New Acridine Thiourea Gold(I) Anticancer Agents: Targeting the Nucleus and Inhibiting Vasculogenic Mimicry

Article abstract of DOI:10.1021/acschembio.7b00090

Two new 1-acridin-9-yl-3-methylthiourea Au(I) DNA intercalators [Au(ACRTU)₂]Cl (2) and [Au(ACRTU) (PPh₃)]PF₆ (3) have been prepared. Both complexes were highly active in the human ovarian carcinoma cisplatin-sensitive A2780 cell line, exhibiting IC₅₀ values in the submicromolar range. Compounds 2 and 3 are also cytotoxic toward different phenotypes of breast cancer cell lines MDA-MB-231 (triple negative), SK-BR-3 (HER2+, ERα-, and ERβ-), and MCF-7 (ER+). Both complexes induce apoptosis through activation of caspase-3 in vitro. While inhibition of some proteins (thiol-containing enzymes) seems to be the main mechanism of action for cytotoxic gold complexes, 2 and 3 present a DNA-dependent mechanism of action. They locate in the cell nucleus according to confocal microscopy and transmission electronic microscopy. The binding to DNA resulted to be via intercalation as shown by spectroscopic methods and viscometry, exhibiting a dose-dependent response on topoisomerase I mediated DNA unwinding. In addition, 2 and 3 exhibit potent antiangiogenic effects and are also able to inhibit vasculogenic mimicry of highly invasive MDA-MB-231 cells.

Full text of DOI:10.1021/acschembio.7b00090

Subscriber access provided by HACETTEPE UNIVERSITESI KUTUPHANESI
Article
New Acridine Thiourea Gold(I) Anticancer Agents:
Targeting the Nucleus and Inhibiting Vasculogenic Mimicry
Sergio A. Pérez, Concepcion de Haro, Consuelo Vicente, Antonio Donaire, Ana
Zamora, Juraj Zajac, Hana Kostrhunova, Viktor Brabec, Delia Bautista, and José Ruiz
ACS Chem. Biol., Just Accepted Manuscript • DOI: 10.1021/acschembio.7b00090 • Publication Date (Web): 07 Apr 2017
Downloaded from http://pubs.acs.org on April 10, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Chemical Biology is published by the American Chemical Society. 1155 Sixteenth
Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 47
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
New Acridine Thiourea Gold(I) Anticancer
Agents: Targeting the Nucleus and Inhibiting
Vasculogenic Mimicry
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
†
†
†
†
Sergio A. Pérez, Concepción de Haro, Consuelo Vicente, Antonio Donaire, Ana
†
‡,#
‡
‡
§
Zamora, Juraj Zajac, Hana Kostrhunova, Viktor Brabec, Delia Bautista, and José
,
†
Ruiz*
†
Departamento de Química Inorgánica, Facultad de Química, Biomedical Research
Institute of Murcia (IMIBꢀArrixacaꢀUMU), Universidad de Murcia, Eꢀ30071 Murcia,
Spain
‡
Institute of Biophysics, Academy of Sciences of the Czech Republic, v.v.i.
Kralovopolska 135, 612 65 Brno, Czech Republic
#
Department of Biophysics, Faculty of Science, Palacky University, Slechtitelu 27,
783 71 Olomouc, Czech Republic
§
SAI., Universidad de Murcia, Eꢀ30071ꢀMurcia, Spain
ABSTRACT
2
Two new 1ꢀacridinꢀ9ꢀylꢀ3ꢀmethylthiourea Au(I) DNA intercalators [Au(ACRTU) ]Cl
(
3 6
2) and [Au(ACRTU)(PPh )]PF (3) have been prepared. Both complexes were highly
active in human ovarian carcinoma cisplatinꢀsensitive A2780 cell line, exhibiting IC50
1
ACS Paragon Plus Environment

ACS Chemical Biology
Page 2 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
values in the subꢀmicromolar range. 2 and 3 are also cytotoxic toward different
phenotypes of breast cancer cell lines MDAꢀMBꢀ231 (triple negative), SKꢀBRꢀ3
(
HER2+, ERα− and ERβ−) and MCFꢀ7 (ER+). Both complexes induce apoptosis
through activation of caspaseꢀ3 in vitro. While inhibition of some proteins (thiolꢀ
containing enzymes) seems to be the main mechanism of action for cytotoxic gold
complexes, 2 and 3 present a DNAꢀdependent mechanism of action. They locate in cell
nucleus according to confocal microscopy and transmission electronic microscopy. The
binding to DNA resulted to be via intercalation as shown by spectroscopic methods and
viscometry, exhibiting a doseꢀdependent response on topoisomerase I mediated DNA
unwinding. In addition, 2 and 3 exhibit potent antiangiogenic effects, and are also able
to inhibit vasculogenic mimicry of highly invasive MDAꢀMBꢀ231 cells.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1. INTRODUCTION
Cisplatin (CDDP), carboplatin and oxaliplatin have become established drugs in
1
ꢀ4
therapy various solid cancers. However, their clinical use has been restricted due to
5
ꢀ7
serious problems such as resistance, nephrotoxicity and neurotoxicity. In order to
circumvent these limitations, different research groups have attracted their attention to
1
,2,4,8ꢀ14
complexes of other precious metals as therapeutic agents.
In this context, gold(I)
and gold(III) compounds have received increasing attention throughout the last decade
15ꢀ29
as anticancer agents.
1
9,22,27
Auranofin and Et PAuCl (Figure 1A)
3
are two interesting complexes that display
their antiproliferative activities specially by interfering with mitochondrial thioredoxin
30ꢀ34
reductase (TrxR).
In addition, more Au(I) compounds have been identified to trigger
their antitumor properties through the inhibition of different enzymes. For example, the
cytotoxic triphenylphosphine complex [Au(NHC)(PPh )]I (Figure 1A) inhibits the
3
3
5
poly(ADPꢀribose) polymerase 1,
and several gold(I) thioureaꢀfunctionalized
2
ACS Paragon Plus Environment

Page 3 of 47
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
3
6
complexes are tyrosine kinase inhibitors while others, such as [Au(TU) ]Cl (TU = 1,3ꢀ
2
37
bis(4ꢀmethoxyphenyl)imidazolidineꢀ2ꢀthione), are thioredoxin reductase inhibitors.
On the contrary, it is well known that some Au(III) compounds interact with DNA
38,39
40
covalently
or induce cell death as effective topoisomerase inhibitors.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
On the other hand, solid tumors need an adequate nutrients and oxygen supply to
41,42
allow continued growth and metastasis.
Thus, inhibition of angiogenesis is a
41,43,44
desirable goal,
and combination of angiogenesis inhibitors with chemotherapy
10,45ꢀ47
would be of great interest.
Several metal complexes have been recently reported
4
8ꢀ58
as angiogenesis inhibitors
including Au(I) and Au(III) complexes shown in Figure
1B.
3
ACS Paragon Plus Environment

