New para-para stilbenophanes: Synthesis by mcmurry coupling, conformational analysis and inhibition of tubulin polymerisation

Article abstract of DOI:10.1002/chem.201002869

The synthesis of a new family of methoxy-substituted [2.7]- and [2.8]paracyclophanes linked by 3-oxapentamethylene-1,5-dioxy and hexamethylene-1,6-dioxy bridges has been carried out by using the McMurry methodology. Related indole compounds were also synthesised. Olefin-to-diol ratios depended on the bridge length, the structure of the aromatic ring and the reaction conditions. Macrocyclisation, the methoxy substituents and the presence of a rigid indole moiety restricted the conformational equilibria, as observed by NMR spectroscopy and according to theoretical calculations. The synthesised compounds display micromolar tubulin polymerisation inhibitory activity. The conformational implications on the tubulin polymerisation inhibitory activity derived from the macrocyclisation when compared with combretastatins, closely related stilbenes, are also discussed. Flip-flop: McMurry coupling of dialdehydes was successfully used to complete the synthesis of macrocyclic stilbene analogues of combretastatins in variable yields (an example of which is depicted). These molecules have several exchanging conformations, as studied by NMR spectroscopy and theoretical calculations, and are the first macrocyclic inhibitors of tubulin polymerisation of their class. Copyright

Full text of DOI:10.1002/chem.201002869

DOI: 10.1002/chem.201002869
New para–para Stilbenophanes: Synthesis by McMurry Coupling,
Conformational Analysis and Inhibition of Tubulin Polymerisation
Raquel ꢀlvarez, Vilmarꢁ Lꢂpez, Carmen Mateo, Manuel Medarde,* and
Rafael Pelꢃez*[a]
Abstract: The synthesis of
a
new
rocyclisation, the methoxy substituents
and the presence of a rigid indole
moiety restricted the conformational
equilibria, as observed by NMR spec-
troscopy and according to theoretical
calculations. The synthesised com-
pounds display micromolar tubulin
polymerisation inhibitory activity. The
conformational implications on the tu-
bulin polymerisation inhibitory activity
derived from the macrocyclisation
when compared with combretastatins,
closely related stilbenes, are also dis-
cussed.
family of methoxy-substituted [2.7]-
and [2.8]paracyclophanes linked by 3-
oxapentamethylene-1,5-dioxy and hex-
amethylene-1,6-dioxy bridges has been
carried out by using the McMurry
methodology. Related indole com-
pounds were also synthesised. Olefin-
to-diol ratios depended on the bridge
length, the structure of the aromatic
ring and the reaction conditions. Mac-
Keywords: combretastatin
ana-
logues · conformation analysis · cy-
clophanes · structure–activity rela-
tionships · tubulin polymerization
inhibition
Introduction
active derivatives in different cell assays.[11] Although it is a
very interesting approach for modulating the activity–selec-
tivity of analogues based on highly potent lead compounds,
it has not yet been applied to active stilbenes and we have
only recently published preliminary work addressing this
issue.[12] The compounds designed for this purpose are stil-
bene and dihydrostilbene derivatives conformationally
blocked by the formation of macrocyclic structures, which
can be classified as cyclophanes,[13] in general, or stilbeno-
phanes, in particular.[14] Stilbenophanes have recently re-
ceived renewed interest by Rajakumarꢀs group,[15] who have
studied the potential applicability and that of related indolo-
phanes[16] in supramolecular chemistry.
The synthetic approach to these stilbene and dihydrostil-
bene cyclophanes can be planned through different ap-
proaches; intramolecular McMurry coupling is the preferred
tool to produce the macrocyclisation step. During our re-
search into new antimitotic agents based on natural prod-
ucts,[17] we described the application of the McMurry meth-
odology[12a] to the synthesis of a new family of doubly sym-
metric (two symmetric benzene rings) stilbenophanes based
on deoxycombretastatin A-4 (Scheme 1),[18] the preparation
Stilbenes are widely studied substructures in organic chemis-
try owing to their presence in a large number of organic
compounds with applications in different fields. One point
of interest is the biological activity displayed by stilbene de-
rivatives,[1] as in the case of the anticancer agents combreta-
ACHTUNGTRENNUNG
statins[2] and resveratrol.[3] Considerable efforts have been
invested in the preparation of analogues of these com-
pounds to unveil the structure–activity relationships and im-
prove the pharmacological profile by increasing the activity
and selectivity.[4] In this sense, the prodrug combreta
ACTHNUTRGNEsNUG tatin
A-4 phosphate[5] appears to be a good drug candidate,
owing to its vascular disrupting activity and adequate water
solubility. Currently, it is in phase II/III clinical trials for the
treatment of anaplastic thyroid cancer and solid tumours.[6]
Conformational restriction is a method for exploring the
influence of molecular geometry on physicochemical and
biological properties[7] that can be achieved by several ap-
proaches, including the macrocyclisation of open-chain
active models.[8] This has been applied to diverse types of
compounds,[9] such as bisindole derivatives,[10] and it has re-
cently been used as a tool for a systematic search for new
[a] Dr. R. ꢁlvarez, Dr. V. Lꢂpez, Dr. C. Mateo, Prof. Dr. M. Medarde,
Dr. R. Pelꢃez
Laboratorio de Quꢄmica Orgꢃnica y Farmacꢅutica
Facultad de Farmacia. Universidad de Salamanca
Campus Miguel de Unamuno, 37007, Salamanca (Spain)
Fax : (+34)923-294515
E-mail: pelaez@usal.es
medarde@usal.es
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201002869.
Scheme 1. Comparison of the structures of previously described stilbeno-
phanes with deoxycombretastatin A-4.
3406
ꢆ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 3406 – 3419

FULL PAPER
the linker structure. The use of
polymer-bound triphenylphos-
phine in the Mitsunobu reac-
tion is recommended to avoid
the presence of phosphine
oxides in the coupling step
mediated by titanium,[21] be-
cause small amounts of phos-
phorous materials elicit a de-
crease in the yields or the ab-
sence of isolable reaction prod-
ucts.
The McMurry olefinations[22]
and pinacol couplings[23] pro-
duced olefins, diols or mixtures
of both types of compounds, de-
pending on the specific condi-
Scheme 2. Comparison of the structures of synthesised stilbenophanes and related indolophanes with the
potent antimitotic agents combretastatin A-4 and deoxycombretastatin A-4.
of some dihydrodihydroxy members of this class of com-
pounds,[12b,d] and we also reported the results of the biologi-
cal assays carried out on some of them.[12c]
Herein, we describe the results obtained upon applying
this chemistry to the preparation of macrocyclic analogues
of combretastatin A-4 (Scheme 2), differently substituted at
the A or B rings, and the conformational analysis of the pro-
duced olefins. The conformational equilibria for these com-
pounds are characterised by the coexistence of differently
preferred conformations in rapid or intermediate exchange.
The existence of favoured conformations and the dynamics
of the conformational equilibria of these compounds can
produce a noticeable effect on their interaction with their
targets and on their cytotoxic activity. A better knowledge
of the spatial arrangement adopted by these molecules can
facilitate the design of new analogues with improved poten-
cy in comparison with the stilbenophanes and indolophanes
already obtained.[12c]
The detailed description and conformational analysis of
diols and their derivatives will be published elsewhere. The
successful synthesis of unsymmetrical stilbenophanes and in-
dolophanes opens the possibility of studying the applicabili-
ty of these fascinating compounds (in terms of their confor-
mational mobility) in supramolecular and medicinal chemis-
try.
Scheme 3. Synthesis of intermediate dialdehydes 1, 2 and 3.[12d] Reagents
and conditions: a) 1) PPh3 resin, aldehyde, CH2Cl2, RT, 1–3 h; 2) di(tert-
butyl) azodicarboxylate (DBAD)/diisopropyl azodicarboxylate (DIAD),
RT, 48 h; b) NaOH, 1,6-dibromohexane, CH2Cl2/H2O 1:1, Bu4NF, RT,
48 h; c) p-hydroxybenzaldehyde, K2CO3, dry DMF, 608C, 48 h, Ar;
d) NaOH, 1,6-dibromohexane, CH2Cl2, Bu4NSO4H, RT, 24 h; e) vanillin
or syringaldehyde, K2CO3, dry DMF, 708C, 48 h, Ar.
tions and structures. Although the yields varied from moder-
ate to high, most of the olefins were isolated in sufficient
quantities to characterise and assay them (Scheme 4).
Under the employed conditions, no dimeric products (as
represented by the general
structure 12) were isolated in
any of the syntheses carried
out, but they were detected as
minor products from dialde-
Results and Discussion
The synthesis of the designed molecules was carried out ac-
cording to the approaches described in Schemes 3 and 4:
two consecutive alkylations of suitable phenolic (indolic) al-
dehydes with the required reactive linker moiety produced
the intermediate dialdehydes (Scheme 3), which were sub-
jected to McMurry internal coupling at room temperature
or under reflux (Scheme 4). The alkylation processes were
performed by the Mitsunobu[19] or catalysed phase-trans-
fer[20] methodologies, depending on the particular cases and
hyde 8.
The product ratios (as esti-
mated by 1H NMR spectrosco-
py) for each crude reaction
Chem. Eur. J. 2011, 17, 3406 – 3419
ꢆ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3407

M. Medarde, R. Pelꢃez et al.
Complete resonance assign-
ments for all of the macrocycles
were achieved by 2D NMR
spectroscopy (COSY, HMQC,
HMBC and ROESY). Diols
and their acetates will be stud-
ied in depth in forthcoming
papers.
Conformational
analysis:[25]
Previously studied macrocyclic
olefins of this class with two
symmetrically
substituted
phenyl rings were seen to have
simple NMR spectra, with
signal averaging of the chemi-
cally equivalent pairs, both in
1
the H and the 13C NMR spec-
tra.[12a,b] In that case, the chemi-
cal exchange of these nuclei
took place readily through heli-
cal changes (double flip, see
Scheme 5)[26] that occur without
the passage of any of the ring
planes through the intra-annu-
lar space, represented by the
double-bond plane.
Deeper insight into the con-
formational equilibria in which
these kinds of compounds are
involved can be obtained from
Scheme 4. Reagents and conditions: a) 1) TiCl4·THF (97%), Zn, dry THF, 08C, refluxed for 30 min; 2) dialde-
hyde, dry THF, 2–24 h, reflux; b) same as a) but 08C or RT, 5–24 h. Products marked with * were not isolated
after chromatography.
Table 1. Results of McMurry cyclisation reactions in Scheme 4, produced under the conditions shown.
[a]
Entry Starting material Reaction conditions Crude yield [%] Product (ratio)
Olefin (yield [%])[b]
1
2
3
4
5
6
7
8
9
4
5
5
6
8
8
10
10
11
5 h; reflux
23 h; reflux
5 h; 08C
24 h; reflux
24 h; reflux
24 h; RT
5 h; reflux
2 h; reflux
5 h; reflux
94
13b/13c (1/1)
–
14b/14c (1/3)
–
[c]
complex mixture
100
complex mixture 15a/15b/15c (1/1/2) 15a (15)
complex mixture
complex mixture
97
70
100
14a (40)
–
[c]
–
–
16a (7)
–
the
compounds
described
[c]
herein, which have at least one
non-symmetrically substituted
aromatic ring. In each case, a
single set of signals is observed
17a/17b/17c (1/2/2) 17a (10)
17a/17b/17c (1/2/2) 17a (7)
18a/18b/18c (3/2/1) 18a (30)
1
[a] Product ratios of the crude mixture were estimated by integration of representative signals in the H NMR
spectra of the crude reaction products. [b] Yield obtained after isolation by flash chromatography. [c] Not es-
1
tablished by H NMR spectroscopy.
mixture and the isolated yields of the olefins are indicated
in Table 1. Olefins (a) were always produced under reflux
(Table 1, entries 2, 4, 5, 7, 8 and 9), except for compound 4
(Table 1, entry 1), but they were not detected in the crude
reaction products when the reactions were carried out at
lower temperatures (Table 1, entries 3 (08C) and 6 (RT)).
Diols (b, c) were produced at lower temperatures (Table 1,
entries 3 and 6), whereas under reflux they were either not
detected (Table 1, entries 2 and 5) or were produced in dif-
ferent ratios (Table 1, entries 1, 4, 7, 8 and 9). The results of
the McMurry reaction were variable depending on the reac-
tion conditions and substrates employed,[24] as observed in
Scheme 5. Definitions and schematic representations of the phenyl ring
motions for the stilbene and the associated helicity changes.[27] Left: one-
ring flip and one-ring rotation, which is a disrotatory movement exempli-
fied herein from the (ꢀ)-helix (at the top) to the (+)-helix (at the
bottom). Right: the two-ring flip, referred to herein as a double flip, is a
conrotatory movement.
Table 1.
The structures of the olefins were readily established by
1H NMR spectroscopy, which showed the presence of the cis
double bond and the usual appearance of only one signal
for the chemically equivalent protons at room temperature.
3408
www.chemeurj.org
ꢆ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 3406 – 3419

