вход по аккаунту


Much AnticipatedЧThe Bioactive Conformation of Epothilone and Its Binding to Tubulin.

код для вставкиСкачать
Much Anticipated—The Bioactive Conformation of
Epothilone and Its Binding to Tubulin
Dirk W. Heinz,* Wolf-Dieter Schubert, and Gerhard Hfle*
conformation analysis · electron crystallography ·
macrocycles · molecular modeling · natural products
A mere ten years ago, Bollag et al.
demonstrated that cytotoxicity of epothilones is a result of their ability to
disrupt the intrinsic dynamics of microtubules. This recognition initiated intense chemical and biological analyses
of this previously neglected class of
substances, the momentum of which
continues to this day.[2] Apart from
Taxol, the epothilones were the first
natural products known to stabilize
microtubules, triggering apoptosis.[1, 3]
This mechanism, resulting in extensive
polymerization of ab-tubulin, is assumed to hold for the clinically highly
successful cytostatic drugs Taxol and
[*] Prof. Dr. D. W. Heinz, Dr. W.-D. Schubert
Division of Structural Biology
Gesellschaft fr Biotechnologische
Forschung (GBF)
38124 Braunschweig (Germany)
Fax: (+ 49) 531-618-763
Prof. Dr. G. Hfle
Department of Natural Product Chemistry
Gesellschaft fr Biotechnologische
Forschung (GBF)
38124 Braunschweig (Germany)
Fax: (+ 49) 531-618-1515
Taxotere and for epothilones, as well.[4]
Furthermore, their significantly higher
activity and their undiminished effectiveness against multiresistant tumor
cells[1, 3, 5] have raised the possibility that
epothilones could potentially replace
taxanes. A major advantage of epothilones is that essentially unlimited quantities can be produced from cultures of
the myxobacterium Sorangium cellulosum.[6] Several natural, semisynthetic,
and fully synthetic epothilones are currently undergoing clinical trials. Of
these, epothilone B lactam[7] (Ixabepilone) has been most successful (phase III).[8] There has been much speculation on the binding mode of epothilone,
but very little hard evidence had been
forthcoming. This situation changed
abruptly following the publication of
the electron-crystallographic structure
of the epothilone A/tubulin complex—
the centerpiece of this highlight.
The early observation that epothilones can displace tubulin-bound Taxol
suggested that both are bound by the
same or by overlapping binding sites.[1]
Structural similarities, such as epoxide
and oxetane rings, geminal dimethyl
groups, and flexible aromatic side
chains, inspired many a mind, and soon
a series of models had been proposed
for a common pharmacophore for both
groups of compounds. These speculations were based on the crystal structure
of epothilone B crystallized from dichloromethane,[10] known since 1995,
and the later structures of epothilones A
and B crystallized from methanol/water
(Figure 1 a).[11] Whereas the macrolide
moiety conformation is extraordinarily
similar in all theses structures, those of
the thiazole side chain are very different, presumably due to its high rota-
tional freedom. This was confirmed by
NMR spectroscopy in DMSO/water. In
addition, the preferred conformation of
the macrocycle in organic solvents was
found to correspond to that in the
crystal.[10] A refined analysis and the
molecular model developed by Taylor
and Zajicek indicated a secondary conformation for the C3–C9 ring segment,
in which the 7-OH and 8-methyl groups
are axial and the 3-OH group assumes
an equatorial position.[12]
In the search for a common pharmacophore by modeling experiments, the
epothilone thiazolyl side chain was
DOI: 10.1002/anie.200462241
Angew. Chem. Int. Ed. 2005, 44, 1298 –1301
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 1. Three-dimensional structures of epothilones in different physical environments:
a) Molecular structure of epothilone B from
dichloromethane/petroleum ether determined
by X-ray crystallography[10] (pink), superimposed upon that of epothilone A from methanol/water[11] (blue); b) structure of epothilone A in the presence of tubulin in aqueous
medium determined by NMR spectroscopy;[22]
c) conformation of epothilone B in a complex
with cytochrome P450epoK;[26] d) conformation of epothilone A in a complex with ab-tubulin in Zn2+-stabilized layers.[9] This Figure
was produced using Pymol (http://pymol.
