Synthesis and NMR-driven conformational analysis of Taxol analogues conformationally constrained on the C13 side chain

Article abstract of DOI:10.1021/jm001103v

Analogues of Taxol (paclitaxel) with the side chain conformationally restricted by insertion of a carbon linker between the 2′-carbon and the ortho-position of the 3′-phenyl ring were synthesized. Biological evaluation of these new taxoids showed that act

Full text of DOI:10.1021/jm001103v

1576
J . Med. Chem. 2001, 44, 1576-1587
Syn th esis a n d NMR-Dr iven Con for m a tion a l An a lysis of Ta xol An a logu es
Con for m a tion a lly Con str a in ed on th e C13 Sid e Ch a in
Luciano Barboni* and Catia Lambertucci
Dipartimento di Scienze Chimiche, Universita` di Camerino Via S. Agostino 1, I-62032 Camerino (MC), Italy
Giovanni Appendino*
DiSCAFF, Universita` del Piemonte Orientale Viale Ferrucci 33, 28100 Novara, Italy
David G. Vander Velde† and Richard H. Himes‡
Departments of Medicinal Chemistry and Molecular Biosciences, University of Kansas, Lawrence, Kansas 66045
Ezio Bombardelli
Indena S.p.A., Viale Ortles 12, 20139 Milano, Italy
Minmin Wang and J ames P. Snyder*
Department of Chemistry, Emory University, Atlanta, Georgia 30322
Received October 24, 2000
Analogues of Taxol (paclitaxel) with the side chain conformationally restricted by insertion of
a carbon linker between the 2′-carbon and the ortho-position of the 3′-phenyl ring were
synthesized. Biological evaluation of these new taxoids showed that activity was dependent
on the length of the linker and the configuration at C2′ and C3′. Two analogues in the homo
series, 9a and 24a , showed tubulin binding and cytotoxicity comparable to that of Taxol.
NAMFIS (NMR analysis of molecular flexibility in solution) deconvolution of the averaged 2-D
NMR spectra for 9a yields seven conformations. Within the latter set, the hydrophobically
collapsed “nonpolar” and “polar” classes are represented by one conformation each with predicted
populations of 12-15%. The five remaining conformers, however, are extended, two of which
correspond to the T-conformation (47% of the total population). The latter superimpose well
with the recently proposed T-Taxol binding conformer in â-tubulin. The results provide evidence
for the existence of two previously unrecognized structural features that support Taxol-like
activity: (1) a reduced torsion angle between C2′ and C3′ and (2) an orthogonal arrangement
of the mean plane through C1′, C2′ and the 2′-hydroxyl and the 3′-phenyl plane, the latter
ring bisected by the former plane. By contrast, epimerization at 2′,3′ and homologation of the
tether to CH2-CH2 were both detrimental for activity. The decreased activity of these analogues
is apparently due to configurational and steric factors, respectively.
In tr od u ction
(docetaxel, 1b),3 a clinically useful compound with a
better solubility and a higher in vitro potency compared
to Taxol. More recent investigations have focused on
analogues active on MDR (multi-drug-resistant) cells,
easier to administer, and endowed with a better profile
of side effects4 as exemplified by the orally active taxoid
IDN 5109 (2).5
The discovery of Taxol-mimicking compounds which
structurally deviate from the taxoid platform and
superficially do not resemble each other has proven of
great significance in the field,6 suggesting the existence
of a shared structural core among the various tubulin
stabilizers. The structural elements underlying the
occupation of an identical binding site are not obvious,
and the pharmacophore models which have emerged
from the vast body of information available are con-
trasting and focus on different functionalities.7 The
problem is complicated by the fact that an endogenous
ligand has not yet been identified, and therefore, it is
not known what endogenous ligand, if any, the various
The discovery that Taxol (paclitaxel, 1a ) binds to
tubulin promoting its aggregation in microtubules and
the disclosure of its potential clinical activity in various
forms of cancer have put this complex diterpenoid at
the center of intense multidisciplinary attention and
extraordinary media attention.1 Within the chemical
arena, the search for improved analogues has generated
an impressive body of structure-activity data (SAR).2
The amino acid side chain and, to a lesser extent, also
the terpenoid core of the natural product turned out to
be amenable to modification for the synthesis of ana-
logues. Early studies led to the discovery of Taxotere
* To whom correspondence should be addressed. Luciano Barboni:
phone, +39 0737 402240; fax, +39 0737 637345; e-mail, barboni@
camserv.unicam.it. Giovanni Appendino: phone, +39 011 670 7684;
fax, +39 011 670 7687; e-mail, appendin@pharm.unito.it. J ames P.
Snyder: phone, 404-727-2415; fax, 404-727-6586; e-mail, snyder@
euch4e.chem.emory.edum.
† Department of Medicinal Chemistry, University of Kansas.
‡ Department of Molecular Biosciences, University of Kansas.
10.1021/jm001103v CCC: $20.00 © 2001 American Chemical Society
Published on Web 04/06/2001

Taxol Analogues Conformationally Constrained at C13
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 10 1577
Sch em e 1a
tubulin stabilizers are supposed to mimic. The discovery
of a second receptor for Taxol8 has added a further layer
of complexity to our understanding of the structural
requirements for cytotoxic activity, since little is known
about the interaction of the biological analogues of Taxol
with this receptor. A further complicating factor is the
poor correlation between tubulin binding and cytotox-
icity that has sometimes been observed for taxoids.9
At the core of the ambiguity in the definition of a
taxoid pharmacophore is the issue of the active confor-
mation of the natural product, and a significant amount
of research has focused on the alleviation of this
problem. The baccatin core of Taxol is conformationally
rigid, but the side chains are not, and the presence of
at least seven easily rotated single bonds within these
moieties translates into a palette of three basic types
of conformations: two globular and an extended one.10
The two globular conformers are characterized by a
clustering of the lipophilic groups on the amino acid side
chain and the terpenoid core and are often referred to
as the nonpolar and the polar forms, since first detected
in CDCl3 and water-DMSO, respectively.10c The im-
portance of the extended conformational type devoid of
intramolecular hydrophobic interactions was recognized
only recently in the deconvolution of the averaged NMR
spectra of Taxol in chloroform10h and at the â-tubulin
binding site.10i,11
a
(a) BnzOCONH2, NaOH, t-BuOCl, (DHQ)2PHAL, K2OsO2(OH)4,
n-PrOH, H2O (45% for 4a , 33% for 4b); (b) i. HCOONH4, Pd(C),
AcOH, ii. PhCOCl, NaHCO3, CH2Cl2-H2O (68% for 5a , 70% for
5b); (c) 2,4-dimethoxybenzaldehyde dimethylacetal, PPTS, toluene,
4; (d) K2CO3, MeOH-H2O; (e) i. DCC, DMAP, toluene, rt, ii. 0.1
M HCl in MeOH (29% from 5a for 9a , 24% from 5b for 9b).
precise pharmacophore model for the interaction of
anticancer taxoids and tubulin.
Ch em istr y
In an ingenious approach to the taxane pharmacoph-
ore, Ojima reported analogues where the amino acid
chain and the lipophilic ester group at C2 are merged
into a macrocyclic diester.7c,12 The activity of some of
these compounds is a highly significant finding, but its
translation into conformational terms is complicated by
the flexibility of the tether. To simplify the recognition
of the still elusive active conformation of anticancer
taxoids, we planned the synthesis of constrained ana-
logues where rotation around the bonds C2′-C3′ and
C3′-phenyl, both pivotal event in the dynamic averag-
ing observed in solution, is blocked or reduced by a short
covalent tether between the 2′-carbon and the ortho-
position of the 3′-phenyl ring.13 These compounds retain
the rich functionality of Taxol but only part of its
conformational complexity. Since the introduction of a
methyl on the 2′-position of Taxol and Taxotere is well-
tolerated in terms of activity,14 the biological evaluation
of these tethered analogues should help define a more
Several methods have been developed to convert
cinnamic acid derivatives to the side chain of anticancer
taxoids,15 and the tethered cinnamates 3a ,b were
identified as logical precursors of the side chains needed
for the synthesis of the target taxoids. The asymmetric
aminohydroxylation (AA)16 of both esters (Scheme 1)
occurred with remarkable regioselectivity but gave
essentially racemic amino alcohols, as evidenced by an
HPLC analysis of the corresponding N-decarbamoyl-N-
menthylcarbonates17 (ee ca. 0 for 4a and <10 for 4b).
Reductive decarbamoylation of the amino alcohols 4a ,b,
followed by remodeling of the substitution pattern of the
heteroatoms via N-acylation and aminal formation with
2,4-dimethoxybenzaldehyde dimethylacetal,18 eventu-
ally gave the oxazolidines 6a ,b. Saponification to the
corresponding acids 7a ,b and coupling with 7-TES-
()triethylsilyl)-baccatin III (8) using the DCC-DMAP
protocol completed the synthesis. Kinetic resolution took

1578 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 10
Barboni et al.
Sch em e 2a
Sch em e 3a
a
(a) ADmix-â, t-BuOH-H2O 1:1 (63%); (b) i. trimethyl orthoben-
zoate, PTSA, ii. AcBr (70%); (c) i. NaN3, DMF, 45 °C, ii. H2, Pd(C),
EtOH (overall 25%); (d) 2,4-dimethoxybenzaldehyde dimethylac-
etal, PPTS, toluene; (e) i. 8, DCC, DMAP, toluene, 70 °C, ii. 0.1 M
HCl in MeOH (11% from 18).
for the synthesis of the side chain of the homo-taxane
9a . To this purpose, 10a was sequentially treated with
phenyl orthoformate and acetyl bromide, cleanly afford-
ing the O-benzoylated bromide 11, next treated with
NaN3 in DMF. Azide to amine reduction with concomi-
tant oxygen to nitrogen acyl rearrangement, followed
by oxazolidine formation and saponification, eventually
afforded the scalemic side chain precursor 15a . Coupling
with 7-TES-baccatin III (8) and deprotection gave a
compound identical to the one obtained from the racemic
acid 7a . Unexpectedly, the orthoester-mediated double-
inversion protocol21 failed with diol 10b, but the re-
quired bromide could be obtained via a sulfate-mediated
double-inversion strategy.22 Thus, treatment of 10b with
thionyl chloride gave a sulfite, which was oxidized to
the corresponding sulfate and nucleophilically opened
with tetrabutylammonium bromide to afford the trans-
bromohydrin 13. The synthesis was completed unevent-
fully as described for 11, providing the scalemic acid 14.
Also in this case, protection as a 2,4-dimethoxybenzyl
acetal and coupling with 7-TES-baccatin III (8) afforded,
after deprotection, a compound identical to the one
obtained from the racemic acid.
With the configurational assignment of 9a ,b secured,
we went on to prepare the corresponding Taxotere
analogues and modify the active (see infra) 2′,2′′-
methylene analogue 9a by changes in the configuration
at C2′, C3′. The 2′S,3′R-acid 19 was prepared using the
same sequence used for the synthesis of its enantiomer
9a , but employing the ADmix-â for the AD step (Scheme
3). The observation that the acid 19 and 7-TES-baccatin
III (8) were essentially unreactive under standard
coupling conditions confirmed that they constitute, in
stereochemical term, a mismatched pair. Nevertheless,
under forcing conditions, coupling could be achieved,
albeit in poor yield, affording the taxoid 20.
a
(a) ADmix-R, t-BuOH-H2O 1:1 (65% for10a , 77% for 10b); (b)
i. trimethyl orthobenzoate, PTSA, ii. AcBr (72%); (c) i. NaN3, DMF,
45 °C, ii. H2, Pd/C, EtOH (overall 21%); (d) 2,4-dimethoxybenzal-
dehyde dimethylacetal, PPTS, toluene, ∆; (e) i. SOCl2, pyridine,
CH2Cl2, ii. NaIO4, RuCl3, MeCN, iii. (n-Bu)4NBr, Me2CO-H2O,
H2SO4 (20% from 10b); (f) i. NaN3, DMF, ii. H2, Pd/C, EtOAc, iii.
BzCl, NaHCO3, CH2Cl2-H2O (overall 87%); (g) i. 8, DCC, DMAP,
toluene, rt, ii. 0.1 M HCl in MeOH (35% from 12 for 9a , 29% from
14 for 9b).
place in the esterification of the 13-hydroxyl affording,
after deprotection, the diastereomerically pure taxoids
9a ,b. Kinetic resolution in the esterification of the 13-
hydroxyl of protected baccatins has been reported with
both acids and â-lactams, affording taxoids with the
natural 2R,3S configuration.19 However, unambiguous
evidence for this configuration could not be obtained by
NMR analysis of 9a ,b. Since the activity of 2′-substi-
tuted taxoids is critically dependent on the configuration
at C2′ and C3′,14a and all reported cases of kinetic
resolution were observed with acylating reagents lack-
ing an R-alkyl substituent,19 the configurational assign-
ment of 9a ,b was secured by an independent asymmet-
ric synthesis.
To this purpose, we capitalized on an asymmetric
dihydroxylation strategy, since the vast body of data
available on the reaction of substituted cinnamates
provides a straightforward configurational assign-
ment.20 Thus, treatment of 3a ,b with the ADmix-R
afforded the diols 10a ,b (Scheme 2) with satisfactory
enantioselectivity (ca. 92% ee, based on HPLC analysis
of the diastereomeric menthylcarbonates17). A straight-
forward oxygen-to-nitrogen substitution via an orthoe-
ster-mediated double inversion protocol21 was employed
The N-debenzoyl-N-Boc()tert-butoxycarbonyl) acids
22a ,b were prepared using the sequence employed for
the synthesis of their racemic N-benzoyl analogues 7a ,b
but replacing benzoyl chloride with Boc-anhydride in
the acylation step (Scheme 4). Once more, complete
kinetic resolution was observed in the coupling with 7,-

