Rhenium(I) tricarbonyl complexes of salicylaldehyde semicarbazones: Synthesis, crystal structures and cytotoxicity

Article abstract of DOI:10.1016/j.jinorgbio.2012.10.011

A series of N,N-disubstituted salicylaldehyde semicarbazones (SSCs), HOC₆H₄CHN-NHCONR₂, and their rhenium(I) tricarbonyl complexes, [ReBr(CO)₃(SSC)], have been synthesised and characterised by IR and ¹H NMR spectroscopy. Crystallographic analysis of the complex [ReBr(CO)₃(H₂Bu₂)] (H₂Bu₂ = SSC where R = Buⁿ) showed that the SSC acts as a bidentate ligand via its imino nitrogen and carbonyl oxygen atoms. The [ReBr(CO)₃(SSC)] complexes exhibit moderate to high cytotoxicities towards MOLT-4 cells (IC₅₀ = 1-24 μM, cf. 18 μM for cisplatin), and the majority of them are virtually non-toxic against non-cancerous human fibroblasts. Apoptotic assays of [ReBr(CO) ₃(H₂Bnz₂)] (Bnz = benzyl) revealed that it mediates cytotoxicity in MOLT-4 cells via apoptosis. The complex [ReBr(CO) ₃(H₂Bnz₂)] reacts with guanosine by proton transfer from the phenolic OH group to N(7) of guanosine. In (CD ₃)₂SO, [ReBr(CO)₃(H₂Bnz ₂)] undergoes facile conversion to the dimeric complex, [Re(CO) ₃(HBnz₂)]₂, via bromide dissociation.

Full text of DOI:10.1016/j.jinorgbio.2012.10.011

Journal of Inorganic Biochemistry 119 (2013) 10–20
Contents lists available at SciVerse ScienceDirect
Journal of Inorganic Biochemistry
journal homepage: www.elsevier.com/locate/jinorgbio
Rhenium(I) tricarbonyl complexes of salicylaldehyde semicarbazones:
Synthesis, crystal structures and cytotoxicity
⁎
Junming Ho, Wan Yen Lee, Kelvin Jin Tai Koh, Peter Peng Foo Lee, Yaw-Kai Yan
Natural Sciences and Science Education, National Institute of Education, Nanyang Technological University, 1 Nanyang Walk, Singapore 637616, Singapore
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 23 July 2012
Received in revised form 10 October 2012
Accepted 23 October 2012
Available online 31 October 2012
A series of N,N-disubstituted salicylaldehyde semicarbazones (SSCs), HOC6H4CH_N-NHCONR2, and their
rhenium(I) tricarbonyl complexes, [ReBr(CO)3(SSC)], have been synthesised and characterised by IR and 1H
NMR spectroscopy. Crystallographic analysis of the complex [ReBr(CO)3(H2Bu2)] (H2Bu2=SSC where R=Bun)
showed that the SSC acts as a bidentate ligand via its imino nitrogen and carbonyl oxygen atoms. The
[ReBr(CO)3(SSC)] complexes exhibit moderate to high cytotoxicities towards MOLT-4 cells (IC50=1–24 μM, cf.
18 μM for cisplatin), and the majority of them are virtually non-toxic against non-cancerous human fibroblasts.
Apoptotic assays of [ReBr(CO)3(H2Bnz2)] (Bnz=benzyl) revealed that it mediates cytotoxicity in MOLT-4 cells
via apoptosis. The complex [ReBr(CO)3(H2Bnz2)] reacts with guanosine by proton transfer from the phenolic
OH group to N(7) of guanosine. In (CD3)2SO, [ReBr(CO)3(H2Bnz2)] undergoes facile conversion to the dimeric
complex, [Re(CO)3(HBnz2)]2, via bromide dissociation.
Keywords:
Salicylaldehyde semicarbazone
Rhenium(I) carbonyl
Cytotoxicity
Anti-cancer
Crystal structure
Apoptosis
© 2012 Elsevier Inc. All rights reserved.
1. Introduction
and aromatic N-substituents were chosen to generate SSCs and com-
plexes that span a wide range of lipophilicities. The cytotoxicity of
Semicarbazones and thiosemicarbazones (Fig. 1) are versatile
compounds that exhibit a wide spectrum of pharmacological activi-
ties, including antitumour activity [1–8]. With the increasing interest
in metal-based drugs, there is also extensive literature on the medic-
inal properties of metal complexes of thiosemicarbazones [9–19]. The
chemistry and biological activity of semicarbazone complexes have
received less attention, however, although the anti-cancer properties
of vanadium, iron, copper and platinum complexes of salicylaldehyde
semicarbazones have been reported [17,20–22]. Significant cancer
cell cytotoxicity has been reported for rhenium(I) tricarbonyl com-
plexes of bis(diphenylphosphinomethyl)amines [23]. Rhenium(I)
carbonyl 2-(dimethylamino)ethoxide complexes were also found to
be very active against suspended cancer cell lines (e.g. MOLT-4 and
HL-60) and solid tumours (e.g. MCF-7 and SK2) [24]. We therefore decid-
ed to extend the above studies by investigating the cytotoxicity of
rhenium(I) tricarbonyl salicylaldehyde semicarbazone complexes. Whilst
rhenium(I) tricarbonyl complexes of thiosemicarbazones are known (but
not their biological activity) [25–28], rhenium(I) semicarbazone com-
plexes have not been reported. It is thus also of interest to study the struc-
ture and solution chemistry of rhenium(I) salicylaldehyde semicarbazone
complexes.
these compounds towards two human cancer cell lines (MOLT-4 and
MCF-7) and non-cancerous human fibroblasts was also determined. In
addition, 1H NMR experiments were conducted on a representative
rhenium complex to investigate its stability in solution and study its
reaction with guanosine. This representative rhenium complex was
also investigated for its ability to trigger apoptosis in human leukaemia
cells (MOLT-4).
2. Experimental
2.1. Materials and apparatus
All experiments were carried out under nitrogen or argon, employing
standard Schlenk techniques. Chemical reagents, unless otherwise stat-
ed, were used directly from commercial sources. Solvents were dried
and distilled under nitrogen before use. The compound [ReBr(CO)5]
was prepared according to a published method [29]. Proton NMR spectra
were recorded on a Bruker CRX400 spectrometer at 25 °C. Elemental
analyses were carried out using an Elementar Vario MICRO CUBE instru-
ment. The IR spectra were recorded on KBr discs using a Perkin–Elmer
Spectrum 100 FT-IR spectrometer.
In this paper, we report the synthesis and characterisation of a se-
ries of N,N-disubstituted salicylaldehyde semicarbazones (SSCs) and
their rhenium(I) tricarbonyl complexes (Fig. 2). A variety of aliphatic
2.2. Preparation of SSC ligands
The preparative protocol was adapted from published methods
[21,30]. For H2Hex2, H2Bnz2 and H2Ph2, the carbamic chloride
⁎
Corresponding author. Tel.: +65 6790 3813; fax: +65 6896 9414.
E-mail address: yawkai.yan@nie.edu.sg (Y.-K. Yan).
0162-0134/$ – see front matter © 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jinorgbio.2012.10.011

J. Ho et al. / Journal of Inorganic Biochemistry 119 (2013) 10–20
11
each) and the combined extract was evaporated under reduced
pressure to give the semicarbazide [R2NC(O)NHNH2]. A solution
of the semicarbazide in toluene (5 mL) was placed on a magnetic
stirrer. To this was added an equimolar amount of salicylaldehyde
and catalytic amounts of p-toluenesulfonic acid. The mixture was
stirred at room temperature for 30 min to give a yellow or white
precipitate of the semicarbazone, which was isolated by filtra-
tion. After washing with cold toluene and drying in vacuo, the
product was recrystallised from acetone–water mixture.
2.2.1. Data for H2Bu2
Fig. 1. The general structure of semicarbazones (X=O) and thiosemicarbazones (X=S);
Yield: (3.3 g, 76%). Anal. Calc. for C16H25N3O2: C, 65.9; H, 8.7; N,
14.4. Found C, 65.7; H, 8.5; N, 14.5. IR (cm−1): ν(C_O) 1640 vs. 1H
NMR [(CD3)2CO, ppm]: 11.80 (1H, s, phenol O\H), 9.65 (1H, s, hydra-
zine N\H), 8.28 (1H, s, imine C\H), 7.24 (2H, m, Ph-H), 6.87 (2H, m,
Ph-H), 3.37 (4H, t, J=7 Hz, α-CH2), 1.61 (4H, m, β-CH2), 1.35 (4H, m,
γ-CH2), 0.95 (6H, t, J=6 Hz, CH3).
