Novel titanocene anti-cancer drugs derived from fulvenes and titanium dichloride

Article abstract of DOI:10.1016/j.jorganchem.2004.04.015

Starting from 6-(p-N,N-dimethylanilinyl)fulvene (1a) or 6-(pentamethylphenyl)fulvene (1b) [1,2-di(cyclopentadienyl)-1,2-di (p-N,N-dimethylaminophenyl)ethanediyl] titanium dichloride (2a) and [1,2-di(cyclopentadienyl)-1,2-bis(pentamethylphenyl)ethanediyl] titanium dichloride (2b) and their corresponding dithiocyanato complexes (3a, 3b) were synthesized. Titanocene 2b did not show a cytotoxic effect, but when 2a was tested against pig kidney carcinoma cells (LLC-PK) or human ovarian carcinoma cells (A2780/cp70) inhibitory concentrations (IC₅₀) of 2.7 × 10^-4 and 1.9 × 10^-4 M, respectively, were observed.

Full text of DOI:10.1016/j.jorganchem.2004.04.015

Journal of Organometallic Chemistry 689 (2004) 2242–2249
www.elsevier.com/locate/jorganchem
Novel titanocene anti-cancer drugs derived from fulvenes
and titanium dichloride
a,
b
a
b
Matthias Tacke *, Lorcan T. Allen , Laurence Cuffe , William M. Gallagher ,
a
a
a
a
Ying Lou , Oscar Mendoza , Helge Muller-Bunz , Franz-Josef K. Rehmann ,
€
Nigel Sweeney
a
a
Chemistry Department, Conway Institute of Biomolecular and Biomedical Research, Centre for Synthesis and Chemical Biology (CSCB),
University College Dublin, Belfield, Dublin 4, Ireland
b
Pharmacology Department, Conway Institute of Biomolecular and Biomedical Research, University College Dublin, Belfield, Dublin 4, Ireland
Received 9 March 2004; accepted 8 April 2004
Abstract
Starting from 6-(p ꢀ N; N-dimethylanilinyl)fulvene (1a) or 6-(pentamethylphenyl)fulvene (1b) [1,2-di(cyclopentadienyl)-1,2-
di(p ꢀ N; N-dimethylaminophenyl)ethanediyl] titanium dichloride (2a) and [1,2-di(cyclopentadienyl)-1,2-bis(pentamethylphe-
nyl)ethanediyl] titanium dichloride (2b) and their corresponding dithiocyanato complexes (3a, 3b) were synthesized. Titanocene 2b
did not show a cytotoxic effect, but when 2a was tested against pig kidney carcinoma cells (LLC-PK) or human ovarian carcinoma
cells (A2780/cp70) inhibitory concentrations (IC50) of 2.7 ꢁ 10ꢀ4 and 1.9 ꢁ 10ꢀ4 M, respectively, were observed.
Ó 2004 Elsevier B.V. All rights reserved.
Keywords: Anti-cancer drug; cis-Platinum; Titanocene; Fulvene; LLC-PK; A2780/cp70
1. Introduction
derivative [10–12], despite the existence of a novel
method starting from titanium dichloride and fulvenes
[13–16], which allows direct access to highly substituted
ansa-titanocenes [17,18]. This paper reports the synthesis
of novel [(1,2-diaryl-1,2-dicyclopentadienyl)-ethanediyl]
titanium dichlorides [19,20], which combine the reac-
tivity of the titanium dichloride moiety with the ability
of hydrogen bonding towards DNA of the ammine li-
gand of cis-platinum, if the aryl group is substituted
accordingly.
Despite the resounding success of cis-platinum and
closely related platinum anti-tumor agents, the move-
ment of other transition-metal anti-cancer drugs to-
wards the clinic has been exceptionally slow [1–3].
Metallocene dichlorides (Cp2MCl2) with M ¼ Ti, V, Nb
and Mo show remarkable anti-tumor activity [4,5].
However, only titanocene dichloride has reached Phase I
clinical trials so far, with a maximum tolerable dose of
315 mg/m2 per week. The dose limiting effects of ti-
tanocene dichloride include nephrotoxicity and eleva-
tion of creatine and bilirubin levels [6,7]. Unfortunately,
the efficacy of Cp2TiCl2 in Phase II clinical trials in
patients with metastatic renal-cell carcinoma [8] or
metastatic breast cancer [9] was too low to be pursued.
Nevertheless, little synthetic effort has been employed to
increase the cytotoxicity of any titanocene dichloride
2. Experimental
Titanium tetrachloride and n-butyl lithium (2 mol
solution in hexane) were obtained commercially from
Aldrich Chemical Co. Potassium thiocyanate (KNCS)
was obtained commercially from Aldrich Chemical Co;
it was well ground before the reaction and dried un-
der vacuum. Acetone was dried over magnesium sul-
fate. THF and toluene were dried over and distilled
from Na/benzophenone prior to use. p-(N; N-dimethyl-
* Corresponding author. Tel.: +35-317-168-428; fax: +35-317-162-
127.
E-mail address: matthias.tacke@ucd.ie (M. Tacke).
0022-328X/$ - see front matter Ó 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2004.04.015

M. Tacke et al. / Journal of Organometallic Chemistry 689 (2004) 2242–2249
2243
amino)benzaldehyde and pentamethylbenzaldehyde
were obtained commercially from Aldrich Chemical Co.
