A step towards development of promising trypanocidal agents: Synthesis, characterization and in vitro biological evaluation of ferrocenyl Mannich base-type derivatives

Article abstract of DOI:10.1016/j.ejmech.2018.12.005

Chagas disease is a neglected chronical parasitosis caused by the parasite Trypanosoma cruzi (T. cruzi). Nine ferrocenyl Mannich base derivatives were synthetized and characterized to explore their in vitro activity on three T. cruzi strains of the parasite and their cytotoxicity on Vero cells to calculate the selectivity index (SI). Compound 2, 1-ferrocenyl-3-(4-(4-(trifluoromethyl)phenyl)piperazin-1-yl)propan-1-one, stood out as the most promising derivative showing a half maximal inhibitory concentration (IC₅₀) value around 5 μM in both amastigote and trypomastigote forms of T. cruzi and SI values higher than 13, being the best value on the trypomastigote forms of the Arequipa strain (SI = 41.7). Moreover, 2 decreased the number of infected cells and was not genotoxic. Furthermore, its possible mechanism of action was studied through the alteration of the metabolites excreted by the parasite during glucose metabolism, the detection of mitochondrial alterations and the inhibition of superoxide dismutase (SOD). Finally, docking studies were executed to analyze the binding mode of the studied compounds to Fe-SOD enzyme.

Full text of DOI:10.1016/j.ejmech.2018.12.005

Accepted Manuscript
A step towards development of promising trypanocidal agents: Synthesis,
characterization and in vitro biological evaluation of ferrocenyl Mannich base-type
derivatives
Rocío Paucar, Rubén Martín-Escolano, Elsa Moreno-Viguri, Nuria Cirauqui, Carlos
Rangel Rodrigues, Clotilde Marín, Manuel Sánchez-Moreno, Silvia Pérez-Silanes,
Mauro Ravera, Elisabetta Gabano
PII:
S0223-5234(18)31042-0
DOI:
https://doi.org/10.1016/j.ejmech.2018.12.005
EJMECH 10938
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 9 November 2018
Revised Date: 29 November 2018
Accepted Date: 2 December 2018
Please cite this article as: Rocí. Paucar, Rubé. Martín-Escolano, E. Moreno-Viguri, N. Cirauqui, C.R.
Rodrigues, C. Marín, M. Sánchez-Moreno, S. Pérez-Silanes, M. Ravera, E. Gabano, A step towards
development of promising trypanocidal agents: Synthesis, characterization and in vitro biological
evaluation of ferrocenyl Mannich base-type derivatives, European Journal of Medicinal Chemistry
(2019), doi: https://doi.org/10.1016/j.ejmech.2018.12.005.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
A Step Towards Development of Promising Trypanocidal Agents:
Synthesis, Characterization and in Vitro Biological Evaluation of
Ferrocenyl Mannich Base-type Derivatives.
a
b
a
c
Rocío Paucar , Rubén Martín-Escolano , Elsa Moreno-Viguri , Nuria Cirauqui , Carlos
c
b
b
a
Rangel Rodrigues , Clotilde Marín , Manuel Sánchez-Moreno , Silvia Pérez-Silanes ,
d
d
Mauro Ravera , Elisabetta Gabano *.
a
Universidad de Navarra, Department of Technology and Chemistry Pharmaceutics,
Instituto de Salud Tropical, Pamplona 31008, Spain.
b
Department of Parasitology, Instituto de Investigación Biosanitaria (ibs.GRANADA),
Hospitales Universitarios De Granada/University of Granada, Granada 18071, Spain
c
Department of Pharmaceutical Sciences, Federal University of Rio de Janeiro, Rio de
Janeiro 21949-900, Brazil
d
Dipartimento di Scienze e Innovazione Tecnologica, Università del Piemonte Orientale,
Viale T. Michel 11, 15121, Alessandria, Italy
*
Corresponding author
Dr. Elisabetta Gabano, E-mail: elisabetta.gabano@uniupo.it. Dipartimento di Scienze e
Innovazione Tecnologica, Università del Piemonte Orientale, Viale T. Michel 11, 15121,
Alessandria, Italy
1

ACCEPTED MANUSCRIPT
Abstract
Chagas disease is a neglected chronical parasitosis caused by the parasite Trypanosoma
cruzi (T. cruzi). Nine ferrocenyl Mannich base derivatives were synthetized and
characterized to explore their in vitro activity on three T. cruzi strains of the parasite and
their cytotoxicity on Vero cells to calculate the selectivity index (SI). Compound 2, 1-
ferrocenyl-3-(4-(4-(trifluoromethyl)phenyl)piperazin-1-yl)propan-1-one, stood out as the
most promising derivative showing a half maximal inhibitory concentration (IC ) value
5
0
around 5 µM in both amastigote and trypomastigote forms of T. cruzi and SI values
higher than 13, being the best value on the trypomastigote forms of the Arequipa strain
(SI = 41.7). Moreover, 2 decreased the number of infected cells and was not genotoxic.
Furthermore, its possible mechanism of action was studied through the alteration of the
metabolites excreted by the parasite during glucose metabolism, the detection of
mitochondrial alterations and the inhibition of superoxide dismutase (SOD). Finally,
docking studies were executed to analyze the binding mode of the studied compounds to
Fe-SOD enzyme.
Keywords: Ferrocene; Mannich bases; Chagas disease, Trypanosoma cruzi, In vitro
studies, Superoxide dismutase.
1. Introduction
Chagas disease (CD) is a neglected chronical parasitosis caused by the protozoan parasite
Trypanosoma cruzi (T. cruzi) [1-3]. According to the World Health Organization (WHO),
approximately 7 million people are infected worldwide, mostly in endemic areas of Latin
American countries [4-5]. Several initiatives and different research partnerships have
2