Taxol Analogues Conformationally Constrained at C13
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 10 1579
Sch em e 4a
Ta ble 2. Mole Fractions (MF) and Selected Geometric
Features for the Seven MM3* Optimized Conformations of 9a
Derived from NAMFIS Analysis in DMSO-d6 Compared with
Taxol and Taxotere Conformers.
selected dihedral angles (deg)
compd MF (%) X-2′-3′-Pha 1′-2′-3′-Ph Od1′-2′-OH
9a 1
9a 2
9a 3
9a 4
9a 5
9a 6
9a 7
25a c
25bd
26e
26
21
15b
15b
12
10
2
-37
33
-35
-35
34
-158
-85
36
27
-156
-157
-83
38
37
57
177
36
41
93
-2
14
-37
-34
66
-155
-157
-58
60
-65
-61
-179
-170
27f
-53
a
b
For 9a 1-9a 7, X ) CH2; for 25-27, X ) H. Differences in
other dihedral angles distinguish these conformations; e.g. C12-
C13-O(ether)-C1′ and O(ether)-C1′-C2′-C3′. c The “polar” form
of Taxol, ref 23. The extended form of Taxol, ref 23. e Taxotere,
d
ref 22. f T-Taxol, ref 10i.
Taxol is critically dependent on the length of the tether
and the configuration at C2′ and C3′. The excellent
antitumor activity and tubulin affinity of 2′-methyl
taxoids were fundamentally maintained in the methyl-
ene-linked compounds 9a and 24a , showing that tether-
ing of the extra C2′ methyl to the ortho-carbon of the
3′-phenyl is compatible with tubulin binding and cyto-
toxicity. A conformational hallmark of these compounds
is the perpendicular location of the mean plane through
C1′, C2′ and the 2′-hydroxyl and the 3′-phenyl plane,
the latter ring bisected by the former plane. This is an
unusual arrangement not found in other taxoids. In the
solid state, the same elements of Taxotere,23 the two
conformations of Taxol,24 and the amino acid side chain
of Taxol25 adopt alternative orientations. Furthermore,
the modest puckering of cyclopentene derivatives means
that in 9a and 24a the CH2(corresponding to H2′ of
Taxol)-C2′-C3′-Ph torsion angle has values restricted
to around +30 and -30° (Table 2). This is at the border
between the gauche and eclipsed domains and signifi-
cantly lower than those detected in the X-ray analysis
of conformationally unconstrained anticancer taxoids
(Taxol 60° and 66° (polar),24 Taxotere -61° 23). The
finding that these changes are well-tolerated in terms
of activity suggests that there may be a taxoid family
of conformers supporting a narrow window of H-C2′-
C3′-Ph dihedral angles that are ideal for tubulin
binding.
a
(a) i. HCOONH4, Pd(C), AcOH, ii. (Boc)2O, NaHCO3, CH2Cl2-
H2O (68% for 21a , 66% for 21b); (b) i. 2,4-dimethoxybenzaldehyde
dimethylacetal, PPTS, toluene, 4 (80%), ii. K2CO3, MeOH-H2O
(80% from 21a for 22a , 78% from 21b for 22b); (c) i. DCC, DMAP,
toluene, rt, ii. 0.1 M HCl in MeOH, iii. Zn, AcOH, MeOH (17% for
24a , 4% for 24b).
Ta ble 1. Biological Evaluation of the 2′,2′′-Conformationally
Constrained Taxoidsa
tubulin binding
cytotoxicity, MCF7 cells
compd
(ED50/ED50(Taxol)
)
(ED50/ED50(Taxol))
9a
4
9
333
1604
2
9b
42
20
24a
24b
inactive
7
30
239
a
The ED50 of Taxol in the tubulin assembly assay was 0.2 to
0.4 µM. The ED50 of Taxol in the cytotoxicity assay was 2 to 4
nM.
10-di-Troc()trichloroethoxycarbonyl)-10-deacetyl-bacca-
tin III (23), affording the diastereomerically pure Tax-
otere analogues 24a ,b.
Resu lts a n d Discu ssion
Biologica l Eva lu a tion . The series of five 2′,2′′-
alkylidene constrained taxoids (9a ,b, 20, 24a ,b) was
tested for tubulin binding and cytotoxicity. Taxol was
used as a reference compound for both assays, and the
results are presented in Table 1. Introduction of a
methylene bridge between the 2′-carbon and the ortho-
position of the 3′-phenyl gave analogues of Taxol and
Taxotere retaining strong tubulin binding and cytotox-
icity (9a and 24a , respectively). Conversely, replacement
of the methylene link with an ethylidene, as in 9b and
24b, led to a dramatic decrease of activity in both
assays, while epimerization of 9a at C2′, C3′ afforded a
practically inactive compound (20). A good correlation
between tubulin binding and cytotoxicity was observed
for all compounds.
The dramatic loss of activity observed upon isomer-
ization of 9a at C2′ and C3′ closely parallels the results
observed in 2′-methyl substituted taxoids and is in sharp
contrast to what is observed in Taxol and Taxotere,
whose 2′S,3′R diastereomers retain a noteworthy activ-
ity in both cytotoxicity and binding assays.26 Apparently,
the presence of an acyclic alkyl substituent at C2′
renders the compounds unable to compensate in con-
formational terms for diastereomeric changes in the side
chain. Similarly, the sharp decrease of activity observed
in the ethylidene-linked compounds 9b and 24b is
striking. Ring puckering is expected to increase from
indans to their corresponding tetrahydronaphthalene
derivatives, translating into higher conformational bar-
riers for 9b and 24b.
Con for m a t ion a l Con sid er a t ion s. The activity of
the C2′, C2′′ conformationally constrained analogues of
Steric factors might also play a role, preventing C2′
substituents larger than a methyl group from being

1580 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 10
Barboni et al.
F igu r e 1. Lowest energy conformations of 9a ,b as derived by NAMFIS analysis with the diterpenoid cores depicted in a similar
orientation (a, 9a 1; b, 9b1). The dotted lines indicate prominent NOE cross-peaks between C3′-H and the cross-cavity protons of
the C4-OAc methyl and C14-H.
accommodated in the binding groove of tubulin which
recognizes the side chain of taxoids. Docking of 9a ,b into
refined models of â-tubulin is supportive of this postu-
late.27
matic ring consistent with an upfield π-electron-induced
chemical shift of the acetate methyl and pronounced
NOEs between these groups.30 Relative to Taxol, the
4-acetyl signal in 9a is shifted upfield δ 0.51 in CDCl3
and δ 0.38 in DMSO. In 3′-alkyl analogues,31 where
π-aromatic effects are absent, and in 2′-alkyl analogues,
where altered side chain conformations appear,14a the
4-acetyl is shifted downfield relative to Taxol. The face-
on approach of the 4-acetate to the bicyclic system can
be achieved in only one possible orientation, namely
with the C3′ hydrogen directed inward with repect to
the concave shape of the diterpenolid core. At the same
time, the C13-O and C1′-C2′ bonds adopt torsional
angles that bring the indan aromatic ring over the
acetate methyl group. In 9a , strong to medium NOEs
are observed between H2′′a and the 18-methyl and
between H3′ and the 4-OAc methyl, as well as between
H3′ and H14R. Structure 9a 1 fulfillls the first two NOE
observations with four proton-proton distances within
the range 2.9-3.9 Å and two H‚‚‚H separations of 2.4-
3.0 Å, respectively. Accordingly, as pictured in Figure
1, the 2′′a hydrogens are syn to Me-18 and the C3′
proton is located within the concave region of the
molecule rather than directed outward from the convex
surface. However, it should be remembered that the
NOEs are averaged across all seven conformations.
Thus, 9a 3 and 9a 4, assigned populations of 15% each
(Table 2), contribute to the H3′-H14R NOE cross-peaks
with distances of 3.6 and 3.1 Å, respectively. Both
likewise exhibit a face-on relationship between 4-OAc
methyl and the indan aromatic ring.
There are three additional distinctive features of the
9a NAMFIS conformations. First, all known families of
conformers are represented. Five forms, 9a 1, 9a 2, 9a 3,
9a 6 and 9a 7, fall into a group characterized by the lack
of hydrophobic collapse between C2 and C3′ hydro-
phobes. Structure 9a 4 (15%) resides in the “nonpolar”
category as a result of aromatic collapse between C3′-
NHCO-Ph and C2-OCO-Ph. By contrast, 9a 5 (12%)
is of the “polar” type exhibiting a tight cluster among
the C2 phenyl, the indan aromatic ring, and the C4 OAc
methyl groups. Second, the indan five-membered ring
in 9a is capable of puckering to give two envelope
conformations, A and B (Figure 2). Conformers 9a 2 and
9a 5 belong in the A class with the benzamidophenyl
moiety in a pseudoequatorial orientation. The other five
forms in the B class present the NHCOPh group in a
pseudoaxial disposition. The observation that the sizes
The convex shape of the baccatin core makes it
possible for R-oriented ester groups at C2, C4, and C13
to engage in hydrophobic interactions.10 In Taxol, the
flexibility of these side chains translates into a palette
of different conformations, two extremes being the so-
called “nonpolar” and “polar” conformations, where the
C2 benzoate phenyl ring clusters with the N-acyl group
and the 3′-phenyl of the amino acid side chain, respec-
tively.10 These conformations are exemplified by the
X-ray geometries of Taxol24 and Taxotere,23 but rapid
equilibration occurs in solution resulting in several
families of rotamers. In chloroform, a bias toward
“nonpolar” forms with small H2′-H3′ torsion angles is
observed,10h while the contribution of the “polar” form
with a large H2′-H3′ torsion angle is increased in
water-DMSO.10 One of the indices of these changes is
rotation around the C2′-C3′ bond. Since the 2′-2′′
tether limits rotation, the conformations of 9a ,b, 20, and
24a ,b might differ substantially from Taxol. Thus, the
side chain conformational behavior of these compounds
in solution was investigated by combining NOESY and
ROESY NMR experiments with NAMFIS analysis.28a,b
2-D NOEs a n d NAMF IS An a lysis. Compound 9a ,
incorporating a five-membered ring between C2′ and the
ortho-position of the C3′ phenyl ring, presented 14 cross-
peaks in its 2-D NOESY NMR spectrum in DMSO-d6.
When the peak volumes were referenced to the confor-
mationally immobile H7-H10 distance of 2.40 Å, 14
averaged intramolecular distances could be calculated
(see Experimental Section). To perform a conformational
deconvolution with the NAMFIS technique, structure
9a was subjected to a Monte Carlo conformational
search with the MM3*/GBSA/H2O force field.29a-c The
resulting 1121 optimized conformers were combined
with the NOE-derived distances by means of the NAM-
FIS least-squares fitting procedure to give seven con-
formations, 9a 1-9a 7, with predicted solution popula-
tions ranging from 2-26%.
The structure predicted to be most populated, 9a 1
(Table 2), is pictured in Figure 1. The C13 side chain is
in an extended conformation; neither terminal phenyl
ring “collapsed” with the C2 benzoyl phenyl. A note-
worthy feature of the structure is the location of the
4-OAc methyl group above the face of the indan aro-