Cyclopentadiene was collected under an atmosphere of
nitrogen from freshly cracked dicyclopentadiene and
pyrrolidine was distilled under argon prior to use. Ma-
nipulation of air and moisture sensitive compounds was
carried out using standard Schlenk techniques under an
argon atmosphere. NMR spectra were measured on a
Varian 300 MHz spectrometer. Chemical shifts are re-
ported in ppm and are referenced to TMS. IR spectra
were recorded on a Perkin–Elmer Paragon 1000 FT-IR
Spectrometer employing a KBr disk. UV/Vis spectra
were recorded on an UNICAM UV/VIS Spectrometer
in either methanol (fulvenes) or acetonitrile (titanoc-
enes). The GC mass spectra were measured on a
FINNIGAN TRACE GC MS 2000Series (70 eV) and a
1 ꢁ 10ꢀ5 M solution in ethyl acetate was used. A Gal-
lenkamp Melting Point Apparatus was used for melting
point measurement.
1b and 3b were collected on a BRUKER Smart Apex
diffractometer at room temperature for 1b and 100 K for
3b, respectively. A semi-empirical absorption correction
on the raw data was performed using the program SAD-
ABS [21]. The crystal structure was then solved by direct
methods (SHELXS-NT 97 [22] and refined by full-matrix
least squares methods against F 2. Further details about
the data collection are listed in Table 1, as well as reli-
ability factors. Further details are available free of charge
from the Cambridge structural database under the
CCDC Nos. 232680 (1b) and 232681 (3b). In the crystal
structure of 3b additionally roughly one-third of a mol-
ecule of chloroform was found per asymmetric unit. All
non-hydrogen atoms except the solvent carbon were re-
fined anisotropically. For the refinement of the partially
occupied CHCl3 positions all chlorine–chlorine distances
and carbon–chlorine bond lengths were restrained to be
equal. The hydrogen atoms were geometrically idealised
and refined using a riding model. Their isotropic tem-
perature factors were fixed to 150% (methyl hydrogens)
or 120% (all other hydrogens) of the equivalent temper-
ature factor of the atom the hydrogen is attached to. A
careful examination of the electron density map of 1b
showed that the methyl groups are disordered.
A single crystal of fulvene 1b suitable for X-ray dif-
fraction experiments was grown by slow evaporation of
petroleum spirit (40–60) from an open sample bottle. For
the ansa-titanocene 3b, slow evaporation of chloroform-
d from an NMR tube was used. X-ray diffraction data for
Table 1
Crystal data and structure refinement for 1b and 3b
Identification code
1b
3b
Empirical formula
Formula weight
Temperature
C17H20
224.33
293(2) K
C36H40N2S2 Ti ꢁ 0.35 CHCl3
654.68
100(2) K
ꢀ
0.71073 A
ꢀ
0.71073 A
Wavelength
Crystal system
Space group
Unit cell dimensions
Monoclinic
P21=n ð#14Þ
Triclinic
ꢁ
P1 ð#2Þ
ꢀ
ꢀ
a ¼ 11:374ð2Þ A
a ¼ 9:4666ð15Þ A
ꢀ
ꢀ
b ¼ 5:7535ð11Þ A
b ¼ 10:9852ð18Þ A
ꢀ
ꢀ
c ¼ 20:851ð4Þ A
c ¼ 17:271ð3Þ A
a ¼ 90°
a ¼ 92:463ð3Þ°
b ¼ 93:033ð4Þ°
c ¼ 104:951ð3Þ°
b ¼ 93:027ð3Þ°
c ¼ 90°
3
3
ꢀ
1362.5(4) A
ꢀ
Volume
Z
1729.9(5) A
4
2
Density (calculated)
Absorption coefficient
F ð000Þ
1.094 Mg/m3
0.061 mmꢀ1
488
1.257 Mg/m3
0.472 mmꢀ1
689
Crystal size
0.40 ꢁ 0.30 ꢁ 0.05 mm3
0.20 ꢁ 0.10 ꢁ 0.05 mm3
h range for data collection
Index ranges
1.96–24.00°
2.20–22.50°
ꢀ13 6 h 6 12, ꢀ6 6 k 6 6, ꢀ23 6 l 6 16
5809
ꢀ10 6 h 6 10, ꢀ11 6 k 6 11, ꢀ18 6 l 6 18
Reflections collected
Independent reflections
Completeness to max. h
Absorption correction
Max. and min. transmission
Refinement method
Data/restraints/parameters
Goodness-of-fit on F 2
Final R indices ½I > 2rðIÞꢂ
R indices (all data)
Largest diff. peak and hole
9459
4388 ½RðintÞ ¼ 0:1379ꢂ
2138 ½RðintÞ ¼ 0:0277ꢂ
99.4%
Semi-empirical from equivalents
96.9%
0.9970 and 0.5061
0.9768 and 0.8526
Full-matrix least-squares on F 2
2138/0/154
1.034
4388/18/412
0.791
R1 ¼ 0:0483, wR2 ¼ 0:1325
R1 ¼ 0:0828, wR2 ¼ 0:1515
R1 ¼ 0:1357, wR2 ¼ 0:2180
R1 ¼ 0:2279, wR2 ¼ 0:2563
ꢀ3
ꢀ3
ꢀ
ꢀ
0.140 and )0.136 e A
0.454 and )0.401 e A

2244
M. Tacke et al. / Journal of Organometallic Chemistry 689 (2004) 2242–2249
With a view to elucidate the structures, spectroscopic
(3 ꢁ 20 ml). The combined organic extracts were washed
with a saturated aqueous NaCl solution. The organic
solution was dried over magnesium sulphate. The crude
product was redissolved in petroleum spirit (40–60) and
purified by column chromatography over silica gel 60
(0.063–0.200) and petroleum spirit (40–60) as the eluent,
yielding 2.0 g (89% yield). m.p. 100 °C.
data, bonding properties and energies of formation, the
application of theoretical methods is advantageous. For
this purpose, the GAUSSIAN 98 Revision A11 [23] run-
ning under Red Hat Linux was used. DFT calculations
were performed at the B3LYP level using the 6-31G**
basis set for the species of interest.