R1, R2, R3 and R4=H, alkyl or aryl groups.
[R2NC(O)Cl] was synthesised from the respective amines, whilst for
the other ligands, the carbamic chlorides were obtained commercially.
(a) Preparation of carbamic chlorides
A
solution of the secondary amine (15 mmol) in
2.2.2. Data for H2Hex2
dichloromethane (25 mL) was added dropwise via a cannula
to a stirred solution of triphosgene (5.25 mmol) and pyridine
(30 mmol) in dichloromethane (25 mL) in an ice-water bath.
The resultant orange solution was left to stand at room tempera-
ture for three days under nitrogen to give a yellow solution of the
carbamic chloride.
Yield: (3.4 g, 65%). Anal. Calc. for C20H33N3O2: C, 69.1; H, 9.6; N,
12.1. Found C, 69.2; H, 9.4; N, 12.3. IR (cm−1): ν(C_O) 1636 vs. 1H
NMR [(CD3)2CO, ppm]: 11.80 (1H, s, phenol O\H), 9.70 (1H, s, hydra-
zine N\H), 8.27 (1H, s, imine C\H), 7.23 (2H, m, Ph-H), 6.88 (2H, m,
Ph-H), 3.37 (4H, t, J=7 Hz, α-CH2), 1.63 (4H, m, β-CH2), 1.33 (12H,
m, γ- to ε-CH2), 0.90 (6H, t, J=6 Hz, CH3).
(b) Preparation of SSCs
To the solution of carbamic chloride was added diethyl ether
(60 mL). The resultant mixture was added dropwise to a solution
of hydrazine monohydrate (60 mmol) in ethanol (30 mL), with
vigorous stirring. The reaction mixture was left to stir for 30 min
before the solvents were removed under reduced pressure. The
residue was extracted three times with dichloromethane (10 mL
2.2.3. Data for H2Bnz2
Yield: (4.0 g, 75%). Anal. Calc. for C22H21N3O2: C, 73.6; H, 5.9; N,
11.6. Found C, 73.7; H, 6.3; N, 11.6. IR (cm−1): ν(C_O) 1645 vs. 1H
NMR [(CD3)2SO, ppm]: 11.55 (1H, s, phenol O\H), 10.85 (1H, s, hy-
drazine N\H), 8.35 (1H, s, imine C\H), 7.2-7.4 (12H, m, Ph-H),
6.88 (2H, m, Ph-H), 4.52 (4H, s, CH2).
Fig. 2. The salicylaldehyde semicarbazones and rhenium(I) carbonyl complexes synthesised in this work.

12
J. Ho et al. / Journal of Inorganic Biochemistry 119 (2013) 10–20
2.2.4. Data for H2MePh
7.86(1H, s, imine C\H), 7.2-7.4 (12H, m, Ph-H), 6.85 (1H, d, J=
8 Hz, Ph-H), 6.69 (1H, t, J=8 Hz, Ph-H), 4.64 (2H, br d, J=16 Hz,
CH2), 4.43 (2H, br d, J=16 Hz, CH2).
Yield: (2.4 g, 60%). Anal. Calc. for C15H15N3O2: C, 66.9; H, 5.6; N,
15.6. Found C, 66.9; H, 5.9; N, 15.4. IR (cm−1): ν(C_O) 1661 vs. 1H
NMR [(CD3)2CO, ppm]: 11.65 (1H, s, phenol O\H), 9.35 (1H, s, hydra-
zine C\H), 8.10 (1H, s, imine C\H), 7.1–7.5 (7H, m, Ph-H), 6.82 (2H,
m, Ph-H), 3.31 (3H, s, CH3).
2.3.4. Data for [ReBr(CO)3(H2MePh)]
Yield: (50 mg, 40%). Anal. Calc. for C18H15BrN3O5Re: C, 34.9; H,
2.4; N, 6.8. Found C, 34.9; H, 2.4; N, 6.8. IR (cm−1): ν(C`O) 2025
vs, 1916 vs, 1881 vs; ν(C_O) 1626 vs. 1H NMR [(CD3)2CO, ppm]:
8.25(1H, s, imine C\H), 7.4–7.6 (7H, m, Ph-H), 7.04 (1H, m, Ph-H),
6.95 (1H, m, Ph-H), 3.47 (3H, m, CH3).
2.2.5. Data for H2Ph2
Yield: (3.0 g, 60%). Anal. Calc. for C20H17N3O2: C, 72.2; H, 5.1; N,
12.9. Found C, 72.3; H, 5.3; N, 12.7. IR (cm−1): ν(C_O) 1691 vs. 1H
NMR [(CD3)2CO, ppm]: 11.58 (1H, s, phenol O\H), 9.61 (1H, s, hydra-
zine N\H), 8.18 (1H, s, imine C\H), 7.1–7.5 (12H, m, Ph-H), 6.84
(2H, m, Ph-H).
2.3.5. Data for [ReBr(CO)3(H2Ph2)]
Yield: (49 mg, 36%). Anal. Calc. for C23H17BrN3O5Re: C, 40.5; H,
2.5; N, 6.1. Found C, 40.3; H, 2.6; N, 5.9. IR (cm−1): ν(C`O) 2027
vs, 1920 vs, 1889 vs; ν(C_O) 1618 m. 1H NMR [(CD3)2CO, ppm]:
8.34 (1H, s, imine C\H), 7.3-7.6 (12H, m, Ph-H), 7.05 (1H, m,
Ph-H), 6.93 (1H, m, Ph-H).
2.2.6. Data for H2BF
Yield: (4.3 g, 81%). Anal. Calc. for C22H19N3O2: C, 73.9; H, 5.4; N,
11.8. Found C, 73.7; H, 5.6; N, 11.6. IR (cm−1): ν(C_O) 1655 vs. 1H
NMR [(CD3)2SO, ppm]: 11.45 (1H, s, phenol O\H), 10.10 (1H, s,
hydrazine N\H), 8.32 (1H, s, imine C\H), 7.42 (2H, m, Ph-H), 7.27
(8H, m, Ph-H), 6.87 (2H, m, Ph-H), 3.08 (4H, s, CH2).
2.3.6. Data for [ReBr(CO)3(H2BF)]
Yield: (78 mg, 55%). Anal. Calc. for C25H19BrN3O5Re: C, 42.4; H,
2.7; N, 5.9. Found C, 42.1; H, 2.9; N, 5.9. IR (cm−1): ν(C`O) 2028
vs, 1915 vs, 1892 vs; ν(C_O) 1623 m. 1H NMR [(CD3)2CO, ppm]:
10.77 (1H, s, hydrazine N\H), 8.31(1H, s, imine C\H), 7.3-7.7
(10H, m, Ph-H), 7.02 (1H, m, Ph-H), 6.92 (1H, m, Ph-H), 3.5 (4H, s,
CH2).
2.3. Preparation of [ReBr(CO)3(SSC)] complexes
A mixture of [ReBr(CO)5] (0.2 mmol), SSC (0.2 mmol) and toluene
(3 mL) was placed in a Schlenk tube with a magnetic stirrer bar. The
mixture was refluxed and the progress of the reaction was monitored
hourly by IR spectroscopy. For the H2Hex2 ligand, the reaction mix-
ture was stirred at 105 °C, slightly below the boiling point of toluene,
because decomposition occurred at refluxing temperature. Generally,
the reaction was complete after 5 h.
Some the complexes (where L=H2Bu2, H2Bnz2 and H2MePh)
began to precipitate out within the first 2 h of reaction. For these
complexes, the crude product was isolated by filtration and washed
with a small volume of cold toluene. The crude product was dissolved
in a minimum amount of dichloromethane and layered with hexane
to afford yellow crystals within three days.
In other cases (L=H2Hex2, H2Ph2 and H2BF), the crude product
was obtained as a waxy solid by removing the toluene under reduced
pressure. Recrystallisation was carried out by dissolving the solid in a
minimum amount of dichloromethane and mixing with approximate-
ly two volumes of hexane to form a homogeneous solution. The vial
was sealed with sealing film and stored at 8 °C for three days to ob-
tain yellow crystals of the complex. The crystals were washed with
a small volume of hexane, crushed and dried under reduced pressure.
2.4. Formation of [Re(CO)3(HBnz2)]2 and [Re(CO)3(HHex2)]2
Hexane was layered over dilute solutions of the mononuclear
complexes, [ReBr(CO)3(H2Bnz2)] and [ReBr(CO)3(H2Hex2)], respec-
tively, in dichloromethane. Crystals of the corresponding dinuclear
complexes, [Re(CO)3(HBnz2)]2 and [Re(CO)3(HHex2)]2, were formed
after two weeks at room temperature.
