Synthesis and biological evaluation of highly potent analogues of Epothilones B and D

Article abstract of DOI:10.1016/S0960-894X(00)00555-2

A series of new epothilone B and D analogues incorporating fused hetero-aromatic side chains have been prepared. The synthetic strategy is based on olefin 3 as the common intermediate and allows variation of the side-chain structure in a highly convergent and stereoselective manner. Epothilone analogues 1a-d and 2a-d are more potent inhibitors of cancer cell proliferation than the corresponding parent epothilones B orD. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0960-894X(00)00555-2

Bioorganic & Medicinal Chemistry Letters 10 (2000) 2765±2768
Synthesis and Biological Evaluation of Highly Potent Analogues
of Epothilones B and D
Karl-Heinz Altmann,* Guido Bold, Giorgio Caravatti, Andreas Florsheimer,
Vito Guagnano and Markus Wartmann
Novartis Pharma AG, TA Oncology Research, WKL-136.4.21, CH-4002 Basel, Switzerland
Received 3 August 2000; revised 28 September 2000; accepted 29 September 2000
AbstractÐA series of new epothilone B and D analogues incorporating fused hetero-aromatic side chains have been prepared. The
synthetic strategy is based on ole®n 3 as the common intermediate and allows variation of the side-chain structure in a highly
convergent and stereoselective manner. Epothilone analogues 1a±d and 2a±d are more potent inhibitors of cancer cell proliferation
than the corresponding parent epothilones B or D. # 2000 Elsevier Science Ltd. All rights reserved.
Cancer represents one of the major causes of mortality
in ®rst world countries and the search for better anti-
cancer drugs represents one of the most dicult and
important challenges in modern drug discovery. Per-
haps the most important anticancer drug developed
over the past decade is Taxol1 (paclitaxel),1 which acts
as a microtubule stabilizer2 and inhibits cancer cell pro-
liferation through interference with microtubule dynam-
ics.3 More recently, the natural products epothilones have
been identi®ed as a new class of microtubule stabilizing
agents (Fig. 1).4,5a
(Fig. 1) have also been reported to possess potent in
vivo antitumor activity in experimental animal models.
These ®ndings have triggered a host of activities in the
chemical community directed at the design and syn-
thesis of novel epothilone analogues, which has led to a
rather comprehensive understanding of the structure±
activity relationship for this class of natural products.8
However, with one exception,9 no epothilone analogues
have been described so far which exhibit more potent
antiproliferative activity than their natural counter-
parts. In this communication we report on a new class
of epothilone B and epothilone D analogues in which
the natural (2-(2-methyl-thiazol-4-yl)-1-methyl)-ethenyl
side-chain has been replaced by a number of benzo-het-
erocyclic moieties (Fig. 2) and which have generally
shown more potent antiproliferative activity than the
corresponding natural epothilones.
Figure 1. Structures of epothilones A (I, R=H), B (I, R=CH3), C (II,
R=H), and D (II, R=CH3). Epothilones C and D are also known as
deoxyepothilone A and deoxyepothilone B, respectively.
Like paclitaxel, these compounds exhibit potent anti-
proliferative activity on drug-sensitive cancer cell lines,
but in contrast to paclitaxel, and also other standard
cytotoxic anticancer drugs, they are equally active on
multidrug-resistant human cancer cells overexpressing
the P-gp 170 e‚ux pump.5 Epothilones B6 and D5c,7
*Corresponding author. Tel+41-61-696-5077; fax: +41-61-696-6246;
e-mail: karl-heinz.altmann@pharma.novartis.com
Figure 2. Epothilone analogues investigated in this study.
