The first titanocenyl dichloride moiety vectorised by a selective estrogen receptor modulator (SERM). Synthesis and preliminary biochemical behaviour

Article abstract of DOI:10.1016/S0022-328X(01)01271-2

The synthesis of 4′, a titanocene derivative of the anticancer drug tamoxifen is presented together with its biochemical properties. Compound 4′ is prepared by a McMurry coupling followed by a ligand exchange. Compound 4′ reveals an unexpected proliferative effect on the hormone-dependent cell line MCF7. Surprisingly this effect is also observed with Cp₂TiCl₂ alone.

Full text of DOI:10.1016/S0022-328X(01)01271-2

Journal of Organometallic Chemistry 643–644 (2002) 350–356
www.elsevier.com/locate/jorganchem
The first titanocenyl dichloride moiety vectorised by a selective
estrogen receptor modulator (SERM). Synthesis and preliminary
biochemical behaviour
S. Top a, E.B. Kaloun a, A. Vessie`res a, I. La¨ıos b, G. Leclercq b, G. Jaouen a,*
a Laboratoire de Chimie Organome´tallique, Ecole Nationale Supe´rieure de Chimie de Paris, UMR CNRS. 7576, 11, rue Pierre et Marie Curie,
75231 Paris Cedex 05, France
b Laboratoire de Cance´rologie Mammaire, Institut Jules Bordet, Rue He´ger Bordet 1, Brussels 1000, Belgium
Received 17 July 2001; accepted 18 September 2001
Abstract
The synthesis of 4%, a titanocene derivative of the anticancer drug tamoxifen is presented together with its biochemical
properties. Compound 4% is prepared by a McMurry coupling followed by a ligand exchange. Compound 4% reveals an unexpected
proliferative effect on the hormone-dependent cell line MCF7. Surprisingly this effect is also observed with Cp2TiCl2 alone.
© 2002 Elsevier Science B.V. All rights reserved.
Keywords: Bioorganometallic; Titanocen; Tamoxifen; Estrogen; SERM; MCF7; Proliferative effect
1. Introduction
Inorganic platinum complexes, particularly cisplatin
1, have become established as highly effective antitu-
moral agents, despite the fact that to date their range of
use remains fairly restricted [1]. They have in fact
revolutionised the treatment of testicular cancer, which
is still one of their most spectacularly successful appli-
cations. This breakthrough has inspired much research
into the application of metal complexes in oncology
[2,3]. In this context, various series of cyclopentadienyl-
type organometallic complexes of Ti, Fe, Mo, V, Re,
have been evaluated. Titanium in particular is showing
promising results, although all stages of clinical trials
have not yet been completed [4]. It appears so far
however, that these metallocenes act via mechanisms
different from those of cisplatin, and thus may lend
themselves to treatment of a wider range of cancers [5].
The species mentioned above are most commonly
used in their simplest form. We however, have under-
taken a program of research focused on attaching them
to a well-defined biological vector in an attempt to
improve their selectivity and their effectiveness on
target cells [6–8]. Our work has led, for example, to
OH-Ferrocifens such as 2 capable of producing an-
tiproliferative effects on breast cancer cells whether they
are estrogen-dependent ER(+) or -independent ER(−)
(ER=estrogen receptor). Resistance to treatment with
tamoxifen 3, encountered in a number of breast can-
cers, remains a major therapeutic and social problem.
This has led to intense activity in research on SERMs
(selective estrogen receptor modulators) that would be
* Corresponding author. Fax: +33-1-43260061.
E-mail address: jaouen@ext.jussieu.fr (G. Jaouen).
0022-328X/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved.
