The Important Role of the Nuclearity, Rigidity, and Solubility of Phosphane Ligands in the Biological Activity of Gold(I) Complexes

Article abstract of DOI:10.1002/chem.201802547

A series of 4-ethynylaniline gold(I) complexes containing monophosphane (1,3,5-triaza-7-phosphaadamantane (pta; 2), 3,7-diacetyl-1,3,7-triaza-5-phosphabicyclo[3.3.1]nonane (3), and PR₃, with R=naphthyl (4), phenyl (5), and ethyl (6)) and diphosphane (bis(diphenylphosphino)acetylene (dppa; 7), trans-1,2-bis(diphenylphosphino)ethene (dppet; 8), 1,2-bis(diphenylphosphino)ethane (dppe; 9), and 1,3-bis(diphenylphosphino)propane (dppp; 10)) ligands have been synthesized and their efficiency against tumor cells evaluated. The cytotoxicity of complexes 2–10 was evaluated in human colorectal (HCT116) and ovarian (A2780) carcinoma as well as in normal human fibroblasts. All the complexes showed a higher antiproliferative effect in A2780 cells, with the cytotoxicity decreasing in the following order 5>6=9=10>8>2>4>7>3. Complex 4 stands out for its very high selectivity towards ovarian carcinoma cells (IC₅₀=2.3 μm) compared with colorectal carcinoma and normal human fibroblasts (IC₅₀>100 μm), which makes this complex very attractive for ovarian cancer therapy. Its cytotoxicity in these cells correlates with the induction of the apoptotic process and an increase of intracellular reactive oxygen species (ROS). The effects of the nuclearity, rigidity, and solubility of these complexes on their biological activity were also analyzed. X-ray crystal structure determination allowed the identification of short N?H???π contacts as the main driving forces for the three-dimensional packing in these molecules.

Full text of DOI:10.1002/chem.201802547

A Journal of
Accepted Article
Title: The important role of the nuclearity, rigidity and solubility of
phosphane ligands on the biological activity of gold(I) complexes.
Authors: Noora Svahn, Artur J Moro, Catarina Rodrigues, Rakesh
Puttreddy, Kari Rissanen, Pedro Viana Baptista, Alexandra
Fernandes, João Carlos Lima, and Laura Rodríguez
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201802547
Link to VoR: http://dx.doi.org/10.1002/chem.201802547
Supported by

10.1002/chem.201802547
Chemistry - A European Journal
The important role of the nuclearity, rigidity and
solubility of phosphane ligands on the biological
activity of gold(I) complexes.
Noora Svahn,a Artur J. Moro,b Catarina Roma-Rodrigues,c Rakesh Puttreddy,d Kari
Rissanen,d Pedro V. Baptista,c Alexandra R. Fernandes,c* João Carlos Lima,b Laura
Rodrígueza,e*
a Departament de Química Inorgànica i Orgànica. Secció de Química Inorgànica.
Universitat de Barcelona, Martí i Franquès 1-11, 08028 Barcelona, Spain.
b LAQV-REQUIMTE, Departamento de Química, CQFB, Universidade Nova de Lisboa,
Monte de Caparica, Portugal
c UCIBIO, Departamento de Ciências da Vida, Faculdade de Ciências e Tecnologia,
Universidade Nova de Lisboa, Campus de Caparica, 2829-516 Caparica, Portugal
d University of Jyvaskyla, Department of Chemistry, Nanoscience Center, P.O. Box 35,
40014 Jyväskylä, Finland.
e Institut de Nanociència i Nanotecnologia (IN2UB). Universitat de Barcelona, 08028
Barcelona (Spain)
*corresponding authors
Laura Rodríguez (laura.rodriguez@qi.ub.es)
Alexandra R Fernandes (ma.fernandes@fct.unl.pt)
1
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
Abstract
A series of 4-ethynylaniline gold(I) complexes containing monophosphanes (1,3,5-triaza-
7-phosphaadamantane, PTA, 2; 3,7-diacetyl-1,3,7-triaza-5-fosfabiciclo[3.3.1]nonane, 3;
and PR3, being R = naphtyl, 4; phenyl, 5 and ethyl, 6) and diphosphanes
(bis(diphenylphosphino)acetylene, dppa, 7; trans-diphenylphosphineethene, dppet, 8;
1,2-bis(diphenylphosphino)ethane, dppe, 9; 1,3-bis(diphenylphosphino)propane, dppp,
10) ligands was synthetized and their efficiency against tumor cells evaluated.
The cytotoxicity of complexes 2-10 was evaluated in human colorectal (HCT116) and
ovarian (A2780) carcinoma, to the effect on normal human fibroblasts. All the complexes
showed a higher antiproliferative effect in A2780 cells, with the following order of
cytotoxicity: 5 > 6 = 9 = 10 > 8 > 2 > 4 > 7 > 3.
Complex 4 stands out for a very high selectivity towards ovarian carcinoma cells (IC50
2.3 M), compared to colorectal carcinoma and normal human fibroblasts (IC50 > 100
M), making this complex very attractive for ovarian cancer therapy. Its cytotoxicity in
these cells correlates with the induction of the apoptotic process and an increase of
intracellular reactive oxygen species (ROS).
The effect of the nuclearity, rigidity and solubility of the complexes on their biological
activity was also analyzed. X-Ray crystal structures determination allowed the
identification of short N-H··· contacts as the main driving forces for the 3D packing in
these molecules.
Keywords: gold(I); X-Ray; antitumoral agents; ovarian carcinoma, intrinsic apoptosis
2
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
Introduction
Cisplatin and other platinum-based metal complexes such as carboplatin and oxaliplatin
are the most widely used and efficient anticancer drugs for the treatment of testicular,
ovarian and bladder cancer. However, the clinical success of cisplatin and its derivatives
is compromised due to inevitable serious side effects, such as nephrotoxicity and
neurotoxicity, ototoxicity and development of drug resistance.[1] Therefore, interest in
medicinal inorganic chemistry has continued to grow with the emergence of gold
complexes as alternatives to cisplatin, as many gold(I) and gold(III) compounds exhibited
potent tumor cell growth inhibitory properties with much lower toxicity and fewer side
effects.[2-5] The recent interest in the biochemistry and bio-inorganic chemistry of gold
has triggered a flood of reports on the use of gold complexes as potential treatment for a
variety of disease models, including rheumatoid arthritis and other chronic inflammatory
and auto-immune diseases, various cancers, as well as bacterial, fungal and viral
infections.[6] Most of these reports focus on gold in either +I or +III oxidation states, as
these forms are comparatively stable in aqueous solution, facilitating their use in
biological research.[7-12] Among the gold complexes with potential anticancer effect,
Auranofin (Ridaura®), a glucopyranosato triethylphosphine gold(I) complex, commonly
used in clinical practice to treat rheumatoid arthritis, has showed a promising effect in the
treatment of several types of cancer (e.g. including ovarian cancer ClinicalTrials.gov
Identifiers: NCT01747798 and NCT03456700). The potential impact of auranofin
triggered the development of new gold compounds. It is known that auranofin and gold(I)
phosphine related compounds selectively inhibit the mitochondrial thioredoxin reductase,
leading to increased ROS production, and consequent mitochondrial damage followed by
internal induction of apoptosis.[13-16] Since the mechanisms of action of anticancer gold
complexes are likely to be different from those of cisplatin, most of the reported cytotoxic
gold complexes are also effective against cisplatin-resistant cancer cells, revealing the
promising prospect in the development of gold chemotherapeutics to resolve the problem
of cisplatin resistance.[17] Ligand displacement is an important reaction of gold(I)
complexes in biological systems. Additionally, this auxiliary ligand can affect the
compounds lipophilicity, stability and the binding affinity of gold(I), and hence the
intracellular transformation of gold(I) species under both in vitro and in vivo conditions.
