated the adoption of the above plan
requiring reduction to the alcohol prior to hydrolysis. Reoxidation of
119 with Dess-Martin periodinane gave aldehyde 120 (93% yield),
which smoothly underwent the Nozaki-Hiyama-Kishi coupling with
86 to furnish hydroxy ester 121 in 71% yield. Cleavage of the TMSE
ester with TBAF now proceeded smoothly, affording hydroxy acid 122
sequence was applied to the trans compound 113, it proved impossible
to hydrolyze the corresponding methyl ester at a practical rate after
the Nozaki-Hiyama-Kishi coupling. Clearly, another protecting group
was needed for the C1 carboxylic acid, and we opted to try a
trimethylsilylethyl (TMSE) ester instead of the methyl ester. In the
event, the aldehyde 113 was reduced to the hydroxy ester 118 (NaBH4,
72% yield), which could now be hydrolyzed to the corresponding

Synthesis of Cyclopropyl and Cyclobutyl Epothilone Analogues
J. Am. Chem. Soc., Vol. 123, No. 38, 2001 9321
Table 1. Cytotoxicity of Epothilones 1 through 12 and Paclitaxel against 1A9 Human Ovarian Carcinoma Cells and â-Tubulin Mutant Cell
Lines Selected with Paclitaxel or Epothilone Aa
cell line
1A9
IC50
A8 (â274)
IC50
PTX10 (â270)
PTX22 (â364)
compound
RR
IC50
RR
IC50
RR
epothilone A (Epo A) 1
epothilone B (Epo B) 2
paclitaxel (Taxol)
cis-(15S)-CP-epo 3
cis-(15S)-CB-epo 4
2.37 ( 0.433
0.095 ( 0.007
1.77 ( 0.227
1.60 ( 0.124
8.8 ( 0.00
117 ( 27.01
2.14 ( 0.072
17.95 ( 3.08
23.43 ( 4.29
196 ( 0.00
>300 (inactive)
>300 (inactive)
48 ( 0.00
>300 (inactive)
53.5 ( 14.57
>300 (inactive)
9.5
49.3
22.5
10.14
14.6
22.2
na
23.35 ( 1.85
9.9
5.8
29.9
6.8
7.1
na
5.21 ( 0.344
2.2
1.7
16.1
1.6
2.3
na
0.548 ( 0.156
52.75 ( 9.4
0.163 ( 0.02
28.5 ( 2.75
2.6 ( 0.2
10.9 ( 1.4
62 ( 0.00
20 ( 0.00
cis-(15R)-CP-epo 5b
cis-(15R)-CB-epo 6b
trans-(15S)-CP-epo 7
trans-(15S)-CB-epo 8
cis-(15S)-CP-py-epo 9
cis-(15R)-CP-py-epo 10
trans-(15S)-CP-py-epo 11
trans-(15R)-CP-py-epo 12
225
>300 (inactive)
>300 (inactive)
14.4 ( 0.00
>300 (inactive)
>300 (inactive)
3.7 ( 0.00
63 ( 0.00
180
na
na
na
2.7 ( 0.100
25.5 ( 1.50
1.40 ( 0.453
>300 (inactive)
0.625 ( 0.175
>300 (inactive)
17.8
>11.7
38.2
na
15.2
na
5.3
5.7
5.8
na
5.6
na
1.4
2.5
0.84
na
0.63
na
146 ( 0.00
8.15 ( 0.565
>300 (inactive)
3.49 ( 0.00
1.17 ( 0.94
>300 (inactive)
0.39 ( 0.00
nd
>300 (inactive)
>300 (inactive)
a The antiproliferative effects of the tested compounds against the parental 1A9 and the paclitaxel- and epothilone-selected drug resistant clones
(PTX10, PTX22, and A8, respectively) were assessed in a 72 h growth inhibition assay using the SRB (sulforhodamine-B) assay.24 IC50 values for
each compound are given in nM and represent the mean of 3-5 independent experiments ( standard error of the mean. Relative resistance (RR)
is calculated as an IC50 value for each resistant subline divided by that for the parental cell line (1A9). b Data from ref 3. CP ) cyclopropyl, CB
) cyclobutyl, na ) not applicable, nd ) not determined, py ) 5-methylpyridine side chain.
Figure 4. Electron microscopy of microtubules. Transmission electron microscopy of tubulin polymerization products induced by paclitaxel (Taxol),
epothilone A (1), epothilone B (2), cis-(15S)-CP-py-epo (9), trans-(15S)-CP-py-epo (11) and trans-(15R)-CP-py-epo (12). Samples were removed
from each polymerization reaction, placed on a 200 mesh Formvar-coated copper grid, stained with 0.5% uranyl acetate, and analyzed with use of
a JEOL 1210 analytical transmission electron microscope at 90 kV.
in high yield. Both the cis and trans isomers 115 and 122 were cyclized
by using the Yamaguchi protocol6 (70% yield), after which the C15
epimers were chromatographically separated, yielding compounds 116,
117, 123, and 124. Desilylation of these compounds finally afforded
the desired cyclopropyl epothilones 9-12 in excellent yields.