ACCEPTED MANUSCRIPT
emerged to control and eradicate this disease; however, there are still some gaps to cover
for the proper management of CD [6-9].
In general, the available chemotherapy is limited to two nitroaromatic drugs,
benznidazole (BZN) and nifurtimox (NFX), where BZN is considered as a first line
treatment (Figure 1). Both drugs show variable efficacy, because they reduce the
parasitic burden in the acute phase and congenital infection. However, their effectiveness
decreases in the chronic phase and in the reactivation of the disease. In addition, their
efficacy varies according to the geographical area, due to differences in drug
susceptibilities against different T. cruzi strains. Moreover, severe toxic side effects and
prolonged treatment may be serious complications, also leading to the suspension of the
therapy [10-12]. Therefore, the research and development of effective, safe and
affordable antiparasitic drugs is one of the main challenges to be faced with to combat
CD.
Figure 1. Chemical structure of Benznidazole (BZN) and Nifurtimox (NFX).
Although metal compounds have been used mainly as anticancer agents, thanks to the
success obtained by cisplatin and its analogues, they can also be exploited in the search of
new drugs to treat infectious diseases. In this framework, metal-containing compounds
3

ACCEPTED MANUSCRIPT
offer several advantages over purely organic molecules, such as multiple mechanisms of
action due to a combination of the pharmacological properties of the metal and the ligand,
a rich redox chemistry that can lead to the production of cytotoxic radical species
resulting in oxidative stress, etc. [13-25].
5
5
In particular, bis(η -cyclopentadienyl)iron ([Fe(η -C H ) ], ferrocene) derivatives have
5
5 2
been largely employed in the design and synthesis of a wide variety of compounds with
enhanced biological activity, due to their versatility and easy functionalization [24, 26-
27]. The incorporation of this fragment into an already known drug has proven to be a
successful strategy in the development of novel chemotherapeutic agents. For example,
the replacement of a phenyl by a ferrocenyl group in tamoxifen, clinically used in the
treatment of estrogen receptor-positive breast cancer, improves the cytotoxic activity of
the organic drug [28]. Similarly, the introduction of the ferrocene group into the anti-
malarial drug chloroquine to yield ferroquine has enhanced its activity against
Plasmodium spp, highlighting its effectiveness against chloroquine-resistant strains [29-
36].
Mannich bases are considered as promising alternatives in the search for potential
antiparasitic agents [37-42]. Our research group has identified Mannich base derivatives,
non-mutagenic and non-genotoxic, that exhibit potential in vitro trypanocidal activity
against different strains and forms of T. cruzi and in a murine in vivo model [43-44]. It is
remarkable to mention that according to our previous studies and literature, the presence
of piperazine is important for the antichagasic activity of these derivatives [43-46]. Also
phenetylamine derivatives [47] and Mannich bases containing anilines [48] have shown
anti-parasitic and anti-infective activity, respectively.
Starting from these premises, the incorporation of a ferrocenyl moiety into Mannich base
derivatives could result in new potential bioorganometallic agents for the treatment of
4

ACCEPTED MANUSCRIPT
CD. Hence, with the aim of exploring the in vitro activity in T. cruzi and cytotoxicity on
Vero cells, nine ferrocenyl Mannich base derivatives containing anilines,
phenethylamines and piperazines were synthesized via straightforward and cost-efficient
methods. Moreover, the possible mechanism of action was studied for the most promising
compound, by analyzing the electrochemical behavior, the alteration of the metabolites
excreted by the parasite during the glycolytic pathway, the detection of mitochondrial
alterations and the inhibition of iron superoxide dismutase (Fe-SOD) enzyme, both in
vitro and in silico.
2. Results and discussion
2.1. Chemistry
The general route for the synthesis of the ferrocenyl Mannich base derivatives (1-9) is
shown in Scheme 1. The ferrocene intermediates were prepared according to literature
procedures, by reaction of ferrocene with acetic anhydride (Friedel-Craft acylation,
acetylferrocene) [49-50] or with 3-chloropropanoyl chloride (i1), or from 3-
chloropropanoylferrocene (i2) [48]. Compounds 1-4 and 7-9 were synthesized by
nucleophilic substitution as previously reported by Martin Escolano et al. [43], in which,
3-chloropropanoylferrocene was condensed with the corresponding arylamine employing
THF anhydrous as solvent and K CO or NaH as base. Compounds 5 and 6 were
2
3
synthesized by the Michael addition reaction described by Damljanovic et al.[48], in
which the acryloylferrocene and the respective arylamine were irradiated with
microwaves in the presence of montmorillonite K10 as a catalyst. The structure of the
5