para–para Stilbenophanes
FULL PAPER
for compounds with either two non-symmetrical rings (17a)
or with one symmetrical ring (14a, 15a, 16a and 18a). In
the case of the latter group, signal averaging of the chemi-
cally equivalent nuclei of the symmetrical ring is usually ob-
served. This can only be explained by either the existence of
a single set of dominating conformations (see below) or by a
conformational exchange involving the passage of ring
planes through the intra-annular space.
The relevant conformational classes, as determined by
molecular mechanics calculations, are depicted in Scheme 6.
The combination of two helical dispositions (+ and ꢀ), with
ent surroundings in each case. The same applies for confor-
mations II and IV.
As shown in Scheme 6, rotation of the aromatic rings can
interconvert the conformations. Ring rotations (L- for the
ring on the left and R- for that on the right) pass the corre-
sponding ring planes through the intra-annular space, while
flip movements pass them through the normal to the
double-bond plane. Usually, flipping or rotating one aromat-
ic ring causes the other ring to flip, in both cases resulting in
a change in helicity (see Scheme 5 for stilbene).[26,28] One-
ring rotation accompanied by one-ring flip (Scheme 5, left;
produced by R-ring rotation
and L-ring flip) changes the ori-
entation of the marked posi-
tions only for the rotating R-
ring, from above to below. A
two-ring flip (Scheme 5, right)
causes no change in their dispo-
sition, both still lying above the
plane after the helix change. Si-
multaneous rotation of both
rings (zero-ring flip, not drawn)
would require an improbable
co-planar disposition of both
rings. In Scheme 6, the confor-
mations are switched horizon-
tally by rotation of the L-ring
(accompanied by R-ring flip);
vertically by rotation of the R-
ring (accompanied by L-ring
flip), and on the external diago-
nals by double flips. Fast rota-
tion of the R- and/or L-rings
through the macrocycle and/or
double flips are required for
the appearance of chemically
Scheme 6. Conformational equilibria for olefins 14a (X=CH2CH2, R=H, Y=OMe), 15a (X=O, R =Me,
Y=H) and 16a (X=CH2CH2, R=Me, Y=H), with one symmetric ring. A schematic representation is also in-
cluded that shows the planes of the double bond (horizontal line) and both aromatic rings (vertical lines), to-
gether with an indication of the helicity (vertical lines angled from left to right indicate a (ꢀ)-helix and from
equivalent positions as a single
signal in NMR spectra (see
Tables 2 and 3).
As stated above, most of the
NMR spectra of these macro-
*
right to left indicate a (+)-helix). Marked positions ( ) facilitate the visualisation of their location above or
below the double-bond plane.
AHCTUNGTRENNUNG
two different situations for the marked position in each ring
(above and below the plane of the double bond), gives rise
to 2ꢇ2ꢇ2=8 conformational populations. Four populations
are mirror images of the other four and only differ in their
helical disposition, and they are indistinguishable from the
NMR spectroscopy point of view. Accordingly, the eight
conformations were labelled I to IV preceded by their heli-
cal sign to simplify the identification of the pairs of mirror
images.
Furthermore, for compounds with one symmetric and one
non-symmetric phenyl ring, such as 14a (X=CH2-CH2, R=
H, Y=OMe), 15a (X=O, R =Me, Y=H) and 16a (X=
CH2-CH2, R=Me, Y=H), conformations I and III are indis-
tinguishable, but the marked positions are located in differ-
non-axially symmetric rings, the presence of only one set of
signals requires a fast exchange between four populations
(I–IV), the presence of a single dominating conformational
population, or intermediate situations (e.g., two rapidly ex-
changing major conformations). For compounds with one
axially symmetric ring, such as 14a, 15a, 16a or 18a, there
are only two types of conformational populations that
would generate different NMR spectra, namely, ꢁI/ꢁIII
and ꢁII/ꢁIV. The presence of a single set of signals in the
NMR spectra requires either the prevalence of one of the
groups or a fast exchange between individual members of
the two groups, as occurs in stilbene. Theoretical calcula-
tions suggest that both groups of conformations are of simi-
lar stability. Furthermore, the signal averaging usually ob-
Chem. Eur. J. 2011, 17, 3406 – 3419
ꢆ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3409

M. Medarde, R. Pelꢃez et al.
Table 2. Independent equilibria remaining if only two motions (as indicated
in the upper row) are fast on the NMR spectroscopy timescale.[a]
served for the chemically equivalent pairs of nuclei requires
a fast exchange of their positions, thus introducing more re-
quirements into the kinematics of the system. It is interest-
ing to note that such requirements would not have been evi-
dent if we had only analysed the previously described sys-
tems with two axially symmetric rings. In the following dis-
cussion, the possible situations accounting for the observed
NMR spectra are first explored, followed by an analysis of
the NMR spectra at different temperatures, and finally, the
implications for the tubulin polymerisation inhibitory activi-
ties are discussed.
Fast L-ring rotation
Fast R-ring rotation
Slow double flip
Fast R-ring rotation
Fast double flip
Slow L-ring rotation
Fast L-ring rotation
Fast double flip
Slow R-ring rotation
Regardless of the nature of the aromatic rings, if the three
types of ring motions shown in Scheme 6 occur rapidly on
the NMR timescale, an averaging of the signals for the four
relevant species (I–IV) and the chemically equivalent pro-
tons (in the particular case of axially symmetric rings) would
one set of signals
one set of signals
one set of signals
[a] The horizontal equilibrium arrows indicate conversions caused by the explain the simple NMR spectra observed. However, more
first fast motion and the vertical ones indicate conversions caused by the
second fast motion. The endo or exo dispositions of the substituent for the
trisubstituted L-rings are indicated by RO-endo/exo just below the confor-
mation number (I–IV) and the corresponding disposition of chemically
stringent situations could also account for the observed re-
sults.
If only two of the three motions were fast on the NMR
timescale, full agreement with the observed signal averaging
*
*
equivalent positions (labelled ( ) and unlabelled ( )) of the symmetrical
rings are indicated underneath. The lower row shows an indication of what
would be observed in the NMR spectra.
would also be produced (Table 2) because the conforma-
tions (ꢁI–ꢁIV) rapidly exchange.
If only one of the three motions were fast on the NMR
timescale, we should observe either two different sets of sig-
nals or two different signals for chemically equivalent pairs
(Table 3 and Scheme 6). To observe a single set of signals
when unsymmetrical L-ring rotation is the only fast motion,
one of the equilibria should predominate over the other. For
the other two cases, the only explanation for a single signal
for chemically equivalent pairs of nuclei would be an acci-
dental degeneration of the signals, arising from similar
chemical environments. This explanation seems more likely
for the fast double flip because when the R-ring rotation is
the fast process the differences caused by the ring-current
effects seem to be too important to be neglected.
Table 3. Representation of the independent equilibria (separated by
dotted lines) remaining if only one motion (as indicated in the upper
row) is fast on the NMR timescale.[a]
Fast L-ring rotation
Slow R-ring rotation Slow L-ring rotation
Slow double flip Slow double flip
Fast R-ring rotation
Fast double flip
Slow R-ring rotation
Slow L-ring rotation
Similarly, eight conformational populations can be pro-
posed for the macrocyclic indole derivatives 17a (Y=H)
and 18a (Y=OMe). The eight conformations have also
been labelled I to IV preceded by their helical sign, as de-
picted in Scheme 7. Owing to the symmetry of the R-ring in
compound 18a, conformations I and III are indistinguisha-
ble, and so are II and IV, but I–IV are different in com-
pound 17a. The same motions depicted in Scheme 5 are re-
sponsible for the conformational transitions. The same con-
siderations made for the appearance of the NMR spectra of
14a, 15a and 16a (see above) apply to 18a. For 17a, with its
two asymmetrically substituted rings, the predominating
conformational populations must exchange rapidly on the
NMR timescale to produce the observed NMR spectra.
This simplified vision of the conformational equilibria for
these compounds becomes more complex if the interactions
and movements of the bridge (3-oxapentamethylene and
hexamethylene) are taken into account. To gain deeper in-
sight into the conformational equilibria for 14a–18a and the
relative ease of the conformational exchanges involved in
the appearance of the NMR spectra, we studied these sys-
two complete sets of one set of signals with one set of signals with
*
signals
different exo and endo different signals for
*
RO-exo¼RO-endo
signals for symmetric
and on the same or
different side from RO
*
*
*
*
-exo/endo¼ -exo/
D-ring ( and )
endo
[a] The endo or exo dispositions of the substituent for the trisubstituted
L-rings are indicated by RO-endo/exo immediately below the conforma-
tion number (I–IV) and the corresponding disposition of chemically
*
*
and unlabeled ) of the symmetrical
equivalent positions (labelled
rings are indicated below. The lower row shows an indication of what
*
*
averaged means that
would be observed in the NMR spectra:
and
chemically equivalent positions of the symmetric ring (R-ring in
Scheme 6) would be observed as a single signal and RO-endo¼exo
means that different sets of signals would be observed for the conforma-
tions with the endo L-ring substituent and the exo L-ring substituent (fast
L ring rotation, left column). RO averaged means that the substituted L-
ring would lead to only one set of signals but a differentiation of chemi-
cally equivalent positions of R-ring would occur (center and right col-
umns)..
3410
www.chemeurj.org
ꢆ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 3406 – 3419