2.4.1. Data for [Re(CO)3(HBnz2)]2
IR (cm−1): ν(C`O) 2021 vs, and 1902 vs; ν(C_O) 1573 s. 1H
NMR [(CD3)2SO, ppm]: 12.40 (2H, s, hydrazine N\H), 7.91 (4H, d,
J=9 Hz, Ph-H), 7.88 (2H, s, imine C\H), 7.4-7.2 (20H, m, Ph-H),
6.78 (4H, q, J=9 Hz, Ph-H), 4.55 (8H, q, J=16 Hz, CH2).
2.4.2. Data for [Re(CO)3(HHex2)]2
IR (cm−1): ν(C`O) 2018 vs, 1910 vs, and 1879 vs; ν(C_O)
1585 m. 1H NMR (CDCl3, ppm): 13.83 (2H, s, hydrazine N\H), 7.82
(2H, s, imine C\H), 7.50 (2H, m, Ph-H), 7.1–7.3 (4H, m, Ph-H), 6.82
(2H, m, Ph-H), 3.65 (2H, m, α-CH2), 3.0–3.3 (4H, m, α-CH2), 2.55
(2H, m, α-CH2), 1.46–1.06 (32H, m, β- to ε-CH2), 0.96–0.84 (12H,
m, CH3).
2.3.1. Data for [ReBr(CO)3(H2Bu2)]
Yield: (58 mg, 45%). Anal. Calc. for C19H25BrN3O5Re: C, 35.6; H,
3.9; N, 6.6. Found C, 35.3; H, 3.9; N, 6.4. IR (cm−1): ν(C`O) 2028
vs, 1914 vs, 1902 vs; ν(C_O) 1625 s. 1H NMR [(CD3)2CO, ppm]:
8.30(1H, s, imine C\H), 7.75 (1H, m, Ph-H), 7.55 (1H, m, Ph-H),
7.22 (1H, d, J=8 Hz, Ph-H), 7.15 (1H, t, J=8 Hz, Ph-H), 3.40 (4H,
m, α-CH2), 1.64 (4H, m, β-CH2), 1.37 (4H, m, γ-CH2), 0.94 (6H, t,
J=6 Hz, CH3).
2.5. X-ray crystallography
Crystals of [ReBr(CO)3(H2Bu2)] were grown by layering hexane
over a dilute solution of the complex in dichloromethane. Single crys-
tals of [Re(CO)3(HBnz2)]2 and [Re(CO)3(HHex2)]2 were obtained as
described in Section 2.4. The crystals were mounted on glass fibres
for data collection at 298(2) K. The data were collected in the θ/2θ
mode using a Siemens P4 diffractometer with Mo Kα radiation (λ=
0.71073 Å), and were corrected for absorption effects using ψ-scan
data. The structures were solved by the heavy atom method and re-
fined by full-matrix least-squares on F2. All non-hydrogen atoms
were refined anisotropically. Hydrogen atoms were introduced in cal-
culated positions and allowed to ride on their carrier atoms. Crystal
and refinement data are given in Table 1.
2.3.2. Data for [ReBr(CO)3(H2Hex2)]
Yield: (59 mg, 42%). Anal. Calc. for C23H33BrN3O5Re: C, 39.6; H,
4.8; N, 6.0. Found C, 39.6; H, 4.9; N, 6.1. IR (cm−1): ν(C`O) 2029
vs, 1930 vs, 1887 vs; ν(C_O) 1626 s. 1H NMR [(CD3)2CO, ppm]:
8.30(1H, s, imine C\H), 7.75 (1H, m, Ph-H), 7.55 (1H, m, Ph-H),
7.17 (2H, m, Ph-H), 3.44 (4H, m, α-CH2), 1.67 (4H, m, β-CH2), 1.35
(12H, m, γ- to ε-CH2), 0.89 (6H, t, J=6 Hz, CH3).
2.3.3. Data for [ReBr(CO)3(H2Bnz2)]
The carbon atoms C(17), C(18), C(19) and C(18′) of [ReBr(CO)3(H2
Bu2)] showed very large displacement parameters, suggesting disorder.
Refinement was performed with two disordered positions (65% and
35% occupancies, respectively) for each of these atoms, with the
Yield: (57 mg, 40%). Anal. Calc. for C25H21BrN3O5Re: C, 42.3; H,
3.0; N, 5.9. Found C, 42.4; H, 3.1; N, 5.9. IR (cm−1): ν(C`O) 2030
vs, 1923 vs, 1878 vs; ν(C_O) 1616 s. 1H NMR [(CD3)2SO, ppm]:

J. Ho et al. / Journal of Inorganic Biochemistry 119 (2013) 10–20
13
disordered butyl chains being restrained to have C\C bond lengths of
1.5400 0.0001 Å. The C\C\C angles involving the disordered atoms
are restrained to be 109.50 0.01°. Adjacent carbon atoms of each dis-
ordered chain are restrained to have similar Uij components, allowing
for a gradual change in magnitude and direction of the anisotropic dis-
placement parameters from the α-carbon outward.
The hexyl chains of [Re(CO)3(HHex2)]2 were restrained to have C\C
bond lengths of 1.5400 0.0001 Å and C\C\C angles of 109.50
0.01°. Adjacent carbon atoms of each chain are restrained to have simi-
lar Uij components, allowing for a gradual change in magnitude and
direction of the anisotropic displacement parameters from the α-carbon
outward.
while the adherent MCF-7 cells and fibroblasts were seeded at 20,000
cells per well and cultured for 24 h prior to treatment with compounds.
All compounds were dissolved in absolute DMSO at a concentra-
tion of 100 mM. These solutions were then diluted to various concen-
trations, using a 10% v/v solution of DMSO in the appropriate culture
medium, before being added to the cell culture wells. The final DMSO
concentration was 1.25% v/v in each well. Six serial dilutions were
made for each compound in order to generate a dose–response
curve, and six replicate wells were set up for each concentration of
the compound. Each plate contained a blank well (cell-free medium-
only well), solvent control wells which contained cells and 1.25% v/v
DMSO, drug colour control wells which contained drug and medium
only, and growth control wells which contained only cells in medium.
Following incubation of cells with test fractions for 24 h at 37 °C
and 5% CO2, 20 μL of a 5 mg/mL solution of MTT was added to each
well (final MTT concentration 1 mg/mL). Three hours later, 100 μL
of lysing solution (20% sodium dodecyl sulfate dissolved in 50% N,
N-dimethylformamide, pH adjusted to 4.7 with acetic acid) was
added to each well. After the microtitre plate was left to stand over-
night, the absorbance of the solution in each well was read at
570 nm. The percent inhibition of growth for each concentration of
compound was calculated from the absorbance [31] and plotted
against the concentration to give a graph from which the IC50 value
(concentration of compound required to inhibit the growth of the
cells by 50%) was determined.
2.6. Cell types and culture conditions
The human cancer cell lines, MOLT-4 (T-lymphoblastic leukaemia)
and MCF-7 (breast carcinoma), were obtained from the American
Type Culture Collection. Human foreskin fibroblasts (as a passage 1
culture) were obtained from the Department of Surgery, National
University of Singapore. Growth of the fibroblasts was subsequently
expanded and a passage 4 culture was used for the cytotoxicity
assay. An RPMI-1640 based medium containing 10% foetal bovine
serum was used to culture MOLT-4 cells, while Dulbecco's Minimal
Essential Medium/Ham's F12 medium (1:1) supplemented with 10%
foetal bovine serum was used for MCF-7 cells. Human fibroblasts were
cultured in Dulbecco's Minimal Essential Medium supplemented with
20% foetal bovine serum. All media also contained 100 units/mL of pen-
icillin and 100 μg/mL of streptomycin. The cells were maintained at
37 °C in a 5% CO2 incubator.
2.8. Assessment of apoptosis induction
The ability of [ReBr(CO)3(H2Bnz2)] to induce apoptosis in MOLT-4
cells was determined by the following experiments. Each experiment
was carried out three times to ensure the reproducibility of results.
2.7. Cytotoxicity assay
2.8.1. Annexin V-FITC staining assay
MOLT-4 cells (106 for each experiment) were grown in 5 cm2
sterile petri dishes and treated with [ReBr(CO)3(H2Bnz2)] (7.3 μM)
or solvent control (1% DMSO) for 3, 6, 12 and 24 h, respectively. Ap-
optotic cell death was determined by staining treated cells with
Annexin-V-FITC and counterstaining with propidium iodide (PI).