Taxol Analogues Conformationally Constrained at C13
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 10 1581
epothilone SAR.7b More recently, a computationally
refined â-tubulin-paclitaxel complex, likewise taking
advantage of the EC structure, has identified an ex-
tended conformation of the drug with a C13 side chain
orientation related to the nonpolar conformation, the
T-conformation, as most consistent with the electron
crystallographic density of the Râ-tubulin dimer.2c,10i,32
To correlate the 9a conformers with these models, the
A-C rings of each of the conformers 9a 5, 9a 4, and 9a 1
(and 9a 3) (Table 2) have been superimposed on the
corresponding atoms in the structures of polar,24
nonpolar,10b and T-forms of Taxol, respectively.10i The
spatial locations of the three phenyl rings at the C2 and
C3′ termini for each of the structure pairs are very
similar. To the extent that these three motifs are
candidates for Taxol binding in microtubules, 9a and
presumably 24a are molecules that appear to sustain
NMR compatible equilibrium mixtures in solution con-
taining all of them. Thus, although C2′-C3′ torsional
opportunities have been severely limited in 9a and 24a ,
the overall structures have compensated by adjustments
around C1′-C2′ and C3′-N. Indeed, while the local
geometries of 9a conformers differ from those of Taxol,
the whole-molecule properties as seen externally (e.g.
by a receptor cavity surface) have not been significantly
affected. Consequently, although cyclization around C2′
and C3′ induces an unusual geometry at the terminus
of the C13 side chains in 9a and 24a , the capacity of
the compounds to adopt each of the three proposed Taxol
binding conformers implies that the present series of
constrained Taxol analogues cannot distinguish among
the protein binding possibilities within the context of
ligand considerations alone. In fact, this observation
appears to be more broadly based. The NAMFIS study
of Taxol in CDCl3 likewise extracted members of the
three conformational families.10h
F igu r e 2. Envelope conformations of the indan system of 9a ,
with the benzamido group pseudoaxial (A) or pseudoequatorial
(B).
of the NOEs from 3′ to both 2′′a protons are unequal is
consistent with an unequal distribution of conformations
in which the benzamide group is partitioned between
pseudoaxial and pseudoequatorial locations. Third, the
available NOE data do not allow the NAMFIS decon-
volution to distinguish (C2′)C3′-NH(CO) rotamers in
which this angle falls near -90 or +90° as is in the case
for 9a 3 and 9a 4.
Although the spectrum of 9b was recorded in CDCl3,
some insights on the two features highlighed for 9a in
Figure 1a can be gleaned for this compound as well. A
Monte Carlo torsional MM3*/GBSA/CHCl3 search for
9b delivered 328 conformations, which, when combined
with seven NOEs and processed by a NAMFIS treat-
ment, provided four conformations with estimated popu-
lations of 59, 35, 5, and 1%. The most populated form
(9b1) is shown in Figure 1. Not surprisingly, it is very
similar to the nonpolar conformation of Taxol, illustrat-
ing hydrophobic collapse between the C2 and C3′ phenyl
groups at the bottom of the figure. Unlike 9a , however,
the 4-OAc methyl group of 9b is not found positioned
above the tetralin aromatic ring. The C13-O and C1′-
C2′ bond torsions position the C3′-amide bond near the
acetate instead. In accord, the chemical shift of this
moiety is found at 2.22 ppm, almost identical to that in
Taxol. At the same time, H3′ evidences a strong NOE
to the 4-acetate methyl indicating that H3′ is directed
into the concave hydrophobic region of the molecule. In
consequence, the 6-ring protons at 2′′a and 2′′b are
oriented syn to the C18 methyl of the terpenoid core as
is the case for 9a . Furthermore, in 9b with a six-
membered ring, the benzamide and C1′ groups appear
to be in pseudoequatorial and equatorial positions,
respectively. The NOEs and J -couplings between the
2′′a and 2′′b protons are all consistent with this ring
conformation. Thus, the preferred conformation of 9b
(Figure 1b) appears to be similar to 9a with respect to
the overall orientation of ring-fused C3′ phenyl and the
terpenoid core fragment.
Finally, do these conformations of 9a ,b correspond to
potential 3-D templates for binding at the â-tubulin/
Taxol binding site? Three competing proposals have
emerged concerning the bioactive conformation of Taxol
at the â-tubulin binding pocket. On the basis of solution
NMR data and molecular modeling of nonataxel, Ojima
and co-workers proposed a pharmacophore that incor-
porates the hydrophobically collapsed polar conforma-
tion of Taxol.7c The same conformation in a different
binding mode has been presented by Li et al. via
REDOR NMR studies which utilized the electron crys-
tallographic coordinates (EC) of â-tubulin obtained
using Zn-stabilized, non-native protein sheets.10g On the
other hand, a pharmacophore model based on the
hydrophobically collapsed nonpolar conformation of
Taxol accommodates a wide range of taxoid and
Su m m a r y a n d Con clu sion s
The introduction of a tether between C2′ and the
ortho-position of the C3′ phenyl ring in Taxol and
Taxotere produces compounds that are only slightly less
active than Taxol in both tubulin polymerization and
cytotoxicity assays (Table 1). Tolerance for structural
deviation, however, is not high. Both tether lengthening
and stereoinversion at C2′ and C3′ compromise activity.
Conformational analysis coupled to 2-D NMR through
the NAMFIS procedure illustrates that the active
structures, 9a and presumably 24a , are characterized
by a family of rapidly interconverting conformations.
Among the seven conformers of 9a compatible with
solution NMR properties are found all three major
types: polar, nonpolar, and extended conformations.
The extended category constitutes 73% of the total
mixture, while 59% of this fraction (47% of the total) is
found in the T-form (i.e. 9a 1, 9a 3, and 9a 7).
Above we reviewed recent reports concerning the
three hypotheses for the binding conformation of Taxol
at the â-tubulin binding site. Two of them, the polar
and nonpolar conformers, derive from pharmacophore
models that consider the SAR of the ligands alone as a
guide to the shape of the protein binding site. The fact
that both nonpolar7b and polar7c models are able to
correlate a large body of taxoid and epothilone SAR
illustrates that such constructs do not converge to a
unique view of the protein-ligand complex. The situa-

1582 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 10
Barboni et al.
(B) ) 19.2 min, purity 97%; 24b: tR(A) ) 14.5 min, tR(B) )
tion is parallel to the point expressed above concerning
9a . In the absence of a constraining protein framework,
individual ligands can adopt a range of relatively low
energy conformations. Thus, in the present context, we
believe that while pharmacophore models are useful,
they incorporate significant limitations in their abilities
to predict the bioactive conformation. This leaves the
Li et al. proposal as the primary support for the polar
conformation based on REDOR measurements within
the tubulin-ligand complex.10g However, T-Taxol like-
wise fulfills the geometric requirements of these mea-
surements.32 Since the polar conformation does not fit
the electron crystallographic density at the binding
cavity of â-tubulin while the T-conformer does,10i we
currently favor the latter as the best candidate for the
bioactive conformation of taxoid drugs. Under this
assumption, the dominant conformation of the con-
strained analogue 9a (9a 1) is superbly suited to mimic
T-taxol in the binding site. As mentioned above, docking
experiments with this structure28 and a refined model
of Râ-tubulin are in full accord with this viewpoint.
Biological analogues of Taxol belonging to six distinct
structural types have been discovered within the natu-
ral product pool (epothilones,6a discodermolide,6b eleuth-
sides,6c rhazinilam,6d laulimalides,6e WS9885B,6f and
polyisoprenylated benzoylphloroglucinols6g). A structur-
ally simple mimic of Taxol has also been discovered
while screening a library of synthetic compounds for
antimitotic activity,33 and a competitive inhibitor of
Taxol binding to microtubules has been reported.34 All
these compounds share a common recognition site on
tubulin, and an enormous amount of work has been
devoted to the identification of a common pharmacoph-
ore, assessing the structural elements responsible for
this type of activity. The discovery of conformationally
constrained analogues of Taxol substantially retaining
the biological activity of the natural product is a
significant addition to this research database and
underscores the relevance of conformational point mu-
tations to focus on the 3-D elements underlying the
microtubule-stabilizing activity of taxoids.
19.5 min, purity 96%.
Biologica l Assa ys. Cytotoxicity and tubulin binding were
carried out according to established literature protocols.35
Ma ter ia ls. Commercially available reagents and solvents
were used without further purification. CH2Cl2 was dried by
distillation from CaH2 and THF by distillation from Na-
benzophenone. 10-Deacetylbaccatin III was a generous gift
from Indena, Milano, Italy. The tethered cinnamates 3a ,b were
prepared according to literature procedures.36,37
Syn th esis of Eth yl 2,2′-Meth ylid en e-N-ben zoylp h e-
n ylisoser in e (5a) an d Eth yl 2,2′-Eth yliden e-N-ben zoylph e-
n ylisoser in e (5b) via Asym m etr ic Am in oh yd r oxyla tion .
Syn th esis of 5b a s Rep r esen ta tive. (a ) To a stirred solution
of benzyl carbamate (2.32 g, 15.3 mmol) in 2-propanol (20 mL),
a solution of NaOH (604 mg) in water (37 mL), tert-butyl
hypochlorite (1.73 mL) and (DHQ)2PHAL (198 mg) in 2-pro-
panol (17.3 mL) were sequentially added. After stirring at room
temperature for 10 min, ethyl 3,4-dihydronaphthalene-2-
carboxylate (1 g, 4.95 mmol) was added, followed by K2OsO2-
(OH)4 (73 mg). After stirring at room temperature for 7 h, the
reaction was worked up by cooling in a ice bath and addition
of saturated Na2SO3 (49.5 mL). After further stirring for 15
min, the reaction mixture was extracted with EtOAc, and the
organic phase was washed with brine, dried (Na2SO4) and
evaporated. The residue was chromatographed on silica gel
25
(hexanes-EtOAc 4:1) to give 610 mg 4b as a gum (33%): [R]D
ca. 0 (CHCl3, c 1.0); IR (neat) 1740, 1650, 1250, 1010, 990 cm-1
;
HRMS m/z 369.1566 (calcd for C21H23NO5, 369.1576); 1H NMR
(300 MHz, CDCl3) δ 8.03 (br m, 1H), 7.52 (br m, 1H), 7.30 (m,
2H), 5.13 (s, 1H), 4.38 (br d, J ) 5.9 Hz, 1H), 4.28 (br s, 1H),
4.20 (q, J ) 7.4 Hz, 2H), 3.11 (m, 1H), 2.70 (m, 1H), 2.23 (m,
1H), 1.20 (t, J ) 7.4 Hz, 3H).
(b) To a solution of 4b (547 mg, 1.49 mmol) in AcOH (23
mL), ammonium formate (375 mg) and 10% Pd on charcoal
(257 mg) were added. After standing at room temperature for
4.5 h, the reaction mixture was filtered over Celite and
evaporated under reduced pressure. The residue was parti-
tioned between aqueous saturated NaHCO3 and EtOAc, and
the organic phase was dried (Na2SO4) and evaporated. The
residue was dissolved in CH2Cl2 (5 mL), and sequentially
treated with an aqueous solution of NaHCO3 (191 mg in 5 mL)
and benzoyl chloride (166 µL, 1.43 mmol, 1.2 mol equiv). After
stirring at room temperature for 3 h, the reaction was worked
up by extraction with CH2Cl2. The organic phase was dried
(Na2SO4) and evaporated. The residue was purified by CC
(hexanes-EtOAc gradient, from 8:2 to 7:3) to give 354 mg 5b
25
(70% from 4b), having [R]D ca. 0. For the other physical and
spectroscopic data of 5b, see 14. The optical purity of 5a ,b was
established as detailed for 10a (see infra), and the observed
ee were ca. 0 and <10%, respectively. Coupling of 5a ,b to
7-triethylsilylbaccatin was carried out as described for 14 (see
infra), employing a 2-fold excess of acid compared to 8.
Exp er im en ta l Section
Gen er a l Met h od s. Anhydrous conditions were achieved
(when indicated) by flame-drying flasks and equipments.
Reactions were monitored by TLC on Merck 60 F254 (0.25 mm)
plates, which were visualized with 5% H2SO4 in EtOH and
heating. Merck silica gel (70-230 mesh) was used for open-
column chromatography (CC). A Waters microPorasil HP
hypersyl column, 5 µm, 200 × 2.1 mm was used for HPLC,
with detection by a Waters diode matrix DAD detector. Melting
points were obtained on a Bu¨chi SMP-20 apparatus and are
uncorrected. 1H NMR (300 MHz) and 13C NMR (75 MHz)
spectra were recorded on a Bruker AC-300 spectrometer at
25 °C. ROESY spectra were recorded at 25 °C in DMSO-d6
and NOESY spectra were recorded at -10 °C in DMSO-d6-
D2O (3:1) on a Bruker AM-500 NMR spectrometer (500.13
(1S,2R)-Eth yl 1,2-Dih yd r oxyin d a n -2-ca r boxyla te (10a ).
To a solution of ADmix-R (14 g) in water-tert-butyl alcohol
(1:1, 100 mL), ethyl indene-2-carboxylate (1.88 g, 10 mmol)
was added in small portions over 5 min. After stirring at room
temperature overnight, the reaction was worked up by the
addition of solid Na2SO3 (12 g), stirring at room temperature
for 1 h, and extraction with CH2Cl2. After drying (MgSO4),
filtration, and removal of the solvent, the residue was purified
by CC (hexanes-EtOAc 8:2) to afford 1.4 g (64.5%) 10a as a
viscous yellowish oil: [R]D -15.0 (CHCl3, c 1.0); IR (neat)
25
3400, 1738, 1370, 1250, 1220 cm-1; HRMS m/z 222.0886 (calcd
1
1
MHz for H). The spectra were fully assigned using HMBC on
for C12H14O4, 222.0892); H NMR (300 MHz, CDCl3) δ 7.50-
a Bruker Advance 400 (400.13 MHz for 1H). The purity of all
the final products (9a , 9b, 20, 24a and 24b) was assayed by
HPLC on a Water microPorasil column (0.8 × 30 cm), with
detection by a Waters differential refractometer 340, using
hexanes-EtOAc 4:6 (system A) and CHCl3-acetone 3:1 (sys-
tem B) as eluant. The values of tR and purity were as follows:
9a : tR(A) ) 14.8 min, tR(B) ) 19.3 min, purity 97%; 9b: tR(A)
) 15.0 min, tR(B) ) 19.6 min, purity 98%; 20: tR(A) ) 14.7
min, tR(B) ) 19.0 min, purity 96%; 24a : tR(A) ) 14.3 min, tR-
7.18 (m, H-4, H-5, H-6, H-7), 5.37 (br s, H-1), 4.34 (q, J ) 7.2
Hz, OEt), 3.78 (s, 2-OH), 3.50 (d, J ) 16.2 Hz, H-3a), 3.10 (d,
J ) 16.2 Hz, H-3b), 2.89 (br s, 1-OH), 1.34 (t, J ) 7.2 Hz, OEt).
The optical purity of 10a was determined as follows: to a
cooled (0 °C) solution of 10a (10 mg, 0.045 mmol) in dry THF
(0.5 mL) TEA (9.4 µL, 6.83 mg, 0.067 mmol, 1.25 mol equiv)
and menthylchloroformate (10.7 µL, 10.9 mg, 0.05 mmol, 1.1
mol equiv in 0.5 mL dry THF) were added dropwise. After
stirring 1 h at 0 °C, the cooling bath was removed, and stirring

