Molecular mechanism of action of 2-ferrocenyl-1,1-diphenylbut-1-ene on hl-60 leukemia cells

Article abstract of DOI:10.1002/cmdc.201402219

The aim of this work was to investigate the mechanism of action of 2-ferrocenyl-1,1-diphenylbut-1-ene (1) on HL-60 human leukemia cells. While inactive against noncancerous cells, 1 provoked a concentration-dependent decrease in viable tumor cells, primarily via apoptosis, as evidenced by analysis of cell morphology, activation of caspases 3 and 7, increased DNA fragmentation, and externalization of phosphatidylserine. Necrosis was observed only at the highest tested concentration (4 μM). Compound 1 interfered with the cell cycle, causing an accumulation of cells in the G₁/G₀ phase. Interaction of 1 with dsDNA and ssDNA was observed by differential pulse voltammetry and confirmed by hyperchromicity in the UV/Vis spectra of dsDNA, with an interaction constant of 2×10⁴ M^-1. Both the organic analogue 1,1,2-triphenylbut-1-ene (2) and ferrocene were inactive against cancer and noncancer cell lines and did not react with DNA. These results reinforce the idea that the hybrid strategy of conjugating ferrocene to the structure of tamoxifen derivatives is advantageous in finding new substances with antineoplastic activity.

Full text of DOI:10.1002/cmdc.201402219

CHEMMEDCHEM
FULL PAPERS
DOI: 10.1002/cmdc.201402219
Molecular Mechanism of Action of 2-Ferrocenyl-1,1-
diphenylbut-1-ene on HL-60 Leukemia Cells
[c, d]
[c]
[e]
Alane Cabral de Oliveira,
Emanuella Gomes da Silva, Danilo Damasceno Rocha,
[a]
[f, g]
[f, g]
[e]
Elizabeth A. Hillard,* Pascal Pigeon, Gꢀrard Jaouen, Felipe A. R. Rodrigues,
[c]
[c]
[c]
Fabiane C. de Abreu, Fabrꢁcia da Rocha Ferreira, Marilia O. F. Goulart, and
[b]
Letꢁcia V. Costa-Lotufo*
The aim of this work was to investigate the mechanism of
action of 2-ferrocenyl-1,1-diphenylbut-1-ene (1) on HL-60
human leukemia cells. While inactive against noncancerous
cells, 1 provoked a concentration-dependent decrease in
viable tumor cells, primarily via apoptosis, as evidenced by
analysis of cell morphology, activation of caspases 3 and 7,
increased DNA fragmentation, and externalization of phospha-
tidylserine. Necrosis was observed only at the highest tested
concentration (4 mm). Compound 1 interfered with the cell
Interaction of 1 with dsDNA and ssDNA was observed by dif-
ferential pulse voltammetry and confirmed by hyperchromicity
in the UV/Vis spectra of dsDNA, with an interaction constant of
4
ꢀ1
2ꢂ10 m . Both the organic analogue 1,1,2-triphenylbut-1-ene
(2) and ferrocene were inactive against cancer and noncancer
cell lines and did not react with DNA. These results reinforce
the idea that the hybrid strategy of conjugating ferrocene to
the structure of tamoxifen derivatives is advantageous in find-
ing new substances with antineoplastic activity.
cycle, causing an accumulation of cells in the G /G0 phase.
1
Introduction
[
5]
Ferrocene (Fc) is receiving considerable attention as a bioiso-
stere in medicinal chemistry, as it can potentiate the biological
egy in experimental chemotherapy. A series of ferrocenyl
compounds, based on the tamoxifen skeleton, show antiproli-
ferative effects against a variety of cancer cell lines, including
[1]
activity of small molecules used in treatment against cancer
[
2]
[3]
[6]
and contagious and parasitic diseases, among others. The
hormone-dependent and -independent breast carcinoma. De-
[4]
ferricenium cation is known to be moderately cytotoxic, and
the combination of Fc with small molecules that target recep-
tors or enzymes has been shown to be an advantageous strat-
rivatives in which one or both of the geminal phenyl rings are
functionalized with an oxidizable group have shown the best
(nanomolar) activity, and reactive quinone or aminoquinone
methide intermediates have been implicated in antiprolifera-
[a] Dr. E. A. Hillard
[6f]
tive effects. Lacking such a functional group, 2-ferrocenyl-
CRPP, UPR 8641, CNRS and Universitꢀ Bordeaux
1
,1-diphenylbut-1-ene (1), is only moderately active against
3
3600 Pessac (France)
E-mail: hillard@crpp-bordeaux.cnrs.fr
[b] Prof. L. V. Costa-Lotufo
Departamento de Fisiologia e Farmacologia
Universidade Federal do Cearꢁ
Campus do Porangabussu, Fortaleza, CE, 60430-270 (Brazil)
E-mail: costalotufo@gmail.com
[c] Dr. A. C. de Oliveira, E. G. da Silva, Prof. F. C. de Abreu,
Dr. F. da Rocha Ferreira, Prof. M. O. F. Goulart
Instituto de Quꢂmica e Biotecnologia, Universidade Federal de Alagoas
Maceio, Al, 57072-970 (Brazil)
[d] Dr. A. C. de Oliveira
Faculdade de Nutriżo, Universidade Federal de Alagoas (Brazil)
[
7]
breast cancer cells and thus has not been studied further in
this context. However, we recently showed that 1 demonstrates
IC50 values below 2 mm for SF-295 human glioblastoma, HCT-8
human colon cancer, and HL-60 acute promyelocytic leukemia
[
e] Dr. D. D. Rocha, F. A. R. Rodrigues
Departamento de Fisiologia e Farmacologia
Universidade Federal do Cearꢁ (Brazil)
[
f] Dr. P. Pigeon, Prof. G. Jaouen
Chimie ParisTech (Ecole Nationale Supꢀrieure de Chimie de Paris)
[8]
cell lines after 72 h. This cytotoxicity, coupled with low hemo-
lytic activity, prompted this further study of 1.