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00555-2

2766
K.-H. Altmann et al. / Bioorg. Med. Chem. Lett. 10 (2000) 2765±2768
Scheme 1. (i) (a) Ole®n 3 (1.25 equiv), 9-BBN, THF, rt, 4 h; (b) add to mixture of Cs2CO3 (1.5 equiv), PdCl2(dppf)2 (0.1 equiv), Ph3As (0.2 equiv),
vinyl iodide 4 (1 equiv), DMF, 10 ꢀC!rt, 16 h, 90%; (ii) LiOH (6 equiv), i-PrOH/H2O 4/1, 50 ꢀC, 7 h, 84%; (iii) TBAF (3 equiv), THF, rt, 18 h,
84%; (iv) (a) 2,4,6-Cl3C6H2C(O)Cl, Et3N, THF, 0 ꢀC, 15 min; (b) dilute with toluene, add to solution of DMAP in toluene, 75 ꢀC, 3 h, 70%; (v)
.
HF pyridine, THF, rt, 17 h, 73%; (vi) MeReO3, H2O2, pyridine, CH2Cl2, rt, 17 h; (vii) H2, Ra-Ni, EtOH, rt, 37% (2 steps).
Our highly convergent general strategy for the synthesis
of these analogues is exempli®ed in Scheme 1 for
quinoline-based structures 1d and 2d. Alkyl-Suzuki
coupling between ole®n 3 and vinyl iodide 410 provided
the fully protected seco-acid 5 as a single double bond
isomer and in excellent yield (90%). Ester saponi®cation
followed by selective deprotection of the 15-hydroxyl
group with TBAF gave the free seco-acid, which under
Yamaguchi conditions11 smoothly cyclized to the pro-
tected macrolactone 6 in 70% yield. Removal of the
the desired diastereoisomer 12 in 57% yield for the two-
step sequence from aldehyde 11. Direct reduction of 12
to aldehdye 13 and subsequent ole®nation with
[Ph3CHICH3]+I
17
gave vinyl iodide 4 (35% from 12).
.
TBS protecting groups with HF pyridine then provided
epothilone D analogue 2d (73%). Epoxidation of the
12
ole®nic double bond with MeReO3/H2O2 proceeded
with ca. 6/1 selectivity in favor of the desired epoxide
isomer, but was slightly complicated by the fact that
N-oxidation of the quinoline moiety was faster than
epoxidation. 2d was thus ®rst converted to the doubly
oxidized species; selective reduction of the N-oxide with
H2/Ra-Ni then provided the desired epothilone B ana-
logue 1d in 37% yield for the two-step sequence from
2d.
The syntheses of ole®n 3 and vinyl iodide 4 are sum-
marized in Schemes 2 and 3, respectively. Brie¯y, the
key step in the preparation of 3 (Scheme 2) consists of
the highly diastereoselective aldol reaction between
aldehyde 713 and ketone 814 which provided the desired
aldol diastereoisomer 9 in 82% yield after FC. Follow-
ing simple protecting group manipulations and func-
tional group transformations ole®n 3 was ®nally
obtained in 20% overall yield for the nine-step sequence
from aldehyde 7.
Scheme 2. (i) (a) 8 (2 equiv), LDA, THF, 78 ꢀC, 80 min; (b) +7
(1 equiv), 78 ꢀC, 75 min, 82%; (ii) PPTS, MeOH, rt, 22 h, 83%; (iii)
TBS-OTf, lutidine, CH2Cl2, rt, 84%; (iv) CSA, MeOH/CH2Cl2, 0 ꢀC,
1 h, 80%; (v) (COCl)2, DMSO, Et3N, CH2Cl2, 78 ꢀC, 93%; (vi)
NaClO2, iso-butene, NaH2PO4, THF, tert±BuOH, H2O, rt, 4 h, 93%;
(vii) DCC, DMAP, MeOH, CH2Cl2, 20 ꢀC!rt, 4 h, 71%; (viii) H2,
Pd-C, MeOH, rt, atm. pressure, 1 h, 80%; (ix) (a) 2-NO2-PhSeCN,
Bu3P, rt, 1 h; (b) NaHCO3, 30% H2O2, rt, 24 h, 88%.