Taxol Analogues Conformationally Constrained at C13
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 10 1583
was continued for 1 h, during which a white precipitate was
formed. The reaction was then worked up by removal of the
solvent and partition between Et2O and 10% NaOH. After
drying (MgSO4) and evaporation, the residue was analyzed by
HPLC (hexanes-EtOAc 95:5 as eluant; flow 0.7 mL/min),
obtaining two peaks having identical UV spectra at tR ) 3.01
and 4.33, in a 96:4 ratio (ee ) 92%): 1H NMR (major isomer,
300 MHz, CDCl3) δ 7.31 (m, 4H), 6.14 (s, 1H), 4.58 (dt, J )
10.7, 4.6 Hz, 1H), 4.33 (q, J ) 7.1 Hz, 2H), 3.53 (d, J ) 16.2
Hz, 1H), 3.19 (d, J ) 16.2 Hz, 1H), 2.10 (m, 1H), 1.98 (m, 1H),
1.70 (m, 2H), 1.44 (m, 2H), 1.32 (t, J ) 7.1 Hz, 3H), 1.12 (m,
2H), 0.95 (d, J ) 7.0 Hz, 3H), 0.89 (d, J ) 5.8 Hz, 3H), 0.81 (d,
J ) 5.8 Hz, 3H).
(1R,2S)-Eth yl 1-Br om o-2-ben zoyloxyin d a n -2-ca r boxy-
la te (11). To a solution of 10a (3.02 g, 13.6 mmol) in dry CH2-
Cl2 (25 mL), p-toluenesulfonic acid monohydrate (25 mg, 0.136
mmol) and trimethyl orthobenzoate (3.04 mL, 3.22 g, 17.7
mmol, 1.3 mol equiv) were added. After stirring at room
temperature for 3 h, the solvent was removed, leaving a
yellowish oil. The latter was dissolved in dry CH2Cl2 (25 mL)
and cooled to -15 °C. Acetyl bromide was then added (1.71
mL, 2 g, 16.3 mmol, 1.1 mol equiv). After stirring at - 15 °C
for 3 h, further trimethyl orthobenzoate (0.29 mL) and acetyl
bromide (0.16 mL) were added, and stirring was continued at
- 15 °C for 1 h. The cooling bath was then removed, and the
reaction mixture was allowed to warm to room temperature,
and worked up by evaporation. The residue was purified by
CC (hexanes-EtOAc gradient, from 95:5 to 9:1) to afford 11
as a foam (3.80 g, 72%): 1H NMR (200 MHz, CDCl3) δ 7.93
(Bz AA′), 7.60-7.20 (m, Bz BB′ + H-4, H-5, H-6, H-7), 5.68 (s,
H-1), 4.33 (q, J ) 7.5 Hz, OEt), 4.27 (d, J ) 17.6 Hz, H-3a),
3.46 (d, J ) 17.6 Hz, H-3b), 1.30 (t, J ) 7.5 Hz, OEt).
taken up in 0.1 M methanolic HCl. After stirring 1 h at room
temperature, the reaction was evaporated, and the residue
purified by CC (silica gel, 2% MeOH in hexanes-EtOAc 1:1)
to afford 47 mg (35% from 12) 9a as a colorless powder: mp
25
254 °C; [R]D -48 (CHCl3, c 0.70); IR (KBr) 1745, 1457, 1383,
1203, 720 cm-1; HRMS-FAB m/z 866.3386 (calcd for C48H51
-
NO14 + H+ 866.3388); 1H NMR (300 MHz, CDCl3) δ 8.08 (AA′
OBz), 7.80 (AA′ NHBz), 7.60-7.40 (m, BB′OBz, C OBz, BB′
NHBz, C NHBz), 7.33 (m, H-3′′, H-4′′, H-5′′, H-6′′), 6.78 (d, J
) 9.0 Hz, NH), 6.42 (br t, J ) 9.2 Hz, H-13), 6.29 (s, H-10),
6.10 (d, J ) 9.0 Hz, H-3′), 5.69 (d, J ) 7.3 Hz, H-2), 4.85 (br d,
J ) 9.9 Hz, H-5), 4.37 (dd, J ) 10.8, 6.8 Hz, H-7), 4.25 (d, J )
8.4 Hz, H-20R), 4.16 (d, J ) 8.4 Hz, H-20â), 3.83 (d, J ) 16.5
Hz, 2′,2′′-CH2R), 3.79 (d, J ) 7.3 Hz, H-3), 3.30 (d, J ) 16.5
Hz, 2,2′′ CH2â), 2.50 (ddd, J ) 14.9, 9.9, 6.8 Hz, H-6R), 2.44
(dd, J ) 15.4, 9.2 Hz, H-14b), 2.35 (dd, J ) 15.4, 9.2 Hz, H-14a),
2.24 (s, 10-OAc), 1.93 (br s, H-18), 1.87 (s, 4-OAc), 1.85 (ddd,
J ) 14.9, 11.0, 2.2 Hz, H-6b), 1.67 (s, H-19), 1.30 (s, H-17),
1.16 (s, H-16); 13C NMR (75 MHz, CDCl3) δ 203.6 (s, C-9), 175.1
(s, C-1′), 171.3 (s, 10-OAc), 170.4 (s, 4-OAc), 168.0 (s, NHBz),
166.9 (s, OBz), 142.0 (s, C-12), 139.1 (s, C-2′′), 138.8 (s,
i-NHBz), 133.6 (d, p-OBz), 133.4 (s, C-1′′), 133.1 (s, C-11), 132.1
(d, C-4′′), 130.2 (d, o-OBz), 129.1 (d, C-5′′), 129.1 (s, i-OBz),
128.7 (d, m-NHBz), 128.7 (d, m-OBz), 127.9 (d, p-NHBz), 127.2
(d, o-NHBz), 125.2 (d, C-3′′), 123.8 (d, C-6′′), 84.4 (d, C-5), 82.1
(d, C-2′), 81.2 (s, C-4), 78.9 (s, C-1), 76.6 (t, C-20), 75.6 (d, C-10),
74.9 (d, C-2), 72.6 (d, C-13), 72.3 (d, C-7), 61.7 (d, C-3′), 58.7
(s, C-8), 45.6 (d, C-3), 43.9 (t, 2′(2′′)-CH2), 43.3 (s, C-15), 35.7
(t, C-6 + C-14), 26.9 (q, C-17), 21.9 (q, C-16), 21.7 (4-OAc),
20.9 (10-OAc), 15.0 (q, C-18), 9.5 (q, C-19).
(1S,2R)-Eth yl 1,2-Dih yd r oxytetr a h yd r on a p h th a len e-
2-ca r boxyla te (10b). To a solution of ADmix-R (14 g) in
water-tert-butyl alcohol (1:1, 100 mL), ethyl 3,4-dihydronaph-
thalene-2-carboxylate (2.02 g, 10 mmol) was added in small
portions over 5 min. After stirring at room temperature
overnight, the reaction was worked up by addition of solid Na2-
SO3 (12 g), stirring at room temperature for 1 h, and extraction
with CH2Cl2. After drying (MgSO4), filtration, and removal of
the solvent, the residue was purified by CC (hexanes-EtOAc
(1S,2R)-Eth yl 1-Ben zoyla m in o-2-h yd r oxyin d a n -2-ca r -
boxyla te (12). To a solution of 11 (4 g, 10.3 mmol) in dry DMF
(16 mL), an excess NaN3 (2.68 g, 41 mmol) was added, and
the solution was heated at 50 °C for 36 h. The reaction was
worked up by dilution with Et2O (50 mL) and washing with
water (3 × 40 mL), After drying (MgSO4) and filtration, the
organic phase was evaporated, and the residue dissolved in
dry EtOH (40 mL) and hydrogenated at ambient pressure in
the presence of 10% Pd on charchoal (1.2 g) for 36 h. After
filtration, the solution was left standing at room temperature
for 48 h, and then evaporated. The residue was purified by
CC (silica gel, hexanes-EtOAc gradient, from 85:15 to 7:3) to
8:2) to afford 1.8 g (77%) 10b as an amorphous solid: [R]D
25
-42.9 (CHCl3, c 1.0); IR (Nujol) 3413, 1720, 1376, 1269, 1235
cm-1; EI-MS (70 eV) 236 (M+), 218, 200, 172, 162, 120; 1H NMR
(300 MHz, CDCl3) δ 7.58 (d, J ) 7.5, H-9), 7.20 (m, H-7, H-8),
7.09 (d, J ) 7.3 Hz, H-5), 5.00 (br d, J ) 11.1 Hz, H-1), 4.31
(q, J ) 7.1 Hz, OEt), 3.59 (s, 2-OH), 3.12 (ddd, J ) 17.0, 12.5,
6.0 Hz, H-4a), 2.74 (ddd, J ) 17.0, 6.0, 2.0 Hz, H-4b), 2.56 (br
d, J ) 11.2 Hz, 1-OH), 2.20 (ddd, J ) 13.7, 12.6, 6.0 Hz, H-3a),
2.02 (ddd, J ) 13.7, 6.0, 2.0 Hz, H-3b), 1.32 (t, J ) 7.1 Hz,
OEt). The optical purity of 10b was determined as described
for 10a , giving an ee value of 93%.
get 12 (700 mg, 21%) as a foam: [R]D +11 (CHCl3, c 1.1); IR
25
(KBr) 3391, 1731, 1651, 1580, 1519, 1339, 1231 cm-1; LC-
MS (ESI) 348.0 (M + Na+); 1H NMR (300 MHz, CDCl3) δ 7.80
(Bz AA′), 7.40-7.60 (m, Bz BB′, Bz C), 7.26 (br s, H-4, H-5,
H-6, H-7), 6.80 (d, J ) 9.3 Hz, NH), 6.09 (d, J ) 9.3 Hz, H-1),
4.36 (q, J ) 7.2, OEt), 3.64 (d, J ) 16.3 Hz, H-3a), 3.14 (d, J
) 16.3 Hz, H-3b), 1.35 (t, J ) 7.2 Hz, OEt); 13C NMR (75.4
MHz, CDCl3) δ 175.0 (s), 167.7 (s), 140.0 (s), 139.0 (s), 133.9
(s), 131.8 (d), 128.6 (d), 128.4 (d), 127.4 (d), 127.2 (d), 124.9
(d),123.9 (d), 82.3 (s), 63.0 (d), 61.1 (t), 43.8 (t), 14.1 (q).
(1R,2S)-Eth yl 1-Br om o-2-h yd r oxytetr a h yd r on a p h th a -
len e-2-ca r boxyla te (13). To a stirred and cooled (0 °C)
solution of 10b (945 mg, 4.0 mmol) in dry CH2Cl2 (60 mL),
pyridine (967 µL, 948 mg, 12 mmol, 3 mol equiv) and thionyl
chloride (408 µL, 5.6 mmol, 1.4 mol equiv) were added. After
stirring at 0 °C for 10 min., the reaction mixture was filtered
on a silica gel pad, and evaporated. The crude sulfite was
dissolved in acetonitrile (40 mL) and treated with NaIO4 (1.284
g, 6 mmol, 1.5 mol equiv) and a solution of RuCl3‚H2O (10 mg,
0.04 mmol) in water (8 mL). After stirring for 15 min at room
temperature, the reaction was worked up by dilution with
water, extraction with ether and filtration over silica gel. After
drying (Na2SO4) and evaporation of the solvent, the crude and
unstable sulfate was dissolved in acetone-water (15 mL, 1:1)
and treated with an excess nBu4NBr (3.380 g, 10.4 mmol, 2.6
mol equiv). After stirring 2 h at room temperature the reaction
was worked up by concentration under vacuum to remove most
acetone and treatment with 20% H2SO4 (8.5 mL) and ether
(10.5 mL). After vigorous stirring for 30 min., the phases were
separated, and the aqueous layer was further extracted with
ether. The pooled organic phases were dried (Na2SO4) and
evaporated, and the residue was purified by CC (silica gel,
hexanes-EtOAc 9:1) to give 243 mg 13 (20%) as an amorphous
glass: EI-MS (70 eV) 301, 299 (M+), 274, 272, 220, 219; 169;
2′,2′′-Meth ylen ep a clita xel (9a ). To a solution of 12 (100
mg, 0.31 mmol, 2 mol equiv) in dry toluene (2.5 mL), pyri-
dinium p-toluenesulfonate (PPTS; 5 mg) and a solution of 2,4-
dimethoxybenzaldehyde dimethylacetal (140 mg, excess) in dry
toluene (0.5 mL) were added. The reaction was refluxed in a
Dean-Stark apparatus for 2 h, and then worked up by removal
of the solvent and partition between CH2Cl2 and water. The
organic phase was dried (MgSO4), filtered and evaporated. The
residue was dissolved in methanol (4.5 mL) and a solution of
K2CO3 (103 mg) in water (2.2 mL) was added. After stirring
at room temperature for 30 min., the reaction was worked up
by removal of the solvent and partition between water and
EtOAc. The aqueous phase was acidified with 5% KHSO4 and
extracted with EtOAc. After washing with brine, drying
(MgSO4) and evaporation, the unstable residue was dissolved
in dry toluene (5 mL) and directly coupled with 7-triethylsi-
lylbaccatin III38 (108.6 mg, 0.155 mmol, 0.5 mol equiv) in the
presence of DMAP (16.5 mg, 0.135 mmol) and DCC (76 mg,
0.37 mmol). After stirring at room temperature for 6 h, the
reaction was worked up by evaporation, and the residue was