The present work is devoted to the investigation of the
mechanism of biological action of 1 as a cytotoxic agent
against HL-60 cells. The influence of 1 on cell morphology, cell
membrane integrity, DNA fragmentation, cell cycle, phosphati-
11 rue Pierre et Marie Curie, 75231 Paris Cedex 05 (France)
[
g] Dr. P. Pigeon, Prof. G. Jaouen
Sorbonne Universitꢀs, Universitꢀ Paris 6, Institut Parisien de Chimie Molꢀcu-
laire (IPCM)–UMR 8232, 4 place Jussieu, 75252 Paris Cedex 05 (France)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201402219.
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 8
&
1
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
dylserine (PS) externalization, and activation of caspases 3 and
likewise inactive against the GBM cell line. However, 2 was also
inactive against HL-60, SF-295, and OVCAR-8 cell lines, with
IC50 values above the maximum concentration tested (25 mm)
for all cell lines, showing that the Fc group is essential for anti-
proliferative effects.
7
was studied. The ability of 1 to interact with DNA was as-
sessed through electrochemical and spectrophotometric ex-
periments using double-stranded DNA (dsDNA)-modified elec-
trodes and dsDNA and single-stranded DNA (ssDNA) in solu-
tion. To elucidate the role of the ferrocenyl group, the antipro-
liferative properties of the phenyl analogue, 1,1,2-triphenylbut-
Mechanistic studies in HL-60 cells
1-ene (2), were also evaluated.
To characterize the cytotoxic effect of 1, a phenotypic evalua-
tion of treated HL-60 cells after 24 h incubation was per-
formed. The tested concentrations were chosen as 1, 2, and
Results
4
mm, based on 24 h IC50 values. Negative control cells exhibit-
Antiproliferative effects
ed a typical non-adherent round morphology after 24 h in cul-
ture, while several 1-induced morphological changes typical of
apoptosis, including a decrease in cell size and the formation
of apoptotic bodies and intracellular vacuoles, were observed
In the present study, we compared the antiproliferative proper-
ties of 1 with its organic analogue, 2, against cancerous and
noncancerous cell lines (Table 1). Compound 1 was not cyto-
toxic against the glomerular basement membrane (GBM) non-
cancerous cell line, with an IC50 value higher than the maxi-
mum concentration tested (343 mm). With IC50 values below
(
Figure 1A). The severity of the effects increased with concen-
tration, and at 4 mm, evidence of necrotic cell death was noted.
The cytotoxic effects of compound 1 were confirmed by
flow cytometry, demonstrating a concentration-dependent de-
crease in cell density (Figure 1B). The mean number of cells
4
mm for HL-60 acute promyelocytic leukemia, HCT-8 human
[
8]
colon cancer, SF-295 human glioblastoma, and OVCAR-8
ovarian carcinoma, 1 can be considered selective for tumori-
genic cells. Compound 2, lacking the ferrocenyl moiety, was
5
ꢀ1
decreased from 4.9ꢂ10 cellsmL in the negative control to
5
ꢀ1
1
.5ꢂ10 cellsmL under exposure to 1 at 4 mm; at this con-
centration, the effect was similar to that observed in
the presence of the positive control doxorubicin at
5
ꢀ1
Table 1. Antiproliferative properties of compounds 1 and 2 after 72 h.
0.5 mm (1.4ꢂ10 cellsmL , Figure 1C). The effect on
cell density was accompanied by a decline in cell
membrane integrity (Figure 1C), although the effects
on proliferation seemed to be more marked and
likely related to more specific pathways than loss of
membrane integrity. At the highest tested concen-
tration, 1 decreased the cell number by 70%, where-
as cell membrane integrity was affected in only 25%
of treated cells.
[
a]
Compd
IC50 [mm]
HL-60
HCT-8
SF-295
MDA-MB-435 OVCAR-8
GBM
[
8]
[8]
[8]
[8]
1
2
1.04ꢁ0.04
>25
0.9ꢁ0.1
1.0ꢁ0.2
>25
16ꢁ4
3.4ꢁ0.6
>25
>343.0
>343.0
ND
ND
[
a] Data are the 95% confidence interval obtained by nonlinear regression for all cell
lines from two independent experiments, performed in duplicate, after 72 h incuba-
tion; ND: not determined.
Figure 1. Effect of 1 on A) HL-60 cell morphology, B) membrane integrity, and C) cell number, as determined by application of hematoxylin/eosin (H/E) stain-
ing and flow cytometry. Data are presented as mean values ꢁSEM of three independent experiments performed in triplicate; *p<0.05 relative to negative
control by ANOVA, followed by Dunnett’s test. HL-60 cells were treated with controls or 1 for 24 h at 1, 2, and 4 mm. DMSO (0.2%) and 0.5 mm doxorubicin
were used as the negative and positive controls, respectively. For H/E staining in panel A), solid-head arrows denote necrotic cells, dashed arrows denote
apoptotic bodies, and open-head arrows denote intracellular vacuoles. Scale bars: 20 mm.
