Exploring the cytotoxic activity of new phenanthroline salicylaldimine Zn(II) complexes

Article abstract of DOI:10.1016/j.jinorgbio.2019.110727

Zinc(II) complexes bearing N-salicylideneglycinate (Sal-Gly) and 1,10-phenanthroline (phen) or phenanthroline derivatives [NN = 5-chloro-1,10-phenanthroline, 5-amine-1,10-phenanthroline (amphen), 4,7-diphenyl-1,10-phenanthroline (Bphen) and 5,6-epoxy-5,6-

Full text of DOI:10.1016/j.jinorgbio.2019.110727

Accepted Manuscript
Exploring the cytotoxic activity of new phenanthroline
salicylaldimine Zn(II) complexes
Cristina P. Matos, Yemataw Addis, Patrique Nunes, Sonia
Barroso, Irina Alho, Marta Martins, António P.A. Matos,
Fernanda Marques, Isabel Cavaco, João Costa Pessoa, Isabel
Correia
PII:
S0162-0134(19)30090-X
DOI:
https://doi.org/10.1016/j.jinorgbio.2019.110727
Article Number:
Reference:
110727
JIB 110727
To appear in:
Journal of Inorganic Biochemistry
Received date:
Revised date:
Accepted date:
13 February 2019
27 May 2019
27 May 2019
Please cite this article as: C.P. Matos, Y. Addis, P. Nunes, et al., Exploring the cytotoxic
activity of new phenanthroline salicylaldimine Zn(II) complexes, Journal of Inorganic
Biochemistry, https://doi.org/10.1016/j.jinorgbio.2019.110727
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.

ACCEPTED MANUSCRIPT
Exploring the cytotoxic activity of new phenanthroline salicylaldimine
Zn(II) complexes
Cristina P. Matos,a Yemataw Addis,a,b Patrique Nunes,a Sonia Barroso,a,c Irina Alho,d Marta
Martins,d António P. A. Matos,e Fernanda Marques,f Isabel Cavaco,a,b João Costa Pessoa,a,* Isabel
Correiaa,*
a Centro de Química Estrutural, Departamento de Engenharia Química, Instituto Superior Técnico,
Universidade de Lisboa, Av Rovisco Pais, 1049-001 Lisboa, Portugal.
b Departamento de Química Bioquímica e Farmácia, Universidade do Algarve, Campus de
Gambelas, 8005-139 Faro, Portugal
c Current address: MARE – Marine and Environmental Sciences Centre, ESTM, Polytechnic
Institute of Leiria, Peniche, Portugal
d Instituto de Medicina Molecular - João Lobo Antunes, Faculdade de Medicina, Universidade de
Lisboa, Av. Professor Egas Moniz, 1649-028 Lisboa, Portugal.
e Centro de Investigação Interdisciplinar Egas Moniz, Campus Universitário, Quinta da Granja,
Monte de Caparica, 2829-511 Caparica, Portugal.
f Centro de Ciências e Tecnologias Nucleares, Instituto Superior Técnico, Universidade de Lisboa,
Estrada Nacional 10, 2695-066 Bobadela LRS, Portugal.
*Corresponding authors:
Isabel Correia: Tel +351-218419270; e-mail address: icorreia@tecnico.ulisboa.pt
João Costa Pessoa: Tel +351-218419269; e-mail address: joao.pessoa@tecnico.ulisboa.pt
Abstract
1

ACCEPTED MANUSCRIPT
Zinc(II) complexes bearing N-salicylideneglycinate (Sal-Gly) and 1,10-phenanthroline (phen) or
phenanthroline derivatives [NN = 5-chloro-1,10-phenanthroline, 5-amine-1,10-phenanthroline
(amphen),
4,7-diphenyl-1,10-phenanthroline
(Bphen)
and
5,6-epoxy-5,6-dihydro-1,10-
phenanthroline] are synthesized. Complexes formulated as [Zn(NN)2(H2O)2]2+ (NN = phen and
amphen), are also prepared. The cytotoxicity of the compounds is evaluated towards a panel of
human cancer cells: ovarian (A2780), breast (MCF7) and cervical (HeLa), as well as non-tumoral
V79 fibroblasts. All compounds display higher cytotoxicity than cisplatin (IC50=22.5±5.0 μM)
towards ovarian cells, showing IC50 values in the low micromolar range. Overall, all compounds
show higher selectivity for the A2780 cells than for the non-tumoral cells and higher selectivity
indexes (IC50(V79)/IC50(A2780) than cisplatin. [Zn(Sal-Gly)(NN)(H2O)] complexes induce caspase-
dependent apoptosis in A2780 cells, except [Zn(Sal-Gly)(Bphen)(H2O)], one of the most cytotoxic
of the series. The cellular uptake in the ovarian cells analyzed by Inductively Coupled Plasma
mass spectrometry indicates different Zn distribution profiles. Transmission electronic microscopy
shows mitochondria alterations and apoptotic features consistent with caspase activation; cells
incubated with [Zn(Sal-Gly)(amphen)(H2O)] present additional nuclear membrane alterations in
agreement with significant association with the nucleus. The increase of reactive oxygen species
and lipid peroxidation forms could be related to apoptosis induction. [Zn(NN)2(H2O)2]2+ complexes
have high ability to bind DNA through intercalation/groove binding, and circular dichroism data
suggests that the main type of species that interact with DNA is [Zn(NN)]2+. Studies varying the %
of fetal bovine serum (1-15%) in cell media show that albumin binding decreases the complex
activity, indicating that distinct speciation of Zn- and phen-containing species in cell media may
affect the cytotoxicity.
Keywords: zinc(II) complexes, polypyridyl ligands, salicylaldimines, , anticancer activity, ROS, lipid
peroxides.
2

ACCEPTED MANUSCRIPT
1. Introduction
The success of platinum-based compounds as anticancer metallodrugs paved the way to an
area of intense research, whose quest is to find new and effective metal-based chemotherapeutic
agents with improved clinical outcomes. In fact, cisplatin is an example of success, although
presenting some constraints as a safe and effective metal-based anticancer drug. Drug toxicity and
acquired resistance during treatment are some of these constraints.
The development of metal-based drugs that make use of bio-essential elements is an area of
current interest. The underlying principle is that metal ion homeostasis of such metals will be better
dealt by human physiology and cause less damaging side effects than when using compounds
based on non-essential metals [1, 2]. Zinc is, after iron, the second most abundant transition metal
in the human body. Zinc is known to be present in thousands of proteins, required for the catalytic
activity of over 200 enzymes [3], and to perform roles related to immune function, protein
synthesis, DNA synthesis and cell division, just to name a few examples [4, 5]. Due to zinc’s
numerous physiological roles, Zn(II) complexes have been used as DNA binding compounds [6-8],
nuclease-mimics [9], radioprotective agents [10], insulin-mimics [11] and antimicrobial agents [12].
Lately, Zn(II) complexes have been gathering attention as potential anticancer drugs with lower
toxicity in vivo, and probably distinct targets and modes of action from the classical Pt-
metallodrugs [13-16]. It is clear that zinc offers interesting chemical and biochemical properties for
the development of anticancer drugs. Combination of the metal ion with bioactive and/or
bioavailable organic ligands is a strategy that may contribute to obtain an effective and more
selective metallodrug.
Schiff base metal complexes have been extensively studied owing to their modular easy
preparation and versatility [17]. The use of salicylaldehyde amino acid derived Schiff bases with
N,O donor sets plays an important role in coordination chemistry related to catalysis and enzymatic
reactions that may be potentially used for many biological applications [18-20]. Metal complexes
based on polypyridyl ligands and metal ions such as Zn(II) [8], Cu(II) [19, 21-26], V(IV) [18, 27],
and Ru(II) [28],[29], have been intensively studied for their features as DNA intercalators, showing
a diverse spectrum of DNA binding/cleavage ability and cytotoxicity against cancer cell lines.
Moreover, some complexes with phenanthroline derivatives were reported to be active against
various pathologic conditions including microbial and fungal infections in addition to cancer
diseases [30-32]. Nuclear DNA has been considered an important target for some of these
complexes. However, more recent studies have shown that the main anticancer action could be
attributed to other factors, such as ROS (reactive oxygen species) production triggering
mitochondria-mediated apoptosis and also indicating the dependence of the cytotoxic effects on
the accumulation of the complexes in different cell compartments [33].
3

ACCEPTED MANUSCRIPT
Our group has been involved in the search for prospective metal-based anticancer drugs. Our
approach has included the use of ternary complexes of vanadium(IV) and copper(II) with Schiff
bases or thiosemicarbazones and polypyridyls [18-20]. The presence of the polypyridyl aromatic
heterocycle seems to be crucial for the anticancer activity, and their binding to Zn(II) may modulate
their activity. Within this objective two groups of complexes were prepared: Zn(II)-complexes with
general formulation [Zn(Sal-Gly)(NN)(H2O)], with Sal-Gly = N-salicylideneglycinate, NN = 1,10-
phenanthroline (phen), 5-chloro-1,10-phenanthroline (Clphen), 5-amine-1,10-phenanthroline
(amphen), 5,6-epoxy-5,6-dihydro-1,10-phenanthroline (epoxyphen) and 4,7-diphenyl-1,10-
phenanthroline (Bphen); and Zn(II)-complexes with formulation [Zn(NN)2]2+ with NN = phen and
amphen (Fig. 1). These compounds and their solution behavior were studied by spectroscopic
methods, and evaluated for their antiproliferative action against several cell lines. Plausible
mechanisms of action were assessed with biochemical techniques.
Fig. 1 General structural formulae and numbering scheme for the Zn(II)-complexes, depicting the
tridentate N-salicylideneglycinate and the 1,10-phenanthroline-derived ligands. The labeling of
protons 3 and 5 in complexes 1-5 are also indicated.
2. Experimental Part
2.1. Materials
Glycine (Gly) was purchased from BDH Chemicals, salicylaldehyde from Aldrich and
phenanthrolines from Alfa Aesar or Sigma-Aldrich. The metal precursors Zn(CH3COO)2.2H2O and
Zn(NO3)2.4H2O were from Panreac and Fluka, respectively. 4-(2-Hydroxyethyl)piperazine-1-
ethanesulfonic acid (HEPES) was purchased from Sigma. Solvents were from Sigma-Aldrich,
Carlo-Erba, Panreac or Fisher and used as received. Phosphate buffered saline (PBS) was
4