mostly associated with one of three
phenyl or even the 10-acetyl side chain
of Taxol.[13–16] Once it emerged that the
C13 side chain of Taxol is not strictly
required for recognition by tubulin, the
models were narrowed down to those
sharing homologous thiazole and 2-benzoyl fragments, as well as the taxane
framework and macrolide ring.[17] Conclusive evidence for common binding
sites in tubulin was provided by crossresistant tubulin mutants, especially
Ala231.[18–20] Nevertheless, an unequivocal model of epothilone bound to
tubulin remained elusive, despite the
publication of the electron-crystallographic structure of the Taxol/tubulin
In an alternative approach, Carlomagno et al. set out to determine the
conformation of tubulin-bound epothilone by NMR spectroscopy.[22] Under
the assumption of a rapid exchange of
free and bound epothilone A, transfer
nuclear Overhauser effects (transferNOEs) and torsion angles were determined by a transfer cross-correlated
relaxation experiment. The results were
used for a force-field calculation to
determine the epothilone A conformation. Overall, the derived conformation
of the macrolide ring in this tubulinassociated epothilone is similar to the
crystal structures of free epothilone.
Only the C2–C3 segment deviates significantly; the 3-OH group is placed in
an equatorial orientation, as proposed
by Taylor et al. and observed in the
crystal structure of epothilone A Noxide.[12, 23] In the side chain, 19-H and
16-Me adopt a syn orientation analogous to that in the crystal structure
obtained from methanol/water (Figure 1). This is also in agreement with
the structure–activity investigations of
pyridine and benzimidazole analogues.
In a control experiment the authors
demonstrate that the observed NMR
effects can indeed be ascribed to the
protein-associated epothilone A. They
found that epothilone B, due to its
twofold higher affinity for tubulin, permanently displaces epothilone A from
the binding sites, canceling the observed
effects. Nevertheless, the small difference in binding affinity does not explain
how small amounts of epothilone B in
solution could induce such strong reAngew. Chem. Int. Ed. 2005, 44, 1298 –1301
sponses. Furthermore it remains unclear
why a hundredfold excess of epothilone A does not in itself lead to the
polymerization of the tubulin.[24] Yet
despite these uncertainties, the conformation described appears to be induced
by the presence of tubulin, and this is
clearly different from those of free
epothilone A in solution.
Crystal structures of epothilone B
and D in complex with cytochrome
P450epoK, which catalyzes the last step
in the biosynthesis of epothilones A and
B, have been refined at 1.9 and 2.7 ,
respectively.[25] Both epothilone B (Figure 1 c) and D are found to adopt
conformations quite distinct from those
discussed before. Apart from the flexible thiazolyl side chain, the conformations bear some resemblance to the
“minor conformer B” proposed by Taylor and Zajicek.[12a] Nevertheless, these
epothilone conformations could not be
modeled into the Taxol-binding pocket
of tubulin with confidence.
With their recent publication of the
electron-crystallographic structure of
the epothilone A/tubulin complex, Nettles et al. were able to solve the puzzle of
the bioactive conformation of epothilone A.[9] As reported for the electroncrystallographic structure determination
of the Taxol/ab-tubulin complex, twodimensional crystals of tubulin heterodimers were obtained by adding Zn2+ in
the presence of an excess of epothilone A. The protofilaments are oriented
antiparallel in the Zn2+-induced tubulin
layers, rather than parallel as in true
microtubules. The 2D crystals diffract
electrons to a resolution of 2.9 parallel to the crystal plane and 4.2 perpendicular to it. Due to the limited
resolution, the conformation of the
epothilone A was derived with the help
of NMR-generated structure ensembles
and molecular dynamics modeling procedures. The structure of the complex is
an important breakthrough as it allows
the interactions of epothilone A with btubulin to be described in detail for the
first time. At a stroke, it also provides
answers to many previously vexing
questions: How can known resistance
mutants in b-tubulin be explained? How
do structural modifications of epothilone affect the interaction with b-tubulin? How does the binding of epothilone
compare to that of Taxol?