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 8
&
2
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
Figure 2. Effect of 1 on the cell-cycle distribution of HL-60 cells treated with controls or 1 for 24 h at 1, 2, and 4 mm. Cell cycle was determined by flow cytom-
etry, and pie graphs depict the percentage of cells in G /G , S, or G /M phases; sub-G denotes fragmentation of DNA. DMSO (0.2%) and 0.5 mm doxorubicin
were used as the respective negative and positive controls.
0
1
2
1
The cell-cycle distribution of HL-60 cells treated with 1 was
analyzed by flow cytometry. Results demonstrated that cells
were arrested at G /G , followed by a progressive increase in
line) showed no redox processes, as to be expected when the
nucleoside bases are primarily paired. In the presence of
1 (black line), the appearance of diagnostic peaks correspond-
ing to guanosine and adenosine oxidation represented a con-
formational change in the dsDNA, resulting from an interaction
between 1 and dsDNA. Compound 1 could also interact with
ssDNA (Figure 4B); here, the intensities of the adenosine and
guanosine current peaks were diminished upon addition of
1 to the solution, without displacement in terms of peak po-
tential. It is worth mentioning that the Fc moiety in 1 under-
went oxidation in these experimental conditions, generating
a positive charge on the molecule. This might cause an electro-
static attraction toward the phosphate backbone of DNA and
could be a drawback in the use of electrochemical sensors for
this system. Therefore, we investigated the possible interaction
of Fc with dsDNA and ssDNA. Figure 4C shows the differential
pulse voltammetry (DPV) scans of a solution containing ssDNA,
with and without Fc. In this case, there was only a slight altera-
tion in the oxidation currents of guanosine, suggesting negligi-
ble interaction with the ferricenium cation. Furthermore, no in-
teraction of the ferricenium cation with dsDNA (Figure S1, Sup-
porting Information) was observed, suggesting that the oxida-
tion of the Fc moiety, per se, did not cause a false positive
result.
0
1
the number of cells with DNA fragmentation with increasing
concentrations of compound (Figure 2). Cell-cycle arrest was
most evident at 1 mm, with 74% of cells in G /G compared
0
1
with 52% in the negative control, while DNA fragmentation
was maximum at the highest tested concentration (4 mm), cor-
responding to 34%, relative to 4% in the negative control.
DNA fragmentation is evidence for an increased number of
dead cells, and is a signature of both necrotic and apoptotic
processes. The externalization of PS, an important biomarker
characteristic of the early stages of cell death by apoptosis,
was observed in treated cells in a concentration-dependent
manner. After 24 h of incubation, only 3% of the negative con-
trol cells presented PS externalization, while in cells treated
with 1 at 4 mm, this number increased to 38% (Figure 3A), sug-
gesting that the effects of 1 are related to apoptosis induction.
Apoptosis was also reinforced by the analysis of caspase 3 and
7
activation (Figure 3B). After 24 h of exposure to 4 mm of 1,
HL-60 cells presented an increase from 1.2 to 14% in caspase 3
and 7 activation, relative to the negative control.
DNA interaction
Compound 2 did not interact with ssDNA, behavior that was
confirmed by the lack of modification of the oxidation currents
of guanosine and adenosine in ssDNA, similar to the behavior
of Fc (Figure 4D). The result was ambiguous toward dsDNA, as
the oxidation of compound 2
Evaluation of the interaction of 1 with dsDNA-modified electro-
des and with ssDNA in solution was carried out using electro-
chemical techniques. For dsDNA (Figure 4A), the blank (grey
occurred at a similar potential
to that of adenosine in dsDNA;
however, the small difference in
currents in both voltammo-
grams suggests a slight interac-
tion at most (Figure S2, Sup-
porting Information).
Spectrophotometric studies
of the effects of compound 1 in
calf thymus dsDNA were carried
out to compare with the elec-
Figure 3. Effect of 1 on: A) phosphatidylserine externalization and B) activation of caspases 3 and 7 of HL-60 cells
treated with controls or 1 for 24 h at 1, 2, and 4 mm. DMSO (0.2%) and 0.5 mm doxorubicin were used as the re-
spective negative and positive controls. Cells were analyzed by flow cytometry, and data are presented as mean
values ꢁSEM of two independent experiments performed in triplicate; *p<0.05 relative to negative control by
ANOVA, followed by Dunnett’s test.
trochemical investigation and to
clarify the role of the pre-oxida-
tion of Fc (Figure 5). In this ex-
periment, a fixed concentration
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 8
&
3
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
Figure 4. A) Differential pulse voltammetry at a dsDNA-modified GCE in acetate buffer+EtOH (4:1, pH 4.5), in the presence and absence of 1 (20 mm). B) Dif-
ferential pulse voltammetry at a GCE in acetate buffer+EtOH (7:3, pH 4.5), and ssDNA in the presence and absence of 1 (10 mm). C) Differential pulse voltam-
metry at a GCE in acetate buffer+EtOH (7:3, pH 4.5), and ssDNA in the presence and absence of ferrocene (10 mm). D) Differential pulse voltammetry at
a GCE in acetate buffer+EtOH (7:3, pH 4.5), and ssDNA in the presence and absence of 2 (5 and 10 mm).
Figure 5. A) UV/Vis spectra for dsDNA (130 mm), in the absence and presence of 1. Inset: The effect of the concentration of 1 (3.2, 4.0, 4.8, 5.6, 6.4, 7.2, and
8
0
mm) on the spectral profile of dsDNA. B) Double reciprocal plot of 1 binding to DNA, where A is the initial absorption of free DNA and A is the absorption
at various concentrations (c) of 1 at 260 nm.
of DNA (0.130 mm) in 30% ethanol/Tris solution was treated
with different concentrations of 1 (3.2–8 mm). Figure 5A shows
the spectra of the free dsDNA and of the complex formed by
compound 1 and dsDNA. In order to display only the dsDNA–
dsDNA peak at 260 nm (Figure 5A). Hyperchromicity is typical
[9,10]
of DNA interactions
and has been attributed, for instance,
to intercalation interactions with small molecules, which results
in exposure of the base pairs to UV/Vis radiation and increas-
[9c,11,12]
1
interaction, the spectrum of 1 was subtracted from the spec-
ing absorbance.