ACCEPTED MANUSCRIPT
purchased from Sigma-Aldrich as tablets readily soluble in water (deionized water) giving 0.010 M
in phosphate (and NaCl 0.138 M; KCl 0.0027 M), pH 7.4 at 25º C. HEPES 10 mM buffer solutions
were prepared in deionized water with pH 7.4 at 25ºC. Calf thymus DNA (ctDNA) was from Sigma-
Aldrich. All materials used were either p.a. or reagent grade. Cell media and supplements were
purchased from Gibco, Thermo Fisher Scientific. Tumor cells A2780 (ovarian), MCF7 (breast),
HeLa (cervical) and a non-tumorigenic cell line V79 (Chinese hamster fibroblasts) were acquired
from Sigma-Aldrich (A2780) and ATCC (American Type Culture Collection).
2.2. Instrumentation
Electronic absorption spectra (UV-Vis) were recorded with a Perkin Elmer Lambda 35
1
spectrometer with 10.0 mm quartz cuvettes. H and 13C NMR spectra were obtained on Bruker
1
Avance+ 400 MHz or 300 MHz Spectrometers and H and 13C NMR chemical shifts (δ) are
expressed in ppm relative to Me4Si or the deuterated solvent residual peak. Elemental analyses
were carried out on a FISONS EA 1108 CHNS-O apparatus. Electrospray Ionization ionization
mass spectra (ESI-MS) were measured on a 500-MS Varian Ion Trap Mass Spectrometer. The
infrared spectra were recorded on a JASCO FT/IR 4100 spectrophotometer. Circular Dichroism
(CD) spectra were measured on a Jasco J-720-CD spectropolarimeter (Jasco, Hiroshima, Japan)
with a 175-800 nm photomultiplier (EXEL-308) using a 10.0 mm quartz cuvette. Fluorescence
measurements were carried out on a SPEX® Fluorolog spectrofluorimeter (Horiba Jobin Yvon) in a
FL3-11 configuration, equipped with a Xenon lamp and in a 10.0 × 10.0 mm quartz cuvette. The
instrumental response was corrected by means of a correction function provided by the
manufacturer. The fluorimetric experiments were carried out at room temperature and are all
steady-state measurements. The X-ray Powder Diffraction (XRPD) data collection was done in a
D8 Advance Bruker AXS θ-2θ diffractometer, with a copper radiation source (Cu Kα, λ = 1.5406 Å)
and a secondary monochromator.
2.3. Synthesis
2.3.1. [Zn(Sal-Gly)(NN)(H2O)] complexes
Complexes were synthesized using the following general procedure: to an ethanolic (or
methanolic) solution of glycine (1 mmol) and NaOH (1 mmol) salicylaldehyde (1 mmol) was added.
The resulting yellow solution was stirred for 30 min at room temperature and a solution of
Zn(CH3COO)2.2H2O (1 mmol in 5 mL of water) was added dropwise. The pH was adjusted to 7
with 1 M NaOH and the mixture was stirred at room temperature for 1 h. The polypyridyl (1 mmol)
in ethanol (10 mL) was then added and the resulting solution stirred at room temperature for 1 h,
subsequently concentrated and left overnight at 4ºC. The precipitate obtained was filtered, washed
with cold ethanol and diethyl ether and dried in vacuum.
5

ACCEPTED MANUSCRIPT
[Zn(Sal-Gly)(phen)(H2O)] 1. Yield: 62 %. Elemental Anal. Calc. for C21H17N3O4Zn: C, 57.22; H,
3.89; N, 9.53. Found: C, 57.31; H, 3.85; N, 9.37. ESI MS m/z: [Calculated (Found)]: 420.04
1
(419.23) (100%) [Zn(Sal-Gly)(phen)-H]−; 422.04 (422.12) (15%) [Zn(Sal-Gly)(phen)+H]+. H NMR
(d6-DMSO, δ/ppm): 8.99 (d, 2H); 8.77 (d, 2H); 8.55 (s, 1H); 8.18 (s, 2H); 7.95 (m, 2H); 7.17 (d, 1H);
13
6.95 (t, 1H); 6.35 (t, 1H); 6.26 (s, 1H); 4.01 (s, 2H). C NMR (d6-DMSO, δ/ppm): 172.17 (COO);
167.23 (CN); 148.89, 138.69, 135.02, 132.61, 128.51, 126.88, 125.03, 122.68, 119.67 (C
aromatic); 58.65 (CH2). UV–Vis in DMSO, λmax/nm (ε/M-1 cm-1): 266 (3.2x104), 371 (6.3 x103). FTIR
(KBr, cm-1): 1386 (s, COO), 1580 (as, COO), 1636 (, C=N).
[Zn(Sal-Gly)(Clphen)(H2O)] 2. Yield: 79%. Elemental Anal. Calc. for C21H16N3O4ClZn: C, 53.08; H,
3.39; N, 8.84. Found: C, 52.75; H, 3.43; N, 8.49. ESI MS m/z: [Calculated (Found)]: 456.00
1
1
(456.11) (20%) [Zn(Sal-Gly)(Clphen)+H]+. H NMR (d6-DMSO, δ/ppm): H NMR (d6-DMSO,
δ/ppm): 9.05 (dd, 2H); 8.88 (d, 1H); 8.70 (d, 1H); 8.51 (s, 1H); 8.45 (s, 1H), 8.07 (m, 1H); 7.96 (m,
1H), 7.16 (m, 1H), 6.98 (t, 1H); 6.35 (t, 2H), 4.00 (s, 2H).13C NMR (d6-DMSO, δ/ppm): 172.18
(COO); 167.34 (CN); 149.78, 149.50, 134.98, 132.64, 129.08, 128.27, 126.34, 122.60, 119.52 (C
aromatic); 58.43 (CH2). UV–Vis in DMSO, λmax/nm (ε/M-1 cm-1): 268 (3.7x104), 374 (6.8x103). FTIR
(KBr, cm-1): 1385 (s, COO), 1601 (as, COO), 1644 (, C=N).
[Zn(Sal-Gly)(amphen)(H2O)] 3. Yield: 74%. Elemental Anal. Calc. for C21H18N4O4Zn·0.3H2O: C,
54.69; H, 4.06; N, 12.15. Found: C, 54.75; H, 3.89; N, 11.95. ESI MS m/z: [Calculated (Found)]:
1
437.05 (437.08) (30%) [Zn(Sal-Gly)(amphen)+H]+. H NMR (d6-DMSO, δ/ppm): 8.95 (d, 2H); 8.59
(d, 1H); 8.54 (s, 1H); 8.35 (d, 1H); 7.96 (t, 1H); 7.73 (t, 1H); 7.17 (d, 1H); 7.00 (s, 1H); 6.94 (t, 1H);
13
6.61 (s, 2H); 6.32 (t, 1H); 6.20 (d, 1H); 4.03 (s, 2H). C NMR (d6-DMSO, δ/ppm): 172.14 (COO);
167.14 (CN); 148.25, 143.50, 143.37, 134.98, 133.78, 132.57, 131.03, 124.88, 123.73, 122.67,
121.85, 119.72, 111.50, 100.85 (C aromatic); 58.88 (CH2). UV–Vis in DMSO, λmax/nm (ε/M-1 cm-1):
277 (1.9x104), 361 (9.9x103). FTIR (KBr, cm-1): 1388 (s, COO), 1599 (as, COO), 1637 (, C=N).
[Zn(Sal-Gly)(epoxyphen)(H2O)] 4. Yield: 90%. Elemental Anal. Calc. for C21H17N3O5Zn·0.8H2O:
C, 53.53; H, 3.98; N, 8.92. Found: C, 53.57; H, 3.84; N, 8.87. ESI MS m/z: [Calculated (Found)]:
1
438.04 (438.11) (15%) [Zn(Sal-Gly)(epoxyphen)+H]+. H NMR (d6-DMSO, δ/ppm): 8.69 (d, 2H);
8.53 (d, 2H); 8.48 (s, 1H); 7.73 (m, 2H); 7.14 (d, 1H); 6.97 (t, 1H); 6.34 (t, 1H); 6.30 (s, 1H); 4.97
(s, 2H); 3.93 (s, 2H). 13C NMR (d6-DMSO, δ/ppm): 172.09 (COO); 167.21 (CN); 148.61, 134.97,
132.61, 130.65, 126.17, 122.74, 119.61, 111.65 (C aromatic); 58.50 (CH2); 53.90 (CH). UV–Vis in
DMSO, λmax/nm (ε/M-1 cm-1): 300 (1.3x104), 371 (6.9x103). FTIR (KBr, cm-1): 1384 (s, COO), 1584
(as, COO), 1639 (, C=N).
[Zn(Sal-Gly)(Bphen)(H2O)] 5. Yield: 80%. Elemental Anal. Calc. for C33H25N3O4Zn·0.5H2O: C,
65.84; H, 4.35; N, 6.98. Found: C, 66.16; H, 4.32; N, 6.77. ESI MS m/z: [Calculated (Found)]:
574.10 (574.11) (30%) [Zn(Sal-Gly)(Bphen)+H]+.1H NMR (CD2Cl2, δ/ppm): 8.83 (d, 2H); 8.47 (s,
6

