Synthesis and structure-activity relationships of ferrocenyl tamoxifen derivatives with modifed side chains

Article abstract of DOI:10.1002/chem.200801108

We report here the synthesis and cell-proliferation properties of derivatives of the breast cancer drug tamoxifen, in which the -O(CH ₂)₂N-(CH₃)₂ side chain, responsible for the drug's antiestrogenic properties,

Full text of DOI:10.1002/chem.200801108

DOI: 10.1002/chem.200801108
Synthesis and Structure–Activity Relationships of Ferrocenyl Tamoxifen
Derivatives with Modified Side Chains
Anh Nguyen,[a] Siden Top,[a] Pascal Pigeon,[a] Anne Vessiꢀres,*[a] Elizabeth A. Hillard,[a]
Marie-Aude Plamont,[a] Michel Huchꢁ,[a] Clara Rigamonti,[b] and Gꢁrard Jaouen[a]
[(h5-C5H4)FeCp],
of a cytotoxic pathway. This compound
is also sensitive to the number of
phenol groups as cell death is dimin-
ished when one of the hydroxyl groups
is omitted (5a and 5b). Molecular
modeling studies of the ligand–ERa
binding stability are broadly consistent
with the experimental binding affinity
results for compounds 2, 2a, 2b, 5, 5a,
and 5b. Electrochemical experiments
show that 1–4, 2a, and 2b are stable to
oxidation on the electrochemical time-
scale, unlike 5, 5a, and 5b, and that cy-
totoxicity is related to less positive
phenol oxidation potentials. The SAR
study shows that the presence of a
ketone group and two phenol groups is
necessary for strong receptor binding
and cytotoxic effects, and that all com-
pounds are estrogenic, despite the pres-
ence of a bulky side chain.
ꢀ
Abstract: We report here the synthesis
and cell-proliferation properties of de-
rivatives of the breast cancer drug ta-
ꢀ
yl function ( O
G
n=1–4) and which show a decreasing
affinity for ERa (ER=estrogen recep-
tor) with increasing chain length, act as
estrogens on MCF-7 cells, and mild cy-
totoxics on PC-3 prostate cancer cells,
with IC50 values around 10 mm. The two
monophenolic derivatives of 2, 2a and
2b, which show a reduced affinity for
ERa compared to 2, are also estrogen-
ic, but are only slightly cytotoxic. Final-
ly, we have reexamined compound 5
and discovered that its antiproliferative
effect against the MCF-7 cell line does
not arise from antiestrogenicity as we
had originally suspected, but by means
moxifen, in which the
ACHTUNGTRENNUNG(CH3)2 side chain, responsible for the
OACHTNGUTERNNU(G CH2)2N-
drugꢀs antiestrogenic properties, has
been modified by a ferrocenyl moiety.
We recently reported the diphenol
compound 5, in which this amino chain
had been replaced with an acyl-ferro-
ꢀ
[(h5-C5H4)FeCp])
cenyl ( OACHTUNGTRENNUNG(CH2)2C(O)ACHUTNGTRENNGUN
group, and which showed antiprolifera-
tive effects against both the hormone-
dependent MCF-7 and -independent
MDA-MB-231 breast cancer cell lines.
We now report the results of a struc-
ture–activity relationship (SAR) study,
in which the lateral chain length has
been varied, the ketone group has been
omitted, and the number of phenol
groups has been varied. Compounds 1–
4, with a side chain lacking the carbon-
Keywords: bioinorganic chemistry ·
bioorganometallics · breast cancer ·
ferrocene
·
relationships
Introduction
The breast cancer drug tamoxifen, as the hydroxylated me-
ꢀ
tabolite, OH Tam, can act as an estradiol (E2) antagonist or
agonist depending on the cellular context.[1] It is currently
accepted that the bioligand–estrogen receptor (ER) complex
is not similarly recognized in all cells, and that tamoxifen re-
sistance or estrogen-like activity in some tissues is related to
this structure.[2–5] According to X-ray diffraction analysis, E2
antagonism is broadly the result of the positioning of helix
12 of the ligand binding domain (LBD), which is displaced
from the agonist position by molecules possessing bulky side
[a] Dr. A. Nguyen, Dr. S. Top, Dr. P. Pigeon, Dr. A. Vessiꢁres,
Dr. E. A. Hillard, M.-A. Plamont, Dr. M. Huchꢂ, Prof. G. Jaouen
Laboratoire de Chimie et Biochimie des Complexes Molꢂculaires
UMR CNRS 7576, Ecole Nationale Supꢂrieure de Chimie de Paris
11 rue Pierre et Marie Curie, 75231 Paris Cedex 05 (France)
Fax (+33)01-43-26-00-61
E-mail: a-vessieres@enscp.fr
[b] Dr. C. Rigamonti
Dipartimento di Chimica Organica e Industriale
Universitꢃ degli Studi di Milano
via Venezian, 21 I-20133 Milan (Italy)
[6–8]
ꢀ
chains, such as OH Tam or pure antiestrogens.
There-
fore, control of the bioligand–ER structure through manipu-
ꢀ
lation of the
O
G
(CH3)2 group has been studied in
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200801108.
view of discovering new pure or partial E2 antagonists. How-
684
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 684 – 696

FULL PAPER
ever, modification of substituents has usually led to a de-
crease in antiestrogenicity.[9–14] For example, substitutions
which diminish the basicity of the amine, by replacing the al-
kylamino side chain with N-oxides, quaternary salts, or by
adding fluorinated tethers, resulted in weakened ER bind-
ing, weakened antiproliferative potency, or even a prolifera-
tive effect on ER+ cells.[12–14] The substitution of the amino
side chain by carboxylic acids, such as in GW5638 and
GW7604, has been the only important functional modifica-
vitro. The ferrocenyl phenols were designed by removing
the hydroxyferrocifen side chain altogether, or by replacing
it with a second hydroxyl group. These compounds are not
antiestrogenic, due to the loss of the lateral chain, but show
potent toxicity against both ER+ and ERꢀ cancer cell
lines.[25,29] The generation of hydroxyl radicals by Fenton
chemistry[27,30–33] and the formation of quinone methide me-
tabolites[34] have been proposed as mechanisms of cytotoxic-
ity. It should be noted that other organometallic substitu-
ents, such as [Re(CO)3(Cp)], [Mn(CO)3(Cp)], and
ꢀ
tion of the OH Tam side chain yielding strong antiestrogen-
ic activity in the breast to date.[15–19]
[Ru(Cp)2], did not lend cytotoxic properties to the OH
ꢀ
The covalent tethering of ferrocene to the OH Tam back-
Tam scaffold[35] or phenolic skeleton.[36]
ꢀ
bone has given rise to the “hydroxyferrocifens” and some
Recognizing the sensitive nature of the side chain on anti-
estrogenicity, and the cytotoxicity of some ferrocenyl com-
pounds, we have recently studied the new ferrocenyl triphe-
nylethylene 5, which showed promising in vitro results
against both ER+ and ERꢀ breast cancer cells.[37,38] In de-
signing this compound, we chose to functionalize the ferro-
cenyl group with a ketone, which has been shown to pro-
active ferrocenyl phenols.[20–26] The former, created by the
ꢀ
replacement of the b phenyl group of OH Tam with ferro-
cene, were designed to combine the antiestrogenicity of the
ꢀ
OH Tam scaffold with the cytotoxicity of a ferrocenyl
group,[27,28] resulting in compounds efficacious both on hor-
mone-dependent and -independent breast cancer cells in
Chem. Eur. J. 2009, 15, 684 – 696
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
685

A. Vessiꢁres et al.
mote double metal–ligand exchange reactions to yield other
organometallic compounds, such as those containing 188Re,
186Re, or 99mTc.[39,40] Molecular modeling experiments
showed that the interaction of 5 with the crystal structure of
the antiestrogenic conformation of ERa is highly thermody-
namically favored, particularly due to the interaction with
the ketone and the LBD residue Asp351,[38] and this was
subsequently reflected in a high relative binding affinity
(RBA) value for ERa of 14%. This good receptor recogni-
tion, and the lability of the CpFe moiety, suggests that this
compound could be a useful precursor in the development
of ER-targeted radiopharmaceuticals or imaging agents.
We describe here the synthesis, receptor binding proper-
ties, proliferative/antiproliferative effects, and electrochemis-
try of the first series of hydroxytamoxifen-like compounds
possessing side chains with organometallic termini. To dis-
cover structure–activity relationships (SARs) based on 5, we
have varied three parameters: the length of the side chain
from one to four carbon atoms, the presence of one (2a, 2b,
5a, 5b) or two (1–5) phenolic groups, and the presence (5,
5a, 5b) or absence (1–4, 2a, 2b) of a ketone group adjacent
to the ferrocene.
Results and Discussion
Synthesis: We generally rely on a synthetic route based on
McMurry cross-coupling to obtain the desired alkenes. Re-
agents 4-hydroxypropiophenone and 4,4’-dihydroxybenzo-
phenone were first transformed into their protected forms, 6
and 7, respectively (Scheme 1). Coupling of 6 with 7, by
using TiCl4/Zn in dry THF, gave 8 as a mixture of Z and E
isomers in 67% yield. Compound 8 reacted with the ferro-
cenyl alcohols 9–11, by the Mitsunobu reaction, in the pres-
ence of triphenylphosphine and DEAD in THF for two
days to give 13–16 in 70 to 80% yield. Deprotection was
then performed by saponification of the pivaloate groups
with sodium hydroxide in a THF/H2O solution to generate
2–4, as a mixture of Z and E isomers, in 70 to 92% yield.
However, we failed to obtain 1 (n=1) from saponification.
The action of sodium hydroxide on 13 immediately pro-
Scheme 1. Synthesis of the ferrocenyl derivatives 1, 2, 3, and 4, obtained as a mixture of Z and E isomers. DEAD=diethyl azodicarboxylate.
686
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 684 – 696

First Derivatives of the Breast Cancer Drug Tamoxifen
FULL PAPER
Scheme 2. Synthesis of the ferrocenyl derivatives 2a and 2b obtained as a mixture of Z and E isomers.
duced a deep purple color, and the workup yielded a com-
plex mixture of compounds, among which 1,1,2-tris-(4-hy-
droxyphenyl)but-1-ene was identified. Therefore, we used
the tert-butyldimethylsilyl protecting group, which allows
milder deprotection conditions, and it proved successful
(Scheme 1).
The synthesis of 5 has been described.[38] Similarly, addi-
tion of a-chloroacetylferrocene to the monosodium salts of
8d and 8b (Scheme 1), respectively, obtained from the reac-
tion with NaH, produced 20a and 20b (Scheme 3). Reflux-
ing of 20a and 20b with NaOH in H2O/THF for 6 h gave
(Z+E)-5a and (Z+E)-5b in 73–75% yield (Scheme 3).
The monophenol 2b was prepared in the same way as 2,
but the synthesis started with
6
and 7a to give 8b
Isomerization: One caveat of the McMurry reaction is that
products are usually obtained as mixtures of Z and E iso-
mers. The separation of isomers was achieved with prepara-
tive HPLC for 5 and 5b. As previously observed in the hy-
droxyferrocifen series,[20] the rate of isomerization of 5 and
5b depends strongly on the nature of solvent; they isomer-
ize rapidly in protic solvents, but neither showed isomeriza-
tion after a week in [D6]DMSO as followed by NMR spec-
troscopy. Therefore, even though a pure isomer was first in-
troduced, the results from the cell culture tests are very
(Scheme 1). After alkylation with 10, saponification of the
protected ferrocenyl intermediate 18 gave 2b, as a mixture
of Z and E isomers, in 79% yield (Scheme 2). We found
that the protection/deprotection steps were important to
maximize the yield of the desired product, because when
the unprotected phenol 8c (Scheme 1) reacted directly with
ferrocenyl ethanol 10, a mixture of monoalkylated (Z+E)-
2a (31%) and dialkylated 19 (32%) was obtained
(Scheme 2).
Scheme 3. Synthesis of the ferrocenyl derivatives 5, 5a and 5b obtained as a mixture of Z and E isomers.
Chem. Eur. J. 2009, 15, 684 – 696
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
687

