Total synthesis and antitumor activity of ZK-EPO: The first fully synthetic epothilone in clinical development

Article abstract of DOI:10.1002/anie.200602785

Going to trial: From about 350 active epothilone analogues synthesized by a highly convergent synthesis, one (ZK-EPO, see picture) has been chosen for clinical development on the basis of its outstanding preclinical data. This compound exhibits higher activity and efficacy than taxanes (e.g. paclitaxel) and second-generation epothilones, a fast and efficient cellular uptake, no recognition by efflux mechanisms, and an improved therapeutic window. (Chemical Equation Presented)

Full text of DOI:10.1002/anie.200602785

Communications
Natural Products
DOI: 10.1002/anie.200602785
Total Synthesis and Antitumor Activity of
ZK-EPO: The First Fully Synthetic Epothilone in
Clinical Development**
Ulrich Klar,* Bernd Buchmann, Wolfgang Schwede,
Werner Skuballa, Jens Hoffmann, and
Rosemarie B. Lichtner
The clinical success of paclitaxel (PT) in the treatment of
ovarian and breast cancer strongly contributed to the assess-
ment that tubulin is one of the best clinically validated targets
in tumor therapy. However, one disadvantage with PT and
other taxane analogues is their recognition by cellular efflux
mechanisms, such as the p-glycoprotein (p-gp), which con-
tribute to a loss of activity in cells overexpressing the
multidrug-resistance (MDR) phenotype.
In 1995 Bollag et al.[1] reported that the natural product
class of epothilones mimics the biological activity of PT with
respect to its action on the tubulin system. Thus, the cell cycle
is arrested in the G2/M phase and the cells are driven into
mitotic catastrophe and/or apoptosis. In contrast to PT,
epothilones possess the potential to overcome MDR in vitro
and, most importantly, in vivo.[2–5] These findings stimulated
intensive research activities in chemistry, pharmacology, and
medicine.
Epothilones were first discovered by Reichenbach, Höfle,
and co-workers who characterized the compounds they had
isolated from the myxobacterial strain Sorangium cellulo-
sum.[6,7] They also described their cytotoxic potential in a
patent application filed in 1991.[8]
The first total syntheses of epothilone A were reported in
1996/1997 by the research groups of Danishefsky, Nicolaou,
and Schinzer.[9–11] Since we had no access to the more potent
compound epothilone B, from the start of our project we
focused on total synthesis to evaluate the potential of this
compound class. During our own investigations, the total
[12–15]
syntheses of epothilones B and Dwere reported.
Epo-
[*] Dr. U. Klar, Dr. B. Buchmann, Dr. W. Schwede, Prof. W. Skuballa,
Dr. J. Hoffmann, Dr. R. B. Lichtner
Schering AG, Research Center Europe
Müllerstrasse 178, 13342 Berlin (Germany)
Fax: (+49)30-468-92635
E-mail: ulrich.klar@schering.de
Homepage: http://www.schering.de
[**] We thank Siegfried Bäsler, Stefan Barm, Angela Bieseke, Klaus
Burmeister, Klaus Cornelius, Gisbert Depke, Ralf Günther, Wolfgang
Halfbrodt, Frank Halwass, Karola Henschel, Hiltraud Hillmann,
Klaus-Dieter Joachim, Frank Kuczynski, Hue-Quan Lao, Günter
Michl, Gerhard Neumann, Christian Pingel, Ingrid Pissarczyk, Peter
Pretsch, Bodo Röhr, Oliver Schenk, Michael Schirner, Jürgen
Skupsch, Marion Slopianka, Daniela Wrobel, and Henk Zimmer-
mann for their excellent scientific and technical assistance.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
7942
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7942 –7948

Angewandte
Chemie
thilone B proved to be highly active in proliferation assays,
but also highly toxic in animal models. On the basis of these
results we started a lead optimization program, with the
primary goal being to maintain the high activity of the natural
compound while improving the therapeutic window.
A highly convergent total synthesis enabled us to intro-
duce structural modifications at nearly every position of the
macrolactone. We found that epothilone Dand most of its
analogues had a significantly reduced activity compared with
epothilone B. Even though they are not good substrates for p-
gp, a delayed cellular uptake and an easy cellular release was
observed, which strengthened our focus on compounds of the
epothilone B type.[16]
EPO (20) as well as the synthetic route developed in our
research.