For this reason, lipophilic/hydrophilic balance should be adjusted through ligand design.
3
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
Gold(I) forms linear complexes, where phosphines[18] and N-heterocyclic carbenes[19]
represent the most common ligands. Thus, complexes containing Au-P bond play an
important role in coordination and organometallic chemistry in the last years. The great
interest for this class of complexes is accounted not only for “pure chemistry” aspects but
also for their applications in homogeneous catalysis,[20] development of anticancer and
antimicrobial metallodrugs and luminophores.[2,3,21] While monodentate tertiary
phosphine ligands have found broad application in the development of anti-cancer gold
complexes, the use of diphosphine ligands is not so extensively exploited.[3,22]
On the other hand, there has been increasing interest over the last 20 years in gold(I)
organometallic complexes containing alkynyl units because of their potential application
in different fields in chemistry, such as molecular electronics, chemical sensors, materials
science, supramolecular chemistry and biological activity among others.[23-27]
Nevertheless, their anticancer activity is just increasing on interest in the last years and
the results are quite promising.[3,28-38]
Taking into consideration all these facts, we have analyzed the effect of the phosphane
unit in the corresponding biological activity of a series of gold(I) complexes containing
the same 4-ethynylaniline chromophore at the second coordination position. Different
factors have been herein considered: i) changes on the solubility of the phosphine in
water; ii) changes on the aromaticity of the PR3 substituents (being R = Ph, Naph or ethyl);
iii) the nuclearity of the complexes (by using diphosphines instead of monophosphines);
iv) flexibility of the molecules (rigid vs. flexible diphosphines). The observed biological
activity in different cancer cell lines has been correlated with the differences on the
resulting chemical structure to obtain some structure-activity-relationship information.
The mechanism underlying cytotoxicity/selectivity for ovarian carcinoma cells of the
most promising complex was further characterized.
4
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
Results and Discussion
Synthesis and Characterization
The reaction of a dichloromethane suspension of [Au(C≡CC6H4NH2)]n (1) with either 1
equivalent of PTA, DAPTA, PNaph3, PPh3, PEt3 or 0.5 equiv of dppa, dppet, dppe and
dppp afforded the desired complexes [Au(C≡CC6H4N)(PR3)] (2-6) and
[Au2(C≡CC6H4NH2)2(μ2-diphosphine)] (7-10) in a yield greater than 60% yield (Schemes
1 and S1-2).
Scheme 1. Chemical structure of the mononuclear (2-6) and dinuclear complexes (7-10)
synthesized in this work.
All the synthesized complexes gave satisfactory elemental analysis and were
characterized by 1H and 31P spectroscopy, ESI-MS(+) spectrometry and IR spectroscopy.
31P{1H} NMR display in all cases a single resonance, which is ca. 50 downfield in
comparison with the free phosphine.[28,34,36-41] Only one peak was recorded for the
diphosphine derivatives 7-10 in agreement with the formation of the symmetrical
dinuclear complexes (see Supporting Information).
1H NMR spectra show the expected signals for α and β protons of the aniline rings as well
as the protons attributable to the phosphine ligands. The characteristic vibration of the
ῡ(C≡C) at ca. 2100 cm-1 is also observed by IR spectroscopy. This band is slightly shifted
5
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
to higher energies with respect to the polymer 1, as previously observed for other
derivatives with analogous chemical structure.[34,40,41] The presence of the molecular peak
recorded in all cases by ESI-MS spectrometry confirms definitively the successful
formation of all complexes.
In order to try to find out the effect of the phosphine conformation of 8 on the chemical
structure and the resulting biological properties, attempts to synthesize the corresponding
cis-phosphine gold(I) isomer were performed. In sharp contrast with the other
phosphines, when the gold polymer [Au(C≡CC6H4NH2)]n (1) was reacted with cis-dppet
(cis-1,2-bis(diphenylphosphino)ethene) in a molar ratio 1:1, under the same conditions as
reported above, a mixture of three different products were obtained being identified as
the ionic compound [Au(cis-dppet)2][Au(C≡CC5H4NH2)2] (Scheme S3) in coexistence
with the desired product and complex 8. This is supported by 31P NMR characterization
(Figure S19) and the ESI-MS analysis that displays the peaks at m/z 989.2 ([Au(cis-
dppet)2]+) and 626.02 ([Au(C≡CC5H4NH2)2]-) in negative and positive mode. Changes on
the concentration and/or the reaction time did not have any influence on the resulting
chemical composition.
In order to obtain complexes with higher solubility in water, similar reactions were
performed by using 3-(diphenylphosphanyl)propanoic acid (Scheme S4-a) and 4-
(diphenylphosphanyl)benzoic acid (Scheme S4-b) but the resulting products could not be
obtained in pure form in agreement with NMR data. A mixture of products can be
identified according to the presence of several peaks in the 31P NMR (Figures S20-S21)
that could not be consequently isolated and purified. This could be attributed to the
possible coordination of the oxygen to the gold metal atom[42-44] giving rise to different
type of derivatives (P-coordination and O-coordination) with similar polarity that
precludes their purification. Final attempts were performed by the synthesis of the
corresponding
phosphine
potassium
salts.
The
reaction
of
4-
(diphenylphosphanyl)benzoic acid with KOH gave KPPh2C6H4CO2 (11), which was
31
characterized by 1H and P NMR, IR and ESI-MS. Approximately 1.5 ppm downfield
shifted signal in 31P NMR, the absence of O-H and C=O vibration in the IR spectrum and
the appearance of characteristic band at 1587 and 1394 cm-1 [45] are in accordance with
the presence of COO- ion. The reaction of 3-(diphenylphosphanyl)propanoic acid with
KOH did not give pure KPC15H14O2, which was proved by 31P NMR showing two
different signals. Unfortunately, the reaction of 11 with AuCl(tht) in dichloromethane
6
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
gave a mixture of products including decomposition of gold(I) complex. Therefore, no
pure anionic chloro-gold(I) complex could be isolated from these reactions and thus, the
resulting aniline products were not possible to obtain.
The solid-state structures of 4, 8, 9 and 10 were analyzed through single crystal X-ray
diffraction studies (Figure 1). Slow evaporation of the dichloromethane/hexane solution
provides single crystals suitable for X-ray crystallography. The complex 8 formed very
low quality and badly twinned single crystals and resulted in abnormal bond distances in
the second Au-PPh3 moiety, thus only the overall structure (dimer) and the Au···Au
distances in 8 can be considered reliable (See Supporting Information Fig. S24 for X-ray
crystal structure of 8). Attempts to obtain crystals for other complexes were unsuccessful.
