Synthesis, characterization and biological activity of gallium(III) complexes with non-symmetrical N2O-donor Schiff bases

Article abstract of DOI:10.1016/j.poly.2016.12.020

Cisplatin and its analogues are some of the most widely used anticancer drugs. However, toxicity and resistance have limited their use, and gallium compounds have emerged as an alternative, due to their remarkable antitumor activity and low toxicity. In this paper, we report on four new mononuclear gallium(III) complexes with the general formula [Ga(bhi-R)₂]⁺, where bhi-R (R?=?–NO₂, –Br, –Cl and –OCH₃) is the deprotonated form of imidazole/phenol-containing tridentate Schiff bases. The molecular structures of C1–C4 were solved by single crystal X-ray diffraction revealing mononuclear and isostructural complexes, with the metal center in distorted octahedral geometries. In all cases, the gallium(III) is coordinated to two ligand molecules arranged in meridional fashion, through the imidazole and imino nitrogen atoms, and to the phenolate oxygen. Complexes C1–C4 were also characterized in solution by spectroscopic and spectrometric techniques. DFT calculations were performed to assist the interpretation of experimental results, and also allowed to establish the reasons why only meridional isomers were obtained. The stability of all complexes was evaluated in PBS buffer (1% DMSO) over 24?h, and results indicate partial hydrolysis in this period. Finally, the biological activity of C1–C4 was evaluated on human tumor cell lines MCF-7 (breast adenocarcinoma), A-549 (lung adenocarcinoma epithelial) and PC-3 (prostate adenocarcinoma). Complex C1 was the most active, being effective and selective for A-549. The IC₅₀(94.12?±?4.62) of C1 is lower than the value (135.10?±?6.50) obtained for cisplatin, which was tested as a metallodrug reference under the same experimental conditions.

Full text of DOI:10.1016/j.poly.2016.12.020

Accepted Manuscript
Synthesis, Characterization and Biological Activity of Gallium(III) Complexes
with non-symmetrical N2O-Donor Schiff Bases
Diego S. Padilha, Yan F. Santos, Letícia C. Giacomin, Frederico A.V. Castro,
Marcos D. Pereira, Alexandre B. Rocha, Jackson A.L.C. Resende, Marciela
Scarpellini
PII:
S0277-5387(16)30671-4
DOI:
http://dx.doi.org/10.1016/j.poly.2016.12.020
POLY 12378
Reference:
Polyhedron
To appear in:
Received Date:
Revised Date:
Accepted Date:
20 September 2016
12 December 2016
16 December 2016
Please cite this article as: D.S. Padilha, Y.F. Santos, L.C. Giacomin, F. A.V. Castro, M.D. Pereira, A.B. Rocha, J.
A.L.C. Resende, M. Scarpellini, Synthesis, Characterization and Biological Activity of Gallium(III) Complexes
with non-symmetrical N2O-Donor Schiff Bases, Polyhedron (2016), doi: http://dx.doi.org/10.1016/j.poly.
2016.12.020
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Synthesis, Characterization and Biological Activity of
Gallium(III) Complexes with non-symmetrical N2O-Donor Schiff
Bases
Diego S. Padilhaa, Yan F. Santosa, Letícia C. Giacomina, Frederico A. V.
Castroa, Marcos D. Pereiraa, Alexandre B. Rochaa, Jackson A. L. C. Resende,b,c
and Marciela Scarpellinia*
a Instituto de Química, Universidade Federal do Rio de Janeiro, 21945-970 Rio
de Janeiro, RJ, Brazil
b Instituto de Química, Universidade Federal Fluminense, 24020–150, Niterói,
RJ, Brazil
c Instituto de Ciências Exatas e da Terra, Centro Universitário do Araguaia,
Universidade Federal do Mato Grosso, 786000-000 , Barra do Garças, MT,
Brasil
*Corresponding Author: phon.: +55 21 3938 7742; fax: +55 21 3938 7559
e-mail address: marciela@iq.ufrj.br
1

Abstract
Cisplatin and its analogues are some of the most widely used anticancer
drugs. However, toxicity and resistance have limited their use, and gallium
compounds have emerged as an alternative, due to their remarkable antitumor
activity and low toxicity. In this paper, we report four new mononuclear
gallium(III) complexes with the general formula [Ga(bhi-R)2]+, where bhi-R (R =
–NO2, –Br, –Cl and –OCH3) is the deprotonated form of imidazole/phenol-
containing tridentate Schiff bases. The molecular structures of C1–C4 were
solved by single crystal X-ray diffraction revealing mononuclear and
isostructural complexes, with the metal center in distorted octahedral
geometries. In all cases, the gallium(III) is coordinated to two ligand molecules
arranged in meridional fashion, through the imidazole and imino nitrogen atoms,
and to the phenolate oxygen atom. Complexes C1–C4 were also characterized
in solution by spectroscopic and spectrometric techniques. DFT calculations
were performed to assist the interpretation of experimental results, and also
allowed to establish the reasons why only meridional isomers were obtained.
The stability of all complexes was evaluated in PBS buffer (1% DMSO) over
24h, and results indicate partial hydrolysis in this period. Finally, the biological
activity of C1–C4 was evaluated on human tumor cell lines MCF-7 (breast
adenocarcinoma), A-549 (lung adenocarcinoma epithelial) and PC-3 (prostate
adenocarcinoma). Complex C1 was the most active, being effective and
selective for A-549. The IC50 (94.12
(135.10 6.50) obtained for cisplatin, which was tested as a metallodrug
reference under the same experimental conditions.
4.62) of C1 is lower than the value
2

Keywords: Gallium complex, metallodrug, DFT calculation, X-ray crystal
structure, tumor cell, MCF-7, A-549, PC-3 .
1. Introduction
Platinum-based anticancer agents have been successfully used in the
clinic for the last 30 years. However, undesirable nephrotoxicity and resistance
remain a challenge [1, 2]. In attempt to overcome these drawbacks, many
researchers have focused on the development of cytotoxic compounds with
improved properties using several metals [3, 4]. Among them, gallium has
proved to be effective against some malignancies, even in the form of simple
salts [5, 6].
The remarkable activity of gallium salts is contrasted by its reduced oral
bioavailability, which was the main obstacle for the development of an oral
formulation, along with renal toxicity [6, 7]. This adverse feature of gallium salts
has led to the design of ligands capable to stabilize the gallium ion and prevent
hydrolysis reactions [8, 9]. Some gallium complexes are now in advanced
development stage, including gallium maltolate (Figure 1a) and KP46 (Figure
1b) [10, 11]. These complexes were designed to present improved oral
bioavailability and have overcome phases I and II clinical trials with promising
outcomes, encouraging researches on other classes of gallium complexes [12].
Recent work on the cytotoxic activity of gallium complexes of different
N,N,N–, N,N,O– and N,O,O–donor ligands includes pyridine [13, 14] and
pyrazole [15] Schiff bases, semicarbazones [16], hydrazones [17], and
thiosemicarbazones [18, 19].
3