We applied a highly convergent strategy based on our
lead-optimization program that offered good flexibility for
introducing structural modifications at nearly every position
of the 16-membered-ring skeleton. This strategy is reflected in
our synthetic route to ZK-EPO (Scheme 2) where we applied
a Wittig reaction to connect C12 and C13, an aldol reac-
tion[10,11] to establish stereocenters C6 and C7, and finally a
Yamaguchi cyclization for macrolactonization.[12]
To increase the chemical as well as the enzymatic stability
we replaced the lactone by a lactam moiety, and were
surprised to discover that the compounds were then well
recognized by MDR mechanisms thus resulting in a taxane-
like profile. Several other structural modifications have since
been identified that also change the compound profile in this
unwanted direction. During extensive studies on structure/
property—as well as structure/activity—relationships we
learned that it is possible to further improve the activity of
the natural compounds in multi-drug resistant cell lines and to
reduce the toxicity in vivo without reducing the antitumor
efficacy. The side chain at C15[17] and the residue at C6 were
identified as key positions for tuning tolerability and effi-
cacy.[18,19] From about 350 active[20] epothilone analogues
synthesized in our laboratories, we have chosen ZK-EPO (20)
on the basis of its outstanding preclinical properties for
clinical development. This compound combines high activity
and efficacy, a fast and efficient cellular uptake, no recog-
nition by efflux mechanisms, and an improved therapeutic
window.
Scheme 2. Strategic bond formation in the construction of ZK-EPO.
Segment A is constructed from a chiral pool synthesis
starting with (ꢀ)-pantolactone (Scheme 3).[23] Even though
11 steps were required, this synthesis allows the preparation
of C6-modified building blocks in large quantities.
Scheme 1 shows the four epothilones currently under-
going advanced clinical development.[21,22] The natural com-
pounds epothilones B and Dare produced by biosynthesis
and the lactam BMS247550 by partial synthesis starting from
epothilone B. We now disclose the chemical structure of ZK-
Scheme 3. Synthesis of the C1–C6 segment (segment A) from
(ꢀ)-pantolactone.
Besides the possibility for structural modifications, seg-
ment B represents a central strategic synthon. By using an
appropriate protecting-group strategy the segments can be
connected either by the sequence C + B!CB + A!CBA or
by the sequence A + B!AB + C!ABC. This flexibility was
achieved during the synthesis of potential drug compounds
through an 11-step synthesis that allowed us to incorporate
the most valuable building blocks A or C last.[19]
After selecting ZK-EPO as the candidate for develop-
ment, and favoring the connection sequence C + B!CB +
A!CBA, the synthesis of segment B can be shortened
considerably (Scheme 4).
The readily available chiral starting material 1 (Roche
ester) is first protected as a tetrahydropyranyl ether. After
reduction of the ester moiety, the resulting primary alcohol is
derivatized to tosylate 2. Coupling of 2 with 2-methyl-1-
butenyl-4-magnesium bromide (4) in the presence of a
catalytic amount of Li2CuCl4[24,25] leads to 5. Bishydroxylation
of the double bond followed by oxidative degradation
provides methyl ketone 6 in an overall yield of 47% starting
from 1.
Scheme 1. Epothilones in advanced clinical trials, their names, produc-
tion method, and development status.
Angew. Chem. Int. Ed. 2006, 45, 7942 –7948ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7943

Communications
stereoisomer that is removed by chromatography (Scheme 7).
The addition of zinc chloride significantly improves the
diastereoselectivity, which normally decreases with increasing
size of residues at C6. The remaining transformations to the
target molecule are well established in the total synthesis of
epothilones.[27–29] Removal of the ketal group in 16 followed
by persilylation generates tetrasilyl ether 17. The primary silyl
ether is removed under mild acidic conditions followed by a
two-step oxidation sequence which establishes carboxylic acid
18. Next, the benzylic alcohol is released and the crude
hydroxy acid subjected to Yamaguchi cyclization conditions
to generate the 16-membered macrolactone. The remaining
protecting groups are removed to yield epothilone Dana-
logue 19. The addition of hexafluorosilicic acid significantly
Scheme 4. Stringent synthetic route to the C7–C12 segment (seg-
ment B). Reagents and conditions: a) 1. DHP, pTsOH, CH2Cl2; 2. LAH,
diethyl ether, RT, 87%; 3. TsCl, pyridine, RT, 97%; b) Mg/Li2CuCl4
(cat.), THF, ꢀ708C to RT, 75%; c) NBS, PPh3, CH2Cl2,
RT, 51%; d) OsO4, NaIO4, H2O, THF, RT, 74%.
