Chemistry Reference
In-Depth Information
on the chemistry developed in our laboratory for these compounds,
126
the synthesis
of such additional analogs should be a straightforward undertaking.
1.4 PHARMACOPHORE MODELING AND
CONFORMATIONAL STUDIES
Our current understanding of the three-dimensional conformation of tubulin is
largely based on the structure of a tubulin/docetaxel complex within a two-dimen-
sional tubulin polymer sheet, which has been solved by electron crystallography at
3.7-
˚
resolution.
58
The availability of this information has significantly improved
our gross understanding of paclitaxel binding to b-tubulin, but the structure-based
design of epothilone analogs or mimics so far has been hampered by the lack of
high (atomic level)-resolution structural data either for tubulin or tubulin/microtu-
bule-epothilone complexes. However, recent studies on the tubulin-bound confor-
mation of Epo A either by NMR spectroscopy on a soluble b-tubulin/Epo A
complex
93
or by a combination of electron crystallography, NMR spectroscopic
conformational analysis, and molecular modeling of a complex between Epo A
and a Zn
2
þ
-stabilized two-dimensional a,b-tubulin sheet (solved at 2.89
˚
resolu-
tion)
127
have provided completely new insights into the bioactive conformation of
the epothilone-class of microtubule inhibitors (vide infra). Before the availability of
these data, various attempts have been described to develop a predictive pharmaco-
phore model for epothilones, which would be of substantial value for the design of
new analogs. Different approaches have been followed to address this problem,
which were generally based on the assumption of a common tubulin binding site
between epothilones and paclitaxel.
57,77-80
For example, the common paclitaxel/
epothilone pharmacophore model presented by Giannakakou et al.
57
is based on
an energy-refined model of the 3.7-
˚
density map of docetaxel bound to b-tubu-
lin.
58
According to this model, the position of the epoxide oxygen in epothilones
within the microtubule binding pocket corresponds with that of the oxetane oxygen
in paclitaxel, whereas the epothilone side chain is located in the same region as
either the C3
0
-phenyl group or, alternatively, the C2-benzoyloxy moiety of pacli-
taxel. The model also suggests that the methyl group attached to C12 in Epo B
is involved in hydrophobic interactions with the side chains of Leu-b273,
Leu-b215, Leu-b228, and Phe-b270, and that this may account for the higher acti-
vity of Epo B versus Epo A. Different conclusions with regard to the relative posi-
tioning of paclitaxel and epothilones within the microtubule binding site have been
reached by Wang et al.
78
In their model, the position of the thiazole moiety in
epothilones within the microtubule binding pocket matches the position of the phe-
nyl group of the C-3
0
-benzamido substituent in paclitaxel. Furthermore, the epoxide
oxygen is concluded not to be involved in interactions with the protein, which is in
line with the experimental data discussed here for cyclopropane-based epothilone
analogs. A model similar to that of Wang et al. has recently been proposed by
Manetti et al.
79
Although these computational models by their very nature are of
limited accuracy, some of them can reproduce at least part of the published
Search WWH ::
Custom Search