Cytotoxic dinuclear titanium-salan complexes: Structural and biological characterization

Article abstract of DOI:10.1016/j.jinorgbio.2011.08.029

Controlled hydrolysis of donor-substituted titanium-salan complexes led to the formation of well-defined dinuclear complexes. Structure determination by means of X-ray and NMR-studies revealed the presence of a single μ-oxo bridge and one labile alkoxide

Full text of DOI:10.1016/j.jinorgbio.2011.08.029

Journal of Inorganic Biochemistry 106 (2012) 68–75
Contents lists available at SciVerse ScienceDirect
Journal of Inorganic Biochemistry
journal homepage: www.elsevier.com/locate/jinorgbio
Cytotoxic dinuclear titanium-salan complexes: Structural and
biological characterization
⁎
Timo A. Immel, Martin Grützke, Ellen Batroff, Ulrich Groth, Thomas Huhn
Fachbereich Chemie and Konstanz Research School Chemical Biology, Universität Konstanz, Universitätsstraße 10, Fach 720, 78457 Konstanz, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 9 May 2011
Received in revised form 24 August 2011
Accepted 26 August 2011
Available online 14 September 2011
Controlled hydrolysis of donor-substituted titanium-salan complexes led to the formation of well-defined
dinuclear complexes. Structure determination by means of X-ray and NMR-studies revealed the presence
of a single μ-oxo bridge and one labile alkoxide ligand per titanium center. Concomitant cytotoxicity assays
of the isolated dinuclear complexes showed cytotoxicities in the low micro-molar region, surpassing in this
respect even their monomeric ancestors, thus making them possible highly active metabolites of titanium-
salan anti-cancer drugs.
Keywords:
© 2011 Elsevier Inc. All rights reserved.
Dinuclear complexes
Titanium salan complexes
Hydrolysis
Antitumor agents
Cytotoxicity
1. Introduction
Nielson and Waters reported the identification of μ-oxo bridged di-
meric complexes bearing a single alkoxide at each titanium center
Titanium complexes have attracted attention due to their encour-
aging antitumor activity in various cell lines. Today, with derivatives
of titanocene dichloride (Cp2TiCl2) [1–5], diketonato-complexes
such as budotitane [Ti(bzac)2(OEt)2; Hbzac=Phenylbutane-1,3-
dione] [6–8] and titanium salan complexes ([ONNO]-type tetraden-
tate diamine-diphenolato ligands) [9–11], three classes of cytotoxic
titanium complexes are known. Whereas for cisplatin the mechanism
of action is well understood, in the case of titanium complexes the
nature of the active species is still unclear. Is partial hydrolysis of the
labile ligands an activating mechanism, as in the cisplatin case, or
does it lead to deactivation by the formation of non-toxic polynuclear
oxo-titanium species?
[17]. Unfortunately, their solubility was quite limited and in some
cases even to low for NMR-measurements thus disqualifying those
complexes for biological assays.
Herein we report the synthesis and structural characterization of a
highly cytotoxic dinuclear titanium species, the first cytotoxic inter-
mediate from partial hydrolysis of a titanium salan complex. By
using methoxy residues at the aromatic rings, we dramatically en-
hanced the solubility of this class of dimeric complexes. Through
this we could show that this intermediate is a potent cytotoxic
agent itself, even surpassing its parent compound in terms of efficacy
and might play an important role in the metabolism of titanium-salan
complexes.
The hydrolysis of budotitane and titanocene derivatives is very
fast, often in the range of minutes yielding a multitude of metabolites
[12, 13]. Titanium salan complexes are remarkably different in this re-
spect. Dependent on the nature of the phenolate substituents, hydro-
lysis rates were recently shown to vary between several minutes and
more than 120 h, with the size of the labile alkoxy ligand exerting a
strong influence on overall cytotoxicity [14]. However, the attempted
isolation or characterization of a bioactive intermediate from partial
hydrolysis has so far proved unsuccessful; only a trinuclear μ2-oxo
bridged titanium salan complex was described as being nontoxic
[15, 16]. With its lack of any labile ligand, it represents a species,
where hydrolysis led ultimately to complete detoxification. Recently,
2. Experimental
2.1. Materials and methods
Titanium tetra(ethoxide) (99%) and ethylenediamine (99%) were
purchased from ABCR GmbH (Karlsruhe, Germany), deuterated sol-
vents were purchased from euriso-top (Saarbrücken, Germany) and
dried where necessary; other solvents were purified according to
standard procedures [18]. All experiments requiring dry atmosphere
were carried out under nitrogen atmosphere using Schlenk tech-
nique. NMR data were recorded on JEOL Eclipse 400 and Bruker
Avance DRX 600 spectrometers at the given frequencies. The 1H and
13C NMR chemical shifts are reported relative to tetramethylsilane;
the resonance of the residual protons of the solvents served as inter-
nal standard for 1H (δ 7.15 benzene; 7.26 chloroform) and the central
⁎
Corresponding author. Tel.: +49 7531 882283; fax: +49 7531 884424.
E-mail address: thomas.huhn@uni-konstanz.de (T. Huhn).
0162-0134/$ – see front matter © 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jinorgbio.2011.08.029

T.A. Immel et al. / Journal of Inorganic Biochemistry 106 (2012) 68–75
69
signal of the solvent peaks for 13C (δ 128.0 benzene; 77.0 chloroform).
The splitting of proton resonances in the reported 1H NMR spectra is
defined as s = singlet, d = doublet, dd = doublet of doublet, m =
multiplet, dq = doublet of quartet and t = triplet. Structure assign-
ments are based on 2D-NMR (COSY, HMBC, HSQC) experiments. Ele-
mental analyses were performed in the microanalytical laboratory of
the University of Konstanz.
NCH2CH2N), 6.27 (d, 3J=8.4 Hz, 2H, H\Ar), 6.52 (d, 3J=8.4 Hz, 2H,
H\Ar), 7.20 (dd, 3J1 = 3J2 =8.4 Hz, 2H, H\Ar), 8.80 ppm (s, 2H,
N_CH);13C-NMR (101 MHz, CDCl3): δ=55.77 (OCH3), 59.75 (NCH2
CH2N), 99.96 (C\Ar), 108.26 (C\Ar), 110.44 (C\Ar), 133.66
(C\Ar), 159.82 (C\Ar), 162.82 (N_CH), 163.99 ppm (C\Ar); m.p.
128.5–129.5 °C (MeOH, yellow crystals); elemental analysis calcd
(%) for C18H20N2O4: C 65.84, H 6.14, N 8.53; found: C 65.70, H 6.16,
N 8.49.
2.2. Synthesis of ligands H4L1–3 via Schiff bases SB1–3
H4L3 was prepared from SB3 according to the general procedure in a
yield of 95%. 1H-NMR (400 MHz, CDCl3): δ=2.85 (s, 4H, NCH2CH2N),
3.76 (s, 6H, OCH3), 4.08 (s, 4H, NCH2C\Ar), 6.37 (d, 3J=8.2 Hz, 2H,
H\Ar), 6.47 (d, 3J=8.2 Hz, 2H, H\Ar), 7.09 ppm (dd, 3J1 =
3J2 =8.2 Hz, 2H, H\Ar); 13C-NMR (101 MHz, CDCl3): δ=45.09
(NCH2CH2N), 48.15 (OCH3), 55.71 (NCH2C\Ar), 101.73 (C\Ar),
109.90 (C\Ar), 110.07 (C\Ar), 128.84 (C\Ar), 157.86 (C\Ar),
159.71 ppm (C\Ar); m.p. 146 °C (CH2Cl2/MeOH); elemental analysis
calcd (%) for C18H24N2O4: C 65.04, H 7.28, N 8.43; found: C 64.99, H
7.21, N 8.43.