para–para Stilbenophanes
FULL PAPER
dominating over the ꢁII/ꢁIV
classes, which were 7.7 kJmolꢀ1
higher in energy. For 17a, the
relative stabilities of I–IV were
1.1, 8.8, 0.0 and 1.7 kJmolꢀ1, re-
spectively. For all of the com-
pounds considered, conforma-
tions close to the double flip
were also frequently found.
Moreover, the conformations
close to ring rotations were
scarcely populated. According-
ly, we considered all of the con-
formational groups in the dis-
cussions of the conformational
equilibria (see the Supporting
Information, p 9, for a summary
of the conformations found for
17a and 18a in the Monte
Carlo searches).
The X-ray crystal structure
obtained
for
olefin
18a
(Figure 1) shows the ꢀI/ꢀIII
conformation in the solid state,
in good agreement with the cal-
Scheme 7. Conformational equilibria for 17a (Y=H), lacking symmetric rings, and 18a (Y=OMe), with one
symmetric ring. Fast rotation of R- and/or I-rings through the macrocycle and/or double flip are required for
the appearance of chemically equivalent positions as a single signal in NMR spectra.
tems using molecular mechanics and molecular dynamics
simulations, and compared the results with those obtained
for related model systems and previously described com-
pounds.
Molecular mechanics: To establish the relevant conforma-
Figure 1. Structure obtained by X-ray diffraction of crystals of olefin 18a.
tions for compounds 14a–18a, Monte Carlo conformational
searches using molecular mechanics (MM3 force field with
chloroform as the solvent) were carried out.[29] The confor-
mations found within 10 kJmolꢀ1 of the global minimum
were considered relevant for the conformational equilibria
studied, and automatically assigned to the conformational
classes ꢁI to ꢁIV, based on the dihedral angles measured
between the ring planes and the double-bond plane (see the
Supporting Information, pp. 7–8, for dihedral angles and
conformational classes definitions). Additional categories
were defined to include conformations representing states
close to transitions between two classes, such as those with
mismatched helical signs for the two aromatic rings, or those
in which the dihedral angles of one or both ring planes with
the double-bond plane are close to (ꢁ108) the change of
quadrants (0, ꢁ90, and ꢁ1808).
Ellipsoids are shown at the 30% probability level.[38]
culations. These conformations also show dihedral angles
between the planes of both rings and a double bond similar
to that found in stilbene.[26,27]
Molecular dynamics: The conformational behaviour of these
macrocyclic stilbenes was investigated by means of molecu-
lar dynamics (MD) simulations at virtual temperatures of
300, 400, 600, 1000, 1200 and 1500 K, with simulation times
of 3 ns (see the Supporting Information, pp. 12–17). The mo-
tions of the hexamethylene- and 3-oxapentamethylene
spacers and the aromatic rings in the resulting trajectories
were analysed in the same way as the results of the Monte
Carlo conformational searches. When the observed confor-
mational class changes were assignable to a single ring rota-
tion, as depicted in Schemes 6 and 7, the transitions were la-
belled accordingly (e.g., if along the trajectory for 18a, a
conformation belonging to class +III was followed by an-
other belonging to ꢀI, such a transition was labelled as
indole-ring rotation, according to the upper row of
Scheme 7). When more than one transition had to be in-
The most stable conformations for compounds 14a and
15a belong to the ꢁI/ꢁIII group, whereas for 16a they
belong to the ꢁII/ꢁIV class. The energy differences with
the alternative conformational groups are 7.3, 2.0 and
3.2 kJmolꢀ1, respectively. For 18a, despite the rigidity intro-
duced into the system by the indole ring, the situation is
similar to that of 14a and 15a, with the ꢁI/ꢁIII group pre-
Chem. Eur. J. 2011, 17, 3406 – 3419
ꢆ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3411

M. Medarde, R. Pelꢃez et al.
voked to explain the changes, no attempt was made to
assign the motions involved.
the conformations corresponding to transitions among con-
formations I–IV. However, the conformations observed for
compounds 14a, 15a, 17a and 18a are the same as those at
300 K. The most remarkable difference was observed for
compound 16a at 1200 K (Figure 2). During the 3 ns simula-
For comparative purposes, identical simulations were car-
ried out on XIX and XX, non-macrocyclic analogues of 17a
and 18a, in which the hexamethylene spacers had been re-
placed by two methyl groups (see the Supporting Informa-
tion, pp. 10–11, for the calculated simulations). The latter
analogue is a very potent inhibitor of tubulin polymeri-
sation, comparable with combretastatin A-4.[17d] At 300 K,
both non-macrocyclic analogues sample all of the conforma-
tional populations (ꢁI to ꢁIV); the I/III group is more
populated than the II/IV group for both compounds. These
preferences can be explained first in terms of the higher sta-
bility of endo rather than exo-methoxy groups (e.g., III is
more stable than I for XIX) and, second, by a disfavoured
simultaneous endo disposition of both the methoxy and
indole (pyrrole moiety) substituents (II less favoured). At
300 K, conformations close to transitions between I and IV
are also observed; the commonest ones are those corre-
sponding to ring rotations. These conformations, stabilised
by conjugation of the rotating ring with the double bond,
were rarely observed for the macrocyclic analogues (see
below), probably due to unfavourable contacts with the
spacer. With the simulation lengths employed, it is already
apparent that the compound with two methoxy groups meta
to the olefinic bridge presents more difficulties for the rota-
tion of the other ring (indole). As expected, simulations at
higher temperatures led to more conformations apart from I
to IV and to more conformational transitions. However, the
relative populations of the conformational groups and tran-
sition frequencies were conserved.
Figure 2. Conformations found along the trajectories of the molecular dy-
namics simulations for olefins 15a (top) and 16a (bottom) at 1200 K. For
the latter, the trisubstituted ring rotation is evidenced by the upper to
lower half crossover.
tion, double flips were often observed within the III/ꢀIV
conformational ensemble, followed by a transition from the
III/ꢀIV to the ꢀI/II ensemble, and later by a reversal of the
process. Experimental support for this result was provided
by the 13C NMR spectra for compound 15a, which showed
two signals for each pair of equivalent carbon atoms of the
disubstituted phenyl ring (Figure 3, on the left), whereas
only one signal for each pair of equivalent carbon atoms of
the disubstituted phenyl ring was observed for 16a
(Figure 3, on the right), according to the ease of exchange
between conformational ensembles.
The molecular dynamics simulations at 300–600 K for the
macrocyclic compounds 14a–18a are very similar to one an-
other. For all of them, the only observed transitions were
double flips. For each compound, the relative stabilities of
the conformations determined their populations in the mo-
lecular dynamics trajectories. Unlike the non-macrocyclic
cases described above, the preferred conformations for 14a–
16a were those with an exo disposition of the substituent of
the trisubstituted ring, due to unfavourable interactions of
the endo ring substituents with the spacer. The situation is
also similar for indolic macrocycles 17a and 18a, with con-
formation I predominating over II. The preference for an
exo disposition of the methoxy groups explains the greater
energy difference for the two conformations found in the
former than in the latter. Intermediate conformations corre-
sponding to double flips were also frequently found along
the trajectories.
The situation described for the 15a and 16a pair is analo-
gous to the case of related macrocyclic diols, in which ring
flipping was observed for
a para-disubstituted phenyl
ring.[12d] However, in the present case, the observed ring flip
was not for the disubstituted phenyl, but for the trisubstitut-
ed one (Figure 4). This observation suggests that in these
systems the interaction of the substituents on the phenyl
rings with the spacer can determine the ease of rotation of
the other ring. In good agreement with this observation,
shortening of the spacer from a hexamethylene to a 3-oxa-
pentamethylene hindered ring rotation for 15a; this is re-
As expected, simulations carried out at higher tempera-
tures (1200 and 1500 K) revealed an increased frequency of
3412
www.chemeurj.org
ꢆ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 3406 – 3419

para–para Stilbenophanes
FULL PAPER
nuclei resonate isochronously, except for 15a and 18a,
which produced broadened signals (for 15a, duplicated sig-
nals are displayed in the 13C NMR for the disubstituted sym-
metric ring). As discussed above, this situation can be ex-
plained if rapid interconversion between conformations I–
IV occurs; rapid rotations are required for at least two of
the three movements taken into account during the confor-
mational analysis.
The observed spectra and the molecular mechanics and
dynamics results are summarised in Table 4, according to
the following rules:
1) Inspection of the molecular dynamics simulations of the
macrocycles suggests that the double flip is always the
fastest, whereas one ring rotations are preferred in open
stilbene-like compounds (XIX, XX and reference [26b]).
It also suggests how easy the other motions are (always
slower than the double flip).
Figure 3. 13C NMR (50.3 MHz) spectra for compounds 15a (left) and 16a
(right). The chemically equivalent methines occurring as a single signal
for 16a and as two signals for 15a are indicated by arrows.
2) The larger hexamethylene spacer allows faster ring rota-
tions (disubstituted or trisubstituted) passing hydrogen
atoms through the intra-anular space. This is the case for
olefin 16a, with two fast motions (double flip and L-ring
rotation).
3) The shorter 3-oxapentamethylene spacer reduces the
intra-anular space and hinders ring rotations of passing
hydrogen atoms through it. Only the double flip is fast
on the NMR spectroscopy timescale and broadened sig-
nals (duplicated in the 13C NMR spectra) are observed
for olefin 15a.
ꢀ
4) The rigidity introduced by the indole N C bond slows
down motions and causes the line broadening observed
for the signals of the other (symmetric) ring in com-
pound 18a.
5) The appearance of only one set of signals in the case of
17a must be explained by a slowed down motion and the
predominance of one conformation (on the contrary, line
broadening of two sets of signals would be observed).
Conformation ꢀI is the most populated one in the mo-
lecular dynamics.
Variable-temperature (VT) 1H NMR spectroscopy: Al-
though the room temperature NMR spectra for all the ole-
fins are simple, the molecular mechanics models and the
molecular dynamics simulations suggest that, at least, one-
Figure 4. Dihedral values between the ring planes and the double-bond
plane along the trajectories of the molecular dynamics simulations for
olefin 16a at 1200 K. The dihedral values for the trisubstituted phenyl
ring are shown on the top, whereas those for the disubstituted phenyl
ring are shown on the bottom, showing the rotation of the former during
the molecular dynamics trajectory.
1
ring flips should be activated processes. VT H NMR spec-
troscopic studies of olefins 14a–16a in CD3OD between 223
and 323 K in 10 K steps revealed complex kinematics. In all
the cases, the observed coalescence processes are consistent
with the freezing of one-ring rotations at lower tempera-
tures.[30] These observations agree with the MD simulations,
which pinpoint this motion as the slower of the fast ones on
the NMR spectroscopy timescale. The effects of this freezing
are most diagnostic for the signals of the other aromatic
ring.
sponsible for signal differentiation for equivalent carbon
atoms. The introduction of more substituents on the phenyl
rings, such as in 14a, or the more rigid indole ring of 17a
and 18a also prevented ring rotations.
Conformational behaviour at room temperature: All of the
macrocyclic olefins described herein (14a–18a) show NMR
spectra with a single set of signals where the exchangeable
Olefins 15a and 16a, differing only in the spacer (3-oxa-
pentamethylene and hexamethylene, respectively), behaved
Chem. Eur. J. 2011, 17, 3406 – 3419
ꢆ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3413