The coordination of acetylene ligands at the Au(I) centers are linear, the P–Au–C angles
range between 170.9(3)° and 179.1(1)°. The Au(I) centers in 4, 9 and 10 adopt linear
geometries and 8 has a T-shape geometry. The P–Au and Au–C bond parameters are in
good agreement with previously reported Au(I)-acetylide complexes.[23,46,47] All four
complexes display, tetrahedral geometry at hinge phosphorus atom, near linear geometry
of P–Au–C≡C–C units and, typical N–H···Csp, C–H···π and π–π interactions in the 3-D
crystal packing similarly to previously reported gold(I) complexes. [23,46,47] The analysis
of short contacts in complexes 4, 8, 9 and 10 in particular N–H···Csp were analysed using
Hirshfeld surfaces. [48-50] The red spots in Fig. S25-S29 (see Supporting Information)
indicate of short contacts between -NH2 and Au(I) bound acetylenic groups.
Complexes 4 and 10 are discrete structures, while 8 and 9 form dimers by aurophilic
interactions as shown in Figures 1 and S24. The packing arrangement in 4 and reported
[Au(C≡C-C6H4NH2){P(3-tolyl)3}[29] are remarkably similar (Fig. S30), both form 1-D
zig-zag polymeric structures, with N–H···Csp interactions (ca. 2.39 and 2.67 Å) connect
discrete units of 4, while the [Au(C≡C-C6H4NH2){P(3-tolyl)3}] units extend via N–
H···Au interactions. The trans- and anti- orientation of backbone P–C–C–P in 9 is the
major structural feature responsible for dimerization via Au···Au interactions. In 8, the
dimers stabilize through cooperative weak Au···Au interactions [3.316(2) Å] as shown
in Figure S24. In complex 9, acetylenic moieties are positioned in the direction similar
to chloride ligands in [Au2Cl2(dppe)]. [51] The Au···Au interactions [3.118(1) Å] in 9 are
similar to [Au2Cl2(dppe)] [3.162(2) Å][51] and shorter when compared to 8 [3.316(2) Å].
The shorter Au···Au interactions in 9 are perhaps due to flexible ligand, dppe, compared
to rigid and planar dppet. The C–HAr···Csp interactions link the dimeric structures in the
7
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
3-D crystal packing of 9 (Fig. S31). In dppp, the more flexible propane chain between
two phosphorus atoms has ability to exhibit various conformers.[52-59] In complex 10, the
propane chain adopts anti-anti conformation with two P–Au–C≡C–C units roughly
parallel to each other with Au(I) ions separated at distances, ca. 5.60 Å. The
intermolecular N–H···C≡C interactions between –NH2 and C≡C groups of discrete units
10 ranging between ca. 2.44 and 2.56 Å play significant role in stabilizing the 3-D lattice.
(Fig. S32) Similar type N–H···C≡C interactions were previously reported between aniline
and dppp-Au(I) macrocyclic compounds.[60] Of note, the 3-D crystal packing of 10 reveals
a complex pattern of numerous short contacts, of prominent were N–H···CAr,
dichloromethane incorporated C–Cl···π and C–H···π interactions (Fig. S27-S29).
Photophysical Characterization
Absorption and emission spectra of all the complexes were recorded in 2·10-5 M solutions
at room temperature and the obtained data are summarized in Table 1.
The electronic absorption spectra of all complexes (Figure 2) show intense high-energy
bands at ca. 300 nm and low energy bands or tails at ca. 353-400 nm. The higher energy
band display the same profile as the corresponding spectrum of the 4-ethynylaniline
ligand (Figure S33), which is red shifted due to the coordination to the metal atom and it
is thus assigned to 휋– 휋∗ transition of the ethynylaniline moiety. Additionally, as
observed for ethynylaniline, is ca. 20 nm red-shifted in the case of THF/H2O mixtures
(compounds 4-10) with respect to pure water solutions (compounds 2 and 3).
8
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
Figure 1. X-Ray crystal structure of 4 (a), 9 (b) and 10 (c). Yellow: gold; orange:
phosphorus; black: carbon; light blue: nitrogen; Solvents are omitted for viewing
clarity. (See Supporting Information Fig. S24 for X-ray crystal structure, 8).
9
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
Table 1. Absorption and emission spectral data of 4-ethynylaniline and complexes 2-10
in water (2-3) and 1: 1 THF:H2O (4-10). a H2O; b H2O/THF 1:1; exc = 285 nm for 2
and 3); exc = 310 nm (for 4-10)
Compound
Absorption 흀풎풂풙(nm)
(ퟏퟎ−ퟑ휺 (M-1 cm-1))
265 (28.3)a, 277 (32.7)b
290 (16.0)
Emission 흀풎풂풙(nm)
4-ethynylaniline
355
352, 470
354, 459
342
2
292 (19.5)
3
298 (41.1), 381 (6.7)
315 (18.4)
4
353
5
310 (20.9)
358
6
307 (20.7), 404 (14.4)
314 (25.7), 400 (4.5)
313 (19.9), 382 (11.9)
310 (22.0), 380 (2.7)
350, 420
350, 420
350, 420
350, 420
7
8
9
10
1.00
0.75
0.50
0.25
0.00
1.00
0.75
0.50
0.25
0.00
2
3
4
5
6
7
8
9
10
300
400
500
600
300
400
500
600
700
Wavelength (nm)
Wavelength (nm)
Figure 2. Absorption spectra of 2·10-5M solutions of mononuclear (left) and dinuclear
(right) complexes in water (2-3) and 1: 1 THF:H2O (4-10). Different colors represent
each compound as indicated in the figure.
10
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
A particular shape is recorded in the case of 4 containing the tris-naphthylphosphine
(PNaph3) with an absorption band centered at 298 nm overlapping the absorption of the
ethynylaniline group and presents a higher intensity and vibronically structured band,
typical of C=C aromatic moieties and is ascribed to the -* (Naph3) transition. The lower
energy absorption band ca. 400 nm is observed as a tail in some cases has been assigned
to aggregates’ absorption from σ*(Au···Au)–π* transitions.[61]
Emission spectra recorded upon excitation at ca. 300 nm displays a highly energy
emission at 350 nm tentatively assigned to 1[휋–휋∗(ethynylaniline)] transition.[62] A lower
3
energy band at 415-450 nm with lower intensity assigned to [휋–휋∗(ethynylaniline)]
emission is observed in some cases (Figures 3 and S34-35). This phosphorescence
emission seems to be more favored in the case of PTA and DAPTA derivatives
(complexes 2 and 3) probably due to the presence of aggregates with a direct influence
on the aniline intersystem crossing process.
7
8
9
10
1.0
0.5
0.0
2
3
4
5
6
1.0
0.5
0.0
350
400
450
500
550
600
300
350
400
450
500
550
600
Wavelength (nm)
Wavelength (nm)
Figure 3. Normalized emission spectra of 2·10-5M solutions of mononuclear (left) and
dinuclear (right) complexes in water (2-3, exc = 300 nm) and 1: 1 THF:H2O (4-10 exc
= 310 nm).
Biological assays
Cell viability
The cytotoxicity of the complexes and of each ligand were assessed by the MTS assay on
representative human cancer cell lines: ovarian cisplatin sensitive, A2780; colorectal
carcinoma, HCT116. Except for trans-dppet, dppe and dppp, the ligands of the dinuclear
11
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
complexes, trans-dppet, dppe and dppp no cytotoxic effect was observed for the other
ligands in concentrations up to 100 M (Figure S36, Table S2). A concentration
dependent effect on cell viability was observed after 48 h incubation for all complexes in
both cancer cell lines. An exception is complex 4, where the IC50 in HCT116 cells was
higher than 100 M (Figure 4, Table 2).