The Schiff-bases SB1–3 were synthesized by stirring of the respec-
tive salicyl aldehydes (2 mmol) in methanol at room temperature and
adding ethylenediamine (1 mmol) [19–22]. After 30 min, the yellow
SB1–3-precipitate was filtered off and suspended in methanol. After
cooling to 0 °C, NaBH4 (8 mmol) was added and the reaction mixture
was stirred for 1 h at room temperature. After addition of water and
extraction with dichloromethane, the organic layer was dried over
MgSO4 and the solvents were evaporated. The resulting ligands
H4L1–3 could be used without further purification.
2.3. Synthesis of ligands H2L1–3
2.2.1. H4L1
SB1 was prepared according to the general procedure in a yield
of 95%. 1H-NMR (400 MHz, CDCl3): δ=3.87 (s, 6H, OCH3), 3.94 (s, 4H,
NCH2CH2N), 6.76 (dd, 3J1=3J2=7.8 Hz, 2H, H\Ar), 6.83 (dd, 3J=
7.8 Hz, 4J=1.4 Hz, 2H, H\Ar), 6.89 (dd, 3J=7.8 Hz, 4J=1.4 Hz, 2H,
H\Ar), 8.31 ppm (s, 2H, N_CH); 13C-NMR (101 MHz, CDCl3):
δ=56.27 (OCH3), 59.69 (NCH2CH2N), 114.29 (C\Ar), 118.26 (C\Ar),
118.61 (C\Ar), 123.31 (C\Ar), 148.49 (C\Ar), 151.63 (C\Ar),
166.87 ppm (N_CH); m.p. 162.5–163.0 °C (MeOH, yellow crystals);
elemental analysis calcd (%) for C18H20N2O4: C 65.84, H 6.14, N 8.53;
found: C 65.66, H 6.28, N 8.48.
Ligands H4L1–3 (10 mmol) were dissolved in acetonitrile/acetic
acid (9: 1, 200 ml). Formaldehyde (37% in H2O, 9 ml) was added
and the reaction mixture was stirred for 1 h at room temperature.
After cooling to 0 °C, NaBH4 (40 mmol) was added in small portions.
The mixture was allowed to warm to room temperature and stirred
for additional 2 h, solvents were evaporated and the remainder sus-
pended in water. After adjusting to pH 6, dichloromethane was
added and the aqueous layer was extracted twice. The combined
organic layers were dried over MgSO4, the solvent was evaporated
and the resulting crude product recrystallized from ethanol. Ligands
H2L1,2 were already synthesized using an alternative approach [23,
24].
H4L1 was prepared from SB1 according to the general procedure
in a yield of 90%. 1H-NMR(400 MHz, CDCl3): δ=2.82 (s, 4H, NCH2
CH2N), 3.87 (s, 6H, OCH3), 3.99 (s, 4H, NCH2C\Ar), 6.62 (d, 3J=
7.7 Hz, 2H, H\Ar), 6.74 (dd, 3J1 = 3J2 =7.7 Hz, 2H, H\Ar), 6.81 ppm
(d, 3J=7.7 Hz, 2H, H\Ar); 13C-NMR (101 MHz, CDCl3): δ=47.95
(NCH2CH2N), 52.12 (OCH3), 56.13 (NCH2C\Ar), 111.13 (C\Ar),
119.08 (C\Ar), 120.92 (C\Ar), 123.03 (C\Ar), 147.09 (C\Ar),
148.14 ppm (C\Ar); m.p. 176 °C (CHCl3/MeOH); elemental analysis
calcd (%) for C18H24N2O4: C 65.04, H 7.28, N 8.43; found: C 65.00, H
7.24, N 8.51.
2.3.1. H2L1
This compound was prepared according to the general procedure
in a yield of 85%. 1H-NMR (400 MHz, CDCl3): δ=2.30 (s, 6H, NCH3),
2.70 (s, 4H, NCH2CH2N), 3.70 (s, 4H, NCH2C\Ar), 3.86 (s, 6H,
OCH3), 6.56 (dd, 3J=7.7 Hz, 4J=1.4 Hz, 2H, H\Ar), 6.72 (dd, 3J1 =
3J2 =7.7 Hz, 2H, H\Ar), 6.80 ppm (dd, 3J=7.7 Hz, 4J=1.4 Hz, 2H,
H\Ar); 13C-NMR (101 MHz, CDCl3): δ=42.09 (NCH3), 54.62 (NCH2
CH2N), 56.10 (NCH2C\Ar), 61.52 (OCH3), 111.42 (C\Ar), 119.01
(C\Ar), 120.81 (C\Ar), 122.04 (C\Ar), 147.21 (C\Ar), 148.13 ppm
(C\Ar); m.p. 115.0–116.0 °C (EtOH, colorless crystals); elemental
analysis calcd (%) for C20H28N2O4: C 66.64, H 7.83, N 7.77; found: C
66.53, H 7.80, N 7.75.
2.2.2. H4L2
SB2 was prepared according to the general procedure in a yield of
95%. 1H-NMR (400 MHz, CDCl3): δ=3.75 (s, 6H, OCH3), 3.94 (s, 4H,
NCH2CH2N), 6.73 (dd, 3J=2.7 Hz, 4J=0.6 Hz, 2H, H\Ar), 6.80–6.95
(m, 4H, H\Ar), 8.31 ppm (s, 2H, N_CH); 13C-NMR (101 MHz,
CDCl3): δ=56.14 (OCH3), 60.05 (NCH2CH2N), 115.16 (C\Ar),
117.89 (C\Ar), 118.46 (C\Ar), 119.75 (C\Ar), 152.25 (C\Ar),
155.36 (C\Ar), 166.47 ppm (N_CH); m.p. 164.5–165.0 °C (MeOH,
yellow crystals); elemental analysis calcd (%) for C18H20N2O4: C
65.84, H 6.14, N 8.53; found: C 65.73, H 6.21, N 8.50.
2.3.2. H2L2
This compound was prepared according to the general procedure
in a yield of 80%. 1H-NMR (400 MHz, CDCl3): δ=2.27 (s, 6H, NCH3),
2.65 (s, 4H, NCH2CH2N), 3.65 (s, 4H, NCH2C\Ar), 3.73 (s, 6H,
OCH3), 6.53 (d, 4J=2.8 Hz, 2H, H\Ar), 6.73 (dd, 3J=8.7 Hz, 4J=
2.8 Hz, 2H, H\Ar), 6.77 ppm (d, 3J=8.7 Hz, 2H, H\Ar); 13C-NMR
(101 MHz, CDCl3): δ=41.93 (NCH3), 54.29 (NCH2CH2N), 55.91
(NCH2C\Ar), 62.02 (OCH3), 113.86 (C\Ar), 114.61 (C\Ar), 116.77
(C\Ar), 122.49 (C\Ar), 151.70 (C\Ar), 152.68 ppm (C\Ar); m.p.
149.0–150.0 °C (EtOH, colorless crystals); elemental analysis calcd
(%) for C20H28N2O4: C 66.64, H 7.83, N 7.77; found: C 66.63, H 7.94,
N 7.72.
H4L2 was prepared from SB2 according to the general procedure
in a yield of 80%. 1H-NMR (400 MHz, CDCl3): δ=2.82 (s, 4H, NCH2
CH2N), 3.74 (s, 6H, OCH3), 3.95 (s, 4H, NCH2C\Ar), 6.27 (dd, 3J=
8.3 Hz, 4J=2.5 Hz, 2H, H\Ar), 6.34 (d, 4J=2.5 Hz, 2H, H\Ar), 6.79
(dd, 3J=8.3 Hz, 2H, H\Ar), ppm; 13C-NMR (101 MHz, CDCl3):
δ=48.09 (NCH2CH2N), 52.92 (OCH3), 56.01 (NCH2C\Ar), 113.96
(C\Ar), 114.60 (C\Ar), 117.00 (C\Ar), 123.03 (C-6), 151.91 (C\Ar),
152.80 ppm
(C\Ar);
m.p.