1H NMR (d ppm CDCl3): 6.49, 6.36, 6.02 (C5H4, 4H
m); 2.27 (o, m-CH3, 12H s); 2.16 (p-CH3, 3H s); 7.30
(Ph–CH–Cp, 1H s).
2.1. 6-(p-N; N-Dimethylanilinyl)fulvene (1a)
13C NMR (d ppm CDCl3): 139.6, 133.7, 131.4, 125.1,
121.9 (C5H4); 18.5, 16.4 (o; m-CH3); 16.8 (p-CH3).
IR absorption (cmꢀ1 KBr): 3064 (w); 2922 (w); 2860
(w); 1636 (C@C m); 1472 (w); 1367 (m); 1066 (m); 898
(m); 768 (s); 616 (s).
The syntheses of fulvenes 1a and 1b were carried out
under argon as outlined in [23]. Pyrrolidine (2.5 ml, 30.0
mmol) was added to a solution of p-(N; N-dimethyla-
mino)benzaldehyde (3.0 g, 30.0 mmol) and cyclopent-
adiene (4.1 ml, 50.0 mmol) in 30 ml of methanol. After
addition, the colour of the solution immediately turned
from colourless to red-orange. Large amounts of an
orange solid precipitated out of the solution. When TLC
analysis showed only one product band after 15 min,
acetic acid (1.8 ml, 32.0 mmol) was added. The reaction
mixture was diluted with 20 ml of a mixture of diethyl
ether and water (1:1). The resultant organic layer was
separated and the aqueous layer was washed with di-
ethyl ether (3 ꢁ 20 ml). The combined organic extracts
were washed with a saturated aqueous NaCl solution.
The organic solution was dried over magnesium sulfate.
When the solvent was removed under reduced pressure,
3.6 g of an orange product were obtained (90% yield).
m.p. 102 °C.
GCMS: 224.2 (M 60%); 209.2 (Mþ ) CH3þ 100%);
194.2 (Mþ ) 2CHþ3 90%); 179.2 (Mþ ) 3CHþ3 70%).
UV/Vis (methanol): kmax ¼ 302 nm.
Anal. Calc. for C17H20: C, 91.01; H, 8.99. Found: C,
90.57; H, 9.06%.
2.3. [1,2-Di(cyclopentadienyl)-1,2-di(p-N,N-dimethyl-
aminophenyl)ethanediyl] titanium dichloride [1,2-(4-
Me2N-C6H4)2 C2H2{g5-C5H4}]TiCl2 (2a)
TiCl4 (0.70 ml, 6.3 mmol) was added to 40 ml of dry
toluene containing 5% dry THF. The solution turned
immediately from colourless to pale yellow. The solu-
tion was stirred and cooled down to )78 °C, followed by
drop wise addition of n-butyl lithium (6.4 ml, 12.7
mmol). The solution turned from yellow to brown
during addition. After this addition, the mixture was
allowed to warm up slowly to room temperature. The
colour of the solution became finally black. After 20 h
stirring at r.t., a solution of 1a (2.5 g, 12.6 mmol) in 35
ml of dry toluene was added to the TiCl2 ꢃ 2THF solu-
tion at r.t. under argon. Then it was stirred under reflux
for another 20 h. The solvent was removed under vac-
uum leaving a black residue. The residue was washed
with chloroform and the solution was filtered through
celite under reduced pressure. The colour of the filtrate
reddened slightly. It was filtered using gravity filtration
for at least four times until no further black precipitate
appeared on the filter paper and the filtrate turned to
dark red. Chloroform was removed to leave dark-red
solid 2a with 1.5 g (3.0 mmol, 47% yield). The ratio of
trans and cis isomers is 60% and 40%, respectively.
1H NMR (d ppm CDCl3): 7.15–6.68 (C6H4, 8H m
and C5H4, 4H m); 6.30–6.12 (C5H4, 4H m); 5.40 (trans-
PhCHCp, 2H s); 4.76 (cis-PhCHCp, 2H s); 2.91 (cis-
N(CH3)2, 12H s); 2.89 (trans-N(CH3)2, 12H s).
1H NMR (d ppm CDCl3): 6.79, 6.63, 6.43, 6.31
(C5H4, 4H m); 7.58, 7.55, 6.71, 6.68 (C6H4, 4H m); 3.03
(N(CH3)2, 6H s); 7.13 (Ph–CH–Cp, 1H s).
13C NMR (d ppm CDCl3): 140.9, 133.7, 128.1, 127.5,
119.5 (C5H4); 151.2, 132.8, 124.8, 112.0 (C6H4); 40.1
(CH3); 139.8 (Ph–CH–Cp).
IR absorption (cmꢀ1 KBr): 3072 (w); 2918 (w); 2816
(w); 1605 (C@C m); 1450 (m); 1373 (m); 816 (s).
GCMS: 197.2 (M 100%); 182.2 (Mþ ) CH3þ 25%);
153.1 (Mþ ) N(CH3)þ2 92%).