ACCEPTED MANUSCRIPT
1H); 7.98 (s, 2H); 7.52 (m, 6H); 7.40 (m, 6H); 7.28 (d, 1H); 7.20 (t, 1H); 6.82 (d, 1H); 6.71 (t, 1H);
3.61 (s, 2H). 13C NMR (CD2Cl2, δ/ppm): 173.86 (COO); 167.54 (CN); 151.79, 148.68, 136.81,
135.19, 133.89, 130.13, 129.88, 129.45, 127.44, 125.51, 125.19, 115.77 (C aromatic); 60.44(CH2).
UV–Vis in EtOH, λmax/nm (ε/M-1 cm-1): 222 (5.4x104), 279 (4.7x104), 358 (6.0x103). FTIR (KBr, cm-
1): 1389 (s, COO), 1597 (as, COO), 1644 (, C=N).
2.3.2. [Zn(NN)2(H2O)2]2+ complexes
The homoleptic zinc complexes with polypyridyls were synthesized following procedures
reported in the literature for similar compounds [34, 35]. The polypyridyl (1 mmol) was dissolved in
ethanol (or methanol, 1 mL) and added dropwise to a stirred ethanolic solution (1.5 mL) of
Zn(NO3)2·4H2O (130.7 mg, 0.5 mmol). The mixture was stirred for 2 h at room temperature or
under reflux and left overnight at 4 ºC. The solid was collected by filtration, washed with small
amounts of cold water, acetone and ethyl ether and dried over silica gel under vacuum.
[Zn(phen)2(H2O)2](NO3)2 6. A white solid was obtained. Yield: 176 mg, 64.0 %. Elemental analysis
for C24H20N6O8Zn: C, 49.20 %; H, 3.44 %; N, 14.35 %. Found: C, 48.86 %; H, 3.11 %; N, 14.12%.
ESI-MS (MeOH) m/z [Calculated (Found)]: 486.04 (486.04) (90 %) [Zn(phen)2(NO3)]+
1
(C24H16N5O3Zn); 212.3 (212.0) (14 %) [Zn(phen)2]2+ (C12H8N2Zn0.5). H NMR (MeOD-d4, δ/ppm):
8.98 (d, 4H, H1 and H11), 8.75 (s, 4H, H5 and H13, br), 8.36 (s, 4H, H7 and H8), 8.08 (s, 4H, H6 and
13
H12, br). C NMR (MeOD-d4, δ/ppm): 149.1 (C5 and C13), 140.5 (C1 and C11), 139.7 (C3 and C10),
129.0 (C2 and C9), 128.0 (C7 and C8), 126.6 (C6 and C12). UV–Vis in DMSO, λmax/nm (ε/M-1 cm-1):
325 (1.17×103), 294 (1.28×104), 272 (3.59×104), 267 (3.76× 104). FTIR (KBr, cm-1): 1620 (, C=N),
1579 (, C=C).
[Zn(amphen)2(H2O)2](NO3)2 7. A yellow solid was obtained. (Yield: 236 mg, 81.4 %). Elemental
analysis for C24H22N8O8Zn: C, 47.81 %; H, 3.60 %; N, 18.19 %. Found: 47.51 %; H, 3.18 %; N,
18.43 %. ESI-MS (MeOH) m/z [Calculated (Found)]: 516.10 (516.10) (10%) [Zn(amphen)2(NO3)]+
1
(C24H18N7O3Zn); 227.00 (227.3) (100%) [Zn(amphen)2]2+ (C12H9N3Zn0.5). H NMR (Methanol-d4,
δ/ppm): 7.15-9.1 (aromatic-Hs). 13C NMR (Methanol-d4, δ/ppm): 103.35-149.38 (aromatic-Cs). UV–
Vis in DMSO, λmax/nm (ε/M-1 cm-1): 363 (1.86×104), 298 (5.19×104), 268 (4.42×104). FTIR (KBr, cm-
1): 1615 (C=N), 1593 (C=C).
−
From this point further the water molecules and NO3 ions included in the solid complex
formulation will be omitted for simplicity, unless relevant. Also, sometimes the complexes will be
referred to as Zn(phen)2 and Zn(amphen)2, not specifying charges.
2.4. General procedures for X-ray crystallography
Crystals suitable for single-crystal X-ray analysis were obtained for a compound designated as
5a from the slow evaporation of an ethanolic solution of 5. The data were collected using a
graphite monochromated Mo-Kα radiation ( = 0.71073 Å) on a Bruker AXS-KAPPA APEX II
7

ACCEPTED MANUSCRIPT
diffractometer equipped with an Oxford Cryo system open-flow nitrogen cryostat. Cell parameters
were retrieved using Bruker SMART software and refined using Bruker SAINT on all observed
reflections. Absorption corrections were applied using SADABS [36]. The structures were solved
using direct methods with program SIR2004 [37] and refined with program SHELXL [38], included
in the WINGX-Version 2014.1 [39] system of programs. All non-hydrogen atoms were refined
anisotropically and the hydrogen atoms were inserted in idealised positions and allowed to refine
riding on the parent carbon atom. SQUEEZE [40] was used for the final refinement due to the
presence of disordered solvent molecules. The electron density modelled by SQUEEZE is
consistent with about 14 ethanol molecules per unit cell (ca. 279 electrons for a volume of ca. 847
Å3), corresponding to about 7 ethanol molecules per molecule of 5a. The molecular diagrams were
drawn with ORTEP-3 for Windows [41] included in the software package.
Crystal Data for C66H46N6O6Zn2 (M =1149.83 g/mol): triclinic, space group P-1, a = 13.1470(5) Å, b
= 14.2290(6) Å, c = 20.0460(9) Å, α = 110.681(2)°, β = 93.023(2)°, γ = 110.093(2)°, V = 3229.5(2)
Å 3 , Z = 2, T = 150(2) K, μ(MoKα) = 0.794 mm-1 , 32134 reflections measured , 13174 unique (Rint
= 0.0449, Rsigma = 0.0672) which were used in all calculations. The final R1 was 0.0376 (I > 2σ(I))
and wR2 was 0.0939 (all data).
For more crystallographic experimental data and structure refinement parameters see Tables S1
S3 in Supplementary Information (SI). CCDC 1840672 contains the supplementary
crystallographic data for this work. The data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/structures.
2.5. Stability studies in aqueous medium
Stock solutions of the [Zn(Sal-Gly)(NN)(H2O)] complexes were prepared in DMSO with ca. 1
mM. 1:20 dilutions were then prepared in PBS buffer (pH 7.4), having a final complex
concentration of ca. 50 μM and 5% DMSO. The UV-Visible spectra (230-500 nm) were recorded
for a few hours (3 to 19 h). Another spectrum was recorded after 24 h.
For the [Zn(NN)2]2+ complexes several techniques were used. In the UV-Vis absorption study, 1.0
mM of 6 or 7 stock solutions of the complexes were prepared in DMSO. 1:20 dilutions were done
in PBS buffer (pH 7.4), having final metal concentrations of 50 μM. UV-Visible absorption spectra
(260-500 nm) were recorded for up to 24 h. Similarly, for the fluorescence studies, stock solutions
of the complexes with concentrations of 500 μM were prepared in DMSO. Dilution in PBS yielded
solutions with 25 μM concentration. The fluorescence emission spectra were recorded at 3 and 24
h, using excitation wavelengths of 326 and 450 nm, for 6 and 7, respectively.
An additional stability study was carried out for 6. A phosphate buffer solution (120 mM and
pH 7.5) was prepared by dissolving 2.4 mg KH2PO4 and 14.4 mg Na2HPO4 in 1 mL of D2O. 12.5
mg of the complex were dissolved in 250 μL of DMSO-d6. Then, 20 μL of this solution were mixed
with 380 μL of the phosphate buffer giving 5 % DMSO and 95 % deuterated PBS. The stability of
8

ACCEPTED MANUSCRIPT
the compound was evaluated for 24 h by 1H-NMR spectroscopy.
2.6. Cell studies
2.6.1. Solutions and cellular media
Cells were cultured in RPMI-1640 (A2780) or DMEM+Glutamax-I (MCF7, HeLa, V79)
supplemented with 10% fetal bovine serum (FBS) and 1% antibiotics in a 5% CO2 incubator
(Heraus, Germany) at 37 °C and in a humidified atmosphere.
2.6.2 Cytotoxicity (MTT assay)
The cytotoxicity of the complexes was assessed against A2780, MCF7, HeLa and V79 cells by
the MTT assay (MTT = 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) using 96-well
plates. The compounds were first dissolved in DMSO and then in medium through serial dilutions
ranging from 0.01 μM to 100 μM. The final DMSO concentration did not exceed 1% (v/v) in either
control or treated cells and, under these conditions, it had no effect on the cellular activity.
For the studies of the effect of FBS on the cytotoxic activity, A2780 cells were incubated in media
containing different concentrations of serum from 1 to 20% and the complex at a concentration
equivalent to its IC50 found at 24 and 48 h. Controls (no treatment) for each concentration of FBS
were included in the assays.
After treatment with the complexes the cell medium was replaced by 200 µL (0.5 mg/mL) of a
MTT solution in PBS and further incubated for 3 h at 37 °C. The resulting purple formazan crystals
from the tetrazolium reduction were solubilized in DMSO. Absorbance in controls and cell-treated
wells was measured at 575 nm using a microplate reader (PowerWave Xs, Biotek Instruments,
USA). The cytotoxicity of the complexes was expressed as the IC50 calculated from dose-response
curves using the GraphPad Prism software (vs. 5.0).
2.6.3. Generation of superoxide radicals (NBT assay)
The nitro blue tetrazolium (NBT) assay [NBT=2,2′-bis(4-Nitrophenyl)-5,5′-diphenyl-3,3′-(3,3′-
dimethoxy-4,4′-diphenylene)ditetrazolium) chloride] was carried out adapting previously described
methods using the 96-well plates model [42-44]. After a pretreatment with the ligands and
corresponding complexes at 1, 10 and 100 µM for 1 h at 37ºC, 20 µL of a NBT solution 10 mg/mL
in water was added to the medium of the treated cells (A2780) and incubation was subsequently
carried out for 1 h at 37 ºC. The culture medium was then gently poured and the blue formazan
particles were dissolved in 200 µL of 90% DMSO (90% DMSO:10% NaOH 0.1 N with 0.1 %SDS).
NBT is reduced by superoxide radical anion to NBT formazan, which is spectrophotometrically
measured at 560 nm. Each experiment was repeated twice and each concentration was tested
with at least six replicates. Results (mean±SD) are expressed as % of controls (untreated cells).
9