ACCEPTED MANUSCRIPT
nine synthesized compounds, together with their predicted drug-likeness properties, are
shown in Table 1. The values, calculated with the DataWarrior program [51], anticipate
-1
good oral bioavailability, with a molar mass lower than 500 g mol , an estimated
partition coefficient (clogP) lower than 5, aqueous solubility (clogS) values higher than -
2
4
, a Polar Surface Area (PSA) lower than 140 Å , a number of hydrogen-bond acceptors
(HBA) lower than 10, and less than 5 hydrogen-bond donors (HBD). These values are in
tune with those required by the Lipinski's Rule of Five, a mnemonic tool for medicinal
chemists to be used as a first approximation in the evaluation of the druglikeness or in the
determination of the potential oral activity of a molecule.
Table 1. Structure and molecular properties of the ferrocenyl Mannich Base-type derivatives 1-8
Molecular properties
Compound
a
b
c
d
e
f
MW
clogP clogS PSA HBA HBD
1
2
3
420.327 0.5064 -2.575 21.51
470.334 1.2539 -3.039 21.51
436.782 1.0116 -2.997 21.51
2
2
2
1
1
1
4
351.201 1.1809 -2.255 29.1
2
1
6

ACCEPTED MANUSCRIPT
5
6
401.208 1.9284 -2.719 29.1
2
1
367.656 1.6861 -2.677 29.1
380.263 0.0957 -2.183 33.68
2
2
1
1
7
8
430.270 0.8432 -2.647 33.68
396.718 0.6009 -2.605 33.68
2
2
1
1
9
a
-1
b
Total Molecular Weight (expressed in g mol like molar mass), logarithm of the partition
c
-1 d
coefficient between n-octanol and water, logarithm of the solubility measured in mol L , Polar
2
e
f
Surface Area (Å ), number of hydrogen-bond acceptors, and number of hydrogen-bond donors
were calculated by using the DataWarrior program [51].
7

ACCEPTED MANUSCRIPT
Scheme 1. General procedure of synthesis. Reagents and conditions: (i) AlCl , CH Cl , rt, 2h; (ii)
3
2
2
CH COOK, ethanol, reflux, 2.5 h; (iii) Montmorillonite K10, MW, 50 W, 1.5 min. (iv) THF,
3
K CO or NaH, 48 h, rt.
2
3
2.2. Electrochemistry
The electrochemical behavior of the 1-9 compounds was investigated by cyclic
voltammetry (CV) in CH Cl solution containing 0.1 M [NBu ][PF ] as a supporting
2
2
4
6
electrolyte and a glassy carbon (GC) as a working electrode. Figure shows the
cyclovoltammograms of acetylferrocene, 1, 4, and 7 as an example.
All compounds showed an anodic process (peak A) with a directly associated reduction
response in the reverse scan (peak B). The A/B redox couple is consistent with the well-
known electrochemically (Nernstian) and chemically reversible 1e transfer of ferrocene
5
+
producing the ferrocenium cation [Fe(η -C H ) ] :
5
5 2
8

ACCEPTED MANUSCRIPT
5
5
+
-
[
Fe(η -C H ) ]
[Fe(η -C H ) ] + 1e
5 5 2
5
5 2
As expected, the presence of the electron withdrawing carbonyl group shifts toward more
negative values the formal potential E°’ of the A/B couple of all complexes with respect
to ferrocene (∆E°’ = +250 mV) (Table 2). The comparison with the redox behavior of an
authentic sample of acetylferrocene confirms this hypothesis (Figure ). Analysis of the
-1
CV responses with scan rates (v) varying from 0.05 to 3.20 V s indicates that: i) the
-½
cathodic-to-anodic peak current ratio (i /i ) is always 1; ii) the current function i v
p,c p,a
p,a
remains substantially constant; iii) the peak-to-peak separation (∆E ) progressively
p
increases from 74 to 95 mV. The trend of ∆E with v, which give information on the
p
extent of structural reorganization accompanying the redox processes, deviates
appreciably from the constant value of 59 mV, theoretically expected for an
electrochemically reversible 1e step (E ) [52]. However, despite the use of the iR
rev
positive feedback compensation during the electrochemical experiments, some
uncompensated solution resistance influences the peak-to-peak separation. In fact, under
5
the same experimental conditions, the 1e oxidation of [Fe(η -C H ) ], added as an
5
5 2
internal standard to the solutions and which is known to involve minimal structural
reorganization, exhibits ∆E separation increasing from 70 mV to 120 mV in the same v
p
range.
Table 2. Electrochemical data of the synthesized compounds acetylferrocene and 1-9.
Experimental conditions: CH Cl 0.1 M [NBu ][PF ]; working electrode: glassy carbon; reference
2
2
4
6
-1
electrode (RE): Ag/AgCl, 3 M KCl. E have been measured at 0.2 V s scan rate.
p,a
Compound
E°’ (V vs. RE) E (V vs. RE)
p,a
acetylferrocene
0.720
-
9

ACCEPTED MANUSCRIPT
1
0.730
0.728
0.728
0.717
0.716
0.720
0.734
0.726
0.728
0.953, 1.28
2
3
4
5
6
7
8
9
1.07
0.974, 1.30
0.970
1.27
1.03
circa 1.1
circa 1.4
circa 1.1
1
0

ACCEPTED MANUSCRIPT
Figure 2. Cyclic voltammetry of a) acetylferrocene, b) 1, c) 4, and d) 7. Experimental conditions:
1
mM in CH Cl 0.1 M [NBu ][PF ]; working electrode: glassy carbon; reference electrode:
2 2 4 6
-1
Ag/AgCl, 3 M KCl, scan rate: 0.2 V s .
1
1