UV/Vis (methanol): kmax ¼ 399 nm.
Anal. Calc. for C14H15N: C, 85.24; H, 7.66; N, 7.10.
Found: C, 84.40; H, 7.69; N, 6.86%.
2.2. 6-(Pentamethylphenyl)fulvene (1b)
Pyrrolidine (1.3 ml, 15.0 mmol) was added to a so-
lution of pentamethylbenzaldehyde (1.8 g, 10.0 mmol)
and cyclopentadiene (2.1 ml, 26.0 mmol) in 30 ml of
methanol. After this addition the solution turned from
colourless to clear orange-yellow. When TLC analysis
showed only one product band after 20 h, acetic acid
(0.9 ml, 16.0 mmol) was added. The reaction mixture
was diluted with 20 ml of a mixture of diethyl ether and
water (1:1). The resultant organic layer was separated
and the aqueous layer was washed with diethyl ether
13C NMR (d ppm CDCl3): 139.6, 138.4, 133.8, 130.3,
129.7, 128.5, 128.2, 126.3, 117.0, 116.9, 114.0, 113.2,
109.9 (C6H4and C5H4); 53.0 (cis-PhCHCp, 2H); 50.7
(trans-PhCHCp); 41.1 (N(CH3)2).

M. Tacke et al. / Journal of Organometallic Chemistry 689 (2004) 2242–2249
2245
IR absorption (cmꢀ1 KBr): 3059 (w); 2956 (w); 2916
(w); 2853 (w); 1520 (s); 1480 (m); 1351 (m); 1261 (m);
1130 (m); 1019 (m); 946 (m); 802 (s); 694 (m); 596 (m).
UV/Vis (MeCN): kmax ¼ 265 nm.
13C NMR (d ppm CDCl3): 149.7, 149.0, 141.4, 129.5,
129.1, 128.3, 126.3, 126.0, 124.4, 116.9, 114.3, 112.4,
110.9 (C6H4 and C5H4); 122.1 (NCS); 53.4 (cis-
PhCHCp, 2H); 50.8 (trans-PhCHCp); 40.5 (N(CH3)2).
IR absorption (cmꢀ1 KBr): 3075 (w); 2952 (w); 2920
(w); 2886 (w); 2852 (w); 2795 (w); 2048 ðmsðCNÞÞ; 2006
ðmasðCNÞÞ; 1611 (m); 1519 (s); 1479 (w); 1351 (m); 1204
(w); 1163 (w); 1059 (w); 946 (m); 820 (s).
Anal. Calc. for C28H30N2Cl2Ti:C, 65.53; H, 5.89; N,
5.46. Found: C, 64.94; H, 6.31; N, 4.94%.
2.4. [1,2-Di(cyclopentadienyl)-1,2-bis(pentamethyl-
phenyl)ethanediyl] titanium dichloride [1,2-(Me5C6)2
C2H2{g5-C5H4}]TiCl2 (2b)
UV/Vis (MeCN): kmax ¼ 285 nm.
Anal. Calc. for C30H30N4S2Ti: C, 64.51; H, 5.41; N,
10.03; S, 11.48. Found: C, 63.89; H, 5.94; N, 10.52; S,
10.07%.
TiCl4 (0.5 ml, 4.5 mmol) was added to 40 ml of dry
toluene containing 5% dry THF. The solution turned
immediately from colourless to pale yellow. The solu-
tion was stirred and cooled down to )78 °C, and then
was treated dropwise with n-butyllithium (4.5 ml, 8.9
mmol). The solution turned from yellow to brown
during the addition. After this addition, the mixture was
allowed to warm up slowly to r.t. The colour of the
solution finally became black. After 20 h stirring, a so-
lution of 1b (2.0 g, 8.9 mmol) in dry toluene was added
to the solution of TiCl2 ꢃ 2THF at r.t. under argon. Then
it was stirred under reflux for another 20 h. After per-
forming the extraction procedure for 2a, the product
was washed with hexane yielding 0.3 g (10%) of a cop-
per-red product 2b. The ratio of trans and cis isomers is
93% and 7%, respectively.
2.6. [1,2-Di(cyclopentadienyl)-1,2-bis(pentamethyl-
phenyl)ethanediyl]titanium dithiocyanate [1,2-(Me5C6)2
C2H2(g5-C5H4)2]Ti(NCS)2 (3b)
KNCS (0.2 g, 2.6 mmol) was added to a solution of
ansa-titanocene dichloride 2b (0.31 g, 0.5 mmol) in 30 ml
acetone. The mixture was refluxed for 3 h, filtered while
still hot and the solvent was removed under reduced
pressure. This gave 0.26 g (79% yield) of a red-brown
solid. The ratio of trans and cis isomers is 93% and 7%,
respectively.
1H NMR (d ppm CDCl3): 6.89–6.56 (C5H4, 8H m);
6.45 (trans-PhCHCp, 2H s); 4.12 (cis-PhCHCp, 2H s);
2.17 (p-CH3,6H s); 2.15 (o; m-CH3, 24H d).
1H NMR (d ppm CDCl3): 6.98–6.56 (C5H4, 8H m);
6.32 (trans-PhCHCp, 2H s); 4.23 (cis-PhCHCp, 2H s);
2.23 (p-CH3, 6H s); 2.15 (o; m-CH3, 24H).