Annexin-V-FITC binds to the exposed phosphatidylserine (PS) on
the outer plasma membrane of apoptotic cells, while PI is excluded by vi-
able cells with intact membranes. Cells positive for annexin-V-FITC but
negative for PI are defined as apoptotic, while cells stained with PI alone
is indicative of necrosis [32,33]. Treated cells were washed twice with
PBS, stained with annexin-V-FITC (BD Pharmingen, 51-65874X) and PI
(Sigma, 20 μg/mL), and analysed on a Dako Cytomation Cyan LX flow
cytometer with 488 nm laser excitation. Emission from FITC was detected
at 525 nm and that for PI at 575 nm.
Cytotoxicity of the compounds was determined using the MTT
[3(3,4-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] assay
[31] on microtitre plates. The suspended MOLT-4 cells were seeded at
40,000 cells per well and exposed to compounds on the same day,
Table 1
Crystallographic data and structure refinement details for [ReBr(CO)3H2Bu2],
[Re(CO)3(HBnz2)]2 and [Re(CO)3(HHex2)]2.
Complex
[ReBr(CO)3(H2Bu2)] [Re(CO)3(HBnz2)]2 [Re(CO)3(HHex2)]2
Chemical formula C19 H25 Br N3 O5 Re C50 H40 N6 O10 Re2 C46 H64 N6 O10 Re2
Formula weight
Crystal system
Space group
a (Å)
b (Å)
c (Å)
641.53
1257.28
Monoclinic
C2/c
19.281(5)
13.684(3)
17.610(2)
90
99.31(1)
90
4585(2)
4
1233.44
Triclinic
P−1
Monoclinic
P2(1)/n
12.525(4)
19.21(1)
19.297(5)
90
94.68(2)
90
4627(4)
8
10.031(2)
10.3750(9)
13.953(1)
108.958(5)
109.44(1)
93.310(9)
1272.5(3)
1
2.8.2. Detection of caspase 3 activity
The caspase 3 activity was quantified according to instructions of
the colorimetric assay kit (Clontech 636217). MOLT-4 cells (106 for
each experiment) were treated with [ReBr(CO)3(H2Bnz2)] (7.3 μM)
or solvent control (1% DMSO) for various durations, after which the
cells were collected by centrifugation at 200 g for 5 min. The cell pel-
let was lysed in 50 μL cell lysis buffer. Aliquots of 50 μL of cell lysate
were mixed with 50 μL of reaction buffer and incubated with the
caspase 3 substrate Ac-DEVD-pNA (50 μM) at 37 °C for 1 h. The
caspase 3 activity was determined spectrophotometrically at 405 nm.
α (°)
β (°)
γ (°)
Volume (Å3)
Z
Crystal size
0.30×0.20×0.10
0.30×0.20×0.10
0.30×0.20×0.10
(mm3)
θ Range (°)
Max. and min.
transmission
1.87 to 25.01
0.9835 and
0.4515
7.012
9893
1.83 to 25.00
0.8037 and 0.5103 0.9301 and 0.6603
2.11 to 25.00
μ (mm−1
)
5.343
4832
4.810
5199
Reflections
collected
2.8.3. Western blot analysis
MOLT-4 cells (106 for each experiment) were incubated with
[ReBr(CO)3(H2Bnz2)] (7.3 μM) for various durations, after which the
cell pellet was collected by centrifugation at 200 g for 5 min at 4 °C.
The cells were lysed by resuspension in lysis solution (7 M urea,
2 M thiourea, 4% CHAPS, 30 mM Tris, pH 9.0) and brief sonication
on ice for 2 min at 30% amplitude, 30 s intervals. The cell lysates
Independent
reflections
Final R indices
[I>2sigma(I)]
R indices
8093 [R(int)
=0.0384]
R1=0.0495,
wR2=0.0974
R1=0.1075,
wR2=0.1119
4030 [R(int)
=0.0299]
R1=0.0322,
wR2=0.0804
R1=0.0436,
wR2=0.0851
4375 [R(int)
=0.0223]
R1=0.0378,
wR2=0.1004
R1=0.0461,
wR2=0.1044
(all data)

14
J. Ho et al. / Journal of Inorganic Biochemistry 119 (2013) 10–20
were centrifuged at 200 g for 5 min, after which the protein-
containing supernatants were quantified using the Bradford protein
assay reagent (Biorad, 500–0201). Equal amounts of proteins (20 μg)
were boiled at 100 °C with 10 μL of 2×Laemmli sample buffer (Biorad,
161–0737) for 2 min, and loaded into each lane of a SDS-PAGE gel (5%
stacking, 10% resolving). After electrophoresis, the separated pro-
teins were transferred to a nitrocellulose membrane (GE Healthcare,
RPN2020D). The blots were blocked in phosphate buffered saline
(PBS) containing 0.05% Tween 20 and 5% non-fat milk for 1 h at
room temperature. After blocking, the blots were probed with rabbit
polyclonal caspase 3 antibody at 1:2500 dilution (US Biological,
C2087-16). After washing twice with PBS Tween (PBS containing
0.05% Tween 20), the blots were incubated with horseradish
peroxidase-conjugated polyclonal goat anti-rabbit secondary antibody
(1:5000, US Biological, 11904-39) for 1 h at room temperature. Blots
were washed again (3×5 min) in PBS Tween and visualised using an en-
hanced chemiluminescence kit (GE healthcare, RPN2124), in accordance
with the manufacturer's recommendations.
The facile formation of these dinuclear complexes is somewhat un-
expected since the soft rhenium(I) centre is expected to bind more
strongly to bromide than to the phenoxide oxygen. Presumably, the for-
mation of [Re(CO)3(HBnz2)]2 and [Re(CO)3(HHex2)]2 is driven by en-
tropy rather than enthalpy, with the formation of 2 mol of gaseous
HBr contributing to the increase in entropy. The reaction most likely
proceeds via the dissociation of bromide as the rate-limiting step,
followed by coordination of the hydroxyl group, and then deproton-
ation. That bromide dissociation is rate-limiting is supported by the
fact that, whereas only 50% of [ReBr(CO)3(H2Bnz2)] is converted to
[Re(CO)3(HBnz2)]2 after 24 h at 37 °C in (CD3)2SO (see Section 3.6),
quantitative conversion occurs within 30 min after addition of one
molar equivalent of AgNO3 to [ReBr(CO)3(H2Bnz2)] at ambient temper-
ature (as shown by 1H NMR spectroscopy). Carballo et al. have ob-
served the similar formation of a dinuclear [ReL(CO)3]2 complex
during slow concentration of an acetone solution of [ReBr(CO)3(HL)]
(HL=ferrocenylcarbaldehyde N-methylthiosemicarbazone) [27], in
which case deprotonation occurs at the hydrazidic nitrogen and
Re-S-Re bridges are formed.
3. Results and discussion
3.2. Infrared spectroscopy
3.1. Syntheses
All the SSCs synthesised show an intense ν(C_O) absorption band
at 1636–1691 cm−1. The carbonyl stretching frequencies are consis-
tently higher for the SSCs bearing aromatic substituents on the terminal
nitrogen (H2Ph2, H2MePh and H2BF). This may be rationalised in terms
of resonance effects (Scheme 1), where the phenyl ring stabilises reso-
nance structure III by competing with the carbonyl group for conjuga-
tion with the amidic nitrogen lone pair.
The carbonyl stretching frequencies decrease on coordination of the
SSCs to rhenium (Table 2). This can be attributed to the stabilisation of
resonance structures IIa and IIb through stabilisation of the negative
charge on the carbonyl oxygen by the Lewis-acidic rhenium centre.
Amongst the complexes, there is negligible difference between the
ν(C_O) of the SSC ligands with aliphatic substituents and that of the
SSCs with aromatic substituents. This implies that resonance structures
IIa and IIb are much preferred over III in the complexes. Interestingly,
the C_O stretching frequencies of [Re(CO)3(HBnz2)]2 and
[Re(CO)3(HHex2)]2 are each lower than those of [ReBr(CO)3(H2Bnz2)]
and [ReBr(CO)3(H2Hex2)], respectively, by ca. 40 cm−1. Presumably,
the phenoxide oxygen, being a poorer σ-donor than bromide, makes
the rhenium atoms in the dinuclear complexes more Lewis-acidic than
those in the mononuclear complexes, thereby further favouring reso-
nance structures IIa and IIb.
The synthetic protocol for SSCs reported by Lee [14] was modified
to improve the efficiency of reaction between the semicarbazide and
salicylaldehyde. In Lee's procedure, this reaction was carried out in
ethanol at 0 °C, and the SSC product precipitated from the solution.