ACCEPTED MANUSCRIPT
In all cases, the metal-centered process A/B is completely independent on the nature of
the organic moiety bound to the ferrocenyl carbonyl fragment, being the experimental E°’
values equal to that of acetylferrocene, within experimental error. However, this redox
process is complicated by further oxidation peaks at more positive potentials (> 0.85 V),
all characterized by the absence of an associated reduction in the reverse scan (EC
mechanism, chemically irreversible reaction following the electron transfer) attributable
to the organic moiety only.
For compounds 1 and 3, two oxidation peaks (C and D) are observed at about 0.95 and
1
.3 V, respectively, with a very limited difference in E (C) and E (D) with respect to
p,a p,a
the halogen. Apparently, peak D has an associated reduction peak (E), but its shape (i.e.,
symmetrical), intensity (i.e., i (E)/i (D) > 1) and peak potential (i.e., E (D) - E (E) >
p,c
p,a
p,a
p,c
140 mV) is compatible with the presence of adsorption complications, that modify the
diffusion-controlled peaks and make the analysis of the electrochemical results quite
complicate. This complication is absent in the case of 2, that, in turn, shows only peak C
at a more positive potential with respect to 1 and 3.
The electrochemical scenario of compounds 4-6 is similar to that of 2, showing a single
chemically irreversible oxidation peak C’ always located at more positive potentials with
respect to the relative 1-3 compounds. Minor reduction peaks associated to fragments
produced during the organic-centered EC step are visible around 0.3-0.4 V.
In the 7-9 series, only ill-defined, minor oxidations are observed at potentials > 0.9 V and
will not further discussed.
No direct reductions are observed for all the compounds in the available window of
potential, and the use of a different electrode material (Pt instead of GC) did not modified
the overall electrochemical response.
1
2

ACCEPTED MANUSCRIPT
The formation of reactive oxygen species (ROS) has been claimed or demonstrated to
explain the biological properties of ferrocene derivatives. However, there is no meeting
of minds among the researchers on whether the redox state of Fe affects the result. For
sure, ROS (e.g. hydroxyl radical and peroxide) may be formed by ferrocene, ferrocenium
or iron ions, and are important in the control of the parasitosis. A direct involvement in
the ROS formation by the ferrocene entity was supposed to explain cell death in T. brucei
and T. cruzi, as well as in L. donovani and L. infantum promastigotes [31, 34, 53-56]. In
the actual series of ferrocenes having the same E°’, the ability to produce ROS through
the formation of ferrocenium cations will be the same, and, hence, differences in the
killing ability will be due to other (overlapped or concurrent) mechanisms.
2.3. In vitro test against epimastigote form of T. cruzi
Considering the recommendations for the development of new promising anti-chagasic
drugs, three different strains of T. cruzi, belonging to the discrete type units TcI, TcV and
TcVI, were selected for the in vitro screening [57-58]. In the first step, the extracellular
epimastigote forms were exposed to compounds 1-9 and BZN to determine their IC50
values after 72 h of exposure at concentration ranging from 6.3 µM to 50 µM values.
Moreover, toxicity values expressed as IC were determined after 72 h of exposure at
5
0
concentration ranging from 10 µM to 200 µM employing mammalian Vero cells. The
selectivity index (SI) was obtained from the data reported above. Results are summarized
in Table 3. Compounds with a SI higher than 5 in epimastigote forms in at least two T.
cruzi strains will move to the following step of the in vitro screening. Thus, compounds
2
, 5 and 6 were selected to determine the IC value and SI using both the amastigote and
50
trypomastigote forms of T. cruzi. (Table 4 and Table 5).
1
3

ACCEPTED MANUSCRIPT
According to the chemical structure of compounds 1-9 and their respective cytotoxicity in
Vero cells, compounds that include an aniline moiety (4-6) showed higher IC values
5
0
than compounds with a piperazine moiety (1-3) that showed slightly higher IC values
5
0
than those with a phenethylamine moiety (7-9).
Table 3. In vitro activity of the studied compounds (1-9) in epimastigote forms.
c
Toxicity IC50
Selectivity Index (SI)
a
Activity IC (µM)
5
0
b
Comp
(µM)
Arequipa
25.5 ± 8.6
18.9 ± 1.6
16.5 ± 2.3
SN3
Tulahuen
25.3 ± 0.9
6.8 ± 0.1
7.5 ± 1.1
15.7 ± 1.7
6.4 ± 2.4
9.8 ± 0.9
6.6 ± 3.0
9.4 ± 2.9
9.5 ± 1.0
8.0 ± 1.9
Vero cells
>200
Arequipa
>7.8
2.2
SN3
>13.1
3.4
Tulahuen
>7.9
6.0
BZN
15.3 ± 1.9
12.1 ± 2.2
18.3 ± 7.6
1
2
3
4
5
6
7
8
9
41.0 ± 7.3
84.0 ± 8.0
27.1 ± 10.6
61.6 ± 10.7
107.0 ± 22.3
94.9 ± 22.8
30.6 ± 2.7
31.3 ± 4.4
38.5 ± 2.0
5.1
4.6
11.2
1.7
32.6 ± 12.5 16.2 ± 5.1
0.8
1.7
18.3 ± 3.5
21.2 ± 2.0
17.5 ± 2.6
19.9 ± 2.6
16.9 ± 4.4
18.0 ± 2.6
22.9 ± 3.0
27.9 ± 9.6
34.4 ± 4.9
15.8 ± 2.2
11.9 ± 1.3
13.6 ± 1.5
3.4
2.7
9.7
5.0
3.8
10.9
14.5
3.3
5.4
2.8
1.5
1.9
1.9
2.6
3.3
2.1
2.8
4.8
IC : Compound concentration that causes a reduction of 50% of parasite or cells growth,
5
0
calculated by Graphpad prism 7 software. Data are the mean of triplicate experiments ± the
standard deviation (SD). (a) IC values against T. cruzi epimastigote forms. (b) IC Vero cells
5
0
50
cytotoxicity. (c) Selectivity Index = IC Vero cells / IC T. cruzi strain.
5
0
50
Table 4. In vitro activity of the compounds 2, 5 and 6 in both amastigote and trypomastigote
forms of T. cruzi.
1
4