Expanding the number of gallium compounds with antitumor activity is a
key step to achieve better understanding of the relevant properties involved in
their biological activity, and thus contribute to the knowledge necessary for an
effective drug design.
In this work, we report on four new gallium(III) complexes (Figure 1 c) of
non-symmetrical N,N,O–donor Schiff base tridentate ligands, and their activities
on the human tumor cell lines MCF-7 (breast adenocarcinoma), A-549 (lung
adenocarcinoma) and PC-3 (prostate adenocarcinoma). Furthermore, in order
to understand the reasons why only the meridional (mer) isomers were isolated,
DFT/B3LYP calculations were employed to optimize the geometries of
meridional (mer) and facial (fac) isomers for all complexes. DFT calculations
were also carried out as an auxiliary tool to understand structural and electronic
properties of complexes C1–C4.
2. Experimental section
2.1. General procedures
Ligands HL1–HL4 were obtained by modifications on previously described
methods for HL1 [20]. To the best of our knowledge, HL2-HL4 are
unprecedented. All other chemicals were analytical grade and used for
syntheses and analyses as received from commercial sources. The analyses of
C1–C4 were performed using only single crystalline samples. Infrared spectra
of ligands and complexes were recorded in the range of 4000 to 400 cm-1, in
KBr pellets (or thin films) on a Nicolet Magna FTIR-760, or in a Shimadzu
IRAffinity-1 spectrophotometer. Elemental analyses were obtained in a Thermo
Scientific FlashEA 1112 CHNS/O Elemental Analyzer. Molar conductivity
4

measurements were performed on a Bel Engineering BE510 conductivimeter.
Electronic spectra were recorded using Varian Cary 50-Bio
a
spectrophotometer, in methanol. Reactivity of C1–C4 in PBS buffer pH 7.4 (1%
DMSO) was evaluated by electronic spectroscopy, over 24h, in the same
equipment described above. Cyclic voltammetry experiments were performed
on a BAS Epsilon potentiostat, at room temperature (25 °C), in anhydrous DMF
and under argon atmosphere. A standard three-electrode cell consisting of
glass carbon working electrode, platinum wire auxiliary electrode and Ag/AgCl
pseudo-reference electrode were employed. LiClO4 [0,1 M] was used as
supporting electrolyte and the ferrocenium-ferrocene couple [21] was employed
as standard reference. Electrospray ionization-mass spectrometry (ESI-MS) of
complexes C1–C4 was carried out with a Perkin Elmer Flexar SQ 300 MS, in
1
the positive mode, in methanol. H (200, 300 or 500 MHz) and 13C (50, 75 or
125 MHz) NMR spectra were obtained on a Bruker-DTX 200 MHz, or on a
Bruker DRX 300MHz spectrophotometer in DMSO-d6 solutions and referenced
1
to TMS (tetramethylsilane). H-1H COSY experiments were also performed for
complexes C1–C4 (Figures S1-S4).
2.2. Synthesis of HL1–HL4
Ligands HL1–HL4 were synthesized by Schiff condensation of equimolar
amounts of the appropriate para-substituted aldehyde with histamine, using the
same general procedure. To a methanolic solution of histamine cooled over ice
bath, a methanolic solution of the aldehyde was added dropwise, under
continuous stirring. After addition, the reaction mixture was stirred over ice bath
for another 3 hours, when the solvent was removed under reduced pressure.
5

2-((2-(1H-imidazol-4-yl)ethylimino)methyl)-4-nitrophenol (Hbhi-NO2, HL1)
Starting from 0.844 g (5 mmol) of 5-nitrosalicylaldehyde, 0.926 g of HL1
were obtained as a yellow powder. Yield: 71%. FTIR (KBr, cm-1): (NHimidazole),
3430; ꢀ(C-HAr/Alif), 3146-2849; ꢀ(C=Nimine), 1643; ꢀ(C=N)/(C=C), 1613-1450;
1
(NO2assim), 1328; ꢀ(NO2sim), 1254; ꢀ(C-O), 1052; δ(CHAr), 890/830. H NMR
(DMSO-d6, 200 MHz): δ 8.67 (s, 1H), 8.38 (d, J = 3.0 Hz, 1H), 8.03 (dd, J = 9.7,
3.1 Hz, 1H), 7.79 (d, J = 1.0 Hz, 1H), 6.97 (d, 1H), 6.61 (d, J = 9.7 Hz, 1H), 3.94
(t, J = 6.6 Hz, 3H), 2.97 (t, J = 6.5 Hz, 3H).
2-((2-(1H-imidazol-4-yl)ethylimino)methyl)-4-bromophenol (Hbhi-Br, HL2)
Starting from 1.005 g (5 mmol) of 5-bromosalicylaldehyde, 1.104 g of HL2
were obtained as a yellow powder. Yield: 75%. FTIR (KBr, cm-1): (NHimidazole),
3432; ꢀ(C-HAr/Alif), 3084-2867; ꢀ(C=Nimine), 1629; ꢀ(C=N)/(C=C), 1573-1478;
1
δ(O-Hphenol), 1384; ꢀ(C-O), 1049; ꢀ(C-Br), 1046; δ(CHAr), 900/824. H NMR
(DMSO-d6, 200 MHz): δ 8.49 (s, 1H), 7.80 (s, 1H), 7.63 (d, J = 2.5 Hz, 1H), 7.44
(dd, J = 8.8, 2.6 Hz, 1H), 6.91 (s, 1H), 6.83 (d, J = 8.8 Hz, 1H), 3.85 (t, J = 6.7
Hz, 2H), 2.90 (t, J = 6.9 Hz, 2H).
2-((2-(1H-imidazol-4-yl)ethylimino)methyl)-4-chlorophenol (Hbhi-Cl, HL3)
Starting from 1.278 g (8 mmol) of 5-methoxysalicylaldehyde, 1.978 g of
HL3 were obtained as a light-brown oil. Yield: 99%. FTIR (KBr, cm-1): ꢀ(C-
HAr/Alif), 3094-2836; ꢀ(C=Nimine), 1638; ꢀ(C=N)/(C=C), 1574-1480; δ(CHAr),
1
874/822; δ(O-Hphenol), 1372; ꢀ(C-O), 1028; ꢀ(C-Cl), 1091. H NMR (300 MHz,
DMSO) δ 10.23 (s, 1H), 8.48 (s, 1H), 7.80 (s, 1H), 7.51 (d, J = 2.7 Hz, 1H), 7.33
6