Figure 2. The Taxol/epothilone binding cavity
in b-tubulin from the corresponding electroncrystallographically obtained structures with
bound a) Taxol and b) epothilone. For b-tubulin the surface of the molecule is depicted (C:
gray, O: dark pink, N: gray-blue). Taxol (C:
green, O: red, N: blue) and epothilone A (C:
violet, O: red, N: blue) are represented as
stick models. This Figure was produced using
Pymol (
As expected, epothilone binds within a cavity on the surface of b-tubulin
(Figure 2 b). This binding site, overlapping with that of Taxol, is partly constituted by the so-called M-loop, which
is also essential for lateral contacts
between protofilaments in the Zn2+induced tubulin layers and in microtubules.[21a] Stabilization of the M-loop,
and hence of lateral contacts between
protofilaments, ultimately stabilizes
growing microtubules, preventing their
Through modeling studies, Nettles et al.
were able to elucidate the influence of
known epothilone modifications on the
interaction with b-tubulin. For example,
an unexpected endo orientation of the
epoxide places the large C12 substituent
(including N-acyl groups of aziridine
analogues) into the extensive hydrophobic indentation below the epothilone
macrolide ring, explaining why larger
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
substituents at C12 and a 9,10-trans
double bond in epothilone can be tolerated. Because of this hydrophobic
indentation, however, the model does
not explain why epothilones of the B
series, which have a methyl substituent
at C12, are generally more active than
those of the A series. It is also not
obvious how the high-energy syn-pentane orientation of the C6 and C8
methyl groups comes about and why
removal of the C8 methyl group leads to
a reduction in activity by a factor of
200.[26] Furthermore the influence of
additional methyl groups at C10 and
C14 is difficult to follow, in particular as
the accepted configuration of the biologically active C14-methyl analogue[27]
is depicted in Figure 4 B of ref. [9] incorrectly. The structure of the complex
is in agreement with some, but not all, of
the known resistance mutation in btubulin. As expected, however, mostly
amino acids that interact directly or
indirectly with epothilone are affected.
Direct comparison of epothilone
binding with Taxol binding is particularly interesting (Figure 2). Not surprisingly, the smaller epothilone molecule
fills only about half of the Taxol-binding
cavity in b-tubulin. One striking aspect
of this is that the binding cavity adjusts
to the substrate by reorienting amino
acid side chains, optimizing the interactions with epothilone and Taxol (“induced fit”). The Taxol/b-tubulin interactions differ fundamentally from those
in the epothilone A/tubulin complex.
The hypotheses concerning a common
pharmacophore can thus finally be laid
to rest. With the exception of the hydrogen bond from 7-OH to Arg282, there
are no interactions of structurally homologous groups in Taxol and epothilone with b-tubulin. Furthermore, the
amino acids in the binding cavity are
involved in entirely unrelated interactions with the two natural products. The
adaptable binding cavity in b-tubulin is
known to also bind further cytostatic
effectors other than taxanes and epothilones, including Eleutherobin/Sarcodictyn, that bear little structural resemblance either to each other or to taxanes
and epothilones.
How does the conformation of epothilone A observed in the presence of
tubulin (Figure 1 b) as determined by
Carlomagno et al. by NMR spectros-
copy fit into this picture? The two
conformations are clearly profoundly
different such that an intermediate
structure that fits both forms at least to
some extent is not readily conceivable.
Instead, it is more likely that each of the
two conformations exists in the specific
environment in which it was studied.