PII: S0022-328X(01)01271-2

S. Top et al. / Journal of Organometallic Chemistry 643–644 (2002) 350–356
351
Scheme 1.
more effective than tamoxifen, the primary example of
this group [9,10]. Favourable indications concerning the
antitumoral potential of dichloro titanocene suggested
it as a good candidate for attachment to a tamoxifen-
type skeleton, allowing us to evaluate the capability of
a complex such as 4 to inhibit cellular proliferation in
breast cancers. The difficulties we encountered in syn-
thesising molecule 4 were surprising, and led us to
develop a novel preparative approach. Even more sur-
prising was our observation of an unanticipated oestro-
genic effect when compound 4 was incubated in
presence of MCF7 (ER+) cancer cells. The results of
this entire body of work are presented here.
tions and had completely disappeared by the end of the
reaction [15]. Pauson et al. have reported the formation
of (C5H4NMe2)Li by a lithium-induced decomposition
of dimethylaminoferrocene [16]. However, application
of this method to OH-Ferrocifen 2 did not lead to the
lithium salt of type 6, the precursor to the complex 4.
On the other hand, Rosenblum et al. were able to
obtain dicyclopentadienylnaphthalene by reacting cy-
clopentadienyl copper dimethyl sulphide on diiodon-
aphthalene [17]. In our case, the reaction of
cyclopentadienyl copper dimethyl sulfide on a vinylic
bromide, an analogue of 7, also failed to result in 5.
Having tried all these known reactions without suc-
cess, we were forced to consider a novel synthetic route
to compound 5. The discovery of this new route is
based on the observation that cymantrene derivatives
decompose slowly through the action of sunlight. By
using irradiation from a UV lamp to accelerate decom-
plexation, and by working in a protic medium, we were
able to establish that the cyclopentadienyl ring was
generated in very good yield (Scheme 3) [18].
2. Results and discussion
2.1. Synthesis
One potential route to 4 is its preparation starting
from 1-phenyl-1-(4-dimethyl aminopropyloxyphenyl)-2-
cymantrenyl-1-butene (5) (Scheme 1). Metallation of 5
by a lithium or thallium reagent would then give the
metal salt 6 which would subsequently be allowed to
react with CpTiCl3 [11]. Efforts were therefore made to
find synthetic routes to the substituted cyclopentadiene
5.
Application of this method should give access to 5
following the synthetic route shown in Scheme 4.
Propionyl cyclopentadienyl manganesetricarbonyl
(15) and the ketone (16) were allowed to react in the
presence of McMurry reagent. After heating for 1 h,
A literature search shows that several possible syn-
thetic routes are possible. The most direct one appears
to be the reaction involving formation of a CꢀC bond
catalysed by palladium. This coupling between a stan-
nylated derivative and a halogenide catalysed by palla-
dium, known as a Stille reaction, is very widely used
[12]. Johnson et al. have shown the viability of coupling
between iodocyclopentenone and a stannylated deriva-
tive by using Pd(Bz)2Cl2–CuI–AsPh3 as the catalyst
[13]. In principle the use of this method in the coupling
between the tin derivative 7 and iodocyclopentenone 8
should give the ketone 9 which is the precursor of 5
(Scheme 2). In practice, the experiment did not in fact
lead to the ketone 9. This negative result may be
attributed to the presence of the ethyl on 7 which could
cause a b-elimination reaction.
Scheme 2.
A second alternative would be a McMurry coupling
reaction between the ketone 10 and the ketone 11 [14].
This reaction did not result in the desired product 12
since the ketone 10 did not tolerate the reaction condi-
Scheme 3.

352
S. Top et al. / Journal of Organometallic Chemistry 643–644 (2002) 350–356
Scheme 4.
the complex 17 was obtained in 54% yield. It appeared
as a yellow oil containing a mixture of the cis and trans
isomers in 50/50 proportion, as estimated from the
NMR spectrum. Next a decomplexation step was per-
formed. UV irradiation of the manganese complex, 17,
in an ethyl ether–methanol mixture gave the cyclopen-
tadienyl derivative, 5, in excellent yield (98%). This
compound, relatively unstable because of a possible
Diels–Alder reaction, was characterised only by NMR.