A. Vessiꢁres et al.
Table 2. Energy variation (DE) values for the binding of the complexes
to ERa.
likely the combined activity of the Z and E mixture, but
remain pure isomers for the low-temperature receptor bind-
ing affinity (RBA) tests. All the compounds of the 1–4
series, including 2a, could be separated by HPLC, but were
found to isomerize quite rapidly, roughly within one hour, as
followed by NMR spectroscopy in CDCl3. Therefore, we did
not separate the isomers preparatively, and all of the biolog-
ical tests were performed with a mixture. Finally, it was not
possible to observe the individual signals of the Z and E iso-
mers of 2b and 5a by analytical HPLC, and thus a mixture
of isomers was used in all tests.
Compound
DE [kcalmolꢀ1
]
Compound
DE [kcalmolꢀ1
]
ꢀ
(Z)-OH Tam
(Z)-5
ꢀ140.6
ꢀ106.9
ꢀ79.2
ꢀ68.2
ꢀ79.7
ꢀ69.4
ꢀ58.8
(E)-5
ꢀ57.1
ꢀ67.6
ꢀ32.2
ꢀ38.7
ꢀ61.9
ꢀ28.8
(Z)-5a
(Z)-5b
(Z)-2
(Z)-2a
(Z)-2b
(E)-5a
(E)-5b
(E)-2
(E)-2a
(E)-2b
sults for receptor binding will now be discussed in terms of
SARs.
The presence of a carbonyl group on the side chain gener-
ally enhanced receptor binding. This is experimentally dem-
RBA and molecular-modeling studies on ligand–ER com-
plexation: The affinities of the compounds were determined
for ERa and the results are summarized in Table 1. These
ꢀ
affinities were not as high as that of OH Tam, probably due
onstrated for (Z)-5/ACTHUGNTERN(UNG Z+E)-2, 5a/2a, and 5b/2b, and these
to the greater steric hindrance of the ferrocenyl group as
compared to a dimethylamine moiety. In the alkyl series 1–
4, RBA values for ERa decreased as the side chains became
longer.
results are in good agreement with theoretical predictions.
This stability seems to arise from hydrogen bonding be-
tween the ketone and Asp351. For example, as previously
reported, DE found for (Z)-5 was ꢀ89 kcalmolꢀ1 with an in-
teraction between Asp351 and Fe, but the binding became
more exothermic (ꢀ106 kcalmolꢀ1) when a hydrogen bond
between Asp351 and C=O was modeled.[38] Figure 1 shows
the theoretical hydrogen-bonding interaction between 5a
and Asp351, which is absent for 2a as it lacks the ketone
function.
Docking experiments for each isomer of 2, 2a, 2b, 5, 5a,
and 5b in the ligand binding domain (LBD), derived from
the structure of ERa crystallized with OH Tam, showed
that the cavity containing the amino side chain is large
enough to host the ferrocenyl group, and all molecules lie
within the LBD similarly to OH Tam, with the side chain
oriented towards Asp351. Bioligand–receptor stability
values are given in Table 2, with more negative values indi-
cating greater stability. The experimental and theoretical re-
ꢀ
ꢀ
Depending on the configuration of the bioligand, hydrox-
yl groups can bind to Glu353, Arg394, and His524, and gen-
erally the loss of a hydroxyl group resulted in the loss of
Table 1. RBA values, logPo/w, and effect on the growth of cancer cells of compounds 1–5.
Compound
R1
R2
R3
RBA for ERa [%][a]
logPo/w
Effect on the growth of cancer cells [%][b]
MCF-7[c]
PC3[d] (IC50 [mm])[e]
17b-E2
100[f]
3.5
253[g]
59[i]
181
173
118
107
166
192
54[k]
62[k]
108
178
164
–
–
38.5[h]
3.2 (Z), 3.4 (E)
6.7 (Z), 5.9 (E)
5.7 (Z), 6.6 (E)
6.2 (Z), 7.1 (E)
6.6 (Z), 7.5 (E)
7.9 (Z), 8.2 (E)
7.9
4.6
5.1
5.8
3.6
5.9
ꢀ
(Z+E)-OH Tam
(Z+E)-1
(Z+E)-2
(Z+E)-3
(Z+E)-4
(Z+E)-2a
(Z+E)-2b
T
OH
OH
OH
OH
H
OH
OH
OH
H
OH
OH
OH
OH
OH
H
OH
OH
OH
H
CH2
A
11.9ꢁ0.2
0.9ꢁ0.3
0.45ꢁ0.05
0.24ꢁ0.02
0.16ꢁ0.02
0.13*
65 (12ꢁ1)
48 (9.8ꢁ0.1)
51 (10.2ꢁ0.3)
76 (12ꢁ2)
90
E
R
N
N
N
U
R
N
N
84
(Z)-5
(E)-5
G
14ꢁ1[j]
49 (7.8ꢁ0.6)
56 (8.3ꢁ0.7)
83
102
96
T
1.19ꢁ0.05[j]
4.1ꢁ0.7
2.3ꢁ0.4
4.6ꢁ0.4
A
ACHTUNGTRENNUNG
(Z)-5b
(E)-5b
OH
OH
T
H
G
[a] Mean of two experiments ꢁ range, except where an asterisk * appears; values for ERb are included in the Supporting Information. [b] Control=
cells without added compound, set at 100% after 5 days of culture in a medium without phenol red. [c] Hormone-dependent breast cancer cells, incuba-
tion with 1 mm except when specified. [d] Hormone-independent prostate cancer cells, incubation with 10 mM; [e] IC50 values were determined when the
percentage of cell growth was lower than 80%, mean of two experiments ꢁ range; [f] Value by definition. [g] Incubation with 1 nm. [h] Value from refer-
ence [20]. [i] Value from reference [25]. [j] Value from reference [38]. [k] Incubation with 10 mm.
688
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 684 – 696

First Derivatives of the Breast Cancer Drug Tamoxifen
FULL PAPER
Compounds 1, 2, 3, and 4
showed an RBA-dependant
proliferative effect on MCF-7
cells, which indicated an estro-
genic character. Conversely,
they had a significant antiproli-
ferative effect on PC-3 cells
with IC50 values around 10 mm,
with no correlation between cy-
totoxicity and chain length.
Compounds 2a and 2b also had
a proliferative effect on MCF-7
cells, but only a modest effect
on PC-3 cells. Quite surprising-
ly none of the complexes
showed an antiproliferative
effect greater than 20% inhibi-
tion at 10 mm for the MBA-MB-
231 cells (Supporting Informa-
Figure 1. Representation of (Z)-5a (left) and (Z)-2a (right) docked in ERa, assuming a direct interaction be-
tween the ketone group of 5a and Asp351 (circled).
theoretical ligand–ER stability. The role of the hydroxyl
groups is well illustrated by the striking difference in bind-
ing observed for the two isomers of 5; the change in config-
uration from Z to E results in a dramatic drop in the affinity
for both receptor isoforms. The Z isomer, which in this case
is also the trans isomer,1 binds more strongly, and this obser-
tion), whereas the effect on PC-3 is more pronounced. This
is the first time that a significant difference has been ob-
served between these two cell lines in our laboratory.
Compounds (Z)- and (E)-5 showed significant and quite
similar antiproliferative effects on both MCF-7 and PC-3
cells (Table 1), and they are the only compounds to inhibit
the proliferation of both cell lines. To determine whether
the antiproliferative effect of 5 on MCF-7 cells was a result
ꢀ
vation is in agreement with previous results with OH Tam,
hydroxyferrocifens, and other triphenylethylenes, in which
ER has a preference for the trans over the cis isomer.[10,20,29]
Molecular modeling on ERa suggests that (Z)-5 is associat-
ed with Asp351 via the ketone, with Glu353 and Arg394 via
the a phenol, and His524 via the b phenol. However, due to
the geometry of (E)-5, it cannot engage in hydrogen bond-
ing with His524, and the predicted stability is reduced. The
situation is similar for (Z)- and (E)-2 from a theoretical per-
spective.
Lipophilicity: Lipophilicity is expressed as the octanol/water
partition coefficient, logACTHNUTRGNEUGN(Po/w), determined by HPLC
(Table 1). As expected, the ferrocenyl derivatives yielded
ꢀ
higher log
G
more unusual, however, is that the E isomers of the com-
pounds are considerably more lipophilic than the corre-
sponding Z isomers. For example, while the difference be-
ꢀ
tween (E)- and (Z)-OH Tam is slight (D=0.2), that of (E)-
and (Z)-5b is significant (D=2.3).
Figure 2. Effect of E2 and of (Z)- and (E)-5 on the proliferation of MCF-
7 cells (hormone-dependent breast cancer cells) after 5 days of culture.
Nontreated MCF-7 cells are used as the control (C) set at 100%. Mean
of two separate experiments ꢁ range.
Cell proliferation: The influence of the compounds on the
proliferation of cancer cells has been tested on the hor-
mone-dependent MCF-7 breast cancer cells, the hormone-
independent MDA-MB-231 breast cancer cells, and the hor-
mone-independent PC-3 prostate cancer cells, and results
are given in Table 1 (MDA-MB-231 results are included as
Supporting Information).
of ER binding, the cells were incubated with 10 mm of 5 in
the presence and absence of 1 nm E2. As shown in Figure 2,
the addition of E2 did not reverse the antiproliferative effect
of (Z)- or (E)-5, which indicated that this effect is cytotoxic
and not antiestrogenic. The estrogenic properties of a com-
pound are known to be expressed at low concentrations
1
The terminology of cis/trans is used to designate the relative position of
the ethyl group to the phenyl bearing the side chain. One should bear
in mind that the trans orientation does not always correspond to the Z
configuration.
Chem. Eur. J. 2009, 15, 684 – 696
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
689

A. Vessiꢁres et al.
(10ꢀ8–10ꢀ10 m), and we found that (Z)- and (E)-5 are indeed
strongly estrogenic at 10ꢀ9 m (Figure 2). Thus, the activity of
5 on the proliferation of MCF-7 cells seems to be a combi-
nation of an estrogenic character and a cytotoxic compo-
nent; at high concentrations (>1 mm), cytotoxicity is domi-
nant, and at low concentrations, the estrogenic effect is
more strongly expressed. Thus, despite the bulky side chain,
5 acts like the ferrocenyl phenols previously described.[25,29]
It should be mentioned that other molecules designed to be
antiestrogens in the breast, for example a series of trifluoro-
methyl-substituted phenylvinyl E2 compounds, exhibited es-
trogenic properties on MCF-7 cells, regardless of their steri-
cally demanding side group.[41]
The ketone and hydroxy functionalities both affect the
biological behavior of the studied compounds. The lack of
the ketone group seems to increase estrogenicity of the
compounds. For example, comparing the antiproliferative
activity of 5 and 2, we observe that they are both cytotoxic
on PC-3 cells, and that they have a similar activity on these
cells (IC50 values between 7.8 and 9.8 mm). However, only 5
inhibited the proliferation of the ER+MCF-7 cells, whereas
2 had a strongly proliferative effect on those cells. The pres-
ence of a second hydroxyl group, on the other hand, mark-
edly increased the compoundsꢀ cytotoxicity, and the com-
pounds with only one phenol have no or only a modest anti-
proliferative effect on PC-3 cells. The importance of the
phenol groups has also been observed in the ferrocenyl phe-
ꢀ
ꢀ
nols, in which Fc diOH is more toxic than Fc OH (IC50 =
0.6 and 1.1 mm, respectively).[22]
Electrochemistry: Since it has been suggested that the cyto-
toxic activity of the ferrocenyl derivatives may originate
from their oxidized forms,[27,30–33] the electrochemical behav-
ior of the compounds was examined. At all scan rates, com-
0/+
pounds 1–4 gave rise to a reversible FeCp2
couple and a
more positive irreversible phenol oxidation wave. The redox
0/+
potentials for the FeCp2 process ranged from 0.432 (3) to
0.506 V (1) versus SCE, and there was no correlation with
the redox potential, the number of carbon atoms in the fer-
rocenyl chain, or the cytotoxicity of the compounds. Com-
pounds 5, 5a, and 5b exhibited more complex behavior. At
low scan rates, the oxidation of ferrocene was irreversible,
although a reduction wave began to appear at higher scan
rates. Comparing the CVs of those compounds possessing
the carbonyl group, to their alkyl analogues (5a/2a, 5b/2b,
5/2; Figure 3), one finds that the ferrocene oxidation waves
of the former were higher in intensity and less reversible
than those of the latter at low scan rates, but the two waves
are similar in intensity and reversibility at high scan rates.
This can be interpreted as a slow degradation of radical
cation, which yields a product that is oxidized at a less-posi-
tive or equal potential to that of the ferrocene moiety.
Figure 3. CVs of the alkylFc compounds (c) compared to their acylFc analogues (a) at low (0.1 Vsꢀ1) and high (20 Vsꢀ1) scan rates in DMF/0.1m
Bu4NBF4.
690
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 684 – 696