Ultimately, both the Zn2+-stabilized tubulin layers and the undefined tubulin
present in the NMR experiments are
artificial systems. The biologically relevant conformation and binding site of
epothilones could thus still differ appreciably from both of these proposed
What experiments could dispel any
remaining doubts? Due to the intermediate resolution of the epothilone/
tubulin structure, the proposed conformation and orientation of epothilone in
the tubulin-binding cavity must be corroborated by independent methods. An
obvious approach would be photoaffinity labeling with epothilone derivatives
having photolabile substituents at the
methyl group of the thiazolyl fragment
or, in the case of aziridine analogues, on
the aziridine nitrogen atom. Both these
positions can be modified without affecting the binding to tubulin.[28, 29] Also
conceivable is a solid-state NMR spectroscopic investigation (MAS-NMR) of
microtubules loaded with 13C- and 19Flabeled epothilones. This would allow
the conformation of the bound molecule
to be determined as well as individual
distance measurements (REDORNMR).
Clearly, attempts will be undertaken
to develop novel epothilones that not
only support the postulated structure
but also have a higher affinity for
tubulin and thus possibly exhibit a more
pronounced cytotoxicity. Epothilones
developed along these lines will, however, presumably not be of therapeutic
value, as a clinical application would
instead require the pharmacological
properties to be optimized to achieve
higher selectivity and lower sensitivity
toward the development of resistance.
This would result in drugs with an
improved efficiency and fewer side effects.
[1] D. M. Bollag, P. A. McQueney, J. Z.
Zhu, O. Hensens, L. Koupal, J. Liesch,
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
M. Goetz, E. Lazarides, C. M. Woods,
Cancer Res. 1995, 55, 2325 – 2333.
a) K. C. Nicolaou, F. Roschangar, D.
Vourloumis, Angew. Chem. 1998, 110,
2120 – 2153; Angew. Chem. Int. Ed.
1998, 37, 2014 – 2045; b) R. M. Borzilleri, G. D. Vite, Drugs Future 2002, 27,
1149 – 1163; c) K.-H. Altmann, Med.
Chem. 2003, 3, 149 – 158.
R. J. Kowalski, P. Giannakakou, E. Hamel, J. Biol. Chem. 1997, 272, 2534 –
a) P. B. Schiff, J. Fant, S. B. Horwitz,
Nature 1979, 277, 665 – 667; b) P. B.
Schiff, S. B. Horwitz, Biochemistry
1981, 20, 3247 – 3252.
A. Wolff, A. Technau, G. Brandner, Int.
J. Oncol. 1997, 11, 123 – 126.
a) K. Gerth, N. Bedorf, G. Hfle, H.
Irschik, H. Reichenbach, J. Antibiot.
1996, 49, 560 – 563; b) K. Gerth, S.
Pradella, O. Perlova, S. Beyer, R. Mller, J. Biotechnol. 2003, 106, 233 – 253.
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. F. Lee, B. H. Long, G. D. Vite,
J. Am. Chem. Soc. 2000, 122, 8890 –
S. Goodin, M. P. Kane, E. H. Rubin, J.
Clin. Oncol. 2004, 22, 2015 – 2025.
J. H. Nettles, H. Li, B. Cornett, J. M.
Krahn, J. P. Snyder, K. H. Downing,
Science 2004, 305, 866 – 869.
G. Hfle, N. Bedorf, H. Steinmetz, D.
Schomburg, K. Gerth, H. Reichenbach,
Angew. Chem. 1996, 108, 1671 – 1673;
Angew. Chem. Int. Ed. Engl. 1996, 35,
1567 – 1569. The crystal structure data
have been available to interested research groups since October 1995.
H.-J. Hecht, G. Hfle, unpublished results; CCDC 241333 and CCDC 241334
contain the supplementary crystallographic data for this paper. These data
can be obtained free of charge via
html (or from the Cambridge Crystallographic Data Centre, 12, Union Road,
Cambridge CB2 1EZ, UK; fax: (+
44) 1223-336-033; or
a) R. E. Taylor, J. Zajicek, J. Org. Chem.
1999, 64, 7224 – 7228; b) R. E. Taylor, Y.
Chen, G. M. Galvin, P. K. Pabba, Org.
Biomol. Chem. 2004, 2, 127 – 132.
J. D. Winkler, P. H. Axelsen, Bioorg.
Med. Chem. Lett. 1996, 6, 2963 – 2966.