13C NMR (d ppm CDCl3): 141.7,139.6, 134.5, 127.1,
117.8, 114.2, 111.2, 108.6 (C5H4); 122.7 (NCS); 49.2
(trans-PhCHCp); 36.9 (cis-PhCHCp); 17.5 (CH3).
IR absorption (cmꢀ1 KBr): 3096 (w); 2921 (m); 2870
(w); 2048 (masðCNÞ vs); 2001 (msðCNÞ vs); 1618 (m); 1452
(m); 1379 (w); 1260 (w); 1063 (m); 819 (m); 747 (w).
UV/Vis (MeCN): kmax ¼ 253 nm.
13C NMR (d ppm CDCl3): 138.7, 134.7, 134.0, 131.2,
126.4, 118.2, 113.6, 111.3 (C5H4); 48.5 (trans-PhCHCp);
31.8 (cis-PhCHCp); 17.5 (CH3).
IR absorption (cmꢀ1 KBr): 3126 (w); 2990 (w); 2920
(s); 2872 (w); 1632 (m); 1570 (m); 1452 (s); 1380 (m);
1260 (w); 1062 (m); 926 (w); 812 (s).
Anal. Calc. for C36H40N2S2Ti: C, 70.57; H, 6.58; N,
4.57; S, 10.47. Found: C, 69.77; H, 7.08; N, 9.95; S,
9.81%.
UV/Vis (MeCN): kmax ¼ 293 nm.
Anal. Calc. for C34H40Cl2Ti: C, 71.96; H, 7.11; Cl,
12.50. Found: C, 71.44; H, 7.68; Cl, 12.05%.
2.7. MTT-based cytotoxicity tests
2.5. [1,2-Di(cyclopentadienyl)-1,2-di(p-N,N-dimethyl-
aminophenyl)ethanediyl]titanium dithio-cyanate [1,2-(4-
Me2N-C6H4)2C2 H2(g5-C5H4)2]Ti(NCS)2 (3a)
The pig kidney carcinoma cell line, LLC-PK, was
obtained from the American Tissue Culture Collection.
The platinum-resistant human ovarian carcinoma cell
line, A2780/cp70, was kindly provided by Prof. Robert
Brown, Centre for Oncology and Applied Pharmacol-
ogy, University of Glasgow, Cancer Research UK
Beatson Laboratories.
The cytotoxic activities of titanocenes 2a and 2b were
determined using an MTT-based assay. In more detail,
cells were seeded into a 96-well plate (5000 cells/well)
and allowed to attach for 24 h. Subsequently, the cells
were treated with various concentrations of the cyto-
toxic agents. After 48 h, the relevant drug was removed,
the cells washed with PBS and fresh medium was added
for another 24 h for recovery. Viability of cells was de-
termined by treatment with MTT in medium (5 mg/11
KNCS (0.1 g, 1.3 mmol) was added to a solution of
ansa-titanocene dichloride 2a (0.3 g, 0.5 mmol) in 30 ml
acetone. The mixture was refluxed for 3 h, filtered while
still hot and the solvent was removed under reduced
pressure. This gave 0.3 g (62% yield) of a dark-brown
solid. The ratio of trans and cis isomers is 60% and 40%,
respectively.
1H NMR (d ppm CDCl3): 7.13–6.69 (C6H4, 8H m
and C5H4, 3H m); 6.83, 6.71, 6.66, 6.53, 6.34, 6.26, 6.07
(C5H4, 5H m); 5.50 (trans-PhCHCp, 2H s); 4.85 (cis-
PhCHCp, 2H s); 2.91 (trans-N(CH3)2, 12H s); 2.89 (cis-
N(CH3)2, 12H s).

2246
M. Tacke et al. / Journal of Organometallic Chemistry 689 (2004) 2242–2249
ml) for 3 h. The purple formazan crystals formed were
dissolved in DMSO and absorbance measured at 540
nm using a VICTOR2 multilabel plate reader (Wallac).
IC50 (inhibitory concentration 50%) values were deter-
mined from the drug concentrations that induced a 50%
reduction in light absorbance.
3.2. Structural discussion
Density functional theory calculations were carried
out for compounds 1b, 2a, and 2b at the B3LYP level
using the 6-31G** basis set. In addition, X-ray diffrac-
tion measurements were carried out for compounds 1b
and 3b (see Table 1 for further details about the data
collection). For compound 3b in accordance with the
ꢁ
3. Results and discussion
centrosymmetric space group P1 the R; R- and S; S-iso-
mers are found in equal amounts in the crystal.
3.1. Synthesis
Selected bond lengths of the optimised structure of
fulvene 1b (Fig. 3) can be found in Table 2. As expected,
the carbon–carbon bond lengths in the cyclopentadiene
system of fulvene 1b vary significantly, demonstrating
the absence of a significant resonance system within the
five-membered ring. This is confirmed by the length of
the regular exocyclic double bond C(1)–C(6) found at
135.4 pm and the single bond carrying the phenyl sub-
stituent C(6)–C(7) is calculated as 148.1 pm (Table 2).
The corresponding values of the crystal structure de-
termination tend to be around 1–3 pm shorter than the
calculated values (Table 2, Fig. 3). Especially out-
standing is the dihedral angle of 61.7°, which shows that
Fulvenes 1a and 1b (Fig. 1) were synthesized, ac-
cording to [24], by reacting the corresponding benzal-
dehyde with cyclopentadiene in the presence of
pyrrolidine as a base. The compounds were formed in
high yields, of about 90%.