(dd, J = 8.8, 2.7 Hz, 1H), 6.91 (s, 1H), 6.87 (d, J = 8.8 Hz, 1H), 3.84 (t, J = 6.8
Hz, 2H), 2.89 (t, J = 6.9 Hz, 2H).
2-((2-(1H-imidazol-4-yl)ethylimino)methyl)-4-methoxyphenol (Hbhi-OCH3, HL4)
Starting from 0.7802 g (5 mmol) of 5-methoxysalicylaldehyde, 0.9865 g of
HL4 were obtained as a yellow-brown oil.
Yield: 80%. FTIR (KBr,
cm-1): ꢀ(C-HAr/Alif), 3137-2836; ꢀ(C=Nimine), 1645; ꢀ(C=N)/(C=C), 1592-1433;
δ(O-Hphenol), 1392; ꢀ(C-O-Cether), 1272; ꢀ(C-O), 1035; δ(CHAr), 870/822. 1H NMR
(DMSO-d6, 200 MHz): δ 10.25 (s, 1H), 8.46 (s, 1H), 7.74 (s, 1H), 7.21 – 6.66
(m, 4H), 5.83 (s, 1H), 3.83 (t, J = 6.9 Hz, 2H), 3.70 (s, 3H), 2.88 (t, J = 7.0 Hz,
2H).
2.3 Synthesis of C1–C4
mer-[Ga(bhi-NO2)2]NO3, C1. Complex C1 was obtained by slow addition of
Ga(NO3)3•6H2O (1 mmol in 30 mL of methanol) to a solution of Hbhi-NO2 (2
mmol in 30 mL of methanol). The reaction mixture was stirred for 2 hours, when
the complex started to precipitate as a pale-yellow amorphous solid. This solid
was filtered off and washed with cold isopropanol and diethyl ether. Yellow
single crystals suitable for X-ray analysis were obtained by recrystallization in
methanol:acetonitrile (1:1). Yield after recrystallization: 103 mg, 16%. MP: >310
ºC. FTIR (KBr, cm-1): (NHimidazole), 3389; ꢀ(C-HAr/Alif), 3158-2930; ꢀ(C=Nimine),
1629; ꢀ(C=N)/(C=C), 1559-1430; (NO2assim), 1318 (NO2sim), 1270; ꢀ(C-O) 1037;
δ(CHAr), 878/832; δ(N-Onitrate), 833. Elemental analysis (calc. for C24H22GaN9O9,
FW: 650.21 g mol-1) calc. (found): C, 44.3 (43.8); H, 3.40 (3.50); N, 19.4 (19.2)
+
%. ESI-MS (CH3OH): m/z+ = 587,2070 (100%, [Ga(bhi-NO2)2]+, C24H22GaN8O6 ,
7

m/z+calc = 587,0918). ΛM = 14.7 Ω-1 mol-1 cm2 (1:1 electrolyte in DMSO) [22]. 1H
NMR (DMSO-d6, 200 MHz): δ 12.96 (s, 1H), 8.89 (s, 1H), 8.42 (d, J = 3.0 Hz,
1H), 8.09 – 7.93 (m, 2H), 7.11 (s, 1H), 6.58 (d, J = 9.3 Hz, 1H), 4.19 (dt, J =
25.9, 12.0, 7.7 Hz, 2H), 3.10 (t, J = 9.8 Hz, 2H). 13C NMR (50 MHz, DMSO-d6) δ
172.02, 171.64, 136.02, 135.78, 135.64, 132.36, 129.51, 122.20, 116.56,
114.47, 60.21, 24.58 (Figure S5).
mer-[Ga(bhi-Br)2]ClO4•3/2H2O•1/2CH3OH,
C2.
Complex
C2
was
synthesized by the slow addition of Ga(NO3)3•6H2O (1 mmol in 30 mL of
ethanol) to a solution of Hbhi-Br (2 mmol in 30 mL of ethanol), followed by the
addition of KOH (2 mmol in 20 mL of ethanol). After 3 hours under stirring, an
excess of NaClO4 was added to the reaction mixture. The resulting solution was
filtered to remove KNO3. Yellow single crystals suitable for X-ray analysis were
obtained from the mother liquor by slow evaporation at 15 °C, after 2 days.
Yield after recrystallization: 90 mg, 12%. MP: 255 ( 3) ºC. FTIR (KBr, cm-1):
(NHimidazole), 3390; ꢀ(C-HAr/Alif), 3154-2934; ꢀ(C=Nimine), 1629; ꢀ(C=N)/(C=C),
1591-1466; (Cl-O), 1100; ꢀ(C-O) 1032; δ(CHAr), 880/827. Elemental analysis
(calc. for C25H28Br2ClGaN6O8; FW: 805.51 g mol-1) calc. (found): C, 37.3 (36.7);
H, 3.50 (3.50); N, 10.4 (10.4) %. ESI-MS (CH3OH): m/z+ = 654,9680 (100%,
+
[Ga(bhi-Br2)2]+, C24H22Br2GaN4O2 , m/z+
= 654,9401). ΛM = 75.2 Ω-1 mol-1
calc
1
cm2 (1:1 electrolyte in methanol) [22]. H NMR (DMSO-d6, 200 MHz): δ 12.71
(s, 1H), 8.56 (s, 1H), 7.81 (s, 1H), 7.44 (d, J = 2.8 Hz, 1H), 7.22 (dd, J = 9.0, 2.7
Hz, 1H), 7.02 (s, 1H), 6.39 (d, J = 8.9 Hz, 1H), 4.09 (dt, J = 34.0, 11.6, 8.2 Hz,
2H), 3.04 (dt, J = 24.3, 9.3, 5.3 Hz, 2H). 13C NMR (50 MHz, DMSO-d6) δ
8

170.85, 165.88, 137.08, 136.18, 135.29, 123.66, 119.19, 114.01, 104.61, 59.84,
24.79, 18.54 (Figure S6).
mer-[Ga(bhi-Cl)2]ClO4•CH3OH, C3. Complex C3 was obtained by slow
addition of Ga(NO3)3•6H2O (1 mmol in 30 mL of methanol) to a methanol
solution of ligand Hbhi-Cl (2 mmol), followed by the addition of KOH (2 mmol in
20 mL of ethanol). The reaction mixture was stirred for 16 hours and then an
excess of NaClO4 was added. Precipitated KNO3 was removed by filtration.
Yellow single crystals suitable for X-ray analysis were obtained by slow
evaporation of the mother liquor after 2 days, at 15 ºC. Yield after
recrystallization: 360 mg, 32%. MP: 238 ( 3) ºC. FTIR (KBr, cm-1): ꢀ(C-HAr/Alif),
3151-2936; ꢀ(C=Nimine), 1630; ꢀ(C=N)/(C=C), 1595-1466; (Cl-O), 1099; ꢀ(C-O),
1030; δ(CHAr), 875/827. Elemental analysis (calc. for C24H28Cl3GaN6O9; FW:
720.60 g mol-1) calc. (found): C, 40.0 (40.2); H, 3.92 (3.90); N, 11.66 (11.8) %.
+
ESI-MS (CH3OH): m/z+ = 567,1660 (100%, [Ga(bhi-Cl)2]+, C24H22Cl2GaN6O2 ,
m/z+
= 567.0995), ΛM = 76,1 Ω-1 mol-1 cm2 (1:1 electrolyte in methanol) [22].
calc
1H NMR (DMSO-d6, 300 MHz): δ 12.72 (s, 1H), 8.57 (s, 1H), 7.82 (s, 1H), 7.32
(d, J = 2.8 Hz, 1H), 7.13 (dd, J = 9.0, 2.8 Hz, 1H), 7.02 (s, 1H), 6.44 (d, J = 9.0
Hz, 1H), 4.30 – 3.92 (m, 2H), 3.16 – 2.89 (m, 2H). 13C NMR (125 MHz, DMSO-
d6) δ 171.44, 166.09, 136.72, 135.86, 135.01, 133.66, 123.77, 118.88, 118.07,
114.56, 60.39, 25.34 (Figure S7).
mer-[Ga(bhi-OCH3)2]ClO4•H2O•CH3OH, C4. Complex C4 was obtained by
the same procedure described for C3. Green single crystals suitable for X-ray
analysis were obtained by slow evaporation of the mother liquor after 1 day at
9