Table 2. Relative IC50 (M) for complexes 2-10 in the human cancer cell lines (A2780,
HCT116) and in normal human fibroblasts (48 h).
A2780
HCT116
Fibroblasts
IC50 (M)
Complexes
1.7 ± 0.4
17.4 ± 1.0
2.3 ± 0.1
0.1 ± 0.0
0.3 ± 0.1
10.9 ± 3.7
0.4 ± 0.2
0.3 ± 0.0
0.3 ± 0.1
18.2 ± 8.1
77.5 ± 0.1
> 100
19.0 ± 7.3
70.4 ± 3.5
> 100
2
3
4
8.7 ± 1.8
6.5 ± 0.1
59.5 ± 17.5
4.9 ± 0.2
1.9 ± 0.3
1.3 ± 0.0
5.3 ± 0.7
1.0 ± 0.4
96.3 ± 4.5
36.4 ± 1.2
15.2 ± 1.8
36.4 ± 0.0
5
6
7
8
9
10
12
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
Figure 4. Cytotoxicity of complexes 2-10 in A2780 cell line (black bars) and HCT116
cell line (grey bars) after 48 h incubation with increasing concentrations of the complex.
Cell viability was determined by MTS assay. Data normalized against the control (0.1
% (v/v) DMSO) and expressed as mean ± SEM of three independent assays. * p < 0.05
relative to control for each cell line.
The effect of the mononuclear complexes 2-6 shows a similar trend in cell viability, and
the IC50 values for A2780 are at least 10 times lower than those for HCT116. This
suggests a considerable selectivity of the mononuclear gold(I) complexes towards ovarian
carcinoma cells. Complex 4, containing tri-1-naphtylphosphine, stands out due to the
higher difference observed between cytotoxicity in A2780 (IC50 2.3 M) relative to
HCT116 (IC50 > 100 M). The cytotoxic effect of mononuclear complexes on primary
human fibroblasts is in the same order of magnitude of that on HCT116 cancer cell line
(Figure 5, Table 2).
13
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
Figure 5. Cytotoxicity of complexes 2-10 in normal human fibroblasts determined by
MTS assay. Fibroblasts were challenged with increasing concentrations of A)
mononuclear complexes 2 (red bars), 3 (black bars), 4 (blue bars), 5 (yellow bars), and 6
(light grey bars), and B) dinuclear complexes 7 (green bars), 8 (dark grey bars), 9
(brown bars) and 10 (light brown bars) after 48 h incubation. Data normalized against
the control (0.1 % (v/v) DMSO) and expressed as mean ± SEM of three independent
assays. * p < 0.05 relative to control for each cell line.
Complex 2 (containing PTA) showed a lower IC50 than complex 3 (containing DAPTA),
suggesting that gold(I) coordination with PTA ligand leads to higher cytotoxicity than
DAPTA in A2780 cells (Table 2) as previously observed in other gold(I) alkynyl
complexes that differ on these phosphanes.[34] However, complex 3 showed lower
cytotoxicity towards healthy human fibroblasts (IC50 70.4 M) compared to complex 2
(IC50 19.0 M) (Figure 5, Table 2). Interestingly, previous analysis of gold(I) complexes
containing DAPTA and PTA in combination with thionate ligands revealed cell growth
inhibition of the cisplatin sensitive cell line A2780 with similar IC50 at 72 h.[28,63]
Our data indicate that, among the mononuclear compounds, complex 5, with PPh3, and
complex 6, which like auranofin is combined with PEt3, showed the highest
antiproliferative activity in A2780 cell line. Moreover, cytotoxicity in A2780 cells of
complex 6 (IC50 0.3 ± 0.1 M after 48h) is slightly higher than that of auranofin (0.5 ±
0.1 M after 72h).[64] Despite this higher cytotoxicity in A2780 tumor cell line, the IC50
of both compounds in healthy fibroblasts at 48 h is lower than the corresponding value
for complex 4. These data seem to indicate that the latter is a more attractive complex for
ovarian cancer therapy, as it allows the use of a higher concentration range with lower
toxicity to normal cells (no effect on cell viability for concentrations up to 100 M).
14
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
Concerning dinuclear complexes 7-10, a decrease of cell viability was observed for the
two tumor cell lines and healthy human fibroblasts in a concentration-dependent manner.
Cytotoxicity of complexes 8-10 follow the same trend, with IC50 ranging from 0.3 to 0.4
M in A2780 cells, 1.3 to 4.9 M in HCT116 cells and 15.2 to 36.4 M in human healthy
fibroblasts. With the exception of complex 7, the dinuclear complexes showed a higher
cytotoxic effect in HCT116 cells compared to the mononuclear complexes 2-6,
independently of the ligand. This higher cytotoxicity of dinuclear complexes is correlated
with the higher cytotoxicity of their respective ligands (Figures 4, 5, S36, Tables 2 and
S2). Nevertheless, the dinuclear complexes 8-10 showed a higher cytotoxicity in normal
healthy fibroblasts compared to their respective ligands, demonstrating their lower
applicability due to the higher possible side-effects in healthy cells.
The flexibility of phosphine conformation showed to have a direct influence on the
biological activity of the complexes. The higher rigidity of 7, conferred by a triple bond
between phosphines (Scheme 1), resulted in an increase of the IC50 for both cancer cell
lines, when compared to the other dinuclear complexes 8-10 (Table 2).
Concerning the cellular uptake of the mononuclear complexes by A2780 cells, results
suggested a correlation between IC50 and complex intracellular concentration, with higher
Au concentrations for complex 5 treated cells (7.2 % Au, IC50 0.1 M) and lower Au for
complex 3 treated cells (0.3 % Au, IC50 17.4M) (Table 2, S3). This correlation can be
translated to dinuclear complexes, where a lower toxicity correlates with intracellular
concentration of the complex 7, probably due to the rigidity of the bond between the
phosphines (Scheme 1, Table S3). However, when comparing mononuclear and dinuclear
complexes, a much higher Au concentration in cells treated with complexes 8-9 was
observed, reaching 46.7 % of intracellular gold in complex 8, with IC50 0.4 M (Table 2,
S3).
Considering the intracellular concentration of all the complexes, one might postulate that
mononuclear compounds show higher cytotoxicity than the dinuclear counterparts
because higher intracellular concentrations of dinuclear compounds resulted in similar
IC50 compared to mononuclear complexes. For example, the mononuclear complex 6 and
dinuclear complex 9 have identical IC50 (0.3 M) but an intracellular gold concentration
of 4.0 % and 41.7 %, respectively (Table 2, S3).
15
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
As such, complex 4 was selected for further biological characterization due to its higher
specificity in A2780 tumor cell line and low toxicity in normal cells (high therapeutic
index).