M. Medarde, R. Pelꢃez et al.
Table 4. Summary of ease of rotations: double flips and ring rotations for olefins 14a–18a compared with the
observed NMR spectra and explained according to molecular dynamics calculations.
conformers in a ratio of about
7:3 ratio. The most affected sig-
nals were those corresponding
to the aromatic protons and the
methoxy group of the monome-
thoxylated phenyl ring. Again,
the magnitude of the chemical
shift differences for the residual
conformers could be assigned
to the anisotropic field effect of
the indole ring. In the minor
conformer (belonging to confor-
mation-type III in Scheme 7,
Y=H), the methoxy group and
its ortho proton were upfield
shifted with respect to those of
Double
flip
L-ring rota-
tion
R-ring rota-
tion
NMR spec- Origin
tra
Explanation
14a +++
+
no
one set
averaged, fast exchange
two fast mo-
tions
shorter spacer
two or three
fast
15a +++
16a +++
+/ꢀ
no
+
one set[a]
one set
intermediate exchange
averaged, fast exchange
+
motions
17a +++
+/ꢀ
+/ꢀ
+/ꢀ
one set
slow exchange,
one conformation predomi-
nates
rigid indole
18a +++
no
one set[a]
intermediate exchange
rigid indole
[a] Signals are broadened.
in accordance with their expected different difficulty in let-
ting the ring planes pass through the intra-annular cavity.
Olefin 15a, with the smaller cavity, already showed a sub-
stantial broadening of the signals integrating two protons of
the axially symmetric ring (R ring in Scheme 6, Y=H) at
room temperature. At lower temperatures, each broadened
signal gave rise to two new ones that integrated to one
proton each, in agreement with slower L-ring rotation. The
chemical shift difference between each one of the pair of
newly generated signals (ca. 0.3 ppm) could be assigned to
the anisotropic field effect of the methoxy group on the op-
posite phenyl ring because the field effect due to the ring
current should still have been averaged by the double flip.
For 16a, the larger intra-annular space gave rise to narrow,
averaged signals at room temperature. A decrease in tem-
perature resulted in substantial line broadening of the sig-
nals of the axially symmetric ring, which were no longer ob-
served below 233 K. Similar behaviour was observed for
14a.
Olefin 18a followed the guidelines indicated for 15a, with
a similar temperature dependency. The aromatic and me-
thoxy protons of the dimethoxy-substituted phenyl ring
were broadened and signal averaged at room temperature
and they were not observable at 273 K. They reappeared
and progressively sharpened up as two new pairs of signals
from 263 down to 213 K. In this case, the chemical shift dis-
persion of about 1 ppm between the chemically equivalent
proton pairs could be assigned to the anisotropic field effect
of the indole ring, whose frozen rotation reveals the pre-
dominance of the endo (I and III) versus exo (II and IV)
conformations, connected by fast double flips.
the major conformer, which corresponded to conformation I
in Scheme 7. The large chemical shift difference suggested a
minor, if any, contribution of conformations II and IV to the
overall equilibrium. These results are in full agreement with
the results of the molecular mechanics for 17a and the non-
macrocyclic analogue XIX, previously discussed.
Tubulin polymerisation inhibition: These macrocyclic stil-
benes are the first of their class to show tubulin polymeri-
sation inhibition (TPI) (Table 5).[31] Comparison of the ac-
Table 5. Tubulin polymerisation inhibitory activity of the macrocyclic
olefins 14a–18a.[a]
Compound
TPI [%] (c [mM])
XXI[12c]
14a
60 (40)
0 (40)
15a
16a
17a
18a
44 (20)
48 (20)
19 (20)
44 (20)
[a] XXI=stilbenophane (Z=CH2-CH2) in Scheme 1.
cessible conformations for the macrocyclic ligands 14a–18a
with the X-ray crystal structures of podophyllotoxin and
DAMA-colchicine complexed with tubulin[32] revealed that
the relevant ones are +I and/or +III (Figure 5). The TPI
assays were carried out at 378C with incubation times of mi-
nutes or more. By taking into account the dynamic behav-
iour shown by the olefins, conformational equilibration
should not then be an issue. Therefore, the similar confor-
mations adopted by 15a and 16a lead to similar TPI poten-
cies, despite their different dynamic behaviour. Further-
more, for all of the studied olefins, conformational classes I/
III are the more stable ones. Accordingly, the observed dif-
ferences in potency should be explained by considering the
structures of the favoured conformations.
Olefin 17a does not have an axially symmetric ring, which
precludes the observation of mutual exchange behaviour as
described above. On the contrary, it shows a non-mutual ex-
change, resulting from the freezing of the same processes as
before. The observed temperature dependency is in between
that of 15a and that of 16a. Signal averaging was observed
at room temperature and a single residual conformer was
seen, the signals of which became substantially line broad-
ened between 273 and 253 K. They were unobservable down
to 233 K and reappeared as two new pairs of broadened sig-
nals at 213 K, which corresponded to two different residual
The great structural similarity observed among the macro-
cyclic stilbenes indicates differences in the substituents to
explain the differences in potency. First of all, the similar be-
haviour of 15a, 16a and XXI, the olefin with one more me-
3414
www.chemeurj.org
ꢆ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 3406 – 3419

para–para Stilbenophanes
FULL PAPER
These compounds were tested against HeLa, HT-29, A-
549 and HL-60 cell lines, but they were not cytotoxic at the
concentrations of ꢂ1 mm assayed (higher concentrations
were not used owing to solubility problems).
Conclusion
The synthesis of new stilbenophanes has been accomplished
by using the McMurry methodology to produce analogues
of antimitotic combretastatins. The presence of a bridge be-
tween para–para positions modifies the dynamics of these
molecules in comparison with those of non-macrocyclic stil-
bene analogues. Conformational analysis of these com-
pounds was carried out by taking into account ring flips,
ring rotations and the effect of the bridge; these results
were in agreement with the molecular calculations and
NMR spectra. These macrocyclic analogues are the first of
their class to inhibit tubulin polymerisation, giving a further
insight into the structural requirements of the colchicines
site of tubulin. The activity results for these macrocycles
have been rationalised in terms of mobility and steric hin-
drance.
Figure 5. Superposition of podophyllotoxin (carbon atoms shown in
green), XX (carbon atoms shown in grey) and 18a (carbon atoms shown
in pink). The structure of podophyllotoxin was obtained by X-ray crystal
structure analysis.[32] The structures of XX and 18a were generated by
docking with Autodock 4.0, following a described procedure.[17e] Com-
pounds 18a and XX belong to the conformational class +I/+III.
thoxy group (the macrocyclic equivalent of deoxyCA-4),
should not pass unnoticed. This is the first case of the re-
placement of the trimethoxyphenyl ring of combretastatin
analogues (the central methoxy group of the dimethoxylated
ring of the macrocyclic olefins is represented by the spacer)
by a moiety with one less methoxy group without severe re-
ductions in potency. This observation suggests a favourable
replacement of such a moiety by the spacer, as we had pre-
viously anticipated.[12c]
On the other hand, for the indolic analogues, the macro-
cycle with a single methoxy group (17a) is less potent than
the dimethoxylated counterpart (18a). The rigidity of the
indole forces the spacer of the +I/III conformations of the
indolic olefins towards the opposite side of the methoxy
group that lies under the phenyl ring of the indole in the
complex with tubulin (Figure 5). In the more flexible phe-
nylic olefins it can go more along the double-bond plane,
thus replacing the aforementioned methoxy group. This sug-
gestion is supported by the X-ray crystal structure of E7030
complexed with tubulin, which shows that the inner me-
thoxy group can be replaced by a hydrogen atom in pyri-
dine-based colchicine site ligands.[33] This observation is im-
portant for the establishment of consistent structure–activity
relationships for colchicine site ligands because the tri-
Experimental Section
General: Reagents were used as purchased without further purification.
Solvents (THF, DMF, CH2Cl2, and toluene) were dried and freshly dis-
tilled before use according to procedures described in the literature.
Chromatographic separations were performed on silica-gel columns by
flash (Kieselgel 40, 0.040–0.063; Merck) or gravity (Kieselgel 60, 0.063–
0.200 mm; Merck) chromatography. TLC was performed on precoated
silica-gel polyester plates (0.25 mm thickness) with a UV fluorescence in-
dicator 254 (Polychrom SI F254). Melting points were determined on a
1
Buchi 510 apparatus and are uncorrected. H and 13C NMR spectra were
recorded in CDCl3 at 258C on a Bruker WP 200-SY spectrometer at 200/
50 MHz or on a Bruker SY spectrometer at 400/100 MHz. Chemical
shifts (d) are given in ppm downfield from tetramethylsilane as the inter-
nal standard and coupling constants (J values) are in hertz. IR spectra
were run on a Nicolet Impact 410 spectrophotometer. GC-MS analyses
were carried out with
a Hewlett–Packard 5890 Series II apparatus
(70 eV). For FAB HRMS analyses, a VG-TS250 apparatus (70 eV) was
used. HPLC separations were run on at least three different columns
(5 mm, 4.6ꢇ150 mm): Waters X-Terra MS C8, Waters X-Terra MS C18
and Waters X-Terra MS CF on an Agilent HP series 1100 with at least
two different solvent gradients (typically acetonitrile/water or methanol/
water). Elemental analyses were run on a Perkin–Elmer 2400 CHN appa-
ratus.
Compounds 1–3 were obtained as previously described,[12] whereas inter-
mediates 4--11 were produced by the Mitsunobu method, phase transfer
or direct alkylations, by following the procedures described below.
ACHTUNGTRENNUNGmethoxyphenyl ring has often been seen as a requirement in
some families of analogues and is absent in other families.
The lack of activity observed for 14a, which had higher
structural resemblance to combretastatin A-4 than 15a and
16a, could also be explained by a preferential binding of
conformational classes +I and/or +III. If the spacer, as
stated above for the indolic macrocycles, fills the concave
side of the molecule, the hydroxyl group must point towards
the convex side of the molecule. The lack of activity ob-
served for 14a, thus arises again from the departure of the
spacer from the plane of double bond, as previously indicat-
ed for indoles.
General procedure for the Mitsunobu reactions producing dialdehydes
4--6: A mixture of the hydroxy aldehyde (1--3), the PPh3 resin (0.76–
1.01 mol/mol of aldehyde, except for the synthesis of 4, 2.32 mol/mol of
1), and the phenolic aldehyde (0.96–1.50 mol/mol of aldehyde, except for
the synthesis of 4, because a large excess of 3.42 mol of 3,4-dihydroxy-
benzldehyde/mol of 1 was used to avoid dialkylations) in dry CH2Cl2 (8–
19 mL/mmol of aldehyde) was stirred for 1–3 h before the slow addition
of DBAD or DIAD (0.76–1.01 mol/mol of aldehyde; in the synthesis of
4, 2.31 mol of DBAD/mol of 1 were used) at RT. After 48 h the reactions
were filtered and the resin was washed with EtOAc. The combined or-
ganic layers were evaporated, dissolved in EtOAc and washed with
Chem. Eur. J. 2011, 17, 3406 – 3419
ꢆ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3415