Preparation of 4 (Scheme 3) was based on the dia-
stereoselective aldol reaction between acetyl sultam 10
and aldehyde 1115 as the key step.16 The resulting aldol
product was obtained as a 5/1 mixture of diastereo-
isomers, which, after silylation of the secondary
hydroxyl group, could be separated by FC to provide

K.-H. Altmann et al. / Bioorg. Med. Chem. Lett. 10 (2000) 2765±2768
2767
Table 1. Induction of tubulin polymerization and antiproliferative activity for compounds 1a±d and 2a±d in comparison to epothilones A, B, D
Cpda
R
X
% Tubulin polymerizationb
IC50 KB-31c,e (nM)
IC50 KB-8511d,e (nM)
1a
1b
1c
1d
CH3
H
CH3
H
S
83
97
99
78
0.13 Æ 0.02
0.13 Æ 0.01
0.14 Æ 0.01
0.11 Æ 0.01
0.09 Æ 0.02
0.46 Æ 0.03
0.38 Æ 0.10
0.10 Æ 0.02
N(CH3)
N(CH3)
CHˆCH
2a
2b
2c
2d
CH3
H
CH3
H
S
76
86
94
90
0.45 Æ 0.03
0.46 Æ 0.09
0.21 Æ 0.01
0.59 Æ 0.04
0.23 Æ 0.03
0.91 Æ 0.10
0.73 Æ 0.04
0.38 Æ 0.03
N(CH3)
N(CH3)
CHˆCH
Epothilone Af
Epothilone Bf
Epothilone Df
Paclitaxel
69
90
83
49
2.15 Æ 0.07
0.19 Æ 0.05
2.70 Æ 0.76
2.92 Æ 0.18
1.91 Æ 0.07
0.18 Æ 0.01
1.44 Æ 0.46
626 Æ 32
aCf. Fig. 2.
bInduction of polymerization of porcine brain-derived microtubule protein by 2 mM of test compound relative to the e€ect of 25 mM epothilone B,
which gave maximal polymerization (85% of protein input). Tubulin polymerization was determined using a modi®ed version of the centrifugation
assay described in ref 18.
cConcentration required to inhibit the growth of paclitaxel-sensitive human epidermoid carcinoma cells KB-31 by 50% (72 h exposure).
dConcentration required to inhibit the growth of paclitaxel-resistant human epidermoid carcinoma cells KB-8511 by 50% (72 h exposure). KB-8511
is a P-gp 170 overexpressing paclitaxel-selected subline of the KB-31 parental line.
eData are presented as mean Æ standard error of the mean of at least three independent experiments.
fCf. Fig. 1.
The biological activities of epothilone B analogues 1a±d
and epothilone D analogues 2a±d are summarized in
Table 1. Antiproliferative activity was assessed against
the drug-sensitive human epidermoid cancer cell line
KB-31 and a P-gp-overexpressing, paclitaxel-resistant
subline KB-8511.19 All epothilone B analogues (epoxides
1a±d) are more potent inhibitors of KB-31 cell pro-
liferation than epothilone B itself; 1a and 1d are also
more active against KB-8511 cells. In addition, quino-
line derivative 1d is more potent than a pyridine-based
analogue (pyridine-epothilone B 149), containing a sim-
ple 2-pyridyl moiety in place of the 2-methyl thiazole
ring in natural epothilones (IC50s for 14 are 0.30 nM
and 0.30 nM against KB-31 and KB-8511 cells,
respectively9).
The increase in growth inhibitory activity conferred by
the benzo-heterocyclic side-chains is more pronounced
for the epothilone D analogues 2a±d, which inhibit the
growth of KB-31 and KB-8511 cells up to 13-fold more
potently than epothilone D. In addition, all epothilone
D analogues are clearly more potent than epothilone A.