155.0–155.5 °C
(CHCl3/MeOH);
elemental analysis calcd (%) for C18H24N2O4: C 65.04, H 7.28, N 8.43;
found: C 64.63, H 7.23, N 8.32.
2.3.3. H2L3
This compound was prepared according to the general procedure
in a yield of 28%. The product was purified by flash chromatography
on silica gel using ethyl acetate with an increasing gradient of etha-
nol. Recrystallization from ethanol gave analytical pure samples.
2.2.3. H4L3
SB3 was prepared according to the general procedure in a yield of
69%. 1H-NMR (400 MHz, CDCl3): δ=3.79 (s, 6H, OCH3), 3.89 (s, 4H,

70
T.A. Immel et al. / Journal of Inorganic Biochemistry 106 (2012) 68–75
1H-NMR (400 MHz, CDCl3): δ=2.30 (s, 6H, NCH3), 2.68 (s, 4H,
2H, NCH2CH2N), 1.28 (t, 3J=7.0 Hz, 6H, OCH2CH3), 2.27 (s, 6H,
NCH3), 2.72 (d, 2J=9.3 Hz, 2H,NCH2CH2N), 3.41 (s, 6H, OCH3), 3.84
(d, 2J=14.1 Hz, 2H, NCH2C\Ar), 4.25 (d, 2J=14.1 Hz, 2H,
NCH2C\Ar), 4.63 (dq, 3J=7.0 Hz, 2J=10.8 Hz, 4H,OCH2CH3), 4.66 (dq,
3J=7.0 Hz, 2J=10.8 Hz, 4H,OCH2CH3), 6.20 (d, 3J=8.2 Hz, 2H, H\Ar),
6.61 (d, 3J=8.2 Hz, H\Ar), 7.11 ppm (dd, 3J1=3J2=8.2 Hz, 2H,
H\Ar); 13C-NMR (101 MHz, C6D6): δ=20.03 (OCH2CH3), 47.84
(NCH3), 52.07 (NCH2CH2N), 55.50 (OCH3), 56.81 (NCH2C\Ar), 71.84
(OCH2CH3), 101.04 (C\Ar), 112.00 (C\Ar), 112.93 (C\Ar), 129.00
(C\Ar), 159.03 (C\Ar), 163.79 ppm (C\Ar); UV/Vis (CHCl3): λmax
(ε)=319 nm (32,048 M−1 cm−1); IR (ATR): v=3064.81 (w), 2966.00
(m), 2924.06 (m), 2859.21 (s), 2839.40 (s), 1896.43 (w), 1793.23 (w),
1591.15 (s), 1572.40 (s), 1460.87 (s), 1420.15 (m), 1371.96 (m),
1351.83 (w), 1298.60 (s), 1275.98 (m), 1242.72 (s), 1196.83 (w),
1089.45 (s), 1006.16 (m), 965.89 (m), 935.97 (m), 909.88 (s), 844.94
(w), 752.97 cm−1 (s); m.p. 163.0–164.0 °C (EtOH, yellow prisms); ele-
mental analysis calcd (%) for C24H36N2O6Ti: C 58.07, H 7.31, N 5.64;
found: C 58.03, H 7.57, N 5.74.
NCH2CH2N), 3.77 (s, 6H, OCH3), 3.80 (s, 4H, NCH2C\Ar), 6.36 (dd,
3J=8.3 Hz, 4J=0.7 Hz, 2H, H\Ar), 6.47 (dd, 3J=8.3 Hz, 4J=0.7 Hz,
2H, H\Ar), 7.09 ppm (dd, 3J1 = 3J2 =8.3 Hz, 2H, H\Ar); 13C-NMR
(101 MHz, CDCl3): δ=42.11 (NCH3), 54.41 (NCH2CH2N, NCH2C\Ar),
55.68 (OCH3), 101.62 (C\Ar), 109.46 (C\Ar), 109.69 (C\Ar), 128.81
(C\Ar), 157.98(C\Ar), 159.60 ppm (C\Ar); m.p. 124.0–125.5 °C
(EtOH, colorless crystals); elemental analysis calcd (%) for
C
20H28N2O4: C 66.64, H 7.83, N 7.77; found: C 66.58, H 7.82, N 7.79.
2.4. Synthesis of mononuclear complexes [TiL1–3(OEt)2]
Ligands H2L1–3 (1.5 mmol) were dissolved in toluene (10 ml) and
titanium ethoxide (1.5 mmol) was added over 10 min under a nitro-
gen atmosphere [14]. The yellow reaction mixture was allowed to
stir overnight at room temperature. After removal of the solvent
under reduced pressure, the complex was obtained as yellow solid
in nearly quantitative yield. Recrystallization from the given solvent
gave analytical pure samples.
2.5. Synthesis of dinuclear complexes [L1,3(OEt)Ti\O\Ti(OEt)L1,3
]
2.4.1. [TiL1(OEt)2]
This compound was prepared according to the general procedure
in a yield of 68%. 1H-NMR (600 MHz, CDCl3): δ=1.28 (t, 3J=7 Hz,
6H, OCH2CH3), 1.81 (d, 2J=9.4 Hz, 2H, NCH2CH2N), 2.45 (s, 6H,
NCH3), 2.98 (d, 2J=9.4 Hz, 2H, NCH2CH2N), 3.14 (d, 2J=13.5 Hz,
2H, NCH2C\Ar), 3.85 (s, 6H, OCH3), 4.59 (d, 2J=13.5 Hz, 2H,
NCH2C\Ar), 4.61 (dq, 3J=7 Hz, 2J=10.8 Hz, 2H, OCH2), 4.72 (dq,
3J=7 Hz, 2J=10.8 Hz, 2H, OCH2), 6.60 (dd, 3J=7.7 Hz, 4J=1.3 Hz,
2H, H\Ar), 6.65 (dd, 3J1 = 3J2 =7.7 Hz, 2H, H\Ar), 6.83 ppm (dd,
3J=7.7 Hz, 4J=1.3 Hz, 2H, H\Ar); 13C-NMR (151 MHz, CDCl3):
δ=19.39 (OCH2CH3), 47.33 (NCH3), 52.11 (NCH2CH2N), 56.72
(C\Ar), 64.35 (NCH2C\Ar), 71.74 (OCH2CH3), 113.13 (C\Ar),
117.69 (C\Ar), 122.14 (C\Ar), 125.50 (C\Ar), 148.58 (C\Ar),
152.01 ppm (C\Ar); UV/Vis (CHCl3): λmax (ε)=332 nm
(8859 M−1 cm−1); IR (ATR (Attenuated Total Reflectance)):
v=3058.52 (w (weak)), 3012.47 (w), 2966.09 (w), 2832.14 (w),
1572.30 (m (medium)), 1483.38 (m), 1371.35 (m), 1301.80 (m),
1243.56 (s (strong)), 1083.33 (s), 1056.49 (s), 1005.67 (m), 905.36
(m), 865.69 (s), 811.18 (m), 765.81 (m) 723.43 cm−1 (s); m.p.
148.0–148.5 °C (EtOH, yellow prisms); elemental analysis calcd (%)
for C24H36N2O6Ti: C 58.07, H 7.31, N 5.64; found: C 58.05, H 7.38, N
5.65.