In a first approach, we took advantage of complex 4 fluorescence (excitation and emission
wavelengths 310/410 nm, respectively in Tris-HCl 5 mM, pH 7.0, NaCl 50 mM) to
evaluate the uptake by A2780 cells. A 2.5-fold increase of fluorescence intensity in cells
treated with complex 4 compared to the non-treated cells was observed after a 3 h
incubation, suggesting its fast internalization in A2780 cells (Figure 6). These
observations corroborate and support the cell uptake data. Interestingly, after 6 h
incubation, a decrease in fluorescence intensity of treated cells to the level of the control
cells suggests that ligand displacement might have occurred within the cells or that the
complex has been cleared (e.g. by efflux transporters). Both effects have been described
for gold(I) complexes in colorectal and ovarian carcinoma cells.[65,66]
Figure 6. Internalization of complex 4 in A2780 cells. A) A2780 cells were incubated
for 3 h with RPMI 1640 medium supplemented with 10 % (v/v) fetal bovine serum and
A1) 0.05 % (v/v) DMSO (control) and A2) 10 x IC50 at 48 h of complex 4. After
replacing the medium with phosphate buffer saline cells were immediately visualized on
a Ti-U Eclipse inverted microscope (Nikon, Tokyo, Japan) with DAPI filter cube
(Nikon) (excitation filter of 340-380 nm and barrier filter at 435-485 nm). B)
Fluorescence intensity of A2780 cells incubated for 3 h, 6 h and 18 h with control (grey
bars) and 10 x IC50 at 48 h of complex 4 (black bars). Fluorescence intensity of cells
was corrected with background fluorescence and normalized to the equivalent of cells
16
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
challenged with vehicle only (DMSO). Error bars represent SEM of 5 independent
images with at least 10 cells per image. * p-value < 0.05 relative to the value of NS-023
treated cells for 6h and 18h.
Hoechst 33258 shows a high affinity to DNA, allowing to detect nuclear modifications
induced by apoptosis, including chromatin condensation or nuclear fragmentation that
ultimately lead to cell death.[67] So, a preliminary analysis was performed using this
strategy to understand whether the decrease in A2780 cells viability by complex 4 was in
fact due to apoptosis, i.e. cell death. Results show that after 9 h incubation with complex
4, an increased number of nuclei showed apoptotic events (Figure 7), that agrees with the
internalization results of complex 4 in A2780 cells and its induced cytotoxicity (Figures
4 and 6).
Figure 7. Apoptotic morphological changes in A2780 cells treated with complex 4. A)
Representative images of A2780 cells incubated for 9 h with RPMI 1640 medium
supplemented with 10 % (v/v) fetal bovine serum and A1) 0.05 % (v/v) DMSO
(control) and A2) the IC50 value at 48 h of complex 4 and stained with Hoechst 33258
for visualization of apoptotic nuclei (excitation and fluorescence emission at 352 nm
and 461 nm, respectively). Three random microscopic fields with circa 30 nuclei were
photographed using an AXIO Scope (Carl Zeiss, Oberkochen Germany). White arrows
point to typical morphological features of apoptosis, including chromatin condensation
and nuclear fragmentation. B) Percentage of apoptotic cells after 9 h incubation of
A2780 cells with complex 4 and respective vector control. * p-value < 0.05 relative to
the control sample.
17
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
The cytotoxic mechanisms of auranofin in tumor cells are not totally understood, but it is
generally accepted that these involve the induction of intrinsic apoptosis triggered by
increased oxidative stress due to inhibition of thioredoxin reductase.[13,14] As such, the
induction of apoptosis by complex 4 was evaluated by calculating the apoptotic index,
BAX/BCL-2 ratio. Intrinsic apoptosis is regulated by the BCL-2 family of proteins that
is composed by pro-apoptotic proteins, such as BAX, and by anti-apoptotic proteins, such
as BCL-2.[68] An increased BAX/BCL-2 ratio is indicative of increased intrinsic
apoptosis.[68] After 3 h incubation, no alterations were registered. However, at 18h, the
apoptotic index increased almost 2 times in complex 4 treated cells relative to the control
(Figure 8). An increased apoptosis index was also observed at 48 h (Figure 8C). Together,
these data indicate that cell death is triggered by apoptotic events in line with those
observed of auranofin.
Figure 8. Expression of BAX and BCL-2 proteins after A2780 cell line incubation with
complex 4. A) Western-Blot for quantification of BAX, BCL-2 and -actin in A2780
cell line after incubation for 3 h, 18 h and 48 h with IC50 at 48 h of 0.05 % (v/v) DMSO
(control) or complex 4. Protein band intensity was first normalized against β-Actin and
then with sample incubated for 3 h with DMSO. B) Protein expression of BAX (black
bars) and BCL-2 (grey bars). C) Apoptotic index of A2780 cells, calculated by
BAX/BCL-2 ratio, after incubation with control (black bars) and complex 4 (grey bars).
* p-value < 0.05 (obtained by t-test analysis).
18
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
To further corroborate this intrinsic apoptotic event, the mitochondrial membrane
potential was evaluated. Mitochondrial membrane permeabilization, which mainly
consist in dissipation of mitochondrial potential and increased production of ROS, is
initiated by pro-apoptotic members of the BCL-2 protein family and is the point of no
return in the events that lead to intrinsic apoptosis.[69] The mitochondrial membrane
potential was assessed by calculating JC-1 monomer/aggregate ratio. JC-1 is a green
fluorescence dye that, when accumulates at the mitochondrial matrix with an intact
electrochemical gradient, suffers a red-shift of the emission peak.[70]
Dissipation of the mitochondrial membrane potential, was observed after 48 h incubation
of A2780 cells at IC50 of complex 4 (Figure 9). Interestingly, mitochondria
hyperpolarization was observed after 3 h and 18 h incubation, regardless of some
heterogeneity.
Figure 9. Mitochondrial membrane potential of A2780 treated with Complex 4
evaluated by the measurement of JC-1 fluorescence. A) A2780 cells were incubated for
3 h with RPMI 1640 supplemented with 10 % (v/v) fetal bovine serum and A1) 0.05 %
(v/v) DMSO (Control) and A2) IC50 at 48 h complex 4 and stained with JC-1. Five
random microscopic fields with circa 30 nuclei were photographed using an AXIO
19
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
Scope (Carl Zeiss, Oberkochen Germany) using for the same field brightfield image
acquisition, rhodamine fluorophore settings (excitation and emission wavelength 540 /
570 nm, respectively) for JC-1 aggregate conformation (JC-1 aggregate), and FITC
settings (excitation and emission wavelength 485 / 535 nm, respectively) for image
acquisition of the monomer conformation (JC-1 monomer). White arrows point to cells
with different membrane potential than the overall cells in the image. B) JC.1 ratio of
monomer vs aggregate fluorescence obtained after 3 h, 18 h and 48 h incubation of
A2780 cells. Values were normalized to the JC-1 ratio of untreated cells. * p-value <
0.05 relative to the value of DMSO at the same time point.
In line with results obtained for mitochondrial potential, ROS production in A2780 cells
under the same experimental conditions increased significantly after 48 h incubation
(Figure 10). These findings suggest that, unlike auranofin, where an increase of ROS at
the first hours of incubation is observed due to an impairment of thioredoxin reductase
activity which in turn triggers apoptosis, in the presence of complex 4, the increase of
intracellular ROS occurs due to the dissipation of mitochondrial.[14]
40
35
*
30
H 2 O 2
25
*
C o n tro l
20
C om plex 4
15
6
4
2
0
h
h
h
3
8
8
1
4
Figure 10. Reactive oxygen species (ROS) production in A2780 cells treated with
compound 4. A2780 cells were incubated for 3 h, 18 h and 48 h with RPMI 1640
medium supplemented with 10 % (v/v) fetal bovine serum and 25 M Hydrogen
peroxide (H2O2) (Black bars), 0.05 % (v/v) DMSO (light grey bars) and IC50 at 48 h of
compound 4 (dark grey bars) and treated with H2DCF-DA (2’,7’-
20
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
dichlorodihydrofluorescein diacetate). ROS production was quantified in an Attune
acoustic focusing cytometer (ThermoFisher Scientific) by the release of the green
fluorescent compound DCF (excitation and emission wavelengths between 492-495 nm
and 517-527 nm, respectively), in a reaction catalyzed by intracellular estereases.