Chemical Biology
The biological activities of the synthesized epothilones were
evaluated through cytotoxicity and tubulin polymerization
assays. Cytotoxicity was first evaluated in a set of ovarian

9322 J. Am. Chem. Soc., Vol. 123, No. 38, 2001
Nicolaou et al.
Table 2. Tubulin Polymerization Potencya and Cytotoxicityb of
Epothilones 1 through 12 and Paclitaxel against Human Epidermoid
Cancer Cell Lines
Remarkably, the trans-cyclopropyl pyridine analogue 11
showed outstanding activity against all of our cell lines, with
IC50 ) 0.6 nM in the 1A9 human ovarian carcinoma cell line.
The cis analogue 9 was also highly active, but was a factor of
3 to 5 less active than 11. Again, the 15R isomeric analogues
(10 and 12) were inactive.
It is noteworthy that the active compounds (3, 4, 7, 8, 9, and
11) display a similar cytotoxicity profile against the â-tubulin
mutants compared to epothilone A (1) (see Table 1). In other
words, these compounds lose some activity against the clones
PTX10 (â270) and A8 (â274), suggesting that residues 270 and
274 are important for the binding of these analogues to tubulin.
However, the most active analogue (11) still retains IC50 < 10
nM for all of these cell lines. Furthermore, we found in the
current study, and in agreement with previous reports,3,20,21 that
the paclitaxel-selected clone PTX22 (â364) retains sensitivity
to the epothilones, especially in the case of the most active
analogues (9 and 11) where the relative resistance (RR) values
are <1.
The cytotoxicity analysis was supplemented with data from
two independent in vitro tubulin polymerization assays. In one
assay, the fraction of tubulin polymerized into microtubules upon
exposure to a given concentration of the respective compound
was determined (see Table 2). In the other assay, tubulin
polymerization kinetics upon exposure to the respective com-
pounds was determined by using purified rat brain tubulin
through measurement of the absorbance at 350 nm (see Figure
3). For this analysis, paclitaxel, epothilone A (1), and epothilone
B (2) were used as controls while compounds 9, 11, and 12
were selected for in vitro analysis. Compound 12 had no in
vitro activity consistent with the lack of cytotoxic activity for
this compound (Table 1). Compounds 9 and 11 exhibited good
in vitro activity although the maximum degree of tubulin
polymerization induced by these compounds was smaller
compared with that induced by epothilone A (1). However, the
increased cytotoxic activity of compounds 9 and 11 relative to
epothilone A (1) could potentially be explained by the faster
kinetics of polymerization induced by compounds 9 and 11 [time
to A350 ) 0.25 is <1 min for compounds 9 and 11, and 2 min
for epothilone A (1)].
compound
% TPa
KB-31b
KB-8511b
epothilone A (Epo A) 1
epothilone B (Epo B) 2
paclitaxel (Taxol)
69
90
49
2.15
0.19
2.92
1.91
0.18
626
cis-(15S)-CP-epo 3
cis-(15S)-CB-epo 4
83
79
0.838
60.7
0.408
29.7
cis-(15R)-CP-epo 5b
cis-(15R)-CB-epo 6b
trans-(15S)-CP-epo 7
trans-(15S)-CB-epo 8
cis-(15S)-CP-py-epo 9
cis-(15R)-CP-py-epo 10
trans-(15S)-CP-py-epo 11
trans-(15R)-CP-py-epo 12
26
29
100
82
100
6
159.5
378
0.971
23.1
66.7
156
0.641
11.5
0.446
>1000
0.676
>1000
0.618
>1000
0.835
930
94
<10
a % TP ) percent tubulin polymerized after incubation of tubulin
with a known concentration of compound (typically 3 µM). b Cyto-
toxicity toward human cancer cell lines as IC50 values given in nM.
KB-31: epidermoid Taxol-sensitive, KB-8511: epidermoid Taxol-
resistant (due to P-gp overexpression).
carcinoma cell lines, including a parental cell line (lA9) and
three drug-resistant cell lines, namely the paclitaxel-resistant
cell lines20 lA9/PTX10 and lA9/PTX22 and the epothilone-
resistant cell line21 1A9/A8. These resistant cell lines harbor
distinct acquired â-tubulin mutations which affect drug-tubulin
interaction and result in impaired taxane and epothilone-driven
tubulin polymerization. The results of these biological investiga-
tions are summarized in Table 1. Further cytotoxicity studies
were undertaken using a set of human epidermoid cancer cell
lines, including a parent cell line (KB-31) and a paclitaxel-
resistant (due to P-gp overexpression) cell line (KB-8511). The
results of these studies are summarized in Table 2.