ACCEPTED MANUSCRIPT
2.6.4. Apoptosis (caspase 3/7 Assay)
Apoptosis was assessed using the Caspase-Glo® 3/7 Assay (Promega), according to
manufacturer´s instructions. With this luminescent assay activities of caspase-3 and caspase-7,
both involved in apoptotic caspase, were measured. The assay provides a proluminogenic
caspase 3/7 substrate containing the tetrapeptide sequence Asp-Glu-Val-Asp (DEVD
aminoluciferin). In the presence of caspase 3/7 this substrate is cleaved; aminoluciferin is
released, consumed by luciferase (provided in the reaction mixture), which generates a
luminescent signal. The luminescent signal is proportional to the caspase activity present in the
cells. The assay was carried out in A2780 cells, plated in 96 wells, treated with the free ligands and
compounds 1-5 for 48 h at concentrations equivalent to their IC50. After 48 h, 100 µL of medium
was removed from each well. Caspase 3/7® reagent was added in a 1:1 ratio and the plate was
shaken in an orbital shaker for 30 s at 300-500 rpm. The plate was incubated at room temperature,
protected from light for 1.5 h. The luminescence intensity was measured using an Infinite 200 Plate
Reader (Tecan). Each experiment was repeated twice and each concentration was tested with at
least three replicates. Results (mean±SD) were expressed as RLU (relative luminescent units).
2.6.5. Cellular Uptake by Inductively coupled plasma mass spectrometry (ICP-MS)
To measure the cellular uptake and distribution of 1, 3 and 6, approx. 1 million A2780 cells
were incubated with the complexes in cell culture medium at a concentration equivalent to the IC50
values found at 24 h incubation. After incubation at 37 °C the cellular pellet obtained was treated
with
a
cell fractionation system (FractionPREPTM, Biovision) to extract cytosolic,
membrane/particulate, nuclear and cystoskeletal fractions. These fractions were dissolved with 0.5
mL nitric acid (65%) with a hot plate at 100 ºC during 12 h. The solution digest is diluted in
ultrapure water to 10 mL, obtained from a MilliQ apparatus. The Zn content (66Zn) was measured
by a Thermo X-Series Quadrupole ICP-MS (Thermo Scientific) equipped with Ni cones and a glass
concentric nebulizer (Meinhard, 1.0 mL min-1) refrigerated with a Peltier system. Indium (115In) at a
concentration of 10 µg L1 was used as internal standard. Standards were prepared from standard
ICP-MS 71 A (Inorganic Venture) with a final concentration of 5.0% nitric acid. The following quality
control was applied: internal standard accepted variation between 80 and 120%; blank’s
concentrations inferior to the limit of detection (1 µg L1) and sample duplicates do not differ more
than 10%. The Zn levels in each fraction were expressed as ng metal content/million A2780 cells.
2.6.6. Morphological analysis by transmission electron microscopy (TEM)
A2780 cells at approx. 70% confluence were treated with two Zn(II) complexes, 1 and 3 at
concentrations equivalent to their IC50 values found at 24 h incubation. Non-treated cells were
used as controls. After incubation, cells were processed following a standard procedure previously
10

ACCEPTED MANUSCRIPT
reported [45]. Briefly, the culture medium was discarded and replaced by a primary fixative
consisting in sodium cacodylate buffer 0.1 M, pH 7.3 containing 3% glutaraldehyde. After primary
fixation for 2 h at 4 °C and a wash step in cacodylate buffer, cells were then scraped, pelleted and
embedded in 2% agar. Samples were then further fixed for 3 h in sodium cacodylate buffer 0.1 M,
pH 7.3 containing 1% osmium tetroxide. Next, samples were washed in acetate buffer (0.1 M, pH =
5.0) and fixed in 0.5% uranyl acetate in the same buffer for 1 h. Dehydration was carried out with
increasing concentrations of ethanol. After passing through propylene oxide, samples were
embedded in Epon-Araldite, using SPI-Pon as an Epon 812 substitute. Glass or diamond knives
were used to cut thin sections, which were stained with 2% aqueous uranyl acetate and Reynold's
lead citrate. The stained sections were analyzed and photographed in a JEOL 1200-EX electron
microscope.
2.6.7. Lipid peroxidation assay
The content of malondialdehyde (MDA) and 4-hydroxynonenal (4-HNE) in lipid extracts of
A2780 cells was measured using a colorimetric kit (BioVision, CA, USA). A2780 cells were seeded
at ~5x105 cells/2 mL medium in 6-well plates for 24 h to allow cells to adhere. Then, after treatment
with the complexes at 10 µM (1, 3, 4 and 5) and 50 µM (2) for 24 h at 37 ºC, the medium was
removed, the cells were washed with PBS, tripsinized and centrifuged to obtain a pellet. After
keeping an aliquot for determination of the number of cells per well, the pellet was treated following
the kit manufacture´s protocol. Briefly, the cells were lysed and centrifuged to remove insoluble
materials and then, to 200 µL of the supernatant, 600 µL of thiobarbituric acid (TBA) was added to
generate the MDA-TBA adduct which was developed incubating the mixture for 1 h at 95 ºC. After
reaction, 200 µL were taken and transferred to a 96-well plate for quantification at 532 nm. To
generate a MDA standard curve and in parallel with the samples, MDA standards were made using
the same procedure. In the absence of cells, lipid peroxides were undetectable in the medium.
Results were expressed as mean±SD from two separate experiments in which triplicate
measurements were made.
2.7. DNA binding studies
Stock solutions of DNA were prepared by dissolving the nucleic acid in PBS or HEPES 10 mM
buffers (pH 7.4), kept at 4 ºC for about 48 h and used within a week. Solutions of DNA gave ratios
of absorbance A260/A280 of ca. 1.9, indicating that the DNA was sufficiently protein free [46, 47]. The
concentration of the prepared ctDNA stock solutions were calculated based on their absorbance at
260 nm by using the per nucleotide extinction coefficient ε260 = 6600 M-1 cm-1 [47].
UV-Vis absorption spectroscopy: Electronic absorption titrations were done by adding aliquots
of the DNA stock solution to solutions of the compounds in 3% DMSO-HEPES (10mM, pH=7.4).
11

ACCEPTED MANUSCRIPT
The DNA solution was also added to the reference cell. Samples were allowed to equilibrate for a
few minutes before recording each spectrum. The initial concentrations were the following: [phen]
= 1.4x10-5, [amphen] = 3.3x10-5 M, [6] = [7] = 1.7x10-5 M. The [DNA]/[Compound] ratio varied
between 0 and 8.0. The data were fitted assuming the formation of a 1:1 association complex
between DNA and the compounds (see section SI4 in SI).
Circular Dichroism spectroscopy: Studies measuring CD spectra were done in quartz
SUPRASIL® cuvettes of 10 mm optical path. The ctDNA solution was prepared by dilution of the
stock solution in HEPES 10 mM buffer to obtain concentrations of ca. 55-80 μM in each sample,
prior to addition of stock solutions of the Zn-complexes' or ligands in DMSO, H2O or MeOH (ca.
2.0–5.6 mM), to obtain each compound:DNA ratio. The samples were allowed to equilibrate for a
few minutes before measurements. HEPES buffer or HEPES/DMSO mixtures were used to obtain
the baseline, which was subtracted from each CD spectrum. The spectra were collected from 200
to 500 nm with a resolution of 1 nm; band width of 1 nm; 3 accumulations; scan speed 200 nm/min
and 2 s response time.
Fluorescence emission spectroscopy: To solutions of the compounds (13.8 to 33 M) in
DMSO:HEPES buffer (pH=7.4) (DMSO 0.4% to 1.2%, v/v), ctDNA (1.5-2.0 mM) in HEPES was
successively added directly to the cuvette, so that the DNA:complex molar ratio varied between
0.05-12. Before each measurement the solutions were kept with gentle stirring for at least 5 min to
equilibrate. The fluorescence emission spectra were recorded at room temperature using the
excitation wavelengths of 295, 333, 297, and 338 nm (slits = 5 nm) for phen, amphen, 6 and 7,
resulting in emission maxima at 359, 570, 350 and 588 nm, respectively. The fluorescence
emission intensities were corrected for the absorption and inner filter effects using UV–Visible
absorption data recorded for each sample [48, 49]. The fluorescence emission of the samples was
recorded from 300 to 660 nm in a 10 mm quartz SUPRASIL® cuvette.
Fluorescence data were analyzed according to the Stern-Volmer equation to determine the
fluorescence quenching and enhancement constant:
(1)
(2)
where I0 and I are the fluorescence intensities in the absence and presence of DNA (Quencher Q
or enhancer E), respectively. Stern-Volmer quenching/enhancing constant, KSV, are determined
from the slope of the plot I0/I versus [DNA].
2.8. Bovine serum albumin (BSA) binding studies
Concentrated stock solutions of BSA were prepared in buffer solution (HEPES with 10 mM, pH 7.4
at 25 ºC) and kept at 4 ºC at least 3 hours to hydrate. The concentration of the protein stock
12