Upon preparation, it was immediately converted into a
lithium salt by action of n-BuLi in THF at −70 °C.
Addition of CpTiCl3 to the organolithium caused the
immediate formation of the titanium complex 4 which
was then transformed into a chlorohydrate salt 4% by
passing over it a flow of HCl gas. A yield of 57% was
obtained.
fore causes an increase in lipophilicity to a level notice-
ably higher than that observed for the addition of a
ferrocene [7] or a CpRe(CO)3 moiety [8].
4. Study of the proliferative/antiproliferative effect on
hormone-dependent cell lines (MCF7 and MVLN)
The effect of 4% on hormone-dependent MCF7 cells,
derived from a breast cancer line containing the estro-
gen receptor ER+, observed after 5 days of incuba-
tion, is shown in Fig. 1. The results are compared to
those found in the presence of estradiol (E2) which is
the standard for the estrogenic effect, and OH-tamox-
ifen (3%), the standard for the antiestrogenic effect. The
results obtained are expressed as a percentage of DNA
compared to control (cells incubated in the absence of
added compound). The result obtained is highly sur-
3. Biochemical data
Table 1
Relative binding affinities (RBA) and partition coefficients (Log Po/
w) of the compounds
3.1. Determination of RBA 6alues and lipophilicity
The relative binding affinity (RBA) value for 4% is
8.5% for estrogen receptor alpha (Table 1), a much
higher value than expected for a compound with no
OH group and more precisely, no phenol ring, which
both OH-tamoxifen (3%), the active metabolite and 2
(OH-Ferrocifen) do possess. It is in fact well known
that this functionality plays a critical role in the binding
of the hormone to its receptor. For Cp2TiCl2, an RBA
value of 0 is found. Lipophilicity of 4% measured by its
octanol–water partition coefficient was found to be
5.83 for the Z isomer and 5.95 for the E isomer. The
addition of the organometallic entity Cp2TiCl2 there-
Compound
RBA (%) (DMSO,
Log Po/w b
25 °C, ERa) a
4%
8.5
0
5.83 (Z), 5.95 (E)
18 (Cp2TiCl2)
E2
3% (OH-Tamoxifen)
2 (OH-Ferrocifen)
100
3.5
38.5 c
11.5 c
3.2 (Z) c; 3.5 (E) c
4.3 (Z) c; 4.5 (E) c
a RBA for the estrogen receptor is determined on lamb uterus
cytosol as described in Ref. [24].
b Octanol–water partition coefficients were determined by reverse
phase HPLC using the method described by Pomper [25].
c Value from Ref. [7].

S. Top et al. / Journal of Organometallic Chemistry 643–644 (2002) 350–356
353
In particular, a hydrolysis of the cyclopentadienyl lead-
ing to a diene was observed, but also a hydrolysis of the
chlorides. Among the species identified were, for exam-
ple, Cp2Ti (H2O)Cl+, Cp2Ti(OH)Cl, (Cp2TiCl2)2O etc.
leading ultimately to the precipitation of polymers
derived from the various hydrolysis products [19,20]. It
is not impossible that one of the products thus formed
may exhibit estrogenic behaviour, considering that the
estrogen activation site is very receptive to phenol
derivatives.
However, an excellent recent study by Sadler [5] has
shed new light on the behaviour of Cp2TiCl2 in biolog-
ical media. According to this work, the most rapid
hydrolysis occurs at the level of the chlorides. The role
of the cyclopentadienes is to slow a too-rapid hydroly-
sis which would lead to inactive polymers, but the Cp
entities are subsequently displaced during the process.