The binding constant (K) of the 1–DNA
trum of the complex at the same concentration. An increase in
complex was obtained by applying the Benesi–Hildebrand
[9c]
the concentration of 1 provoked a hyperchromic effect on the
equation [Eq. (1)] to the band at 260 nm:
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 8
&
4
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
[
19]
l
1
1
ty of the damage, cell death pathways can be activated.
¼
þ
ð1Þ
DA D K De ½Sꢂ D De
Indeed, cells treated with 1 underwent characteristic apoptotic
cell death, as demonstrated by the morphology of the treated
cells, the increasing number of cells with DNA fragmentation,
and by PS externalization and caspase 3 activation, all pheno-
t
DS
DS
t
DS
in which l is the path length (1 cm), D is the total DNA con-
t
centration, DeDS is the molar absorptivity of the interaction
complex, [S] is the analyte concentration, and KDS is the stabili-
ty constant for complex formation (DS). Thus, the ratio of the
intercept to the slope in the double reciprocal plot in Fig-
[20]
typic and biochemical features related to apoptosis.
It is
useful to compare these results with what is already known
about the biochemical mechanism of the ferrocifens. The
effect on cell-cycle progression of the ferrocenyl analogue of
hydroxytamoxifen (Fc-OH-TAM) depends on the ER expression
[
9c]
ure 5B gave the binding constant (KDS). The binding constant
4
ꢀ1
for this interaction was found to be 2ꢂ10 m , consistent with
literature data for interaction of dsDNA with other intercala-
[21,6e]
pattern of treated cells;
in ER+ cells, Fc-OH-TAM provokes
[
9,13]
[9a,14]
arrest in the G /G phase, while ERꢀ cells are arrested in the S
tors,
including ferrocenyl compounds.
0
1
phase. The authors furthermore linked the antiproliferative ef-
fects of Fc-OH-TAM with ROS production and senescence,
which apparently is not the case for 1, as previous studies
demonstrated that it was unable to elicit ROS production in
Discussion
[8]
The antitumor properties of the ferrocifens and derivatives
HL-60 cells. This observation, as well as the observation that
have been well studied, particularly against ER+ (MCF-7) and
1 provoked G /G arrest, in spite of the low ER expression in
HL-60 cells, suggest that the mechanisms of action for 1 and
0
1
[
6]
ERꢀ (MDA-MB-231) breast cancer cells. The antiproliferative
activity of some of the most potent compounds has been at-
tributed to formation of electrophilic intermediates and re-
quires the presence of an oxidizable group conjugated to the
Fc-OH-TAM differ significantly.
[22]
Interestingly, Osella et al. showed that Fc-OH-TAM cytotox-
icity cannot be directly attributed to oxidative stress, although
it is known to provoke DNA damage. Indeed, one of the most
important aspects in the discovery of antineoplastic agents is
the interaction of these substances with DNA. Electrochemical
experiments on 1 using a dsDNA biosensor or ssDNA in solu-
tion evidenced such an interaction. The obtained binding con-
[
6f]
ferrocenyl group. In this case, 1 possesses no such functional
groups and was indeed less active on MDA-MB-231 cells (IC =
50
[
15]
7
.54 mm)
than those compounds possessing phenol func-
[7]
tionalities (e.g., IC =0.6 mm for the 4,4’-dihydroxy analogue).
50
Nonetheless, we felt that 1 merited further study, based on
a recent discovery of its antiproliferative activity against SF-
4
ꢀ1
stant of 2ꢂ10 m , as well as the hyperchromic effect upon
addition of 1 to dsDNA, is consistent with weak intercalation
295, HCT-8, and HL-60 tumor cells in the low micromolar
[9a,11,12,23]
range, the lack of unspecific toxicity on erythrocytes, and its
between base pairs.
Studies to evaluate the interaction
[
8]
lack of oxidative stress induction properties.
of Fc or the organic moiety 2 with dsDNA and ssDNA gave
Compound 1 was selective toward tumor cells when com-
negative results, reinforcing the importance of the hybrid skel-
[24]
pared with noncancerous cells; the mean IC5 value (4.5 mm)
eton on the activity of 1.
0
[
8]
against HL-60, HCT-8, SF-295, MDA-MB-435, and OVCAR-8
tumor cells compared favorably with that obtained using the
noncancerous GBM monkey cell line (IC >343.00 mm). Such
50
Conclusions
selectivity toward tumor cells is a key factor in the develop-
ment of new anticancer agents, with the goal of a decrease in
side effects and an improvement in the quality of life of pa-
tients under chemotherapy. Often, such selectivity is sugges-
tive of specific interactions with receptors present in cancer
cells, such as the estrogen receptor (ER), which is targeted by
hydroxytamoxifen. Although there is evidence of the presence
of functional membrane ERa and ERb in HL-60 leukemia
2-Ferrocenyl-1,1-diphenylbut-1-ene (1) was not cytotoxic
toward GBM monkey cells but was able to decrease, in a con-
centration-dependent manner, the number of viable cells in
the HL-60 line. Cell death was caused by apoptosis, as evi-
denced by morphological analyses, activation of effector cas-
pases 3 and 7, increased DNA fragmentation, externalization of
PS, and a necrosis process at the highest concentration. Fur-
thermore, 1 interfered with the cell cycle, leading to accumula-
[
16]
cells, as well as functional ERb for colon cancer cells, includ-
[
17]
[18]
ing HCT-8, analysis of the molecular data from the NCI
tion of cells in the G /G phase. Interaction with dsDNA and
0 1
showed that all cell lines used in the present work indeed ex-
press low levels of ERs relative to some breast (MCF-7 and
T47D) and ovarian (SK-OV-3) overexpressing cells. Based on the
present findings, the activity of 1 seems to be independent of
an ER expression pattern.
ssDNA suggests that the mechanism of action involves molec-
ular damage to DNA via intercalation.