M. Medarde, R. Pelꢃez et al.
NaOH (4%) and brine until the pH of the solution was neutral. Once
dried and evaporated, the crude reaction products were purified by flash
chromatography (hexane/EtOAc mixtures).
0.59 mmol) in CH2Cl2 (40 mL) were stirred for 1 h, after which 1,6-dibro-
mohexane (4.25 mL, 6.74 g, 27.60 mmol) was added and the mixture was
maintained for 48 h at RT The crude product was extracted from a SiO2
column with hexane and CH2Cl2 to yield bromoaldehyde 9 (3.04 g,
9.90 mmol, 71.7%). 1H NMR (200 MHz): d=1.28–1.32 (m, 2H; CH2),
1.36–1.38 (m, 2H; CH2), 1.76–1.78 (m, 2H; CH2), 1.80–1.82 (m, 2H;
CH2), 3.32 (t, J=6.6 Hz, 2H; CH2), 4.09 (t, J=7.0 Hz, 2H; CH2); 6.62 (d,
J=3.3 Hz, 1H; Ar-H); 7.15 (d, J=3.3 Hz, 1H; Ar-H), 7.37 (d, J=8.4 Hz,
1H; Ar-H), 7.74 (dd, J=8.4, 1.5 Hz, 1H; Ar-H), 8.10 (d, J=1.5 Hz, 1H;
Ar-H), 9.99 ppm (s, 1H; CHO); 13C NMR (50.3 MHz): d=26.1 (CH2),
27.7 (CH2), 30.1 (CH2), 32.6 (CH2), 33.8 (CH2), 46.5 (CH2), 103.4 (CH),
110.0 (CH), 121.7 (CH), 126.5 (CH), 128.4 (C), 129.3 (C), 129.9 (CH),
139.3 (C), 192.4 ppm (CH).
Compound 4: 1H NMR (400 MHz): d=3.86–3.88 (m, 2H; CH2), 3.91 (s,
6H; 2ꢇCH3), 3.97–3.99 (m, 2H; CH2), 4.30–4.32 (m, 4H; 2ꢇCH2), 6.99
(d, J=7.9 Hz, 1H; Ar-H), 7.37 (d, J=1.8 Hz, 1H; Ar-H), 7.41–7.43 (m,
1H; Ar-H), 7.42 (s, 2H; Ar-H), 9.84 (s, 1H; CHO), 9.86 ppm (s, 1H;
CHO); 13C NMR (100.6 MHz): d=56.5 (2ꢇCH3), 69.5 (2ꢇCH2), 71.0
(CH2), 72.6 (CH2), 107.0 (CH), 113.3 (CH), 115.3 (CH), 124.2 (CH),
131.3 (C), 132.2 (C), 142.7 (C), 147.5 (C), 151.6 (C), 154.0 (2ꢇ C), 191.4
(CH), 191.4 ppm (CH); FTIR: v˜ =3341, 1689, 1607, 1507, 1127 cmꢀ1
.
1
Compound 5: H NMR (400 MHz): d=1.59–1.61 (m, 4H; 2ꢇCH2), 1.79–
1.81 (m, 4H; 2ꢇCH2), 3.91 (s, 6H; 2ꢇCH3), 3.99–4.01 (m, 2H; CH2),
4.09–4.10 (m, 2H; CH2), 6.95 (d, J=8.3 Hz, 1H; Ar-H), 7.12 (s, 2H; 2ꢇ
CH), 7.40 (d, J=8.3 Hz, 1H; Ar-H), 7.43 (brs, 1H; Ar-H), 9.82 (s, 1H;
CHO), 9.86 ppm (s, 1H; CHO); 13C NMR (100.6 MHz): d=25.5 (2ꢇ
CH2), 28.8 (CH2), 30.0 (CH2), 56.1 (2ꢇCH3), 69.5 (CH2), 73.2 (CH2),
106.7 (2ꢇCH), 111.1 (CH), 114.0 (CH), 124.6 (CH), 130.1 (C), 131.5 (C),
142.8 (C), 142.8 (C), 151.9 (2ꢇC), 153.3 (C), 171.2 (C), 191.1 (CH),
Direct alkylation synthesis of 10:
A solution of vanillin (1.72 g,
11.30 mmol) and K2CO3 (10.00 g, 7.25 mmol) in dry DMF (40 mL) was
stirred for 0.5 h. Then, bromoaldehyde 9 (2.90 g, 9.44 mmol) in DMF
(10 mL) was added and the reaction was maintained for 48 h at 708C
under argon. The mixture was poured into hexane (150 mL), filtered, and
the solvent was evaporated in vacuo. The crude reaction mixture was re-
dissolved in CH2Cl2, washed with NaOH (4%) and brine, and evaporated
to yield dialdehyde 10 (3.41 g, 9.34 mmol; 98%). 1H NMR (400 MHz):
d=1.35–1.43 (m, 2H; CH2), 1.45–1.47 (m, 2H; CH2), 1.82–1.84 (m, 2H;
CH2), 1.85–1.87 (m, 2H; CH2), 3.89 (s, 3H; CH3), 4.05 (t, J=6.6 Hz, 2H;
CH2), 4.17 (t, J=7.0 Hz, 2H; CH2), 6.62 (d, J=3.3 Hz, 1H; Ar-H), 6.91
(d, J=8.4 Hz, 1H; Ar-H), 7.18 (d, J=3.3 Hz, 1H; Ar-H), 7.39–7.43 (m,
3H; Ar-H), 7.70 (dd, J=8.4, 1.4 Hz, 1H; Ar-H), 8.13 (d, J=1.4, 1H; Ar-
H), 9.83 (s, 1H; CHO), 10.00 ppm (s, 1H; CHO); 13C NMR (100.6 MHz):
d=25.4 (CH2), 26.4 (CH2), 28.7 (CH2), 30.0 (CH2), 46.3 (CH2), 55.8
(CH3), 68.7 (CH2), 103.2 (CH), 109.3 (CH), 110.0 (CH), 111.4 (CH),
121.5 (CH), 126.4 (CH), 126.6 (CH), 128.3 (C), 129.2 (C), 129.8 (CH),
129.8 (C), 139.2 (C), 149.7 (C), 154.0 (C), 190.7 (CH), 192.2 ppm (CH);
191.1 ppm (CH); FTIR: v˜ =3309, 2276, 1688, 1508, 1127 cmꢀ1
.
Compound 6: 1H NMR (400 MHz): d=3.65 (s, 3H; CH3), 3.73–3.77 (m,
4H; 2ꢇ CH2), 3.98–4.06 (m, 4H; 2ꢇ CH2), 6.78 (d, J=8.8 Hz, 2H; 2ꢇ
CH), 7.16–7.20 (m, 3H; Ar-H), 7.59 (d, J=8.8 Hz, 2H; 2ꢇ CH), 9.61 (s,
1H; CHO), 9.64 ppm (s, 1H; CHO); 13C NMR (100.6 MHz): d=55.7
(CH3), 67.7 (CH2), 68.4 (CH2), 69.5 (2ꢇCH2), 109.3 (CH), 111.8 (CH),
114.8 (2ꢇCH), 126.4 (CH), 129.9 (C), 130.1 (C), 131.8 (2ꢇCH), 149.7
(C), 153.7 (C), 163.7 (C), 190.7 (CH), 190.9 ppm (CH); FTIR: v˜ =1683,
1595, 1511, 1268, 1132 cmꢀ1
.
Phase-transfer synthesis of 7: Vanillin (5.52 g, 36.3 mmol), NaOH (2.27 g,
56.8 mmol), 1,6-dibromohexane (12.0 mL, 19.0 g, 78.0 mmol) and tetrabu-
tylammonium bromide (1.90 g, 5.89 mmol) were added to a heterogene-
ous mixture of CH2Cl2 (190 mL) and water (190 mL). The reaction was
stirred for 48 h., separated and the aqueous layer extracted with CH2Cl2.
The combined organic layers were successively washed with NaOH (4%)
and brine, dried (Na2SO4) and evaporated. After column chromatography
(SiO2) with hexanes and CH2Cl2, bromoaldehyde 7 (8.13 g, 25.8 mmol,
71%) was obtained as a brown solid, which was crystallised in CH2Cl2/
hexane. M.p. 458C; 1H NMR (400 MHz): d=1.55–1.57 (m, 4H; 2ꢇCH2),
1.86–1.93 (m, 4H; 2ꢇCH2), 3.40 (t, J=6.6 Hz, 2H; CH2), 3.89 (s, 3H;
CH3); 4.08 (t, J=6.6 Hz, 2H; CH2), 6.94 (d, J=8.0 Hz, 1H; Ar-H), 7.38
(d, J=1.8 Hz, 1H; Ar-H), 7.41 (dd, J=1.8, 8.0 Hz, 1H; Ar-H), 9.81 ppm
(s, 1H; CHO); 13C NMR (100.6 MHz): d=25.1 (CH2), 27.8 (CH2), 28.7
(CH2), 32.6 (CH2), 33.8 (CH2), 55.9 (CH3), 68.8 (CH2), 109.2 (CH), 111.4
(CH), 126.6 (CH), 129.8 (C), 149.7 (C), 154.0 (C), 190.7 ppm (CHO);
FTIR: v˜ =1684, 1589, 1511, 1269, 1134 cmꢀ1; MS: m/z (%): 314 (4), 316
(4) [M+], 152 (100).
FTIR: v˜ =1693, 1596, 808, cmꢀ1
.
Direct alkylation synthesis of 11: A solution of syringaldehyde (2.16 g,
11.87 mmol) and K2CO3 (10.00 g, 7.25 mmol) in dry DMF (40 mL) was
stirred for 0.5 h. Then, bromoaldehyde 9 (3.05 g, 9.89 mmol) in DMF
(10 mL) was added and the mixture was maintained for 48 h at 708C
under argon. The mixture was poured into hexane (150 mL), filtered and
the solvent was evaporated in vacuo. The crude reaction mixture was re-
dissolved in CH2Cl2, washed with NaOH (4%) and brine, and evaporated
to yield dialdehyde 11 (3.30 g, 8.35 mmol; 85%). 1H NMR (400 MHz):
d=1.20–1.22 (m, 2H; CH2), 1.40–1.43 (m, 2H; CH2), 1.64–1.67 (m, 2H;
CH2), 1.71–1.74 (m, 2H; CH2), 3.71 (s, 6H; 2ꢇCH3), 3.91–3.95 (m, 2H;
CH2), 3.98–4.01 (m, 2H; CH2), 6.48 (d, J=3.3 Hz, 1H; Ar-H), 6.97 (s,
2H; 2ꢇ Ar-H), 7.07 (d; J=3.3 Hz, 1H; Ar-H), 7.26 (d, J=8.4 Hz, 1H;
Ar-H), 7.61 (d, J=8.4 Hz, 1H; Ar-H), 7.96 (s, 1H; Ar-H), 9.70 (s, 1H;
CHO), 9.85 ppm (s, 1H; CHO); 13C NMR (100.6 MHz): d=25.4 (CH2),
26.5 (CH2), 29.9 (CH2), 30.2 (CH2), 46.5 (CH2), 56.1 (2ꢇCH3), 73.2
(CH2), 103.2 (CH), 106.6 (2ꢇCH), 110.0 (CH), 121.4 (CH), 126.4 (CH),
128.2 (C), 129.1 (C), 129.9 (CH), 131.6 (C), 139.2 (C), 142.7 (C), 153.8
(2ꢇC), 191.1 (CH), 192.3 ppm (CH).
Direct alkylation synthesis of 8: A solution of p-hydroxybenzaldehyde
(1.22 g, 10.0 mmol), K2CO3 (4.84 g, 35.0 mmol) and bromoaldehyde 7
(2.07 g, 6.57 mmol) in dry DMF (30 mL) was maintained for 48 h at 608C
under argon. The solvent was evaporated in vacuo and the crude reaction
mixture was extracted with CH2Cl2 and 2n HCl, and the organic layer
was washed with brine and dried. By column chromatography (SiO2)
using CH2Cl2/AcOEt (10:1) and 1% of triethylamine (TEA), dialdehyde
8 (1.08 g, 3.03 mmol; 46%) was obtained as a white solid, which was re-
crystallised in CH2Cl2/hexane. M.p. 788C; 1H NMR (400 MHz): d=1.48–
1.51 (m, 4H; 2ꢇCH2), 1.77–1.80 (m, 4H; 2ꢇCH2), 3.86 (s, 3H; CH3),
3.96–4.07 (m, 4H; 2ꢇCH2), 6.92 (d, J=8.0 Hz, 1H; Ar-H), 6.93 (d, J=
8.8 Hz, 2H; 2ꢇ Ar-H), 7.35 (d, J=1.8 Hz, 1H; Ar-H), 7.38 (dd, J=1.8,
8.0 Hz, 1H; Ar-H), 7.76 (d, J=8.8 Hz, 2H; 2ꢇ Ar-H), 9.78 (s, 1H;
CHO), 9.81 ppm (s, 1H; CHO); 13C NMR (100.6 MHz): d=25.8 (2ꢇ
CH2), 28.9 (2ꢇCH2), 56.0 (CH3), 68.2 (CH2), 68.9 (CH2), 109.3 (CH),
111.4 (CH), 114.8 (2ꢇCH), 126.8 (CH), 129.8 (C), 130.0 (C), 132.0 (2ꢇ
CH), 149.8 (C), 154.1 (C), 161.2 (C), 190.8 (CHO), 190.9 ppm (CHO);
FTIR: v˜ =1683, 1596, 1511, 1267, 1160, 1135, 1012 cmꢀ1; MS: m/z (%):
356 (65) [M+], 152 (100).
General procedure for McMurry reactions: Mixtures of TiCl4 98% (5–
10 mol/mol of dialdehyde) and Zn (10–20 mol/mol of dialdehyde) in dry
THF (20–100 mL/mmol of dialdehyde) were prepared at 08C and heated
at reflux for 30 min. Then, solutions of dialdehydes (4, 5, 6, 8, 10 or 11)
in dry THF (20–40 mL/mmol of dialdehyde) were added and maintained
at either room temperature or under reflux conditions. The reactions
were poured into mixtures of EtOAc and 2m HCl; the aqueous layer was
extracted and the combined organic layers were worked up. Olefins were
separated by chromatography (SiO2, hexane/EtOAc mixtures with 0.1%
triethylamine). For details about temperature, time, crude product and
isolated yields, see Table 1.
Reaction of dialdehyde 4 (Table 1, entry 1): By treatment of dialdehyde 4
(260 mg, 0.6 mmol) with a mixture of Zn (430 mg, 6.6 mmol) and TiCl4
(0.4 mL, 3.5 mmol) in THF (25 mL) under reflux for 5 h, the crude reac-
tion product (245 mg) was obtained after the described workup. Olefin
13a was not detected in the crude reaction mixture, which contained a
mixture of diols 13b and 13c.
Phase-transfer synthesis of 9: By following the procedure described
above, 1H-indole-5-carbaldehyde (2.00 g, 13.80 mmol), NaOH (1.10 g,
27.60 mmol) and tetrabutylammonium hydrogen sulfate (200 mg,
3416
www.chemeurj.org
ꢆ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 3406 – 3419