ACCEPTED MANUSCRIPT
Activity IC (µM)
5
0
T. cruzi Arequipa
T. cruzi SN3
T. cruzi Tulahuen
Compounds
Amastigote
form
Trypomastigote Amastigote Trypomastigote
Amastigote
form
Trypomastigote
form
form
form
form
BZN
3.0 ± 0.4
4.5 ± 0.5
7.3 ± 0.5
10.5 ± 0.2
6.2 ± 1.7
2.0 ± 0.1
11.1 ± 2.5
5.2 ± 1.2
4.4 ± 0.2
3.8 ± 0.6
10.6 ± 1.2
8.6 ± 0.5
33.3 ± 1.6
4.8 ± 0.9
3.9 ± 0.2
6.3 ± 1.1
3.2 ± 0.1
6.0 ± 0.4
6.7 ± 1.0
5.1 ± 0.4
32.0 ± 2.2
13.0 ± 2.4
2
5
6
13.0 ± 5.3
10.2 ± 1.5
In vitro activity against amastigote and trypomastigote forms of T. cruzi, expressed as IC . Data
5
0
are the mean of duplicate (amastigotes form) and triplicate (trypomastigote forms) experiments ±
SD.
Table 5. Selectivity index (SI) of the compounds 2, 5 and 6 in both amastigote and
trypomastigote forms of T.cruzi.
Selectivity Index (SI)
T. cruzi Arequipa
T. cruzi SN3
T. cruzi Tulahuen
Compounds
Amastigote
Trypomastigote
Amastigote
Trypomastigote
form
Amastigote
form
Trypomastigote
form
form
form
form
BZN
>67.2
>32.5
>45.6
>6.0
17.4
8.2
>51.7
13.4
33
>29.9
16.6
3.3
2
5
6
18.7
14.7
9.1
41.7
9.6
21.8
10.1
11.0
18.2
9.3
15.9
7.3
Considering the activity results reported in Table 4, the compound 2 stands out for its
IC as the most promising ferrocenyl Mannich base derivative showing an IC value
5
0
50
1
5

ACCEPTED MANUSCRIPT
around 5 µM in both amastigote and trypomastigote forms. These values meet one of the
established criteria (IC < 10 µM) for the selection of the most promising compounds
5
0
against CD [59-61].
The SI is an important criterion for the selection of an antichagasic compound. Although,
none of the compounds studied meets the established criteria to be selected for the
following trypanocidal activity assays (SI > 50) [59-61], compound 2 showed a
substantially important result with its SI values in the trypomastigote forms of the
Arequipa strain of T. cruzi (Table 5).
For a more accurate study of the in vitro antiparasitic activity, the effect of 2 was studied
at different concentration levels. BZN was also evaluated as a reference standard (Figure
3
). A remarkable reduction of infected cells was observed while the concentration of 2
was increasing gradually. For example, the rates of infected cells treated with 2 decreases
to 10% at 50 µM and with an IC value of 6.4 µM. BZN presented a similar behavior
5
0
(
approximately to 15% at 50 µM, IC value of 4.1 µM). In addition, a decrease in the
50
number of amastigote and trypomastigote forms was observed. So, compound 2 could
inhibit the multiplication of the parasite, as well as cause its death (trypanocidal activity).
1
6

ACCEPTED MANUSCRIPT
Figure 3. Reduction of the infection of T. cruzi Arequipa strain concerning the decrease of
amastigote, trypomastigote forms and infected Vero cells treated with BZN and 2. Data are the
mean of two (infected cells an amastigote forms) or three (trypomastigote forms) separate
determinations. IC value (in parenthesis) were calculated using Graphpad prism 7 software.
5
0
2.4. In vitro genotoxicity screening. SOS/umu test.
As 2 has demonstrated to be the most promising derivative of the studied compounds, the
SOS/umu test has been employed to determine its genotoxic potential [44]. In this study,
2
was considered non-genotoxic both with and without metabolic activation, whereas
1
7

ACCEPTED MANUSCRIPT
BZN induced genotoxicity in the same conditions [44]. Therefore, the results confirm the
promising profile of compound 2.
2
.5. Studies of the mechanism of action.
.5.1. Metabolite excretion.
2
T. cruzi is unable to completely degrade glucose to CO in the presence of oxygen, and
2
catabolizes glucose at a high rate, acidifying the medium when excreting incompletely
oxidized acids [62]. The final products in this catabolism are mainly pyruvate, acetate,
1
and succinate, among the others [63]. Therefore, the H NMR spectra of the mediums of
non-treated parasites and parasites treated either with BZN or compound 2 at IC50
concentrations were registered (see supporting information, together with the chemical
shifts of the peaks for each metabolite). The qualitative and quantitative analysis of the
spectra was performed to compare the effect of BZN and 2 on the glucose metabolism of
T. cruzi Arequipa epimastigote form. Figure 4 shows the results obtained in comparison
with non-treated parasites.
1
8