Experimental Section
Studies on the mechanism of action of 1 suggest that it in-
duces cell-cycle arrest at the G /G phase, followed by an in-
Preparation of 2-ferrocenyl-1,1-diphenylbut-1-ene (1) and
0
1
1
,1,2-triphenylbut-1-ene (2)
creasing number of cells with DNA fragmentation. Usually, in
the presence of DNA damage, cells can trigger a complex se-
quence of events, including checkpoint activation, leading to
cell-cycle arrest, allowing DNA repair. Depending on the severi-
[7]
The synthesis of 1 has been previously reported. Briefly, 1 was
obtained by a Friedel–Crafts acylation of Fc, followed by a McMurry
cross-coupling of propionyl Fc with benzophenone.
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 8
&
5
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
[25]
The preparation and characterization of 2 has been reported.
DNA interaction studies
However, we here used a one-step McMurry method, starting from
commercially available propiophenone and benzophenone. Titani-
um chloride (7.6 g, 4.4 mL, 40 mmol) was added dropwise to a sus-
pension of zinc powder (3.9 g, 60 mmol) in dry THF at 108C. The
mixture was heated at reflux for 2 h. A second solution was pre-
pared by dissolving propiophenone (1.34 g, 1.33 mL, 10 mmol) and
benzophenone (1.82 g, 10 mmol) in dry THF. This latter solution
was added dropwise to the first solution, and the solution was
stirred at reflux for an additional 2 h. After cooling to room tem-
perature, the mixture was stirred with water and dichloromethane,
acidified with dilute hydrochloric acid until the dark color disap-
peared, and decanted. The aqueous layer was extracted with di-
chloromethane, and the organic layers were combined and dried
over magnesium sulfate. After concentration under reduced pres-
sure, the crude product was separated by preparative HPLC with
acetonitrile as the eluent to yield pure 2 in 55% yield. Physico-
Electrochemistry with dsDNA biosensors: The ability of 1 and 2 to in-
teract dsDNA was investigated using an electrochemical approach
employing DNA-modified glassy carbon electrode (GCE) biosen-
[28,29]
sors.
The electrochemical experiments were performed with
a conventional undivided three electrode cell using an Autolab
PGSTAT-30 potentiostat (Echo Chemie, Utrecht, the Netherlands)
coupled to a microcomputer, interfaced by GPES 4.9 software. A
working GCE (diameter=3 mm), a Pt wire counter electrode, and
ꢀ
a AgjAgCl, Cl (saturated) reference electrode were used. All ex-
periments were conducted at room temperature (25ꢁ28C) after
purging with argon. The DNA used was type I calf thymus (Sigma
Aldrich, Saint Louis, USA), highly polymerized, containing 6.2% Na
and 13% H O, dried, and stored at 88C. The electrochemical proce-
2
dure for the investigation of dsDNA interaction with compounds
1 and 2 involved three steps: preparation of the GCE surface, im-
mobilization of dsDNA gel, and voltammetric transduction. Initially,
the GCE was polished with alumina (BAS polishing kit). The elec-
trode was then electrochemically pre-treated with a sequence of
[25]
chemical data were identical to those in the literature.
Compounds 1 and 2 were purified by semi-preparative HPLC
13
before use. Elemental analysis of 1 and C NMR of 2 confirmed
95% purity.
5
cyclic potential scans from 0 to +1.4 V versus AgjAgCl, KCl
>
(0.1m) in acetate buffer, washed thoroughly with distilled/deion-
ized water, and dried. In order to immobilize the dsDNA, the sur-
face of the electrode was coated with 10 mL of calf thymus DNA
solution (containing 12.0 mg of dsDNA in 1.0 mL of acetate buffer,
pH 4.5, stored at ꢀ208C for 48 h), the gel was allowed to dry at
room temperature under a stream of nitrogen, and the biosensor
was subsequently immersed in 5 mL of aqueous-ethanolic (30%)
acetate buffer, pH 4.5. For each series of experiments, an identical
dsDNA-GCE was prepared as a reference blank. Before analysis, the
modified electrode was left for 10 min in contact with the buffer
solution containing either 20 mL of ethanol (negative control), Fc
Biochemistry
Cell lines: HL-60 (acute promyelocytic leukemia), OVCAR-8 (ovarian
carcinoma), and SF-295 (glioblastoma) cell lines were provided by
the U.S. National Cancer Institute (Bethesda, MD). Glomerular base-
ment membrane (GBM) monkey cells were used as a noncancerous
cell line for comparison. Cells were maintained in RPMI 1640
medium supplemented with 10% fetal bovine serum and 1% anti-
ꢀ1
(10 mm), 1 (20 mm), or 2 (10 mm). DPV scans (n=10 mVs , pulse
amplitude=50 mV and pulse width=70 ms) and baseline correc-
biotics at 378C in a humidified atmosphere of 5% CO and 95%
air.