Values were normalized to the fluorescence of untreated cells. * p-value < 0.05 relative
to the value of DMSO at the same time point.
These results suggest that, like other gold(I) compounds,[14] complex 4 induces intrinsic
apoptosis. Accumulation of the complex in A2780 cells is higher for the first hours of
incubation (Figure 6), when cells start to respond to the presence of the complex, resulting
in mitochondrial hyperpolarization (Figure 9). The increased membrane polarization
might be due to an increased metabolism induced by the complex. In fact, such response
has been suggested by proteomic analysis of cisplatin-resistant A2780 cells exposed to
gold(III) compounds, where an alteration of proteins involved in glycolytic pathway was
observed.[64]
The hyperpolarization of the mitochondria is still observed after 18 h, when the apoptotic
events have already started (Figures 7, 8). A slight increase of ROS is then observed
(Figure 10). After 48 h, most cells are in apoptosis, resulting in a decrease of
mitochondrial membrane potential and increase of ROS.
21
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
Conclusions
The synthesis of a series of mono- and dinuclear gold(I) complexes containing the same
chromophoric unit (4-ethynylanile) and that differ on the phosphine located at the second
coordination position has been successfully obtained by making react the [Au(4-
ethynylaniline)]n polymer with the corresponding phosphane. The presence of carboxylic
derived phosphanes and diphosphanes in cis-conformation avoids the formation of the
products in pure form.
The presence of the same choromophoric unit in all complexes let us to perform a
systematic analysis of the effect of phosphanes on the biological activity of gold(I)
alkynyl complexes and retrieve some important conclusions: i) it is possible to modulate
the selectivity of the treated tumor since mononuclear complexes are more cytotoxic
against A2780 ovarian cells while dinuclear complexes are effective also against HCT116
colorectal cells; ii) the presence of PNaph3 phosphine has a great selectivity against
A2780 ovarian cells, having no cytotoxic effect neither on HCT116 colorectal cells nor
on normal fibroblasts, making it a very interesting candidate of ovarian cancer with very
low secondary effects; iii) the flexibility of the diphosphines has a direct influence on the
biological activity, since the only dinuclear complex that do not present any denoted
effect is that containing the higher rigid bridging ligand (dppa, complex 7); iv) the results
obtained for mitochondrial potential and ROS production in A2780 cells under the same
experimental conditions suggest that, unlike auranofin, the increase of intracellular ROS
could be due to an increased apoptosis, and not to thioredoxin reductase inhibition.
The results reported herein are of great importance to improve the understanding of the
mechanism of activity of gold(I) complexes for future applications in ovarian cancer
therapy.
22
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
Experimental Section
General procedures
All manipulations have been performed under prepurified N2 using standard Schlenk
techniques. All solvents have been distilled from appropriated drying agents. Commercial
reagents 1,3,5-triaza-7-phosphatricyclo[3.3.1.13.7]decane (PTA, Aldrich 97%), 3,7-
diacetyl-1,3,7-triaza-5-phosphabicyclo[3.3.1]nonane (DAPTA, Aldrich 97%), tri-1-
naphthylphosphine (PNaph3, Aldrich, 97%), triphenylphosphine (PPh3, Aldrich, 97%),
triethylphosphine (PEt3, Strem Chemicals, 99%), bis(diphenylphosphino)acetylene
(dppa, Aldrich, 98%), trans-1,2-bis(diphenylphosphino)ethylene (dppet, Aldrich, 97%),
1,3-bis(diphenylphosphino)propane (dppp, Aldrich 97%), and 4-ethynylaniline (Aldrich,
97%) were used as received. Literature methods were used to prepare AuCl(tht)[71] and
1,2-bis(diphenylphosphino)ethane (dppe).[72]
Physical measurements
Infrared spectra have been recorded on a FT-IR 520 Nicolet Spectrophotometer. 1H NMR
31
(δ(TMS) = 0.0 ppm) and P{1H} NMR (δ(85% H3PO4) = 0.0 ppm) spectra have been
obtained on a Varian Mercury 400 and Bruker 400. ElectroSpray-Mass spectra (+) has
been recorded on a Fisons VG Quatro spectrometer. Absorption spectra have been
recorded on a Varian Cary 100 Bio UV- spectrophotometer and emission spectra on a
Horiba-Jobin-Ybon SPEX Nanolog spectrofluorimeter. Elemental analyses of C, H and
N were carried out at the CCiTUB in Barcelona.
X- ray crystal structure determination.
Single-crystal X-ray data for 4 and 10 was measured using a Rigaku SuperNova dual-
source Oxford diffractometer[73] equipped with an Atlas detector using mirror-
monochromated Cu-Kα (λ = 1.54184 Å) radiation. The data for 8 was measured using
Rigaku SuperNova single-source diffractometer[73] fitted with an Atlas EoS CCD detector
using mirror-monochromated Mo-Kα (λ = 0.71073 Å) radiation. The data for 9 were
measured on a Bruker-Nonius Kappa CCD diffractometer with an APEX-II CCD detector
using graphite-monochromated Mo-Kα (λ = 0.71073 Å) radiation. The data collection
and reduction for 4, 8 and 10 collected using Rigaku instruments were performed using
the program CrysAlisPro[73] and Gaussian face index absorption correction method[73] was
applied. For the data obtained from Bruker Nonius Kappa diffractometer were performed
23
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
using the program COLLECT[74] and HKL DENZO and SCALEPACK.[75] The intensities
for data collected using Bruker Nonius Kappa diffractometer were corrected for
absorption using SADABS[76] with multi-scan absorption correction type method. All
structures were solved with direct methods (SHELXS)[77] and refined by full-matrix least
squares on F2 using the OLEX2 software,[78] which utilizes the SHELXL-2013 module.[77]
Constraints and restraints were used where appropriate for disordered models. Hirsfeld
surface analysis details.[79]
X-ray crystallographic data for 4, 8, 9 and 10 are given in Supplementary information.
CCDC-1844226 - 1844229 contains the supplementary crystallographic data for this
paper. These data can be obtained free of charge from The Cambridge Crystallographic
Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
Cytotoxicity in human cell lines
Cell culture
Human ovarian carcinoma (A2780) and colorectal carcinoma (HCT116) were grown in
RPMI 1640 and Dulbecco’s modified Eagle’s medium (DMEM), respectively (Invitrogen
Corp., Grand Island, NY, USA) both supplemented with 10% fetal bovine serum and 1%
antibiotic/antimycotic solution (Invitrogen Corp.) and maintained at 37 ºC in a humidified
atmosphere of 5 % (v/v) CO2.[80,81] Normal human primary fibroblasts were grown in the
same conditions as HCT116 cell line, supplemented with 1 % MEM non-essential amino
acids (Invitrogen Corp.).[82,83] All cell lines were purchase from ATCC (www.atcc.org).
Complexes exposure for dose-response curves
Cells were plated at 7500 cells/well in 96-well plates. Media was removed 24 h after
platting and replaced with fresh media containing: 0.01 - 100 µM of each complex 2-10
or each ligand or 0.1 % (v/v) DMSO (vehicle control). Complexes and ligands solutions
were prepared from a concentrated stock solution (in DMSO).