ACS Chemical Biology
Page 4 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 1. Some antitumor active gold complexes (A) and dual antiangiogenic and
33,55,58
antitumor gold complexes (B)
The soꢀcalled PTꢀACRAMTU and AUꢀ
ACRAMTU prepared by Bierbach (vide infra) (C).
4
1,59,60
In addition, vascular mimicry
(VM) enables an additional dissemination way
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
1ꢀ63
within the tumor,
and in fact, current antiꢀvascular therapies targeting endothelial
cells are not relevant in VM tumors. Therefore dual action compounds for inhibiting
both vasculogenic mimicry and angiogenesis are required to improve the clinical
64,65,66
results.
Prompted by the success of the soꢀcalled PTꢀACRAMTU (ACRAMTU = 1ꢀ[2ꢀ
(acridinꢀ9ꢀylamino)ethyl]ꢀ1,3ꢀdimethylthiourea), a DNAꢀtargeted platinumꢀacridine
67,68
conjugate,
and the dicationic gold(I) analogue AUꢀACRAMTU (Figure 1C) which
69
shows a potent activity against Mycobacterium tuberculosis, we have prepared two
closeꢀrelated
acridine
gold(I)
complexes,
2
[Au(ACRTU) ]Cl
and
[
Au(ACRTU)(PPh
3
)]PF (ACRTU= 1ꢀacridinꢀ9ꢀylꢀmethylthiourea), in order to explore
6
their antitumor activity together with their possible antiꢀangiogenic properties. We have
aimed to more thoroughly elucidate their mechanism of action in ovarian and breast
cancer cell lines, with a wide variety of in vitro and biochemical techniques, including
apoptosis studies and intracellular localization studies by confocal microscopy and
transmission electron microscopy. In addition, their DNA intercalating characteristics
have been studied by spectroscopic methods, viscosity measurements and
topoisomerase I inhibition studies. Finally, the angiogenesis inhibiting properties of the
new Au(I) compounds in the immortalized humanendothelial cell line EA.hy926 and
the vasculogenic mimicry effect in the highly invasive MDAꢀMBꢀ231 cells have been
studied.
4
ACS Paragon Plus Environment

Page 5 of 47
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
2. RESULTS AND DISCUSSION
2.1. Synthesis and Characterization of the Au(I) Compounds. 2.1.1. Synthesis of
the New Complexes. The gold complexes 2 and 3 (Chart 1) were prepared from the
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
6
8
reaction of the previously reported 1ꢀacridinꢀ9ꢀylꢀ3ꢀmethylthiourea ACRTU ligand (1)
in the proper molar ratio and solvent with [AuCl(tht)] (tht = tetrahydrothiophene) or
[AuClPPh ], respectively. Complexes 2 and 3 were characterized by NMR in DMSOꢀd6
3
31
(
Figures S1ꢀS8 in the Supporting Information), IR and UVꢀvis spectroscopy. The
P
NMR spectrum of 3 shows a unique resonance for coordinated PPh
3
ligand (Figure S5).
The characteristic optical properties of complexes 2 and 3 were assigned mainly to
intraligand (IL) transitions centered on the acridine group (Figure S9). Elemental
analyses showed that the purity of the compounds was at least 95%, and their positiveꢀ
+
ion ESIꢀMS displayed the [M] peaks in agreement with the calculated isotopic pattern.
HPLC chromatogram of 2 shows a single peak at retention time of 11.1 min (Figure
S11A) corresponding to complex 2 according to MS TOF spectrum, while the HPLC
chromatogram of 3 followed by MS TOF studies indicated that it undergoes partial
+
disproportionation in homoleptic complexes 2 and [Au(PPh ) ] (Figure S11B), maybe
3
2
because the lack of stability for 3 in the column and/or the acidic media used for this
experiment.
5
ACS Paragon Plus Environment

ACS Chemical Biology
Page 6 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2.1.2. Structure Determination of Complex 3. The crystal structure of 3 has been
established by Xꢀray diffraction (Figure 2). The structure confirms the formation of a
1
:1 conjugate with the ligand being connected to an almost linear gold through thiourea
sulfur. Intermolecular πꢀπ interactions between acridine ligands are observed and
molecules of 3 are stacked to give centrosymmetric pairs (Figure S12). Intermolecular
interactions contacts C–H···F–P and N–H···F–P are also observed (Table S3 and Figure
S13).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Molecular structure of gold cation with atom numbering schemes for 3 (30%
thermal ellipsoids). Selected bond lengths (Å) and angles (deg): Au(1)−P(1) =
2
.2639(5), Au(1)−S(1) = 2.3017(5), P(1)−Au(1)−S(1) = 171.027(17).
2.2. Biological Activity. 2.2.1. Cytotoxicity Studies. The cytotoxicity of 2 and 3 was
evaluated (Table 1 and Figures S14 and S15) towards A2780 cells and a panel of human
breast cancer cell lines, MDAꢀMBꢀ231 (triple negative), SKꢀBRꢀ3 (HER2+, ERα− and
ERβ−) and MCFꢀ7 (ER+). The cytotoxicity of CDDP, the free ligand 1 and AuClPPh
3
was also determined for comparison purposes.
6
ACS Paragon Plus Environment