room temperature. Yield after recrystallization: 338 mg, 48%. MP: 244 ( 3) ºC.
FTIR (KBr, cm-1): ꢀ(C-HAr/Alif), 3150-2953; ꢀ(C=Nimine), 1632; ꢀ(C=N)/(C=C),
1584-1437; ꢀ(C-O-Cether), 1266; (Cl-O), 1092 ꢀ(C-O), 1025; δ(CHAr), 866/811.
Elemental analysis (calc. for C27H34ClGaN6O10; FW: 707.77 g mol-1) calc.
(found): C, 45.8 (44.3); H, 4.8 (4.9); N, 11.8 (11.7) %. ESI-MS: m/z+ = 557,1000
+
(100%, [Ga(bhi-OCH3)2]+, C26H28GaN6O4 , m/z+
= 557,1423). ΛM = 77.9 Ω-1
calc
1
mol-1 cm2 (1:1 electrolyte in methanol) [22]. H NMR (DMSO-d6, 200 MHz): δ
12.61 (s, 1H), 8.55 (s, 1H), 7.75 (s, 1H), 7.00 (s, 1H), 6.90 – 6.71 (m, 2H), 6.40
(d, J = 9.0 Hz, 1H), 4.12 (dt, J = 42.6, 11.7, 11.7 Hz, 2H), 3·64 (s, 3H), 3.15 –
2.87 (m, 2H). 13C NMR (50 MHz, DMSO-d6) δ 171.18, 161.93, 148.48, 136.34,
135.06, 123.87, 122.07, 116.30, 115.75, 113.69, 59.66, 55.47, 48.59, 24.97
(Figure S8).
2.3. X-ray crystallography
Suitable single crystals for X-ray diffraction experiments of C1 were
obtained by recrystallization in methanol:acetonitrile (1:1) solution. For
complexes C2–C4, crystals were obtained by slow evaporation of mother
liquors. Selected crystals of C1, C2 and C4 were analyzed on a Bruker Nonius
Kappa CCD diffractometer using MoKα radiation (λ = 0.71073) at room
temperature. The data collection was performed using Collect [23] and reduced
using EvallCCD software [24]. For complex C3, the data acquisition was
performed in a Bruker D8 Venture diffractometer with MoKα radiation, at room
temperature. The ApexII software [25] was used in the processing and data
reduction. Structures were solved by direct methods using SHELXS-13 software
[26]. All non-hydrogen atoms were refined anisotropically on F2 by the full-
10

matrix least-square technique using SHELXL-13 software [26]. Hydrogen atoms
were generated geometrically and treated by a mixture of independent and
constrained refinement.
2.4. In vitro biological assays
The human cancer cell lines A-549, MCF-7, and PC-3, were cultured in
DMEM supplemented with 10% fetal bovine serum (FBS), 2 mmol/L of L-
glutamine, 100 U/mL of penicillin and 100 µg/mL of streptomycin (Invitrogen,
Carlsbad, CA, USA), at 37 °C, and 5% CO2 humidified atmosphere. Complexes
C1–C4 were initially dissolved in DMSO at a final concentration of 50 mmol/L,
and then further diluted directly in DMEM prior the tests. In addition, final DMSO
concentration did not exceed 1% during cytotoxic tests. For determination of
C1–C4 toxicity, 1.25 × 104 cells/mL were seeded in quadruplicate into 96-well
plates and incubated at 37 °C, in a 5% CO2 humidified atmosphere (Tecnal, TE-
399). The culture medium was replaced by a freshly prepared DMEM medium
containing increasing concentration of C1–C4 (10, 25, 50, 75, 100, 150 and 200
µmol/L), and incubated as described above. After 24 h, cells were washed and
0.100 mL of a 0.5 mg/mL solution of [3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide] (MTT) in DMEM FBS-free was added. Then, the
plates were incubated for 3 h with MTT and after this period, the solution was
removed and formazan crystals were solubilized with 0.1 mL/ well DMSO.
Survival was analyzed spectrophotometrically at 570 nm using a HT
PowerWave XS microplate reader (BioTeK). IC50 was calculated by a non-linear
regression from log transformation of the dose-response curves. Data were
obtained from three independent experiments [27].
11

2.5. DFT calculations
Complexes geometries were optimized with cc-pVDZ basis set and zero
point corrections were performed in each case, at the equilibrium geometries.
Since heavy atoms are present, the influence of relativistic effects was also
studied at the equilibrium geometry by the Douglas-Kroll-Hess Hamiltonian
corrected to third order [28, 29]. In the last case, a larger basis set cc-pVTZ-DK
was used, which is more appropriated to this kind of calculation. All calculations
were performed with GAMESS suite of programs.
3. Results and Discussion
3.1 Synthesis and Characterization
Ligands HL1–HL4 were synthesized in desirable purity and yield by
condensation reactions of the appropriate para-substituted salicylaldehyde with
histamine, using modifications on previously reported methodology for HL1 [20].
Ligands HL1 and HL2 were obtained as yellow powders, while HL3 and HL4
were isolated as light-brown and yellow-to-brown oils, respectively, which
solidified after two days in desiccator. Complexes C1–C4 were obtained in low
to moderate yields by treating each ligand with Ga(NO3)3•6H2O (2:1 molar
ratio), in methanol or ethanol, in the presence of an auxiliary base (KOH). In the
absence of KOH, the complexation reaction leads to a mixture of the desired
complex and the nitrate salt of the protonated ligand. These species were
characterized by X-ray single crystal crystallography and data are not shown
here. This indicates that part of the ligand is acting as a base to deprotonate the
phenol and assisting the complexation reaction.
12

Characterization of C1–C4, including single crystal X-ray diffraction,
supports the obtaining of mononuclear and isostructural gallium(III) complexes,
in pseudo-octahedral environments. The molecular ion peak [Ga(bhi-R)2]+ was
observed on the mass spectrum of compounds C1–C4, in good agreement
between experimental and calculated isotopic distribution patterns. Molar
conductivity values for C1–C4 are in the range found for 1:1 electrolytes [22], as
expected for monocationic complexes. Compounds C1–C4 are soluble in most
of polar solvents, but C1 is only slightly soluble in water and methanol.
3.2 Crystal structures and structural properties
Crystal structures of C1–C4 were solved by single crystal X-ray diffraction.
Crystallographic and refinement data are shown in Table 1. Thermal ellipsoid
drawing for C1 is presented in Figure 2, and in Figures S9 – S11 for C2–C4.
Results for C1–C4 show the gallium(III) center surrounded by two molecules of
the respective ligand in a meridional pseudo-octahedral environment. Ligands
coordinate to the metal center through the imidazole and the imine nitrogen
atoms,
and
the
phenolate
oxygen,
with
an
[Ga<Nimine1Nimine2><Nimidazole1Ophenol2> <Nimidazole2Ophenol1>] arrangement [30, 31].
Complexes C1 and C2 crystallize in the monoclinic crystal system, belonging to
the P21/n space group. Complex C4 is also in the monoclinic system, but in the
Cc space group. Complex C3 crystallizes in the orthorhombic crystal system, in
Pbca space group. The asymmetric unit of C1 is composed by one molecule of
the cation complex [Ga(bhi-NO2)2]+ and one nitrate molecule as counter ion.
The nitrate forms hydrogen bonds with the N–H group of the imidazole ring
(Figure 2) that are relevant to uphold the crystal lattice. For complex C2, the unit
13