First Derivatives of the Breast Cancer Drug Tamoxifen
FULL PAPER
Irreversible phenol oxidation potentials ranged between
0.867 and 1.17 V (at 0.1 Vsꢀ1) and were bimodal in distribu-
tion. The compounds with the least positive phenol oxida-
tion potentials, 1, 2, 3, and 4 possess two phenol groups,
whereas those with more positive oxidation potentials, 2a
and 2b, have only one. The lowering of the phenol oxidation
potential is probably due to the additional resonance stabili-
zation that the second phenol group imparts to the electro-
chemically generated phenoxy radical. Thus, the presence of
two phenol groups gave rise to more accessible phenol oxi-
dation potentials, which correlate with the cytotoxicity of
the compounds; this suggests that the generation of active
phenoxy radicals or quinones could play a role in the cyto-
toxicity of these compounds.
Experimental Section
General considerations: All air-sensitive reactions were carried out under
an argon atmosphere, by using standard Schlenk and vacuum-line tech-
niques. “Standard workup” refers to extraction of the reaction mixture
with dichloromethane, washing of the organic phase with water, drying
over MgSO4, filtering, removal of the solvent under reduced pressure,
and purification by flash chromatography. Dry THF and diethyl ether
were obtained by distillation from sodium/benzophenone. Preparative
TLC chromatography was performed on silica gel 60 GF254. Flash chro-
matography was performed on silica gel Merck 60 (0.040–0.060 mm), or
when necessary on aluminum oxide. HPLC spectra were measured by a
Shimadzu instrument. HPLC system: Kromasil C18 columns (analytical:
4.6ꢅ250, preparative: 20ꢅ250), eluent: water/acetonitrile mixture. IR
spectra were obtained on a FTIR BOMEM Michelson-100 spectrometer
1
equipped with a DTGS detector. H and 13C NMR spectra were recorded
on a 300 MHz Bruker spectrometer and the results d given in ppm. Mass
spectrometry was performed with a Nermag R 10–10C spectrometer.
HRMS was carried out with a MStation 700 (JEOL) spectrometer. Ele-
mental analyses were performed by the Microanalysis Service of ICSN
(Institut de Chimie des Substances Naturelles), Gif-sur-Yvette. Com-
pounds 6a, 7a, and 7c were purchased from Acros and used as received.
The syntheses of compounds Fc-diOH,[20] 5,[38] 7,[42] 7b,[43] 8c,[25] 8d,[42] 9,
10, 11, and 12[44] have been previously described.
Conclusions
We have described the first series of compounds in which
ꢀ
the amino side chain of OH Tam has been replaced by an
organometallic moiety. Although this work was inspired by
preliminary molecular-modeling results, which suggested
that these compounds should act as strong antiestrogens, all
of the compounds gave rise to estrogenic effects. Clearly, a
“good fit” of the bioligand with the antiestrogenic form of
the LBD crystal structure of ERa is not predictive of the
antiestrogenic activity of these molecules.
Synthesis and characterization
(p-Trimethylacetoxy)propiophenone (6): Sodium hydride (60% in oil,
4.8 g, 0.12 mol) was slowly added to p-hydroxypropiophenone (15 g,
0.1 mol) in dry THF (200 mL) and the resulting reaction mixture was
stirred for 30 min. Trimethylacetyl chloride (14 mL, 0.12 mol) was added
and the mixture stirred for a further 2 h. The mixture was poured into
water (200 mL) and underwent the standard workup to give 6 as a white
solid (quant.). M.p. 378C; 1H NMR (300 MHz, CDCl3): d=7.99 (d, J=
8.8 Hz, 2H; CHarom), 7.14 (d, J=8.8 Hz, 2H; CHarom), 2.96 (q, J=7.2 Hz,
2H; CH2), 1.36 (s, 9H; CH3 of tBu), 1.22 ppm (t, J=7.2 Hz, 3H; CH3 of
Et); 13C NMR (75 MHz, CDCl3): d=199.6 (CO), 176.6 (COO), 154.7 (C),
134.3 (C), 129.5 (2CHarom), 121.7 (2CHarom), 39.2 (C, tBu), 31.8 (CH2),
27.0 (3CH3, tBu), 8.2 ppm (CH3, Et); IR (KBr): n˜ =2987, 2929, 2872
(CH2, CH3), 1750 cmꢀ1 (CO); ESI-MS (H2O/MeOH 1:9): m/z: 257.5
[M+Na]+, 289.3 [M+Na+MeOH]+, 491.7 [2M+Na]+; elemental analysis
calcd (%) for C14H18O3: C 71.76, H 7.74; found: C 71.71, H 7.64.
The influence of the side-chain length, phenol groups, and
electron-withdrawing ketone group adjacent to the ferrocen-
yl moiety was studied. In the series of compounds lacking
the ketone, a longer ferrocenyl side chain corresponded to
lower binding affinities and a lower activity on the prolifera-
tion of the MCF-7 cells, although no relationship was ob-
served for cytotoxic effects or electrochemical behavior. The
tethering of a ketone function adjacent to the ferrocenyl
entity conveyed an additional stabilizing interaction with the
ER, accounting, in part, for the good affinity of 5 with ERa
found experimentally. The ketone group is also responsible
for irreversible ferrocene oxidation behavior. Although this
group contributed to the stronger cytotoxic activity of 5 rel-
ative to its analogues, its mere presence is not sufficient.
The loss of one hydroxyl group significantly weakens the cy-
totoxic activity, and indeed, the presence of two phenol
groups seems to be the primary factor in the cytotoxicity of
these types of compounds.
Therefore, it requires the presence of both the ketone
function adjacent to the ferrocene group and the presence
of two phenols to yield a cytotoxic molecule, with good
binding affinity, and a noteworthy antiproliferative activity.
This molecule 5 is an interesting prototype for further ex-
ploitation, especially for radioimaging and radiotherapy ap-
plications as it has been shown that keto–ferrocenyl deriva-
tives can be used as stable precursors of rhenium and tech-
netium derivatives.[39,40]
4-(tert-Butyldimethylsilyloxy)propiophenone (6b): tert-Butyldimethyl-
chlorosilane (7.54 g, 50 mmol) and imidazole (8.51 g, 125 mmol) were
added to a solution of 4-hydroxypropiophenone (7.51 g, 50 mmol) in dry
DMF (30 mL), and the reaction mixture was stirred for 3 h. The solution
was poured into a 5% solution of NaHCO3 (200 mL) and underwent the
standard workup to give 17 as a colorless oil. Another synthesis of 6b
has been published and the characterization of 6b was identical to that
reported.[45]
1-(4-Hydroxyphenyl)-1,2-bis(4-trimethyacetoxyphenyl)but-1-ene
(8):
TiCl4 (5.3 mL, 48 mmol) was added dropwise under an inert atmosphere
to a suspension of Zn (5.36 g, 82 mmol) in dry THF (100 mL). After the
Zn/TiCl4 suspension had been refluxed for 2 h, 6 (2.3 g, 10 mmol) and 7
(3.58 g, 12 mmol) dissolved in dry THF (50 mL) were added. The mixture
was heated at reflux for 2 h. After cooling, the mixture was hydrolyzed
by acidified water (200 mL), followed by the standard workup. The crude
product was recrystallized in ethanol, yielding 8 as a white powder
(3.36 g, 67%; isomer ratio: 1:5). Major isomer: 1H NMR (300 MHz,
CDCl3): d=7.22 (d, J=8.6 Hz, 2H; CHarom), 7.09 (d, J=8.6 Hz, 2H;
CHarom), 7.05 (d, J=8.6 Hz, 2H; CHarom), 6.87 (d, J=8.6 Hz, 2H;
CHarom), 6.71 (d, J=8.6 Hz, 2H; CHarom), 6.47 (d, J=8.6 Hz, 2H;
CHarom), 2.45 (q, J=7.3 Hz, 2H; CH2), 1.37 (s, 9H; CH3 of tBu), 1.34 (s,
9H; CH3 of tBu), 0.91 ppm (t, J=7.3 Hz, 3H; CH3 of Et); 13C NMR
(75 MHz, CDCl3): d=177.3 (CO), 153.8 (C), 149.7 (C), 149.2 (C), 141.0
(C), 140.8 (C), 139.7 (C), 137.8 (C), 135.1 (C), 132.1 (2CHarom), 130.5
(2CHarom), 130.4 (2CHarom), 121.1 (2CHarom), 120.9 (2CHarom), 114.5
(2CHarom), 39.0 (2C, tBu), 28.9 (CH2), 27.1 (2ꢅ3CH3, tBu), 13.5 ppm
ꢀ
ꢀ
(CH3, Et); IR (KBr): n˜ =3407 (O H), 2977 (C Harom), 1749, 1726 (CO),
1610 (C=C), 1504 cmꢀ1 (C=C arom); MS (EI): m/z: 500 [M]+, 57 [tBu]+;
Chem. Eur. J. 2009, 15, 684 – 696
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
691