Page 7 of 47
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
The results in Table 1 show a distinct antiproliferative activity of the compounds 2
and 3 according to the tested cell lines. Compound 3 resulted to exhibit a higher potency
than 2 in all the studied cell lines, being both complexes extremely cytotoxic
(submicromolar IC50) toward A2780 cells. This effect, although high, is not so acute for
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
the breast cancer cell lines. Interestingly, both compounds exhibit the higher activity in
the negative MDAꢀMBꢀ231 cell line, becoming approximately 13x more cytotoxic than
CDDP. It is to be noted that the ligand 1 is not cytotoxic (IC50
>
50
ꢀM except for
MDAꢀMBꢀ231) and complex AuClPPh shows relatively low IC values (5.60
−
14
3
50
ꢀM), although always higher than 3.
Table 1. IC50
(ꢀM) for CDDP, AuClPPh
3
and Compounds 1
−
3 at 48 h.
Compound
A2780
MDAꢀMBꢀ231
SKꢀBRꢀ3
MCFꢀ7
LLCꢀPK1
EA.hy926
1
> 50
10.94 ± 0.56
2.75 ± 0.40
2.69 ± 0.87
10.33 ± 3.81
34.90 ± 7.54
> 50
> 50
ꢀ
ꢀ
2
3
0.88 ± 0.20
0.40 ± 0.18
14 ±1
3.62 ± 0.14
1.98 ± 0.33
5.60 ± 0.24
7.79 ± 0.56
5.32 ± 0.65
2.35 ± 0.39
8.68 ± 0.29
8.56 ± 2.64
2.57 ± 0.38
ꢀ
3.57 ± 0.16
1.17 ± 0.07
ꢀ
AuClPPh
3
CDDP
2.88 ± 0.09
12.58 ± 1.88
3.10 ± 0.07
9.86 ± 0.36
1
Due to the high stability of 2 and 3 in solution, no free ligand was detected by H
NMR after 24 h incubation in mixtures of DMSOꢀd /D O (1:1) (Figure S16) and DMSOꢀ
d /D O (100 mM NaCl) (1:1) (Figure S17), the cytotoxic effect was exclusively
6
2
6
2
attributed to the gold complexes or its metabolites.
2.2.2. CellꢀSelectivity of Au(I)ꢀAcridine Complexes
2 and 3 in A2780 and MDAꢀMBꢀ
2
31. We have found that 2 and 3 were more selective for cancerous cells over healthy
cells than CDDP. Table 1 show their IC50 values in a kidney healthy cell line, LLCꢀ
47
PK1. Thus, selectivity factors for 2 and 3 (Figure S18) range from 6−9x in A2780 to
1
0−35x in MDAꢀMBꢀ231 cells. Likewise, IC50 values for 2 and 3 in the EA.hy926 cell
7
ACS Paragon Plus Environment

ACS Chemical Biology
Page 8 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
line were also determined (Table 1), showing that their ability to inhibit angiogenesis
4
7
took place at subꢀcytotoxic concentrations (vide infra, Section 2.4).
2.2.3. Apoptosis Studies. Apoptotic studies were undertaken in MDAꢀMBꢀ231 cell
4
7
line by flow cytometric assay. As it can be seen in Figure 3, 2 and 3 caused an
elevated percentage of apoptosis (82 and 98% respectively) in MDAꢀMBꢀ231 cells.
Similar values were found in MCFꢀ7 cells after 24 h of incubation with 2 and 3 (Figure
S19). TEM images (Figure 7) supported also apoptosis (vide infra).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. Apoptosis analysis of MDAꢀMBꢀ231 cells after 24 h for untreated cells (A)
and treatment with 2 (C) and 3 (D) using an annexin V/PI assay (at concentrations 2
IC50). Caspaseꢀ3 activation in MDAꢀMBꢀ231 cells caused by exposure to complexes 2,
and CDDP (B). *p < 0.05 was considered to be statistically significant. The results are
×
3
expressed as mean values ± SD for two independent samples for each compound.
To clarify the mechanism by which the complexes induce apoptosis, the activity of
7
0
caspaseꢀ3, a cysteine protease, was studied using the caspaseꢀ3 fluorescence detection
8
ACS Paragon Plus Environment

Page 9 of 47
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
kit (FITCꢀDEVDꢀFMK) in MDAꢀMBꢀ231 (Figure 3B) and MCFꢀ7 (Figure S19B) cell
line after 24 h exposure to the complexes 2 and 3 (1 or 2 ꢀM). As shown in Figure 3B,
an increase of caspaseꢀ3 activity is noted for the compounds with respect to the control
and CDDP. Therefore, 2 and 3 induce apoptosis through activation of caspaseꢀ3 in vitro.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
.2.4. Cell Cycle Arrest. The effect of Au(I) compounds on the cell cycle was
examined. MDAꢀMBꢀ231 cells were exposed to the compounds at concentrations
corresponding to their IC50 for 24 h at 37 ºC, collected then treated with 70% (vol)
EtOH, RNase and dyed with PI and finally, analyzed by flow cytometry. The
histograms obtained for 2 and 3 are shown in Figure 4. As can be observed in Figure 4B
an increased amount of cells stalled in the G2/M phase ocurred due to the presence of 2
and 3. Therefore, the G2/M checkpoints impedes the entry into mitosis as result of DNA
71,72
damaging agents,
leading the cells to die through apoptosis.
9
ACS Paragon Plus Environment

ACS Chemical Biology
Page 10 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 4. Cell cycle analysis of MDAꢀMBꢀ231 cells. IP fluorescence intensity (FL2)
for negative control (A) and cells exposed to 2 (C) and 3 (D). Effect of 2 and 3 in the
studied cell cycle distribution after 24 h incubation (B).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2.2.5. Cellular Localization of Ligand
1
and Complexes
2 and 3 by Confocal
Microscopy. Intracellular distribution of 1
−
3 was investigated in MCFꢀ7 cells.
Mitochondria and nuclei were stained for 30 min with 100 nM of the far redꢀfluorescent
dye MitoTracker Red and DAPI (binds strongly to AꢀT rich regions), respectively. After
2
h incubation with 2 (5 ꢁM) and 3 (2 ꢀM), the complexes were excited at a wavelength
that does not interfere with the other dyes. The luminescence of 2 and 3 was highly
visible in the tumor cells (Figure 5F). It can be observed that the area of the emission of
2
or 3 (Figure 5 E) does not superimpose with MitoTracker Red dye (Figure 5G).
10
ACS Paragon Plus Environment

Page 11 of 47
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
Figure 5. Determination of intracellular localization of 1ꢀ3 by costaining experiments
with DAPI and MitoTracker Red in MCFꢀ7 cells. Phase contrast (A) and (D). Ligand 1
(B) and overlay of phase contrast and 1 (C). Overlay images of DAPI and phase
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
contrast (E). Overlay image of 2 or 3 and phase contrast (F). MitoTracker Red
fluorescence images (G). Dual excitation images of DAPI and MitoTracker Red (H).
Dual excitation image of 2 or 3 and MitoTracker Red (I). Scale bar
= 30 ꢀm.
In order to avoid possible fluorescence bleed through from DAPI in previous
experiment, we have undertaken also confocal microscopic studies of MCFꢀ7 cells
treated with 2 or 3 alone (Figure S20). It shows clearly that the complexes accumulate
mainly in the nucleus. In addition, as shown in Figures 5B and 5C, ligand 1 is a strong
emitter in the cytosol of MCFꢀ7 cells, but surprisingly not in the nucleus in the same
experimental conditions used for complexes 2 and 3. This rather unexpected result
could be one of the reasons for the lack of cytotoxicity of 1 in this cancer cell line.
Indeed, the non fluorescence signal in the nucleus for ligand 1 may be explained
assuming that 2 h was not enough time to accumulate into the nucleus, although another
73
possibility might be a quenching process in this organelle. Notably, the lack of
extranuclear emission in Figure 5F indicates that no significant dissociation of acridine
ligand in the cytoplasm is observed.
2.2.6. Transmission Electron Microscopy (TEM). This technique can be very useful
for the study of cancer cells treated with the gold complexes, as is a thirdꢀrow transition
74
metal and thus it has a high electron density. None of the conventional staining agents
was added, therefore images were as consequence of the added gold compound.
Ultrathin sections of MDAꢀMBꢀ231 and MCFꢀ7 cells treated with 2 and 3 (2 ꢀM) at 37
ºC for 24 h were observed under TEM as shown in Figure 6. A higher contrast, in
11
ACS Paragon Plus Environment