In agreement with previous reports,3,4 we found that the
cyclopropyl epothilone A (3) inhibits slightly more potently the
1A9 and KB-31 cell growth than the parent compound
epothilone A (1). The 15S-cyclobutyl epothilone A (4) retains
good activity but is less potent than either 1 or 3. It is noteworthy
that the 15R-isomers (5 and 6) of both compounds are inactive
at low concentrations against the parental sensitive 1A9 and
KB-31 cells. Interestingly, even the (12R,13S)-trans-substituted
epothilones 7 and 8 showed good activity, again with the
cyclopropyl analogue being the most potent. These results are
in agreement with our previous report concerning trans-epoxide
analogues of epothilones A and B.22 In another study,23 we found
that (13R)-cyclopropyl epothilones 125 and 126 (see Figure 2),
originally incorrectly assigned as (13S)-diastereomers, were
inactive. Thus, we have now prepared and tested all four possible
diastereomers of 12,13-cyclopropyl epothilone A, and on the
basis of these results, it would appear that while the configu-
ration at C12 has relatively little influence on the cytotoxicity,
the 13S configuration is essential.
Finally, tubulin polymerization products of these compounds
were examined by electron microscopy (Figure 4) to rule out
the potential increase in absorbance due to the formation of
nonmicrotubule polymers. As seen in Figure 4, all compounds
tested induced the formation of microtubule polymers with the
exception of compound 12 where no microtubules were
observed.
Conclusion
In conclusion, we have constructed by total chemical synthesis
seven 12,13-cyclopropyl and three 12,13-cyclobutyl epothilone
analogues, and evaluated their biological activities against
tubulin and a series of cancer cell lines. Among the several
bioactive analogues, the novel, highly potent trans-cyclopropyl
pyridine epothilone A (11) stands out. This compound is one
of the most active epothilone analogues reported to date, with
IC50 ) 0.6 nM against the 1A9 ovarian carcinoma cell line.
While the other 15S compounds synthesized also exhibited
potent cytotoxicity against tumor cells, the 15R isomers were
devoid of such actions, re-enhancing the view that while the
oxygen atom at the C12-C13 site is not essential2j,3,4 for
biological activity, the proper configuration at C15 (15S) is
crucial for such action. Continuing investigations in this area
aim at further elucidation of the mechanism of action of the
epothilones and their structure-activity relationships.
(20) Giannakakou, P.; Sackett, D. L.; Kang, Y.-K.; Zhan, Z.; Buters, J.
T. M.; Fojo, T.; Poruchynsky, M. S. J. Biol. Chem. 1997, 272, 17118-
17125.
(21) Giannakakou, P.; Gussio, R.; Nogales, E.; Downing, K. H.;
Zaharevitz, D.; Bollbuck, B.; Poy, G.; Sackett, D.; Nicolaou, K. C.; Fojo,
T. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 2904-2909.
(22) Nicolaou, K. C.; Vourloumis. D.; Li, T.; Pastor, J.; Winssinger, N.;
He, Y.; Ninkovic, S.; Sarabia, F.; Vallberg, H.; Roschangar, F.; King, N.
P.; Finlay, M. R. V.; Giannakakou, P.; Verdier-Pinard, P.; Hamel, E. Angew.
Chem., Int. Ed. Engl. 1997, 36, 2097-2103.
(23) Nicoloau, K. C.; Finlay, M. R. V.; Ninkovic, S.; King, N. P.; He,
Y.; Li, T.; Sarabia, F.; Vourloumis, D. Chem. Biol. 1998, 5, 365-372.
(24) Skehan, P.; Storeng, D.; Monks, A.; McMahon, J.; Vistica, D.;
Warren, J. T.; Bokesch, H.; Kenney, S.; Boyd, M. R. J. Natl. Cancer Inst.
1990, 82, 1107-1112.

Synthesis of Cyclopropyl and Cyclobutyl Epothilone Analogues
J. Am. Chem. Soc., Vol. 123, No. 38, 2001 9323
Experimental Section
STINT Scholar Program (to A.R.), the Norwegian Research
Council (NFR) (to T.U.), and grants from Abbott, Amgen,
ArrayBiopharma, Boehringer-Ingelheim, Glaxo, Hoffmann-
La Roche, DuPont, Merck, Novartis, Pfizer, and Schering
Plough.
Full experimental details are provided as Supporting Information.
Acknowledgment. We thank Dr. Robert Apkarian (Emory)
for assistance with electron microscopy, and Drs. D. H. Huang
and G. Siuzdak (TSRI) for NMR and mass spectrometric
assistance, respectively. Financial support for this work was
provided by The Skaggs Institute for Chemical Biology, the
National Institutes of Health (USA), and CaPCURE, with
fellowships from the Naito Foundation, Japan (to M.S.), the
Supporting Information Available: Full experimental
details of all compounds (PDF). This material is available free
of charge via the Internet at http://pubs.acs.org.
JA011338B