Some of the SSCs used in this work are quite soluble in ethanol, how-
ever, and did not precipitate out even after a significant portion of
ethanol was evaporated off. This resulted in poor yields for these
ligands (b50%). A protocol by Noblia [21] was thus used, in which
the reaction between the semicarbazide and salicylaldehyde was car-
ried out in the presence of catalytic amounts p-toluenesulfonic acid in
toluene at room temperature. Carrying out the reaction at room tem-
perature rather than in an ice-bath also increased the rate of reaction,
and the crude product usually precipitated out of solution almost in-
stantaneously. This approach resulted in significantly higher yields
(at least 60%) for all the SSCs.
The rhenium(I) tricarbonyl complexes of the SSCs were prepared by
refluxing an equimolar mixture of the ligand and [ReBr(CO)5] in toluene.
Toluene was used because it is a non-coordinating solvent with high
boiling point. Coordinating solvents such as tetrahydrofuran were
found to compete with the SSCs for complexation and resulted in poor
yields of the desired complexes.
During an attempt to grow single crystals of the complexes
[ReBr(CO)3(H2Bnz2)] and [ReBr(CO)3(H2Hex2)], crystals of the dinuclear
complexes [Re(CO)3(HBnz2)]2 and [Re(CO)3(HHex2)]2 were obtained
instead (Fig. 3). These crystals were grown under conditions similar to
those for recrystallising the mononuclear complexes, but more dilute so-
lutions were used and consequently crystals only appeared after approx-
imately two weeks. Crystallographic analysis (vide infra) revealed that
the phenolic oxygen atoms have been deprotonated, and have displaced
the bromide ligands of the rhenium centres to form a bridged structure.
3.3. Proton NMR spectroscopy
The signals of the SSC protons generally shift downfield on coordi-
nation. A single imine proton peak appears in the spectrum of each
complex, indicating that the SSC ligand is coordinated to the rhenium
atom in a single configuration. Interestingly, the methylene protons
of the H2Bnz2 ligand give two broad doublets in the spectrum of
[ReBr(CO)3(H2Bnz2)], as opposed to a sharp singlet in the spectrum
Fig. 3. Formation of the dinuclear complexes [Re(CO)3(HBnz2)]2 and [Re(CO)3(HHex2)]2.

J. Ho et al. / Journal of Inorganic Biochemistry 119 (2013) 10–20
15
Table 3
Selected bond lengths (Å) and angles (°) for [ReBr(CO)3(H2Bu2)].
Bond lengths
Re(1)-C(1)
Re(1)-C(2)
Re(1)-C(3)
Re(1)-Br(1)
Re(1)-N(1)
Re(1)-O(5)
C(10)-N(1)a
C(11)-O(5)b
C(11)-N(3)c
C(11)-N(2)d
1.90(1)
1.90(1)
1.90(1)
2.615(1)
2.167(7)
2.160(6)
1.28(1)
1.25(1)
1.32(1)
1.37(1)
Re(1′)-C(1′)
Re(1′)-C(2′)
Re(1′)-C(3′)
Re(1′)-Br(1′)
Re(1′)-N(1′)
Re(1′)-O(5′)
C(10′)-N(1′)a
C(11′)-O(5′)b
C(11′)-N(3′)c
C(11′)-N(2′)d
1.89(1)
1.89(1)
1.91(1)
2.627(1)
2.172(7)
2.138(6)
1.28(1)
1.25(1)
1.33(1)
1.37(1)
Bond angles
C(11)-N(3)-C(12)
C(11)-N(3)-C(16)
C(12)-N(3)-C(16)
119.9(9)
123.6(8)
116.5(7)
C(11′)-N(3′)-C(12′)
C(11′)-N(3′)-C(16′)
C(12′)-N(3′)-C(16′)
118.7(8)
123.8(7)
117.4(7)
Scheme 1. Four of the resonance structures of the carbamide group of semicarbazones.
a
Imine C_N.
Amide C_O.
Amide C\N.
Hydrazide C\N.
b
Table 2
Comparison of carbonyl stretching frequencies (ν˜/cm−1
[ReBr(CO)3(SSC)] complexes.
c
)
between SSCs and
d
Ligand
ν˜ (ligand)
ν˜ (complex)
ν˜ (ligand)–ν˜ (complex)
semicarbazone, HOC6H4CH_N\NHCONH2 [34]. The three angles
about the terminal nitrogen atom of both molecules in the unit cell
[N(3) and N(3′)] sum up to 360°, indicating a trigonal planar geometry
for the nitrogen atoms. This is consistent with the dominance of reso-
nance structure IIa (Scheme 1), where the terminal nitrogen is sp2
hybridised. The chelate rings are virtually planar, with maximum devi-
ations of 0.01 [for C(11)] and 0.03 Å [for N(1′) and N(2′)] from the
respective mean planes. The coordination plane of each terminal nitro-
gen is almost parallel to the mean plane of the chelate ring to which the
nitrogen is attached (angles of deviation are 3° each), providing further
evidence for the conjugation of the nitrogen lone pair with the C_O
double bond. Notably, an intramolecular hydrogen bond exists between
the phenolic oxygen and the hydrazine NH group of each molecule
(Fig. 4). This is made possible by the adoption of the cis configuration
by the imine group in the complex (as opposed to the trans configura-
tion expected for the free ligand [34]). The imine groups are not copla-
nar with the phenyl rings bearing them, thus the imino protons deviate
significantly from the mean planes of the phenyl rings [by 0.30 Å for the
proton on C(10) and 0.29 Å for that on C(10′)]. The Re-Br, Re-CO and
Re-NC(imine) distances are close to those observed in analogous
rhenium(I) tricarbonyl complexes [25,26,28].
H2Bu2
H2Hex2
H2Bnz2
H2MePh
H2Ph2
1640
1636
1645
1661
1691
1655
1625
1626
1616
1626
1618
1623
15
10
29
35
73
32
H2BF
of the free ligand. This may be attributed to the stabilisation of reso-
nance structure IIa on complexation (see Section 3.2), which results
in hindered rotation of the Bnz2N−CO bond and hence non-
equivalence of the methylene protons. Similar effects are observed
in the methylene proton signals of [ReBr(CO)3(H2Hex2)] and
[ReBr(CO)3(H2Bu2)].
3.4. Crystal structures
3.4.1. Crystal structure of [ReBr(CO)3(H2Bu2)]
There are two independent molecules of the complex
[ReBr(CO)3(H2Bu2)] in the unit cell (Fig. 4). The H2Bu2 ligand acts as
a bidentate ligand, binding to the rhenium centre via the imine nitro-
gen and carbonyl oxygen. This binding mode to the {Re(CO)3} unit is
analogous to that of thiosemicarbazones, wherein the sulfur and
imine nitrogen atoms are coordinated to rhenium [25,26,28]. The
lengths of the C_O, C_N and C\N bonds within the H2Bu2 ligand
(Table 3) are similar to those observed in the parent salicylaldehyde
3.4.2. Crystal structures of [Re(CO)3(HBnz2)]2 and [Re(CO)3(HHex2)]2
The deprotonated SSC ligands adopt a tridentate coordination
mode in [Re(CO)3(HBnz2)]2 (Fig. 5) and [Re(CO)3(HHex2)]2 (Fig. 6),
with the imine nitrogen and carbonyl oxygen binding to one rhenium
Fig. 4. The two independent molecules in the crystal structure of [ReBr(CO)3(H2Bu2)] (40% probability ellipsoids); hydrogen atoms bonded to the butyl and phenyl groups are
omitted for clarity. An intramolecular hydrogen bond exists between O(4) and N(2) and between O(4′) and N(2′) [O(4)∙∙∙H(2A) 1.98 Å, O(4)∙∙∙N(2) 2.64 Å, ∠O(4)∙∙∙H(2A)-N(2)
133°; O(4′)∙∙∙H(2′A) 2.03 Å, O(4′)∙∙∙N(2′) 2.65 Å, ∠O(4′)∙∙∙H(2′A)-N(2′) 128°, N\H 0.860 Å].

16
J. Ho et al. / Journal of Inorganic Biochemistry 119 (2013) 10–20
Fig. 5. Crystal structure of [Re(CO)3(HBnz2)]2; all hydrogen atoms, except those of the imino groups [on C(7) and C(7A)], are omitted for clarity. The molecule has a crystallographic
centre of symmetry.
Fig. 6. Crystal structure of [Re(CO)3(HHex2)]2; hydrogen atoms are omitted for clarity. The molecule has a crystallographic centre of symmetry.