para–para Stilbenophanes
FULL PAPER
Reaction of dialdehyde 5: Synthesis of olefin 14a (Table 1, entry 2): By
treatment of dialdehyde 5 (800 mg, 2.0 mmol) with a mixture of Zn
(1300 mg, 20.0 mmol) and TiCl4 (1.1 mL, 9.7 mmol) in THF (130 mL)
under reflux for 24 h, the crude reaction product (785 mg) was obtained
after the described workup. Chromatographic separation afforded 14a
(290 mg, 0.80 mmol, 40%). 1H NMR (400 MHz): d=1.1–1.6 (m, 8H; 4ꢇ
CH2), 3.65 (s, 6H; 2ꢇ CH3), 4.11–4.08 (m, 4H; 2ꢇ CH2), 6.13 (s, 2H; 2ꢇ
Ar-H), 6.20 (dd, J=1.9, 8.4 Hz, 1H; Ar-H), 6.51 (d, J=8.4 Hz, 1H; Ar-
H), 6.70 (d, J=1.9 Hz, 1H; Ar-H), 6.88 ppm (brs, 2H; 2ꢇOlef-H);
13C NMR (100.6 MHz): d=26.5 (CH2), 27.0 (CH2), 29.6 (CH2), 30.0
(CH2), 55.8 (2ꢇCH3), 69.8 (CH2), 71.6 (CH2), 106.8 (2ꢇCH), 114.3 (CH),
115.4 (CH), 121.8 (CH), 133.1 (C), 133.3 (CH), 133.7 (CH), 134.0 (C),
132.8 (C), 142.1 (C), 142.5 (C), 152.6 ppm (2ꢇC); FTIR: v˜ =3420, 1580,
J=2.7 Hz, 1H; Ar-H) 7.14 (d, J=10.2 Hz, 1H; Olef-H), 7.40 ppm (s, 1H;
Ar-H); 13C NMR (100.6 MHz): d=23.2 (CH2), 25.9 (CH2), 27.2 (CH2),
29.5 (CH2), 46.1 (CH2), 55.4 (CH3), 70.3 (CH2), 101.0 (CH), 109.2 (CH),
113.9 (CH), 117.9 (CH), 120.5 (CH), 122.5 (CH), 124.6 (CH), 128.2 (C),
128.2 (CH), 129.8 (C), 132.9 (CH), 133.6 (C), 134.8 (CH), 135.0 (C),
144.9 (C), 150.3 ppm (C); FTIR: v˜ =1509, 1262 cmꢀ1
.
Reaction of dialdehyde 11: Synthesis of olefin 18a (Table 1, entry 9): By
treatment of dialdehyde 11 (570 mg, 1.44 mmol) with a mixture of Zn
(2040 mg, 30.7 mmol) and TiCl4 (1.8 mL, 16.3 mmol) in THF (300 mL)
under reflux for 5 h, the crude reaction product (680 mg) was obtained
after the described workup. Chromatographic separation afforded 18a
(150 mg, 0.43 mmol, 26.7%) and diols 18b and 18c. 18a: 1H NMR
(400 MHz): d=0.90–0.94 (m, 2H; CH2), 1.23–1.26 (m, 4H; 2ꢇCH2),
1.61–1.64 (m, 2H; CH2), 3.45 (brs, 6H; 2ꢇCH3), 3.98 (t, J=6.4 Hz, 2H,
CH2), 4.06 (t, J=5.9 Hz, 2H; CH2), 6.05 (brs, 2H; 2ꢇAr-H), 6.39 (d, J=
3.1 Hz, 1H; Ar-H), 6.45 (dd, J=8.2, 1.4 Hz, 1H; Ar-H), 6.87 (d, J=
8.2 Hz, 1H; Ar-H), 6.92 (d, J=10.1 Hz, 1H; Olef-H), 6.97 (d, J=3.1 Hz,
1H; Ar-H), 7.16 (d, J=10.1 Hz, 1H; Olef-H), 7.40 ppm (s, 1H; Ar-H);
13C NMR (100.6 MHz): d=23.4 (CH2), 26.0 (CH2), 27.8 (CH2), 29.9
(CH2), 45.7 (CH2), 55.6 (2ꢇCH3), 71.8 (CH2), 101.0 (CH), 107.2 (2ꢇCH),
109.3 (CH), 120.4 (CH), 124.2 (CH), 128.0 (C), 128.1 (CH), 129.8 (C),
133.4 (CH), 133.9 (C), 134.5 (C), 134.9 (CH), 135.1 (C), 152.4 ppm (2ꢇ
1503, 1123 cmꢀ1
.
Reaction of dialdehyde 5 (Table 1, entry 3): By treatment of dialdehyde 5
(315 mg, 0.8 mmol) with a mixture of Zn (523 mg, 8.0 mmol) and TiCl4
(0.5 mL, 4.4 mmol) in THF (25 mL) at 08C for 5 h, the crude reaction
product (370 mg) was obtained after the described workup. Olefin 14a
was not detected in the crude reaction mixture and the mixture (14b/14c
1/3) could not be separated by chromatography.
Reaction of dialdehyde 6: Synthesis of olefin 15a (Table 1, entry 4). By
treatment of dialdehyde 6 (640 mg, 1.86 mmol) with a mixture of Zn
(3820 mg, 58.4 mmol) and TiCl4 (2.0 mL, 18.2 mmol) in THF (300 mL) at
reflux for 24 h, the crude reaction product (580 mg) was obtained after
the described workup. Chromatographic separation afforded 15a (87 mg,
0.28 mmol, 15%) and diols 15b and 15c. 15a: 1H NMR (400 MHz): d=
3.55–3.58 (m, 4H; 2ꢇCH2), 3.66 (s, 3H; CH3), 4.23–4.26 (m, 4H; 2ꢇ
CH2), 6.26 (dd, J=8.5, 1.8 Hz, 1H; Ar-H), 6.27 (d, J=1.8 Hz, 1H; Ar-
H), 6.5–6.7 (m, 4H; Ar-H), 6.69 (d, J=8.5 Hz, 1H; Ar-H), 6.97 (d, J=
10.0 Hz, 1H; Olef-H), 7.00 ppm (d, J=10.0, 1H; Olef-H); 13C NMR
(100.6 MHz): d=55.5 (CH3), 68.0 (CH2), 72.0 (CH2), 72.4 (CH2), 72.5
(CH2), 113.9 (CH), 115.0 (CH), 115.7 (CH), 118.5 (CH), 122.6 (CH),
130.0 (CH), 130.8 (CH), 131.7 (C), 134.2 (C), 134.4 (2ꢇCH), 146.9
(C),150.3 (C), 157.1 ppm (C); FTIR: v˜ =1609, 1505, 1262, 1228,
C); FTIR: v˜ =1577, 1126 cmꢀ1
.
Tubulin isolation: Calf brain microtubule protein (MTP) was purified by
two cycles of temperature-dependent assembly/disassembly, according to
the method of Shelanski et al.[34] modified as described in the litera-
ture.[35] The solution of MTP was stored at ꢀ808C. Protein concentrations
were determined by the method of Bradford[36] by using bovine serum al-
bumin (BSA) as standard. Four different MTP preparations were used in
the tubulin assembly assays.
Tubulin assembly: In vitro tubulin self-assembly was monitored turbidi-
metrically at 450 nm by using a thermostated Thermospectronic Helios a
spectrophotometer fitted with a Peltier temperature controller and a cir-
culating water carrousel system. The ligands were dissolved in DMSO
and stored at ꢀ208C. The amount of DMSO in the assays was 4%,
which has been reported not to interfere with the assembly process.[37]
The increase in turbidity was followed simultaneously in a batch of six
cuvettes (containing 1.0 mgmLꢀ1 MTP in 0.1m MES buffer, 1 mm ethyl-
ene glycol tetraacetic acid (EGTA), 1 mm MgCl2, 1 mm, b-ME, 1.5 mm
guanosine triphosphate (GTP), pH 6.7, and the measured ligand concen-
tration), with a control (i.e., with no ligand) always being included. The
samples were preincubated for 30 min at 208C to allow binding of the
ligand, and were cooled on ice for 10 min. The cuvettes were then placed
in the spectrophotometer at 48C. The assembly process was initiated by a
shift in the temperature to 378C.
1118 cmꢀ1
.
Reaction of dialdehyde 8: Synthesis of olefin 16a (Table 1, entry 5): By
treatment of dialdehyde 8 (850 mg, 2.39 mmol) with a mixture of Zn
(4600 mg, 70.3 mmol) and TiCl4 (2.7 mL, 24.6 mmol) in THF (300 mL)
under reflux for 24 h, the crude reaction product (670 mg) was obtained
after the described workup. Chromatographic separation afforded 16a
(52 mg, 0.16 mmol, 7%). 1H NMR (400 MHz): d=1.29–1.21 (m, 4H; 2ꢇ
CH2), 1.46–1.39 (m, 4H; 2ꢇ CH2), 3.71 (s, 3H; CH3), 4.05 (t, J=7.8 Hz,
2H; CH2), 4.11 (t, J=7.5 Hz, 2H; CH2), 6.25 (dd, J=8.2, 1.9 Hz, 1H; Ar-
H), 6.52 (d, J=1.9 Hz, 1H; Ar-H), 6.56 (d, J=8.2 Hz, 1H; Ar-H), 6.62
(d, J=8.6 Hz, 2H; 2ꢇAr-H), 6.77 (d, J=8.6 Hz, 2H; 2ꢇ Ar-H), 6.84 (d,
J=10.8 Hz, 1H; Olef-H), 6.87 ppm (d, J=10.8 Hz, 1H; Olef-H);
13C NMR (100.6 MHz): d=23.6 (2ꢇCH2), 25.6 (CH2), 26.0 (CH2), 55.6
(CH3), 68.1 (CH2), 69.7 (CH2), 113.3 (CH), 116.1 (2ꢇCH), 116.9 (CH),
122.5 (CH), 130.4 (2ꢇCH), 131.4 (C), 132.7 (C), 132.8 (2ꢇCH), 144.5
(C), 150.3 (C), 155.9 ppm (C); FTIR: v˜ =1601, 1505, 1257, 1233,
Acknowledgements
1131 cmꢀ1
.
We thank the Junta de Castilla y Leꢂn (ref. SA067A09, SA063A09), the
Spanish MEC/MICINN (ref. SAF2008-04242) and the EU (Structural
Funds) for financial support. C.A. and R.A. thank the Spanish MEC for
FPI predoctoral grants.
Reaction of dialdehyde 8 (Table 1, entry 6): By treatment of dialdehyde 8
(62 mg, 0.17 mmol) with a mixture of Zn (343 mg, 5.25 mmol) and TiCl4
(0.20 mL, 1.77 mmol) in THF (100 mL) at room temperature for 24 h,
the crude reaction product (57 mg; 35/65 mixture of 16b+16c) was ob-
tained after the described workup.
Reaction of dialdehyde 10: Synthesis of olefin 17a (Table 1, entries 7 and
8): By treatment of dialdehyde 10 (1660 mg, 4.54 mmol) with a mixture
of Zn (6430 mg, 96.8 mmol) and TiCl4 (5.3 mL, 47.1 mmol) in THF
(300 mL) at reflux for 5 h, the crude reaction product (1610 mg) was ob-
tained after the described workup. Chromatographic separation afforded
17a (158 mg, 0.46 mmol, 10.1%) and diols 17b and 17c. 17a: 1H NMR
(400 MHz): d=0.92–0.95 (m, 2H; CH2), 1.17–1.19 (m, 2H; CH2), 1.32–
1.35 (m, 2H; CH2), 1.61–1.64 (m, 2H; CH2), 3.55 (s, 3H; CH3), 3.98 (t,
J=7.7 Hz, 2H; CH2), 4.05 (t, J=6.2 Hz, 2H; CH2), 6.25 (dd, J=8.1,
1.8 Hz, 1H; Ar-H), 6.38 (dd, J=7.9, 1.8 Hz, 1H; Ar-H), 6.40 (d, J=
2.7 Hz, 1H; Ar-H), 6.44 (brs, 1H; Ar-H), 6.47 (d, J=8.1 Hz, 1H; Ar-H),
6.83 (d, J=7.9 Hz, 1H; Ar-H), 6.92 (d, J=10.2 Hz, 1H; Olef-H), 6.96 (d,
[1] a) T. Shen, X.-N. Wang, H.-X. Lou, Nat. Prod. Rep. 2009, 26, 916–
935; b) I. Chrzascik, Crit. Rev. Anal. Chem. 2009, 40, 70–80; c) K.
Xiao, H.-J. Zhang, L.-J. Xuan, J. Zhang, Y.-M. Xu, D.-L. Bai, Stud.
Nat. Prod. Chem. 2008, 34, 453–646; d) F. Orsini, G. Sello, Stud.
Nat. Prod. Chem. 2008, 34, 77–127.
[2] a) R. Singh, H. Kaur, Synthesis 2009, 2471–2491; b) G. C. Tron, T.
Pirali, G. Sorba, F. Pagliai, S. Busacca, A. Genazzani, J. Med. Chem.
2006, 49, 3033–3044.
[3] a) R. F. Guerrero, M. C. Garcia-Parrilla, B. Puertas, E. Cantos-
Villar, Nat. Prod. Commun. 2009, 4, 635–658; b) F. Orallo, Curr.
Med. Chem. 2008, 15, 1887–1898; c) R. Chillemi, S. Sciuto, C. Spata-
fora, C. Tringali, Nat. Prod. Commun. 2007, 2, 499–513; d) F.
Chem. Eur. J. 2011, 17, 3406 – 3419
ꢆ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3417