cell shows the presence of two crystallographically independent [Ga(bhi-Br)2]+
molecules that are slightly different in bond distances and angles. Two
perchlorate anions, three water and one methanol molecules are also observed.
In the case of complexes C3 and C4 the structure exhibits a single molecule of
the cation complex surrounded by one perchlorate and one methanol. For
complex C4, a water molecule is also present. Complex C3 shows a vibrational
crystallographic disorder in the ethylene group of one of the bhi-Cl ligands and
a rotational disorder in the perchlorate counter ion. Complexes C1–C4 show
comparable bond lengths and angles, with average distances of 1.940 Å for
Ga–Ophenol, 2.056 Å for Ga–Nimine and 2.074 Å for Ga–Nimidazole. These values
are in good agreement with similar gallium(III) complexes reported in the
literature [13, 15, 31]. Correlations between the Hammett parameter of phenol
ring substituents in para position (σp) and Ga–Nimine or Ga–Ophenol bond lengths
were observed. For Ga–Nimine, the increase in electron withdrawing effect
(–OCH3 < –Br ≈ –Cl < –NO2) leads to a decrease in the bond length, while the
opposite effect is observed for Ga–Ophenol (Table 2). Shortening of C–O bond
lengths was also observed. This trend can be illustrated by the decrease of pKa
values for para-substituted phenols as the electron withdrawing effect
increases. This effect was experimentally and theoretically observed before [32,
33], where good agreement between σp values and pKa was found.
As already mentioned, DFT calculations were carried out in order to
understand the preference for the meridional coordination mode of the
tridentate ligand observed here. Meridional (mer) and facial (fac) isomers of
gallium(III) complexes with HL1–HL4 were evaluated and the calculated relative
energies are presented in Table 3. In all cases, the mer isomer is the most
14

stable. The relative energy values in column A are the difference of total
energies calculated at equilibrium geometry. In column B, relative energies
were calculated by the same procedure adopted in column A, but total energies
were corrected for relativistic effects through the Douglas-Kroll-Hess
Hamiltonian. In column C, relative energies were calculated by the same
procedure, but total energies were corrected with zero point energy (ZPE), i.e.
taking into account the vibrational ground state energy. Finally, in column D,
both relativistic and ZPE corrections were considered. Differences in relative
energies depend on the correction applied but, in all cases, these differences
are higher than 30 Kcal/mol. This is a remarkable value and is higher than that
found in other cases where mer/fac relative energies were calculated [31].
Figure S12 presents the final optimized structures of mer and fac isomers for
C1. It can be observed that the fac isomer presents large steric constraints that
enforce the break of one Ga–Nimine bond. It results in a less stable
pentacoordinated species, whereas the mer isomer remains six coordinated.
The absence of one chemical bond explains such a large difference in energy
between the two isomers and may explain the reason why the fac isomer was
not able to be isolated.
3.3 Spectroscopic and redox behavior
1H NMR data for ligands and complexes are in good agreement with
proposed structures. The presence of a single set of signals (Figure S13) is an
indicative of the complexes stability in DMSO solution, since no sign of ligand
release was observed. 1H-1H COSY (Figures S1–S4) experiments were carried
out for complexes C1–C4 and results were helpful to perform the assignments.
15

Comparing the data (Table S1) of free and complexed ligands, a downfield
shift is observed for hydrogen atoms H5 and H7, indicating that there is
coordination with nearby nitrogen atoms of the imidazole and imine groups in
solution [34]. The same behavior is observed for H4, but less pronounced. On
the other hand, an inverse behavior occurs for H3 and an upshift is detected for
C2-C4 and is almost null for C1. In general, the –NO2 promotes a downshift of
1
the HNMR spectrum compared to the other substituents in this series of
complexes. For –Br, –Cl and –OCH3 no clear trends are observed for the
chemical shifts of H4, H5, H7 and H8.
Electronic spectra of HL1–HL4 (Figure S14) and complexes C1–C4
(Figure 3) were recorded in methanol and Table 4 summarizes the data. A
strong band around 400 nm is observed in the spectra of HL1–HL4 and is blue-
shifted in the spectra of C1–C4. This band can be attributed to π→π* intraligand
charge transfer transitions (ILCT) and, as its position is sensitive to the nature of
phenol ring substituent, it might involve the imine moiety. In both, free and
complexed compounds, this band undergoes a redshift as the electron
withdrawing power of the substituent decreases. This band ranges from 350 nm
for C1 (–NO2) to 370 nm, for C2 (–Br) and C3 (–Cl), and 390 nm for C4 (–
OCH3). The same behavior is observed for the band around 265 nm that ranges
from 250 nm for C1 to 275 for C4. All the bands are attributed to intraligand
charge transfer transition processes since the gallium(III) ion is a d10 system
and do not present metal involved charge transfer transitions. Hyperchromic
effect is observed for all bands after coordination.
Considering that redox processes are one of the modes of action of
cytotoxic substances [35], cyclic voltammetry experiments were performed to
16

assess and compare the redox behavior of free and complexed ligands, since
gallium(III) is well known to be redox inactive in biological systems. DMF was
used as solvent due to its broader work window, when compared to methanol or
water. Voltammograms were recorded in cathodic scans, at several scan rates,
and in the potential range of +1.0 to -2.5 V using complexes concentration of
1.0 x 10-3 mol L-1. Voltammograms of C1–C4 are presented in Figure S15 and
the data are summarized in Table 4. The CV of all ligands and complexes show
an irreversible reduction process around -2.4 V vs Fc+/Fc (Peak B). This
process undergoes cathodic shift (more negative values) of about 0.1 V vs
Fc+/Fc after coordination, unless for HL2 and C2. The similarities among the
values suggest it should arise from the reduction of the imine. This group is
common to all compounds and seems not to be affected by the substituents in
the phenol ring. HL1 and C1 present an additional irreversible reduction around
-1.9 V vs Fc+/Fc (Peak A). It can be tentatively assigned to the one-electron
•-
reduction of –NO2 to the nitro radical anion NO2 [36]. Irreversible oxidation
processes are observed for ligands and complexes. Peak C observed for C2 in
-0.9 V vs Fc+/Fc (Figure S15) is tentatively attributed to the oxidation of the
species containing the NO2•- group. In all complexes, peaks D and E are subtle
to the substituent in the phenol ring, and should be related to the oxidation of
the two phenol groups yielding phenoxyl radicals [37]. However, no correlation
was found with the σ Hammet parameters and steric effects of the substituents
p
can overcome the electronic ones.
17