THP=tetrahydropyranyl, DHP=dihydropyran,
pTsOH=toluene-4-sulfonic acid, LAH=lithium alumi-
num hydride, Ts=toluene-4-sulfonyl, NBS=N-bromo-
succinimide.
Benzothiazole 8 is obtained from benzoic
acid 7 in a one-pot reaction (Scheme 5). Alde-
hyde 9 is prepared by a reduction–oxidation
sequence and then submitted to an Evans aldol
addition of 3-acetyl-(4S,5R)-4-methyl-5-phenyl-
2-oxazolidinone (10) to establish the stereogen-
ic center at C15. The excess of chiral auxiliary
can be considerably reduced by the presence of
zinc chloride to give a 8:2 mixture of diastereo-
isomers from which pure 11 is isolated directly
by crystallization.[26] The relative stereochemis-
try was determined by X-ray analysis of 11,
thereby confirming the correct absolute config-
uration at C15.
The most efficient way to recover the chiral
Scheme 5. Synthesis of the C13–C15 segment (segment C). Reagents and conditions:
auxiliary was transesterification to 12 and
subsequent reduction to the corresponding
alcohol, which is then transformed to crystalline
phosphonium salt 13 (segment C). The overall
yield starting from 7 is 16%.
a) Na2S , AcO, HOAc, reflux, 71%; b) 1. LAH, THF, RT to reflux, 71%; 2. SO3·pyridine,
2
Et3N, CH2Cl2/DMSO, RT, 95%; c) nBuLi, ZnCl2, THF, ꢀ708C; d) 1. TBDMSCl, imid-
azole, DMF, RT, 91%; 2. Ti(OEt)4, EtOH, reflux, 91%; e) 1. DIBAH, toluene, CH2Cl2,
ꢀ408C, crude; 2. I2, PPh3, CH2Cl2, RT, 92%; 3. PPh3, toluene, 1008C, 87%.
TBDMS=tert-butyldimethylsilyl, DIBAH=diisobutylaluminum hydride.
Segments B (6) and C (13) are connected by
a Wittig reaction to afford a nearly 1:1 mixture
of E/Z isomers.[12] After removal of the tetra-
hydropyranyl ether the isomers can be sepa-
rated by chromatography to yield stereochem-
ically pure (Z)-14 and (E)-14 (Scheme 6). The
configuration of the double bond was assigned
unambiguously by NOE experiments. (E)-14
can be isomerized by irradiation with light with
a wavelength greater than 280 nm to give a 6:4
mixture of (E)-14 and (Z)-14 in a total yield of
over 90%. Finally, alcohol (Z)-14 is oxidized to
aldehyde 15 (segment BC). The total yield is
36% (photochemical recycling not included).
Aldehyde 15 was treated with segment A in
an aldol reaction to yield 16 with good selec-
tivity, along with minor amounts of its 6,7 dia-
Scheme 6. Synthesis of C7–C15 segment (segment BC). Reagents and conditions:
a) 1. NaHMDS, THF, 08C to RT, 83%; 2. pTsOH (cat.), EtOH, RT, 86%; b) hn,
toluene/acetone, RT, 92%; c) Swern oxidation, crude. HMDS=1,1,1,3,3,3-hexamethyl-
disilazane.
7944 www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7942 –7948

Angewandte
Chemie
thilone Danalogue—the latter showing
significantly reduced antiproliferative
activity relative to its 14(S) epimer
which adopts a conformation similar to
epothilone B.[31] These investigations sug-
gest that the conformation of ZK-EPO in
the crystal lattice might not be the con-
formation adopted at the common
taxane/epothilone tubulin binding site.