ACS Chemical Biology
Page 12 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
particular in the nucleus and nucleolus, is observed for cells treated with 2 (Figures 6B
and 6E) and 3 (Figures 6C and 6F) compared to the control. In addition, more electron
density is observed both in euchromatin and heterochromatin in treated cells with both
complexes in comparison with the control, so the packing of DNA does not affect to the
complexation of DNAꢀgold complexes. The results are in agreement with the confocal
microscopy and indicate that these compounds mainly interact with nuclear DNA.
Furthermore, it is noteworthy that some cells showed changes related to apoptosis, such
as blebs at the cell surface (Figure 6B) and cytoplasmic vacuolization (Figure 6F).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. TEM images of untreated MDAꢀMBꢀ231 cells (A) and treated with 2 (B), 3
(C). Images of untreated MCFꢀ7 cells (D) or treated with 2 (E), 3 (F). Nucleus (ꢀ),
nucleolus (ꢀ), mitochondria (ꢀ), cytoplasmic vacuolization (ꢀ) and cytoplasm (ꢀ)
can be distinguished. Cells in an early (E) or late (C) stage of apoptosis can be
distinguished.
2.2.7. Cellular Internalization and Accumulation Studies. Internalization studies for
1−3 (5 µM) in A2780 and MCFꢀ7 cells have been undertaken after different incubation
times (2 min, 30 min and 1 h) by flow cytometry. As observed in Figure S21 the free
ligand 1 and the new complexes are able to internalize into cancer cells very easily
12
ACS Paragon Plus Environment

Page 13 of 47
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
(even after 2 min). Both A2780 and MCFꢀ7 cells exhibited very low level of
autofluorescence (negative controls in Figure S21). The total cellular accumulation of
gold complexes and CDDP in A2780 cell line was studied in order to investigate a
possible relationship between the cellular uptake and in vitro cytotoxicity.
Concentrations of metals in cells exposed to complexes (1 ꢁM) for 24 h were measured
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7
5
by ICPꢀMS. The results are shown in Table 2. A direct relationship between Au
accumulation and activity in A2780 cell line is clearly observed in our measurements.
10,76
Accordingly to log P (Table 2) and RPꢀHPLC studies
(Figure S11), the most
hydrophobic complex 3, which is 2ꢀfold more cytotoxic than 2 (Table 1), accumulates
to the highest extent in A2780 cells. Therefore, there is also a clear relationship between
a higher hydrophobicity and an increase in cellular accumulation. As an increase in
lipophilicity causes enhanced cellular accumulation, it is probably that passive diffusion
7
5
plays a major role in the mechanism of cellular accumulation for the compounds.
Table 2. Accumulation of gold compounds and CDDP in A2780 cells treated for 24 h at
1 µM, and log P values.
6
a
6
a
b
pmol/10 cells
ng/10 cells
log P
2
3
39 ± 2
59 ± 4
10 ± 2
7.6 ± 0.4
12 ± 1
1.9 ± 0.4
0.57 ± 0.07
0.79 ± 0.03
ꢀ2.30 ± 0.04
CDDP
a
Total cellular uptake of tested compounds.
b
75
Log
P (octanol/water) values measured at room temperature.
All results are expressed as the mean ± SD from at least three independent experiments.
The results of cellular localization for 2 and 3 by confocal microscopy and
transmission electron microscopy (sections 2.2.5. and 2.2.6.) show that both complexes
accumulate preferentially in the cell nucleus, where DNA is localized. Thus, it is
reasonable to assume that 2 and 3, after they accumulate in tumor cells, preferentially
interact with DNA in the cell nucleus and not with RNA, since most of RNAs are
localized in the cytoplasm. Therefore, further studies were focused on understanding
interactions of 2 and 3 with DNA in cellꢀfree media.
13
ACS Paragon Plus Environment

ACS Chemical Biology
Page 14 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2.2.8. No MitochondrialꢀDisrupting Activity. The possible mitochondrialꢀdestabilizing
activity of the compounds was rejected after the study of the intracellular reactive
oxygen species (ROS) production and the mitochondrial membrane potential. 2',7'ꢀ
Dichlorodihydrofluorescein diacetate (DCFHꢀDA) was used to measure by flow
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
7ꢀ79
cytometry
the effect of 2 and 3 on the intracellular production of ROS in MCFꢀ7
and MDAꢀMBꢀ231 cells (Figure S22). After 24 h incubation of cells with 2 or 3 (1 or 2
µM, a concentration closed to their IC ) and 30 min after the addition of DCFHꢀDA,
50
the resulting green fluorescence of the oxidation product DCF (2′,7′ꢀ
dichlorofluorescein) was measured. No increments of ROS production were observed
after the exposure of 2 and 3 with respect to the control.
Mitochondria maintain a high negative transmembrane potential at their inner
membrane due to generation of ATP through electron transport chain. This unique
physical property can be exploited for mitochondrial targeting by using positively
charged moieties. For example, Rhoꢀ123, a fluorescent lipophilic dye with a delocalized
positive charge (pK
a
> 10) on the xanthenes ring, is highly specific for mitochondria in
80ꢀ82
living cells.
The small variation in the fluorescence of Rhoꢀ123 after the treatment
of MCFꢀ7 and MDAꢀMBꢀ231 cells with the complexes (Figure S23), indicate no
changes in the mitochondrial membrane potential. These results support the idea that the
complexes exhibit their antitumor activity through a DNAꢀdependent mechanism of
action.
2
.3. Interaction with DNA. TEM and confocal studies indicate nuclear localization
of 2 and 3. However they do not modify the electrophoretic mobility of pUC19 plasmid
Figure S24), probably because the compounds in a relatively strong electric field
(
produced during the electrophoresis could dissociate from DNA. Due to the new
83
cationic compounds contain a ligand with increasased
π
ꢀsurface, further studies on
14
ACS Paragon Plus Environment