ACCEPTED MANUSCRIPT
Figure 4. Variation (%) among peaks of catabolites excreted by epimastigotes of T. cruzi
Arequipa strain exposed to BZN and compound 2 at IC concentrations in comparison with non-
5
0
treated parasites incubated 72 h.
The excretion of all the metabolites was disturbed in both treatments. Minor alterations
were observed for BZN, but 2-treated parasites showed an increase of about 80% for all
the metabolites. Most of them are final products of catabolic routes that the parasite can
use in conditions of energy deficit [62]. Therefore, it is interesting to note that these
increases may be the basis of a change in the succinate pathway that could be related to a
mitochondrial dysfunction [64]: an alteration at the mitochondrial level may be the
ultimate reason for the changes observed in the excreted products. Hence, mitochondrial
studies were conducted.
2.5.2. Mitochondrial membrane potential
+
In normal cells, active pumping of H is produced to maintain the electrochemical
gradient, the integrity and the function of mitochondria. Disturbances in this potential
lead to a decrease in ATP production, causing apoptosis and/or necrosis [65]. Therefore,
the alteration in the glycolytic pathway could be the result of a membrane depolarization,
which produces a cellular accumulation of succinate, pyruvate and malate. This causes a
blockage of the glycolytic pathway, and the energy is obtained through other catabolic
routes. In order to evaluate if the effect of 2 in the glycolytic pathway is due to a
mitochondrial dysfunction, the fluorescence intensity of rhodamine 123 (Rho 123) was
quantified by cytometry flow. Figure 5 shows the results obtained after the analysis of
the mitochondrial membrane potential. BZN caused a decrease of the membrane potential
(15.2%) but higher alteration was observed for 2-treated parasites, reaching a
depolarisation of 75.1% in the mitochondrial membrane. Therefore, it can be suggested
1
9

ACCEPTED MANUSCRIPT
that the trypanocidal activity of compound 2 is related to an effect at the mitochondrial
level.
Figure 5. Cytometry analysis of the mitochondrial membrane potential from epimastigotes of T.
cruzi Arequipa strain exposed to BZN and compound 2 at their IC in comparison to non-treated
5
0
2
0

ACCEPTED MANUSCRIPT
parasites incubated 72 h. (a) blank, (b) control (non-treated), (c) BZN, (d) compound 2. Percent of
variation in mitochondrial membrane potential: BZN (-15.2 %) and compound 2 (-75.1 %).
2.5.3. DNA and RNA levels
The depolarisation of the mitochondrial membrane potential produces an imbalance in the
+
NADH/NAD and ATP/ADP ratios [62], which may compromise the DNA and RNA
levels of the parasite. Therefore, DNA and RNA levels were quantified by cytometry
flow using acridine orange (AO). The results are shown in Figure 6, and non-significant
differences were observed for the parasites treated with BZN and 2 with respect to the
untreated ones.
Figure 6. Cytometry analysis of the inhibition in DNA and RNA levels of epimastigotes of T.
cruzi Arequipa strain exposed to BZN and compound 2 at their IC in comparison with non-
5
0
treated parasites incubated 72 h: (a) blank, (b) control (non-treated), (c) BZN and (d) compound
. Variation (%) at DNA and RNA levels: BZN (-4.7 %) and compound 2 (5.9 %).
2
2
1

ACCEPTED MANUSCRIPT
2.5.4. T. cruzi Fe-SOD enzyme
Finally, inhibition studies were performed in order to evaluate whether the effect at the
mitochondrial level is ultimately due to the redox stress produced by the inhibition of the
mitochondrion-resident Fe-SOD, a key enzyme in the defense of T. cruzi but absent in
other eukaryotic cells,[64, 66]. Fe-SOD is one of the most studied therapeutic targets
because it shows structural and biochemical differences with respect to its human
homologue Cu/Zn-SOD and eliminates ROS, which allows T. cruzi to protect itself from
the damage produced by oxidative stress [67-68].
Figure 7 shows the inhibition data of 2, with the corresponding calculated IC for Fe-
5
0
SOD from T. cruzi and its human homologue Cu/Zn-SOD from erythrocytes. The results
indicate that 2 showed a highly inhibitory activity for the Fe-SOD enzyme, whereas there
was no inhibitory activity for the human Cu/Zn-SOD until 50 µM concentration. The
inhibitory activity of 2 was so high to reach 100% inhibition of Fe-SOD at a 6.3 µM
concentration with an IC value of 1.2 µM. Therefore, being 2 a remarkable selective
50
inhibitor of Fe-SOD, it can be suggested that Fe-SOD enzyme can be one of the targets of
this compound. In addition, this significant inhibitory activity could be the final cause of
the mitochondrial dysfunction, because the Fe-SOD is a crucial component of the redox
stress response [69]. In any case, some other mechanisms of action and the chance of a
multitarget compound should not be discarded.
2
2