1584 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 10
Barboni et al.
1H NMR (300 MHz, CDCl3) δ 7.35 (m, H-8), 7.19 (m, H-6, H-7),
7.12 (m, H-5), 5.19 (s, H-1), 4.31 (q, J ) 7.1 Hz, OEt), 3.03 (m,
H-4a, H-4b), 2.65 (m, H-3a), 2.16 (m, H-3b), 1.33 (t, J ) 7.1
Hz, OEt).
72.3, 72.2 (d, C-13 + C-7), 58.6 (s, C-8), 51.9 (d, C-3′), 45.5 (d,
C-3), 43.2 (s, C-15), 35.8, 35.6 (t, C-6, C-14), 31.8 (t, 2′,2′′-
ethylidene), 26.9(q, C-17), 24.6 (t, 2′,2′′-ethylidene), 22.7 (OAc),
22.3 (C-16), 20.9 (OAc), 14.9 (C-18), 9.6 (C-19).
(1S,2R)-E t h yl
1-Ben zoyla m in o-2-h yd r oxyt et r a h y-
2′,2′′-Meth ylen e-2′,3′-ep ip a clita xel (20). To a solution of
19 (138 mg, 0.31 mmol, 2 mol equiv, prepared from 3a as
described for 9a but employing ADmix-â for the dihydroxyla-
tion step (ee of the diol: 92%)) in dry toluene (12.5 mL), DMAP
(37 mg, 0.31 mmol, 2 mol equiv), 7-triethylsilylbaccatin III (8)
(108 mg, 0.155 mmol) and DCC (166 mg, 0.8 mmol, 5.1 mol
equiv) were added, and the reaction was stirred at room
temperature overnight and then heated at 70 °C for 1 h. After
cooling, the reaction was filtered, and evaporated. The residue
was taken up in 0.1 M methanolic HCl (21 mL). After stirring
1.5 h at room temperature, the reaction was evaporated, and
the residue purified by CC (silica gel, hexanes-EtOAc 1:1) to
d r on a p h th a len e-2-ca r boxyla te (14). To a stirred solution
of 13 (543 mg, 1.82 mmol) in dry DMF, NaN3 (471 mg, 7.24
mmol, 4 mol equiv) was added, and the solution was stirred
10 min. at 50 °C. The reaction was worked up by addition of
ice and extraction with Et2O. After drying (MgSO4), the organic
phase was evaporated, and the residue was purified by CC
(silica gel, hexanes-EtOAc 4:1) to afford 60 mg (13%) of the
corresponding azide. To a solution of the latter (284 mg, 1.19
mmol) in EtOAc (20 mL), Pd on charcoal (50 mg) was added,
and the mixture was hydrogenated at room temperature and
atmospheric pressure for 4 h. After filtration over Celite and
evaporation, the residue was dissolved in CH2Cl2 (5 mL), and
an aqueous solution of NaHCO3 (191 mg in 5 mL) and benzoyl
chloride (166 µL, 201 mg, 1.43 mmol, 1.2 mol equiv) were
added. After vigorous stirring for 3 h, the phases were
separated, and the aqueous phase was extracted with CH2-
Cl2. The pooled organic phases were dried (Na2SO4) and
evaporated, to afford 536 mg (87%) 14 as an amorphous solid:
25
afford 31 mg (11%) 20 as a foam: [R]D -100 (CHCl3, c 0.80);
IR (KBr) 1716, 1653, 1488, 1373, 1241, 751 cm-1; HRMS-FAB
m/z 866.3380 (calcd for C48H51NO14 + H+ 866.3388); 1H NMR
(300 MHz, CDCl3) δ 8.08 (AA′ OBz), 7.63-7.40 (overlapped
m, AA′ NHBz, BB′OBz, C OBz, BB′ NHBz, C NHBz), 7.33 (m,
H-3′′, H-4′′, H-5′′, H-6′′), 6.81 (d, J ) 8.0 Hz, NH), 6.38 (br t,
J ) 8.8 Hz, H-13), 6.28 (s, H-10), 6.18 (d, J ) 8.8 Hz, H-3′),
5.66 (d, J ) 7.1 Hz, H-2), 4.88 (br d, J ) 9.6 Hz, H-5), 4.40
(dd, J ) 10.8, 6.7 Hz, H-7), 4.25 (d, J ) 8.8 Hz, H-20R), 4.12
(d, J ) 8.8 Hz, H-20â), 3.80 (d, J ) 7.3 Hz, H-3), 3.65 (d, J )
16.3 Hz, 2′,2′′-CH2R), 3.27 (d, J ) 16.3 Hz, 2,2′′ CH2â), 2.51
(m, H-6R), 2.28 (m, H-14a + H-14b), 2.21 (s, 10-OAc), 2.00 (br
s, H-18), 1.91 (s, 4-OAc), 1.83 (m, H-6â), 1.65 (s, H-19), 1.30
(s, H-17), 1.14 (s, H-16); 13C NMR (75 MHz, CDCl3) δ 203.7 (s,
C-9), 175.1 (s, C-1′), 171.2 (s, 10-OAc), 169.8 (s, 4-OAc), 167.5
(s, NHBz), 167.2 (s, OBz), 142.8 (s, C-12), 139.7 (s, C-2′′), 138.2
(s, i-NHBz), 133.8 (d, p-OBz), 132.5 (s, C-1′′), 132.0 (d, C-4′′),
130.9 (s, C-11), 130.0 (d, o-OBz), 129.1 (d, C-5′′), 129.0 (s,
i-OBz), 128.7 (d, m-NHBz), 128.7 (d, m-OBz), 128.0 (d, p-
NHBz), 127.1 (d, o-NHBz), 124.9 (d, C-3′′), 124.4 (d, C-6′′),
84.5 (d, C-5), 82.2 (d, C-2′), 81.2 (s, C-4), 79.4 (s, C-1), 76.5 (t,
C-20), 75.5 (d, C-10), 75.0 (d, C-2), 72.9 (d, C-13), 72.2 (d, C-7),
61.1 (d, C-3′), 58.6 (s, C-8), 45.5 (d, C-3), 44.5 (t, 2′(2′′)-CH2),
43.2 (s, C-15), 35.9, 35.6 (t, C-6 + C-14), 27.0 (q, C-17), 22.1
(q, C-16), 21.7 (4-OAc), 20.8 (10-OAc), 15.1 (q, C-18), 9.6 (q,
C-19).
[R]D 45.2 (CHCl3, c 1.4); IR (Nujol) 3385, 1730, 1655, 1340,
25
1225 cm-1; LC-MS (ESI) 362.0; 1H NMR (300 MHz, CDCl3) δ
7.77 (m, 2H), 7.26 (m, 1H), 7.15 (m, 3H), 6.73 (d, J ) 9.5 Hz,
1H), 5.86 (d, J ) 9.5 Hz, 1H), 4.27 (m, 2H), 3.19 (ddd, J )
16.8, 12.8, 5.7 Hz, 1H), 2.83 (ddd, J ) 16.8, 5.8, 2.0 Hz, 1H),
2.36 (ddd, J ) 13.7, 12.8, 5.8 Hz, 1H), 2.04 (ddd, J ) 13.7, 5.7,
2.2 Hz, 1H), 1.28 (t, J ) 7.2 Hz, 3H).
2′,2′′-Eth ylid en ep a clita xel (9b). To a solution of 14 (406
mg, 1.2 mmol, 2 mol equiv) in dry toluene (5 mL), PPTS (10
mg) and a solution of 2,4-dimethoxybenzaldehyde dimethy-
lacetal (537 mg) in dry toluene (1 mL) were added. The reaction
was refluxed in a Dean-Stark apparatus for 2 h, and then
worked up by removal of the solvent and partition between
CH2Cl2 and water. The organic phase was dried (MgSO4),
filtered and evaporated. The residue was dissolved in methanol
(10 mL) and a solution of K2CO3 (395 mg) in water (8.5 mL)
was added. After stirring at room temperature for 30 min.,
the reaction was worked up by removal of the solvent and
partition between water and EtOAc. The aqueous phase was
acidified with 5% KHSO4 and extracted with EtOAc. After
washing with brine, drying (MgSO4) and evaporation, the
unstable residue was dissolved in dry toluene (19 mL) and
directly coupled with 7-triethylsilylbaccatin III (414 mg, 0.60
mmol) in the presence of DMAP (63 mg, 0.51 mmol) and DCC
(291 mg, 1,42 mmol, 1.2 mol equiv). After stirring at room
temperature overnight, the reaction was worked up by evapo-
ration, and the residue was taken up in 0.1 M methanolic HCl.
After stirring 1 h at room temperature, the reaction was
evaporated, and the residue purified by CC (silica gel, 2%
MeOH in hexanes-EtOAc 1:1) to afford 301 mg (29%) 9b as a
Syn th esis of (()-2,2′-Meth ylid en e-N-(ter t-bu toxyca r -
bon yl)p h en ylisoser in e (21a ) a n d (()-2,2′-Eth ylid en e-N-
(ter t-bu toxyca r bon yl)p h en ylisoser in e (21b). Syn th esis of
21b a s Rep r esen ta tive. Compound (()4b was reductively
deacylated as described for the synthesis of 5b. To a solution
of the crude deacylated product (307 mg, 1.3 mmol) in CH2Cl2
(4.5 mL), aqueous NaHCO3 (158 mg in 4.5 mL) and tert-
butoxycarbonylpyrocarbonate (Boc2O; 284 mg, 1.3 mmol, 1 mol
equiv) were added. After stirring at room temperature for 5
h, further solid NaHCO3 (23 mg) and Boc2O (38 mg) were
added, and stirring was continued overnight. After separation
of the two phases, the water phase was extracted with CH2-
Cl2, and the pooled organic phases were dried (Na2SO4) and
evaporated. The residue was purified by CC (hexanes-EtOAc
gradient, from 9:1 to 8:2), 289 mg 21b were obtained (66%) as
an amorphous solid: [R]D25 9.2 (CHCl3, c 1.1); IR (Nujol) 1731,
1678, 1528, 1250, 1173 cm-1; LC-MS (ESI) 358.4 (M + Na+);
1H NMR (300 MHz, CDCl3) δ 7.26 (m, 1H), 7.16 (m, 2H), 7.09
(m, 1H), 5.26 (br d, J ) 10.2 Hz, 1H), 5.07 (br d, J ) 10.2 Hz,
1H), 4.27 (m, 2H), 3.12 (ddd, J ) 16.8, 12.7, 5.7 Hz, 1H), 2.76
(ddd, J ) 16.8, 6.0, 2.2 Hz, 1H), 2.29 (ddd, J ) 13.5, 12.7, 6.0
Hz, 1H), 1.97 (ddd, J ) 13.5, 5.7, 2.2 Hz, 1H), 1.31 (t, J ) 7.1
Hz, 3H). Physical data for 21a (43% yield from 5a ): gum; IR
(neat) 1735, 1680, 1610, 1510, 1247, 1089 cm-1; HRMS m/z
321.1588 (calcd for C17H23NO5, 321.1576); 1H NMR (300 MHz,
CDCl3) δ 7.23 (m, 4H), 5.53 (br d, J ) 9.8 Hz, 1H), 5.13 (br d,
J ) 9.8 Hz, 1H), 4.31 (q, 2H), 3.60 (br s, 1H), 3.56 (d, J ) 16.1
Hz, 1H), 3.06 (d, J ) 16.1 Hz, 1H), 1.45 (s, 9H), 1.32 (t, J )
6.9 Hz, 3H); 13C NMR (75.4 MHz, CDCl3) δ 174.9 (s), 155.8
(s), 140.1 (s), 138.9 (s), 128.2 (d), 127.2 (d), 124.8 (d), 123.9
(d), 82.3 (s), 79.8 (s), 62.7 (d), 62.6 (t), 43.1 (t), 28.3 (q), 14.1
(q).
colorless powder: mp 254 °C; [R]D 3.3 (CHCl3, c 0.73); IR
25
(KBr) 1745, 1457, 1383, 1203, 720 cm-1; HRMS-FAB m/z
880.3577 (calcd for C49H54NO14 + H+ 880.3544); 1H NMR (300
MHz, CDCl3) δ 8.22 (AA′ OBz), 7.71 (AA′ NHBz), 7.60-7.40
(m, BB′OBz, C OBz, BB′ NHBz,), 7.34 (C NHBz), 7.23 (m, H-3′′,
H-4′′, H-5′′, H-6′′), 6.76 (d, J ) 9.6 Hz, NH), 6.35 (br t, J ) 9.2
Hz, H-13), 6.26 (s, H-10), 5.96 (d, J ) 9.0 Hz, H-3′), 5.66 (d, J
) 7.3 Hz, H-2), 4.86 (br d, J ) 8.9 Hz, H-5), 4.38 (dd, J )
10.8, 6.8 Hz, H-7), 4.25 (d, J ) 8.4 Hz, H-20R), 4.20 (d, J ) 8.4
Hz, H-20â), 3.80 (d, J ) 7.3 Hz, H-3), 3.27 (m, 2′,2′′-ethylidene),
2.92 (dd, J ) 16.8, 4.5 Hz, 2′,2′′-ethylidene), 2.48 (m, H-6R,
Η-14R, 2′,2′′-ethylidene), 2.32 (s, 10-OAc), 2.22 (s, 4-OAc), 2.20
(m, H-14â, 2′,2′′ethylidene), 1.87 (br s, H-18), 1.85 (m, H-6â),
1.68 (s, H-19), 1.24 (s, H-17), 1.12 (s, H-16); 13C NMR (75 MHz,
CDCl3) δ 203.7 (s, C-9), 175.6 (s, C-1′), 171.3 (s, 10-OAc), 171.0
(s, 4-OAc), 167.3, 167.1 (s, NHBz + OBz), 142.1 (s, C-12), 135.6
(s, C-2′′), 133.8 (d, p-OBz), 133.7, 133.6, (s, NHBz + C-1′′),
132.9 (C-11), 131.9 (d, pNBz), 130.4 (d, o-OBz), 129.2 (s, i-OBz),
128.8, 128.7 (d, m-NHBz + m-OBz), 127.6, 127.3, 127.1, 126.9
(d, C-3′′ + C-4′′ + C-5′′ + C-6′′), 84.5 (d, C-5), 81.1 (d, C-4),
79.2 (s, C-1), 76.7 (t, C-20), 75.5 (d, C-10), 75.1 (d, C-2 + C-2′),