2
[28,29]
tions were performed using GPES 4.9.
Cytotoxicity: Cytotoxicity was determined using the MTT (3-(4,5-di-
Electrochemistry of ssDNA in solution: To produce ssDNA by acid–
base treatment, dsDNA (1 mg) was dissolved in 1m HCl (100 mL) by
heating at 1008C in a sealed glass tube immersed in a boiling
[26]
methylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay.
Ac-
cordingly, GBM, HL-60, OVCAR-8, and SF-295 cell lines were seeded
in 96-well plates and incubated with 1 (0.02–343 mm) or vehicle
control (DMSO, 0.5%) for 72 h. After the incubation period, a solu-
tion of MTT (1 mm) was added to each well and incubated for 3 h.
Plates were spectrophotometrically evaluated at 570 nm, and IC50
values were determined using the GraphPad Prism 5.0 software.
[29,30]
water bath for 1 h, followed by neutralization with 1m NaOH.
A freshly prepared solution, consisting of 4 mL ssDNA and 950 mL
[28–30]
of ethanol,
was added to the electrochemical cell. Single-scan
DPV experiments were conducted between 0 and +1.4 V versus
ꢀ
1
AgjAgCl, KCl (0.1m), (n=10 mVs , pulse amplitude=50 mV and
pulse width=70 ms). Peaks corresponding to the oxidation of gua-
nine and adenine appeared at +0.815 V and +1.164 V, respective-
ly. After rinsing the surface, the GCE was inserted into a solution
containing 1 or 2 (50 mL, corresponding to a concentration of
Mechanistic studies: For the following experiments, HL-60 cells
5
ꢀ1
were seeded at a density of 3ꢂ10 cellsmL and incubated with
, 2, or 4 mm of 1 for 24 h. The vehicle (DMSO, 0.5%) and doxorubi-
1
cin (0.5 mm) were used as negative and positive controls, respec-
tively. Statistical analyses of results was performed using the
GraphPad Prism 5.0 software.
1
0 mm) or Fc (10 mm), and the DPV experiment was repeated. A
clean GCE was also employed in DPV experiments involving
a 0.1 mm buffered solution of 1 alone (pH 4.5), in order to observe
possible interference of oxidation waves from compounds 1 or 2.
Baseline corrections were made using the software GPES 4.9.
Analysis of morphological changes—(H/E stain): Untreated or 1-
treated cells were examined for morphological changes by light
microscopy (Olympus, Tokyo, Japan). To evaluate nuclear morphol-
ogy, cells were harvested, transferred to cytospin slides, fixed with
methanol for 1 min, and stained with hematoxylin and eosin.
Spectrophotometric evaluation of interaction with dsDNA: Spectro-
photometric studies were carried out with a Shimadzu Multi-
spec 1501 or diode array Hewlett Packard 8453 spectrophotometer.
The interaction of compound 1 with dsDNA was monitored by the
Flow cytometry analysis: All experimental procedures adopted in
[9,10]
flow cytometry analyses essentially followed the methodology de-
absorption titration method,
in which a fixed concentration of
[27]
scribed by Montenegro et al. The evaluated parameters were cell
membrane integrity, DNA fragmentation, cell cycle, PS externaliza-
tion and caspase 3/7 activation. For all experiments, 5000 events
were evaluated using a Guava EasyCyte Mine flow cytometer and
Guava Express Plus software. DNA fragmentation and the cell cycle
were analyzed by ModFit LT for Win32 version 3.1.
DNA (130 mm) was treated with different concentrations of com-
pound 1 (3.2-8 mm) in Tris-HCl buffer (pH 7.4, 0.1m, 30% ethanol).
A stock solution of DNA was prepared by dissolving 12 mg of calf
thymus DNA per mL of acetate buffer. An aliquot (10 mL) was then
dissolved in Tris-HCl buffer and kept at 88C for 24 h, stirring at fre-
quent intervals to ensure the homogeneity of the solution. The
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 8
&
6
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
final concentration of the dsDNA solution was measured spectro-
photometrically at 260 nm, using the molar extinction coefficient
[8] A. C. de Oliveira, E. A. Hillard, P. Pigeon, D. D. Rocha, F. A. R. Rodrigues,
R. C. Montenegro, L. V. Costa-Lotufo, M. O. F. Goulart, G. Jaouen, Eur. J.
Med. Chem. 2011, 46, 3778–3787.
ꢀ
1
ꢀ1
[9a,d]
of e260 =6600 cm
m
per nucleotide for calf thymus DNA.
The
[
9] a) A. Shah, R. Qureshi, A. M. Khan, R. A. Khera, F. L. Ansari, J. Braz. Chem.
Soc. 2010, 21, 447–451; b) N. K. Janjua, A. Shaheen, A. Yaqub, F.
Perveen, S. Sabahat, M. Mumtaza, C. Jacob, L. A. Ba, H. A. Mohammed,
Spectrochim. Acta Part A 2011, 79, 1600–1604; c) S. Agarwal, D. K.
Jangir, R. Mehrotra, J. Photochem. Photobiol. B 2013, 120, 177–182;
d) H. R. Mahler, B. Kline, B. P. Mehrotra, J. Mol. Biol. 1964, 9, 801–811.
concentration of the DNA solution was adjusted to 130 mm. The
ratio between the absorbance values at 260 and 280 nm in the
DNA spectra (A260/A280) was measured as 1.81, indicative of DNA
[9a,31]
free of protein contamination.
The same procedure was per-
formed for compound 2.