J. Ho et al. / Journal of Inorganic Biochemistry 119 (2013) 10–20
17
Table 4
are the only ones with appreciable toxicity towards MCF-7 cells. They
are also significantly more active than cisplatin against both MOLT-4
and MCF-7 cells.
Selected bond lengths (Å) and angles (°) for [Re(CO)3(HBnz2)]2 and
[Re(CO)3(HHex2)]2.
[Re(CO)3(HBnz2)]2
[Re(CO)3(HHex2)]2
The cytotoxicity of the complexes towards MOLT-4 cells appears
to be correlated with the lipophilicity of the ligands (Table 5). Cyto-
toxicity increases with increasing lipophilicity from H2MePh to
H2Bnz2/H2BF, and decreases thereafter with a further increase in lipo-
philicity. This suggests that absorption across lipid membranes in the
cells is an important factor determining the cytotoxicity of the com-
pounds [39].
All the compounds tested except [ReBr(CO)3(H2BF)] exhibit negligi-
ble toxicity against fibroblasts, showing that most of these compounds
are selective against cancer cells. This is an important property for any
chemotherapeutic agent, since harmful side effects of the treatment
will be minimised. The complex [ReBr(CO)3(H2Bnz2)] stands out as
the only one with appreciable activity against both MOLT-4 and
MCF-7 cells, but negligible toxicity towards fibroblasts.
Bond lengths
Re(1)-C(11)
Re(1)-C(12)
Re(1)-C(13)
Re(1)-O(2A)
Re(1)-N(1A)
Re(1)-O(1)
C(7)-N(1)a
C(8)-O(2)b
C(8)-N(3)c
C(8)-N(2)d
1.909(7)
1.911(6)
1.904(6)
2.145(5)
2.189(4)
2.158(4)
1.300(7)
1.269(6)
1.322(7)
1.373(7)
Re(1)-C(1)
Re(1)-C(2)
Re(1)-C(3)
Re(1)-O(5)
Re(1)-N(1)
Re(1)-O(4A)
C(10)-N(1)a
C(11)-O(5)b
C(11)-N(3)c
C(11)-N(2)d
1.899(8)
1.925(8)
1.897(8)
2.144(5)
2.190(5)
2.156(4)
1.272(9)
1.247(9)
1.335(9)
1.352(9)
Bond angles
C(8)-N(3)-C(9)
C(8)-N(3)-C(10)
C(9)-N(3)-C(10)
118.5(5)
123.9(5)
117.3(5)
C(11)-N(3)-C(21)
C(11)-N(3)-C(31)
C(21)-N(3)-C(31)
119.9(6)
123.4(6)
116.7(5)
a
Imine CN.
Amide CO.
Amide CN.
Hydrazide CN.
3.6. Solution chemistry
b
c
d
The solution chemistry of [ReBr(CO)3(H2Bnz2)] was investigated
by 1H NMR spectroscopy. This complex was chosen as it shows the
best activity profile in the cytotoxicity assays. Since the solubility of
[ReBr(CO)3(H2Bnz2)] in water is too low for NMR studies, investiga-
tions were conducted in (CD3)2SO.
centre and the phenoxo oxygen to the other. In general, the lengths of
the C_O, C_N and C\N bonds within the SSC ligands (Table 4)
do not differ significantly from those in HOC6H4CH_N\NHCONH2
[34], although there is noticeable variation in the lengths of these
bonds between [Re(CO)3(HBnz2)]2 and [Re(CO)3(HHex2)]2. Like those
of [ReBr(CO)3(H2Bu2)], the amide nitrogens of both [Re(CO)3(HBnz2)]2
and [Re(CO)3(HHex2)]2 have trigonal planar coordination geometry
(sum of bond angles=360°). The imino protons of [Re(CO)3(HBnz2)]2
and [Re(CO)3(HHex2)]2 deviate more from the mean planes of the
salicylaldehyde phenyl rings than those of [ReBr(CO)3(H2Bu2)], with
the proton of [Re(CO)3(HBnz2)]2 deviating by 0.41 Å and that of
[Re(CO)3(HHex2)]2 by 0.42 Å.
The spectral changes that occurred after incubating a solution
of [ReBr(CO)3(H2Bnz2)] at 37 °C for 24 h (Fig. 7) are consistent
with the conversion of the compound to the dinuclear complex
[Re(CO)3(HBnz2)]2. In particular, signals attributed to the dinuclear
complex (quartets at 4.54 and 6.78 ppm) appeared while those of
the mononuclear complex (broad doublets at 4.43 and 4.64, triplet
at 6.70, and doublet at 6.83 ppm) decreased in intensity. Since the
quartet at 6.78 ppm overlaps with the doublet at 6.83 ppm but not
the triplet at 6.70 ppm, the percent conversion of [ReBr(CO)3(H2Bnz2)]
to [Re(CO)3(HBnz2)]2 could be calculated to be about 50% from the
expression [(Ia−Ib)/(Ia+Ib)], where Ia is the total integrated intensity
of the overlapping signals at 6.78 and 6.83 ppm and Ib is the integrated
intensity of the triplet at 6.70 ppm. These observations are consistent
with the facile conversion of [ReBr(CO)3(H2Bnz2)] to [Re(CO)3(HBnz2)]2
during crystal growing experiments (Section 3.1).
To investigate the possible interaction of [ReBr(CO)3(H2Bnz2)]
with DNA, we also examined the reaction of [ReBr(CO)3(H2Bnz2)]
with guanosine. Addition of an equimolar amount of guanosine to a
freshly-prepared solution of [ReBr(CO)3(H2Bnz2)] causes immediate
changes to the 1H NMR signals of [ReBr(CO)3(H2Bnz2)] (Fig. 8),
most notably an upfield shift of the imine proton signal from 7.87 to
7.77 ppm and an upfield shift of the doublet at 6.83 ppm (assigned
to the proton ortho to the hydroxyl group) to 6.78 ppm. These shifts
are consistent with the deprotonation of the hydroxyl group of
[ReBr(CO)3(H2Bnz2)]. Correspondingly, the N(1)-H (10.63 ppm),
C(8)-H (7.93 ppm) and NH2 (6.46 ppm) signals of guanosine are
shifted downfield to 10.98, 8.46 and ca. 6.7 ppm, respectively, indi-
cating the protonation of guanosine, predominantly at N(7). Addition
of 1.0, 6.7 and 13.4 molar equivalents of HCl (aq) to guanosine in
(CD3)2SO causes the chemical shift of C(8)-H to plateau at
9.33 ppm. Taking this to indicate 100% protonation of guanosine,
the extent of protonation of guanosine observed in the presence of
[ReBr(CO)3(H2Bnz2)] is calculated to be ca. 36%, assuming that the
C(8)-H signal at 8.46 ppm is the weighted average of unprotonated
and protonated guanosine. No change is observed for the C\H proton
resonances of the ribose moiety [5.75, 4.39, 4.12 and 3.92 ppm for
protons on C(1′)–C(4′), respectively; signals of protons on C(5′) are
masked by the water peak], hence there is probably negligible inter-
action between the ribose unit and [ReBr(CO)3(H2Bnz2)].
3.5. Cytotoxicity
The cytotoxicities of the SSCs and their rhenium tricarbonyl com-
plexes were evaluated against MOLT-4 (cisplatin-sensitive [35]) and
MCF-7 (cisplatin-resistant [36–38]) cells. To assess the selectivity of
these compounds towards cancer cells, their cytotoxicities against
fibroblasts (non-cancerous) were also determined. Cisplatin was
included in the cytotoxicity assays for comparison. The results are
shown in Table 5.
Generally, the [ReBr(CO)3(SSC)] complexes exhibit similar or
higher activity than the corresponding SSC ligands against MOLT-4
cells, but are generally less active than the ligands against MCF-7
cells. The complexes [ReBr(CO)3(H2Bnz2)] and [ReBr(CO)3(H2BF)]
Table 5
IC50 values of the SSC ligands and their rhenium carbonyl complexes. Standard errors
are shown in parentheses (N=6). The LogP values (logarithm of the partition coeffi-
cient between n-octanol and water) were calculated using the ChemDrawUltra 11.0
software.