Complexes [TiL1,3(OEt)2] were suspended in a mixture of ethanol/
water (95:5). The suspension was heated to 60 °C and a mixture of
ethanol/water as before was added dropwise until the complexes
had dissolved. The reaction mixture was kept for 2 days at room tem-
perature; during that time the product crystallized as very thin yellow
platelets which were filtered off and washed with cold ethanol.
2.5.1. [L1(OEt)Ti\O\Ti(OEt)L1]
This compound was prepared according to the general procedure
in a yield of 80%. 1H-NMR (600 MHz, C6D6): δ=1.04 (d, 2J=9.9 Hz,
2H, NCH2CH2N), 1.17 (t, 3J=7.0 Hz, 6H, OCH2CH3), 1.19 (d, 2J=
9.9 Hz, 2H, NCH2CH2N), 2.17 (s, 6H, NCH3), 2.71 (d, 2J=13.4 Hz, 2H,
NCH2C\Ar), 2.74–2.90 (m, 4H, NCH2CH2N), 2.85 (s, 6H, NCH3), 3.01
(d, 2J=14.1 Hz, 2H, NCH2C\Ar), 3.55 (s, 6H, OCH3), 3.65 (s, 6H,
OCH3), 4.56 (dq, 2J=11.1 Hz, 3J=7.0 Hz, 2H, OCH2CH3), 4.62 (d, 2J=
13.4 Hz, 2H, NCH2C\Ar), 4.88 (dq, 2J=11.1 Hz, 3J=7.0 Hz, 2H,
OCH2CH3), 5.83 (d, 2J=14.1 Hz, 2H, NCH2C\Ar), 6.55 (dd, 3J=7.7 Hz,
4J=1.8 Hz, 2H, H\Ar), 6.58 (dd, 3J=7.7 Hz, 4J=1.8 Hz, 2H, H\Ar),
6.66 (t, 3J=7.7 Hz, 2H, H\Ar), 6.75 (t, 3J=7.7 Hz, 2H, H\Ar), 6.80
(dd, 3J=7.7 Hz, 4J=1.8 Hz, 2H, H\Ar), 6.82 ppm (dd, 3J=7.7 Hz,
4J=1.8 Hz, 2H, H\Ar); 13C-NMR (151 MHz, C6D6): δ=19.57
(OCH2CH3), 47.52 (NCH3), 47.84 (NCH3), 51.68 (NCH2CH2N), 52.88
(NCH2CH2N), 55.73 (OCH3), 57.25 (OCH3), 64.97 (NCH2C\Ar), 65.13
(NCH2C\Ar), 72.23 (OCH2CH3), 112.31 (C\Ar), 114.18 (C\Ar),
117.26 (C\Ar), 117.36 (C\Ar), 122.51 (C\Ar), 123.45 (C\Ar),
126.26 (C\Ar), 127.29 (C\Ar), 149.17 (C\Ar), 149.64 (C\Ar),
153.17 (C\Ar), 153.94 ppm (C\Ar); UV/Vis (CHCl3): λmax (ε)=343
(11,069), 240 nm (19,606 M−1 cm−1); IR (ATR): v=2976.21 (s),
2849.26 (s), 2282.61 (w), 2050.20 (w), 1980.69 (w), 1593.26 (s),
1573.92 (s), 1462.45 (m), 1373.04 (m), 1299.24 (s), 1242.31 (s),
1144.58 (m), 1088.57 (s), 1008.03 (s), 908.74 (m), 751.65 (s),
706.64 cm−1 (s); m.p. 210 °C (EtOH, yellow rods); elemental analysis
calcd (%) for C44H62N4O11Ti: C 57.52, H 6.80, N 6.10; found: C 57.54, H
6.73, N 6.08.
2.4.2. TiL2(OEt)2
This compound was prepared according to the general procedure
in a yield of 51%. 1H-NMR (400 MHz, CDCl3): δ=1.25 (t, 3J=7 Hz,
6H, OCH2CH3), 1.80 (d, 2J=9.3 Hz, 2H, NCH2CH2N), 2.46 (s, 6H,
NCH3), 2.99 (d, 2J=9.3 Hz, 2H, NCH2CH2N), 3.08 (d, 2J=13.6 Hz,
2H, NCH2C\Ar), 3.74 (s, 6H, OCH3), 4.48–4.60 (m, 6H, H-8,
OCH2CH3), 6.54 (d, 4J=3.0 Hz, 2H, H\Ar), 6.67 (d, 3J=8.8 Hz, 2H,
H\Ar), 6.74 ppm (dd, 4J=3.0 Hz, 3J=8.8 Hz, 2H, H\Ar); 13C-NMR
(101 MHz, CDCl3): δ=19.52 (OCH2CH3), 47.40 (NCH3), 52.10
(NCH2CH2N), 56.00 (OCH3), 64.58 (NCH2C\Ar), 71.54 (OCH2CH3),
114.26 (C\Ar), 115.37 (C\Ar), 118.07 (C\Ar), 125.23 (C\Ar),
151.90 (C\Ar), 155.88 ppm (C\Ar); UV/Vis (CHCl3): λmax (ε)=
312 nm (12,139 M−1 cm−1); IR (ATR): v=3013.72 (w), 2964.96
(w), 2899.87 (w), 2831.35 (w), 1486.31 (s), 1413.60 (m), 1369.74
(w), 1316.46 (w), 1258.79 (s), 1223.05 (s), 1148.95 (m), 1110.87
(m), 1045.56 (s), 1004.86 (s), 904.67 (s), 859.64 (s), 824.68 (s),
799.21 (s), 752.71 (s), 669.66 cm−1 (m); m.p. 160.0–161.0 °C
(EtOH, yellow prisms); elemental analysis calcd (%) for
2.5.2. [L3(OEt)Ti\O\Ti(OEt)L3]
This compound was prepared according to the general procedure
in a yield of 85%. 1H-NMR (400 MHz, C6D6): δ=1.04 (d, 2J=9.7 Hz,
2H, NCH2CH2N), 1.13 (t, 3J=7.0 Hz, 6H, OCH2CH3), 1.15 (d, 2J=
9.7 Hz, 2H, NCH2CH2N), 2.17 (s, 6H, NCH3), 2.67–2.85 (m, 4H, NCH2
CH2N), 2.88 (s, 6H, NCH3), 3.32 (s, 6H, OCH3), 3.39 (s, 6H, OCH3),
3.77 (d, 2J=14.0 Hz, 2H, NCH2C\Ar), 4.00 (d, 2J=14.4 Hz, 2H,
NCH2C\Ar), 4.22 (d, 2J=14.0 Hz, 2H, NCH2C\Ar), 4.55 (dq, 2J=
11.0 Hz, 3J=7.0 Hz, 2H, OCH2CH3), 4.62 (dq, 2J=11.0 Hz, 3J=
7.0 Hz, 2H, OCH2CH3), 5.31 (d, 2J =14.4 Hz, 2H, NCH2C\Ar), 6.17
C24H36N2O6Ti: C 58.07, H 7.31, N 5.64; found: C 57.94, H 7.16, N 5.68.