3.4 Reactivity
Considering the strong tendency of gallium(III) complexes to hydrolyze in
solution, C1–C4 were spectroscopically monitored over 24h, in PBS buffer, pH
7.4, containing 1% DMSO, at 1.0 × 10-5 mol L-1. At the region around 400 nm, a
decrease in the absorption is observed for all complexes, as shown in Figure
S16. Results can be rationalized in terms of Hammett parameter (σp) and
hydrolysis. As σp increases, the tendency to undergo hydrolysis reactions
reduces. For complex C1 (σp = 0.78), a decrease of 17% in the absorption was
observed, while 76% was found for complex C4 (σp = -0.27). These results
suggest that gallium(III) complexes with electron withdrawing ligands are prone
to be more stable in the biological media, which could lead to higher biological
activities. In addition, lower hydrolysis rates are desirable for application of
gallium-based chemotherapeutics, which should be effective in short time
exposure.
3.5 Biological activity
Human cancer cell lines (MCF-7, PC-3 and A-549) were exposed to
increasing concentrations of ligands HL1–HL4, complexes C1–C4 and
Ga(NO3)3•6H2O over 24 h. This time exposure was chosen taking into account
the stability/hydrolysis data discussed above. The anticancer drug cisplatin was
used as standard metallodrug in cytotoxic assays. After the time exposure, cell
survival was evaluated and data are presented in Figure 4. At low
concentrations, none of the ligands and complexes significantly affects cell
survival, which only occurs above 150 µM. Thus, this value was taken as the
18

lower threshold for comparisons. For MCF-7 cell line (Figure 4A and D), at 150
µM, the anticancer activity of compounds follow the order C4 < C2 < C3 < C1 <
cisplatin, and Ga(NO3)3•6H2O < HL1 < HL2 ≈ HL4 < HL3. Considering the PC-
3 cell line (Figure 4B and E) at the same drug concentration, the order C4 < C3
≈ cisplatin < C2 < C1, and Ga(NO3)3•6H2O < HL3 < HL1 ≈ HL2 ≈ HL4 is
observed. Finally, in A-549 cell line (Figure 4C and F), the cytotoxic profile of
tested compounds is C2 < C3 < C4 < cisplatin < C1, and HL1 ≈ HL2 ≈ HL4 <
Ga(NO3)3•6H2O < HL3. Then, considering data at 150 µM of cell exposure, C1
is the most active complex against all lineages, showing cell survivals of 61.5
2.1 % in MCF-7, 53.5 3.5 % in PC-3 and 40.9 0.5 % in A-549. For PC-3 and
A-549 cell lines, C1 proved to be even better than cisplatin, which displayed
survival rates of 70.0 4.5 and 46.5 3.7 % for PC-3 and A-549, respectively.
However, analyzing IC50 (Table 5), all tested compounds presented values
higher than 200 µM for both MCF-7 and PC-3 cell lines. In the case of A-549,
results point to an effectiveness and selectivity of C1, with lower IC50 (94.1
4.6) than cisplatin (134.3 7.6). When the activities of complexes are compared
to that of the free ligands, a significant increase after coordination is observed
only for C1, which presented an average decrease of 35% in survival rates. For
C2 – C4 the difference is within the standard deviation.
The higher activity of C1 compared to the other tested compounds might
be related to the presence of the –NO2 substituent, which is known to undergo
redox reactions in the biological media, generating cytotoxic products [38].
Other possible explanation for the enhanced activity of complex C1 is its low
solubility in polar solvents along with its lower tendency to undergo hydrolysis
reactions, that might lead to a higher cell uptake of C1 compared to the other
19

complexes. These IC50 values are higher than those reported for other
gallium(III) complexes tested against the same cell lines, but differences in time
incubation must be taken into account. Indeed, we believe that the activity
strength of an antitumor drug is better evaluated during short exposure times,
when it is expected to kill most of tumor cells. Gallium(III) complexes with
semicarbazones ligands of salicylaldehyde derivatives [16] showed IC50 values
between 23 and 164 µM after 72h of incubation. Values in the range of 13.3 to
52 µM were also observed after 72h for similar gallium(III) complexes with
tridentate ligands pyrazol-imino-phenol type reported by Silva and coworkers
[15].
4. Conclusions
In this work four new mononuclear and isostructural gallium(III) complexes
with non-symmetric tridentade ligands were obtained. Complexes C1–C4 show
ligands coordinated in a meridional fashion, which was found to be the more
thermodynamically favorable configuration, according to DFT calculations. The
redox and electronic behavior of C1–C4 are ligand-centered and does not
involve the gallium(III) ion. The reduction of imine to amine is suggested to be
the main observed redox process. Reduction of the NO2 group was also
observed and oxidation processes are tentatively attributed to the generation of
phenoxyl radicals.
Complexes C1–C4 proved not to be highly cytotoxic against the tested
tumor lineages MCF-7 and PC-3. Complexes C1 and C4 were the most active
compounds and capable to inhibit the A-549 cell growth in concentrations lower
than 94.12 4.62 µM and 174.70 8.41 µM, respectively. Complexes C1 and
20

C4 were active only against A-549 cell line, hence indicating a possible
selectivity. In addition, C1 is more active than the clinically standard metallodrug
cisplatin (IC50 = 135.10 6.5 µM) in the same experimental conditions. The
nature of the phenol ring substituent was found to be relevant for the cytotoxic
activity, however the mechanisms of action require further evaluation.
Future directions include the obtaining of novel gallium complexes
containing the related amine ligands and the replacement of imidazole ring to
pyridine in order to evaluate the effect of these modifications on the antitumor
activity. These results are promising and can guide further research aiming at
the development of gallium-based antitumor metallodrugs.
Appendix A. Supplementary data CCDC 1441117 - 1441120 contain the
supplementary crystallographic data for C1 - C4. These data can be obtained
free of charge via http://www.ccdc.cam.ac.uk/conts/retrieving.html, or from the
Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ,
UK; fax: (+44) 1223-336-033; or e-mail: deposit@ccdc.cam.ac.uk.
5. Acknowledgements
The authors are grateful to Professors Nicolás A. Rey and Ricardo Aucélio
(PUC-RJ) for elemental analyses, Professor Mauricio Lanznaster (UFF) for ESI-
MS analyses, and grants from Capes, CNPq (476269/2011-3) and Faperj (E-
26/111.177/2011, E-26/110.157/2014) Financial Agencies.
21

6. References
[1] Graf, N., & Lippard, S. J. (2012). Redox activation of metal-based prodrugs
as a strategy for drug delivery. Advanced drug delivery reviews, 64(11), 993-
1004. ( DOI: 10.1016/j.addr.2012.01.007)
[2] Gianferrara, T., Bratsos, I., & Alessio, E. (2009). A categorization of metal
anticancer compounds based on their mode of action. Dalton Transactions,
(37), 7588-7598. (DOI: 10.1039/B905798F)
[3] Kaluderovic, G. N., & Paschke, R. (2011). Anticancer metallotherapeutics in
preclinical development. Current medicinal chemistry, 18(31), 4738-4752.
(DOI: http://dx.doi.org/10.2174/092986711797535308)
[4] Komeda, S., & Casini, A. (2012). Next-generation anticancer metallodrugs.
Current
topics
in
medicinal
chemistry, 12(3),
219-235.
(DOI: http://dx.doi.org/10.2174/156802612799078964)
[5] Bernstein, L. R. Therapeutic Gallium Compounds. In: Gielen, M., & Tiekink,
E. R. (Eds.). (2005). Metallotherapeutic drugs and metal-based diagnostic
agents: the use of metals in medicine. Padstow: John Wiley & Sons, 2005. p.
259-278. (DOI: 10.1002/0470864052.ch14)
[6] Chitambar, C. R. (2012). Gallium-containing anticancer compounds. Future
medicinal chemistry, 4(10), 1257-1272. (DOI: 10.4155/fmc.12.69)
[7] Dabrowiak, J. C. Metals in Medicine. Chippenham: John Wiley & Sons,
2009. p. 177-181. (DOI: 10.1002/9780470684986)
22