ACCEPTED MANUSCRIPT
solutions was ca. 46 μM, estimated considering the molar extinction coefficient at 279 nm (44309
M−1cm−1).[50] A stock solution of 1,10-phenanthroline was prepared by dissolving the compound in
spectroscopic DMSO to obtain a final concentration of 550 μM.
The protein stock solution was diluted to the working concentration of ca. 2 M in HEPES buffer
with a final volume of 3.0 mL placed in a quartz cuvette with a 1 cm path length. The tryptophan
fluorescence quenching experiments were done by incremental additions of 2-67 L of 1,10-
phenanthroline stock solution to obtain final concentrations ranging from 0.4 to 29 M (DMSO <
5% v/v). After each addition the solutions were equilibrated for ca. 5 min before the acquisition of
spectra. The fluorescence emission spectra were recorded between 300 and 450 nm by applying
an excitation wavelength of 295 nm. UV-visible absorption spectra were recorded for each sample
to correct the absorption and inner filter effects [48, 49] of the fluorescence emission intensities.
Due to the intrinsic emission fluorescence of the quencher (1,10-phenanthroline), the fluorescence
emission spectra of the ligand alone, under the same experimental conditions used in the BSA
titration assays, were recorded and subtracted to each corresponding emission spectrum. The
increasing concentrations of phen induce a decrease in the emission fluorescence intensity of BSA
tryptophan residue (max = 343 nm) and the fluorescence quenching was fitted with the Stern-
Volmer equation 1.
3. Results and discussion
A series of ternary Zn(II) complexes with the salycylaldimine N-salicylidene-glycinate (Sal-Gly)
as one of the ligands, and five different polypyridyls was synthesized in good yields by reaction of
Zn(II) acetate with stoichiometric amounts of the tridentate Sal-Gly, generated in situ, and the co-
ligand. Reaction of Zn(II) nitrate with 2 moles of selected polypyridyls at room temperature in
methanol yielded [Zn(NN)2]-type complexes. Complexes 1 and 6 are not new, but were included
here for comparison with the new compounds. The crystal structure of 1 was previously
determined [51] showing a six coordinate Zn-center, with a water molecule completing the
coordination sphere. For 6 several different crystal structures have been obtained [52, 53] (a
search in CSD database yielded 36 hits), all showing the Zn metal center with a coordination
number of 6. To the best of our knwoledge complexes 2-5 and 7 are new. The compounds were
characterized in the solid state and in solution using analytical and spectroscopic techniques.
1
Results obtained from elemental analysis, UV-Vis electronic absorption, ESI MS, FTIR and H
NMR spectroscopy (see below) are in agreement with the proposed formulations: [Zn(Sal-
Gly)(NN)(H2O)](H2O)x and [Zn(NN)2(H2O)2](NO3)2. Their structural formulae are depicted in Fig. 1.
3.1. FTIR spectroscopy
13

ACCEPTED MANUSCRIPT
Tentative assignments of the main FTIR bands were made based on previous reports [51, 54,
55] and also on the vibrational behavior of vanadium and copper complexes of salicylaldimines
[18, 20, 56]. The FTIR spectra obtained for the [Zn(Sal-Gly)(NN)(H2O)] complexes display the
presence of strong imine stretching bands (C=N) in the range of 1636 – 1644 cm−1, as predicted
in the literature for the coordination of a Schiff base to a metal ion (Table S1 in Electronic
Supplemental Information) [20, 56-58]. A strong band found between 1580 – 1601 cm−1 is
consistent with an asymmetric carboxylate stretch as(COO), the symmetric stretch, s(COO), being
observed between 1384 – 1389 cm−1The difference between these two values (around 200 cm−1)
is characteristic of monodentate coordination of the O-carboxylate atom [56]. The phenolic (CPh-
O) around 1300 cm−1, probably coupled to vibrations of the aromatic ring [58], was also identified.
The presence of phenanthroline in the coordination sphere of the complexes was confirmed by
several bands corresponding to stretching and deformation vibrations of the N-heterocyclic ligands.
Differences amongst vibrational spectra of the complexes were found, as expected due to
electronic effects of the various phenanthroline ring substituents.
The FTIR spectra of the phen and amphen ligands show characteristic bands at 1636 and
1667 cm−1, respectively, associated with vibrations of the -C=N-C- moiety. In the FTIR spectra of
the corresponding [Zn(NN)2(H2O)2]2+ complexes, they were shifted, due to coordination through the
nitrogen atoms.
3.2. Electronic absorption spectroscopy
Electronic spectral data for the Zn-complexes in solution are given in Table S2. An absorption
band in the range 355-375 nm was observed in the electronic spectra of the Zn(Sal-Gly)-
complexes, which is attributed to π→π* and n→π* transitions originating mainly in the azomethine
chromophore. The aromatic ring absorption bands from the coordinated phenanthroline ligands
and benzene ring are observed in the UV region between 266 and 300 nm for all the complexes
(including 6 and 7), and are due to π→π* ring transitions [57, 59]. These are in agreement with the
observed for the same type of complexes with Cu(II) [19] and with VIVO [60], which showed strong
intraligand bands at 230–250 and at 270–300 nm and intense bands at 375–385 nm.
3.3. NMR spectroscopy
1
The H NMR spectra of 1-5 are somewhat diverse. When compared to the NN ligand
precursors, all complexes’ spectra show downfield shifts for the proton signals, consistent with
coordination to the Zn(II) center. The aromatic protons belonging to the Sal-Gly benzene ring are
observed in the range 7.3 – 6.2 ppm, upfield and in most cases in a clearly separate region of the
spectra from the phenanthroline proton chemical shifts, which appear between 9.2 – 7.2 ppm. The
azomethine proton H(5) (Figure 1 and Table S3) appears as a singlet at 8.90-8.48 ppm. The CH2
protons adjacent to the carboxylate moiety H(3) are observed also as a singlet at 4.17 – 3.68 ppm.
14

ACCEPTED MANUSCRIPT
Characteristic signals associated with phenanthroline derivatives’ structural features were also
observed. Some examples are the CH-CH proton signal recorded at 4.99 ppm as a singlet in 4, or
the NH2 singlet recorded at 6.61 ppm in 3.
The homoleptic [Zn(NN)2(H2O)2]2+ complexes and respective free ligands were also
1
13
1
characterized by H and C NMR spectroscopy. In the H NMR spectra of 1,10-phenanthroline in
DMSO-d6 four types of aromatic protons appear in the range of 7.7 - 9.1 ppm. In its corresponding
Zn-complex 6, these are shifted downfield, appearing between 8.1 - 9.0 ppm. 5-Amino-1,10-
phenanthroline in CD3OD shows the presence of seven types of proton peaks (Fig. S1). However,
in the spectra measured for 7 additional peaks are found, which cannot be assigned simply to the
presence of free ligand. Thus, additional NMR studies were done (13C, COSY and HMBC) to
assign the peaks and understand the solution behavior, which are described in the Supplemental
Information. Overall, equilibrium takes place in the NMR time scale involving species with different
ligand to metal ratios: 2:1 and 1:1, however, no free ligand was found upon dissolution of the
complex, all species detected being Zn(II) complexes.
3.4. Single crystal X-ray diffraction
Slow evaporation of an ethanolic solution of 5 yielded single crystals suitable for single crystal
X-Ray diffraction (XRD) studies, revealing a very interesting molecular structure, 5a, shown in Fig.
2. Crystallographic details are summarized in Table S6 and selected interatomic distances and
angles are presented in Table S7.
Fig. 2 ORTEP representation of the molecular structure of [Zn(Sal-Gly)2Zn(Bphen)2] (5a), drawn
with 50% thermal ellipsoids. Hydrogen atoms are omitted for clarity. Selected distances (Å): Zn1-
N1 2.135(2), Zn-N2 2.157(2), Zn1-N3 2.145(2), Zn1-N4 2.134(2), Zn1-O1 2.111(2), Zn1-O4
15

ACCEPTED MANUSCRIPT
2.106(1), Zn2-O1 2.126(1), Zn2-O2 2.064(2), Zn2-O4 2.101(2), Zn2-O5 2.172(2), Zn2-N5 2.090(2),
Zn2-N6 2.076(2).
The molecular structure of 5a reveals a dinuclear complex with both Zn(II) centers presenting
approximately octahedral geometry. One zinc atom (Zn2) is bound to two Sal-Gly ligands while the
other (Zn1) is bound to two phenanthroline ligands and to the two phenolate O-atoms (Ophenolate) of
the Sal-Gly ligands, which are shared between the Zn-atoms. The dinuclear complex is formally
neutral being composed by an anionic [Zn(Sal-Gly)2]2− and a cationic [Zn(Bphen)2]2+ species that
are bound through the Ophenolate bridges. This type of arrangement of the phenanthroline ligands,
giving rise to two metal centers with different coordination spheres, is not common and only one
example of a similar Ni dinuclear compound was found in the literature [61]. The elemental
composition of monomer 5 and dinuclear 5a are indistinguishable. However, we do not expect this
type of dinuclear arrangement to correspond to the formulation of 1-5, as most structures found for
ternary Zn(L)(phen) complexes are monomeric and 5a was obtained by slow evaporation of an
ethanolic solution of complex 5, which took several weeks. Only a few crystals were obtained and
thus, no other analytical analyses were made for 5a, but a X-ray powder diffraction spectrum
(XRPD) was simulated and compared with the one measured for complex 5, confirming that they
correspond to different species (Fig. S4).
The Zn-NBphen bond lengths, ranging from 2.135(2) to 2.157(2) Å, are in agreement with those
found in the literature for this type of bonds [62-64]. A monomeric cationic Zn(II)-complex
containing two phen molecules and a bidentate acetate showed similar Zn-N bond lengths (2.133
to 2.140 Å) [65] .The Zn-OPhenolate and Zn-Nimine bond distances [2.106(1) and 2.126(1); 2.076(2)
and 2.090(2) Å, respectively] are longer than those observed for Zn-Sal-Gly complexes, while the
Zn-Ocarboxylate lengths [2.064(2) and 2.172(2) Å] are within the values previously reported [51, 66].
3.5. Stability in aqueous media
All complexes are stable in DMSO, some up to 72 h (Fig. S5). DMSO may replace the
coordinated H2O molecule, but this substitution may be so fast that we cannot detect it by UV-Vis
spectroscopy. Since one of the goals of the work is to evaluate the therapeutic potential of the Zn-
complexes as antitumor drugs, it is important to monitor their hydrolytic behavior in aqueous media
in the M range. Therefore, the complexes were dissolved in organic solvents, diluted in
physiological pH buffer and their stability evaluated (Figs. S6-S10). The UV–visible absorption
spectra measured in 5% DMSO–PBS over time for 2 are shown in Fig. 3. Changes are seen within
the first minutes, stabilizing after one hour. The band around 360-370 nm, due to transitions in the
imine bond, decreased with time, which is consistent with hydrolysis of this bond. This behavior,
16