ACCEPTED MANUSCRIPT
-
1
Figure 7. In vitro inhibition of Fe-SOD (activity 32.3 ± 6.7 U mg ) and CuZn-SOD (activity
-1
7
81.5 ± 3.0 U mg ) for compound 2. Values are the average of three separate rate determinations
±
standard deviation. IC value (in parenthesis) were calculated using Graphpad prism 7.
50
2.6. Docking study of the new compounds in the T.cruzi Fe-SOD enzyme
To further analyze the observed inhibition of the Fe-SOD enzyme by compound 2 and to
study the potential binding of the whole series presented here, docking studies of the nine
compounds on the mitochondrial T. cruzi enzyme were performed. The predicted binding
modes, together with the MolDock scores [70] are shown in Figure S5.1 (see supporting
information). As it can be observed in Figure 8a, the results show that 2 is bound in the
dimer interface of the enzyme, in a way very similar to that obtained for other series of
compounds [43-44]. This interface and the amino acids showing interactions with the
ferrocene compounds are conserved in Fe-SOD enzymes from different pathogenic
protozoans [71]. The ferrocene moiety performs π-stacking interactions with Tyr36, a
conserved amino acid that is nitrated upon peroxynitrite treatment, leading to inactivation
of T. cruzi [72]. At the same time, a hydrogen bond is observed between the carbonyl
group of the compounds and either Tyr36 or Tyr169, suggesting that this group is
important for affinity. The ligand amino group binds close to Arg177, performing, in
2
3

ACCEPTED MANUSCRIPT
some compounds, hydrogen bonds with the backbone carbonyl group of Val120. This
differs from what observed with other Mannich-type compounds, which presented this
group involved in hydrogen bonds with Glu166 [43]. The ferrocene and polar groups are
bound deeper in the dimer interface, whereas the phenyl moiety extends more outside,
between Phe123 and Lys39. Concerning the docking scores, the predicted values
correlate very well with the activity of compounds against T. cruzi strains (Figure 8b),
with compounds 2 and 8 presenting the best-predicted affinity. This observation
reinforces the hypothesis about the mechanism of action of the investigated ferrocene
derivatives, which would act at the mitochondrial level by inhibiting Fe-SOD enzyme. In
general, the results predicted a better affinity on the T. cruzi Fe-SOD enzyme for
compounds presenting the piperazine scaffold and the trifluoro-phenyl substituent at the
phenyl moiety.
Figure 8. Docking study of the new compounds on the T.cruzi Fe-SOD enzyme (PDB entry
4
DVH). (a) Proposed binding mode for compound 2 on the enzyme, with protein and ligand
colored according to: oxygen (red), nitrogen (blue), fluoride (pale blue), iron (brown), carbon
pale yellow for compound 2, and green or cyan for each protein protomer). The enzyme iron ions
(
are depicted as brown spheres. Hydrogen bonds between protein and ligand are shown as yellow
dashed lines. The figure was created with the PyMOL Molecular Graphics System, Version 1.7
2
4

ACCEPTED MANUSCRIPT
Schrödinger, LLC. (b) Comparison between the MolDock score (red) and the media of the in
vitro activity values for the three strains (blue).
3. Conclusion
The simple and cost-efficient synthesis, the characterization and the preliminary
antitrypanosomal in vitro screening of nine ferrocenyl Mannich base derivatives were
performed and here reported. In particular, the compound 2, 1-ferrocenyl-3-(4-(4-
(trifluoromethyl)phenyl)piperazin-1-yl)propan-1-one, got attention for its attractive
results in terms of trypanocidal activity. Moreover, this compound is non-genotoxic and
the results obtained from the studies of the mechanism of action (electrochemistry,
metabolites excretion, mitochondrial membrane potential and in silico tests) have
revealed that it could act at the mitochondrial level by selective inhibition of Fe-SOD.
Hence, these encouraging results allow the researchers to move towards finding novel
bioorganometallic compounds as very valuable, affordable and safe new candidates for
the development of future trypanocidal drugs.
4
4
4
. Experimental
.1. Chemistry
.1.1. Material and methods
All the chemicals (Acros Organics, Alfa Aesar-Thermo Fisher Scientific, Fluka, Panreac
and Sigma-Aldrich) were used without further purification. The intermediate ferrocene
2
5

ACCEPTED MANUSCRIPT
complexes (i1-i2 and acetylferrocene
)
and all the studied compounds were prepared
according to literature procedures as described below.
The microwave-assisted synthesis of compounds 5 and 6 was performed using a
microwave synthesizer Discover SP® (CEM Corporation). The synthesis of studied
compounds was monitored by thin layer chromatography (TLC) employing SIL G/UV ,
2
54
0.2 mm of layer thickness (Macherey-Nagel GmbH & Co. KG. Germany), organic
solvents in different volume to volume proportion (CH Cl /methanol or hexane/ethyl
2
2
acetate), and UV light (254 nm) or iodine chamber or a combination of both were
employed to spots detection. Column chromatography was carried out on a
CombiFlash®RF instrument (Teledyne Isco. USA) with Silica RediSep Rf gold®
columns employing a normal phase gradient. Infrared spectra (IR), proton nuclear
1
13
magnetic resonance ( H NMR), carbon nuclear magnetic resonance ( C NMR), reversed
phase HPLC (RP-HPLC) and elemental microanalyses (CHN) were carried out for the
chemical characterization of these compounds. IR measurements (KBr pellets or NaCl
plates) were performed on a Nicolet Nexus FTIR (Thermo Madison, USA). IR
-1
1
frequencies (cm ) were reported as medium (m) and strong (s). The NMR spectra ( H
1
3
NMR, C NMR, and 2D NMR) were measured on a Bruker 400 Ultrashield operating at
1
13
4
5
4
00 MHz ( H) and 100 MHz ( C), NMR-Bruker Avance III spectrometer operating at
1
13
00 MHz ( H) and 125.7 MHz ( C), or on a JEOL Eclipse Plus spectrometer operating at
1
13
00 MHz ( H) and 100.6 MHz ( C) respectively. Chemical shifts (δ) were expressed in
parts per million (ppm) referenced to solvent resonances, coupling constants (J) were
given in hertz (Hz) and multiplicity was reported as singlet (s), bs (broad singlet), doublet
(d), doublet of doublets (dd), triplet (t) and multiplet (m).
The purity of acetylferrocene and 3-chloropropanoylferrocene (i1) was assessed by
analytical RP-HPLC. Chromatographic analyses were carried out using a C18
2
6