2.4.3. TiL3(OEt)2
This compound was prepared according to the general procedure
in a yield of 42%. 1H-NMR (400 MHz, C6D6): δ=1.11 (d, 2J=9.3 Hz,

T.A. Immel et al. / Journal of Inorganic Biochemistry 106 (2012) 68–75
71
(d, 3J=7.8 Hz, 2H, H\Ar), 6.24 (d, 3J=7.8 Hz, 2H, H\Ar), 6.52 (d,
3J=7.8 Hz, 2H, H\Ar), 6.73 (d, 3J=7.8 Hz, 2H, H\Ar), 7.10 (dd,
3J1=3J2=7.8 Hz, 2H, H\Ar), 7.22 ppm (dd, 3J1=3J2=7.8 Hz, 2H,
H\Ar); 13C-NMR (101 MHz, C6D6): δ=19.44 (OCH2CH3), 47.88
(NCH3), 47.96 (NCH3), 51.78 (NCH2CH2N), 52.64 (NCH2CH2N), 55.42
(OCH3), 55.50 (OCH3), 56.37 (NCH2C\Ar), 56.81 (NCH2C\Ar), 72.40
(OCH2CH3), 100.86 (C\Ar), 100.87 (C\Ar), 111.34 (C\Ar), 111.76
(C\Ar), 113.58 (C\Ar), 114.22 (C\Ar), 128.31 (C\Ar), 128.87
(C\Ar), 159.11 (C\Ar), 159.14 (C\Ar), 164.11 (C\Ar), 164.18 ppm
(C\Ar); UV/Vis (CHCl3): λmax (ε)=316 (23,504 M−1 cm−1); IR
(ATR): v=3013.47 (w), 2855.37 (w), 1589.70 (m), 1573.66 (m),
1462.85 (s), 1376.22 (w), 1301.44 (m), 1242.82 (s), 1088.65 (s),
1145.00 (w), 1056.59 (s), 1007.67 (m), 964.41 (m), 903.85 (m),
752.53 (s), 718.35 (s), 695.70 cm−1 (s); m.p. 240.0–242.0 °C (EtOH/
toluene, yellow platelets); elemental analysis calcd (%) for
[IN2σ(I)]=0.0548, wR2 for all=0.1123. Single crystals of [TiL2(OEt)2]
were grown from solutions in ethanol by slow evaporation at 6 °C.
2.6.2. Crystal data for [L1(OEt)Ti-O-Ti(OEt)L1]
C
44H62N4O11Ti, M=918.72, monoclinic, C 2/c, a=44.5019(18),
b=11.0477(5), c=19.3169(9) Å, β =113.729(3)°, V=8694.2(6) Å3,
T=100(2) K, Z=8, ρcalcd. =1.404 g cm−3, μ(MoKα)=0.433 mm−1
,
63,223 reflections collected, 9270 unique (Rint=0.0756), R1 for
[IN2σ(I)]=0.0579, wR2 for all=0.1442. Single crystals of [L1(OEt)Ti-
O-Ti(OEt)L1] were grown from a saturated solution of the complex in
ethanol at −20 °C.
2.7. Time resolved hydrolysis studies using 1H-NMR-spectroscopy
Hydrolysis experiments followed by 1H-NMR spectroscopy were
conducted in a mixture of 95% [D8]THF and 4.8% D2O and 0.2%
DMSO at 37 °C. Spectra were recorded at regular intervals. Data
analysis was achieved by monitoring the decrease of at least two
well-isolated signals of the titanium bound salan backbone and the
increase of the evolving signals of unbound ethoxide over time.
Resulting integrals were normalized against the internal standard
(DMSO) and plotted against elapsed time. Control measurements
without DMSO showed no significant alteration of hydrolysis rate
and products formed. Plotted data for the hydrolysis of mono- and
binuclear complexes [TiL1(OEt)2], [TiL3(OEt)2], [L1(OEt)Ti\O\Ti
(OEt)L1] and [L3(OEt)Ti\O\Ti(OEt)L3] can be found as figures S1
and S2 in the electronic supplements to this manuscript. Table 4
summarizes the calculated t1/2-values of all complexes.
C44H62N4O11: C 57.52, H 6.80, N 6.10; found: C 57.36, H 6.71, N 6.05.
2.6. X-ray crystallographic studies
Data collection was performed with a STOE IPDS-II diffractometer
equipped with
a
graphite monochromated radiation source
(λ=0.71073 Å), an image plate detection system and an Oxford
Cryostream 700 with nitrogen as coolant gas. The selection, integra-
tion, and averaging procedure of the measured reflex intensities, the
determination of the unit cell by a least-squares fit of the 2Θ values,
data reduction, LP correction, and the space group determination
were performed using the X-Area software package delivered with
the diffractometer [25].
A semi-empirical absorption correction
method was used after indexing of the crystal faces. Structures were
solved by direct methods either with SHELXS-97 [26] ([TiL2(OEt)2])
or SIR-97 [27] ([L1(OEt)Ti\O\Ti(OEt)L1]) and refined by standard
Fourier techniques against F2 with a full-matrix leastsquares algo-
rithm using SHELXL-97 [26] and the WinGX (1.80.05) [28] software
package. All non-hydrogen atoms were refined anisotropically. Hy-
drogen atoms were placed in calculated positions and refined with
2.8. Cytotoxicity assay
Cytotoxicity was estimated in cells of human HeLa S3 cervix
carcinoma and Hep G2 liver carcinoma cells obtained from European
Collection of Cell Cultures (ECACC) using an AlamarBlue based assay
[31, 32]. AlamarBlue was purchased from BioSource Europe.
Cells were cultivated at 37 °C in humidified 5% CO2 atmosphere
using Dulbecco's DMEM-medium (Invitrogen) containing 10% fetal
calf serum, 1% penicillin and 1% streptomycin. Cells were split every
three days. Both cell lines were tested for mycoplasma infections
using a mycoplasma detection kit (Roche Applied Science).
AlamarBlue, the dark blue colored sodium salt of resazurin (7-
hydroxy-3H-phenoxazin-3-one-10-oxide) was used to measure
growth and viability of cells which are capable of reducing it to the
fluorescent, pink colored resorufin (7-hydroxy-3H-phenoxazin-3-
one). Cells were seeded in 96-well plates (4000 HeLa S3 cells/well
a
riding model. Graphical representations were prepared with
ORTEP-III [29]. The program PLATON [30] was used to check the
results of the X-ray crystal structure determination.
2.6.1. Crystal data for [TiL2(OEt)2]
C
24H36N2O6Ti, M=496.45, monoclinic, P 21/c, a=12.4419(11),
b=12.6233(7), c=19.0955(17) Å, β=125.596(6)°, V=2438.7(3) Å3,
T=100(2) K, Z=4, ρcalcd=1.352 g cm−3, μ(MoKα)=0.392 mm−1
,
31,005 reflections collected, 4806 unique (Rint=0.0870), R1 for
R1
R3
R1
R1
NH2
OH
N
HO
N
OH
H2N
NaBH4
MeOH
R2
R2
R2
MeOH
R3
R3
O
SB1-3
R1
R1
R1
R1
1. H2COaq.
2. NaBH4
OH
HO
OH
HO
R2
R2
AcOH, MeCN
R2
R2
R3 HN
NH R3
H4L1-3
R3
N
N
R3
H2L1-3
Scheme 1. Synthesis of donor substituted salans H2L1–3 (L1: R1 =OMe, R2 =R3 =H; L2: R1 =H, R2 =OMe, R3 =H; L3: R1 =R2 =H, R3 =OMe).

72
T.A. Immel et al. / Journal of Inorganic Biochemistry 106 (2012) 68–75
R1
R2
R3
R1
R3
R1
O
OH
N
HO
Ti(OEt)4
toluene
N
N
Ti
O
R2
R2
EtO
OEt
R3
R3
R2
N
R1
H2L1-3
[TiL1(OEt)2]: R1 = OMe, R2 = R3 = H
[TiL2(OEt)2]: R1 = H, R2 = OMe, R3 = H
[TiL3(OEt)2]: R1 = R2 = H, R3 = OMe
Scheme 2. Synthesis of [TiL1–3(OEt)2].
or 8000 Hep G2 cells/well) and allowed to attach and grow for 24 h.