A panel of 20 different human tumor
cell lines was exposed to different con-
centrations of ZK-EPO and compared
with PT (Figure 2, top), doxorubicin
(adriamycin ADM), cisplatin (CDDP),
and camptothecin (CPT, Figure 2,
bottom). The degree of growth inhibition
was measured three days after initiation
of drug treatment. The activity of each
compound is given as a concentration
that inhibits cell proliferation by 50%
(IC50 values). Cell lines marked in red
show significant resistances to standard
cytotoxic drugs. ZK-EPO proved to be
highly active in all of these proliferating
human tumor cell lines, irrespective of
their MDR phenotype, and showed a
mean IC50 value below 1 nm.
Scheme 7. Synthesis of the C1–C15 segment (segment ABC) and completion of the synthesis
of ZK-EPO (20). Reagents and conditions: a) LDA, ZnCl2, THF, ꢀ708C; 64%; b) 1. pTsOH
(cat.), EtOH, RT, 97%; 2. TBDMSOTf, 2,6-lutidine, CH2Cl2, ꢀ708C to 08C, 96%; c) 1. CSA,
CH2Cl2, MeOH, RT, 80%; 2. Swern oxidation, crude; 3. NaOCl2, NaH2PO4, 2-methyl-2-butene,
THF, H2O, tBuOH, 08C to 158C, 85%; d) 1. TBAF, THF, RT, crude; 2. Yamaguchi cyclization,
60%; 3. HF·pyridine, hexafluorosilicic acid, THF, RT, 87%; e) DMDO, acetone, CH2Cl2,
ꢀ788C, 71% + 10% b-epoxide. LDA=lithium diisopropylamide, Tf=trifluoromethanesul-
fonyl, CSA=camphorsulfonic acid, TBAF=tetrabutylammonium fluoride, DMDO=3,3-dime-
thyldioxirane.
To determine whether ZK-EPO also
influences nonproliferating nontumor
cells, ZK-EPO was tested on proliferat-
ing or confluent (nonproliferating)
human immortal keratinocyte HaCat
cells and the results compared with
those with PT.[32] The growth of prolifer-
ating HaCat cells was potently inhibited
facilitates the cleavage of both silyl ethers in terms of reaction
time, yield, and insensitivity to the quality of different batches
of HF·pyridine. Epoxidation of the double bond affords the a-
epoxide ZK-EPO (20) with high stereoselectivity and in 15%
overall yield along with minor amounts of the biologically less
active b-epoxide.
The longest linear sequence (C!CB!ZK-EPO)
involves 22 steps and occurs in an overal yield of 0.9%.
With this synthetic route we succeeded in producing 36 g of
compound, which was sufficient to perform nearly all the
preclinical investigations.
by PT and ZK-EPO while only minor effects were observed
on confluent cells. This observation indicates a very low toxic
potential on normal, nondividing cells.
Even modest structural modifications may have consid-
erable effects on the recognition through MDR-related
pathways; this is illustrated by comparing the antiproliferative
activity in the sensitive MDR(ꢀ) human breast cancer cell
line MCF-7 (PT-sensitive) with the multidrug-resistant
MDR(+) cancer cell line NCI/ADR (PT-insensitive;
Figure 3). Epothilone B is highly active towards the
The relative and absolute config-
uration of ZK-EPO was confirmed
by X-ray analysis (Figure 1, left). It is
interesting to note that the observed
conformation differs considerably
from that of epothilone B (Figure 1,
right):[30] in ZK-EPO, both hydroxy
groups as well as the epoxide are
oriented towards the center of the
macrocyle. A similar conformation
was postulated for 14(S)-methyl-
epothilone B, as deduced from its
corresponding 14(R)-methyl-epo-
Figure 1. The absolute and relative configuration of ZK-EPO (20; left) and comparison with the conformation
of epothilone B (right).
Angew. Chem. Int. Ed. 2006, 45, 7942 –7948ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7945

Communications
In the next experiment we tested whether ZK-EPO or
epothilone B are still recognized by efflux mechanisms. The
action of verapamil on drug accumulation in cells is often used
to show that the efflux is mediated by the MDR1 p-gp. Thus,
10 mm verapamil was added together with paclitaxel, epothi-
lone B, or ZK-EPO to NCI/ADR cells and the cellular growth
estimated after three days of exposure. As expected, verapa-
mil renders NCI/ADR cells responsive to growth inhibition
by 100 nm PT. Furthermore, verapamil decreases the
IC50 value for growth inhibition by epothilone B from 2.4 to
0.3 nm, but does not further enhance the potent antiprolifer-
ative effects of ZK-EPO on p-gp that overexpress NCI/ADR
cells.[32] These data further indicate that ZK-EPO is not
recognized by cellular efflux pumps such as p-gp.