Page 15 of 47
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
DNA interactions of 2 and 3 were undertaken by competitive binding experiments,
circular and linear dichroism and viscometry.
2.3.1. Competitive Binding Experiments on Calf thymus DNA. Fluorescence
competition experiments with ethidium bromide (EB) are a useful tool to investigate the
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
interaction mode between 2 and 3 with DNA. EB intercalates into DNA, forming a
6
−1 84
strong emitter adduct (KEB = 1.2
× 10 M ) when excited near 520 nm. However, free
EB is only weakly fluorescent so that quenching of the EBꢀDNA fluorescence is used to
85,86
calculate the magnitude of the interaction between the quencher and DNA.
The
successive additions of 2 to DNA yielded a stronger reduction in the fluorescence at 597
5
−1
nm than for 3 (Figure S25). As expected, complex 2 (KSV = 1.9
×
10 M ) is a stronger
4
−1 14
DNA binder than 3 (KSV = 1.9
×
10 M ). The apparent binding constant values (Kapp)
4
4
−1
for complexes 2 and 3 were 3.7
×
10 and 1.1 × 10 M , respectively (Figures S26 and
8
5
S27).
.3.2. Circular and Linear Dichroism Studies. CD spectrometry of ctꢀDNA treated
2
with 2 or 3 at varying r values (r = [complex]/[DNA] ratio) was used to investigate how
Au(I) complexes bind to DNA and affect its secondary structure (Figure S28). As 2 or 3
are not chiral compounds they yield no intrinsic CD signals. Below 300 nm (where
DNA absorbs), alterations in the CD spectra as a consequence of interactions of the
DNA with 2 and 3 can be due to disturbances in DNA conformation. As 2 and 3 absorb
at the wavelengths lower than 300 nm, the interpretation of CD spectra in terms of
alterations in DNA conformation is impossible. The part of the CD spectrum obtained at
the wavelengths higher than 300 nm (at ligand transition wavelengths) contains the
band corresponding to induced CD due to the interaction of 2 and 3 with a chiral, rightꢀ
handed structure of DNA. The less intense, broad and positive induced CD bands in the
regions of 330−450 nm or 380−450 nm, respectively (Figure S28) are consistent with
15
ACS Paragon Plus Environment

ACS Chemical Biology
Page 16 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
8
7
specific binding of the Au(I) complexes to DNA. Theoretical calculations have shown
that large induced CD signals are expected for groove binding and smaller signals for
88
intercalation. Thus, small induced CD signals of 2 or 3 bound to ctꢀDNA might be
consistent rather with their intercalative DNA binding mode. However, small induced
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
89,90
CDs have been also observed for some DNA groove binders in certain positions
so
that the exact orientation of the ligand bound to DNA with respect to the double helix
axis based on its induced CD may not be unambiguous.
In this respect flow linear dichroism (LD) has been shown to be more useful than CD.
LD (defined as the difference in absorption of light polarized parallel and perpendicular
9
1,92
to an orientation axis
the interaction of complexes 2 or 3 with DNA. Flow orientation of DNA molecules
approximately 250 or more base pairs long) can be achieved by Couette flow where
two concentric cylinders with a small annular gap (0.5 mm) are aligned and one of them
when the sample is subject to flow) was also used for studying
(
93
rotates. Absence of a flow LD at the ligand transition wavelengths is generally an
indication that the small molecules do not interact with DNA or are bound randomly.
On the other hand, the appearance of a ligand LD signal indicates an interaction with
DNA and binding in a specific orientation towards DNA. Thus, our observation (Figure
7
) that sample of ctꢀDNA modified with complexes 2 or 3 yielded the LD ligand signals
indicates that 2 and 3 may intercalate between the bases of double strands of DNA. The
sign of the measured LD at ligand transition wavelengths (at 370−450 nm) and at
wavelengths below 300 nm (where DNA absorbs) was negative (Figure 7) which is
consistent with DNA remaining in a Bꢀconformation when it interacts with 2 or 3. More
importantly, the negative DNA LD signal at 260 nm is markedly enhanced if even low
concentrations of the Au(I) complexes was added (at these loadings the 260 nm LD
signal can be completely ascribed to the DNA LD). This increase of DNA LD is usually
16
ACS Paragon Plus Environment

Page 17 of 47
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
connected with either elongation of DNA molecule when it is unwound as a
consequence of intercalation or stiffening the DNA. The appearance of weaker negative
LD signals at 300−500 nm is consistent with the angle of long axis of the studied
complexes to the axis of DNA double helix greater than 54°. This result supports the
hypothesis and is consistent with the view that arrangement of the acridineꢀthiourea
moiety of 2 or 3 and the DNA base pairs is coplanar; this arrangement is in agreement
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
94
with intercalative DNA binding mode.
ꢀ4
Figure 7. Linear dichroism spectra of ctꢀDNA (3.5 x 10 M) in the presence of
increasing concentrations of complexes 2 (A) and 3 (B). Spectra were run in TrisꢀHCl
buffer (10 mM, pH 7.4) at 25 °C.
2.3.3. Viscometry studies. The results of the competition experiments together with
flow LD suggest for both compounds intercalative DNA binding mode accompanied by
an increase of their rigidity. The classical intercalative DNA binding mode can be
17
ACS Paragon Plus Environment

ACS Chemical Biology
Page 18 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
proved by measuring viscosity of DNA samples, which provides information on
9
5
hydrodynamic properties of DNA. Therefore, the relative specific viscosity of ctꢀ
DNA modified by complexes 2 or 3 was examined as well (Figure 8). It has been shown
that EB, a well known DNA intercalator, increases the relative specific viscosity as a
consequence of its intercalation which results in the increase of persistence length of the
DNA molecule. In contrast to EB, the minorꢀgroove binding drugs Hoechst 33258 does
not significantly affect viscosity of DNA. Increasing the amounts of 2 or 3 in the
reaction with DNA resulted in a continuous increase of the relative viscosity of the
DNA, which is consistent with the lengthening of DNA typical for the classical
intercalation model.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ꢀ4
Figure 8. Relative specific viscosity of ctꢀDNA (3.5 x 10 M) in the presence of
complex 2 (
) or 3 (▲). Viscosity was measured in TrisꢀHCl buffer (10 mM, pH 7.4) at
7 °C. As a positive control, relative viscosity of ctꢀDNA in the presence of intercalator