A. Vessiꢁres et al.
HRMS (EI, 70 eV): m/z: calcd for C32H36O5: 500.2563 [M]+; found:
500.2574.
137.8 (C), 135.7, 135.0 (C), 132.0, 131.7 (2CHarom), 130.6, 130.5 (2CHarom),
130.5, 130.4 (2CHarom), 121.1, 121.0 (2CHarom), 120.9, 120.4 (2CHarom),
114.1, 113.5 (2CHarom), 84.8 (C, C5H4), 68.6 (5CH, Cp+2CH, C5H4), 68.4,
68.2 (OCH2), 67.5, 67.4 (2CH, C5H4), 39.0 (2C, tBu), 29.5, 29.0 (CH2),
27.1 (2ꢅ3CH3, tBu), 13.6 ppm (CH3); IR (KBr): n˜ =1752 cmꢀ1 (CO); MS
(CI, NH3) : m/z: 712 [M+H]+, 730 [M+NH4]+; elemental analysis calcd
(%) for C44H48FeO5: C 74.15, H 6.78; found: C 74.03, H 6.84.
1,2-Bis[4-(tert-butyl-dimethylsilyloxy)phenyl]-1-(4-hydroxyphenyl)but-1-
ene (8a): The same procedure as that of 8 was used with 6b (2.64 g,
10 mmol) and 7b (3.28 g, 10 mmol). After standard workup, the crude
product was chromatographed on a silica-gel column with dichlorome-
thane as the eluent to yield pure 8a as an oil (75%; isomer ratio: 1:1).
1H NMR (300 MHz, CDCl3): d=7.09, 7.08 (d, J=8.5 Hz, 2H; CHarom),
6.95 and 6.94 (d, J=8.6 Hz, 2H; CHarom), 6.81 and 6.79 (d, J=8.5 Hz,
2H; CHarom), 6.75–6.60 (m, 4H; CHarom), 5.13, 4.84 (s, 1H; OH), 2.45 (q,
J=7.3 Hz, 2H; CH2), 1.01, 0.97 (s, 9H; tBuSi), 0.99 (t, J=7.3 Hz, 3H;
CH3), 0.97, 0.94 (s, 9H; tBuSi), 0.48, 0.45 (d, J=8.6 Hz, 2H; CHarom),
0.23, 0.17 (s, 6H; SiMe2), 0.17, 0.12 ppm (s, 6H; SiMe2); 13C NMR
(75 MHz, CDCl3): d=154.2 (C), 153.7 (C), 153.4, 153.3 (C), 140.7 (C),
137.4 (C), 137.0, 136.6 (C), 136.5, 136.2 (C), 135.6, 135.5 (C), 132.1, 131.9
(2CHarom), 130.8, 130.7 (2CHarom), 130.7, 130.6 (2CHarom), 119.6, 119.5
(2CHarom), 119.5, 118.8 (2CHarom), 114.9, 114.2 (2CHarom), 28.9, 28.8
(CH2), 25.7 (2tBu), 18.2 (2C, tBuSi), 13.7 (CH3), ꢀ4.4 ppm (2SiMe2); IR
(KBr): n=3428 (OH), 1260 cmꢀ1 (SiCH3); MS (EI, 70 eV): m/z: 560
[M]+, 545 [MꢀCH3]+, 57 [tBu]+; HRMS (EI, 70 eV): m/z: calcd for
C34H48O3Si2: 560.3142 [M]+; found: 560.3132.
1-[4-(3-Ferrocenylpropoxy)phenyl]-1,2-bis(4-trimethylacetoxyphenyl)but-
1-ene (15): The reaction was accomplished with ferrocenylpropanol 11
(0.586 g, 2.4 mmol). Yield: 70%; 1H NMR (300 MHz, CDCl3): d=7.24–
6.74 (m, 10H; CHarom), 6.73, 6.57 (d, J=8.8 Hz, 2H; CHarom), 4.14–4.02
(m, 9H; CpFeC5H4), 3.99, 3.86 (t, J=6.3 Hz, 2H; OCH2), 2.59–2.39 (m,
4H; 2CH2), 2.09–1.86 (m, 2H; CH2), 1.37, 1.34 (s, 9H; tBu), 1.34, 1.30 (s,
9H; tBu), 0.91, 0.90 ppm (t, J=7.4 Hz, 3H; CH3); 13C NMR (75 MHz,
CDCl3): d=176.2 (2CO), 157.1, 156.2 (C), 152.5, 148.8 (C), 151.6, 148.4
(C), 140.2 (C), 139.8 (C), 138.7 (C), 136.9 (C), 134.1 (C), 131.1, 130.9
(2CHarom), 129.6 (2CHarom), 129.5 (2CHarom), 120.2, 119.5 (2CHarom),
120.0 (2CHarom), 113.2, 112.6 (2CHarom), 87.4 (C, C5H4), 67.6 (5CH, Cp),
67.2 (2CH, C5H4), 66.4, 66.3 (2CH, C5H4), 66.2 (OCH2), 38.2 (2C, tBu),
29.6 (CH2), 28.8, 28.1 (CH2), 26.3 (2ꢅ3CH3, tBu), 25.0 (CH2), 12.7 ppm
(CH3); IR (KBr): n˜ =1752 cmꢀ1 (CO); MS (EI): m/z: 726 [M]+, 661
[MꢀCp]+, 199 [CpFe
(h5-C5H4)CH2]+, 121 [CpFe]+, 57 [tBu]+; elemental
ACHTUNGTRENNUNG
analysis calcd (%) for C45H50FeO5: C 74.37, H 6.93; found: C 74.28, H
6.99.
1-(4-Hydroxyphenyl)-1-(4-trimethylacetoxyphenyl)-2-phenylbut-1-ene
(8b): The same procedure as that of 8 was used with 4-hydroxybenzophe-
none (4 g, 20 mmol) and 6 (4.69 g, 20 mmol) to give 8b (70%; isomer
ratio: 1:1). 1H NMR (300 MHz, CDCl3): d=7.30–6.46 (m, 13H; CHarom),
4.99, 4.74 (s, 1H; OH), 2.40, 2.38 (q, J=7.4 Hz, 2H; CH2), 1.34, 1.33 (s,
9H; CH3 of tBu), 0.86, 0.85 ppm (t, J=7.4 Hz, 3H; CH3 of Et); 13C NMR
(75 MHz, CDCl3): d=177.2 (CO), 154.4, 153.6 (C), 149.2 (C), 143.6,
143.1 (C), 141.0, 140.5 (C), 139.8, 139.7 (C), 138.7, 138.6 (C), 136.0, 135.4
(C), 132.1, 130.8 (2CHarom), 130.7, 130.6 (2CHarom), 129.4, 128.1
(2CHarom), 127.4, 126.6 (2CHarom), 125.8, 125.6 (CHarom), 120.9, 120.8
(2CHarom), 115.0, 114.4 (2CHarom), 39.1 (Cq, tBu), 28.9 (CH2), 27.1 (CH3,
tBu), 13.6 ppm (CH3, Et); MS (EI): m/z: 400 [M]+, 57 [tBu]+; elemental
analysis calcd (%) for C27H28O3: C 80.90, H 6.99; found: C 80.77, H 6.96.
1-[4-(4-Ferrocenylbutoxy)phenyl]-1,2-bis(4-trimethylacetoxyphenyl)but-
1-ene (16): The reaction was accomplished with ferrocenylbutanol 12
(0.620 g, 2.4 mmol). Yield: 74%; 1H NMR (300 MHz, CDCl3): d=7.30–
6.50 (m, 12H; CHarom), 4.12, 4.09 (s, 5H; Cp), 4.06, 4.05 (s, 2H; C5H4),
4.04, 4.03 (s, 2H; C5H4), 3.99, 3.85 (t, J=6.2 Hz, 2H; OCH2), 2.55–2.30
ꢀ
(m, 4H; 2CH2), 1.90–1.52 (m, 4H; CH2 CH2), 1.38, 1.35, 1.30, 1.27 (s,
18H; 2tBu), 0.94, 0.92 ppm (t, J=7 Hz, 3H; CH3); 13C NMR (75 MHz,
CDCl3): d=177.2, 177.1, 177.0, 176.9 (2CO), 157.2, 155.0 (C), 149.7 (C),
149.2 (C), 141.1 (C), 140.6, 140.5 (C), 139.6 (C), 137.9, 137.8 (C), 135.6,
134.9 (C), 132.0, 131.7 (2CHarom), 130.5 (2CHarom), 130.5, 130.4
(2CHarom), 121.1, 120.9 (2CHarom), 120.9, 120.4 (2CHarom), 114.0, 113.4
(2CHarom), 89.0 (C, C5H4), 68.5 (5CH, Cp), 68.1 (2CH, C5H4), 67.7, 67.4
(OCH2), 67.1 (2CH, C5H4), 39.1 (C, tBu), 39.0 (C, tBu), 29.7, 29.0 (CH2),
29.3 (CH2), 29.2 (CH2), 27.5 (CH2), 27.2 (2ꢅ3CH3, tBu), 13.6 ppm (CH3);
General procedure for the preparation of 13, 14, 15, and 16: A solution
of DEAD (0.42 g, 2.4 mmol) in dry THF (3 mL) was added dropwise at
08C to a solution of ferrocenyl alcohol 9, 10, 11, or 12 (2.4 mmol), respec-
tively. Compound
8 (1 g, 2 mmol) and triphenylphosphine (0.74 g,
ACHTUNGTRENNUNG
IR (KBr): n˜ =1754 cmꢀ1 (CO); MS (EI): m/z: 740 [M]+, 199 [CpFe(h5-
2.8 mmol) in dry THF (12 mL) were then added. The reaction was stirred
at room temperature for 48 h. The solvent was evaporated under vacuum
and the residue was purified by aluminum oxide column chromatography
(petroleum ether) to give 13, 14, 15, or 16 (isomer ratio: 1:1) as yellow
solids. These compounds were recrystallized from ether/pentane.
C5H4)CH2]+, 121 [CpFe]+; elemental analysis calcd (%) for C46H52FeO5:
C 74.58, H 7.07; found: C 74.51, H 7.36.
1-[4-(2-Ferrocenylethoxy)phenyl]-1-phenyl-2-(4-trimethylacetoxyphenyl)-
but-1-ene (18): The same procedure as that for 13 was used to synthesize
18, except that the substituted butene 8 was replaced by 8b (0.801 g,
2 mmol). Yield: 96%; isomer ratio: 55:45; 1H NMR (300 MHz, CDCl3):
d=7.40–6.75 (m, 11H; CHarom), 6.87 and 6.58 (d, J=8.5 Hz, 2H; CHarom),
4.25–4.00 (m, 9H; CpFeC5H4), 4.11, 3.97 (t, J=7.0 Hz, 2H; OCH2), 2.85,
2.75 (t, J=7.0 Hz, 2H; CH2), 2.51, 2.45 (q, J=7.4 Hz, 2H; CH2), 1.35,
1.34 (s, 9H; tBu), 0.96, 0.94 ppm (t, J=7.4 Hz, 3H; CH3); 13C NMR
(75 MHz, CDCl3): d=177.0 (CO), 157.7, 156.9 (C), 149.2 (C), 143.7,
143.2 (C), 141.0, 140.4 (C), 139.7 (C), 138.8, 138.7 (C), 135.7, 135.0 (C),
131.9, 130.8 (2CHarom), 130.6 (2CHarom), 130.6, 129.4 (2CHarom), 128.1,
127.4 (2CHarom), 126.6, 125.8 (CHarom), 120.9, 120.8 (2CHarom), 114.1,
113.5 (2CHarom), 84.7 (C, C5H4), 68.5 (5CH, Cp+2CH, C5H4), 68.4, 68.2
(OCH2), 67.5, 67.4 (2CH, C5H4), 39.0 (C, tBu), 29.6, 29.5 (CH2), 29.0
(CH2), 27.1 (3CH3, tBu), 13.6 ppm (CH3); IR (KBr): n˜ =1756, 1747 cmꢀ1
(CO); MS (CI, NH3): m/z: 612 [M+H]+, 630 [M+NH4]+; elemental anal-
ysis calcd (%) for C39H40FeO3: C 76.46, H 6.58; found: C 76.56, H 6.61.
1-[4-(Ferrocenylmethoxy)phenyl]-1,2-bis[4-(trimethylacetoxy)phenyl]but-
1-ene (13): The reaction was accomplished with 0.519 g (2.4 mmol) of fer-
rocenylmethanol 9. Yield: 83%; 1H NMR (300 MHz, CDCl3): d=7.25–
6.50 (m, 12H; CHarom), 4.81, 4.67 (s, 2H; OCH2), 4.34, 4.26 (t, J=1.8 Hz,
2H; C5H4), 4.21, 4.16 (t, J=1.8 Hz, 2H; C5H4), 4.20, 4.15 (s, 5H; Cp),
2.46 (q, J=7.4 Hz, 2H; CH2), 1.37, 1.35, 1.34, 1.30 (s, 18H; 2tBu),
0.92 ppm (t, J=7.4 Hz, 3H; CH3); 13C NMR (75 MHz, CDCl3): d=177.1
(CO), 177.0 (CO), 157.8, 157.0 (C), 149.8, 149.3 (C), 149.3, 149.0 (C),
141.2, 141.1 (C), 140.7, 140.5 (C), 139.7, 139.5 (C), 137.9, 137.8 (C), 135.8,
135.1 (C), 131.9, 131.7 (2CHarom), 130.5, 130.4 (2ꢅ2CHarom), 121.1, 120.9
(2CHarom), 120.9, 120.4 (2CHarom), 114.3, 113.8 (2CHarom), 82.6 (C, C5H4),
69.2 (2CH, C5H4), 68.6 (2CH, C5H4), 68.5 (5CH, Cp), 66.6, 66.3 (OCH2),
39.1 (C, tBu), 39.0 (C, tBu), 29.1, 29.0 (CH2), 27.1 (6CH3, tBu), 13.5 ppm
(CH3); IR (KBr): n˜ =1752 cmꢀ1 (CO); MS (EI): m/z: 698 [M] +, 199
[CpFeACHTUNGTRENNUNG
(h5-C5H4)CH2]+; elemental analysis calcd (%) for C43H46FeO5: C
73.92, H 6.63; found: C 73.58, H 6.66.
General procedure for the preparation of 2, 3, 4, and 2b: Sodium hydrox-
ide (0.80 g, 20 mmol) was added to a solution of 14, 15, 16, and 18
(2 mmol), respectively, dissolved in THF (30 mL) and water (40 mL).
The reaction mixture was heated under reflux for 24 h. The solution was
hydrolyzed, acidified to pH 1, and underwent the standard workup. The
residue was purified on an aluminum oxide column (eluent: dichlorome-
thane/acetone 95:5). The products were further purified by preparative
HPLC with a solution of acetonitrile and water or pure acetonitrile. The
isomers (distinctly separate on TLC plates, eluent: dichloromethane, in
an approximately 1:1 ratio) could be easily separated but re-isomerized
1-[4-(2-Ferrocenylethoxy)phenyl]-1,2-bis(4-trimethylacetoxyphenyl)but-1-
ene (14): The reaction was accomplished with ferrocenylethanol 10
(0.552 g, 2.4 mmol). Yield 85%; 1H NMR (300 MHz, CDCl3): d=7.20–
6.65 (m, 10H; CHarom), 6.66, 6.49 (d, J=8.8 Hz, 2H; CHarom), 4.20–3.90
(m, 9H; CpFeC5H4), 4.01, 3.89 (t, J=7.0 Hz, 2H; OCH2), 2.76, 2.66 (t,
J=7.0 Hz, 2H; CH2), 2.41, 2.38 (q, J=7.4 Hz, 2H; CH2), 1.29, 1.26, 1.25,
1.22 (s, 18H; 2tBu), 0.86, 0.84 ppm (t, J=7.4 Hz, 3H; CH3); 13C NMR
(75 MHz, CDCl3): d=177.0 (CO), 157.7, 156.9 (C), 149.7, 149.3 (C),
149.2, 149.0 (C), 141.2, 141.1 (C), 140.7, 140.5 (C), 139.6, 139.5 (C), 137.9,
692
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 684 – 696