Whether or not this is the case, the result is a monomer
of TiIV, a hard metal similar to FeIII, which is quickly
captured and stabilised by transferrine, one of the
iron-carrying proteins in the blood. It is interesting to
note that whatever the initial titanium complex,
whether Cp2TiCl2 or Budotitane, the result is the cap-
ture of the TiIV ion, preferentially in the case of trans-
ferrine where it replaces FeIII and binds to aspartate,
histidine, tyrosine and carbonate residues. It has been
shown that the level of transferrine receptor is greater
at the surface of cancerous cells, perhaps due to their
increased demand for iron [5]. Because of the change in
pH at this level the TiIV is then released and is trans-
ported to the centre of the cell by binding to ATP. This
work attempts to elucidate the nature of the active
species, TiIV, no matter what the precursor may be, or
how it is delivered to tumour cells. It is clear from the
above that depending on the specific nature of the
target cells, differentiated behaviour may be expected.
And in fact, although titanium has not yet been studied
in this context, the role of certain metal ions has
already been the subject of occasional analyses in rela-
tion to estrogen receptors [21,22]. It is known that
estrogen receptor alpha, which makes up by far the
largest proportion of the ER present in MCF7 cells,
contains zinc in its natural state. It is these zinc fingers
that mediate the association of the DNA binding do-
main to an estrogen response element (ERE) [23]. Re-
Fig. 1. Effect of E2 (Estradiol), 3% (OH-Tamoxifen), 4% and 18
(Cp2TiCl2) on the proliferation of MCF7 cells (estrogen receptor-pos-
itive cells). The results are expressed as the percentage of DNA in the
sample versus the DNA value of the control. The values are the
means of two independent experiments in quadruplicate9limits of
confidence (P=0.1, t=1.415) The values in brackets correspond to
the log of molarity of incubation.
prising. For while E2 and OH-tamoxifen give the ex-
pected effects (proliferative for the former, antiprolifer-
ative for the latter), 4% acts as an estrogen almost as
active as estradiol itself, at molarities as low as 1×
10−7 M. This powerful estrogenic effect is strong
enough to reverse the antiestrogenic effect produced by
the dimethylamino side chain, an effect that has already
been observed for tamoxifen derivatives bearing
organometallic entities, for example ferrocene [7] or
CpRe(CO)3 [8]. This unusual behaviour led us to exam-
ine the effect of the Cp2TiCl2 entity alone. We observed
a proliferative effect greater than that of 4%. It seems
therefore that this effect must essentially be due to the
entity itself or to one of its hydrolysates. These results
were confirmed on MVLN cells, another hormone de-
pendent cell line, in which expression of luciferase is
proportional to the estrogenicity of the tested com-
pounds. After 24 h of incubation, 4% and Cp2TiCl2
showed an estrogenic effect at 1 mM (148, 334%, respec-
tively with E2 0.1 nM at 261%). Within this short
period of incubation the effect observed with 4% is less
pronounced. This is probably due to competition be-
tween the antiproliferative effect of the organometallic
unit and the tamoxifen moiety.
placing zinc by either nickel (Ni2+) or copper (Cu2+
)
5. Discussion
inhibits this association while Co+ or Cd+ have no
effect. On the other hand, treatment of the receptor
with Cd+ in very low concentration (1 nM) results in
coordination of this metal at the level of the ligand
binding domain. The complex formed with this exoge-
nous metal activates ER alpha just as binding to estra-
diol would. It must be concluded that the effects of
Cd2+ and Ti4+ are similar, in that both lead to prolif-
eration of MCF7 type cancer cells.
In order to try to understand this surprising result,
namely that both 4% and Cp2TiCl2 demonstrate strong
estrogenic activity, it is necessary first of all to establish
the behaviour of these complexes in that most universal
of biological solvents, water. In fact, the hydrolysis of
Cp2TiCl2 has been the subject of a number of studies,
and appears to involve a complex series of events that
depends largely on the conditions that come into play.