It should be noted, however, that di€erences in cellular
activity do not directly correlate with the tubulin poly-
merization data shown in Table 1, which de®ne the
ability of a compound to induce tubulin polymerization
in vitro at a ®xed concentration (2 mM) relative to the
e€ect of 25 mM epothilone B. However, these data pro-
vide only a gross measure for tubulin anity and in
order to gain a better understanding of the relationship
between the induction of tubulin polymerization in vitro
and antiproliferative activity at the cellular level, we
have determined accurate EC50-values for in vitro
tubulin polymerization for a series of selected com-
pounds. Thus, the concentrations required to induce
50% polymerization of porcine brain-derived micro-
tubule protein (EC50) are 0.67 mM for epothilone B,
1.12 mM for epothilone A, 0.88 mM for epothilone D,
and 0.93 mM for compound 2a. These data suggest that
at least for 2a the improved antiproliferative activity
over epothilone D is not the result of more pronounced
e€ects on tubulin/microtubules, but may rather be rela-
ted to parameters such as cellular uptake or intracellular
processing.
Scheme 3. (i) (a) Et3B, CF3SO3H, hexane, rt, 20 min, 40 ꢀC, 10 min;
(b) +Hunig's base, 0 ꢀC; (c) +10, 5 ꢀC, 5 min; (d) +11, 78 ꢀC, 4 h,
82% (5/1 mixture); (ii) TBS-Cl, imidazole, DMF, 40 ꢀC, 16 h, 70%.
(iii) DIBAL-H, CH2Cl2, 78 ꢀC, 6 h, 79%; (iv) (a) [Ph3PCHICH3]+I ,
NaHMDS (0.95 equiv), THF, 78 ꢀC, 30 min, 15 ꢀC, 20 min; (b)
+13 (0.8 equiv), 78 ꢀC, 40 min, 44%.
In summary, we have developed a highly convergent
route for the synthesis of epothilone B analogues 1a±d
and epothilone D analogues 2a±d. Several of these

2768
K.-H. Altmann et al. / Bioorg. Med. Chem. Lett. 10 (2000) 2765±2768
compounds are more potent inhibitors of human cancer
cell growth than the respective parent epothilone and
this is particularly true in the epothilone D series.20
Given the fact that epothilone D has been suggested to
be a promising anticancer drug candidate with a poten-
tially improved therapeutic index over epothilone B
(due to the absence of the epoxide moiety as the pur-
ported source of unspeci®c toxicity),5c,7 our improved
epothilone D analogues could represent interesting can-
didates for anticancer drug development. We are cur-
rently in the process of evaluating the in vivo antitumor
activity of these compounds and the results of these
studies will be reported in due course.
7. Chou, T.-C.; Zhang, X.-G.; Harris, C. R.; Kuduk, S. D.;
Balog, A.; Savin, K. A.; Bertino, J. R.; Danishefsky, S. J.
Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 15798.
8. For reviews, cf.: (a) Nicolaou, K. C.; Roschangar, F.;
Vourloumis, D. Angew. Chem. 1998, 110, 2120; Nicolaou, K.
C.; Roschangar, F.; Vourloumis, D. Angew. Chem., Int. Ed.
Engl. 1998, 37, 2014. (b) Harris, C. R.; Danishefsky, S. J. J.
Org. Chem. 1999, 64, 8434.
9. Nicolaou, K. C.; Scarpelli, R.; Bollbuck, B.; Werschkun,
B.; Manuela, M.; Pereira, A.; Wartmann, M.; Altmann,
K.-H.; Zaharevitz, D.; Gussio, R.; Giannakakou, P. Chem.
Biol. 2000, 7, 593. As suggested by one of the reviewers, it
should be noted that 14 is not the most active pyridine-based
epothilone analogue described in the above reference. Analo-
gues with a 4- or 5-methyl substituent on the pyridine ring are
in fact more potent (IC50s of 0.16 nM and 0.11 nM, respec-
tively). However, to highlight the e€ect of the benzo-heterocycle
modi®cation discussed in this paper, 14 represents the most
appropriate reference compound for a comparison with 2d.