First Derivatives of the Breast Cancer Drug Tamoxifen
FULL PAPER
rapidly, before the solvents could be removed. Recrystallization failed to
occur because the solutions became oily at low temperature, and at room
temperature the compounds degraded in a few days.
(2CHarom), 114.9, 114.3 (2CHarom), 114.8 (2CHarom), 114.0, 113.3
(2CHarom), 89.1 (C, C5H4), 68.7, 68.6 (5CH, Cp), 68.2, 68.1 (2CH, C5H4),
67.7, 67.5 (OCH2), 67.3, 67.2 (2CH, C5H4), 29.3, 29.1 (CH2), 29.2 (CH2),
28.9 (CH2), 27.5, 27.4 (CH2), 13.7 ppm (CH3); IR (KBr): n˜ =3422 cmꢀ1
(OH); HRMS (CI, NH3): m/z: calcd for C36H37FeO3 [M+H]+: 573.2093;
found: 573.2089.
1-[4-(2-Ferrocenylethoxy)phenyl]-1,2-bis(4-hydroxyphenyl)but-1-ene (2):
The reaction was accomplished with 14 (1.425 g, 2 mmol). The product
was purified by preparative HPLC with acetonitrile/water 90:10 as the
eluent. Compound 2 was retrieved as a yellow solid (92%). 1H NMR
(300 MHz, CDCl3): d=7.11, 7.05 (d, J=8.6 Hz, 2H; CHarom), 6.94 (d, J=
8.6 Hz, 2H; CHarom), 6.85, 6.76 (d, J=8.6 Hz, 2H; CHarom), 6.75, 6.70 (d,
J=8.6 Hz, 2H; CHarom), 6.60, 6.59 (d, J=8.6 Hz, 2H; CHarom), 6.53, 6.45
(d, J=8.6 Hz, 2H; CHarom), 4.25–4.02 (m, 9H; CpFeC5H4), 4.08, 3.93 (t,
J=7.0 Hz, 2H; OCH2), 2.80, 2.69 (t, J=7.0 Hz, 2H; CH2), 2.42 (q, J=
7.3 Hz, 2H; CH2), 0.90 ppm (t, J=7.3 Hz, 3H; CH3); 13C NMR (75 MHz,
CDCl3): d=158.8, 158.0 (C), 155.8, 154.9 (C), 155.3 (C), 141.9 (C), 138.7
(C), 138.0, 137.8 (C), 137.6, 137.4 (C), 136.1 (C), 133.5, 133.4 (2CHarom),
132.3 (2CHarom), 132.2, 132.0 (2CHarom), 116.4, 115.8 (2CHarom), 116.3
(2CHarom), 115.5, 114.8 (2CHarom), 86.3 (C, C5H4), 70.2 (5CH, Cp+2CH,
C5H4), 69.9, 69.7 (OCH2), 69.1, 69.0 (2CH, C5H4), 31.0, 30.9 (CH2), 30.3
(CH2), 15.1 ppm (CH3); IR (KBr): n˜ =3417 (OH), 2869, 2928, 2962 cmꢀ1
(CH2, CH3); HRMS (CI, CH4): m/z: calcd for C34H33FeO3: 545.1780
[M+H]+; found: 545.1786.
1-[4-(Ferrocenylmethoxy)phenyl]-1,2-bis[4-(t-butyl-dimethylsilyloxy)phe-
nyl]but-1-ene (17): A solution of DEAD (0.72 g, 2.7 mmol) in dry THF
(3 mL) was dropped at 08C into a solution of ferrocenylmethanol 9
(0.59 g, 2.75 mmol), 8a (1.1 g, 1.96 mmol) and PPh3 (0.72 g, 2.7 mmol) in
dry THF (12 mL). The reaction was stirred at room temperature for 96 h.
The solvent was evaporated under vacuum and the residue was purified
by alumina-gel column chromatography (petroleum ether) to give 17 as a
yellow solid (78%; isomer ratio: 1:1). This compound was recrystallized
from an ether/pentane solution. 1H NMR (300 MHz, CDCl3): d=7.23–
6.45 (m, 12H; CHarom), 4.83, 4.68 (s, 2H; OCH2), 4.22, 4.14 (t, J=1.8 Hz,
2H; C5H4), 4.08, 4.04 (t, J=1.8 Hz, 2H; C5H4), 4.07, 4.03 (s, 5H; Cp),
2.50, 2.49 (q, J=7.3 Hz, 2H; CH2), 0.89, 0.85 (s, 9H; tBuSi), 0.88, 0.87 (t,
J=7.3 Hz, 3H; CH3), 0.86, 0.82 (s, 9H; tBuSi), 0.26, 0.20 (s, 6H; SiMe2),
0.21, 0.15 ppm (s, 6H; SiMe2); 13C NMR (75 MHz, CDCl3): d=157.9,
157.1 (C), 154.5, 154.1 (C), 154.1, 153.7 (C), 140.9 (C), 137.8 (C), 137.3,
137.0 (C), 136.8, 136.4 (C), 136.0, 135.9 (C), 132.3 (2CHarom), 131.0
(2CHarom), 130.9 (2CHarom), 119.9, 119.8 (2CHarom), 119.8, 119.2
(2CHarom), 114.5, 113.8 (2CHarom), 83.0 (C, C5H4), 69.5 (2CH, C5H4), 68.9
(5CH, Cp), 68.8 (2CH, C5H4), 66.8, 66.6 (OCH2), 29.2, 29.1 (CH2), 26.0
(2tBu), 18.6 (C, tBuSi), 18.5 (C, tBuSi), 14.0 (CH3), ꢀ4.0 (SiMe2),
ꢀ4.1 ppm (SiMe2); IR (KBr): n˜ =3087, 3032, 2956, 2929, 2896, 2857
(CH2, CH3), 1254 cmꢀ1 (SiCH3); MS (CI, NH3): m/z: 759 [M+H]+, 776
1-[4-(2-Ferrocenylethoxy)phenyl]-1-phenyl-2-(4-hydroxyphenyl)but-1-ene
(2b): The reaction was accomplished with 18 (1.225 g, 2 mmol). The
product was purified by preparative HPLC with pure acetonitrile. Com-
pound 2b was retrieved as a yellow solid (79%). 1H NMR (300 MHz,
CDCl3): d=7.35–6.75 (m, 9H; CHarom), 6.75–6.42 (m, 4H; CHarom), 4.56,
4.54 (s, 1H; OH), 4.15–3.95 (m, 9H; C5H4FeCp), 4.04, 3.89 (t, J=7.0 Hz,
2H; OCH2), 2.76, 2.66 (t, J=7.0 Hz, 2H; CH2), 2.37, 2.36 (q, J=7.4 Hz,
2H; CH2), 0.87, 0.85 ppm (t, J=7.4 Hz, 3H; CH3); 13C NMR (75 MHz,
CDCl3): d=157.5, 156.7 (C), 153.7 (C), 144.0, 143.6 (C), 141.3, 140.7 (C),
138.0, 137.9 (C), 136.2, 135.7 (C), 134.8 (C), 131.9, 130.9 (2CHarom), 130.9,
130.8 (2CHarom), 130.6, 129.5 (2CHarom), 128.1, 127.3 (2CHarom), 126.4,
125.5 (CH), 114.9, 114.8 (2CHarom), 114.0, 113.4 (2CHarom), 84.8 (C,
C5H4), 68.6 (5CH, Cp+2CH, C5H4), 68.5, 68.3 (OCH2), 67.5, 67.4 (2CH,
C5H4), 29.6, 29.5 (CH2), 28.9 (CH2), 13.6 ppm (CH3); IR (KBr): n=
3432 cmꢀ1 (OH); HRMS (CI, CH4): m/z: calcd for C34H33FeO2: 529.1830
[M+H]+; found: 529.1829.
[M+NH4]+, 199 [CpFeCpCH2]+.
; elemental analysis calcd (%) for
C45H58FeO3Si2: C 71.21, H 7.70; found: C 70.87, H 7.56.
1-[4-(Ferrocenylmethoxy)phenyl]-1,2-bis(4-hydroxyphenyl)but-1-ene (1):
Compound 17 was dissolved in dry THF (30 mL) and a 1m solution of
tetrabutylammonium fluoride (2.8 mL, 2.8 mmol) was added. The solu-
tion was stirred for 25 min, hydrolyzed, and underwent the standard
workup. The residue was purified on semipreparative HPLC with aceto-
nitrile/water 80:20 as the eluent to give pure 1 (71%). The isomers
(isomer ratio: 1:1) were separated, but rapidly isomerized before evapo-
ration of acetonitrile under reduced pressure. The mixture was extracted
with dichloromethane and water, decanted, dried on MgSO4, and concen-
trated under reduced pressure. 1H NMR (300 MHz, CDCl3): d=7.15,
7.09 (d, J=8.6 Hz, 2H; CHarom), 7.03–6.42 (m, 10H; CHarom), 4.82, 4.68 (s,
2H; OCH2), 4.34, 4.26 (t, J=1.8 Hz, 2H; C5H4), 4.21, 4.17 (t, J=1.8 Hz,
2H; C5H4), 4.20, 4.15 (s, 5H; Cp), 2.46 (q, J=7.4 Hz, 2H; CH2), 0.94,
0.93 ppm (t, J=7.4 Hz, 3H; CH3); 13C NMR (75 MHz, CDCl3): d=157.5,
156.7 (C), 154.2, 153.7 (C), 153.7, 153.3 (C), 140.5 (C), 137.4 (C), 136.6
(C), 136.2, 136.1 (C), 134.9 (C), 132.2, 131.9 (2CHarom), 130.9 (2CHarom),
130.8, 130.6 (2CHarom), 115.0, 114.9 (2CHarom), 114.9, 114.3 (2CHarom),
114.3, 113.7 (2CHarom), 82.6 (C, C5H4), 69.2 (2ꢅ2CH, C5H4), 68.6 (5CH,
Cp), 66.6, 66.4 (OCH2), 28.9 (CH2), 13.7 ppm (CH3); IR (KBr): n˜ =
3414 cmꢀ1 (OH); HRMS (CI, CH4): m/z: calcd for C33H31FeO3 [M+H]+:
531.1623; found: 531.1625.
1-[4-(3-Ferrocenylpropoxy)phenyl]-1,2-bis(4-hydroxyphenyl)but-1-ene
(3): The reaction was accomplished with 15 (1.453 g, 2 mmol). The prod-
uct was purified by preparative HPLC with acetonitrile/water 90:10 as
1
the eluent. Compound 3 was retrieved as a yellow solid (70%). H NMR
(300 MHz, CDCl3): d=7.11, 7.06 (d, J=8.7 Hz, 2H; CHarom), 6.94 (d, J=
8.7 Hz, 4H; CHarom), 6.86, 6.76 (d, J=8.7 Hz, 2H; CHarom), 6.75, 6.71 (d,
J=8.7 Hz, 2H; CHarom), 6.60 (d, J=8.7 Hz, 2H; CHarom), 6.55, 6.45 (d,
J=8.7 Hz, 2H; CHarom), 4.14–4.01 (m, 9H; CpFeC5H4), 3.97, 3.85 (t, J=
6.3 Hz, 2H; OCH2), 2.55–2.36 (m, 4H; 2CH2), 2.06–1.75 (m, 2H; CH2),
0.91, 0.90 ppm (t, J=7.3 Hz, 3H; CH3); 13C NMR (75 MHz, CDCl3): d=
157.6, 156.7 (C), 154.2, 153.3 (C), 153.7 (C), 140.5 (C), 137.3 (C), 136.6,
136.4 (C), 136.2, 136.0 (C), 134.9 (C), 132.2, 132.0 (2CHarom), 130.9
(2CHarom), 130.8, 130.6 (2CHarom), 115.0, 114.3 (2CHarom), 114.8
(2CHarom), 114.1, 113.4 (2CHarom), 88.6, 88.4 (C, C5H4), 68.7, 68.6 (5CH,
Cp), 68.2 (2CH, C5H4), 67.4, 67.2 (OCH2), 67.3 (2CH, C5H4), 30.6, 30.4
(CH2), 28.9 (CH2), 26.0, 25.9 (CH2), 13.7 ppm (CH3); IR (KBr): n˜ =
3407 cmꢀ1 (OH); HRMS (CI, CH4): m/z: calcd for C35H35FeO5 [M+H]+:
559.1936; found: 559.1926.
Preparation of 2a and 19: A solution of DEAD (0.42 g, 2.4 mmol) in dry
THF (3 mL) was added dropwise to a 08C solution of ferrocenyl alcohol
10 (2.4 mmol), the known diphenol 8d (0.633 g, 2 mmol), and PPh3
(0.74 g, 2.8 mmol) in dry THF (12 mL). The reaction mixture was stirred
at room temperature for 48 h. The solvent was evaporated under vacuum
and the residue was purified on an aluminum oxide column with petrole-
um ether to give 2a (isomer ratio: 1:1) and 19 as yellow solids. Com-
pound 19 was recrystallized from petroleum ether and 2a was re-purified
on semipreparative HPLC with acetonitrile/water 90:10 as the eluent to
give pure 2a. The isomers were separated but remixed in the same flask
(because of rapid isomerization) before evaporation of the maximum of
acetonitrile under reduced pressure. The mixture was extracted with di-
chloromethane and water, decanted, dried on MgSO4, and concentrated
under reduced pressure.
1-[4-(4-Ferrocenylbutoxy)phenyl]-1,2-bis(4-hydroxyphenyl)but-1-ene (4):
The reaction was accomplished with 16 (1.481 g, 2 mmol). The product
was purified by preparative HPLC with acetonitrile/water 90:10 as the
eluent. Compound 4 was retrieved as a yellow solid (83%). 1H NMR
(300 MHz, CDCl3): d=7.13, 7.08 (d, J=8.7 Hz, 2H; CHarom), 6.96 (d, J=
8.7 Hz, 2H; CHarom), 6.87, 6.79 (d, J=8.7 Hz, 2H; CHarom), 6.77, 6.73 (d,
J=8.7 Hz, 2H; CHarom), 6.63, 6.61 (d, J=8.7 Hz, 2H; CHarom), 6.61, 6.60
(d, J=8.7 Hz, 2H; CHarom), 6.56, 6.48 (d, J=8.7 Hz, 2H; CHarom), 4.19–
4.03 (m, 9H; CpFeC5H4), 3.99, 3.85 (t, J=6.3 Hz, 2H; OCH2), 2.52–2.30
ꢀ
(m, 4H; 2CH2), 1.93–1.54 (m, 4H; CH2 CH2), 0.93 ppm (t, J=7.3 Hz,
3H; CH3); 13C NMR (75 MHz, CDCl3): d=157.7, 156.8 (C), 154.3, 153.7
(C), 153.7, 153.4 (C), 140.4 (C), 137.3 (C), 136.5, 136.3 (C), 136.2, 135.9
(C), 134.9 (C), 132.1, 131.9 (2CHarom), 130.9 (2CHarom), 130.8, 130.6
1-[4-(2-Ferrocenylethoxy)phenyl]-1-(4-hydroxyphenyl)-2-phenylbut-1-ene
(2a): Yield: 31%; 1H NMR (300 MHz, CDCl3): d=7.22–7.05 (m, 7H;
CHarom), 6.89, 6.81 (d, J=8.8 Hz, 2H; CHarom), 6.77, 6.73 (d, J=8.8 Hz,
Chem. Eur. J. 2009, 15, 684 – 696
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
693