3
EB (dotted line) is shown. Data represent means ± SD of three separate experiments.
18
ACS Paragon Plus Environment

Page 19 of 47
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
2.3.4. Emission spectra of complexes
2 and 3 in the presence of different types of
DNA. Since 2 and 3 are fluorescent in DMSO at RT (Figure S10), the possible baseꢀ
preference was studied by following the changes in their emission spectra after the
addition of ctꢀDNA and the synthetic polynucleotides poly AT and poly GC. Indeed, 2
and 3 exhibit relatively intense emission in buffer (5 mM TrisꢀHCl, 50 mM NaCl, pH 7)
when excited at 390 nm. Upon the addition of either polynucleotide or ctꢀDNA a strong
quenching of the fluorescence of 2 and 3 is observed with little modification of the
emission bands position (Figure 9). As stated above the luminiscent properties of 2 and
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
were assigned mainly to IL transitions centered on the ACRTU group and early
studies have shown that such excited states are emissive in water and the intensity of the
emissions can be diminished on DNA interaction, indicating an intercalative
83,96
interaction.
The poly GC and ctꢀDNA addition results in a significantly more severe
fluorescence quenching than poly AT, indicating that both complexes have a higher
affinity for GC than AT bases. Although, the affinity of platinumꢀacridinylthiourea
97
conjugates for the adenines in the minor groove has been previously reported, it is not
likely the primary contributor to DNA binding for the Au(I) complexes 2 and 3.
19
ACS Paragon Plus Environment

ACS Chemical Biology
Page 20 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 9. Emission spectra of 2 (A) and 3 (B) in the presence of buffer, polyꢀAT, polyꢀ
GC and calf thymus DNA. [Complexes]: 10
.0), λexc 390 nm.
ꢀM, [DNA]: 50 ꢀM, TrisꢀHCl buffer (pH
7
=
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Finally, the possible covalent interaction between the complexes and DNA was also
1
studied by HꢀNMR. Both complexes were incubated in a DMSOꢀd /D O (1:1) mixture
6
2
at RT with 9ꢀethylguanine in 1:5 molar ratio (Figure S29). After 24 h no reaction was
observed in the assayed conditions, which suggest that covalent binding is not favored
and at the same time support the idea of a nonꢀclassical interaction between the Au(I)
complexes and DNA.
2.3.5. Inhibition of topoisomerase I (topo I). DNA topoisomerases represent a unique
class of enzymes participating in the processes associated with functions of the genome
by tuning the level of DNA supercoiling. Moreover, topoisomerase inhibitors have
98ꢀ101
become clinically important targets for cancer chemotherapy treatments.
and
mechanism of action of inhibitors of topo I are often connected with their intercalative
102,103
DNA binding mode.
A DNAꢀrelaxation assay was used to assess the activity of eukaryotic ct topo I in
104
the presence of the Au(I) complexes. The in vitro assay for topo I activity was based
on monitoring relaxation of negatively supercoiled pSP73KB plasmid in the presence of
the Au(I) complexes with the aid of agarose gel electrophoresis. The effect of DNA
modification by 2 or 3 on topo I catalyzed DNA relaxation was assessed by comparing
the extent of relaxation of nonmodified supercoiled plasmid with plasmid modified by
105
the Au(I) complex. Like other DNA intercalating agents, the Au(I) complexes 2 or 3
exhibited a doseꢀdependent effect on topo I mediated DNA unwinding (Figure 10).
Complex 2 demonstrated somewhat higher inhibitory activity on the relaxation activity
20
ACS Paragon Plus Environment

Page 21 of 47
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
of topo I. Thus, the results of the experiments based on the use of topo I assay are
consistent with the view that both complexes 2 or 3 may bind to DNA via intercalation.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 10. Topo Iꢀmediated relaxation of negatively supercoiled pSP73KB plasmid
ꢀ5
(6.25 x 10 M) in the absence or presence of increasing concentrations in the range of
−12 µM of complexes 2 (A) and 3 (B) analyzed by agarose gel electrophoresis. SC =
supercoiled; OC = open circular.
0
2.4. Inhibition of Tube Formation by Complexes 2 and 3 in EA.hy926
47,106
Endothelial Cells. In vitro angiogenesis assays
were undertaken after incubating
EA.hy926 cells for 16 h with 2 and 3 at sub cytotoxic concentrations (Table 1, Figure
1A for 2). CDDP (1 μM) was used for comparison purposes. The quantification of
resulting formation or inhibition of tubular structures was measured (Figures 11B for 2
1
2
47
and Figure S30 for 3) (μm ) .
21
ACS Paragon Plus Environment

ACS Chemical Biology
Page 22 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 11. Angiogenesis assay for complex 2. Typical images after an ImageJ process
of EA.hy926 cells (A). Quantification of number of meshes and total meshes area of the
capillary structures of EA.hy926 cells treated with compound 2 and CDDP (B). *p <
0.05 was considered to be statistically significant. The results are expressed as mean
values ± SD for two independent samples.
Both complexes 2 and 3 inhibit the formation of vascular tubes at 0.25 µM (14 and 4×
lower concentration than their IC50 in the EA.hy926 cell line, respectively). The
inhibitory effect of both compounds remains when a reduction of the concentration until
0.06 µM is carried out. And, it is to be noted that 2 effectively inhibit the tube formation
at 0.03 µM, i.e. 119× lower concentration than its IC50. Therefore, inhibition of
angiogenesis could contribute to the anticancer capabilities of these compounds. It must
take into account that 2 inhibits angiogenesis at 29× lower concentrations than IC50 in
A2780 cells.
2.5. In Vitro Inhibition of Vascular Mimicry in MDAꢀMBꢀ231 Cells. Highly
metastatic MDAꢀMBꢀ231 cells display the capability to mimic the pattern of embryonic
vasculogenic networks found in aggresive tumors, a phenomenon known as
22
ACS Paragon Plus Environment

Page 23 of 47
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
vasculogenic mimicry (VM). Considering the potent antiproliferative activity against
metastatic breast cancer as well as the antiꢀangiogenic activity, we further evaluated the
mode of actions of 2 and 3 at subꢀIC50 concentrations in MDAꢀMBꢀ231 cells. For that,
similar experiments used to study angiogenesis, such as inhibition of tube formation
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
Figure 12A and 12B) and cell migration were performed (Figure 12C and 12D). The
concentration reduction, from 2 to 1 M, was undertaken in order to completely discard
ꢀ
the cytotoxicity effect as responsible of the inhibition. First, tube formation was
measured after 16 h incubation and quantified as above. As can be observed in Figure
1
2B, both complexes significantly inhibited the capillaryꢀlike formation structure at 1
47,107
ꢀ
M. Subsequently the wound healing assay
was carried out for 2 and 3 after
treatment for 8 h. Figure 12C shows that 2 diminished the mobility of MDAꢀMBꢀ231
cells with respect to the control.
Thus, these compounds markedly inhibited capillaryꢀlike structure formation and cell
migration at concentrations, where no significant cytotoxicity was observed, in MDAꢀ
MBꢀ231 cells, indicating that these effects were not due to their antiproliferative and/or
proapoptotic effects.
23
ACS Paragon Plus Environment