The following examples demonstrate that the activity seen
in vitro also translate into in vivo models. Human cervix
cancer cells (HeLa/MaTu) from large cervix tumors were
xenotransplanted into nude mice. The growth of these HeLa/
MaTu cervix tumors could be efficiently inhibited by treat-
ment with PT (12 mgkgꢀ1 once daily for 5 consecutive days;
data not shown). A single treatment with ZK-EPO resulted in
an almost complete inhibition of tumor growth whereas the
untreated control showed rapid growth (Figure 4, top). Treat-
ment with ZK-EPO resulted in a dose-dependent inhibition
of the tumor growth. Response was seen at all doses after one
initial treatment. Further tumor growth was completely
inhibited at a dose of 8 mgkgꢀ1. At the highest dose level
(8 mgkgꢀ1, application at days 6 and 25) the compound was
well tolerated, as demonstrated by a modest reduction in
body weight (about 5%) and a time to recover of less than
seven days (Figure 4, bottom).
Figure 2. ZK-EPO efficiently inhibits cell proliferation of different
human tumor types in vitro.
In the multidrug-resistant HeLa/MaTu/ADR cervix
cancer model PT did not show an antitumor effect, while
treatment with ZK-EPO resulted in an almost complete
inhibition of tumor growth, whereas the untreated control
showed rapid growth (Figure 5, left). A single application of
6 mgkgꢀ1 of ZK-EPO resulted in a complete regression in
50% of these highly resistant tumors. A second dose was not
given in this experiment to determine the time to progression.
Severe toxic side effects were not observed with these
treatment schedules.
Figure 3. ZK-EPO maintains its high antiproliferative activity even in
tumor cell lines overexpressing the MDR phenotype.
MDR(ꢀ) cell line and also displays considerable activity
towards the MDR(+) cell line, although the activity is
reduced about tenfold. If the lactone is replaced by a lactam
moiety (for example, in BMS247550[33,34]), the activity at the
MDR(ꢀ) cell line is significantly reduced and is nearly lost
with the MDR(+) cell line, thus resulting in a PT-like profile.
Replacement of the methyl group on the thiazole ring in the
side chain of epothilone B by an aminomethylene group (for
example, as in BMS310705[35]) generates a compound that
shows nearly the same activity towards the MDR(ꢀ) cell line
as epothilone B, but the activity towards the MDR(+) cell
line is reduced about 100-fold. The natural product epothil-
one Dpossesses comparable activity in both cell lines, but
with a significantly reduced overall activity.[36] ZK-EPO,
however, maintains its high activity also towards the
MDR(+) cell line. This finding is an indication that, in
contrast to other epothilones or PT, ZK-EPO is not exported
by the p-gp.
The models discussed are selected from tumor indications
suitable for chemotherapy and include breast, ovarian, small
cell lung, non-small cell lung, prostate, and colorectal cancers.
ZK-EPO shows high efficacy in all these tumor types,
including both MDR-negative and MDR-positive tumors.[37]
In addition, ZK-EPO displays high efficacy in tumor types
normally not sensitive to chemotherapy. Among these tumor
types are pancreatic, cholangio, renal cell, and hepatocellular
cancers, melanoma, and brain tumors. An example of a
melanoma cancer model demonstrating the high efficacy of
ZK-EPO is depicted in Figure 5 (right).
Our preclinical data indicate that ZK-EPO exhibits a
more potent antitumor activity than taxanes (for example,
paclitaxel) and second-generation epothilones such as ixabe-
pilone (BMS247550). ZK-EPO eludes the cellular efflux
pumps normally responsible for MDR and has activity against
both taxane-resistant and taxane-sensitive tumor models. ZK-
EPO uptake may be unaffected by MDR-pump activity, thus
7946 www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7942 –7948

Angewandte
Chemie
suggesting efficient maintenance within tumor cells. There-
fore, ZK-EPO may be clinically effective as a first-line
therapy that delays or prevents MDR-based resistance.