Taxol Analogues Conformationally Constrained at C13
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 10 1585
2′,2′′-Meth ylen ed oceta xel (24a ). (a ) Am id o a lcoh ol p r o-
tection a n d ester h yd r olysis of 21a : To a solution of 21a
(321 mg, 1 mmol) in dry toluene (6.5 mL) PPTS (14 mg) and
2,4-dimethoxybenzaldehyde dimethylacetal (424 mg, excess)
were added, and the solution was refluxed in a Dean-Stark
apparatus for 2 h. After evaporation of the solvent, the residue
was taken up in CH2Cl2 and washed with brine. The organic
phase was dried (Na2SO4) and evaporated. The residue was
filtered through a short column of silica gel to remove the
excess acetal, and then hydrolyzed by treatment with K2CO3
in methanol (234 mg in 5 mL). After stirring 30 min at room
temperature, the reaction was worked up by evaporation, and
the residue taken up in a mixture of EtOAc and water (1:1,
10 mL). The water phase was next acidified with 5% KHSO4
and extracted with EtOAc. After drying (Na2SO4) and removal
of the solvent, crude 22a was obtained (356 mg, 80% from 21a )
as an unstable glass, directly used for the coupling step.
C-9), 175.7 (s, C-1′), 170.8 (s, 4-OAc), 167.1 (s, OBz), 156.1 (s,
OBoc), 138.6, 135.7, 135.5, 134.5 (s, C-11, C-12, C-1′′, C-2′′),
133.5 (d, p-OBz), 130.3 (d, o-OBz), 129.2 (s, i-OBz), 128.7 (d,
m-OBz), 127.6 127.3 (d, 3′-Ph), 127.0, 126.6 (d, C-3′′ + C-4′′ +
C-5′′ + C-6′′), 84.3 (d, C-5), 81.0 (s, C-4), 79.8 (s, NHBoc), 79.0
(s, C-1), 76.8 (t, C-20), 75.2 (d, C-2′), 74.9 (d, C-2), 74.4 (d, C-10),
72.1 (d, C-13, C,7), 57.6 (s, C-8), 53.4 (d, C-3′), 46.5 (d, C-3),
43.1 (s, C-15), 37.0 (t, C-6), 35.9 (t, C-14), 31.5 (t, 2′,2′′-
ethylidene), 28.2 (q, NHBoc), 26.6 (q, C-17), 24.7 (t, 2′,2′′-
ethylidene), 22.6 (q, OAc), 21.04 (q, C-16), 14.3 (q, C-18), 9.9
(q, C-19).
Com p u ta tion a l P r oced u r es. Structure 9a was subjected
to a Monte Carlo conformational search40 with the MM3* force
field and the GB/SA/H2O continuum model in MacroModel6.5
with default parameters.29 The Monte Carlo treatment was
employed since it is one of the most efficient procedures
available for exploring the conformational energy surface of
small to medium sized molecules.41 Similarly, the semiana-
lytical GB/SA/H2O model was chosen to mimic polar solvent
(b) Cou p lin g a n d d ep r otection : To a stirred solution of
22a (450 mg, 1.02 mmol, 1.6 mol equiv) and 7,10-bis(trichlo-
roacetyl)-10-deacetylbaccatin III (23)39 (576 mg, 0.64 mmol)
in dry toluene (22 mL), DMAP (67 mg, 0.55 mmol, 0.86 mmol)
and DCC (305 mg, 1.48 mmol, 2.3 mol equiv) were added, and
the mixture was stirred at room temperature overnight. The
reaction was worked up by evaporation, and the residue was
treated with 0.1 N methanolic HCl (5 mL) to hydrolyze the
acetal moiety on the side chain. After evaporation, the residue
was dissolved in dry methanol (22.5 mL) and treated with zinc
(previously activated by treatment with dilute HCl, filtration,
and washing with MeOH, 600 mg) and AcOH (13.5 mL). After
stirring at 60 °C for 2 h, the reaction mixture was filtered,
evaporated, and partitioned between water and EtOAc. The
organic phase was dried (Na2SO4) and evaporated. The residue
was purified by CC (hexanes-EtOAc gradient, from 35:65 to
3:7) to afford 159 mg 24a (17%) as a foam: [R]D25 -47 (CHCl3,
c 0.70); IR (KBr) 3435, 1707, 1244, 1166, 1094, 1053, 1025
cm-1; HRMS-FAB m/z 820.3551 (calcd for C44H53NO14 + H+
866.3544); 1H NMR (300 MHz, CDCl3) δ 8.05 (AA′ OBz), 7.55
(C OBz), 7.45 (BB′ OBz), 7.27 (m, H-3′′, H-4′′, H-5′′, H-6′′), 6.36
(br t, J ) 8.9 Hz, H-13), 6.29 (s, H-10), 5.67 (d, J ) 7.0 Hz,
H-2), 5.56 (d, J ) 9.6 Hz, H-3′), 5.20 (s, H-10), 5.16 (d, J ) 9.6
Hz, NH), 4.84 (br d, J ) 9.0 Hz, H-5), 4.24 (d, J ) 8.7 Hz,
H-20R), 4.19 (m H-7), 4.14 (d, J ) 8.7 Hz, H-20â), 3.89 (d, J )
7.0 Hz, H-3), 3.73 (d, J ) 16.5 Hz, 2′,2′′-CH2â), 3.20 (d, J )
16.5 Hz, 2,2′′ CH2R), 2.50 (m, H-6R), 2.29 (m, H-14a + H-14b),
1.96 (br s, H-18), 1.89 (s, 4-OAc), 1.80 (m, H-6â), 1.73 (s, H-19),
1.42 (s, NHBoc), 1.26 (s, H-17), 1.12 (s, H-16); 13C NMR (75
MHz, CDCl3) δ 211.4 (s, C-9), 175.1 (s, C-1′), 170.4 (s, 4-OAc),
167.0 (s, OBz), 156.2 (s, NHBoc), 139.5, 138.5, 138.4, 135.9 (s,
C-11, C-12, C-1′′, C-2′′), 133.7 (d, p-OBz), 130.1 (d, o-OBz),
129.1 (s, i-OBz), 128.7 (d, m-OBz), 128.8, 127.6, 125.0, 123.9
(d, C-3′′ + C-4′′ + C-5′′ +C-6′′), 84.2 (d, C-5), 82.2 (d, C-2′),
81.1 (s, C-4), 80.29 (s, NHBoc), 78.8 (s, C-1), 76.7 (t, C-20),
74.8 (d, C-2), 74.6 (d, C-10), 72.4 (d, C-13), 72.1 (d, C-7), 63.1
(d, C-3′), 58.7 (s, C-8), 46.4 (d, C-3), 43.3 (s, C-15), 43.2 (t, 2′-
(2′′)-CH2), 37.1 (t, C-6), 35.8 (t, C-14), 28.3 (q, Boc), 26.6 (q,
C-17), 21.7 (q, OAc), 20.72 (q, C-16), 14.6 (q, C-18), 9.8 (q, C-19).
because it is well-parametrized for
a range of organic
structures.29b At worst, the method dampens unrealistic gas-
phase dipole-dipole interactions. Under these conditions, the
9a search was run for a total of 5000 steps. The global
minimum was located 6 times. The set of 1121 optimized
conformers was treated to a NAMFIS conformer deconvolution
analysis28 by fitting 14 NOE-determined distances derived in
DMSO-d6. The NOE cross-peak volumes were scaled by
comparison with the volume of the H7-H10 cross-peak as-
signed a distance of 2.40 Å. The NAMFIS fitting procedure
provided seven conformations with predicted populations
ranging from 2-26%, i.e., 26, 21, 15, 15, 12, 10 and 2%, with
a sum of square differences (SSD)28 of 128. A Monte Carlo
torsional search was likewise performed for the six-membered
ring analogue 9b, this time, however, with the MM3*/GBSA/
CHCl3 prescription.29c The calculation delivered 328 conforma-
tions (global minimum found 49 times), which when combined
with seven NOEs and processed by a NAMFIS treatment
provided four conformations with estimated populations of 59,
35, 5 and 1% (SSD ) 5). Consistent with the polarity of the
solvent, the two most abundant conformers correspond to the
“nonpolar” Taxol variant, while the less abundant structures
are uncollapsed forms. All structures were visualized on SGI
O2 workstations and further manipulated in MacroModel: for
example, for evalutation of the ring-flip character of 9a A and
B as well as superposition of 9a conformers with those of Taxol,
Taxotere, and T-Taxol.
Refer en ces
(1) Three books have been devoted to the chemistry and medicinal
chemistry of Taxol: (a) Taxol: Science and Application; Suffness,
M., Ed.; CRC: Boca Raton, FL, 1995. (b) Taxane Anticancer
Agents: Basic Science and Current Status; Georg, G. I., Chen,
T. T., Ojima, I., Vyas, D. M., Eds.; American Chemical Society:
Washington, DC, 1995. (c) The Chemistry and Pharmacology of
Taxol and its Derivatives; Farina, V., Ed.; Elsevier: Amsterdam,
1995. For an earlier review, see: Nicolaou, KL. C.; Dai, W. M.;
Guy, R. K. Chemistry and Biology of Taxol. Angew. Chem., Int.
Ed. Engl. 1994, 33, 15-44.
(2) For major reviews, see: (a) Chen, S. H.; Farina, V. Paclitaxel
(Taxol) Chemistry and Structure-Activity Relationships. In The
Chemistry and Pharmacology of Taxol and its Derivatives;
Farina, V., Ed.; Elsevier: Amsterdam, 1995; pp 165-253. (b)
Georg, G. I. The Medicinal Chemistry of Taxol. In Taxol: Science
and Application; Suffnes, M., Ed.; CRC: Boca Raton, FL, 1995;
pp 317-375. (c) Wang, M.; Cornett, B.; Nettles, J .; Liotta, D.
C.; Snyder, J . P. The Oxetane Ring of Taxol. J . Org. Chem. 2000,
65, 1059-1068.
(3) For a review, see: Gue´nard, D.; Gue´ritte-Voegelein, F.; Potier,
P. Taxol and Taxotere: Discovery, Chemistry, and Structure-
Activity Relationships. Acc. Chem. Res. 1993, 26, 160-167.
(4) For a review, see: Ojima, I.; Lin, S.; Tao, W. Recent Advances
in the Medicinal Chemistry of Taxoids with Novel â-Amino Acid
Side Chains. Curr. Med. Chem. 1999, 6, 927-954.
(5) (a) Ojima, I.; Slater, J . C.; Kuduk, S. D.; Takeuchi, C. S.; Gimi,
R. H.; Sun, C.-M.; Park, Y. H.; Pera, P.; Veith, J . M.; Bernacki,
R. J . Syntheses and Structure-Activity Relationships of Taxoids
Derived from 14â-Hydroxy-10-deacetylbaccatin III. J . Med.
Chem. 1997, 40, 267-278. (b) Nicoletti, M. I.; Colombo, T.; Rossi,
2′,2′′-Eth ylid en ed oceta xel (24b). 21b (530 mg) was pro-
tected, hydrolyzed and coupled with 7,10-bistrichloroacetyl-
10-deacetylbaccatin III (23) as described for 21a . After depro-
tection, the final product was purified by CC (hexanes-EtOAc
25
1:1) to afford 9 mg 24b (4% from 22b) as a foam: [R]D -26.9
(CHCl3, c 0.5) (CHCl3, c 0.70); IR (KBr) 3450, 1740, 1715, 1660,
1640, 1240 cm-1; HRMS-FAB m/z 834.3695 (calcd for C45H55
-
NO14 + H+ 834.3701); 1H NMR (300 MHz, CDCl3) δ 8.13 (AA′
OBz), 7.58 (C OBz), 7.48 (BB′ OBz), 7.20 (m, H-3′′, H-4′′, H-5′′,
H-6′′), 6.38 (br t, J ) 9.0 Hz, H-13), 6.29 (s, H-10), 5.66 (d, J
) 7.2 Hz, H-2), 5.34 (d, J ) 10.1 Hz, H-3′), 5.18 (s, H-10), 5.12
(d, J ) 10.1 Hz, NH), 4.86 (br d, J ) 9.3 Hz, H-5), 4.25 (d, J
) 8.5 Hz, H-20R), 4.19 (m, H-7), 4.17 (d, J ) 8.5 Hz, H-20â),
3.91 (d, J ) 7.0 Hz, H-3), 3.20 (m, 2′,2′′-ethylidene), 2.86 (dd,
J ) 17.2, 4.5 Hz, 2′,2′′-ethylidene), 2.53 (m, H-6R), 2.38 (m,
2′,2′′-ethylidene), 2.27 (s, 4-OAc), 2.15 (m, H-14R,â), 1.91 (br
s, H-18), 1.82 (m, H-6â), 1.74 (s, H-19), 1.32 (s, Boc), 1.26 (s,
H-17), 1.11 (s, H-16); 13C NMR (75 MHz, CDCl3) δ 211.4 (s,