Titanocenes 2a and 2b (Fig. 2) were synthesized by
reductive dimerisation of fulvenes 1a and 1b with tita-
nium dichloride, respectively. TiCl2 was obtained by
reduction of TiCl4 with nBuLi as described in [17,18].
The determined cis–trans ratio at the bridge is 60:40 for
2a, and 93:7 for 2b. The high trans-ratio in 2b reflects the
steric bulk of the pentaphenyl rings making the trans
geometry highly favoured. This steric strain might also
be the cause for the very small yield of only 10% ob-
tained for 2b, compared with 47% for 2a. Substitution of
the chloride ligands is easily achieved by reaction of 2a
and 2b with potassium thiocyanate in acetone under
reflux to obtain the corresponding dithiocyanato com-
plexes 3a and 3b (Fig. 2), respectively [25].
C13
C2
C3
C4
C14
C1
C6
C5
C8
C7
C9
C12
C10
C11
(a)
(b)
C15
N
C16
C15
Fig. 3. Molecular structure of 1b; thermal ellipsoids are drawn on the
50% probability level; from disordered methyl groups only one ori-
entation is shown.
Fig. 1. Structures of fulvenes 1a and 1b.
Me2N
Table 2
Selected bond lengths of the DFT-calculated structure and crystal
structure of fulvene 1b
H
H
X
X
Bond length (pm) (1b) Bond length (pm) (1b)
DFT structure
Ti
Ti
X
crystal structure
X
C(1)–C(2)
C(2)–C(3)
C(3)–C(4)
C(4)–C(5)
C(5)–C(1)
C(1)–C(6)
C(6)–C(7)
147.1
135.5
147.2
135.6
147.1
135.4
148.1
146.2
132.1
143.8
133.6
143.9
134.5
148.7
Me2N
H
H
2a: X = Cl
3a: X = NCS
2b: X = Cl
3b: X = NCS
Fig. 2. Structures of ansa-titanocene dichlorides 2a/b, and ansa-
titanocene dithiocyanates 3a/b.

M. Tacke et al. / Journal of Organometallic Chemistry 689 (2004) 2242–2249
2247
the bulky aryl group does not allow co-planarity of the
full aromatic ring system. The crystal structure confirms
this twist with an even larger angle of 82.2°. The sig-
nificant difference between calculated and experimental
angle may be due to packing effects in the crystal.
Optimised structures were also calculated for tita-
nocenes 2a and 2b (Scheme 1) at the B3LYP level using
the 6-31G** basis set. Selected bond lengths of these
structures are listed in Table 3.
The length of bonds between the metal centre and the
carbon atoms of the cyclopentadienyl rings bound to the
metal ion are similar for both titanocene complexes.
They vary between 236.8 and 244.5 pm for 2a and
237.0–245.1 pm for 2b with values slightly different for
the different cyclopentadienyl rings. The same applies
for the carbon-carbon bonds of the cyclopentadienyl
rings with bonds length between 140.3 and 143.0 pm for
2a and 140.2–142.9 pm for 2b. The higher steric bulk in
titanocene 2b leads to a elongated bond of the carbon
bridge (C(6)–C(60)) compared to 2a with values of 157.1
and 156.2 pm, respectively. The titanium–chlorine bond
length are almost identical for 2a and 2b, with values for
234.7 pm and 235.1 pm for 2a and identical values for 2b
of 234.8 pm. For 2a the TiCl2 angle was calculated to be
97.5° and the dihedral angle between the aryl rings to be
62.1°. The corresponding values for 2b are 97.5° and
83.8°, respectively. For 2b, the angles formed by the
bonds between C(1), C(6) and C(60) is 106.0°, between
C(7), C(6), and C(60) is 120.6, and between C(7), C(6)
and C(1) 114.9°, compared with the values found for the
X-ray structure of 110.4°, 125.6°, and 115.0° for 3b
(Fig. 4), the dithiocyanato derivative of 2b. The corre-
sponding calculated values of 2a are 108.0°, 113.9°, and
113.5°. The significantly larger angle formed between
C(7), C(6), and C(60) in 2b compared with 2a demon-
strates the bulky overload given by the pentamethyl-
phenyl substituent. In contrast, the angles formed
between the centroids of the cyclopentadienyl rings are
almost identical, with values of 128.0° for 2a and 128.3°
in 2b. The carbon–carbon bonds of the cyclopentadienyl
rings found in the crystal structure of complex 3b tend to
be around 2–3 pm shorter than the calculated values of
2b. And also the titanium–carbon bond distances of
232.4–237.9 pm found in the crystal structure of 3b tend
to be significantly shorter (up to 10 pm) compared to the
values calculated for 2b. This difference is even more
accentuated for the carbon–carbon bridge, which was
measured by X-ray diffraction with 146.3 pm for 3b
compared to the calculated value of 157.1 pm in 2b.
Due to the large, unexpected differences between the
DFT-calculated structure of 2b and the crystal structure
3
2
R
4
6
1
H
Cl
5
Ti
Cl'
2'
R'
6'
1'
3'
4'
H'
5'
Scheme 1. Numbering scheme of atoms for the discussed titanocene
complexes 2a/b and 3b.