354
S. Top et al. / Journal of Organometallic Chemistry 643–644 (2002) 350–356
A first step in elucidating the role of Cd2+ was
g, 14.70 mmol) and propionyl chloride (1.35 g, 14.70
mmol) were dissolved in CS2 (30 ml). AlCl3 (0.977 g,
7.35 mmol) was added in small portions to the solution
of cymantrene, and the solution became red. After 1.5
h, the CS2 was removed under vacuum and the oil
obtained was hydrolysed with water at 0 °C. After
extraction with Et2O (3×20 ml), and drying over
MgSO4, the solution was concentrated, whereupon ad-
dition of pentane yielded (propionylcyclopentadi-
enyl)tricarbonylmanganese (3.540 g, 92%) as a yellow
powder. Recrystallization from ether/pentane furnished
yellow crystals, m.p. 250 °C. 1H-NMR (200 MHz,
CDCl3): l 5.44 (t, 2H, J=2.1 Hz, h5-C5H4), 4.85 (t, 2H,
J=2.1 Hz, h5-C5H4), 2.64 (q, 2H, CH2), 1.17 (t, 3H,
CH3). 13C-NMR (50 MHz, CDCl3): l 197.9 (CO), 91.5
(C1 of h5-C5H4), 86.4 (2C, C5H4), 83.4 (2C, C5H4), 32.2
(CH2), 7.9 (CH3). IR (CH2Cl2): wCO at 2030 (s), 1948 (s),
1686 cm−1 (m). Anal. Calc. for C11H9MnO4: C, 50.79;
H, 3.49. Found: C, 50.83; H, 3.41%.
reported very recently [22]. Cadmium is able to coordi-
nate with cysteins 381 and 447, glutamic acid 523,
histidine 524 and aspartic acid 538 of the receptor ligand
binding domain. In so doing it inhibits the binding of
estradiol. It has been suggested that the hormone bind-
ing domain acts somewhat like a mousetrap which is
closed by creation of a saline bridge between the two
helices H4 and H12 of the receptor. This bridge, after
estradiol capture, positions helix 12 such that the tran-
scriptional machinery can be activated. Cysteins 381
and 447 are localised on helices H4 and H8, glutamic
acid 523 and histidine 524 are on helix H11 close to the
bioligand, and aspartic acid 538 is localised at the
interface of helix H12. The idea that this type of
coordination may cause the receptor to act in the same
way as the natural ligand, by snapping the trap shut, is
an appealing one that will be confirmed or denied by
further studies. It also remains to be seen whether the
explanation advanced for Cd2+ holds good for Ti4+
.
It should in any case be borne in mind that the
estrogen receptor is a critical mediator in the body, and
that species that can bind to it and activate it have the
potential to increase the risk of breast cancer. The
preliminary data presented here suggest that titanium
(TiIV) in the environment could have estrogenic activity
and thus be a new risk factor. In addition, if these
results were to be confirmed, the future of Cp2TiCl2
currently in phase II clinical trials, would have to take
into account this previously-unknown parameter.
6.2.2. 4-Dimethylaminopropyloxy benzophenone (16)
In a Schlenk tube purged with Ar, 4-hydroxyben-
zophenone (3.96 g, 0.02 mol) was dissolved in
CH3COCH3 (80 ml). Powdered NaOH (1.60 g, 0.04
mol) was added into the solution, and the mixture
heated at reflux for 15 min. After that time, dimethy-
laminopropionylchloride hydrochloride (3.16 g, 0.02
mol) was added into the mixture in one portion. The
heating was maintained for 15 h. After cooling to room
temperature (r.t.), the solution was filtrated and evapo-
rated. The crude product obtained was chro-
matographed on a silica gel column. The column was
first eluted with CH3COCH3 to remove the unreacted
hydroxybenzophenone, and then with CH3COCH3–
Et3N (10:1). Dimethylaminopropyloxy benzophenone
was isolated as an colourless oil which became a solid
6. Experimental
6.1. General data
1
Anhydrous THF and Et2O were distilled from
sodium–benzophenone. Methanol (Prolabo) was used
without any further purification. Thin layer chromatog-
raphy was performed on Silica Gel 60 GF254. UV-pho-
tolyses were carried out with a Heraeus TQ150, 150 W
high pressure Hg lamp. FT-IR spectra were recorded on
upon addition of pentane (4.40 g, 77%). H-NMR (200
MHz, CDCl3): l 7.84 and 6.96 (d, d, 2H, 2H, J=8.9
Hz, C6H4), 7.75 and 7.50 (m, m, 2H, 3H, C6H5), 4.11 (t,
2H, J=6.4 Hz, OCH2), 2.49 (t, 2H, J=7.4 Hz, NCH2),
2.28 (s, 6H, NMe2), 2.00 (m, 2H, CH2). MS (EI, 70 eV)
m/z, 283 [M+]. Anal. Calc. for C18H21NO2: C, 76.30; H,
7.47; N, 4.94. Found: C, 76.30; H, 7.58; N, 4.92%.