[8] Bratsos, I.; Gianferrara, T.; Alessio, E.; Hartinger, C. G.; Jakupec, M. A.;
Keppler, B. K. Ruthenium and Other Non-platinum Anticancer Compounds. In:
Alessio, E. (Ed.). (2011). Bioinorganic medicinal chemistry. Weinheim: Wiley-
VHC, 2011. p. 151-174. (DOI: 10.1002/9783527633104.ch5)
[9] Lessa, J. A., Parrilha, G. L., & Beraldo, H. (2012). Gallium complexes as
new promising metallodrug candidates. Inorganica Chimica Acta, 393, 53-63.
(DOI: 10.1016/j.ica.2012.06.003)
[10] Chitambar, C. R., Purpi, D. P., Woodliff, J., Yang, M., & Wereley, J. P.
(2007). Development of gallium compounds for treatment of lymphoma: gallium
maltolate, a novel hydroxypyrone gallium compound, induces apoptosis and
circumvents lymphoma cell resistance to gallium nitrate. Journal of
Pharmacology
and
Experimental
Therapeutics, 322(3),
1228-1236.
(DOI: 10.1124/jpet.107.126342)
[11] Bonchi, C., Imperi, F., Minandri, F., Visca, P., & Frangipani, E. (2014).
Repurposing of gallium‐based drugs for antibacterial therapy. Biofactors,40(3),
303-312. (DOI: 10.1002/biof.1159)
[12] Enyedy, É. A., Dömötör, O., Bali, K., Hetényi, A., Tuccinardi, T., & Keppler,
B. K. (2014). Interaction of the anticancer gallium (III) complexes of 8-
hydroxyquinoline and maltol with human serum proteins. JBIC Journal of
Biological Inorganic Chemistry, 20(1), 77-88.
(DOI: 10.1007/s00775-014-1211-9)
[13] Shakya, R., Peng, F., Liu, J., Heeg, M. J., & Verani, C. N. (2006).
Synthesis, structure, and anticancer activity of gallium (III) complexes with
23

asymmetric tridentate ligands: growth inhibition and apoptosis induction of
cisplatin-resistant neuroblastoma cells. Inorganic chemistry, 45(16), 6263-6268.
(DOI: 10.1021/ic060106g)
[14] Chen, D., Frezza, M., Shakya, R., Cui, Q. C., Milacic, V., Verani, C. N., &
Dou, Q. P. (2007). Inhibition of the proteasome activity by gallium (III)
complexes contributes to their anti–prostate tumor effects. Cancer
research,67(19), 9258-9265. (DOI: 10.1158/0008-5472.CAN-07-1813)
[15] Silva, F., Marques, F., Santos, I. C., Paulo, A., Rodrigues, A. S., Rueff, J.,
& Santos, I. (2010). Synthesis, characterization and cytotoxic activity of gallium
(III) complexes anchored by tridentate pyrazole-based ligands.Journal of
inorganic biochemistry, 104(5), 523-532. (DOI: 10.1016/j.jinorgbio.2010.01.003)
[16] Gambino, D.; Fernández, M.; Santos, D.; Etcheverría, G. A.; Piro, O. E.;
Pavan, F. R.; Leite, C. Q. F.; Tomaz, I.; Marques, F. (2011). Searching for
gallium bioactive compounds: Gallium (III) complexes of tridentate
salicylaldehyde semicarbazone derivatives. Polyhedron, 30(7), 1360-1366.
(DOI: 10.1016/j.poly.2011.02.037)
[17] Machado, I.; Fernández, M.; Becco, L.; Garat, B.; Brissos, R.F. ; Zabarska,
N.; Gamez, P.; Marques, F.; Correia, I.; Costa Pessoa, J.; Gambino, D. (2014).
New metal complexes of NNO tridentate ligands: Effect of metal center and co-
ligand on biological activity. Inorganica Chimica Acta, 420, 39-46.
(DOI: 10.1016/j.ica.2013.10.022)
[18] Kumar, K.; Schniper, S.; Gonzalez-Sarrías, A.; Holder, A. A.; Sanders, N.;
Sullivan, D.; Jarrett, W. L.; Davis, K.; Bai, F.; Seeram, N. P.; Kumar, V. (2014).
24

Highly potent anti-proliferative effects of a gallium (III) complex with 7-
chloroquinoline thiosemicarbazone as a ligand: Synthesis, cytotoxic and
antimalarial evaluation. European journal of medicinal chemistry, 86, 81-86.
(DOI: 10.1016/j.ejmech.2014.08.054)
[19] Parrilha, G. L.; Ferraz, K. S. O.; Lessa, J. A.; Oliveira, K. N.; Rodrigues, B.
L.; Ramos, J. P.; Souza-Fagundes, E. M.; Ott, I.; Beraldo, H. (2014). Metal
complexes with 2-acetylpyridine-N (4)-orthochlorophenylthiosemicarbazone:
Cytotoxicity and effect on the enzymatic activity of thioredoxin reductase and
glutathione reductase. European journal of medicinal chemistry, 84, 537-544.
(DOI: 10.1016/j.ejmech.2014.07.055)
[20] Baldwin, M. J., Kampf, J. W., Kirk, M. L., & Pecoraro, V. L. (1995).
Structural and Magnetic Studies of Manganese (II) Complexes of the Imidazole-
Containing Ligand 5-NO2-salimH [5-NO2-salimH2= 4-(2-((5-nitrosalicylidene)
amino) ethyl) imidazole] with Varying Nuclearity. Inorganic Chemistry, 34(21),
5252-5260. (DOI: 10.1021/ic00125a026)
[21] Gagne, R. R., Koval, C. A., & Lisensky, G. C. (1980). Ferrocene as an
internal standard for electrochemical measurements. Inorganic Chemistry,19(9),
2854-2855. (DOI: 10.1021/ic50211a080)
[22] Geary, W. J. (1971). The use of conductivity measurements in organic
solvents for the characterisation of coordination compounds. Coordination
Chemistry Reviews, 7(1), 81-122. (DOI: 10.1016/S0010-8545(00)80009-0)
[23] Bruker (2008). COLLECT. Bruker AXS Inc., Madison, Wisconsin, USA.
25