ACCEPTED MANUSCRIPT
which is observed for all Zn(Sal-Gly)-complexes, as well as the presence of isobestic points,
corroborates the formation of new species. For 5 precipitation was also observed.
The UV–visible absorption spectra of complexes 6 and 7 depict almost no changes within 24 h
in PBS buffered solutions (Fig. S12); probably changes do occur, but these happen immediately
upon dissolution. In fact, the speciation diagrams calculated for the Zn-phen system in the 1-100
M Zn-concentration range (see Figure S26A) indicate that a mixture of Zn(II), [Zn(phen)]2+,
[Zn(phen)2]2+ and [Zn(phen)3]2+ complexes exist, whose concentrations depend on the total amount
of Zn(II) and Zn:phen molar ratio. The precursor [Zn(Sal-Gly)(H2O)2] was also prepared and its
stability evaluated, showing extensive hydrolysis occurring within one hour after dissolution in PBS
buffer (Fig. S11).
a
b
Fig. 3 a) Electronic absorption spectra recorded for 2 in PBS with 5% DMSO, measured over time;
b) changes observed in the absorbance at 320 and 349 nm with time; [2] = 56 M.
The stability was also evaluated by measuring the fluorescence emission spectra of the
complexes with time (Figs. S14-S16). Again, the Zn-Sal-Gly complexes are stable in DMSO but
not in aqueous buffered solutions, while complexes 6 and 7 show only minor fluorescence intensity
1
changes in both solvents. This was further corroborated for 1 and 6 by H NMR spectroscopy.
While 6 shows no changes in deuterated PBS buffer for 24 h (Fig. S18), complex 1, when
dissolved in DMSO-d6 and diluted with PBS (Fig. S17), showed the appearance of a new peak at
10.15 ppm, which is assigned to an aldehyde CHO proton. Also, the peak assigned to the CHN
imine proton (at 8.54 ppm) decreased its intensity and new peaks, probably corresponding to
aromatic salicylaldehyde protons, appeared in the 6.5-8.0 ppm region. Overall, this confirms that
the Schiff bases hydrolyze in aqueous media. At lower concentrations this becomes even more
critical, as evidenced by ESI-MS, in which solutions of 1 at ca. 10 M were evaluated with time,
17

ACCEPTED MANUSCRIPT
showing that no peaks assignable to [Zn(Sal-Gly)(phen)(H2O)] species were present (data not
shown). Speciation diagrams calculated for the systems Zn-Sal-Gly (taking stability constants from
refs. [67, 68]) and Zn-Sal-Gly-phen in the Zn-concentration range 1-1000 M (section SI-7)
corroborate the extensive hydrolysis.
These experiments demonstrate that at the concentrations relevant for the cytotoxicity studies,
in the low M range, the active species is no longer the originally [Zn(Sal-Gly)(NN)(H2O)] dissolved
complex, but may be the bioactive free NN, the products of the hydrolyzed Sal-Gly Schiff base or
the result of a combined effect of the ligands and Zn(II), or a new Zn(II) species formed with other
bioligands present in the cell culture media. This was one of the motivations for the preparation
and biological evaluation of the [Zn(NN)2]2+ complexes.
3.6. Cell Studies
3.6.1. Cytotoxicity assays
Despite the low hydrolytic stability of complexes 1-5, their cytotoxicity was evaluated in order
to have data for comparison with the ligands and the [Zn(NN)2]2+ complexes. Thus, the cytotoxic
activity of the Zn-compounds and respective ligands was evaluated against a panel of human
cancer cells: ovarian (A2780), breast (MCF7) and cervical carcinoma (HeLa), as well as a non-
tumoral cell line (V79 fibroblasts). The precursor [Zn(Sal-Gly)(H2O)2] and cisplatin were also
assessed for comparison (Table 1).
In general, all compounds are more cytotoxic towards the A2780 cell line, showing the
following cytotoxicity order: A2780 > MCF7 > HeLa. In the ovarian cells the IC50 values upon 48 h
incubation, ranging between 0.5 and 26 μM, are lower than or comparable to cisplatin (IC50 = 22.5
μM) under the same conditions. The [Zn(Sal-Gly)(H2O)2] precursor shows no cytotoxicity,
emphasizing the active contribution of the polypyridyls to the biological activity of the complexes.
However, the polypyridyl ligands per se displayed considerable activity, the IC50 values being in
general of the same order of magnitude or even lower than the corresponding complexes.
When comparing the IC50 values obtained in the ovarian A2780 cell lines for the pairs [Zn(Sal-
Gly)(phen)(H2O)] (1) and [Zn(phen)2(H2O)2]2+ (6) and/or [Zn(Sal-Gly)(amphen)(H2O)] (3) and
[Zn(amphen)2(H2O)2]2+ (7) it appears that the values obtained for the [Zn(NN)2(H2O)2]2+ in all
cancer cells are roughly half the values measured for [Zn(Sal-Gly)(NN)(H2O)]. Considering that
each [Zn(NN)2(H2O)2]2+ contains two moles of the NN derivative, this suggests that the active
species is the NN ligand. This is further substantiated by the results obtained with phen and 6 i.e.,
in A2780 cell line the IC50 value obtained for the phen ligand (5.842.3 M) is roughly twice the one
obtained for 6 (2.41.3 M). The same is observed for amphen and 7. The low hydrolytic stability
18

ACCEPTED MANUSCRIPT
observed for the Zn(Sal-Gly)-complexes gives further support to these considerations. Higher
activity of the amphen when compared with its Ru(II)-complex [(6-p-cymene)RuCl(2-amphen)]Cl
[69] upon 96 h incubation were also reported in experiments with A2780 and MCF7 cells.
Regarding free Zn2+ ions (ZnCl2), no cytotoxic activity was observed (IC50 > 100 µM) in the
ovarian cells during 48 h incubation in the same concentration range of the complexes (0.01 μM-
100 μM). This is compatible with the reported statement that Zn2+ is hydrophilic and hence cannot
cross the plasma membrane or the other membranes of intracellular components to induce an
effect [70].
The selectivity of the complexes towards cancer cells was evaluated by testing the complexes
against the non-tumoral V79 hamster fibroblasts. Despite the non-malignant cell line being of a
different species and tissue it can give an overall idea of the compounds’ selectivity. As shown in
Table 1, for most Zn-complexes the selectivity index (A2780 vs. V79 cells) is higher (SI > 2) than
for the control drug cisplatin (SI = 1) and in the case of complexes 6 and 7, much higher than their
respective ligands. The most selective (SI > 4) are compounds 1, 4, 6 and 7. One of the most
active complexes, 5, which shows IC50 values in the low M range in all cancer cell lines, is also
highly toxic towards the fibroblasts, showing low selectivity.
Table 1 In vitro cytotoxicity measured as the half-inhibitory concentration (IC50) after 48 h
incubation period, using the MTT assay, against three human tumor cell lines: ovarian A3780,
breast MCF7 and cervical HeLa, as well as non-tumoral hamster fibroblasts V79. IC50 values are
reported in μM (±SD) (SD = standard deviation).
Compounds
A2780
>100
MCF7
>100
HeLa
>100
V79
SI*
-
[Zn(Sal-Gly)(H2O)2]
>100
[Zn(Sal-Gly)(phen)(H2O)] 1
[Zn(Sal-Gly)(Clphen)(H2O)] 2
[Zn(Sal-Gly)(amphen)(H2O)] 3
[Zn(Sal-Gly)(epoxyphen)(H2O)] 4
[Zn(Sal-Gly)(Bphen)(H2O)] 5
5.8 ± 3.0
25.9 ± 5.3
3.40 ± 2.8
5.6 ± 0.9
1.73 ± 0.4
2.4 ± 1.3
0.75 ± 0.13
5.84 ± 2.3
5.70 ±3.0
1.84 ± 0.56
14.4 ±3.5
0.50 ±0.5
22.5 ± 5.0
17.6 ± 10.0
26.9 ± 9.5
23.9 ± 6.5
28.1 ± 5.9
10.8 ± 6.5
33.6 ± 11
4.06 ± 0.55
14.2 ± 7.8
6.78 ± 3.9
7.10 ± 3.0
9.50 ±2.3
3.90 ±1.8
18.7 ±5.0
1.20 ±0.7
23.5 ± 5.0
4.1
1.1
3.2
6.0
2.3
5.9
9.0
1.2
1.7
2.1
1.3
2.4
1.0
21.8 ± 12.6 37.7 ± 15.5
12.3 ± 6.11
18.2 ± 4.5
3.04 ± 1.2
9.10 ± 3.7
5.42 ± 2.1
6.21 ± 2.2
10.0 ±2.8
4.41 ± 2.8
10.0 ±4.2
3.20 ± 0.9
20.7 ± 6.4
15.9 ± 8.6
23.7 ± 6.5
4.58 ± 1.0
16.6 ± 7.0
8.16 ± 4.3
11.8 ±3.6
12.0 ±2.2
7.20 ±1.3
18.4 ±4.2
2.00 ±0.5
3.59 ± 2.2
[Zn(phen)2)(H2O)2]2+
6
[Zn(amphen)2)(H2O)2]2+
7
phen
Clphen
amphen
epoxyphen
Bphen
Cisplatin
19