1
a BOMEM Michelson-100 spectrometer. H- and 13C-
NMR spectra were acquired on Bruker 200, 250 or 400
MHz spectrometers. Mass spectrometry was performed
by laboratoire de spectrome´trie de masse (ENSCP) and
by CRMPO (Universite´ de Rennes 1). Melting points
were measured with a Kofler device. Elemental analyses
were performed by the Regional Microanalysis Depart-
ment of Universite´ Pierre et Marie Curie.
6.2.3. 1-Phenyl,-1-(4-dimethylaminopropyloxyphenyl)-
2-cymantrenyl-1-butene (17)
TiCl4 (1.71 g, 9 mmol) was added dropwise to a
suspension of Zn powder (1.17 g, 12 mmol) in THF (30
ml) at 0 °C. The blue mixture obtained was heated at
reflux for 2 h, the solution became black, and the oil
bath was removed. A second solution was prepared by
dissolving 4-dimethylaminopropyloxy benzophenone,
16, (0.849 g, 3 mmol) and propionylcymantrene, 15,
(0.780 g, 3 mmol) in THF (15 ml). The latter solution
was added dropwise to the first solution, and the result-
ing mixture was again heated for 1 h. After cooling to
6.2. Synthesis of the compounds
6.2.1. Propionylcyclopentadienyltricarbonylmanganese
(15)
In a Schlenk tube purged with Ar, cymantrene (3.00

S. Top et al. / Journal of Organometallic Chemistry 643–644 (2002) 350–356
355
r.t., the mixture was hydrolysed with 100 ml of a 10%
7.1. Determination of the relati6e binding affinity
(RBA) of the compounds for the estrogen receptor
alpha (ERh)
Na2CO3 solution. After CHCl3 extraction and solvent
removal, the crude product was chromatographed on
silica gel plates with THF–Et3N 10:1 as eluent to give
17 (Z+E) as an orange oil (0.828 g, 54% yield).
1H-NMR (200 MHz, CDCl3): l 7.32–6.78 (m, 9H,
aromatic rings), 4.50 (m, 4H, C5H4), 3.98 (t, 2H,
OCH2), 2.45 (t, 2H, CH2ꢀN), 2.30 (q, 2H, CH2CH3,
hindered by NMe2 signal), 2.25 (s, 6H, NMe2), 1.94 (t,
2H, CH2), 1.05 (t, 3H, CH2CH3). MS (EI, 70 eV) m/z,
511 [M+], 427 [(M−3CO)+].
Aliquots (200 ml) of sheep uterine cytosol prepared as
described in Ref. [24] were incubated for 3 h at 0 °C
with [6,7-3H]-estradiol (2×10−9 M, specific activity
1.96 TBq mmol−1) in the presence of nine concentra-
tions of the hormones to be tested. Stock solutions
(1×10−3 M) of the compounds were prepared in
DMSO. At the end of the incubation period, the free
and bound fractions of the tracer were separated by
protamine sulphate precipitation. The percentage re-
duction in binding of [3H]-estradiol (Y) was calculated
using the logit transformation of Y (logitY: ln[y/1−Y]
versus the log of the mass of the competing steroid. The
concentration of unlabeled steroid required to displace
50% of the bound [3H]-estradiol was calculated for each
steroid tested, and the results expressed as RBA. The
RBA value of estradiol is by definition equal to 100%.