[
10] R. Vijayalakshmi, M. Kanthimathi, V. Subramanian, B. U. Nair, Biochem.
Biophys. Res. Commun. 2000, 271, 731–734.
[
[
11] R. Hajian, T. G. Huat, J. Spectrosc. 2013, DOI: 10.1155/2013/380352.
12] A. H. Hegde, S. N. Prashanth, J. Seetharamappa, J. Pharm. Biomed. Anal.
Acknowledgements
2
012, 63, 40–46.
This research was supported by grants from the Conselho Nacio-
nal de Desenvolvimento Cientꢂfico e Tecnolꢃgico (CNPq), CAPES,
CAPES/PNPD, RENORBIO, UFAL, and the National Cancer Institute
of Brazil. The authors are also indebted to PRONEX-CNPq-FAPEAL
and the CNRS.
[
[
[
13] A. Rescifina, C. Zagni, M. G. Varrica, V. Pistarꢆ, A. Corsaro, Eur. J. Med.
Chem. 2014, 74, 95–115.
14] F. Asghar, A. Badshah, A. Shah, M. K. Rauf, M. I. Ali, M. N. Tahirb, E. Nos-
heen, Zia-ur-Rehman, R. Qureshi, J. Organomet. Chem. 2012, 717, 1–8.
15] M. Gormen, P. Pigeon, S. Top, E. A. Hillard, M. Huchꢀ, C. G. Hartinger, F.
de Montigny, M.-A. Plamont, A. Vessiꢄres, G. Jaouen, ChemMedChem
2
010, 5, 2039–2050.
16] M. A. Kauss, G. Reiterer, R. P. Bunaiciu, A. Yen, Exp. Cell Res. 2008, 314,
999–3006.
[
Keywords: antitumor
agents
·
apoptosis
·
2
bioelectrochemistry · ferrocenes · HL-60 cells
[17] G. Fiorelli, L. Picariello, V. Martineti, F. Tonelli, M. L. Brandi, Biochem. Bio-
phys. Res. Commun. 1999, 261, 521–527.
[
18] Molecular Target Data, NCI Wiki, U.S. National Cancer Institute, https://
wiki.nci.nih.gov/display/NCIDTPdata/Molecular+Target+Data (accessed
July 2014).
[
1] a) “Ferrocene-functionalized endocrine modulators as anticancer
agents”: E. A. Hillard, A. Vessiꢄres, G. Jaouen in TopOrgan, Vol. 32,
Springer, Berlin/Heidelberg, 2010, pp. 81–117; b) G. Gasser, I. Ott, N.
Metzler-Nolte, J. Med. Chem. 2011, 54, 3–25; c) S. S. Braga, A. M. S. Silva,
Organometallics 2013, 32, 5626–5639; d) D. Plazuk, J. Zakrzewski, M.
Salmain, A. Blauz, B. Rychlik, P. Strzelczyk, A. Bujacz, G. Bujacz, Organo-
metallics 2013, 32, 5774–5783.
[
19] a) H. Okada, T. W. Mak, Nat. Rev. Cancer 2004, 4, 592–603; b) H. Rodri-
guez-Rocha, A. Garcia-Garcia, M. I. Panayiotidis, R. Franco, Mutat. Res.
2
011, 711, 158–166.
[
[
[
20] a) K. Nishida, O. Yamaguchi, K. Otsu, Circ. Res. 2008, 103, 343–351;
b) S. J. Martin, D. R. Green, Cell 1995, 82, 349–352.
[
[
2] a) F. Dubar, C. Slomianny, J. Khalife, D. Dive, H. Kalamou, Y. Guꢀrardel, P.
Grellier, C. Biot, Angew. Chem. Int. Ed. 2013, 52, 7690–7693; Angew.
Chem. 2013, 125, 7844–7847; b) A. Mahajan, L. Kremer, S. Louw, Y. Guꢀr-
ardel, K. Chibale, C. Biot, Bioorg. Med. Chem. Lett. 2011, 21, 2866–2868;
c) K. Kumar, S. Carrere-Kremer, L. Kremer, Y. Guerardel, C. Biot, V. Kumar,
Organometallics 2013, 32, 5713–5719.
3] a) K. Kowalski, J. Skiba, L. Oehninger, I. Ott, J. Solecka, A. Rajnisz, B. Ther-
rien, Organometallics 2013, 32, 5766–5773; b) B. Maity, M. Roy, S. Saha,
A. R. Chakravarty, Organometallics 2009, 28, 1495–1505; c) D. R. van Sta-
veren, N. Metzler-Nolte, Chem. Rev. 2004, 104, 5931–5986.
21] A. Vessiꢄres, C. Corbet, J. M. Heldt, N. Lories, N. Jouy, I. Laios, G. Le-
clercq, G. Jaouen, R. A. Toillon, J. Inorg. Biochem. 2010, 104, 503–511.
22] D. Osella, H. Mahboobi, D. Colangelo, G. Cavigiolio, A. Vessiꢄres, G.
Jaouen, Inorg. Chim. Acta 2005, 358, 1993–1998.
[
[
23] U. Pindur, M. Jansen, T. Lemster, Curr. Med. Chem. 2005, 12, 2805–2847.
24] E. H. Kerns, L. Di, Drug-like Properties: Concepts, Structure, Design and
Methods, Academic Press, London, 2008.
[
25] M. R. Schneider, E. von Angerer, H. Schonenberger, R. T. Michel, H. P. For-
tmeyer, J. Med. Chem. 1982, 25, 1070–1077.
[
26] T. Mosmann, J. Immunol. Methods 1983, 65, 55–63.