Complexes to be tested were dissolved in a suitable amount of
DMSO. Different concentrations were prepared by serial dilution
with medium to give final concentrations with a maximum DMSO
content of 1%. The cells were then incubated for 48 h with 100 μl
each of above dilution series. AlamarBlue (10 μl) was added and the
cells were incubated for another hour. After excitation at 530 nm,
fluorescence at 590 nm was measured using a Synergy 2 HT Fluores-
cence Microplate Reader (BioTek). Cell viability is expressed as a per-
centage with respect to a control containing only pure medium and
1% DMSO incubated under identical conditions. All experiments
were repeated for a minimum of three times with each experiment
done in four replicates. The resulting curves were fitted using Sigma
plot 10.0 [33]. Viability charts (Figures S3–S7) of mono- and dinuclear
complexes [TiL1(OEt)2], [TiL2(OEt)2], [TiL3(OEt)2], [L1(OEt)Ti\O\Ti
(OEt)L1], and [L3(OEt)Ti\O\Ti(OEt)L3] after 48 h of incubation in
HeLa S3 and Hep G2 cells respectively can be found in the electronic
supplements to this manuscript.
3. Results and discussion
The synthesis of the methoxy salans H2L1–3 was achieved by a se-
quence of two subsequent reductive aminations (Scheme 1).
Starting from ethylenediamine and the respective methoxy salicy-
laldehyde gave ligand precursors H4L1–3 in good yields [34]. Reduc-
tive methylation of H4L1,2 proceeded smoothly to yield H2L1,2
whereas the formation of a red polymeric byproduct decreased the
yield of H2L3 because of heavy cross-linking via both activated o-
and p-positions. Metalation of the three salans H2L1–3with titanium
(IV) ethoxide in toluene at room temperature led to the formation
of [TiL1–3(OEt)2] as single (racemic) isomers in nearly quantitative
yield (Scheme 2) [35].
1H-NMR spectra of [TiL1–3(OEt)2] showed the familiar AB pattern
of the benzylic protons consistent with C2 symmetry and a fac-fac
wrapping of the salan around the titanium center. Single crystals of
[TiL2(OEt)2] suitable for X-ray crystal structure determination were
grown from ethanol.
[TiL2(OEt)2] crystallizes in the monoclinic system in the centrosym-
metric space group P21/c, with one molecule in the asymmetric unit and
no additional solvent. The crystal structure confirmed the C2 symmetry
with the labile ethoxy ligands bound in a cis fashion at the equatorial
plane and the phenolates occupying the bis-trans-axial positions of
the slightly distorted octahedral complex (Fig. 1). With 1.82–1.83 Å
for the titanium-alkoxide and 1.90 Å for the titanium-phenolate
distances, bond-lengths around the titanium center are well compara-
ble with the methyl substituted members of this class of complexes
[14]. The electron richness usually attributed to methoxy-substituted
Table 1
IC50 [μM] values in Hela S3 and Hep G2 cells estimated by AlamarBlue assay after 48 h
incubation with [TiL1–3(OEt)2] or ligands H2L1–3 respectively. All IC50 values given in
μM are means from at least three independent experiments each done in four
replicates.
Ligand/complex
IC50 in HeLa S3
IC50 in Hep G2
[TiL1(OEt)2]
[TiL2(OEt)2]
[TiL3(OEt)2]
H2L1
6.2 0.5
4.0 0.6
6.2 0.4
46.9 15.6
N100
13.0 1.7
5.4 0.8
7.6 3.4
48.6 13.1
N100
H2L2
H2L3
37.6 8.4
1.2 0.4
69.5 11.2
3.0 1.3
Cisplatina
Fig. 1. ORTEP diagram of the molecular structures of [TiL2(OEt)2]. Displacement ellip-
soids are drawn at the 50% probability level. Hydrogen atoms are omitted for clarity.
a
Cisplatin served as the reference compound in all assays.

T.A. Immel et al. / Journal of Inorganic Biochemistry 106 (2012) 68–75
73
R2
R2
R1
R2
N
Ti
O
O
N
R1
O
N
Ti
O
O
up to 1000 eq. H2O
EtOH
O
O
N
R1
O
R2
O
Ti
N
N
O
R2
R1
R1
O
R2
R1
[TiL1(OEt)2]: R1 = OMe, R2 = H
[TiL3(OEt)2]: R1 = H, R2 = OMe
[L1(OEt)Ti-O-Ti(OEt)L1]: R1 = OMe, R2 = H, 80%
[L3(OEt)Ti-O-Ti(OEt)L3]: R1 = H, R2 = OMe, 85 %
Scheme 3. Partial hydrolysis resulting in the exclusive formation of dinuclear [L1,3(OEt)Ti\O\Ti(OEt)L1,3].
arenes is not apparent in the bonding parameters of the complexes
[TiL1–3(OEt)2].
new compounds were isolated as microcrystalline material in ≥80%
yield. Signals from the salan ligands appeared doubled in the 1H-
NMR but no liberation of either free salans H2L1 or H2L3 was ob-
served. Most strikingly, the ratio of labile to salan ligand had changed.
From the integral ratio, it seemed that each new compound had lost
one of its labile ethoxy groups. Hence, it was anticipated that a dinuc-
lear species similar to the complexes described by Nielson and Waters
[17] with one bridging oxo-ligand in place of the former labile ligand
had formed by partial hydrolysis, thus, giving reason for the symme-
try reflected by the NMR spectra (Scheme 3).
Attempts to grow single crystals suitable for X-ray crystal struc-
ture determination proved difficult because of both compounds' pro-
nounced tendency to form very thin platelets. Finally, after keeping a
solution of [L1(OEt)Ti\O\Ti(OEt)L1] for several weeks at freezer
temperature, suitable crystals began to separate from the solution.
The structure solved showed nearly perfect C2-symmetry around
the μ2-oxo bridge (O5, Fig. 2) as center of symmetry.
Both hemispheres of the dinuclear complex still feature the lightly
distorted octahedral geometry known from the mononuclear salan
complexes. That is, the phenolates (O1\Ti1\O2 and O4\Ti2\O3)
are oriented in a bis-trans-axial fashion while the amino and oxo li-
gands are bound pair-wise in cis-fashion at the equatorial plane.
Bond length and angles show very little variation when comparing
the mononuclear [TiL2(OEt)2] with the dinuclear [L1(OEt)Ti\O\Ti
(OEt)L1] (Selected bond length and angles of both new complexes
are tabulated in Tables 2 and 3).
To estimate the cytotoxicity of these donor substituted complexes,
their efficacy was screened in two human tumor cell lines (Hela S3
and Hep G2) using the AlamarBlue assay [31, 32]. Cisplatin served
as the reference compound in all assays. To verify that the observed
cytotoxicity does not originate from the free ligands those were also
tested. Data are summarized in Table 1.
All three complexes showed cytotoxicity in a low μ-molar range
and therefore belong to the group of highly bioactive titanium com-
plexes, whereas the ligands H2L1–3 exhibited only limited toxicity.
The formation of dinuclear complexes [L1,3(OEt)Ti\O\Ti(OEt)
L1,3] was achieved by suspending [TiL1,3(OEt)2] in an ethanol/water
mixture (95:5) at 60 °C. To this additional solvent was slowly added
until all starting material had dissolved. Intriguingly, 1H-NMR spectra
recorded from the crude reaction mixtures revealed the very rapid
formation of an apparently highly symmetric single new compound
in both cases. The reaction proved surprisingly tolerant against the
amount of water being added, similar spectra were recorded when
the amount of water varied between 10 and 1000 equivalents, the
The half-life of mononuclear and dinuclear complexes under hy-
drolytic conditions at 37 °C was estimated routinely by NMR spec-
troscopy in a mixture consisting of 95% [D8]THF, 4.8% D2O and 0.2%
DMSO as internal standard [14]. All methoxy salan complexes studied
showed a remarkable sensitivity towards hydrolysis. Compared with
methyl-substituted salan complexes with half-lives of several hours
or halogen-substituted ones with half-lives of more than 120 h [14],
[TiL1,2(OEt)2] and the respective dinuclear complexes [L1,3(OEt)
Table 2
Selected bond lengths [Å] and angles [°] for [TiL2(OEt)2].