6.2.4. Titanium complex 4%
Compound 17 (0.200 g, 0.39 mmol) was dissolved in
a mixture of technical grade ethyl ether (20 ml) and
MeOH (10 ml). The pyrex tube containing the solution
was placed under a UV lamp (TQ150) for 30 min
irradiation. Significant gas evolution was observed and
a brown powder precipitated. The solution was filtered
and evaporated. TLC on silica gel with 20:1 THF–
Et3N as eluent gave 5 (0.144 g) as a beige oil. Com-
pound 5 was immediately used to prepare titanium
complex. Compound 5 was dissolved in THF (7 ml)
and cooled to −70 °C. n-BuLi (0.34 ml of a 1.2 M
solution in hexane, 0.41 mmol,) was added dropwise.
The lithium salt solution was stirred for 1 h 15 min.
while the temperature was allowed to rise slowly to
−20 °C, and the solution became orange. The solution
was re-cooled to −70 °C, then CpTiCl3 (0.085 g, 0.39
mmol) was added as a solid in one portion, and the
solution immediately turned dark red. After stirring for
1.5 h, a flow of HCl gas was passed into the mixture for
5 min. The THF was removed under vacuum and the
dark brown solid obtained was washed twice by Et2O.
The solid was redissolved in CH2Cl2 saturated with HCl
(15 ml). The solution was filtered and concentrated to 2
ml. Et2O was added yielding a dark red precipitate of 4%
7.2. Measurement of the lipophilicity of 4%
Log Po/w were measured by reverse phase HPLC
using the method described by Pomper et al. [25].
7.3. Test on MCF7 cells
7.3.1. Culture materials
Earle’s based minimal essential medium (MEM), fe-
tal bovine serum (FBS),
L-glutamine, penicillin, gen-
tamicin, streptomycin were obtained from Gibco
(Ghent, Belgium), plastic culture materials from Falcon
(Ghent, Belgium).
7.3.2. Culture conditions
1
(0.130 g, 57%). H-NMR (200 MHz, CD3OD): l 7.32–
MCF7 cells were from the Michigan Cancer Founda-
tions (Detroit). Cells are maintained in monolayer cul-
ture in Dulbecco–MEM added with 10% thermally
6.83 (m, 9H, aromatic rings), 6.53 and 6.52 (s, s, 5H,
C5H5), 6.32 (m, 2H, C5H4), 6.06 and 6.00 (t, t, 2H,
C5H4), 4.10 (q, 2H, OCH2), 2.93 (s, 6H, NMe2), 2.41
(m, 2H, CH2CH3), 2.21 (m, 2H, CH2), 0.98 and 0.99 (t,
t, 3H, CH2CH3). MS (FAB) 556 [M+H]+, 520 [M−
Cl]+, 372 [M−CpTiCl2]+. High resolution MS: Calc.
[M+H]+ 556.1657, Found: [M+H]+ 556.1651. Anal.
Calc. for C31H35Cl2NOTi·HCl: C, 62.80; H, 6.12; N,
2.36. Found: C, 61.60; H, 6.66; N, 2.41%.
inactivated FBS,
L
-glutamine (0.6 mg ml−1) and a
cocktail of antibiotics (gentamicin 40 mg ml−1, peni-
cillin 100 U ml−1, streptomycin 100 mg ml−1). Growth
of the cells was assessed by measuring the DNA con-
tent of treated and untreated (control) cells after 5 days
of culture [26].
Acknowledgements
7. Biochemical studies
Financial support from CNRS and MRT (France)
and from DeBiopharm (Switzerland) is gratefully ac-
knowledged. We wish to thank A. Cordaville for tech-
nical assistance and B. McGlinchey for her assistance in
translating the manuscript.