I. Ojima, S. Chakravarty, T. Inoue, S.
Lin, L. He, S. B. Horwitz, S. D. Kuduk,
S. J. Danishefsky, Proc. Natl. Acad. Sci.
USA 1999, 96, 4256 – 4261.
M. Wang, X. Xia, Y. Kim, D. Hwang,
J. M. Jansen, M. Botta, D. C. Liotta, J. P.
Snyder, Org. Lett. 1999, 1, 43 – 46.
Angew. Chem. Int. Ed. 2005, 44, 1298 –1301
[16] F. Manetti, S. Forli, L. Maccari, F.
Corelli, M. Botta, Il Farmaco 2003, 58,
357 – 361.
[17] L. He, P. G. Jagtap, D. G. I. Kingston,
H.-J. Shen, G. A. Orr, S. B. Horwitz,
Biochemistry 2000, 39, 3972 – 3978.
[18] P. Giannakakou, R. Gussio, E. Nogales,
K. H. Downing, D. Zaharevitz, B. Bollbuck, G. Poy, D. Sackett, K. C. Nicolaou,
T. Fojo, Proc. Natl. Acad. Sci. USA 2000,
97, 2904 – 2909.
[19] N. M. Verrills, C. L. Flemming, M. Liu,
M. T. Ivery, G. S. Cobon, M. D. Norris,
M. Haber, M. Kavallaris, Chem. Biol.
2003, 10, 597 – 607.
[20] L. He, C.-P. H. Yang, S. B. Horwitz, Mol.
Cancer Ther. 2001, 1, 3 – 10.
[21] a) E. Nogales, S. G. Wolf, K. H. Downing, Nature 1998, 391, 199 – 203; b) J.
Lwe, H. Li, K. H. Downing, E. No-
Angew. Chem. Int. Ed. 2005, 44, 1298 –1301
gales, J. Mol. Biol. 2001, 313, 1045 –
1057; c) J. P. Snyder, J. H. Nettles, B.
Cornett, K. H. Downing, E. Nogales,
Proc. Natl. Acad. Sci. USA 2001, 98,
5312 – 5316.
[22] T. Carlomagno, M. J. J. Blommers, J.
Meiler, W. Jahnke, T. Schupp, F. Petersen, D. Schinzer, K.-H. Altmann, C.
Griesinger, Angew. Chem. 2003, 115,
2615 – 2619; Angew. Chem. Int. Ed.
2003, 42, 2511 – 2515.
[23] G. Hfle, N. Glaser, M. Kiffe, H.-J.
Hecht, F. Sasse, H. Reichenbach, Angew. Chem. 1999, 111, 2090 – 2093; Angew. Chem. Int. Ed. 1999, 38, 1971 –
[24] According to HPLC analysis, a solution
in D2O, saturated at room temperature,
contains 0.5 mg mL 1 epothilone A and
0.06 mg mL 1 epothilone B; G. Hfle,
unpublished results.
S. Nagano, H. Li, H. Shimizu, C. Nishida, H. Ogura, P. R. Ortiz de Montellano,
T. L. Poulos, J. Biol. Chem. 2003, 278,
44 886 – 44 893.
A. Balog, P. Bertinato, D.-S. Su, D.
Meng, E. Sorensen, S. J. Danishefsky,
Y.-H. Zheng, T.-C. Chou, L. He, S. B.
Horwitz, Tetrahedron Lett. 1997, 38,
4529 – 4532.
R. E. Taylor, Y. Chen, A. Beatty, D. C.
Myles, Y. Zhou, J. Am. Chem. Soc. 2003,
125, 26 – 27.
N. Glaser, PhD Thesis, Technische Universitt Braunschweig, 2001.
A. Regueiro-Ren, R. M. Borzilleri, X.
Zheng, S.-H. Kim, J. A. Johnson, C. R.
Fairchild, F. Y. F. Lee, B. H. Long, G. D.
Vite, Org. Lett. 2001, 3, 2693 – 2696.
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Без категории
Размер файла
819 Кб
conformational, much, tubulin, anticipatedчthe, binding, bioactive, epothilone
Пожаловаться на содержимое документа