[TiL2(OEt)2]
Ti(1)\O(5)
Ti(1)\O(6)
Ti(1)\O(1)
Ti(1)\O(3)
Ti(1)\N(2)
Ti(1)\N(1)
O(5)\Ti(1)\O(6)
O(5)\Ti(1)\O(1)
O(6)\Ti(1)\O(1)
O(5)\Ti(1)\O(3)
O(6)\Ti(1)\O(3)
1.817(2)
1.834(2)
1.8988(18)
1.9014(19)
2.314(2)
2.322(2)
105.40(10)
92.75(9)
95.47(9)
95.97(9)
91.42(8)
O(1)\Ti(1)\O(3)
O(5)\Ti(1)\N(2)
O(6)\Ti(1)\N(2)
O(1)\Ti(1)\N(2)
O(3)\Ti(1)\N(2)
O(5)\Ti(1)\N(1)
O(6)\Ti(1)\N(1)
O(1)\Ti(1)\N(1)
O(3)\Ti(1)\N(1)
N(2)\Ti(1)\N(1)
167.06(9)
88.03(9)
165.43(9)
89.39(8)
81.39(8)
163.28(9)
90.97(9)
82.13(8)
86.83(8)
76.07(8)
Fig. 2. ORTEP diagram of the molecular structures of [L1(OEt)Ti\O\Ti(OEt)L1] with
the two remaining labile ligands (O11,O10) pointing towards the front. Displacement
ellipsoids are drawn at the 50% probability level. Hydrogen atoms are omitted for
clarity.

74
T.A. Immel et al. / Journal of Inorganic Biochemistry 106 (2012) 68–75
Table 3
Table 5
Selected bond lengths [Å] and angles [°] for [L1(OEt)Ti\O\Ti(OEt)L1].
[L1(OEt)Ti\O\Ti(OEt)L1]
IC50 [μM] values in Hela S3 and Hep G2 cells after 48 h incubation with complexes [L1,3
(OEt)Ti\O\Ti(OEt)L1,3] estimated by AlamarBlue assay. All IC50 values given in μM are
means from at least three independent experiments each done in four replicates.
N(1)\Ti(1)
N(2)\Ti(1)
N(3)\Ti(2)
N(4)\Ti(2)
O(1)\Ti(1)
O(2)\Ti(1)
O(3)\Ti(2)
O(4)\Ti(2)
O(5)\Ti(2)
O(5)\Ti(1)
O(10)\Ti(1)
O(11)\Ti(2)
Ti(2)\O(5)\Ti(1)
O(10)\Ti(1)\O(5)
O(10)\Ti(1)\O(2)
O(5)\Ti(1)\O(2)
O(10)\Ti(1)\O(1)
O(5)\Ti(1)\O(1)
O(2)\Ti(1)\O(1)
O(10)\Ti(1)\N(1)
O(5)\Ti(1)\N(1)
O(2)\Ti(1)\N(1)
2.342(3)
2.354(3)
2.346(3)
2.343(3)
1.922(2)
1.898(2)
1.900(2)
1.952(2)
1.815(2)
1.822(2)
1.816(2)
1.812(2)
169.15(14)
106.58(11)
90.94(11)
96.29(10)
94.98(11)
94.57(10)
165.62(11)
164.87(11)
86.71(10)
80.22(10)
O(1)\Ti(1)\N(1)
O(10)\Ti(1)\N(2)
O(5)\Ti(1)\N(2)
O(2)\Ti(1)\N(2)
O(1)\Ti(1)\N(2)
N(1)\Ti(1)\N(2)
O(11)\Ti(2)\O(5)
O(11)\Ti(2)\O(3)
O(5)\Ti(2)\O(3)
O(11)\Ti(2)\O(4)
O(5)\Ti(2)\O(4)
O(3)\Ti(2)\O(4)
O(11)\Ti(2)\N(4)
O(5)\Ti(2)\N(4)
O(3)\Ti(2)\N(4)
O(4)\Ti(2)\N(4)
O(11)\Ti(2)\N(3)
O(5)\Ti(2)\N(3)
O(3)\Ti(2)\N(3)
O(4)\Ti(2)\N(3)
N(4)\Ti(2)\N(3)
91.03(10)
91.92(11)
161.38(11)
85.29(10)
81.43(10)
75.24(10)
106.74(11)
93.90(11)
96.73(11)
95.09(11)
93.86(10)
163.53(11)
165.34(10)
87.18(10)
79.60(10)
88.39(10)
90.84(11)
162.19(10)
84.63(11)
81.47(10)
75.56(10)
Complex
IC50 in HeLa S3
IC50 in Hep G2
[L1(OEt)Ti\O\Ti(OEt)L1]
[L3(OEt)Ti\O\Ti(OEt)L3]
Cisplatina
3.0 0.9
19.3 2.0
1.2 0.4
5.0 0.4
20.8 2.3
3.0 1.3
a
Cisplatin served as the reference compound in all assays.
certainly doubtful. However, the characterized dinuclear complexes
represent the first of their kind tested in biological assays and it is
yet not clear if other salan complexes behave in a similar manner.
Moreover, further research efforts are required to answer the ques-
tion if such dinuclear complexes might form under biological
conditions.
In summary, herein we report the synthesis of mononuclear,
methoxy substituted titanium salan complexes. Their controlled hy-
drolysis afforded μ-oxo bridged dinuclear complexes in good yields
still bearing one labile alkoxy ligand at each metal center. These par-
tially hydrolyzed dinuclear complexes are the first of their kind to dis-
play a high degree of cytotoxicity when tested in two different human
cancer cell lines.
Ti\O\Ti(OEt)L1,3] showed a drastic accelerated speed of hydrolysis
and the formation of higher aggregates with no liberation of salan.
Both mononuclear complexes show a t1/2 of less than 1 h. Surprising-
ly, the dinuclear complexes hydrolyze with comparable speed, with
[L3(OEt)Ti\O\Ti(OEt)L3] being slightly more stable than [L1(OEt)
Ti\O\Ti(OEt)L1]. Table 4 summarizes the results.
Both dinuclear complexes were screened for their bioactivity in
the AlamarBlue assay as the mononuclear compounds had been be-
fore. Knowing from former studies [14] that cytotoxicity is extremely
diminished when complex size is increased, we were quite impressed
that [L3(OEt)Ti\O\Ti(OEt)L3] showed only a threefold decreased ac-
tivity and [L1(OEt)Ti\O\Ti(OEt)L1] had an even twofold increased
activity compared to their respective mononuclear starting com-
pounds (Table 5). The bioactivity of both dinuclear complexes corrob-
orates our hypothesis that the presence of labile ligands is a
prerequisite for cytotoxicity.
Abbreviations
COSY
Cp
DMEM Dulbecco's Modified Eagle Medium
ECACC
Hbzac
Hela S3 human cervix adenocarcinoma cell-line
87110901
Hep G2 human hepatocyte carcinoma — ECACC No. 85011430
HMBC
HSQC
ORTEP
THF
correlation spectroscopy
η5-cyclopentadienyl
European Collection of Cell Cultures
benzoylacetone=phenylbutane-1,3-dione
— ECACC No.
heteronuclear multiple-bond correlation
heteronuclear single quantum coherence
Oak Ridge Thermal Ellipsoid Plot Program
tetrahydrofurane
TMS
tetramethylsilane
Acknowledgments
4. Conclusions
We gratefully acknowledge financial and scientific support from
the Konstanz Research School Chemical Biology (KoRS-CB). T.A.I.
thanks the KoRS-CB for a personal scholarship. The authors would
like to express their gratitude to Malin Bein for help with the biolog-
ical assays and to Anke Friemel for recording the time-resolved and
2D-NMR spectroscopy data.