A. Vessiꢁres et al.
2H; CHarom), 6.54, 6.47 (d, J=8.8 Hz, 2H; CHarom), 4.72, 4.49 (s, 1H;
OH), 4.25–4.00 (m, 9H; C5H4FeCp), 4.11, 3.95 (t, J=7.0 Hz, 2H; OCH2),
2.84, 2.73 (t, J=7.0 Hz, 2H; CH2), 2.49 (q, J=7.4 Hz, 2H; CH2),
0.93 ppm (t, J=7.4 Hz, 3H; CH3); 13C NMR (75 MHz, CDCl3): d=155.7,
154.8 (C), 152.4, 151.6 (C), 140.8 (C), 139.2 (C), 135.9 (C), 134.5, 134.4
(C), 134.1, 133.9 (C), 130.2, 130.1 (2CHarom), 128.9, 128.7 (2CHarom), 127.8
(2CHarom), 126.0 (2CHarom), 124.0 (CH), 113.1, 112.4 (2CHarom), 112.2,
111.4 (2CHarom), 82.9 (C, C5H4), 66.7 (5CH, Cp+2CH, C5H4), 66.6, 66.4
(OCH2), 65.6, 65.5 (2CH, C5H4), 27.8, 27.6 (CH2), 27.1 (CH2), 11.8 ppm
(CH3); IR (KBr): n˜ =3416, 3262 (OH), 2962, 2928, 2870 cmꢀ1 (CH2,CH3);
HRMS (EI, 70 eV): m/z: calcd for C34H32FeO2: 528.1752 [M]+; found:
528.1765; elemental analysis calcd (%) for C34H32FeO2: C 77.27, H 6.1;
found: C 77.02, H 6.12.
[M]+, 121 [FeCp]+, 57 [tBu]+; HRMS (EI, 70 eV): m/z: calcd for
C39H38O4Fe: 626.2120 [M]+; found: 626.2117.
1-[4-(2-Ferrocenyl-2-oxoethoxy)phenyl]-2-(4-trimethylacetoxyphenyl)-1-
phenylbut-1-ene (20b): Yield: 45%, isolated as a mixture of both isomers
(isomer ratio: 3:2); 1H NMR (300 MHz, CDCl3): d=7.36–6.66 (m, 13H;
ꢀ
CHarom), 4.97, 4.82 (s, 2H; O CH2), 4.94, 4.88 (t, J=1.9 Hz, 2H; C5H4),
4.55, 4.54 (t, J=1.9 Hz, 2H; C5H4), 4.23, 4.16 (s, 5H; Cp), 2.41 (q, J=
7.3 Hz, 2H; CH2), 1.34, 1.32 (s, 9H; CH3 of tBu), 0.93, 0.92 ppm (t, J=
7.4 Hz, 3H; CH3 of Et); 13C NMR (75 MHz, CDCl3): d=199.4 (CO),
177.0 (COO), 157.1 (C), 156.3 (C), 149.3 (C), 141.2 (C), 140.8 (C), 139.6
(C), 138.6, 138.5 (C), 132.0 (2CHarom), 130.8 (2CHarom), 130.7, 130.5
(2CHarom), 129.4, 128.1 (2CHarom), 127.4, 126.6 (CHarom), 121.0, 120.8
(2CHarom), 114.5, 113.9 (2CHarom), 77.2 (C, C5H4), 72.6, 72.5 (2CH, C5H4),
ꢀ
71.5 (O CH2), 70.0 (Cp), 69.4 (2CH, C5H4), 39.0 (C, tBu), 29.0 (CH2),
1,1-Bis[4-(2-ferrocenylethoxy)phenyl]-2-phenylbut-1-ene (19): Yield:
32%; 1H NMR (300 MHz, CDCl3): d=7.22–7.06 (m, 7H; CHarom), 6.88
(d, J=8.7 Hz, 2H; CHarom), 6.77 (d, J=8.7 Hz, 2H; CHarom), 6.54 (d, J=
8.7 Hz, 2H; CHarom), 4.18 (t, J=1.8 Hz, 2H; C5H4), 4.15 (s, 5H; Cp),
4.12–4.07 (m, 9H; Cp+C5H4), 4.11 (t, J=7.0 Hz, 2H; OCH2), 4.06 (t, J=
1.8 Hz, 2H; C5H4), 3.95 (t, J=7.0 Hz, 2H; OCH2), 2.84 (t, J=7.0 Hz,
2H; CH2), 2.73 (t, J=7.0 Hz, 2H; CH2), 2.49 (q, J=7.4 Hz, 2H; CH2),
0.93 ppm (t, J=7.4 Hz, 3H; CH3); 13C NMR (75 MHz, CDCl3): d=157.5
(C), 156.7 (C), 142.7 (C), 141.0 (C), 136.3 (C), 134.8 (C), 133.3 (C), 131.9
(2CHarom), 130.6 (2CHarom), 129.7 (2CHarom), 127.8 (2CHarom), 125.9
(CHarom), 114.0 (2CHarom), 113.3 (2CHarom), 84.7 (2C, C5H4), 68.5 (2ꢅ
5CH, Cp+2ꢅ2CH, C5H4), 68.4 (OCH2), 68.2 (OCH2), 67.5 (2CH,
C5H4), 67.4 (2CH, C5H4), 29.6 (CH2), 29.5 (CH2), 29.0 (CH2), 13.6 ppm
(CH3); IR (KBr): n˜ =2360, 2867, 2928, 2956, 3092 cmꢀ1 (CH2, CH3); MS
(EI, 70 eV) m/z: 740 [M]+, 741 [M+H]+, 199 [CpFeCpCH2]+, 121
[CpFe]+; elemental analysis calcd (%) for C46H44Fe2O2: C 74.60, H 5.98;
found: C 74.42, H 5.94.
27.1 (3CH3, tBu), 13.6 ppm (CH3, Et); MS (EI): m/z: 626 [M]+, 121
[FeCp]+, 57 [tBu]+; HRMS (EI, 70 eV): m/z: calcd for C39H38O4Fe:
626.2120 [M]+; found: 626.2120.
General procedure for the preparation of 5a and 5b: Esters 20a and 20b
(0.3 mmol) were dissolved in THF (5 mL), respectively. NaOH (220 mg,
excess) in water (5 mL) was added. The mixture was allowed to stir
under reflux for 6 h, after which time, it underwent the standard workup.
By flash chromatography, orange/red solids of 5a and 5b were isolated as
a mixture of Z and E isomers.
1-(2-Ferrocenyl-2-oxoethoxyphenyl)-1-(4-hydroxyphenyl)-2-phenylbut-1-
ene (5a): Yield: 75% (isomer ratio: 55:45); H NMR (300 MHz, CDCl3):
d=7.19–6.47 (m, 13H; CHarom), 4.95, 4.87 (t, J=1.9 Hz, 2H; C5H4), 4.98,
4.81 (s, 2H; O CH2), 4.60, 4.55 (t, J=1.9 Hz, 2H; C5H4), 4.23, 4.16 (s,
5H; Cp), 2.41, 2.39 (q, J=7.4 Hz, 2H; CH2), 0.90 ppm (t, J=7.4 Hz, 3H;
CH3); 13C NMR (75 MHz, CDCl3): d=199.9 (CO), 156.9, 156.1 (C),
154.7, 153.8 (C), 142.6 (C), 141.3, 141.2 (C), 137.7 (C), 137.3, 136.8 (C),
136.0, 135.6 (C), 132.1 (2CHarom), 130.7 (2CHarom), 129.7 (2CHarom),
127.9, 127.8 (2CHarom), 126.0 (CHarom), 115.1, 114.4 (2CHarom), 114.4,
1
ꢀ
General procedure for the preparation of 20, 20a, and 20b: In a Schlenk
tube, under inert atmosphere, KH (25–35% in oil, 0.03 mL, 1.2 mmol;
1.2 equiv) was dispersed in dry THF (10 mL). After the reaction mixture
had been stirred for 10 min, a solution of 8, 8d, or 8b (1 mmol), respec-
tively, in dry THF (10 mL) was added. The mixture was stirred under
reflux for 15 min and then a-chloroacetylferrocene (443 mg, 1.5 mmol) in
dry THF (10 mL) was added. The solution was heated under reflux over-
night. After hydrolysis and standard workup, orange solids of 20, 20a, or
20b were obtained.
ꢀ
113.7 (2CHarom), 75.8 (C, C5H4), 72.8, 72.7 (2CH, C5H4), 71.4, 71.3 (O
CH2), 70.2, 70.0 (5CH, Cp), 69.4 (2CH, C5H4), 29.1 (CH2), 13.6 ppm
(CH3); IR (KBr): n˜ =1665 (CO), 1608 (C=C), 1508 cmꢀ1 (C=C arom);
MS (ESI): m/z: 565 [M+Na]+, 541 [M+H]+; HRMS (EI, 70 eV): m/z:
calcd for C34H30O3Fe: 542.1545 [M]+; found: 542.1549.
1-(2-Ferrocenyl-2-oxoethoxyphenyl)-2-(4-hydroxyphenyl)-1-phenylbut-1-
ene (5b): Yield: 73%, isolated as a mixture of both isomers (isomer
ratio: 55:45).
(Z)-Isomer: 1H NMR (400 MHz, [D6]DMSO): d=9.25 (s, 1H; OH), 7.34
(d, J=7.5 Hz, 2H; meta-CH of a-C6H4), 7.26 (d, 2H; para-CH of a-
C6H4), 7.16 (d, J=7.2 Hz, 2H; ortho-CH of a-C6H4), 6.91 (d, J=8.3 Hz,
2H; b-C6H4), 6.74 (d, J=8.7 Hz, 2H; a’-C6H4), 6.66 (d, J=8.7 Hz, 2H;
1-[4-(2-Ferrocenyl-2-oxoethoxy)phenyl]-1,2-bis-(4-trimethylacetoxyphe-
nyl)but-1-ene (20): Yield: 300 mg, 41%, isolated as a mixture of both iso-
mers (isomer ratio: 2:1); 1H NMR (300 MHz, CDCl3): d=7.23–6.65 (m,
ꢀ
12H; CHarom), 4.87 (t, J=1.9 Hz, 2H; C5H4), 4.82 (s, 2H; O CH2), 4.54
(t, J=1.9 Hz, 2H; C5H4), 4.16 (s, 5H; Cp), 2.42 (q, J=7.3 Hz, 2H; CH2),
1.36, 1.33 (s, 9H; CH3 of tBu), 1.33, 1.29 (s, 9H; CH3 of tBu), 0.89 ppm
(t, J=7.3 Hz, 3H; CH3 of Et); 13C NMR (75 MHz, CDCl3): d=199.3
(CO), 177.0 (COO), 156.4 (C), 149.8 (C), 149.3 (C), 141.1 (C), 140.9 (C),
139.5 (C), 137.6 (C), 136.0 (C), 132.1, 131.7 (2CHarom), 130.5, 130.4 (2ꢅ
2CHarom), 121.1, 121.4 (2CHarom), 121.0, 120.0 (2CHarom), 114.5, 113.9
ꢀ
ꢀ
CH CO of a’-C6H4), 6.56 (d, J=8.3 Hz, 2H; CH COH of b-C6H4), 5.04
ꢀ
ꢀ
(s, 2H; O CH2 CO), 4.88 (t, J=1.9 Hz, 2H; C5H4), 4.61 (t, J=1.9 Hz,
2H; C5H4), 4.23 (s, 5H; Cp), 2.30 (q, J=7.4 Hz, 2H; CH2), 0.83 ppm (t,
J=7.4 Hz, 3H; CH3); the Z isomer was identified by 2D NMR spectros-
copy.
(E)-Isomer: 1H NMR (400 MHz, [D6]DMSO): d=7.33–6.61 (m, 13H;
CHarom), 4.94 (t, J=1.9 Hz, 2H; C5H4), 4.81 (s, 2H; O CH2 CO), 4.55 (t,
J=1.9 Hz, 2H; C5H4), 4.16 (s, 5H; Cp), 2.45 (q, J=7.4 Hz, 2H; CH2),
0.93 ppm (t, J=7.4 Hz, 3H; CH3).
ꢀ
(2CHarom), 77.2 (C, C5H4), 72.6 (2CH, C5H4), 72.4 (2CH, C5H4), 71.4 (O
CH2), 70.0, 69.3 (5CH, Cp), 39.1, 39.0 (2C, tBu), 29.0 (CH2), 27.1 (2ꢅ
ꢀ
ꢀ
ꢀ
3CH3, tBu), 13.6 ppm (CH3, Et); IR (KBr): n˜ =2972 (C H Ph), 1751
(CO), 1685 (FcCO), 1507 cmꢀ1 (C=C arom); MS (EI): m/z: 726 [M]+,
121 [CpFe]+, 57 [tBu]+; MS (CI, NH3): m/z: 744 [M+NH4]+, 727
[M+H]+; elemental analysis calcd (%) for C44H46O6Fe: C 72.66, H 6.33;
found: C 72.38, H 6.36.
AHCTUNGTRENNUNG
(Z+E)-Mixture: 13C NMR (100.61 MHz, CDCl3): d=200.5 (CO), 160.6
(C), 153.9 (C), 143.4 (C), 141.6 (C), 139.5 (C), 138.4 (C), 137.6 (C), 132.0,
130.9 (2CHarom), 130.8, 129.5 (2CH, C5H4), 128.1, 127.6 (2CH, C5H4),
127.4, 126.5 (2CH, C5H4), 125.7, 125.6 (CH, C5H4), 114.9, 114.8 (2CH,
C5H4), 114.5, 113.8 (2CH, C5H4), 77.2 (C, C5H4), 72.6 (2CH, C5H4), 71.5
1-[4-(2-Ferrocenyl-2-oxoethoxy)phenyl]-1-(4-trimethylacetoxyphenyl)-2-
phenylbut-1-ene (20a): Yield: 50%, isolated as a mixture of both isomers
(isomer ratio: 5:1); 1H NMR (300 MHz, CDCl3): d=7.25–6.61 (m, 13H;
CHarom), 4.86 (t, J=1.9 Hz, 2H; C5H4), 4.80 (s, 2H; O-CH2), 4.54 (t, J=
1.9 Hz, 2H; C5H4), 4.16 (s, 5H; Cp), 2.42 (q, J=7.4 Hz, 2H; CH2), 1.36
(s, 9H; CH3 of tBu), 0.88 ppm (t, J=7.4 Hz, 3H; CH3 of Et); 13C NMR
(75 MHz, CDCl3): d=199.3 (CO), 177.1 (COO), 156.3 (C), 149.7 (C),
142.2 (C), 142.0 (C), 141.0, 140.8 (C), 137.2 (C), 136.2, 136.1 (C), 132.0
(2CHarom), 130.4 (2CHarom), 129.6 (2CHarom), 127.9 (2CHarom), 126.1
(CHarom), 121.1 (2CHarom), 113.7 (2CHarom), 77.2 (C, C5H4), 72.5 (2CH,
ꢀ
(O CH2), 70.1 (5CH, Cp), 69.4 (2CH, C5H4), 28.9 (CH2), 13.6 ppm
(CH3); IR (CH2Cl2): n˜ =3595 (OH), 1686 (CO), 1606 (C=C), 1509 cmꢀ1
(C=C arom); MS (CI, NH3): m/z: 560 [M+NH4]+, 543 [M+H]+; elemen-
tal analysis calcd (%) for C34H30O3Fe: C 75.22, H 5.53; found: C 75.35, H
5.57.
Biochemical experiments
Materials: Stock solutions (1ꢅ10ꢀ3 m) of the ferrocenyl complexes to be
tested were prepared in DMSO and were kept at 48C in the dark; under
these conditions they are stable for at least two months. Serial dilutions
ꢀ
C5H4), 71.5 (O CH2), 70.0 (5CH, Cp), 69.4 (2CH, C5H4), 39.1 (C, tBu),
29.1 (CH2), 27.1 (3CH3, tBu), 13.5 ppm (CH3 Et); MS (EI): m/z: 626
694
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 684 – 696