[
[
4] a) P. Kçpf-Maier, H. Kçpf, E. W. Neuse, Angew. Chem. Int. Ed. Engl. 1984,
[
27] R. C. Montenegro, A. J. Araffljo, M. T. Molina, J. D. B. Marinho-Filho, D. D.
Rocha, E. Lopez-Montero, M. O. F. Goulart, E. S. Bento, A. P. N. N. Alves,
C. Pessoa, M. O. Moraes, L. V. Costa-Lotufo, Chem. Biol. Interact. 2010,
23, 456–457; Angew. Chem. 1984, 96, 446–447; b) P. Kçpf-Maier, H.
Kçpf, E. W. Neuse, J. Cancer Res. Clin. Oncol. 1984, 108, 336–340.
5] a) J. Spencer, A. P. Mendham, A. K. Kotha, S. C. W. Richardson, E. A. Hill-
ard, G. Jaouen, L. Male, M. B. Hursthouse, Dalton Trans. 2009, 918–921;
b) S. Top, J. Tang, A. Vessiꢄres, D. Carrez, C. Provot, G. Jaouen, Chem.
Commun. 1996, 8, 955–956; c) J. J. Cꢅzares-Marinero, S. Top, A. Ves-
siꢄres, G. Jaouen, Dalton Trans. 2014, 43, 817–830.
1
84, 439–448.
[
28] A. J. Araffljo, A. A. de Souza, E. N. da Silva Jfflnior, J. D. B. Marinho-Filho,
M. A. B. F. de Moura, D. D. Rocha, M. C. Vasconcellos, C. O. Costa, C.
Pessoa, M. O. de Moraes, V. F. Ferreira, F. C. Abreu, A. V. Pinto, R. C. Mon-
tenegro, L. V. Costa-Lotufo, M. O. F. Goulart, Toxicol. In Vitro 2012, 26,
[
6] a) S. Top, A. Vessiꢄres, G. Leclercq, J. Quivy, J. Tang, J. Vaissermann, M.
Huchꢀ, G. Jaouen, Chem. Eur. J. 2003, 9, 5223–5236; b) E. A. Hillard, A.
Vessiꢄres, S. Top, P. Pigeon, K. Kowalski, M. Huchꢀ, G. Jaouen, J. Organo-
met. Chem. 2007, 692, 1315–1326; c) A. Vessiꢄres, S. Top, P. Pigeon, E. A.
Hillard, L. Boubeker, D. Spera, G. Jaouen, J. Med. Chem. 2005, 48, 3937–
585–594.
[
[
29] F. C. de Abreu, F. S. de Paula, D. C. M. Ferreira, V. B. Nascimento, J. C. D.
Lopes, A. M. C. Santos, M. M. Santoro, C. E. Salas, M. O. F. Goulart, Sen-
sors 2008, 8, 1519–1538.
30] C. O. L. Costa, E. O. C. Costa, F. R. Ferreira, L. S. Viana, F. A. S. Silva, F. C.
Abreu, I. M. Figueiredo, Jr., W. Pinho, E. S. Bento, T. L. G. Lemos, O. D. L.
Pessoa, B. Aguilera-Venegas, E. Norambuena, C. Olea-Azar, M. O. F. Gou-
lart, J. Braz. Chem. Soc. 2012, 23, 1174–1185.
3940; d) P. Pigeon, S. Top, O. Zekri, E. A. Hillard, A. Vessiꢄres, M. A. Pla-
mont, O. Buriez, E. Labbꢀ, M. Huchꢀ, S. Boutamine, C. Amatore, G.
Jaouen, J. Organomet. Chem. 2009, 694, 895–901; e) A. L. Lainꢀ, E.
Adriaenssens, A. Vessiꢄres, G. Jaouen, C. Corbet, E. Desruelles, P. Pigeon,
R. A. Toillon, C. Passirani, Biomaterials 2013, 34, 6949–6956; f) E. Hillard,
A. Vessiꢄres, L. Thouin, G. Jaouen, C. Amatore, Angew. Chem. Int. Ed.
[
31] N. K. Janjuaa, A. Siddiqa, A. Yaqub, S. Sabahat, R. Qureshi, S. Ul Haque,
Spectrochim. Acta Part A 2009, 74, 1135–1137.
2
006, 45, 285–290; Angew. Chem. 2006, 118, 291–296.
7] E. A. Hillard, P. Pigeon, A. Vessiꢄres, C. Amatore, G. Jaouen, Dalton Trans.
007, 5073–5081.
[
Received: May 27, 2014
Published online on && &&, 0000
2
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 8
&
7
&
ÞÞ
These are not the final page numbers!

FULL PAPERS
A. C. de Oliveira, E. G. da Silva,
D. D. Rocha, E. A. Hillard,* P. Pigeon,
G. Jaouen, F. A. R. Rodrigues,
Who is the ferrous one of all? 2-Ferro-
cenyl-1,1-diphenylbut-1-ene, a ferro-
cene–tamoxifen derivative, provokes
DNA fragmentation and apoptosis in
HL-60 cells, but is inactive against non-
cancerous cells. These results suggest
that this ferrocene hybridization strat-
egy may have utility in the future
F. C. de Abreu, F. da Rocha Ferreira,
M. O. F. Goulart, L. V. Costa-Lotufo*
&& – &&
Molecular Mechanism of Action of 2-
Ferrocenyl-1,1-diphenylbut-1-ene on
HL-60 Leukemia Cells
design of antineoplastic therapeutics.
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 8
&
8
&
ÞÞ
These are not the final page numbers!