ACCEPTED MANUSCRIPT
* The selectivity index (SI) was calculated for the A2780 cells: IC50(V79)/IC50(A2780).
Since the activity at 48 h appears to arise from the NN aromatic heterocycle, additional studies
were made with complexes 6 and 7, as well as their corresponding NN ligands, incubating A2780
and MCF7 cells with the compounds for shorter time periods. Although in some cases the SD
values are relatively high the results depicted in Fig. S19 globally show that: (i) the cytotoxicity is
time dependent, as expected, with the same trend for both cells; additionally, (ii) at 3 h, all
compounds display similar cytotoxicity in both cell lines, with 7 being more active in the ovarian
than the breast cells; (iii) at 24 h the amphen is equally active in both cells but its corresponding
complex 7 is more cytotoxic in the breast than the ovarian cells; (iv) for the MCF7 cells at 24 h the
complexes are much more cytotoxic than the NN ligands.
This additional study at different incubation times shows that some of the conclusions based
on the results obtained at 48 h of incubation were misrepresentative of the activity of the
complexes compared with their corresponding NN ligands. Considering shorter incubation periods,
in the ovarian cancer cells the ligands and respective complexes perform similarly, however in the
breast cancer cells, the complexes are more active than the ligands, thus not only the metal plays
a relevant role but the type of cell line also. Therefore, to directly conclude that the active species
is the NN ligand is putting the observations in too simplistic terms.
3.6.2. ROS production
Most of the anticancer agents (e.g., cisplatin) induce the formation of highly reactive oxygen
species (ROS) revealed by the elevation of lipid peroxidation products and reduction of the
antioxidant status. ROS at high levels can activate cell death mechanisms and induce apoptosis
[71, 72]. Generally Zn(II) has a protective role against oxidative stress [73]. However, it has been
shown that increased levels of intracellular zinc may impair antioxidant mechanisms such as the
glutathione system [74], or mitochondrial functions [70], which may result in increased production
of ROS [75]. ROS such as superoxide can be detected with NBT, a cell permeant dye that can
cross the cell membrane and be reduced by intracellular free radicals (superoxide anions) to form
formazan compounds, detected as blue precipitates in lysed cells. As shown in Fig. 4A, complexes
2-5 (not clear for 1) increased the ROS levels in a dose-dependent manner with ROS production
being considerably higher when compared with controls (untreated cells) and with the precursor.
Complexes 2 and 3 at 10 M showed greater effect on the ROS levels than 5 under the same
conditions, suggesting that ROS production is not completely correlated with the cytotoxic activities
displayed by these complexes. The free NN ligands also increased the ROS levels but not in a
dose-dependent manner and to a lesser extent than the complexes. The free Zn(II) ions in the form
of ZnCl2 did not induce ROS production in the ovarian cells in a wide concentration range (10−7-
20

ACCEPTED MANUSCRIPT
10−3 M) (Figure S20). This result correlates with the statement that Zn2+ is not expected to enter
inside the cells [81]. Tetrazolium salts like MTT and NBT are converted to colored formazan
crystals and this reduction occurs inside metabolically active cells. The Zn-complexes entered the
cells and therefore a reduction effect of NBT occurred.
Fig. 4 NBT dye assay demonstrating increase in NBT reduction by the Zn(II)-complexes (A) and
the NN ligands (B) in a concentration-dependent manner after treatment of A2780 cells with the
compounds. Zn.2H2O corresponds to [Zn(SalGly)(H2O)2].
3.6.3. Caspase assays
21

ACCEPTED MANUSCRIPT
Caspases, a family of cysteine proteases, regulate ordered processes such as apoptotic
signaling pathways. Among them, caspase-3 and caspase-7 are considered executioners of the
apoptosis pathways. Activation of the caspase-3/7 pathway can be used in cellular assays to
quantify activators and inhibitors of the cell death cascade [76-78]. We explored the use of the
active form of caspase-3 and 7 for the detection of apoptotic events induced by the Zn(Sal-
Gly)(NN)-complexes, since there have been reports that cells undergo apoptosis when exposed to
metal polypyridyl complexes [79]. As depicted in Fig. 5 the ability of the compounds to activate
caspase-3/7 in vitro is not directly correlated with their ability to induce cell death. Complex 2,
which exhibited moderate cytotoxicity and is the less active compound in the series, showed the
highest caspase activation in contrast to 5, that showed no sign of apoptosis in relation to controls
or even inhibited apoptosis. While 1-4 induce cell death by apoptosis, results suggest that complex
5 may induce cell death by other mechanisms. The free ligands followed, in general, the same
trend i.e., their ability to activate caspase-3/7 did not correlate with their cytotoxic activity.
22

ACCEPTED MANUSCRIPT
Fig. 5 Caspase Assays. A) Complexes 1-4, but not 5 induce apoptosis in A2780 cells. B) All the
free ligands induced apoptosis in particular phen and Clphen to a higher extent. The apoptotic
effect was evaluated by CaspaseGlo assay after cell treatment for 48 h, using for each compound
the concentration corresponding to its IC50 value. Data was obtained from three replicates per
assay and was expressed as the mean±SD, in relative luminescent units. *p<0.05 (graph A).
Results are representative of two independent experiments. Although in Fig. 5A some of the
values are not significantly different, some global trends may be concluded.
3.6.4. Cellular uptake and distribution
Cellular uptake studies were carried out to evaluate the distribution and localization of Zn
inside the A2780 cells upon incubation with three complexes (1, 3 and 6), and also to assess their
potential cellular targets. These complexes present high cytotoxicity (at 48 h incubation) and
induce apoptosis (6 was not evaluated). For this study A2780 cells were incubated for 24 h at 37°C
with the complexes at concentrations equivalent to their IC50 values (at 24 h), respectively 100, 30
and 100 µM. The Zn content was determined in cytoplasm, membranes, nucleus and cytoskeleton
using ICP-MS. As depicted in Fig.6, all complexes (or Zn-containing species derived from them)
enter the cells but present quite different distribution profiles; for 3 the Zn is found in roughly the
same amount (~25% of total uptake) in all the cellular fractions, while for complex 1 the metal ion is
mainly retained in the cytosol fraction (60% total uptake) and for 6 mainly in the membranes
fraction (56% total uptake) (Figure 6B). Importantly, Zn was also found in the nucleus, at a higher
percentage (and absolute amount) for complex 3 (the most cytotoxic at 24 h of the three
compounds), when compared to 1 (26 vs. 9%) or 6 (26 vs. 14%). Noteworthy, the absolute amount
of Zn entering the cells is globally higher in the case of incubation with 30 M of complex 3 than
upon incubation with 100 M of complex 6.
Overall, these results also support the idea that the cytotoxicity might not be solely/simply due
to the polypyridyl ligand, as complexes 1 and 6, both containing the phen molecule, accumulate
preferentially in different compartments. In fact, the distinct nature of the NN ligand is also causing
distinct distribution of Zn in the cellular compartments.
a
b
23

ACCEPTED MANUSCRIPT
Fig. 6 Cellular uptake and distribution profile of Zn obtained by ICP-MS for complexes [Zn(Sal-
Gly)(phen)(H2O)] (1), [Zn(Sal-Gly)(amphen)(H2O)] (3) and [Zn(phen)2(H2O)2](NO3)2 (6). a) the
content of Zn (ng) per million A2780 cells; b) the content of Zn expressed as percentage of total
uptake in the cellular fractions.
3.6.5. Morphological analysis by transmission electron microscopy (TEM)
The ultrastructural alterations at cellular organelle level were also evaluated for complexes 1
and 3 by TEM studies, in the conditions described in Figure 7. Results showed that, upon
treatment with 1 (Figure 7 c, d and S21B), A2780 cells suffered marked mitochondrial changes and
morphological alterations typical of cell death, including the presence of electron-dense floccular
material in the matrix, in contrast with the cell morphology depicted in Figure 7 a, b (control cells,
see also Fig. S21A) where a normal variation of the mitochondria and nucleus structures were
observed.
Treatment with 3 induced milder mitochondrial changes, but the nuclear membrane showed
marked indentations and larger nucleoli (Fig. 7 e, f and Fig. 21C). Apoptotic features were present
in both treatments, but in a more advanced stage for 1. These effects are consistent with caspase
activation detected for both complexes, in particular complex 1, and perhaps in agreement with the
higher increased cytosol association of 1 and significantly higher nuclear association of 3.
24

ACCEPTED MANUSCRIPT
Fig. 7 Thin section transmission electron microscopy of A2780 cells treated for 24 h with 1 and 3 at
a concentration equivalent to their IC50 at 24 h (100 and 30 µM, respectively) (see Experimental
Part for details). a) and b): control (no treatment); c) and d): 1 treatment; e) and f): 3 treatment.
Cells treated with 1 showed mitochondrial (M) alterations (c) and apoptosis (d). Cells treated with 3
showed deep nuclear indentation (N) and prominent nucleolus (e) and swollen mitochondria (M)
and lysosomes (L) with variable contents (f).
3.6.6 Membrane lipid peroxidation
Quantification of lipid peroxidation is very important to assess oxidative stress in several
processes. Lipid peroxides are one of the most widely used indicators of free radical formation, a
key biomarker of oxidative stress. These products are unstable and decompose to form a complex
series of compounds, which include reactive carbonyl compounds, such as MDA and lipid alkenals
such as 4-HNE [80]. Measurement of lipid peroxidation has relied on the detection of thiobarbituric
acid (TBA) reactive compounds generated from the decomposition of lipid peroxidation products.
This reaction, which takes place under acidic conditions at 95-100 ºC, results in the formation of a
MDA-TBA adduct that can be quantified by its absorbance at 532 nm.
25