First Derivatives of the Breast Cancer Drug Tamoxifen
FULL PAPER
in DMSO were prepared just prior to use. A stock solution (1ꢅ10ꢀ3 m) of
17b-E2 was prepared in ethanol. Dulbeccoꢀs modified eagle medium
(DMEM) was purchased from Gibco BRL, fetal calf serum from Dutsch-
er, Brumath (France), glutamine, E2, and protamine sulfate were from
Sigma. MCF-7 and MDA-MB-231 cells were from the Human Tumor
Cell Bank. Sheep uteri weighing approximately 7 g were obtained from
the slaughterhouse at Mantes–la–Jolie (France). They were immediately
frozen and kept in liquid nitrogen prior to use.
Molecular modeling: Theoretical docking experiments with the two iso-
mers of 2, 2a, 2b, 5, 5a, and 5b in the ER were performed by using the
ꢀ
LBD structure of ERa bound to OH Tam (Protein Data Bank code:
3ERT,)[7] and Mac Spartan Pro(Wavefunction Co., Irvine CA 92612,
USA). The affinity of the bioligand for the cavity was determined by
using MMFF molecular mechanics, with calculations for the bioligand–
ER cavity combination, and for the ER cavity and the bioligand per-
formed separately, each retaining the conformation previously deter-
mined for the molecular complex. This gives a value of the energy varia-
tion DE of the reaction: ligand+cavity!ligandꢀcavity complex.
Determination of the relative binding affinity (RBA) of the compounds
for ERa: RBA values were measured on ERa from lamb uterine cytosol
prepared in buffer A (0.05m Tris-HCL, 0.25 m sucrose, 0.1% b-mercapto-
Electrochemistry: Cyclic voltammograms (CVs) were obtained by using
a three-electrode cell with a 0.5 mm Pt working electrode, gold-plated
nickel mesh counter electrode, and saturated calomel reference electrode,
with an Autolab PGStat20 potentiostat driven by GPES software (Gen-
eral Purpose Electrochemical System, Version 4.8, EcoChemie B.V.,
Utrecht, the Netherlands) Solutions consisted of DMF (6 mL), analyte
(1 mm), and TBABF4 supporting electrolyte (0.1 m). Variable scan rate
CVs were obtained from 0.05 to 20 Vsꢀ1. Between each scan the working
electrode was gently polished with a sheet of “kimwipe light” (Kimberly–
Clark Co.).
ACHTUNGTRENNUNG
ethanol, pH 7.4 at 258C) as described previously.[30] Aliquots (200 mL) of
cytosol were incubated for 3 h at 08C with [6,7-3H]-E2 (2ꢅ10ꢀ m, specific
activity 1.62 TBqmmolꢀ1, NEN Life Science, Boston MA) in the pres-
ence of nine concentrations of the ferrocenyl complexes to be tested (be-
tween 6ꢅ10ꢀ7 and 6ꢅ10ꢀ9 m for the complexes with RBA values higher
than 5% and between 6ꢅ10ꢀ6 and 6ꢅ10ꢀ8 m for the compounds with
RBA values lower than 5%) or of 17b-E2 (between 8ꢅ10ꢀ8 and 7.5ꢅ
10ꢀ10 m). At the end of the incubation period, the fractions of [3H]-E2
bound to the estrogen receptors (Y values) were precipitated by addition
of a 200 mL of a cold solution of protamine sulfate (1 mgmLꢀ1 in water).
After a 10 min period of incubation at 48C, the precipitates were recov-
ered by filtration on 25 mm circle glass microfibre filters GF/C filters by
using a Millipore 12 well filtration ramp. The filters were rinsed twice
with cold phosphate buffer and then transferred in 20 mL plastic vials.
After addition of 5 mL of scintillation liquid (BCS Amersham) the radio-
activity of each fraction was counted in a Packard tri-carb 2100TR liquid
scintillation analyzer. The concentration of unlabeled steroid required to
displace 50% of the bound [3H]-E2 was calculated for 17b-E2 and for
Acknowledgements
We thank M.-N. Rager for 2D analysis, A. Cordaville for technical assis-
tance, and the Agence Nationale de la Recherche for financial support
(No. ANR-06-BLAN-0384-01, “FerVect”).
each complex by plotting the logit values of Y (logit Y=ln
G
versus the mass of the competing complex. The RBA (relative binding af-
finity) was calculated as follows: RBA of a compound=concentration of
E2 required to displace 50% of [3H]-E2 ꢅ100/concentration of the com-
pound required to displace 50% of [3H]-E2. The RBA value of E2 is by
definition equal to 100%.
[1] V. C. Jordan, Curr. Probl. Cancer 1992, 16, 129–176.
[2] V. C. Jordan, J. Med. Chem. 2003, 46, 883–908.
[3] J. I. MacGregor, V. C. Jordan, Pharmacol. Rev. 1998, 50, 151–196.
[4] V. C. Jordan, J. Med. Chem. 2003, 46, 1081–1108.
[5] J. S. Lewis, V. C. Jordan, Mutat. Res. 2005, 591, 247–263.
[6] A. M. Brzozowski, A. C. Pike, Z. Dauter, R. E. Hubbard, T. Bonn,
O. Engstrom, L. Ohman, G. L. Greene, J.-A. Gustafsson, M. Carl-
quist, Nature 1997, 389, 753–758.
[7] A. K. Shiau, D. Barstad, P. M. Loria, L. Cheng, P. J. Kushner, D. A.
Agard, G. L. Greene, Cell 1998, 95, 927–937.
[8] A. C. W. Pike, A. M. Brzozowski, J. Walton, R. E. Hubbard, A. G.
Thorsell, Y. L. Li, J.-A. Gustafsson, M. Carlquist, Structure 2001, 9,
145–153.
[9] M. J. Meegan, D. G. Lloyd, Curr. Med. Chem. 2003, 10, 181–210.
[10] R. A. Magarian, L. B. Overacre, S. Singh, K. L. Meyer, Curr. Med.
Chem. 1994, 1, 61–104.
[11] D. W. Robertson, J. A. Katzenellenbogen, J. R. Hayes, B. S. Katze-
nellenbogen, J. Med. Chem. 1982, 25, 167–171.
[12] A. B. Foster, R. McCague, A. Seago, G. Leclercq, S. Stoessel, F.
Roy, Anticancer Drug Des. 1986, 1, 245–257.
[13] M. Jarman, O. T. Leung, G. Leclercq, N. Devleeschouwer, S. Stoes-
sel, R. C. Coombes, R. A. Skilton, Anticancer Drug Des. 1986, 1,
259–268.
[14] V. Agouridas, I. Laios, A. Cleeren, E. Kizilian, E. Magnier, J.-C.
Blazejewski, G. Leclercq, Bioorg. Med. Chem. 2006, 14, 7531–7538.
[15] T. M. Willson, J. D. Norris, B. L. Wagner, I. Asplin, P. Baer, H. R.
Brown, S. A. Jones, B. Henke, H. Sauls, S. Wolfe, D. C. Morris , D. P.
McDonnell, Endocrinology 1997, 138, 3901.
[16] T. M. Willson, B. R. Henke, T. M. Momtahen, P. S. Charifson, K. W.
Batchelor, D. B. Lubahn, L. B. Moore, B. B. Oliver, H. R. Sauls, J. A.
Triantafillou, S. G. Wolfe, P. G. Baer, J. Med. Chem. 1994, 37, 1550–
1552.
[17] Y.-L. Wu, X. Yang, Z. Ren, D. P. McDonnell, J. D. Norris, T. M.
Willson, G. L. Greene, Mol. Cell 2005, 18, 413–424.
[18] K. S. Kraft, P. C. Ruenitz, M. G. Bartlett, J. Med. Chem. 1999, 42,
3126–3133.
[19] V. N. Rubin, P. C. Ruenitz, F. D. Boudinot, J. L. Boyd, Bioorg. Med.
Chem. 2001, 9, 1579–1587.
Measurement of the octanol/water partition coefficient (logPo/w) of the
compounds: The logPo/w values of the compounds were determined by
reverse-phase HPLC on a C-8 column (Nucleosil 5.C8, from Macherey
Nagel, France) according to a previously described method.[46,47] Mea-
surement of the chromatographic capacity factors (kN) for each com-
pound was done at various concentrations in the range of 85–60% meth-
anol (containing 0.25% octanol) and an aqueous phase consisting of
0.15% n-decylamine in 0.02m MOPS (3-morpholinopropanesulfonic
acid) buffer pH 7.4 (prepared in 1-octanol/saturated water). These ca-
pacity factors (kN) are extrapolated to 100% of the aqueous component
given the value of k0w ꢅlogPo/w (y) is then obtained by the formula: y=
0.13418+0.98452ꢅlogkw.
Culture conditions: Cells were maintained in monolayer culture in
DMEM with phenol red/Glutamax I, supplemented with 9% of decom-
plemented fetal calf serum and 0.9% kanamycine, at 378C in a 5% CO2
air humidified incubator. For proliferation assays, cells were plated in 24-
well sterile plates at a density of 1.1ꢅ104 cells for PC-3 or MDA-MB-231
and of 3ꢅ104 cells for MCF-7 in 1 mL of DMEM without phenol red,
supplemented with 9% of fetal calf serum desteroided on dextran char-
coal, 0.9% Glutamax I, and 0.9% kanamycine, and were incubated for
24 h. The following day (D0), 1 mL of the same medium containing the
compounds to be tested diluted in DMSO was added to the plates (final
volumes of DMSO: 0.1% ; 4 wells for each conditions). After 3 days
(D3), the incubation medium was removed and 2 mL of fresh medium
containing the compounds was added. At different days (D3, D4, D5,
and D6), the protein content of each well was quantified by methylene
blue staining as follows. Cell monolayers were fixed and stained for 1 h
in methanol with methylene blue (2.5 mgmLꢀ1), and then washed thor-
oughly with water. Two milliliters of HCl (0.1m) was then added, and the
plate was incubated for 1 h at 378C. Then the absorbance of each well
was measured at 655 nm with a Biorad spectrophotometer (microplate
reader). The results are expressed as the percentage of proteins versus
the control. Experiments were performed at least in duplicate.
Chem. Eur. J. 2009, 15, 684 – 696
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
695