10. The usefulness of an alkyl-Suzuki coupling to establish the
C11±C12 bond in epothilones was ®rst recognized by Pro-
fessor Danishefsky's laboratory: Balog, A.; Meng, D.; Kame-
necka, T.; Bertinato, P.; Su, D.-S.; Sorensen, E. J.;
Danishefsky, S. J. Angew. Chem., Int. Ed. Engl. 1996, 35, 2801.
11. Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yama-
guchi, M. Bull. Chem. Soc. Jpn. 1979, 52, 1989.
Acknowledgements
The excellent technical assistance by Kurt Hauenstein,
Jurgen Koppler, Michele Lartigot, Jacqueline Loretan,
Robert Reuter, Werner Vetterli, Rene Vogelsanger, and
Severine Wojeik is gratefully acknowledged.
References and Notes
12. Rudolph, J.; Reddy, K. L.; Chang, J. P.; Sharpless, K. B.
J. Am. Chem. Soc. 1997, 119, 6189.
1. Rowinsky, E. K. Ann. Rev. Med. 1997, 48, 353. Through-
out this paper we shall refer to the compound as `paclitaxel',
which is the active drug substance of Taxol1.
13. Aldehyde 7 was obtained through Swern oxidation of the
corresponding primary alcohol: Schinzer, D.; Bauer, A.;
Schieber, J. Chem. Eur. J. 1999, 2492.
2. Schi€, P. B.; Horwitz, S. B. Proc. Natl. Acad. Sci. U.S.A.
1980, 77, 1561.
14. Schinzer, D.; Limberg, A.; Bohm, O. M. Chem. Eur. J.
1996, 2, 1477.
3. Wilson, S.; Jordan, M. A. Chem. Biol. 1995, 2, 569.
4. (a) Ho¯e, G.; Bedorf, N.; Gerth, K.; Reichenbach, H.
German patent disclosure DE 4138042, 5 May, 1993 (Priority
19 November, 1991). (b) Gerth, K.; Bedorf, N.; Ho¯e, G.;
Irschik, H.; Reichenbach, H. J. Antibiotics 1996, 49, 560.
5. (a) Bollag, D. M.; McQueney, P. A.; Zhu, J.; Hensens, O.;
Koupal, L.; Liesch, J.; Goetz, M.; Lazarides, E.; Woods, C. A.
Cancer Res. 1995, 55, 2325. (b) Kowalski, R. J.; Giannakakou,
P.; Hamel, E. J. Biol. Chem. 1997, 272, 2534. (c) 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.
15. Kingsbury, W. D.; Pendrak, I.; Leber, J. D.; Boehm, J. C.;
Mallet, B.; Sarau, H. M.; Foley, J. J.; Schmidt, D. B.; Daines,
R. A. J. Med. Chem. 1993, 36, 3308.
16. Bond, S.; Perlmutter, P. J. Org. Chem. 1997, 62, 6397.
17. Chen, J.; Wang, T.; Zhao, K. Tetrahedron Lett. 1994, 35,
2827.
18. Lin, C. M.; Yian, Y. Q.; Chadhary, A. G.; Rimoldi, J. M.;
Kingston, D. G. I.; Hamel, E. Cancer Chem. Pharm. 1996, 38,
136.
19. Akiyama, S.; Fojo, A.; Hanover, J. A.; Pastan, I.; Got-
tesmann, M. M. Somat. Cell. Mol. Genet. 1985, 11, 117.
20. This phenomenon is not limited to the KB lines described
in this paper, but extends to a broad range of other human
cancer cell lines (Wartmann, M., unpublished).
6. Altmann, K.-H.; Wartmann, M.; O'Reilly, T. Biochim.
Biophys. Acta 2000, 1470, M79.