ACCEPTED MANUSCRIPT
The basal level of lipid peroxides in the A2780 cells was 0.5 nmol/106 cells (Table 2). Complexes
increased lipid peroxide level in the ovarian cells by more than three times for 1 and 2 and more
than 10 times for 3, 4 and 5.
Table 2. Lipid peroxide (MDA) content in the A2780 cells. Cells were treated with complexes 1-5
for 24 h at a concentration equivalent to 2x their IC50. Results are expressed as mean±SD from two
separate experiments in which triplicate measurements were made.
MDA
Culture conditions
Cells only (no treatment)
1 (10 µM)
(nmol /106 cells)
0.47±0.06
1.81±0.32
1.91±0.20
6.90±0.91
6.21±0.54
8.42±0.66
2 (50 µM)
3 (10 µM)
4 (10 µM)
5 (10 µM)
3.7. DNA binding studies
Although we tried to evaluate the ternary [Zn(Sal-Gly)(NN)(H2O)] complexes, their low stability
in aqueous media precluded making a reliable study. In section SI-10 we present a few speciation
diagrams that help demonstrating the very extensive hydrolysis expected for the [Zn(Sal-
Gly)(phen)(H2O)] complex in the typical low concentration conditions used in studies of binding of
metal complexes to ctDNA. The homoleptic Zn-polypyridyl complexes, 6 and 7, were evaluated
towards their ability to bind DNA. It has been reported that metal complexes containing planar
aromatic heterobases, such as the present NN derivatives, are able to bind DNA by intercalative
and non-intercalative modes [21-23, 81]. Spectroscopic techniques, namely UV-Vis absorption and
fluorescence quenching measurements have been usually applied to determine the binding mode
and estimate binding constants. We should emphasize that, as demonstrated in Fig. S29A, if Zn-
phen complexes are dissolved in aqueous solution, they undergo speciation and several Zn-phen
complexes will co-exist [Zn(phen)]2+, [Zn(phen)2]2+ and [Zn(phen)3]2+, and in theory all may bind to
DNA.
3.7.1. UV-Vis absorption spectroscopy
The conditional binding constants for the free phen and amphen, as well as overall constants
for complexes 6 and 7 were determined by monitoring changes in the UV-vis absorption spectra of
solutions containing the compounds that were titrated with increasing amounts of ctDNA (see Fig.
S22). The spectra measured for 7 are shown in Fig. 8, as well as changes observed at maximum
26

ACCEPTED MANUSCRIPT
wavelengths. Isobestic points are clearly detected at 300, 315 and 355 nm, indicating equilibrium
between two spectroscopic forms of the complex, free and associated with DNA. The variation of
molar absorptivity is plotted in the inset at two selected wavelengths. The spectroscopic results
were used to determine binding constants for the association of phen, amphen, 6 and 7 with DNA
assuming the formation of a 1:1 complex (see details in SI-4). The values of the determined
conditional binding constants are compiled in Table S8. Similar binding affinities with Ka values
between 103-104 M-1 were obtained for both ligands and 7, while 6 showed a much lower value of
Ka.
Fig. 8 UV-vis absorption spectra measured for the titration of a solution of 7 (1.7x10-5 M) with
increasing amounts of ctDNA (arrows indicate changes with DNA addition), after dilution
correction. The same amount of DNA was added in the reference cuvette. Inset: variation of the
molar absorptivity at two wavelengths (338 and 287 nm) with the [DNA]/[complex] ratio; the lines
show the fitted model to determine the binding constants of the complex with DNA.
3.7.2. Circular dichroism spectroscopy
DNA shows circular dichroism spectra in the UV arising from the coupling of the stacked
planar bases, which adopt a chiral helical arrangement. Calf Thymus DNA circular dichroism (CD)
spectrum displays a positive band centered at ca. 275 nm (due to base stacking), and a negative
band centered at ca. 245 nm (associated with right handed helicity), typical of ctDNA in the right
handed B-form [82]. Spectral changes in this region upon addition of compounds to solutions of
DNA can be due to electrostatic interactions, major or minor grove binding or intercalation of the
added compounds. Non-chiral molecules, such as the free ligands in this study, or racemic
mixtures such as the presently studied Zn-complexes, have no intrinsic CD signal. However, if they
bind to a chiral macromolecule such as DNA, an induced CD signal (ICD) may arise in the range
where they absorb [83]. Typically, the ICD signals of intercalators are of low intensity (< 10 M-
27

ACCEPTED MANUSCRIPT
1cm-1) and those of groove-binders are stronger [84]. If one of the enantiomers of the racemic
mixture of a Zn-complex binds preferentially to ctDNA, additional CD bands may develop.
Fig. 9 shows the circular dichroism spectra measured in systems containing ctDNA and
increasing amounts of complexes 6 and 7, for the [Zn-complex]/[DNA] molar ratios indicated; Fig.
S23 displays CD spectra measured upon adding the free ligands phen and amphen to ctDNA
solutions. It is clear from Fig. 9 that in both systems strong and distinct changes are observed,
even for low molar ratios of added Zn-complexes. For 6 the CD bands, particularly in the range
270-310 nm, increase with additions of the complex (Fig. 9a), at least up to a molar ratio of ca. 1.0.
This increase is much steeper than when only the free phen ligand is added to ctDNA (Fig. S23a);
in this latter system there is an initial increase in the  values up to a molar ratio of 0.4, and
almost no further changes take place upon further additions. This is in agreement with intercalation
of the phen molecules in between DNA base pairs since both the 275 and the 245 nm bands
increase its intensity indicating lengthening of the DNA helix. It also corroborates that when
complex 6 is added to ctDNA, it is a Zn-containing species that binds to the macromolecule. The
ICD bands are superimposed on the ctDNA spectrum, since this is the spectral region in which
both DNA and the complex absorb (between 250-300 nm). It presents positive bands with maxima
at 277 nm and a shoulder at 296 nm and negative bands at ca. 257 nm. The bands are quite
strong, particularly the one at 277 nm. Fig. S24 shows the circular dichroism spectra upon
additions of an aqueous solution containing 1:1 molar concentrations of Zn(NO3)2 and phen, this
simulating the addition of a [Zn(phen)2+] complex. The CD spectra are very similar to the ones
obtained with 6, but those of Fig. S24 are less intense; these changes are also intermediate
between the effects observed for complex 6 and those imposed by the phen ligand alone. Overall
our data, namely the absence of clear isodichroic points, support the assumption that several
phen-containing species exist in solution interacting with DNA. Moreover, the Zn-phen species
undergo speciation when added to the solution containing ctDNA (as indicated in Fig. S29A) and
suggests that probably the main species that is binding to ctDNA is the [Zn(phen)2+] complex. For
the very low 6:ctDNA molar ratios it is probably mostly phen that binds to ctDNA.
Upon addition of 7 to ctDNA, the changes in the CD spectra are more drastic than for the Zn-
phen complexes in similar conditions (Fig. 9b): up to a molar ratio 7:ctDNA of ca. 0.4 positive
bands at ~303 and 279 nm develop, and the negative band at 245 nm disappears, and upon
further additions of 7 a negative strong band at ~286 nm appears and the 303 nm band
disappears. For the free amphen ligand up to a molar ratio of amphen:ctDNA of ca. 1.0, the CD
spectra measured resemble those reported by Valladolid et al. [69] for the amphen:ctDNA system
up to a molar ratio of 3. In this latter work the ctDNA used was sonicated to obtain DNA fragments
with 1000 base-pairs, this probably explaining why we obtain distinct CD spectra for the higher
molar ratios of amphen:ctDNA. Overall, checking changes upon subtraction of the ctDNA
28

ACCEPTED MANUSCRIPT
spectrum, for ratios <1 the measured spectra show weak ICD bands at ca. 302 nm (+) and 280 nm
(-). For ratios amphen:ctDNA > 1 the 302 nm band disappears and negative bands at ca. 267 and
290 nm develop.
When a solution containing Zn(NO3)2 and amphen in 1:1 molar ratio is added to ctDNA, the
global evolution of the CD spectra is similar, but the negative strong band at ~286 nm only appears
for higher amphen:ctDNA molar ratios (ca. 0.8-1.0) (Fig. S25), while when adding free amphen
(Fig. S23) this bands only develops for still higher amphen:ctDNA molar ratios (ca. 1.2-1.4). This
indicates that the Zn-amphen species undergo speciation when added to the solution containing
ctDNA (as in the case of the Zn-phen system) and suggests that probably it is mostly the
[Zn(amphen)]2+ complex that is binding to ctDNA. The rather distinct CD spectra obtained for the
Zn-amphen system, when compared with Zn-phen, indicates a distinct type of binding. Additionally,
the changes observed in the CD spectra of Fig. 9b/c indicate that there is more than one binding
mode for complex 7 to DNA. 7 contains amino groups that may interact with the DNA phosphate
groups through H bonds, in the minor or major groove, which for ratios above 0.5 probably become
important producing the strong negative CD bands observed.
a
b
c
Fig. 9 Circular dichroism spectra in the UV range of solutions containing of ctDNA in HEPES
buffer, pH 7.4, and increasing amounts of the zinc complexes: a) [ctDNA] ~65 M and addition of 6
dissolved in water; b) and c) [ctDNA] ~83 M and addition of 7 dissolved in MeOH (max. of 5.3%
29