Viability assays
After 48 h of cell incubation in the presence or absence of the complexes or each ligand,
cell viability was evaluated with CellTiter 96® Aqueous One Solution Cell Proliferation
Assay (Promega, Madison, WI, USA), using 3-(4,5-dimethylthiazol-2-yl)-5-(3-
carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt (MTS) as
previously described.[83,84] In brief, this is a homogeneous, colorimetric method for
24
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
determining the number of viable cells in proliferation, cytotoxicity or chemosensitivity
assays. The Cell Titer 96® Aqueous Assay is composed of solutions of MTS and an
electron coupling reagent (phenazine ethosulfate, PES). MTS is bioreduced by cells into
a formazan product that is soluble in tissue culture medium. The absorbance of the
formazan product at 490 nm can be measured directly from 96-well assay plates without
additional processing. The conversion of MTS into the aqueous soluble formazan product
is accomplished by dehydrogenase enzymes found in metabolically active cells. The
quantity of formazan product was measured in a Bio-Rad microplate reader Model 680
(Bio-Rad, Hercules, CA, USA) at 490 nm, as absorbance is directly proportional to the
number of viable cells in culture.
Cellular uptake of complexes 2-10 by A2780 cells
To quantify the internalization of complexes 2-10 by A2780, 5x105 cells were exposed to
RPMI supplemented with a concentration of the complex at 10 x the respective IC50, and
incubated for 3 h at 37 ºC, 5 % (v/v) CO2 and 99 % (v/v). The supernatant was then
transferred to a new tube, cells were washed with phosphate buffer saline (PBS) that was
added to the supernatant. Cells were then detached and pelleted by centrifugation at 500
xg, 5 min. The cell pellets were treated with Acqua regia, and the amount of Au quantified
in each sample by Inductively Coupled Plasma-Atomic Emission Spectrometry (ICP-
AES) - contracted service. The fraction of intracellular Au was obtained by dividing the
Au concentration in the cell pellet by the sum of Au concentrations in cell pellet and Au
concentration on the respective supernatant.
Complex 4 internalization in A2780 cells
Taking advantage of the fluorescence of complex 4 (excitation and emission wavelengths
310 nm and 410 nm, respectively in Tris-HCl 5 mM, pH 7.0, NaCl 50 mM), the
internalization of the complex in A2780 cells was investigated by fluorescence
microscopy. With that purpose, A2780 cells were seeded on a 24-well plate with a density
of 1 x 105 cells/well and incubated for 24 h. The medium was replaced by fresh medium
supplemented with 23 M complex 4 (10 x IC50 complex 4 in A2780 cells), 0.05 % (v/v)
DMSO (vector control) or non-supplemented (untreated cells). After 3 h, 6 h and 18 h,
the medium was replaced by PBS and cells were immediately visualized on a Ti-U
Eclipse inverted microscope (Nikon, Tokyo, Japan) with DAPI filter cube (Nikon)
(excitation filter of 340-380 nm and barrier filter at 435-485 nm). Images were acquired
25
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
using NIS Elements Basic software (Nikon). The corrected total cell fluorescence (CTCF)
was quantified for each cell using the formula: CTCF = integrated density – (area of cell
x background mean fluorescence) and fluorescence intensity of DMSO and complex 4
treated samples was normalized with fluorescence intensity of untreated cells.
Represented values correspond to the average and SEM of normalized fluorescence
intensity observed in 20 cells analyzed in five random microscopic fields.
Apoptosis evaluation using Hoechst 33258 staining
A2780 cells were seeded in a 24 well plate containing a cover slip with 12 mm diameter,
at a cell density of 37,500 cells/well. After 24 h, culture medium was replaced by fresh
medium containing 2.3 M complex 4 (IC50 of complex 4 in A2780 cells) or 0.05 % (v/v)
DMSO (vehicle control) and incubated for 9 h at 37 ºC, 5 % (v/v) CO2, 99 % (v/v) relative
humidity. After washing 3 times with PBS, cells were fixed with 4 % (w/v) formaldehyde
in PBS for 20 min in the dark, washed 3 times with PBS and stained for 15 min with 7.5
g/mL Hoechst 33258 (LifeTechnologies). After wash 3 times with PBS coverslips
containing the stained cells were placed on top of a glycerol drop, sealed. Cell nuclei were
visualized on an AxioImager D2 microscope (Carl Zeiss, Oberkochen Germany), and
acquired and processed using microscope respective software (ZEN Blue edition, 2011).
Normal nuclei showed uniformly distributed chromatin; condensed or fragmented
chromatin was visualized on the nuclei and/or when apoptotic bodies were imaged
indicated apoptotic cells. Values correspond to the average and SEM of apoptotic events
observed in five random microscopic fields containing 20 nuclei each.
Western-Blot for BAX and BCL-2 quantification
A2780 cells were plated on T-flasks with 25 cm2 with a cell density of 7.5 x 105 cells.
After 24 h incubation to allow cells to adhere, the medium was replaced by fresh medium
supplemented with 2.3 M complex 4 (IC50 complex 4 in A2780 cells) or 0.05 % (v/v)
DMSO (vector control) and incubated at 37 ºC, 5 % (v/v) CO2 and 99 % (v/v) relative
humidity for 3 h, 18 h or 48 h. Cell monolayer was washed 3 times with PBS and cells
were detached using a cell scraper in 2 mL PBS. Cells were pelleted using a centrifugation
500 xg 5 min and following procedure was identical to a previously described,[85] with
the exception that a PVDF membrane (GE Healthcare) was used and signal acquisition
was obtained in a Hyperfilm ECL (GE Healthcare). The quantification of protein band
26
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
intensity was performed using FIJI software.[86] To calculate the percentage of BAX and
BCL-2 the band intensity was first normalized to β-actin values and then to the value of
the sample incubated for 3 h with DMSO.
Mitochondrial membrane potential
The evaluation of mitochondrial membrane potential was based on measurement of JC-1
dye fluorescence. Briefly, 37500 A2780 cells were seeded on a 24 well plate with a cover
sleep and let to adhere for 24 h. The medium was then replaced with fresh medium
supplemented with 2.3 M complex 4 (IC50 complex 4 in A2780 cells), 0.05 % (v/v)
DMSO (vector control) and with non-supplemented medium (untreated cells). After 3 h,
18 h and 48 h, cells were stained with JC-1 dye (Life Technologies, ThermoFisher
Scientific, Waltham, Massachusetts, EUA) according to manufacturer’s instructions.
After 20 min incubation at 37 ºC, 5 % (v/v) CO2 and 99 % (v/v) relative humidity,
coverslips were placed on a microscopic slide and immediately visualized in an
AxioImager D2 microscope and respective software. The JC-1 is a green fluorescent dye
(emission fluorescence 529 nm) that forms red fluorescent aggregates (emission
fluorescence 590 nm) when accumulates in the mitochondria with higher membrane
potential.[87] Five random microscopic fields were analyzed using in the same field,
autoexposure for brigthfield image acquisition, 150 ms exposure for image acquisition of
JC-1 dye in aggregate conformation, using rhodamine fluorophore settings (excitation
and emission wavelength 540 nm and 570 nm, respectively) and 150 ms exposure for
image acquisition of the monomer conformation, using FITC settings (excitation and
emission wavelength 485 nm and 535 nm, respectively). The relative fluorescent cell
intensity of each cell was measured using FIJI software[86] and CTCF was quantified for
each JC-1 conformation. After calculating the monomer/aggregate CTCF ratio, values
were normalized with ratios obtained for untreated cells.