M. Medarde, R. Pelꢃez et al.
Orallo, Curr. Med. Chem. 2006, 13, 87–98; e) J. A. Baur, D. A. Sin-
clair, Nat. Rev. Drug Discovery 2006, 5, 493–506.
9055; c) J. Gierschner, H. G. Mack, D. Oelkrug, I. Waldner, H. Rau,
J. Phys. Chem. A 2004, 108, 257–263; d) L. Ernst, Prog. Nucl. Magn.
Reson. Spectrosc. 2000, 37, 4–190; e) D. Tanner, O. Wennerstrom,
Tetrahedron Lett. 1981, 22, 2313–2316.
[4] a) S. Y. Liao, J. C. Chen, T. F. Miao, Y. Shen, K. C. Zheng, J.
Enzyme Inhib. Med. Chem. 2010, 25, 421–429; b) D. Simoni, R. Ro-
magnoli, R. Baruchello, R. Rondanin, M. Rizzi, M. G. Pavani, D.
Alloatti, G. Giannini, M. Marcellini, T. Riccioni, M. Castorina,
M. B. Guglielmi, F. Bucci, P. Carminati, C. Pisano, J. Med. Chem.
2006, 49, 3143–3152; c) T. L. Nguyen, C. McGrath, A. R. Hermone,
J. C. Burnett, D. W. Zaharevitz, B. W. Day, P. Wipf, E. Hamel, R.
Gussio, J. Med. Chem. 2005, 48, 6107–6116; d) S. Ducki, G. Macken-
zie, N. J. Lawrence, J. P. Snyder, J. Med. Chem. 2005, 48, 457–465.
[5] a) D. W. Siemann, D. J. Chaplin, P. A. Walicke, Expert Opin. Invest.
Drugs 2009, 18, 189–197; b) A. Gaya, F. Daley, N. J. Taylor, G.
Tozer, U. Qureshi, A. Padhani, R. B. Pedley, R. Begent, D. Wellsted,
J. J. Stirling, G. Rustin, Br. J. Cancer 2008, 99, 321–326; c) D. W. Sie-
mann, W. Shi, Anticancer Res. 2008, 28, 2027–2031; d) D. M. Patter-
son, G. J. S. Rustin, Drugs Future 2007, 32, 1025–1032.
[6] a) S. A. Hill, D. J. Chaplin, G. Lewis, G. M. Tozer, Int. J. Cancer
2002, 102, 70–74; b) C. J. Mooney, G. Nagaiah, P. Fu, J. K. Wasman,
M. M. Cooney, P. S. Savvides, J. A. Bokar, A. Dowlati, D. Wang,
S. S. Agarwala, S. M. Flick, P. H. Hartman, J. D. Ortiz, P. N. Lavertu,
S. C. Remick, Thyroid 2009, 19, 233–240; c) D. J. Chaplin, M. R.
Horsman, D. W. Siemann, Curr. Opin. Invest. Drugs 2006, 7, 522–
528; d) http://www.clinicaltrials.gov/ct2/results?term=fosbretabulin;
http://www.clinicaltrials.gov/ct2/results?term=ZYBRESTAT; http://
www.oxigene.com/pipeline.php.
[15] a) P. Rajakumar, S. Selvam, Tetrahedron 2007, 63, 8891–8901; b) P.
Rajakumar, R. Kanagalatha, Tetrahedron Lett. 2007, 48, 8496–8500.
[16] a) P. Rajakumar, M. G. Swaroop, S. Jayavelu, K. Murugesan, Tetra-
hedron 2006, 62, 12041–12050; b) P. Rajakumar, M. G. Swaroop,
Tetrahedron Lett. 2006, 47, 3019–3022; c) P. Rajakumar, M. G. Swar-
oop, Tetrahedron Lett. 2005, 46, 8543–8546; d) P. Rajakumar, M. G.
Swaroop, Tetrahedron Lett. 2004, 45, 6165–6167; e) R. Gibe, J. R.
Green, G. Davidson, Org. Lett. 2003, 5, 1003–1005; f) R. Gibe, J. R.
Green, G. Davidson, Org. Lett. 2003, 5, 1003–1005; g) G. J. Bodwell,
J. Li, Org. Lett. 2002, 4, 127–130.
[17] a) Lignans: B. Madrigal, P. Puebla, R. Pelꢃez, E. Caballero, M. Med-
arde, J. Org. Chem. 2003, 68, 854–864; b) indolocarbazole alkaloids:
M. Adeva, H. Sahagun, E. Caballero, R. Pelꢃez, M. Medarde; F.
Tomꢅ, J. Org. Chem. 2000, 65, 3387–3394; F. Tomꢅ, J. Org. Chem.
2000, 65, 3387–3394; c) combretastatins: C. Perez-Melero, A. B. S.
Maya, B. del Rey, R. Pelaez, E. Caballero, M. Medarde, Bioorg.
Med. Chem. Lett. 2004, 14, 3771–3774; d) combretastatins: A. B. S.
Maya, C. Perez-Melero, C. Mateo, D. Alonso, J. L. Fernandez, C.
Gajate, F. Mollinedo, R. Pelaez, E. Caballero, M. Medarde, J. Med.
Chem. 2005, 48, 556–568; e) isocombretastatins: R. Alvarez, C. Al-
varez, F. Mollinedo, B. G. Sierra, M. Medarde, R. Pelaez, Bioorg.
Med. Chem. 2009, 17, 6422–6431; f) phenstatins: C. Alvarez, R. Al-
varez, P. Corchete, C. Perez-Melero, R. Pelaez, M. Medarde, Bioorg.
Med. Chem. 2008, 16, 8999–9008; g) phenstatins: C. Alvarez, R. Al-
varez, P. Corchete, C. Perez-Melero, R. Pelaez, M. Medarde, Eur. J.
Med. Chem. 2010, 45, 588–597.
[7] a) I. Barrett, M. Carr, N. OꢀBoyle, L. M. Greene, A. J. S. Knox,
D. G. Lloyd, D. M. Zisterer, M. J. Meegan, J. Enzyme Inhib. Med.
Chem. 2010, 25, 180–194; b) B. Pandit, Y. Sun, P. Chen, D. L. Sack-
ett, Z. Hu, W. Rich, C. Li, A. Lewis, K. Schaefer, P.-K. Li, Bioorg.
Med. Chem. 2006, 14, 6492–6501; c) C. G. Wermuth in The Practice
of Medicinal Chemistry, 2nd ed. (Ed.: C. G. Wermuth) Academic
Press, London, 2003, pp. 215–231.
[18] M. Cushman, D. Nagarathnam, D. Gopal, A. K. Chakraborti, C. M.
Lin, E. Hamel, J. Med. Chem. 1991, 34, 2579–2588.
[19] a) O. Mitsunobu, Synthesis 1981, 1–28; b) D. L. Hughes, Org. React.
1992, 42, 335–656; c) D. L. Hughes, Org. Prep. Proced. 1996, 28,
127–164; d) K. C. Kumara Swamy, N. N. Bhuvan Kumar, E. Balara-
man, K. V. P. Pavan Kumar, Chem. Rev. 2009, 109, 2551–2651.
[20] a) E. V. Dehmlow, S. S. Dehmlow, Phase Transfer Catalysis, 3rd ed.,
VCH, Weinheim, 1993; b) C. M. Starks, C. L. Liotta, M. Halpern,
Phase-Transfer Catalysis, Chapman & Hall, New York, 1994; c) Y.
Sasson, R. Neumann, Handbook of Phase-Transfer Catalysis, Black-
[8] E. K. Singh, R. P. Sellers, L. D. Alexander, S. R. McAlpine, Curr.
Opin. Drug Discovery Dev. 2008, 11, 544–552.
[9] a) O. Querolle, J. Dubois, S. Thoret, F. Roussi, F. Guꢅritte, D. Guꢅ-
nard, J. Med. Chem. 2004, 47, 5937–5944; b) C. J. Dinsmore, M. J.
Bogusky, J. Ch. Culberson, J. M. Bergman, C. F. Homnick, C. Blair
Zartman, S. D. Mosser, M. D. Schaber, R. G. Robinson, K. S.
Koblan, H. E. Huber, S. L. Graham, G. D. Hartman, J. R. Huff,
T. M. Williams, J. Am. Chem. Soc. 2001, 123, 2107–2108; c) A. K.
Ghosh, L. M. Swanson, H. Cho, S. Leshchenko, K. Azhar Hussain,
S. Kay, D. E. Walters, Y. Koh, H. Mitsuya, J. Med. Chem. 2005, 48,
3576–3585; d) X. Hu, K. T. Nguyen, V. C. Jiang, D. Lofland, H. E.
Moser, D. Pei, J. Med. Chem. 2004, 47, 4941–4949.
[10] S. Bartlett, G. S. Beddard, R. M. Jackson, V. Kayser, C. Kilner, A.
Leach, A. Nelson, P. R. Oledzki, P. Parker, G. D. Reid, S. L. Warrin-
er, J. Am. Chem. Soc. 2005, 127, 11699–11708, and references there-
in.
[11] Y. Kim, M. A. Arai, T. Arai, J. O. Lamenzo, E. F. Dean, N. Patter-
son, P. A. Clemons, S. L. Schreiber, J. Am. Chem. Soc. 2004, 126,
14740–14745.
ie Academic
& Professional, London, 1997; d) M. E. Halpern,
Phase-Transfer Catalysis, ACS Symposium Series 659, ACS, Wash-
ington, 1997; e) R. A. Jones, Aldrichimica Acta 1976, 9, 35–45; f) A.
McKillop, J.-C. Fiaud, R. P. Hug, Tetrahedron 1974, 30, 1379–1382.
[21] a) A. R. Tunoori, D. Dutta, G. I. Georg, Tetrahedron Lett. 1998, 39,
8751–8754; b) J. C. Pelletier, S. Kincaid, Tetrahedron Lett. 2000, 41,
797–800; c) A. J. Reynolds, M. Kassiou, Curr. Org. Chem. 2009, 13,
1610–1632; d) R. Dembinski, Eur. J. Org. Chem. 2004, 2763–2772;
e) S. Dandapani, D. P. Curran, Chem. Eur. J. 2004, 10, 3130–3138;
f) E. Valeur, D. Roche, Tetrahedron Lett. 2008, 49, 4182–4185.
[22] a) R. Shukla, D. M. Brody, S. V. Lindeman, R. Rathore, J. Org.
Chem. 2006, 71, 6124–6129; b) J. E. McMurry, Chem. Rev. 1989, 89,
1513–1524; c) M. Ephritikhine, C. Villiers, The McMurry Coupling
and Related Reactions in Modern Carbonyl Olefination: Methods
and Applications (Ed.: T. Takeda), Wiley-VCH, New York, 2004;
d) T. Kawase, N. Ueda, K. Tanaka, Y. Seirai, M. Oda, Tetrahedron
Lett. 2001, 42, 5509–5511.
[12] a) C. Mateo, C. Perez-Melero, R. Pelaez, M. Medarde, Tetrahedron
Lett. 2005, 46, 7055–7057; b) C. Mateo, C. Perez-Melero, R. Pelaez,
M. Medarde, J. Org. Chem. 2005, 70, 6544–6547; c) C. Mateo, R. Al-
varez, C. Perez-Melero, R. Pelꢃez, M. Medarde, Bioorg. Med.
Chem. Lett. 2007, 17, 6316–6320; d) C. Mateo, V. Lopez, M. Med-
arde, R. Pelaez, Chem. Eur. J. 2007, 13, 7246–7256.
[23] a) T. Mukaiyama, T. Sato, J. Hanna, Chem. Lett. 1973, 1041–1044;
b) J. E. McMurry, J. G. Rico, Tetrahedron Lett. 1989, 30, 1169–1172;
c) T. Li, W. Cui, J. Liu, J. Zhao, Z. Wang, Chem. Commun. 2000,
139–140; d) A. Gansꢈuer, Synlett 1997, 363–364; e) F. T. Ladipo,
Curr. Org. Chem. 2006, 10, 965–980; f) A. Gansꢈuer, H. Bluhm,
Chem. Rev. 2000, 100, 2771–2788; g) X. F. Duan, J. X. Feng, G. F.
Zi, Z. B. Zhang, Synthesis 2009, 277–282.
[13] a) F. Vçgtle, P. Neumann, Tetrahedron 1970, 26, 5847–5873; b) Cy-
clophanes, Vols. 1 and 2 (Eds.: P. M. Keehn, S. M. Rosenfeld), Aca-
demic Press, New York, 1983; c) F. Vçgtle, Cyclophane Chemistry,
Wiley, New York, 1993; d) Cyclophanes (Ed.: E. Weber), Springer,
Berlin, 1994; e) Modern Cyclophane Chemistry (Eds.: R. Gleiter, H.
Hopf), Wiley-VCH, Weinheim, 2004; f) L. Ernst, Annu. Rep. NMR
Spectrosc. 2006, 60, 77–143.
[24] a) G. M. Robertson, Comprehensive Organics Synthesis: Carbon-
Carbon s-Bond Formation, Pergamon Press, New York. 1991, 563–
611; b) T. A. Lipski, A. Hilfiker, S. G. Nelson, J. Org. Chem. 1997,
62, 4566–4567.
[14] a) H. R. Darabi, M. A. Arani, A. Motamedi, R. Firouzi, R. Herges,
A. R. Mohebbi, S. M. Nasseri, C. Nather, Supramol. Chem. 2009, 21,
632–637; b) T. Sawada, M. Morita, K. Chifuku, Y. Kuwahara, H.
Shosenji, M. Takafuji, H. Ihara, Tetrahedron Lett. 2007, 48, 9051–
3418
www.chemeurj.org
ꢆ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 3406 – 3419