ACCEPTED MANUSCRIPT
Phenomenex Phenosphere-NEXT (5-µm, 250×4.6 mm ID) column on a Waters HPLC-
MS instrument (equipped with Alliance 2695 separations module, 2487 dual lambda
absorbance detector, and 3100 mass detector). The UV-visible detector was set at 210
nm, the eluant was a 30/70 mixture of 15 mM formic acid/CH OH (10/90 for
3
-1
acetylferrocene) and the flow rate was 0.5 mL min . Mass spectra of intermediate
complexes were recorded using source and desolvation temperatures set to 150 and 250
°
C, respectively, with nitrogen used as both drying and nebulizing gas. The cone and the
capillary voltages were usually +30 V (positive ion mode) and 2.70 kV, respectively. The
+
quasi-molecular ion peak [M+H] of the synthesized complexes or the molecular ion
+
peak [M] of the corresponding ferrocenium ions formed into the mass spectrometer[73]
were assigned on the basis of the m/z values and of the simulated isotope distribution
patterns. The purity of compounds 1-9 and acryloylferrocene (i2) was assessed by
elemental microanalyses of carbon, hydrogen and nitrogen on a CHN-900 elemental
analyzer (Leco, TresCantos, Spain). The elemental analysis results were within ±0.4
respect to theoretical value. Melting point (uncorrected) was determined using a Mettler
FP82 + FP80 (Greifensee, Switzerland).
4.1.2. Synthesis and characterization of the intermediate ferrocenes.
4.1.2.1. Synthesis of acetylferrocene. Acetylferrocene was synthesized according to
previously reported procedures.[49] Briefly, 85% phosphoric acid (4 mL) was slowly
added to a stirred solution of ferrocene (4.0 g, 21.5 mmol) and acetic anhydride (13.3 mL,
140.0 mmol). The mixture was heated in a water bath maintained at 50 °C for 10 min and
then cooled in ice. Water (25 mL) was added to the mixture, that was then neutralized
with NaHCO until the end of CO formation. CH Cl (50 mL) was added and the orange
3
2
2
2
organic layer was separated and retained in the dark, whereas the brown-coloured
2
7

ACCEPTED MANUSCRIPT
aqueous layer washed with a further 20 mL of CH Cl . The combined organic fractions
2
2
were washed twice with water and then dried with anhydrous magnesium sulfate. Crude
acetyl ferrocene, obtained removing the solvent under reduced pressure, was purified by
silica gel column chromatography using a 95/5 mixture of petroleum ether/ethyl acetate
as starting eluent to first separate the unreacted ferrocene. Once the ferrocene was eluted,
the eluent was replaced by 100% ethyl acetate to collect acetyl ferrocene, that was then
obtained by removing the solvent on a rotary evaporator.
4.1.2.2. Synthesis of 3-chloropropanoylferrocene (i1). 3-chloropropanoylferrocene was
prepared according to slight modifications of previously published procedures.[50]
Briefly, a solution of 3-chloropropanoyl chloride (3.2 g, 25.0 mmol) in CH Cl (10 mL)
2
2
was added dropwise to a suspension of AlCl (3.7 g, 27.4 mmol) in CH Cl (10 mL) at
3
2
2
room temperature. After 2h stirring, the solution was cooled to 0°C and ferrocene (4.7 g,
5.0 mmol) in CH Cl (60 mL) was added dropwise. The resulting solution was allowed
2
2
2
to warm to room temperature and then left stirring overnight. The crude was purified by
silica gel column chromatography using a step gradient as follows: 100% petroleum
ether, 60/40 petroleum ether/ethylacetate, 40/60 petroleum ether/ethylacetate and, finally,
100% ethylacetate. Unreacted ferrocene was firstly eluted, followed by the product;
finally, if any, the disubstituted derivative could be eluted.
4.1.2.3. Synthesis of acryloylferrocene (i2). Acryloylferrocene was prepared according to
slight modifications of previously reported.[48] 3-chloropropanoylferrocene (1.0 mmol)
and CH CO K (3.0 mmol) was diluted in ethanol (25 mL). The reaction mixture was
3
2
refluxed for 2.5 h. Ethanol was removed by evaporation under vacuum, and the residue
was diluted in CH Cl (30 mL). The organic phase was washed with water, dried with
2
2
anhydrous Na SO , filtered, and evaporated. The final compound was purified by
2
4
2
8