Quantification of Reactive Oxygen Species (ROS)
A2780 cells were seeded in a 6 well plate in a density of 1 x 105 cells/well. After 24 h
incubation, medium was discarded and replaced with fresh medium supplemented with
2.3 M complex 4 (IC50 complex 4 in A2780 cells), 0.05 % (v/v) DMSO (vector control),
25 M H2O2 (positive control for ROS induction), and with non-supplemented medium
(untreated cells). After 3 h, 18 h and 48 h incubation, cells were detached, centrifuged 5
27
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
min at 500 xg, washed 2 times with PBS and resuspended in a pre-warmed solution of 10
M H2DCF-DA (LifeTechnologies) in PBS. H2DCF-DA (2’,7’-
dichlorodihydrofluorescein diacetate) is a cell permeant non-fluorescent reduced form of
fluorescein that after cleavage of acetate groups by intracellular estereases and oxidation,
forms green fluorescent DCF (2’,7’-dichlorofluorescein) (excitation and emission
wavelengths between 492-495 nm and 517-527 nm, respectively). Data was acquired
using an Attune acoustic focusing cytometer (ThermoFisher Scientific) and respective
software (Attune Cytometric Software, vs 2.1). Data was represented as the mean
fluorescence intensity of each sample normalized to the fluorescence intensity of
untreated cells.
Statistical analysis
All data were expressed as mean ± SEM from at least three independent experiments.
Statistical significance was evaluated using the Student’s t-test; p < 0.05 was considered
statistically significant.
Synthesis and characterization
Synthesis of [Au(C≡C-C6H4NH2)]n (1)
Complex 1 was prepared by introducing slight modifications of the procedure previously
reported on the literature.[29] NEt3 (0.54 ml, 3.87 mmol) and AuCl(tht) (0.70 g, 2.19
mmol) were added into a stirring solution of 4-ethynylaniline (0.26 g, 2.20 mmol) in
dichloromethane (35 ml) under N2 atmosphere at room temperature. After stirring for 30
min the resulting orange suspension was filtered, washed with dichloromethane (3×15
ml) and dried under vacuum. Yield: 92% (0.63 g). IR (KBr, cm-1): 3363 s, 3202 s (N-H);
1994 w (C≡C).
Synthesis of [Au(4-ethynylaniline)(PTA)] (2)
Solid PTA (50.5 mg, 0.32 mmol) was added into a stirring suspension of 1 (0.10 g, 0.32
mmol) in dichloromethane (20 ml) under N2 atmosphere at room temperature. The
reaction mixture turned white after 10 min of stirring. The resulting suspension was
concentrated in half volume after 2 h of stirring and hexane (17 ml) was added to favor
precipitation. The resulting white solid was filtrated, washed with hexane (3 ml) and dried
under vacuum. Purification of the solid was performed by stirring in
dichloromethane:methanol:acetonitrile (10:3:3) mixture. The suspension was
28
This article is protected by copyright. All rights reserved.

10.1002/chem.201802547
Chemistry - A European Journal
concentrated and precipitated with hexane (10 ml) to obtain a white solid in 63% yield
(95.3 mg). 1H NMR (400 MHz, CDCl3, ppm): δH 7.26 (dt, J1 = 8.4, J2 = 2.4 Hz, 2H, 퐻β),
6.55 (dt, J1 = 8.4, J2 = 2.4 Hz, 2H, 퐻α), 4.54 (d, J = 13.2 Hz, 3H, N-CH2-N), 4.47 (d,
31
J=13.2 Hz, 3H, N-CH2-N), 4.22 (s, 6H, N-CH2-P), 3.68 (br, 2H, NH2). P{1H} NMR
(CDCl3, 161.9 MHz, ppm): -49.4. IR (KBr, cm-1): 3424 s, 3355 s (N-H); 2102 w (C≡C).
ESI-MS m/z 471.095 ([M+H]+, calc.: 471.094). Anal. Calc. for C14H18AuN4P C 35.76, H
3.86, N 11.91; found C 35.82, H 3.78, N 11.98.
Synthesis of [Au(4-ethynylaniline)(DAPTA)] (3)
Solid DAPTA (73.7 mg, 0.32 mmol) was added into a stirring suspension of 1 (0.10 g,
0.32 mmol) in dichloromethane (20 ml) under N2 atmosphere at room temperature. The
reaction mixture turned clear yellow after stirring for 10 min. The resulting yellow
suspension was concentrated after 2 h of stirring and hexane (17 ml) was added to favor
the precipitation. The yellow product was filtrated, washed with hexane (6 ml), dried
under vacuum and recrystallized with acetonitrile (10 ml)/ diethylether (7 ml) to give a
1
yellow solid in 51% yield (88.7 mg). H NMR (400 MHz, CDCl3, ppm): 7.27 (dt, J1 =
7.7, J2 = 2.7 Hz, 2H, 퐻β), 6.56 (dt, J1 = 8.5, J2 = 2.4 Hz, 2H, 퐻α), 5.79 (d, J = 22.3 Hz,
1H, N-CH2-N), 5.60 (dd, J1 = 15.6, J2 = 7.0, 1H, P-CH2-N), 4.95 (d, J = 14.3 Hz, 1H, N-
CH2-N), 4.56–4.64 (m, 2H, N-CH2-N and P-CH2-N), 4.09 (dt, J1 =15.9, J2 = 3.0 Hz, 1H,
P-CH2-N), 4.04 (d, J = 14.8 Hz, 1H, N-CH2-N), 3.82 (s, 2H, P-CH2-N), 3.71 (s, 2H, C-
NH2), 3.49 (dt, J1 =16.0, J2 = 4.0 Hz, 1H, P-CH2-N), 2.10 (d, J = 2.2 Hz, 6H, CO-CH3)
ppm. 31P{1H} NMR (CDCl3, 161.9 MHz, ppm): -22.5 (s, 1P) ppm. IR (KBr, cm-1): 3456
s, 3353 s (N-H); 2099 w (C≡C). ESI-MS m/z 543.115 ([M+H]+, calc.: 543.115). Anal.
Calc. for C17H22AuN4O2P C 37.65, H 4.09, N 10.33; found C 37.51, H 4.15, N 10.38.
Synthesis of [Au(4-ethynylaniline)(tri-1-naphthylphosphine)] (4)
Solid tri-1-naphthylphosphine (0.13 g, 0.32 mmol) was added at once into a stirring
suspension of 1 (0.10 g, 0.32 mmol) in dichloromethane (20 ml) under N2 atmosphere at
rt. The solution turned pale yellow and turbid after 10 min of stirring. The resulting
suspension was concentrated to the half after stirring for 2 h, and hexane (15 ml) was
added. The pale-yellow suspension was filtrated, washed with hexane (5 ml) and dried in
vacuum. The crude product was purified by stirring with acetonitrile and adding
1
diethylether to precipitate a pale-yellow product in 56% yield (0.13 g). H NMR (400
MHz, CDCl3, ppm): 8.82 (d, J = 8.0 Hz, 3H, HNaph), 8.02 (d, J = 8.0 Hz, 3H, HNaph), 7.95
29
This article is protected by copyright. All rights reserved.