para–para Stilbenophanes
FULL PAPER
[25] Conformational studies on combretastatins have been carried with
different members of this family: a) L. Padmaja, C. Ravikumar, D.
Sajan, I. H. Joe, V. S. Jayakumar, G. R. Pettit, O. F. Nielsen, J.
Raman Spectrosc. 2009, 40, 419–428; b) D. Sajan, J. P. Abraham, I.
Hubert Joe, V. S. Jayakumar, J. Aubard, O. Faurskov Nielsen, J. Mol.
Struct. 2008, 889, 129–143; c) C. James, G. R. Pettit, O. F. Nielsen,
V. S. Jayakumar, I. Hubert Joe, Spectrochim. Acta Part A 2008, 70,
1208–1216; d) S. Kim, S. Y. Min, S. K. Lee, W.-J. Cho, Chem.
Pharm. Bull. 2003, 51, 516–521.
[26] a) Z. Rappoport, S. E. Biali, Acc. Chem. Res. 1997, 30, 307–314;
b) E. Gur, M. Kaftory, S. E. Biali, Z. Rappoport, J. Org. Chem.
1999, 64, 8144–8148.
[27] The preferred helical disposition of cis-stilbene in the ground state
has been often described: a) D. Yusheng, R. E Allen, J. Chem. Phys.
2003, 119, 10658–10666; b) R. Improta, F. Santoro, J. Phys. Chem.
A 2005, 109, 10058–10067; c) V. Molina, M. Merchꢃn, B. O. Roos,
Spectrochim. Acta 1999, 55, 433–446; d) G. Hohlneicher, M. Muller,
M. Demmer, J. Lex, J. H. Penn, L.-X. Can, P. D. Loesell, J. Am.
Chem. Soc. 1988, 110, 4483–4494.
[30] Line-shape analysis of the mutual exchange for the methoxy and ar-
omatic signals of 18a gave reasonably good simulations with an acti-
vation enthalpy of about 10 kcalmolꢀ1. The simulations were carried
out with the following program: g-NMR v.5.0.6.0, Ivory soft, Bethes-
da, MD, 2006. These results are in good agreement with theoretical
calculations.
[31] The limited water solubility of the compounds did not allow us to
calculate IC50 values.
[32] R. B. Ravelli, B. Gigant, P. A. Curmi, I. Jourdain, S. Lachkar, A.
Sobel, M. Knossow, Nature 2004, 428, 198–202.
[33] A. Dorleans, B. Gigant, R. B. Ravelli, P. Mailliet, V. Mikol, M.
Knossow, Proc. Natl. Acad. Sci. USA 2009, 106, 13775–13779.
[34] M. L. Shelanski, F. Gaskin, C. R. Cantor, Proc. Natl. Acad. Sci. USA
1973, 70, 765–768.
[35] C. Dumortier, M. Gorbunoff, J. M. Andreu, Y. Engelborghs, Bio-
chemistry 1996, 35, 4387–4395.
[36] M. M. Bradford, Anal. Biochem. 1976, 72, 248–254.
[37] Y. Han, A. G. Chaudhary, M. D. Chordia, D. L. Sackett, B. Pꢅrez-
Ramꢄrez, D. G. Kingston, S. Bane, Biochemistry 1996, 35, 14173–
14183.
[38] CCDC-813154 contains the supplementary crystallographic data for
this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
[28] a) F. Momicchioli, I. Baraldi, M. C. Bruni, Chem. Phys. 1982, 70,
161–176; b) J. Almlçf, Chem. Phys. 1974, 6, 135; c) A. Wolf, H. H.
Scbmidtke, J. V. Knop, Theor. Chim. Acta 1978, 48, 37–45.
[29] Systematic conformational searches by using Monte Carlo and MM2
and MM3 molecular force fields were performed with Macromodel
v. 5.1 and clustering of the resulting conformations with XCluster on
Silicon Graphics workstations.
Received: October 5, 2010
Published online: February 23, 2011
Chem. Eur. J. 2011, 17, 3406 – 3419
ꢆ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3419