1586 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 10
Barboni et al.
C.; Monardo, C.; Stura, S.; Zucchetti, M.; Riva, A.; Morazzoni,
P.; Donati, M. B.; Bombardelli, E.; D’Incalzi, M.; Giavazzi, R.
IDN5109, a Taxane with Oral bioavailability and Potent Anti-
tumor Activity. Cancer Res. 2000, 60, 842-846.
and Organic Solvent. J . Am. Chem. Soc. 1993, 115, 11650-
11651. (d) Gomez Paloma, L.; Guy, R. K.; Wrasidlo, W.; Nicolaou,
K. C. Conformation of a Water-Soluble Derivative of Taxol in
Water by 2D-NMR Spectroscopy. Chem. Biol. 1994, 1, 107-112.
(e) Ojima, I.; Scott, D. K.; Chakravarty, S.; Ourevitch, M.; Begue´,
J . P. A Novel Approach to the Study of Solution Structures and
Dynamic Behavior of Paclitaxel and Docetaxel Using Fluorine-
Containing Analogues as probes. J . Am. Chem. Soc. 1997, 119,
5519-5527. (f) Milanesio, M.; Ugliengo, P.; Viterbo, D.; Appen-
dino, G. Ab Initio Conformational Study of the Phenylisoserine
Side Chain of Paclitaxel. J . Med. Chem. 1999, 42, 291-299. (g)
Li, Y.; Poliks, B.; Cegelski, L.; Poliks, M.; Gryczynski, Z.;
Piszczek, G.; J agtap, P. G.; Studelska, D. R.; Kingston, D. G. I.;
Schaefer, J .; Bane, S. Conformation of Microtubule-Bound
Paclitaxel Determined by Fluorescence Spectroscopy and RE-
DOR NMR. Biochemistry 2000, 39, 281-291. (h) Snyder, J . P.;
Nevins, N.; Cicero, D. O.; J ansen, J . The Conformation of Taxol
in Chloroform. J . Am. Chem. Soc. 2000, 122, 724-725. (i)
Snyder, J . P.; Nettles, J . H.; Cornett, B.; Downing, K. H.;
Nogales, E. Proc. Natl. Acad. Sci. U.S.A., in press.
(6) (a) Epothilones (Bollag, D. M.; McQueney, P. A.; Zhu, J .;
Hensens, O.; Koupal, L.; Liesch, J .; Goetz, M.; Lazarides, E.;
Woods, C. M. Epothilones, a New Class of Microtubule-Stabiliz-
ing Agents with a Taxol-Like Mechanism of Action. Cancer Res.
1995, 55, 2325-2333). (b) Discodermolide (ter Haar, E.; Kow-
alski, R. J .; Hamel, E.; Lin, C. M.; Longley, R. E.; Gunasekera,
S. P.; Rosenkranz, H. S.; Day, B. W. Discodermolide, a Cytotoxic
marine Agent that Stabilizes Microtubules more Potently than
Taxol. Biochemistry 1996, 35, 243-250). (c) Eleuthesides (Lindel,
T.; J ensen, P. R.; Fenical, W.; Long, B. H.; Casazza, A. M.;
Carboni, J .; Fairchild, C. R. Eleutherobin, a New Cytotoxin that
Mimics Paclitaxel (Taxol) by Stabilizing Microtubules. J . Am.
Chem. Soc. 1997, 119, 8744-8745). (d) Rhazinilam (Dupont, C.;
Gue´nard, D.; Tchertanov, L.; Thoret, S.; Gue´ritte, F. D-Ring
Substituted Rhazinilam Analogues: Semisynthesis and Evalu-
ation of Antitubulin Activity. Bioorg. Med. Chem. 1999, 7, 2961-
2969. (e) Laulimalides (Mooberry, S. L.; Tien, G.; Hernandez,
A. H.; Plubrukarn, A.; Davidson, B. S. Laulimalide and Isolauli-
malide, New Paclitaxel-Like Microtubule-Stabilizing Agents.
Cancer Res. 1999, 59, 653-660). (f) WS9885B (Muramatsu, H.;
Miyauchi, M.; Sato, B.; Yoshimura, S. 40th Symposium on The
Chemistry of Natural Products, Fukuoka, J apan, 1998; Paper
83, p 487. See also: Vanderwal, C. D.; Vosburg, D. A.; Weiler,
S.; Sorensen, E. J . Postulated Biogenesis of WS9885B and
Progress toward an Enantioselective Synthesis. Org. Lett. 1999,
1, 645-648). (g) Polyisoprenylated benzophenones (Roux, D.;
Hadi, H. A.; Thoret, S.; Gue´nard, D.; Thoison, O.; Pa¨ıs, M.;
Se´venet, T. Structure-Activity Relationship of Polyisoprenyl
Benzophenones from Garcynia pyrifera on the tubulin/Microtu-
bule System. J . Nat. Prod. 2000, 63, 1070-1076).
(7) (a) Winkler, J . D.; Axelsen, P. H. A Model for the Taxol
(Paclitaxel)/Epothilone Pharmacophore. Bioorg. Med. Chem. Lett.
1996, 6, 2963-2966. (b) Wang, M.; Xia, X.; Kim, Y.; Hwang, D.;
J ansen, J . M.; Botta, M.; Liotta, D. C.; Snyder, J . P. A Unified
and Quantitative Receptor Model for the Microtubule Binding
of Paclitaxel and Epothilone. Org. Lett. 1999, 1, 43-46. (c) Ojima,
I.; Chakravarty, S.; Inoue, T.; Lin, S.; He, L.; Horwitz, S. B.;
Kuduk, S. D.; Danishefsky, S. J . A Common Pharmacophore for
Cytotoxic Natural Products that Stabilize Microtubules. Proc.
Natl. Acad. Sci. U.S.A. 1999, 96, 4256-4261. (d) Giannakakou,
P.; Gussio, R.; Nogales, E.; Downing, K. H.; Zaharevitz, D.;
Bollbuck, B.; Poy, G.; Sackett, D.; Nicolaou, K. C.; Fojo, T. A
Common Pharmacophore foe Epothilone and Taxanes: Molec-
ular Basis for Drug Resistance Conferred by Tubulin Mutations
in Human Cancer Cells. Proc. Natl. Acad. Sci. 2000, 97, 2904-
2909. (e) He, L.; J agtap, P. G.; Kingston, D. G. I.; Shen, H.-J .;
Orr, G. A.; Horwitz, S. B. A Common Pharmacophore for Taxol
and the Epothilones Based on the Biological Activity of a Taxane
Molecule lacking a C-13 Side Chain. Biochemistry 2000, 39,
3972-3978.
(11) As pointed out in ref 10h, there appear to be three conforma-
tional extremes with respect to the orientation of the C13 side
chain of Taxol. The polar and apolar forms correspond to
hydrophobic collapse between the C2 benzoyl phenyl and the
C3′ phenyl and C2′ benzamide rings, respectively. We refer to
the third, uncollapsed, family as “extended” or “open” forms.
T-Taxol is one of these in which the centroid of the C2 phenyl is
nearly equidistant from the centroids of the two terminal phenyl
rings emanating for C3′.10i
(12) Ojima, J .; Lin, S.; Inoue, T.; Miller, M. L.; Borella, C. P.; Geng,
X.; and J ohn J . Walsh, J . J . Macrocycle Formation by Ring-
Closing Metathesis. Application to the Syntheses of Novel
Macrocyclic Taxoids. J . Am. Chem. Soc. 2000, 122, 5343-5353.
(13) Barboni, L.; Lambertucci, C.; Ballini, R.; Appendino, G.; Bom-
bardelli, E. Synthesis of a Conformationally Restricted Analogue
of Paclitaxel. Tetrahedron Lett. 1998, 39, 7177-7180.
(14) (a) Denis, J .-N.; Fkyerat, A.; Gimbert, Y.; Coutterez, C.; Man-
tellier, P.; J ost, S.; Greene, A. E. Docetaxel (Taxotere) Deriva-
tives: Novel NbCl3-Based Stereoselective Approach to 2′-
Methyldocetaxel. J . Chem. Soc., Perkin Trans. 1 1995, 1811-
1815. (b) Kant, J .; Schwartz, W. S.; Fairchild, C.; Gao, Q.; Huang,
S.; Long, B. H.; Kadow, J . F.; Langley, D. R.; Farina, V.; Vyas,
D. Diastereoselective Addition of Grignard Reagents to Azeti-
dine-2,3-dione: Synthesis of Novel Taxol Analogues. Tetrahedron
Lett. 1996, 37, 6495-6498.
(15) For a review, see: Kant, J . The Chemistry of the Taxol Side
Chain: Synthesis, Modifications and Conformational Studies.
In The Chemistry and Pharmacology of Taxol and its Derivatives;
Farina, V., Ed.; Elsevier: Amsterdam, 1995; pp 255-300.
(16) Li, G.; Agert, H. H.; Sharpless, K. B. N-Halocarbamate Salts
Lead to More Efficient Catalitic Asymmetric Aminohydroxyla-
tion. Angew. Chem., Int. Ed. Engl. 1996, 35, 2813-1817. For
an early application of the aminohydroxylation to the synthesis
of the Taxol side chain, see: Li, G.; Sharpless, K. B. Catalytic
Asymmetric Aminohydroxylation Provides a Short Taxol Side-
chain Synthesis. Acta Chem. Scand. 1996, 50, 649-651. A recent
and remarkable one-step synthesis of the isopropyl ester of the
Taxol side chain capitalizes on the asymmetric aminohydroxy-
lation (Song, C. E.; Oh, C. R.; Roh, E. J .; Lee, S.; Choi, J . H.
One-step Synthesis of Paclitaxel Side-chain Precursor: Benza-
mide-based Asymmetric Aminohydroxylation of Isopropyl trans-
Cinnamate. Tetrahedron: Asymmetry 1999, 10, 671-674).
(17) Seeman, J . I.; Chavdarian, C. G.; Secor, H. V. Synthesis of the
Enantiomers of Nornicotine. J . Org. Chem. 1985, 50, 5419-5421.
(18) J ohnson, R. A.; Nidy, E. G.; Dobrowolski, P. J .; Gebhard, I.;
Qualls, S. J .; Wicnienski, N. A.; Kelly, R. C. Taxol Chemistry.
7-O-Triflates as Precursors to Olefins and Cyclopropanes. Tet-
rahedron Lett. 1994, 35, 7893-7896. See also Upjohn PCT
Application WO 9413655, J une 23, 1994.
(19) (a) Denis, J .-N.; Kanezawa, A. M.; Greene, A. E. Taxotere by
Esterification with Stereochemically “Wrong” (2S,3S)-Phenyli-
soserine Derivatives. Tetrahedron Lett. 1994, 35, 105-108. (b)
Park, H.; Hepperle, M.; Boge, T. C.; Himes, R. H.; Georg, G. I.
Preparation of Phenolic Paclitaxel Metabolites. J . Med. Chem.
1996, 39, 2705-2709.
(20) Kolb, H. C.; VanNieuwenhze, M. S.; Sharpless, K. B. Chem. Rev.
1994, 94, 2483-2547. For the asymmetric dihydroxylation of
compounds related to 3a ,b, see: Nakajima, M.; Tomioka, K.;
Koga, K. Tetrahedron 1993, 49, 10807-10816.
(8) Rodi, D. J .; J anes, R. W.; Sanganee, H. J .; Holton, R. A.; Wallace,
B. A., Makowski, L. Screening of a Library of Phage-Displayed
Peptides Identifies Human bcl-2 as a Taxol-binding Protein. J .
Mol. Biol. 1999, 285, 197-203.
(9) This was first observed by Kingston (Samaranayake, G.; Magri,
N. F.; J itransri, C.; Kingston, D. G. I. Modified Taxols. 5.
Reaction of Taxol with Electrophilic Reagents and Preparation
of
a Rearranged Taxol Derivative with Tubulin Assembly
Activity. J . Org. Chem. 1991, 56, 5114-5119. See also: Chordia,
M. D.; Kingston, D. G. I.; Hamel, E.; Lin, C. M.; Long, B. H.;
Fairchild, C. A.; J ohnston, K. A.; Rose, W. C. Synthesis and
Biological Activity of A-nor-Paclitaxle Analogues. Bioorg. Med.
Chem. 1997, 5, 941-947). A partial separation of the tubulin
mechanism from the cytotoxic activity has been achieved with
some totally synthetic 15(11f12)abeo-2,11-cyclo-2,11-epoxy C-
aromatic taxoids derived from borneol (Klar, U.; Graf, H.;
Schenk, O.; Ro¨hr, B.; Schulz, H. New Synthetic Inhibitors of
Microtubule Depolymerization. Bioorg. Med. Chem. Lett. 1998,
8, 1397-1402). Among the biological analogues of Taxol, also
rhazinilam and the polyisoprenyl benzophenones xanthochymol
and guttiferone E show poor cytotoxicity (see refs 6d,g, respec-
tively).
(10) For leading studies in the area, see: (a) Dubois, J .; Gue´nard,
D.; Gue´ritte-Voegelein, F.; Guedira, N.; Potier, P.; Gillet, B.;
Beloeil, J . C. Conformation of Taxotere and Analogues Deter-
mined by NMR Spectroscopy and Molecular Modelling Studies.
Tetrahedron 1993, 49, 6533-6544. (b) Williams, H. J .; Scott, A.
I.; Dieden, R. A.; Swindell, C. S.; Chirlian, L. E.; Francl, M. M.;
Heerding, J . M.; Krauss, N. E. NMR and Molecular Modelling
Study of the Conformation of Taxol and its Side Chain Methyl
Ester in Aqueous and Non-Aquous Solution. Tetrahedron 1993,
49, 6545-6560. (c) Vander Velde, D. G.; Georg, G. I.; Grunewald,
G. L.; Gunn, C. D.; Mitscher, L. A. “Hydrophobic Collapse” of
Taxol and Taxotere Solution Conformations in Mixtures of Water
(21) Hu, Z.; Erhardt, P. W. Utilization of a Benzoyl Migration to
Effect an Expeditious Synthesis of the Paclitaxel C-13 Side
Chain. Org. Proc. Res. Dev. 1997, 1, 387-390.
(22) Koskinen, A. M. P.; Karvinen, E. K.; Siirila¨, J . P. Enantioselec-
tive Synthesis of the Taxol and Taxotere Side Chain. J . Chem.
Soc., Chem. Commun. 1994, 21, 22.