ACS Chemical Biology
Page 24 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 12. Effect of gold complexes on the development of VM in MDAꢀMBꢀ231 cells
in vitro. Typical images after an ImageJ process for culture medium as a control and
with 2 (1 µM for 16 h) (A). Quantification of number of meshes and total meshes area
of MDAꢀMBꢀ231 cells treated with 2 and 3 (B). Effect of 2 on the migration of MDAꢀ
MBꢀ231 cells. Typical images of the wound at the beginning of the experiment and after
47
8 h for a control and with 2 (1 ꢁM) (C). % occupancy of 2 and 3 on the migration of
MDAꢀMBꢀ231 cells (D). *p < 0.05 was considered to be statistically significant. Error
bars indicate ± standard deviations for two sets of three data points for each compound.
3. CONCLUSIONS
Two new monocationic 1ꢀacridinꢀ9ꢀylꢀ3ꢀmethylthiourea Au(I) DNA intercalators
2 3 6
[Au(ACRTU) ]Cl (2) and [Au(ACRTU)(PPh )]PF (3) have been prepared. Both
compounds are highly active in cisplatinꢀsensitive A2780 cells, exhibiting IC values in
50
the subꢀmicromolar range. 2 and 3 are also cytotoxic towards the breast tumor cell lines
24
ACS Paragon Plus Environment

Page 25 of 47
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
MDAꢀMBꢀ231 (triple negative), SKꢀBRꢀ3 (HER2+, ERα− and ERβ−) and MCFꢀ7
(
ER+). We have completed a thorough body of work that contributes to a deeper
understanding of the mechanism of action of a potentially interesting class of new antiꢀ
cancer agents. Thus, the new compounds increased caspaseꢀ3 activity in MCFꢀ7 and
MDAꢀMB231 cells. Compounds 2 and 3 are not able to increase the production of ROS
in MCFꢀ7 and MDAꢀMBꢀ231 cells, the mitochondria membrane potential being
unaffected. Ligand 1 and complexes 2 and 3 are strongly emissive in cells, and flow
cytometry studies reveal that all of them are able to internalize into cancer cells very
quickly. 2 and 3 were localized in the nucleus according to TEM and confocal
microscopy in MCFꢀ7 cells. This is in contrast to ligand 1, which is a strong emitter in
the cytosol but surprisingly not in the nucleus. The new complexes were able to bind to
DNA via intercalation as shown by spectroscopic methods and viscometry, exhibiting a
doseꢀdependent effect on topoisomerase I mediated DNA unwinding. Additionally, 2
and 3 effectively inhibit angiogenesis in the inmortalized human endothelial cell line
EA.hy926 at subcytotoxic concentrations. Furthermore, the new compounds showed
vasculogenic mimicry and antiꢀmigratory properties in the highly invasive MDAꢀMBꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
231 cell line. In summary, in this work we highlight the value of targeting angiogenesis
and vasculogenic mimicry as a strategy for antiꢀcancer therapy and discuss the benefits
of developing small molecules that exhibit these properties in addition to cytotoxicity.
4. EXPERIMENTAL SECTION
Instrumental Measurements. The C, H, N, and S analyses were performed with a
Carlo Erba model EA 1108 microanalyzer. Decomposition temperatures were
determined with a SDT 2960 simultaneous DSCꢀTGA of TA instruments Other details
14
of these methods were the same as in the previously published article. Infrared spectra
1
13
31
were recorded on a PerkinꢀElmer 100 FTꢀIR spectrometer. The H, C and P spectra
25
ACS Paragon Plus Environment

ACS Chemical Biology
Page 26 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
were recorded on a Bruker AV 400 or Bruker AV 600 MHz spectrometer. ESI mass
(positive mode) analyses were performed on a HPLC/MS TOF 6220. Absorption
spectra were measured with a Varian Cary 4000 UVꢀVis spectrophotometer or on a
Perkin Elmer Lambda 750 S spectrometer. Fluorescence measurements were carried out
with a Perkin Elmer LS 55 50 Hz Fluorescence Spectrometer.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Starting materials and reagents. The starting complex [AuCl(tht)] (tht
=
108
tetrahydrothiophene) was prepared by a procedure described elsewhere.
compound [AuCl(PPh
phosphine in 1:1 molar ratio in dichloromethane. The 1ꢀacridinꢀ9ꢀylꢀ3ꢀmethylthiourea
The
3
)] was prepared by reacting [AuCl(tht)] with the corresponding
6
8
ligand (ACRTU) 1 was synthesized by using procedures adapted from the literature.
Ethidium bromide (EB), TrisꢀHCl buffer, pUC19 plasmid, and cisplatin [purity ≥
99.9%] were purchased from SigmaꢀAldrich. Cisplatin stock solution was prepared by
®
dissolving cisplatin in ultrapure water (MilliꢀQ , Merck Millipore) to a final
concentration of 3 mM. ctꢀDNA (42% G + C, mean molecular mass ca. 20 000 kDa)
109,110
was prepared as previously described.
PolyꢀAT (ATATATATATATATATATAT)
and polyꢀGC (GCGCGCGCGCGCGCGCGCGC) were obtained from Eurofins
Genomics. The purity of all biologically evaluated molecules, based on elemental
analysis, is >95%.
Preparation and Characterization of Au(I) Compounds
Synthesis of [Au(ACRTU)
ligand (ACRTU) (267 mg, 1 mmol) were mixed in a 50 mL Schlenck flask.
CH Cl /MeOH (20 mL, 1:1 v/v) were added. The resulting mixture was stirred at RT for
48 h under nitrogen. The yellowꢀorange solid was collected by filtration, washed with
2
]Cl (2). [AuCl(tht)] (160 mg, 0.500 mmol) and the thiourea
2
2
26 6 2 2 2
diethylether and air dried. Yield: 62%. Anal. calcd. for 2 C30H N ClS Au⋅CH Cl : C,
26
ACS Paragon Plus Environment