[24] Duisenberg, A. J., Kroon-Batenburg, L. M., & Schreurs, A. M. (2003). An
intensity evaluation method: EVAL-14. Journal of applied crystallography,36(2),
220-229. (DOI: 10.1107/S0021889802022628)
[25] Bruker (2014). ApexII. Bruker AXS Inc., Madison, Wisconsin, USA.
[26] Sheldrick, G. M. (2007). A short history of SHELX. Acta Crystallographica
Section
A:
Foundations
of
Crystallography, 64(1),
112-122.
(DOI: doi:10.1107/S0108767307043930)
[27] Freshney, R. (2005). Culture of animal cells: A manual of basic technique
(5th ed.). Hoboken, N.J.: Wiley-Liss.
[28] Douglas, M., & Kroll, N. M. (1974). Quantum electrodynamical corrections
to the fine structure of helium. Annals of Physics, 82(1), 89-155.
(DOI: 10.1016/0003-4916(74)90333-9)
[29] Hess, B. A. (1986). Relativistic electronic-structure calculations employing a
two-component no-pair formalism with external-field projection operators.
Physical Review A, 33(6), 3742. (DOI: 10.1103/PhysRevA.33.3742)
[30] Miessler, G.L; Tarr, D. A. Inorganic Chemistry; Pearson-Prentice Hall:
Upper Saddle River, NJ, 2004; pp 311-315.
[31] Imbert, C., Hratchian, H., Lanznaster, M., Heeg, M., Hryhorczuk, L.,
Mcgarvey, B., Verani, C. Influence of Ligand Rigidity and Ring Substitution on
the Structural and Electronic Behavior of Trivalent Iron and Gallium Complexes
with Asymmetric Tridentate Ligands. Inorganic Chemistry, 44, 7414-7422, 2005.
(DOI: 10.1021/ic050658j)
26

[32] Gross, K. C., & Seybold, P. G. (2001). Substituent effects on the physical
properties and pKa of phenol. International Journal of Quantum
Chemistry,85(4‐5), 569-579. (DOI: 10.1002/qua.1525)
[33] Liptak, M. D., Gross, K. C., Seybold, P. G., Feldgus, S., & Shields, G. C.
(2002). Absolute p K a determinations for substituted phenols. Journal of the
American Chemical Society, 124(22), 6421-6427. (DOI: 10.1021/ja012474j)
[34] Souza, E. T., Castro, L. C., Castro, F. A. V., do Canto Visentin, L., Pinheiro,
C. B., Pereira, M. D., Machado, S. P., Scarpellini, M. (2009). Synthesis,
characterization and biological activities of mononuclear Co (III) complexes as
potential
bioreductively
activated
prodrugs. Journal
of
inorganic
biochemistry, 103(10), 1355-1365. (DOI: 10.1016/j.jinorgbio.2009.07.008)
[35] Wilson, W. R., & Hay, M. P. (2011). Targeting hypoxia in cancer therapy.
Nature Reviews Cancer, 11(6), 393-410. (DOI: 10.1038/nrc3064)
[36] Ramogida, C. F., Pan, J., Ferreira, C. L., Patrick, B. O., Rebullar, K., Yapp,
D. T. T., Lin, Kuo-S., Adam, M. J., Orvig, C. (2015). Nitroimidazole-Containing
H2dedpa and H2CHXdedpa Derivatives as Potential PET Imaging Agents of
Hypoxia with 68Ga. Inorganic Chemistry, 54, 4953-4965. (DOI:
10.1021/acs.inorgchem.5b00554)
[37] Lanznaster, M., Hratchian, H. P., Heeg, M. J., Hryhorczuk, L. M.,
McGarvey, B. R., Schlegel, H. B., Verani, C. N. (2006). Structural and Electronic
Behavior of Unprecedented Five-Coordinate Iron(III) and Gallium(III)
Complexes with a New Phenol-Rich Electroactive Ligand. Inorganic Chemistry,
45, 955-957. (DOI: 10.1021/ic050809i)
27

[38] Denny, W. A. (2001). Prodrug strategies in cancer therapy. European
Journal of Medicinal Chemistry, 36 (7), 577-595. (DOI: 10.1016/S0223-
5234(01)01253-3)
28

Tables
Table 1. Crystal data and structure refinement for complexes C1–C4.
C1
C2
C3
C4
Empirical formula
C24H22GaN9O9
650.22
C25H28Br2ClGaN6O8 C25H26Cl3GaN6O7
C27H33ClGaN6O10
706.76
Formula weight
805.52
293(2)
698.59
293(2)
Temperature/K
293(2)
293(2)
Crystal system
Monoclinic
P21/n
Monoclinic
P21/n
Orthorhombic
Pbca
Monoclinic
Cc
Space group
a/Å
b/Å
c/Å
α/°
β/°
γ/°
8.6545(17)
13.987(3)
21.911(4)
90
18.846(4)
18.893(4)
19.287(4)
90
16.0687(7)
17.3818(9)
20.7575(10)
90
13.675(3)
14.476(3)
15.612(3)
90
95.12(3)
90
118.65(3)
90
90
95.95(3)
90
90
Volume/Å3
2641.7(9)
4
6026(3)
8
5797.6(5)
8
3073.7(11)
Z
4
ρcalcg/cm3
1.635
1.776
1.601
1.527
µ/mm-1
1.114
3.712
1.281
1.048
F(000)
1328.0
3216.0
2848.0
1460.0
Crystal size/mm3
0.276 × 0.2 × 0.109 0.33 × 0.31 × 0.21
0.38 × 0.118 × 0.11 0.42 × 0.16 × 0.09
Radiation
MoKα
(λ = 0.71073)
MoKα
(λ = 0.71073)
MoKα
(λ = 0.71073)
MoKα
(λ = 0.71073)
2Θ range for data collection/° 4.734 to 50.694
6.04 to 50.7
3.97 to 52.046
7.546 to 50.698
Index ranges
-10 ≤ h ≤ 10,
-16 ≤ k ≤ 16,
-26 ≤ l ≤ 26
-22 ≤ h ≤ 22,
-22 ≤ k ≤ 22,
-23 ≤ l ≤ 22
-19 ≤ h ≤ 15,
-21 ≤ k ≤ 21,
-25 ≤ l ≤ 25
-16 ≤ h ≤ 16,
-15 ≤ k ≤ 17,
-18 ≤ l ≤ 18
Reflections collected
53899
32857
68332
8973
Independent reflections
4842
[Rint = 0.0564,
Rsigma = 0,0298]
10875
[Rint = 0.0777,
Rsigma = 0.1006]
5703
[Rint = 0.0803,
Rsigma = 0.0313]
5056
[Rint = 0.0631,
Rsigma = 0.0594]
Data/restraints/parameters
4842/0/388
10875/4/776
5703/12/446
4982/5/421
Goodness-of-fit on F2
1.174
1.013
1.120
1.067
Final R indexes [I>=2σ (I)]
R1 = 0.0375,
wR2 = 0.1016
R1 = 0.0583,
wR2 = 0.1115
R1 = 0.0535,
wR2 = 0.1005
R1 = 0.1397,
wR2 = 0.1291
R1 = 0.0554,
wR2 = 0.1318
R1 = 0.0910,
wR2 = 0.1561
R1 = 0.0463,
wR2 = 0.1012
R1 = 0.0678,
wR2 = 0.1249
Final R indexes [all data]
Largest diff. peak/hole / e Å-3 0.40/-0.44
0.81/-1.05
0.77/-0.55
0.54/-0.36
29