Taxol Analogues Conformationally Constrained at C13
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 10 1587
(23) Gue´ritte-Voegelein, F.; Gue´nard, D.; Mangatal, L.; Potier, P.;
Guilhem, J .; Cesario, M.; Pascard, C. Structure of a Synthetic
Taxol precursor: N-tert-Butoxycarbonyl-10-deacetyl-N-deben-
zoyltaxol. Acta Crystallogr. 1990, C46, 781-784.
(32) Boge, T. C.; Wu, Z.-J .; Himes, R. H.; Vander Velde, D. G.; Georg,
G. I. Conformationally Restricted Paclitaxel Analogues: Mac-
rocyclic Mimics of the “Hydrophobic Collapse” Conformation.
Bioorg. Med. Chem. Lett. 1999, 9, 3047-3052.
(33) Mayer, T. U.; Kapoor, T. M.; Haggarty, S. J .; King, R. W.:;
Schreiber, S. L.; Mitchison, T. J . Small Molecule Inhibitor of
(24) Mastropaolo, D.; Camerman, A.; Luo, Y.; Brayer, G. D.; Cam-
erman, N. Crystal and Molecular Structure of Paclitaxel (Taxol).
Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 6920-6924.
Mitotic Spindle Bipolarity Identified in
a Phenotype-Based
(25) Peterson, J . R.; Do, H. D.; Rogers, R. D. X-ray Structure and
Crystal Lattice Interactions of the Taxol Side-Chain Methyl
Ester. Pharm. Res. 1991, 8, 908-912.
(26) Gue´ritte-Voegelein, F.; Gue´nard, D.; Lavelle, F.; Le Goff, M.-T.;
Mangatal, L.; Potier, P. Relationship Between the Structure of
Taxol Analogues and Their Antimitotic Activity. J . Med. Chem.
1991, 34, 992-998.
Screen. Science 1999, 286, 971-974. For a partial description
of the structure, see: Chem. Eng. News 1999, J uly 26, 44-46.
(34) Miglietta, A.; Bocca, C.; Gadoni, E.; Gabriel, L.; Rampa, A.; Bisi,
A.; Valenti, P.; Da Re, P. Interaction of Geiparvarin and Related
Compounds with Purified Microtubular Protein. Anti-Cancer
Drug Des. 1996, 11, 35-48.
(27) Wang, M.; Cornett, B.; Nettles, J .; Snyder, J . P. Unpublished.
(28) (a) Cicero, D. O.; Barbato, G.; Bazzo, R. NMR Analysis of
Molecular Flexibility in Solution: A New Method for the Study
of Complex Distributions of Rapidly Exchanging Conformations.
Application to a 13-Residue Peptide with an 8-Residue Loop. J .
Am. Chem. Soc. 1995, 117, 1027-1033. (b) Nevins, N.; Cicero,
D.; Snyder, J . P. A Test of the Single-Conformation Hypothesis
in the Analysis of NMR Data for Small Polar Molecules: A Force
Field Comparison. J . Org. Chem. 1999, 64, 3979-3986.
(29) (a) Mohamadi, F.; Richards, N. G. J .; Guida, W. C.; Liscamp,
R.; Lipton, M.; Caufield, C.; Chang, G.; Hendrickson, T.; Still,
W. C. MacroModel - an integrated software system for modeling
organic and bioorganic molecules using molecular mechanics.
J . Comput. Chem. 1990, 11, 440. (b) Still, W. C.; Tempczyk, A.;
Hawley, R. C.; Hendrickson, T. Semianalytical Treatment of
Solvation for Molecular Mechanics and Dynamics. J . Am. Chem.
Soc. 1990, 112, 6127-6129. (c) cf. http://www.schrodinger.com/
macromodel2.html.
(30) Anisotropic shifts of this type are well-documented in taxoids,
the most striking example being the negative (-0.21 ppm) (!)
chemical shift value of H-6R observed for Taxine A in CDCl3
(Barboni, L.; Gariboldi, P.; Appendino, G.; Enriu`, R.; Gabetta,
B.; Bombardelli, E. New Taxoids from Taxus baccata L. Liebigs
Ann. 1995, 345-349). Like Taxol, also Taxine A adopts different
conformations in polar and apolar media.
(31) See for instance: Ali, S. M.; Hoemann, M. Z.; Aube´, J .; Mitscher,
L. A.; Georg, G. I.; McCall, R.; J ayasinghe, L. R. Novel Cytotoxic
3′-(tert-Butyl) 3′-Dephenyl Analogues of Paclitaxel and Doc-
etaxel. J . Med. Chem. 1995, 38, 3821-3828.
(35) (a) Sengupta, S.; Boge, T. C.; Georg, G. I.; Himes, R. H.
Interaction of a Fluorescent Paclitaxel Analogue with Tubulin.
Biochemistry 1995, 34, 11889-11894. (b) See ref 31, except that
the MCF-7 breast cancer cell line was used in the cytotoxicity
assay.
(36) Brown, D. S.; Marples, B. A.; Smith, P.; Walton, L. Epoxidation
with Dioxiranes Derived from 2-Fluoro-2-substituted-1-tetral-
ones and -1-Indanones. Tetrahedron 1995, 51, 3587-3606.
(37) Murphy, J . A.; Patterson, C. W. Regioselectivity of Radical-
Induced Bond Cleavages in Epoxides. J . Chem. Soc., Perkin
Trans 1 1993, 405-410.
(38) Denis, J .-N.; Greene, A. E.; Gue´nard, D.; Gue´ritte-Voegelein, F.;
Mangatal, L.; Potier, P. A Highly Efficient Practical Approach
to Natural Taxol. J . Am. Chem. Soc. 1988, 110, 5917-5919.
(39) Gue´ritte-Voegelein, F.; Se´nilh, V.; David, B.; Gue´nard, D.; Potier,
P. Chemical Studies of 10-Deacetyl Baccatin III. Hemisynthesis
of Taxol Derivatives. Tetrahedron 1986, 42, 4451-4460.
(40) Chang, G.; Guida, W. C.; Still, W. C. An Internal Coordinate
Monte Carlo Method for Searching Conformational Space. J . Am.
Chem. Soc. 1989, 111, 4379-4386.
(41) Saunders: M.; Houk, K. N.; Wu, Y.-D.; Still, W. C.; Lipton, M.;
Chang, G.; Guida, W. C. Conformations of Cycloheptadecane. A
Comparison of Methods for Conformational Searching. J . Am.
Chem. Soc. 1990, 112, 1419-1427.
J M001103V