Ligand (L) LogP IC50 (μM)
L
[ReBr(CO)3L]
MOLT-4 MCF-7 Fibroblast MOLT-4 MCF-7 Fibroblast
H2MePh
H2Ph2
H2Bu2
H2Bnz2
H2BF
2.59 56(3)
3.63 >63a
4.43 5(3)
4.69 11(1)
>125
>63a
5(4)
>125
>63a
>125
24(6)
15(1)
15(1)
>125
>125
>125
>125
>125
>125
8(3)
100(10)
7.3(0.4) 24 (4) >125
1.0 (0.1) 35 (4) 9 (1)
b
b
b
5.02
H2Hex2
Cisplatin
6.55 56(6)
8(4)
71(8)
>125
28(2)
22(2)
18(1)
>125
71(8)
>125
28(2)
–
18(1)
a
Highest concentration tested as the compound has poor solubility in the medium
used for the cytotoxicity assay.
Interestingly, no reaction was observed between guanosine and the
free ligand, H2Bnz2, within 24 h, suggesting that H2Bnz2 is a much
b
Not soluble in the medium used for the cytotoxicity assay.

18
J. Ho et al. / Journal of Inorganic Biochemistry 119 (2013) 10–20
Fig. 7. Proton NMR spectra of [ReBr(CO)3(H2Bnz2)] (12 mM) in (CD3)2SO: (a) obtained from a freshly-prepared solution, (b) recorded after the solution was incubated at 37 °C for
24 h. The peak at 5.76 ppm is due to adventitious CH2Cl2.
weaker acid than [ReBr(CO)3(H2Bnz2)]. The higher acid strength of
[ReBr(CO)3(H2Bnz2)] may be due to the presence of an intramolecular
(hydrazine)NH···O(phenol) hydrogen bond in the complex, assuming
that hydrogen bonding similar to that observed in [ReBr(CO)3(H2Bu2)]
(see Section 3.4.1) occurs in [ReBr(CO)3(H2Bnz2)]. Intramolecular hydro-
gen bonding between phenolic oxygen atoms and NH protons is known
to increase the acid strength of phenols [40]. The existence of the
NH···O interaction is supported by the fact that the benzyl proton signals
at 4.43 and 4.64 ppm shift upfield to 4.39 and 4.58 ppm, respectively (see
Fig. 8), when guanosine is added to [ReBr(CO)3(H2Bnz2)]: deprotonation
of the phenolic oxygen increases the electron density on the hydrazine
nitrogen, which increases the dominance of resonance structure IIb
(Scheme 1) at the expense of structure IIa, thereby increasing the
shielding of the benzyl protons. The {Re(CO)3} group probably also con-
tributes to the higher acidity of the complex since the negative charge
on the phenoxo oxygen can be delocalised onto the imino nitrogen,
where it is stabilised by the Lewis acidic rhenium(I) centre.
Incubating the guanosine-[ReBr(CO)3(H2Bnz2)] mixture at 37 °C
for 24 h resulted in the appearance of resonances due to
[Re(CO)3(HBnz2)]2 (Fig. 8). Taking the multiplet at 4.39 ppm to be com-
posed of resonances of the guanosine C(2′)-H proton and two of the ben-
zyl protons of [ReBr(CO)3(H2Bnz2)], and that at 4.54 ppm to be composed
of resonances due to all four of the benzyl protons of [Re(CO)3(HBnz2)]2
and two of the benzyl protons of [ReBr(CO)3(H2Bnz2)], the percent
conversion of [ReBr(CO)3(H2Bnz2)] to [Re(CO)3(HBnz2)]2 was calculated
to be about 50% from the intensity ratio of the multiplets. The similar
Fig. 8. Proton NMR spectra of: (a) guanosine, (b) [ReBr(CO)3(H2Bnz2)], (c) freshly-prepared mixture of guanosine and [ReBr(CO)3(H2Bnz2)], (d) guanosine-[ReBr(CO)3(H2Bnz2)]
mixture incubated at 37 °C for 24 h. All solutes were dissolved in (CD3)2SO at 12 mM concentration. The guanosine C(1′)-H signal overlaps with that of adventitious CH2Cl2 at
5.76 ppm.

J. Ho et al. / Journal of Inorganic Biochemistry 119 (2013) 10–20
19
complex to be present in significant amounts in the cells, however,
since its rate of formation would be negligible under conditions of
high dilution in the assays. Hence, the unsaturated intermediate
[Re(CO)3(H2Bnz2)]+ is likely to accept electron pairs from donor groups
of biomolecules instead. The NMR spectra of [ReBr(CO)3(H2Bnz2)] and
guanosine-[ReBr(CO)3(H2Bnz2)] mixture show no apparent signals as-
signable to uncomplexed H2Bnz2, hence ligand displacement does not
occur to a significant extent within 24 h at 37 °C. Given that DMSO is
a strong donor solvent, the lack of ligand displacement in DMSO sug-
gests that ligand displacement is not involved in the mechanism of cy-
totoxicity of rhenium salicylaldehyde semicarbazone complexes.
3.7. Assessment of apoptosis induction by [ReBr(CO)3(H2Bnz2)]
MOLT-4 cells treated with [ReBr(CO)3(H2Bnz2)] showed an increas-
ing degree of Annexin-V-FITC positivity over time (Fig. 9), indicating
that the cells were undergoing apoptosis. Occurrence of apoptosis was
further confirmed by measuring the caspase 3 activity of treated cells.
Activation of caspase 3 activity is integral to the initiation of apoptosis
[41,42]. MOLT-4 cells treated with [ReBr(CO)3(H2Bnz2)] showed in-
creased caspase 3 activity over the 24-h period of treatment (Fig. 10a),
with activity reaching its maximum at 6 h. Correspondingly, western
blotting with antibodies against caspase 3 also revealed the presence of
an active cleaved 17 kDa caspase 3 subunit (Fig. 10b).
Fig. 9. Flow cytometric assessment of apoptosis in MOLT-4 cells treated with
[ReBr(CO)3(H2Bnz2)] for 3, 6, 12 and 24 h, respectively: (a) cells stained with
Annexin-V-FITC only; (b) cells stained with propidium iodide only.
4. Summary and conclusion
A
series of N,N-disubstituted salicylaldehyde semicarbazones
percent conversion to that observed in the absence of guanosine is consis-
tent with the earlier deduction (Section 3.1) that the rate-limiting step for
dimerisation is bromide dissociation (and not deprotonation).
The relatively strong acidity of [ReBr(CO)3(H2Bnz2)] suggests that
the deprotonated complex [ReBr(CO)3(HBnz2)]− is an important spe-
cies interacting with cells during the cytotoxicity assay, while the facile
formation of [Re(CO)3(HBnz2)]2 suggests that bromide dissociation
(i.e., solvolysis) occurs quite readily. It is unlikely for the dinuclear
(SSCs) and their rhenium(I) tricarbonyl complexes, [ReBr(CO)3(SSC)],
have been synthesised and screened for cytotoxicity against MOLT-4,
MCF-7 and human fibroblast cells. The dinuclear complexes,
[Re(CO)3(HBnz2)]2 and [Re(CO)3(HHex2)]2, are formed during
slow evaporation of dichloromethane-hexane solutions of
[ReBr(CO)3(H2Bnz2)] and [ReBr(CO)3(H2Hex2)], respectively. Crys-
tal structures have been determined for [ReBr(CO)3(H2Bu2)],
[Re(CO)3(HBnz2)]2 and [Re(CO)3(HHex2)]2. On the whole, the
Fig. 10. Detection of caspase 3 activity in MOLT-4 cells treated with [ReBr(CO)3(H2Bnz2)]: (a) colorimetric measurement of caspase-specific cleavage of Ac-DEVD-pNA; (b) western
blot analysis of proteins separated by SDS-PAGE.

20
J. Ho et al. / Journal of Inorganic Biochemistry 119 (2013) 10–20
[9] M. Belicchi-Ferrari, F. Bisceglie, G. Pelosi, P. Tarasconi, Polyhedron 27 (2008)
1361–1367.
[10] M. Baldini, M. Belicchi-Ferrari, F. Bisceglie, G. Pelosi, S. Pinelli, P. Tarasconi, Inorg.
Chem. 42 (2003) 2049–2055.
[11] M. Belicchi-Ferrari, F. Bisceglie, C. Casoli, S. Durot, I. Morgenstern-Badara, G.
Pelosi, E. Pilotti, S. Pinelli, P. Tarasconi, J. Med. Chem. 48 (2005) 1671–1675.
[12] M. Belicchi-Ferrari, G. Gasparri-Fava, E. Leporati, G. Pelosi, R. Rossi, P. Tarasconi, R.
Albertini, A. Bonati, P. Lunghi, S. Pinello, J. Inorg. Biochem. 70 (1998) 145–154.
[13] M. Belicchi-Ferrari, F. Bisceglie, G. Pelosi, P. Tarasconi, R. Albertini, A. Bonati, P.
Lunghi, S. Pinelli, J. Inorg. Biochem. 83 (2001) 169–179.