Page 27 of 47
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
43.67; H, 3.31; N, 9.86; S, 7.53. Found: C, 44.28; H, 3.41; N, 10.30; S, 7.80. Mp: 204
−
1
1
°
C (dec). IR (cm ): ν(NH) 3219; ν(C=S) 748; ν(AuꢀS) 452. H NMR (400 MHz,
DMSOꢀd , TMS): 13.62 (m, 1 H, NH), 8.54 (m, 1 H, NH), 7.80 (m, 6H, ACRTU), 7.40
6
13
(m, 2H, ACRTU), 2.85 (m, 3H, CH
3
6
). C NMR (100.4 MHz, DMSOꢀd ): 173.30,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
+
1
63.30, 143.27, 138.19, 130.76, 126.95, 121.78, 120.72, 34.85. ESI mass spectra
+
ꢀ1
ꢀ1
(
DMSO): m/z +731.13 [C H N S Au] . ε402 = 23663 M cm (in DMSO).
30
26
6 2
Synthesis of [Au(ACRTU)PPh
3 6 3
]PF (3). [AuClPPh ] (100 mg, 0.202 mmol) in acetone
(
15 mL) was placed in a Schlenck flask under nitrogen, then a solution of AgPF (51.14
6
mg, 0.202 mmol) in acetone (15 mL) was added. AgCl immediately was formed. The
resulting suspension was stirred in the dark for 1 h after which was filtered over a
suspension of thiourea ligand (ACRTU) (54.19 mg, 0.202 mmol) in acetone (20 mL)
under nitrogen. After 3 h stirring, in the dark, the reaction mixed was again filtered
through celite and the solvent was removed under reduced pressure to a minimum.
Diethyl ether was added to precipitate a yellow solid, which was washed with diethyl
ether and dried in vacuo. Yield: 50%. Anal. calcd. for 3 C33
28 3 2 6
H N SP F Au: C, 45.48; H,
3
.24; N, 4.82; S, 3.68. Found: C, 45.16; H, 3.19; N, 4.74; S, 3.64. Mp: 273 °C (dec). IR
−1
ꢀ
1
(cm ): ν(NH) 3418, 3349; (PF
6 3
) 857; ν(C=S) 748; (PPh ) 558ꢀ498; ν(AuꢀS) 452. H
NMR (400 MHz, DMSOꢀd , TMS): 12.96 (m, 1H, NH), 9.54 (m, 1 H, NHꢀCH ), 8.09
6
3
(
d, 2H, ACRTU, J
.41 (m, 9H, PPh
DMSOꢀd , H PO
(100.4 MHz, DMSOꢀd
=
8 Hz), 7.81 (t, 2H, ACRTU, J = 7.2 Hz), 7.57 (m, 4H, ACRTU),
31
7
3
), 6.93 (m, 6H, PPh
3
), 2.96 (m, 3H, CH
3
). P NMR (121.5 MHz,
13
6
3
4
): 36.81(s, PPh ), ꢀ144.14 (sept, PF
3
6
, JPꢀF = 710.87 Hz). C NMR
6
): 171.11, 156.54, 139.70, 134.85, 133.28, 132.13, 129.45,
+
1
28.01, 127.36, 123.74, 118.33, 116.37, 31.59. ESI mass spectra (DMSO): m/z
+
ꢀ1
ꢀ1
+
28 3
726.14 [C33H N SPAu] . ε402 = 7473 M cm (in DMSO).
27
ACS Paragon Plus Environment

ACS Chemical Biology
Page 28 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Xꢀray Crystal Structure Analysis. Suitable crystals of 3 were grown from
dichloromethane/hexane. The crystal and molecular structure of 3 have been determined
by Xꢀray diffraction studies (Table S1). A summary of crystal data collection and
refinement parameters for compound are given in Tables S1ꢀS3 in the Supporting
Information. Crystals were mounted on glass fibers and transferred to the cold gas
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
stream of the diffractometer Bruker SMART D8 Quest using monochromated Mo K
radiation (
SADABS). Structures were refined anisotropically using SHELXLꢀ97.
α
λ
= 0.71073 Å). Absorption corrections were based on multiꢀscans (program
1
11
Hydrogen
atoms were included using rigid methyl groups or a riding model, except for H of NH
bonds, which were refined freely.
Cell lines and Culture Media. Human breast cancer (MDAꢀMBꢀ231, SKꢀBRꢀ3 and
MCFꢀ7), pig kidney epithelial LLCꢀPK1, and EA.hy926 cell lines were acquired from
the American Type Tissue Culture Collection (ATCC, USA), while the cell line A2780
was obtained from European Collection of Animal Cell Culture (ECACC, Salisbury,
U.K.). This last cell line was cultured in the RPMIꢀ1640 medium, in the Dulbecco’s
Modified Eagle Medium (DMEM) with a high content of glucose (4.5 g/L), the human
breast cancer cells were all cultured in DMEM medium, but with a low glucose (1 g/L)
concentration. Finally, LLCꢀPK1 cells were cultured in the medium M199. All media
were supplemented with fetal bovine serum (FBS) 10 %, glutamine 1 mM, and
pennicillin and streptomycin. DMEM media was also supplemented with pyruvate 2
mM. Cells were subculture and medium change once a week, and in all case trypsine
0.12%ꢀEDTA 0.05 mM were used. All cell lines were micoplasmaꢀfree, as determinated
1
12
the Hoechts DNA stain method.
Cytotoxicity Assays. Cell proliferation was evaluated by MTT assay in the same way
47
as described in our recently published work. The maximum % of DMSO used was 0.4
28
ACS Paragon Plus Environment

Page 29 of 47
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
(except for CDDP, water diluted) and the measurements were corrected with the
corresponding control.
74
Apoptosis Experiments. The method described elsewhere has been followed.
A
Beckman Coulter Epics XL flow cytometer was used.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Caspaseꢀ3 Assay. MCFꢀ7 and MDAꢀMBꢀ231 cells were plated in 96ꢀwell sterile
4
plates at a density of 1×10 cells/well with 200 µL of complete medium and incubated
overnight. Then, 1 and 2 ꢁM of compounds 2 and 3 were added for 24 h. For caspaseꢀ3
assay, 200 µL of a solution (1 µg/mL) of FITCꢀDEVDꢀFMK was added to each well
and the plate was incubated for 1 h at 37 °C in the dark. Control cells were not treated
with any compound. CDDP (1
ꢀM) was used for comparison purposes. Finally,
fluorescence was measured at 535 nm (excitation at 485 nm) in Fluorstar Omega
spectrophotometer.
5
Cell Cycle Arrest Assays. MDAꢀMBꢀ231 cells (1×10 ) were seeded in 6ꢀwell
plates and left at 37
2
°C in a 5% or 10% CO atmosphere for 24 h. After that, the
corresponding compound at an IC concentration was added to each well (except for
5
0
the control). Cells were incubated in these conditions for 24 h and then, removed from
the wells with trypsinꢀEDTA, collected and centrifuged (250g, 10 min). The supernatant
was removed, cells were washed with PBS and centrifuged (250g, 10 min). The
supernatant was also removed and the cells treated with EtOH at 70% (phosphate
buffered saline, PBS, 30%) for 45 min at 4 ºC and finally EtOH was removed by
centrifugation. Cells were resuspended in PBS and finally, RNase solution and PI was
added at a final concentration of 0.1 mg/ml and 40 mg/ml, respectively. Cells were
incubated in for 30ꢀ60 min in this condition, and PI fluorescence was measured for each
29
ACS Paragon Plus Environment