Following successful phase I clinical trials,[38] the potential
of ZK-EPO in the treatment of different human tumors is
now beeing evaluated in an extensive phase II clinical trials
program.
Received: July 13, 2006
Published online: September 27, 2006
Keywords: clinical development · epothilones ·
.
natural products · total synthesis
[1] D. M. Bollag, P. A. McQueeney, J. Zhu, O. Hensens, L. Koupal, J.
Liesch, M. Goetz, E. Lazarides, C. M. Woods, Cancer Res. 1995,
55, 2325.
[2] T.-C. Chou, X.-G. Zhang, A. Balog, D.-S. Su, D. Meng, K. Savin,
J. R. Bertino, S. J. Danishefsky, Proc. Natl. Acad. Sci. USA 1998,
95, 9642.
[3] T.-C. Chou, X.-G. Zhang, C. R. Harris, S. D. Kuduk, A. Balog, K.
Savin, J. R. Bertino, S. J. Danishefsky, Proc. Natl. Acad. Sci. USA
1998, 95, 15798.
[4] L. Sepp-Lorenzino, A. Balog, D.-S. Su, D. Meng, N. Timaul, H. I.
Scher, S. J. , Danishefsky, N. Rosen, Prostate Cancer Prostatic
Dis. 1999, 2, 41.
[5] R. Christina, C. R. Harris, S. J. Danishefsky, J. Org. Chem. 1999,
64, 8434.
[6] K. Gerth, N. Bedorf, G. Höfle, H. Irschik, H. Reichenbach, J.
Antibiot. 1996, 49, 560.
[7] G. Höfle, N. Bedorf, H. Steinmetz, D. Schomburg, K. Gerth, H.
Reichenbach, Angew. Chem. 1996, 108, 1671; Angew. Chem. Int.
Ed. Engl. 1996, 35, 1567.
[8] G. Hoefle, N. Bedorf, K. Gerth, H. Reichenbach (GBF),
DE4138042C2, 1993.
Figure 4. Therapeutic effect of ZK-EPO in nude mice bearing human
cervix xenografts HeLa/MATU after intravenous injection. Changes in
tumor area over time are shown (top). Evaluation of body weight as a
parameter for drug-related toxicity during treatment with ZK-EPO.
During the treatment period the body weight is measured twice
weekly. The change in body weight over time is shown (bottom). The
red arrows indicate when ZK-EPO was administered.
[9] A. Balog, D. Meng, T. Kamenecka, P. Bertinato, D.-S. Su, E. J.
Sorensen, S. J. Danishefsky, Angew. Chem. 1996, 108, 2976;
Angew. Chem. Int. Ed. Engl. 1996, 35, 2801.
[10] K. C. Nicolaou, Y. He, D. Vourloumis, H. Vallberg, Z. Yang,
Angew. Chem. 1997, 109, 170; Angew. Chem. Int. Ed. Engl. 1997,
36, 166.
Figure 5. Tumor-inhibiting effect of ZK-EPO and PT in the human multidrug-resistant HeLa/MaTu/ADR cervix tumor (left) and A-375 melanoma
(right) xenograft models in nude mice after intravenous injection. The red arrows indicate when ZK-EPO was administered.
Angew. Chem. Int. Ed. 2006, 45, 7942 –7948ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 7947

Communications
[11] D. Schinzer, A. Limberg, A. Bauer, O. M. Böhm, M. Cordes,
Angew. Chem. 1997, 109, 543; Angew. Chem. Int. Ed. Engl. 1997,
36, 523.
[12] K. C. Nicolaou, S. Ninkovic, F. Sarabia, D. Vourloumis, Y. He, H.
Vallberg, M. R. V. Finlay, Z. Yang, J. Am. Chem. Soc. 1997, 119,
7974.
[13] D. Meng, P. Bertinato, A. Balog, D.-S. Su, T. Kamenecka, E. J.
Sorensen, S. J. Danishefsky, J. Am. Chem. Soc. 1997, 119, 10073.
[14] D. Schinzer, A. Bauer, J. Schieber, Chem. Eur. J. 1999, 5, 2492.