Table 3
Selected bond lengths from the DFT-calculated structures of com-
plexes 2a/b, and from the crystal structure determination of complex 3b
Bond length
(pm) (2a) DFT (pm) (2b) DFT (pm) (3b)
Bond length
Bond length
structure
structure
crystal structure
Ti–C(1)
Ti–C(2)
241.2
236.8
243.7
244.5
241.4
242.6
237.3
243.1
244.9
240.8
143.0
142.0
140.5
142.4
141.4
242.2
237.0
244.0
245.2
240.1
242.2
237.0
244.0
245.2
240.1
142.9
142.1
140.2
142.3
142.0
142.9
142.1
140.2
142.3
142.0
234.8
234.8
157.1
151.8
235.3
235.2
Ti–C(3)
Ti–C(4)
237.9
237.7
Ti–C(5)
233.2
236.6
Ti–C(10)
Ti–C(20)
Ti–C(30)
Ti–C(40)
Ti–C(50)
C(1)–C(2)
C(2)–C(3)
C(3)–C(4)
C(4)–C(5)
C(5)–C(1)
232.4
237.8
236.2
234.5
144.1
140.0
C7
C4
C1
140.3
139.3
C5
C3
C2
C6
C6’
140.8
140.9
C18’
0C(10)–C(20) 142.3
N’
Ti
S’
C7’
C(20)–C(30)
C(30)–C(40)
C(40)–C(50)
C(50)–C(10)
Ti–Cl
Ti–Cl0
C(6)–C(60)
C(1)–C(6)
142.4
140.3
142.4
141.9
235.1
234.7
156.2
151.4
139.9
138.7
C2’
C1’
139.3
140.0
C3’
N
C5’
C18
C4’
Ti–N: 204.8
Ti–N0: 201.8
146.3
S
Fig. 4. Molecular structure of 3b showing the atom numbering scheme;
thermal ellipsoids are drawn on the 25% probability level.
150.6

2248
M. Tacke et al. / Journal of Organometallic Chemistry 689 (2004) 2242–2249
found for 3b, the refinement of 3b will be discussed in
detail. The aromatic substituents form a channel in
which solvent molecules are situated. As adjacent
CHCl3 locations are very close to each other (the
shortest intermolecular Cl–Cl-distance is 155 pm), these
locations must be partially occupied. As the solvent
positions are so close to each other, a molecule can
easily hop from one position to the next. Therefore, at
room temperature a solution failed completely due to
this easy movement of the solvent. Even at 100 K large
thermal parameters indicate a high mobility of the
CHCl3 molecules, which causes a general haziness of the
electron density map. All components in this crystal
structure arrange in layers, which leads to plate-shaped
crystals with a low mechanical stability. The growth of
crystals is significantly hindered by such an arrange-
ment. In fact, the size of the crystal was at the lower
limit of being suitable for an X-ray structure analysis.
The most important consequence of these two diffi-
culties is the very short C(6)–C(60) distance, compared
with the calculated value for the chlorine compound 2b
(Table 3). A close look on their thermal ellipsoids re-
veals that both are elongated along the C–H bond axis
which probably is a hint on a slight disorder. This is
supported by the position of these carbon atoms, which
are found to be nearly in the plane of their non-hydro-
gen-neighbours. However, the low quality of the crystal
did not allow the refinement of any disorder of this kind.
1,0
0,8
0,6
0,4
0,2
0,0
cis-Pt, IC50: (3.3+/-0.5)E-6
Cp2TiCl2, IC50: (2.0+/-1.0)E-3
2a, IC50: (2.7+/-0.1)E-4
1E-10
1E-9
1E-8
1E-7
1E-6
1E-5
1E-4
1E-3
log10 of drug concentrations
Fig. 5. Cytotoxicity curves from typical MTT assays showing the effect
of cis-platin, Cp2TiCl2, and 2a on the viability of pig kidney carcinoma
(LLC-PK) cells.
1,2
1,0
0,8
0,6
0,4
3.3. Cytotoxicity studies
cis-Pt, IC50: (1.5+/-0.3)E-6
0,2
The in vitro cytotoxicity of compounds 2a and 2b
were determined by an MTT-based assay [26] involving
a 48 h drug exposure period, followed by 24 h of re-
covery time. Initially, compounds 2a and 2b were tested
for their activity on pig kidney carcinoma (LLC-PK)
cells. Due to the low solubility of 2b even in DMSO, the
saturated solution of 2b in DMSO was applied in a di-
lution series with medium, starting with a concentration
of 1% of the 2b-saturated DMSO solution in medium.
However, no cytotoxic effect was observed even at the
highest concentration of the drug. In contrast, 2a
showed a significant inhibition of cell growth at higher
concentrations, with an IC50 value of 2.7 ꢁ 10ꢀ4 M.
Under identical conditions, cis-platin showed an IC50
value of 3.3 ꢁ 10ꢀ6 M, whereas the activity of Cp2TiCl2
was at least one order of magnitude lower (Fig. 5). The
cytotoxicity of 2a is therefore very promising, since it is
shows significantly higher activity compared to
Cp2TiCl2, the latter compound already having reached
use in Phase I/II clinical trials.
Cp2TiCl2, IC50: (6.0+/-0.6)E-4
: (
2a, IC50 1.9+/-0.3)E-4
0,0
1E-11 1E-10
1E-9
1E-8
1E-7
1E-6
1E-5
1E-4
1E-3
log10 drug concentrations
Fig. 6. Cytotoxicity curves from typical MTT assays showing the effect
of cis-platin, Cp2TiCl2, and 2a on the viability of platinum-resistant
human ovarian carcinoma (A2780CP/cp70) cells.
be 1.5 ꢁ 10ꢀ6 and 6.0 ꢁ 10ꢀ4 M, respectively, which are
similar to values published previously [11,27]. Conse-
quently, for human ovarian carcinoma cells, the observed
activity for 2a is higher than for Cp2TiCl2, but still sig-
nificantly less than for cis-platin.