In order to obtain reproducible results new stock
solutions (1×10−3 M) have to be prepared for each
experiment with freshly purified organometallic
complexes.

356
S. Top et al. / Journal of Organometallic Chemistry 643–644 (2002) 350–356
[14] J.E. McMurry, Acc. Chem. Res. 16 (1983) 405.
References
[15] M. Graff, A. Al Dilaimi, P. Seguineau, M. Rambaud, J. Vil-
lieras, Tetrahedron Lett. 14 (1986) 1577.
[16] J.M. Osgerby, P.L. Pauson, J. Chem. Soc. (1961) 4604.
[17] B.M. Foxman, D.A. Gronbeck, N. Khrushova, M. Rosenblum,
J. Org. Chem. (1993) 4078.
[18] S. Top, E.B. Kaloun, G. Jaouen, J. Am. Chem. Soc. 122 (2000)
736.
[19] E. Samuel, Bull. Soc. Chim. Fr. (1966) 3548.
[20] J.H. Toney, T. Marks, J. Am. Chem. Soc. 107 (1985) 947.
[21] Y. Maaroufi, A. Cleeren, G. Leclercq, J. Biol. Inorg. Chem. 3
(1998) 508.
[22] A. Stoica, B.S. Katzenellenbogen, M.B. Martin, Mol. Endo. 14
(2000) 545.
[23] P.F. Predki, B. Sarkar, J. Biol. Chem. 267 (1992) 5842.
[24] A. Vessie`res, S. Top, A.A. Ismail, I.S. Butler, M. Louer, G.
Jaouen, Biochemistry 27 (1988) 6659.
[25] M.G. Pomper, H. Van Brocklin, M.M. Thieme, R.D. Thomas,
D.O. Kiesewetter, K.E. Carlson, C.J. Mathias, M.J. Welch, J.A.
Katzenellenbogen, J. Med. Chem. 33 (1990) 3143.
[26] G. Leclercq, N. Devleeschouwer, J.C. Heuson, J. Steroid
Biochem. 19 (1983) 75.
[1] E. Wong, C.M. Giandomenico, Chem. Rev. 99 (1999) 2451.
[2] M.J. Clarke, F. Zhu, D.R. Frasca, Chem. Rev. 99 (1999) 2511.
[3] P. Ko¨pf-Maier, Eur. J. Clin. Pharmacol. 47 (1994) 1.
[4] Z. Guo, P.J. Sadler, Angew. Chem. Int. Ed. Engl. 38 (1999)
1512.
[5] M. Guo, S. Hongzhe, H.J. McArdle, L. Gambling, P.J. Sadler,
Biochemistry 39 (2000) 10023.
[6] S. Top, H. El Hafa, A. Vessie`res, J. Quivy, J. Vaisserman, D.W.
Huges, M.J. McGlinchey, J.P. Mornon, E. Thoreau, G. Jaouen,
J. Am. Chem. Soc. 117 (1995) 8372.
[7] G. Jaouen, A. Vessie`res, G. Leclercq, J. Quivy, L. Jin, A. Croisy,
C.R. Acad. Sci. Paris, Se´rie IIc (2000) 89.
[8] G. Jaouen, S. Top, A. Vessie`res, P. Pigeon, G. Leclercq, I. La¨ıos,
Chem. Commun. (2001) 383.
[9] K. Dhingra, Invest. New Drugs 17 (1999) 285.
[10] C.K. Osborne, H. Zhao, S.A.W. Fuqua, J. Clin. Oncol. 18
(2000) 3172.
[11] M.F. Sullivan, W.F. Little, J. Organomet. Chem. 8 (1967) 277.
[12] J.K. Stille, Angew. Chem. Int. Ed. Engl. 25 (1986) 508.
[13] C.R. Johnson, J.P. Adams, M.P. Braun, C.B.W. Senanayake,
Tetrahedron Lett. 33 (1992) 919.