[14] Z. Afrasiabi, E. Sinn, J. Chen, Y. Ma, A.L. Rheingold, L.N. Zakharov, Inorg. Chim. Acta
357 (2004) 271.
[15] N.C. Kasuga, K. Sekino, M. Ishikawa, A. Honda, M. Yokoyama, S. Nakano, N.
Shimada, C. Koumo, K. Nomiya, J. Inorg. Biochem. 96 (2003) 298–310.
[16] C.R. Kowol, R. Berger, R. Eichinger, A. Roller, M.A. Jakupec, P.P. Schmidt, V.B. Arion,
B.K. Keppler, J. Med. Chem. 50 (2007) 1254–1265.
[17] J. Patole, S. Padhye, M.S. Moodbidri, N. Shirsat, Eur. J. Med. Chem. 40 (2005)
1052–1055.
[18] U.K. Mazumder, M. Gupta, S.S. Karki, S. Bhattacharya, S. Rathinasamy, S. Thangavel,
Chem. Pharm. Bull. 52 (2004) 178–185.
[ReBr(CO)3(SSC)] complexes exhibit similar or higher activity than
the corresponding SSC ligands against MOLT-4 cells, with two com-
plexes showing particularly high activities. Most of the tested com-
pounds demonstrate negligible activity against non-cancerous
fibroblasts. Of these compounds, [ReBr(CO)3(H2Bnz2)] shows the
best cytotoxicity profile, being the only one that has high activity
against both MOLT-4 and MCF-7 cells, and negligible toxicity towards
fibroblasts. It also induces apoptosis in MOLT-4 cells. In (CD3)2SO,
[ReBr(CO)3(H2Bnz2)] reacts with guanosine by proton transfer from
the phenolic OH group to N(7) of guanosine. The relatively strong acidity
of the OH group may be attributed to intramolecular (hydrazine)
NH···O(phenol) hydrogen bonding, which is made possible by the adop-
tion of the cis configuration by the imine group in [ReBr(CO)3(H2Bnz2)].
Displacement of the SSC ligand is probably not involved in the mechanism
of cytotoxicity of rhenium salicylaldehyde semicarbazone complexes, but
interaction with biomolecules after loss of bromide is plausible.
[19] S. Grguric-Sipka, C.R. Kowol, S.-M. Valiahdi, R. Eichinger, M.A. Jakupec, A. Roller, S.
Shova, V.B. Arion, B.K. Keppler, Eur. J. Inorg. Chem. (2007) 2870–2878.
[20] P. Noblia, E.J. Baran, L. Otero, P. Draper, H. Cerecetto, M. Gonzalez, O.E. Piro, E.E.
Castellano, T. Inohara, Y. Adachi, H. Sakurai, D. Gambino, Eur. J. Inorg. Chem.
(2004) 322–328.
Acknowledgements
[21] W.Y. Lee, P.P.F. Lee, Y.K. Yan, M. Lau, Metallomics 2 (2010) 694–705.
[22] J. Rivadeneira, D.A. Barrio, G. Arrambide, D. Gambino, L. Bruzzone, S.B. Etcheverry,
J. Inorg. Biochem. 103 (2009) 633–642.
[23] J. Zhang, J.J. Vittal, W. Henderson, J. Wheaton, I.H. Hall, T.S. Hor, Y.K. Yan, J.
Organomet. Chem. 650 (2002) 123–132.
The authors would like to thank Dr. Shu Uin Gan (Department of
Surgery, National University of Singapore) for the provision of human
fibroblasts. Financial support from the National Institute of Education,
Singapore (grant number RI 4/05 LPF) is gratefully acknowledged. We
also thank Ms. Ying-Chang Ting for the technical assistance.
[24] W. Wang, Y.K. Yan, T.S. Hor, J.J. Vittal, J.R. Wheaton, I.H. Hall, Polyhedron 21
(2002) 1991–1999.
[25] A. Nunez-Montenegro, R. Carballo, E.M. Vazquez-Lopez, Polyhedron 27 (2008)
2867–2876.
[26] G. Pereiras-Gabian, E.M. Vazquez-Lopez, U. Abram, Z. Anorg. Allg. Chem. 630
(2004) 1665–1670.
Appendix A. Supplementary data
Crystallographic data (excluding structure factors) for the com-
plexes [ReBr(CO)3(H2Bu2)], [Re(CO)3(HBnz2)]2 and [Re(CO)3(HHex2)]2
may be obtained from the Cambridge Crystallographic Data Centre
(CCDC, 12 Union Road, Cambridge CB2 1EZ, UK; fax: _/44-1223-336-
033; e-mail: deposit@ccdc.cam.ac.uk or www: http://www.ccdc.cam.
ac.uk) on quoting the depository numbers CCDC 763433, 763434 and
763435, respectively.
[27] R. Carballo, J.S. Casas, E. Garcia-Martinez, G. Pereiras-Gabian, A. Sanchez, J. Sordo,
E.M. Vazquez-Lopez, J.C. Garcia-Monteagudo, U. Abram, J. Organomet. Chem. 656
(2002) 1–10.
[28] I.G. Santos, U. Abram, R. Alberto, E.V. Lopez, A. Sanchez, Inorg. Chem. 43 (2004)
1834–1836.
[29] S.P. Schmidt, W.C. Trogler, F. Basolo, Inorg. Synth. 28 (1990) 160–165.
[30] P.F. Lee, C.T. Yang, D. Fan, J.J. Vittal, J. Ranford, Polyhedron 22 (2003) 2781–2786.
[31] M.B. Hansen, T. Fojo, J. Immunol. Methods 119 (1989) 203–210.
[32] J.F. Kerr, A.H. Wyllie, A.R. Currie, Br. J. Cancer 26 (1972) 239–257.
[33] I. Vermes, C. Haanen, H. Steffens-Nakken, C. Reutelingsperger, J. Immunol.
Methods 184 (1995) 39–51.
Supplementary data to this article can be found online at doi http://
dx.doi.org/10.1016/j.jinorgbio.2012.10.011.
[34] J. Valdes-Martinez, R.A. Toscano, R. Salcedo, R. Cea-Olivares, A. Melendez,
Monatsh. Chem. 121 (1990) 641–647.
[35] D.M. Fan, C.T. Yang, J.D. Ranford, J.J. Vittal, P.F. Lee, J. Chem. Soc. Dalton Trans.
(2003) 2680–2685.
References
[1] H. Beraldo, D. Gambino, Minirev. Med. Chem. 4 (2004) 31–39.
[2] A. Quiroga, C. Ranninger, Coord. Chem. Rev. 248 (2004) 119–133.
[3] Z. Afrasiabi, E. Sinn, W. Lin, Y. Ma, C. Campana, S.B. Padhyé, J. Inorg. Biochem. 99
(2005) 1526–1531.
[4] Z. Afrasiabi, E. Sinn, S. Padhye, S. Dutta, C. Newton, C.E. Anson, A.K. Powell, J. Inorg.
Biochem. 95 (2003) 306–314.
[36] W. Friebolin, G. Schilling, M. Zoller, E. Amtmann, J. Med. Chem. 48 (2005)
7925–7931.
[37] Y.J. Chua, C. Steer, D. Yip, Cancer Treat. Rev. 30 (2004) 521–543.
[38] F.M. Muggia, T. Fojo, J. Chemother. 16 (Suppl. 4) (2004) 77–82.
[39] C.A. Lipinski, F. Lombardo, B.W. Dominy, P.J. Feeney, Adv. Drug Deliv. Rev. 46
(2001) 3–26.
[5] R.W. Byrnes, M. Mohan, W.E. Antholine, R.X. Xu, D.H. Petering, Biochemistry 29
(1990) 7046–7053.
[40] D. Kanamori, A. Furukawa, T. Okamura, H. Yamamoto, N. Ueyama, Org. Biomol.
Chem. 3 (2005) 1453–1459.
[6] I.H. Hall, S.Y. Chen, B.J. Barners, D.X. West, Metal-Based Drugs 6 (1999) 143–147.
[7] D.L. Klayman, J.F. Bartosovich, T.S. Griffin, C.J. Mason, J.P. Scovill, J. Med. Chem. 22
(1979) 855–862.
[41] Z. Tao, J. Goodisman, H.S. Penefsky, A.K. Souid, Mol. Pharm. 4 (2007) 583–595.
[42] N.A. Thornberry, Y. Lazebnik, Science 281 (1998) 1312–1316.
[8] R.C. Condit, R. Easterly, R.F. Pacha, Z. Fathi, R.J. Meis, Virology 185 (1991) 857–861.