4. Conclusion and outlook
The water-insoluble, non-cytotoxic titanocene 2b was
synthesised using the aryl-substituted fulvene 1b. In
contrast, when the NMe2 substituted fulvene 1a was
used, compound 2a, which resembles cis-platinum to a
certain degree, was produced. In 2a, the two NMe2
substituents resemble the ammine ligands of cis-plati-
num and the TiCl2 is equivalent to the PtCl2 group. This
Additionally, the cytotoxic effect of 2a was also tested
against the cis-platin-resistant human ovarian carcinoma
cell line, A2780/cp70, with an IC50 value of 1.9 ꢁ 10ꢀ4 M
being obtained (Fig. 6). Under equivalent conditions, the
IC50 values of cis-platin and Cp2TiCl2 were determined to

M. Tacke et al. / Journal of Organometallic Chemistry 689 (2004) 2242–2249
2249
[8] G. Lummen, H. Sperling, H. Luboldt, T. Otto, H. Rubben,
Cancer Chemother. Pharmacol. 42 (1998) 415.
compound 2ais significantly more cytotoxic against LLC-
PK and A2780/cp70 cells than titanocene dichloride, for
which Phase I/II clinical trials have been performed. It is
intended to perform further in vitro cellular assays with
the compound to evaluate its potential for testing in ani-
mal models and additionally to search for differently
substituted titanocenes also derived from fulvenes.
€
[9] N. Kroger, U.R. Kleeberg, K. Mross, L. Edler, G. Saß, D.K.
Hossfeld, Onkologie 23 (2000) 60.
[10] G. Mokdsi, M.M. Harding, Metal-Based Drugs 5 (1998) 207.
[11] O.R. Allen, L. Croll, A.L. Gott, R.J. Knox, P.C. McGowan,
Organometallics 23 (2004) 288.
[12] R.J. Boyles, M.C. Baird, B.G. Campling, N. Jain, J. Inorg.
Biochem. 84 (2001) 159.
[13] R. Teuber, G. Linti, M. Tacke, J. Organomet. Chem. 545–546
(1997) 105.
Acknowledgements
[14] F. Hartl, L. Cuffe, J.P. Dunne, S. Fox, T. Mahabiersing, M.
Tacke, J. Mol. Struct. 559 (2001) 331.
[15] M. Tacke, J.P. Dunne, S. Fox, G. Linti, R. Teuber, J. Mol. Struct.
570 (2001) 197.
The authors thank the Higher Education Authority
(HEA) and the Centre for Synthesis and Chemical Biol-
ogy (CSCB) for funding through the HEA PRTLI cycle 3.
[16] S. Fox, J.P. Dunne, R. Dronskowski, D. Schmitz, M. Tacke, Eur.
J. Inorg. Chem. (2002) 3039.
[17] J.J. Eisch, X. Shi, F.A. Owuor, Organometallics 17 (1998) 5219.
[18] J.J. Eisch, F.A. Owuor, X. Shi, Organometallics 18 (1999) 1583.
[19] K.M. Kane, P.J. Shapiro, A. Vij, R. Cubbon, A.L. Rheingold,
Organometallics 16 (1997) 4571.
References
[20] S. Fox, J.P. Dunne, M. Tacke, J.F. Gallagher, Inorg. Chim. Acta
357 (2004) 225.
[1] A. Gelasco, S.J. Lippard, Top. Biol. Inorg. Chem. 1 (1999) 1.
[2] E.R. Jamieson, S.J. Lippard, Chem. Rev. 99 (1999) 2467.
[3] N. Farrell, Y. Qu, J.D. Roberts, Top. Biol. Inorg. Chem. 1 (1999)
99.
€
[21] G.M. Sheldrick, SADABS, Version 2.03, University of Gottingen,
Germany, 2002.
[22] G.M. Sheldrick, SHELXS 97 and SHELXL 97
, University of
€
€
[4] P. Kopf-Maier, H. Kopf, Chem. Rev. 87 (1987) 1137.
€
Gottingen, Germany, 1997.
€
€
[5] P. Kopf-Maier, H. Kopf, Struct. Bond. 70 (1988) 105.
[6] C.V. Christodoulou, D.R. Ferry, D.W. Fyfe, A. Young, J. Doran,
T.M. Sheehan, A. Eliopoulos, K. Hale, J. Baumgart, G. Saß, D.J.
Kerr, J. Clin. Oncol. 16 (1998) 2761.
[23] Gaussian Inc., Carnegie Office Park, Building 6, Suite 230,
Carnegie, PA 15106, USA.
[24] K.J. Stone, R.D. Little, J. Org. Chem. 49 (1984) 1849.
[25] S.A. Giddings, Inorg. Chem. 6 (1967) 849.
[7] A. Korfel, M.E. Scheulen, H.J. Schmoll, O. Grundel, A.
Harstrick, M. Knoche, M. Fels, M. Skorzec, F. Bach, J.
Baumgart, G. Saß, S. Seeber, E. Thiel, W.E. Berdel, Clin. Cancer
Res. 4 (1998) 2701.
[26] T. Mosmann, J. Immunol. Methods 65 (1983) 55.
[27] C.V. Christodoulou, A.G. Eliopoulos, L.S. Young, L. Hodgkins,
D.R. Ferry, D.J. Kerr, Br. J. Cancer 77 (1998) 2088.