[15] E. J. Sorensen, S. J. Danishefsky, Y.-H. Zheng, T.-C. Chou, L. He,
S. B. Horwitz, J. Am. Chem. Soc. 1997, 119, 2733.
[16] R. B. Lichtner, A. Rotgeri, T. Bunte, B. Buchmann, J. Hoffmann,
W. Schwede, W. Skuballa, U. Klar, Proc. Natl. Acad. Sci. USA
2001, 98, 11743.
[17] In parallel, colleagues from Novartis had also replaced the side
chain at C15 with bicyclic heteroaromatic groups: K.-H.
Altmann, G. Bold, G. Caravatti, A. Flörsheimer, V. Guagnano,
M. Wartmann, Bioorg. Med. Chem. Lett. 2000, 10, 2765.
[18] Since the IUPAC nomenclature of epothilone Dis completely
different from epothilone B we prefer to use the numbering
depicted in Scheme 2 throughout the following discussions.
[19] “Synthesis and biological activity of epothilones”: U. Klar, W.
Skuballa, B. Buchmann, W. Schwede, T. Bunte, J. Hoffmann,
R. B. Lichtner in Anticancer Agents. Frontiers in Cancer Chemo-
therapy (Eds.: I. Ojima, G. D. Vite, K.-H. Altmann), Oxford
University Press, New York, 2001, pp. 131 – 147.
[20] Compounds that inhibit the proliferation of tumor cells by 50%
at concentrations below 100 nm are defined as active.
[21] Three other epothilones have been investigated in early clinical
studies (phase I): Kos-1584 (Kosan/Roche), ABJ879 (Novartis),
BMS310705 (BMS).
[22] J. M. G. Larkin, S. B. Kaye, Expert Opin. Invest. Drugs 2006, 15,
691 – 702.
[23] U. Klar, B. Röhr, F. Kuczynski, W. Schwede, M. Berger, W.
Skuballa, B. Buchmann, Synthesis 2005, 301.
[24] J. Wu, K. Mori, Agric. Biol. Chem. 1991, 55, 2667.
[25] G. Fouquet, M. Schlosser, Angew. Chem. 1974, 86, 50; Angew.
Chem. Int. Ed. Engl. 1974, 13, 82.
[26] Only stereochemically homogeneous material obtained from the
direct crystallization was used in the calculation. Additional
product can be obtained from the mother liquor.
[27] E. Watkins, A. G. Chittiboyina, J.-C. Jung, M. A. Avery, Curr.
Pharm. Des. 2005, 11, 1615.
[28] K.-H. Altmann, Org. Biomol. Chem. 2004, 2, 2137.
[29] R. Zhang, Z. Liu, Curr. Org. Chem. 2004, 8, 267.
[30] G. Höfle, N. Bedorf, H. Steinmetz, D. Schomburg, K. Gerth, H.
Reichenbach, Angew. Chem. 1996, 108, 1671; Angew. Chem. Int.
Ed. Engl. 1996, 35, 1567.
[31] R. E. Taylor, Y. Chen, A. Beatty, D. C. Myles, Y. Zhou, J. Am.
Chem. Soc. 2003, 125, 26.
[32] See supporting information.
[33] F. Y. F. Lee, R. Borzilleri, C. R. Fairchild, S.-H. Kim, B. H. Long,
C. Reventos-Suarez, G. D. Vite, W. C. Rose, R. A. Kramer, Clin.
Cancer Res. 2001, 7, 1429.
[34] R. M. Borzilleri, X. Zheng, R. J. Schmidt, J. A. Johnson, S.-H.
Kim, J. D. DiMarco, C. R. Fairchild, J. Z. Gougoutas, F. Y. Lee,
B. H. Long, G. D. Vite, J. Am. Chem. Soc. 2000, 122, 8890.
[35] R. M. Borzilleri G. D. Vite, Drugs Future 2002, 27, 1149.
[36] K.-H. Altmann, M. Wartmann, T. OꢀReilly, Biochim. Biophys.
Acta 2000, 1470, M79.
[37] J. Hoffmann, B. Buchmann, W. Schwede, W. Skuballa, R.B.
Lichtner, G. Siemeister, S. Hammer, K. Mittelstaedt, S. Winsel, J.
Eschenbrenner, C. Demarche, U. Klar, unpublished results.
[38] The first clinical data for ZK-EPO were presented at the 41st
Annual Meeting of the American Society of Clinical Oncology
(ASCO) in spring 2005.
7948 www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7942 –7948