We recently showed that by increasing the sterical demand of
complexes their cytotoxicity decreased dramatically [14]. Interesting-
ly, the herein described dinuclear complexes [L1,3(OEt)Ti\O\Ti(OEt)
L1,3] formed by partially hydrolysis of [TiL1,3(OEt)2] show strong cyto-
toxicity even so they are quite bulky. In contrast to known cyclic tri-
nuclear titanium(IV) species recently described as being nontoxic
[15, 16], the dinuclear [L1,3(OEt)Ti\O\Ti(OEt)L1,3] still feature re-
placeable ligands, thus potentially permitting the coordinative inter-
action with biomolecules. This is in contrast to mononuclear
complexes which feature flat aromatic moieties and therefore might
allow DNA intercalation [36], X-ray structure determination showed
that [L1(OEt)Ti\O\Ti(OEt)L1] adopts an almost spherical shape.
The pronounced toxicity of the dinuclear complexes thus makes a
proposed involvement of DNA intercalation in the biological activity
Appendix A. Supplementary data
Supplementary data to this article can be found online at doi:10.
1016/j.jinorgbio.2011.08.029.
References
[1] H. Köpf, P. Köpf-Maier, Angewandte Chemie (International Ed. in English) 18
(1979) 477–478.
[2] P. Köpf-Maier, F. Preiss, T. Marx, T. Klapötke, H. Köpf, Anticancer Research 6
(1986) 33–37.
[3] N.J. Sweeney, O. Mendoza, H. Müller-Bunz, C. Pampillón, F.-J.K. Rehmann, K.
Strohfeldt, M. Tacke, Journal of Organometallic Chemistry 690 (2005) 4537–4544.
[4] P.M. Abeysinghe, M.M. Harding, Dalton Transactions (2007) 3474–3482.
[5] K. Strohfeldt, M. Tacke, Chemical Society Reviews 37 (2008) 1174–1187.
[6] F. Caruso, M. Rossi, J. Tanski, R. Sartori, R. Sariego, S. Moya, S. Diez, E. Navarrete, A.
Cingolani, F. Marchetti, C. Pettinari, Journal of Medicinal Chemistry 43 (2000)
3665–3670.
Table 4
Half-life of complexes [TiL1,3(OEt)2] and [L1,3(OEt)Ti\O\Ti(OEt)L1,3] under hydrolytic
conditions determined by time resolved NMR at 37 °C.
Complex
t 1/2 [min]
Complex
t
[min]
1/2
[TiL1(OEt)2]
[TiL2(OEt)2]
50
60
[L1(OEt)Ti\O\Ti(OEt)L1]
[L3(OEt)Ti\O\Ti(OEt)L3]
50
130
[7] T. Schilling, K.B. Keppler, M.E. Heim, G. Niebch, H. Dietzfelbinger, J. Rastetter, A.R.
Hanauske, Investigational New Drugs 13 (1996) 327–332.

T.A. Immel et al. / Journal of Inorganic Biochemistry 106 (2012) 68–75
75
[8] H. Bischoff, M.R. Berger, B.K. Keppler, D. Schmähl, Journal of Cancer Research and
Clinical Oncology 113 (1987) 446–450.
[22] B. Schweder, D. Walther, T. Dohler, O. Klobes, H. Gorls, Journal für Praktische Chemie
341 (1999) 736–747.
[9] E.Y. Tshuva, D. Peri, Coordination Chemistry Reviews 253 (2009) 2098–2115.
[10] E.Y. Tshuva, J.A. Ashenhurst, European Journal of Inorganic Chemistry 15 (2009)
2203–2218.
[11] T.A. Immel, M. Debiak, U. Groth, A. Bürkle, T. Huhn, ChemMedChem 4 (2009)
738–741.
[12] J.H. Toney, T.J. Marks, Journal of the American Chemical Society 107 (1985)
947–953.
[13] H. Köpf, S. Grabowski, R. Voigtländer, Journal of Organometallic Chemistry 216
(1981) 185–190.
[23] M.J. Caulfield, T. Russo, D.H. Solomon, Australian Journal of Chemistry 53 (2000)
545–549.
[24] C.J. Whiteoak, G.J.P. Britovsek, V.C. Gibson, A.J.P. White, Dalton Transactions
(2009) 2337–2344.
[25] Stoe & Cie GMBH, IPDS Software, Darmstadt, Germany, 1996.
[26] G.M. Sheldrick, SHELX-97, University of Göttingen, Göttingen, 1997.
[27] A. Altomare, M.C. Burla, M. Camalli, G.L. Cascarano, C. Giacovazzo, A. Guagliardi,
A.G.G. Moliterni, G. Polidori, R. Spagna, Journal of Applied Crystallography 32
(1999) 115–119.
[14] T.A. Immel, U. Groth, T. Huhn, Chemistry
A
European Journal 16 (2010)
[28] L.J. Farrugia, Journal of Applied Crystallography 32 (1999) 837–838.
[29] L.J. Farrugia, ORTEP-III for Windows, Journal of Applied Crystallography 30
(1997) 565.
[30] A.L. Spek, Journal of Applied Crystallography 36 (2003) 7–13.
[31] R.D. Fields, M.V. Lancaster, American Biotechnology Laboratory 11 (1993) 48–49.
[32] R. Hamid, Y. Rotshteyn, L. Rabadi, R. Parikh, P. Bullock, Toxicology in Vitro 18
(2004) 703–710.
[33] Systat Software, Inc. 2006 (http://www.systat.com).
[34] J. Balsells, P.J. Carroll, P. Walsh, Journal of Inorganic Chemistry 40 (2001)
5568–5574.
[35] S. Gendler, S. Segal, I. Goldberg, Z. Goldschmidt, M. Kol, Inorganic Chemistry 45
(2006) 4783–4790.
2775–2789.
[15] D. Peri, S. Meker, M. Shavit, E.Y. Tshuva, Chemistry A European Journal 15 (2009)
2403–2415.
[16] D. Peri, S. Meker, C.M. Manna, E.Y. Tshuva, Inorganic Chemistry 50 (2011)
1030–1038.
[17] A.J. Nielson, J.M. Waters, Polyhedron 29 (2010) 1715–1726.
[18] W.L.F. Armarego, C.L.L. Chai, Purification of Laboratory Chemicals, 5th ed. Elsevier
Science, 2003.
[19] I. Correia, J.C. Pessoa, M.T. Duarte, M.F.M. da Piedade, T. Jackush, T. Kiss, M.M.C.A.
Castro, C.F.G.C. Geraldes, F. Avecilla, European Journal of Inorganic Chemistry 11
(2005) 732–744.
[20] P.E. Aranha, M.P. dos Santos, S. Romera, E.R. Dockal, Polyhedron 26 (2007)
1373–1382.
[36] M. Shavit, D. Peri, C.M. Manna, J.S. Alexander, E.Y. Tshuva, Journal of the American
Chemical Society 129 (2007) 12098–12099.
[21] S.R. Doctrow, K. Huffman, C.B. Marcus, G. Tocco, E. Malfroy, C.A. Adinolfi, H. Kruk,
K. Baker, N. Lazarowych, J. Mascarenhas, B. Malfroyt, Journal of Medicinal Chem-
istry 45 (2002) 4549–4558.