A. Vessiꢁres et al.
[20] S. Top, A. Vessiꢁres, G. Leclercq, J. Quivy, J. Tang, J. Vaissermann,
M. Huchꢂ, G. Jaouen, Chem. Eur. J. 2003, 9, 5223–5236.
[21] S. Top, A. Vessiꢁres, C. Cabestaing, I. Laios, G. Leclercq, C. Provot,
G. Jaouen, J. Organomet. Chem. 2001, 637, 500–506.
[22] G. Jaouen, S. Top, A. Vessiꢁres, G. Leclercq, J. Quivy, L. Jin, A.
Croisy, C. R. Acad. Sci. Ser. IIc 2000, 89–93.
[23] S. Top, B. Dauer, J. Vaissermann, G. Jaouen, J. Organomet. Chem.
1997, 541, 355–361.
[24] S. Top, J. Tang, A. Vessiꢁres, D. Carrez, C. Provot, G. Jaouen, Chem.
Commun. 1996, 955–956.
[25] A. Vessiꢁres, S. Top, P. Pigeon, E. A. Hillard, L. Boubeker, D.
Spera, G. Jaouen, J. Med. Chem. 2005, 48, 3937–3940.
[26] A. Nguyen, A. Vessiꢁres, E. A. Hillard, S. Top, P. Pigeon, G. Jaouen,
Chimia 2007, 61, 716–724.
[27] P. Kçpf-Maier, H. Kçpf, E. W. Neuse, Angew. Chem. 1984, 96, 446–
447; Angew. Chem. Int. Ed. Engl. 1984, 23, 456–457.
[28] P. Kçpf-Maiyer, H. Kçpf, E. W. Neuse, J. Cancer Res. Clin. Oncol.
1984, 108, 336–340.
[29] E. A. Hillard, P. Pigeon, A. Vessiꢁres, C. Amatore, G. Jaouen,
Dalton Trans. 2007, 5073–5081.
[30] G. Tabbi, C. Cassino, G. Cavigiolio, D. Colangelo, A. Ghiglia, I.
Viano, D. Osella, J. Med. Chem. 2002, 45, 5786–5796.
[31] D. Osella, M. Ferrali, P. Zanello, F. Laschi, M. Fontani, C. Nervi, G.
Cavigiolio, Inorg. Chim. Acta 2000, 306, 42–48.
[35] A. Vessiꢁres, S. Top, W. Beck, E. A. Hillard, G. Jaouen, Dalton
Trans. 2006, 4, 529–541.
[36] E. A. Hillard, A. Vessiꢁres, S. Top, P. Pigeon, K. Kowalski, M.
Huchꢂ, G. Jaouen, J. Organomet. Chem. 2007, 692, 1315–1326.
[37] A. Nguyen, V. Marsaud, C. Bouclier, S. Top, A. Vessiꢁres, P. Pigeon,
R. Gref, P. Legrand, G. Jaouen, J.-M. Renoir, Int. J. Pharm. 2008,
347, 128–135.
[38] A. Nguyen, S. Top, A. Vessiꢁres, P. Pigeon, M. Huchꢂ, E. A. Hillard,
G. Jaouen, J. Organomet. Chem. 2007, 692, 1219–1225.
[39] S. Masi, S. Top, L. Boubeker, G. Jaouen, S. Mundwiler, B. Spingler,
R. Alberto, Eur. J. Inorg. Chem. 2004, 2013–2017.
[40] S. Top, S. Masi, G. Jaouen, Eur. J. Inorg. Chem. 2002, 1848–1853.
[41] K. W. Nettles, J. B. Bruning, G. Gil, E. E. OꢀNeill, J. Nowak, A.
Hughs, Y. Kim, E. R. DeSombre, R. Dilis, R. N. Hanson, A. Joachi-
miak, G. L. Green, EMBO Rep. 2007, 8, 563–568.
[42] S. Gauthier, J. Mailhot, F. Labrie, J. Org. Chem. 1996, 61, 3890–
3893.
[43] D. G. Lloyd, H. M. Smith, T. OꢀSullivan, D. M. Zisterer, M. J.
Meegan, Med. Chem. 2005, 1, 335–353.
[44] W. L. Davis, R. F. Shago, E. H. G. Langner, J. C. Swarts, Polyhedron
2005, 24, 1611–1616.
[45] S. G. A. Moinuddin, S. Hishiyama, M.-H. Cho, L. B. Davin, N. G.
Lewis, Org. Biomol. Chem. 2003, 1, 2307–2313.
[46] D. J. Minick, J. H. Frenz, M. A. Patrick, D. A. Brent, J. Med. Chem.
1988, 31, 1923–1933.
[32] A. M. Joy, D. M. L. Goodgame, I. J. Stratford, Int. J. Radiation On-
cology Biol. Phys. 1989, 16, 1053–1056.
[33] H. Tamura, M. Miwa, Chem. Lett. 1997, 1177–1178.
[34] E. A. Hillard, A. Vessiꢁres, L. Thouin, G. Jaouen, C. Amatore,
Angew. Chem. 2006, 118, 291–296; Angew. Chem. Int. Ed. 2006, 45,
285–290.
[47] M. G. Pomper, H. VanBrocklin, A. M. Thieme, R. D. Thomas, D. O.
Kiesewetter, K. E. Carlson, C. J. Mathias, M. J. Welch, J. A. Katze-
nellenbogen, J. Med. Chem. 1990, 33, 3143–3155.
Received: June 7, 2008
Published online: December 3, 2008
696
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 684 – 696