Thiosemicarbazones as Aedes aegypti larvicidal

Article abstract of DOI:10.1016/j.ejmech.2015.04.061

Abstract A set of aryl- and phenoxymethyl-(thio)semicarbazones were synthetized, characterized and biologically evaluated against the larvae of Aedes aegypti (A. egypti), the vector responsible for diseases like Dengue and Yellow Fever. (Q)SAR studies were useful for predicting the activities of the compounds not included to create the QSAR model as well as to predict the features of a new compound with improved activity. Docking studies corroborated experimental evidence of AeSCP-2 as a potential target able to explain the larvicidal properties of its compounds. The trend observed between the in silico Docking scores and the in vitro pLC50 (equals -log LC50, at molar concentration) data indicated that the highest larvicidal compounds, or the compounds with the highest values for pLC50, are usually those with the higher docking scores (i.e., greater in silico affinity for the AeSCP-2 target). Determination of cytotoxicity for these compounds in mammal cells demonstrated that the top larvicide compounds are non-toxic.

Full text of DOI:10.1016/j.ejmech.2015.04.061

Accepted Manuscript
Thiosemicarbazones as Aedes aegypti Larvicidal
João Bosco P. da Silva, Daniela Maria do A.F. Navarro, Aluizio G. da Silva, Geanne
K.N. Santos, Kamilla A. Dutra, Diogo Rodrigo Moreira, Mozart N. Ramos, José
Wanderlan P. Espíndola, Ana Daura T. de Oliveira, Dalci José Brondani, Ana Cristina
L. Leite, Marcelo Zaldini Hernandes, Valéria R.A. Pereira, Lucas F. da Rocha, Maria
Carolina A.B. de Castro, Beatriz C. de Oliveira, Que Lan, Kenneth M. Merz, Jr.
PII:
S0223-5234(15)30024-6
10.1016/j.ejmech.2015.04.061
EJMECH 7876
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 17 November 2014
Revised Date: 28 April 2015
Accepted Date: 29 April 2015
Please cite this article as: J.B.P. da Silva, D.M.d.A.F. Navarro, A.G. da Silva, G.K.N. Santos, K.A. Dutra,
D.R. Moreira, M.N. Ramos, J.W.P. Espíndola, A.D.T. de Oliveira, D.J. Brondani, A.C.L. Leite, M.Z.
Hernandes, V.R.A. Pereira, L.F. da Rocha, M.C.A.B. de Castro, B.C. de Oliveira, Q. Lan, K.M. Merz Jr.,
Thiosemicarbazones as Aedes aegypti Larvicidal, European Journal of Medicinal Chemistry (2015), doi:
10.1016/j.ejmech.2015.04.061.
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
Thiosemicarbazones as Aedes aegypti Larvicidal
a,*
a,*
a
João Bosco P. da Silva , Daniela Maria do A. F. Navarro , Aluizio G. da Silva , Geanne K. N.
a
a
a,1
a
Santos , Kamilla A. Dutra , Diogo Rodrigo Moreira , Mozart N. Ramos , José Wanderlan P.
b
b
b
b
Espíndola , Ana Daura T. de Oliveira , Dalci José Brondani , Ana Cristina L. Leite , Marcelo Zaldini
b
c
c
c
Hernandes , Valéria R. A. Pereira , Lucas F. da Rocha , Maria Carolina A. B. de Castro , Beatriz C.
c
d
e,2
de Oliveira , Que Lan , Kenneth M. Merz Jr.
a
Departamento de Química Fundamental, Centro de Ciências Exatas e da Natureza, Universidade
Federal de Pernambuco, 50670-901, Recife, PE, Brazil.
b
Departamento de Ciências Farmacêuticas, Centro de Ciências da Saúde, Universidade Federal de
Pernambuco, 50740-521, Recife, PE, Brazil.
c
Departamento de Imunologia, Centro de Pesquisas Aggeu Magalhães, Fundação Oswaldo Cruz,
5
0670-420, Recife, PE, Brazil.
d
Department of Entomology, University of Wisconsin-Madison, 1630 Linden Drive, Madison, WI
3706, USA.
5
e
Quantum Theory Project, University of Florida, 2234 New Physics Building, Gainesville, PO Box
18435, Florida and Institute for Cyber Enabled Research, Department of Chemistry and the
1
Department of Biochemistry and Molecular Biology, Michigan State University, 578 S. Shaw Lane,
East Lansing, MI 48824, USA
KEYWORDS: Aedes aegypti, larvicide, thiosemicarbazones, sterol carrier protein-2, QSAR, Docking.
*
Corresponding authors: Departamento de Química Fundamental, Centro de Ciências Exatas e da
Natureza, Universidade Federal de Pernambuco, 50670-901, Recife, PE, Brazil. Tel. +55 81 2126-
419/-7468; Fax: +55 81 2126-8442. e-mails: paraiso@ufpe.br (J.B.P. da Silva) and dmafn@ufpe.br
D.M.A.F. Navarro).
7
(

ACCEPTED MANUSCRIPT
1
Present address: D.R. Moreira - Centro de Pesquisas Gonçalo Moniz - CPqGM, Rua Waldemar
Falcão, 121, Candeal, 40296-710, Salvador, BA, Brazil.
2
Present address: K. M. Merz Jr. - Institute for Cyber Enabled Research, Department of Chemistry and
the Department of Biochemistry and Molecular Biology, Michigan State University, 578 S. Shaw Lane,
East Lansing, MI 48824, USA.
ABSTRACT
A set of aryl- and phenoxymethyl-(thio)semicarbazones were synthetized, characterized and
biologically evaluated against the larvae of Aedes aegypti (A. aegypti), the vector responsible for
diseases like Dengue and Yellow Fever. (Q)SAR studies were useful for predicting the activities of the
compounds not included to create the QSAR model as well as to predict the features of a new
compound with improved activity. Docking studies corroborated experimental evidence of AeSCP-2 as
a potential target able to explain the larvicidal properties of its compounds. The trend observed between
the in silico Docking scores and the in vitro pLC50 (equals −log LC50, at molar concentration) data
indicated that the highest larvicidal compounds, or the compounds with the highest values for pLC50,
are usually those with the higher docking scores (i.e., greater in silico affinity for the AeSCP-2 target).
Determination of cytotoxicity for these compounds in mammal cells demonstrated that the top larvicide
compounds are non-toxic.

ACCEPTED MANUSCRIPT
1
. Introduction
A large number of diseases are transmitted by mosquitoe vectors such as filariasis (Culex
quinquefasciatus) [1], malaria (Anopheles gambiae) [2], West Nile fever (Aedes albopictus) [3] and
dengue, Chikungunya and yellow fever (Aedes aegypti) [4]. These four last diseases are transmitted by
a virus
.
Nowadays, Dengue Fever (DF) is considered one of the most rapidly spreading diseases, being
disseminated mainly but not limited to tropical and subtropical regions [5]. Since DF correlates with
the vector expansion [6], the possibility of occurrence of DF in non-disease-endemic areas like the
continental United State (US) cannot be neglected. The epidemics in Puerto Rico in the 1990’s [7] and
in Hawaii in 2001 [8] support this possibility. The large international mobility throughout the US
territory (the so called imported cases) may partially be responsible but the occurrence of DF was
related mainly due to the presence of two principal vectors (A. albopictus and A. aegypti) in southern
and central US [9]. For some of the reasons above mentioned, Australia [10], Spain [11] and Germany
[
12] may also be considered areas of potential risk for DF infection.
Recent estimates of the World Health Organization (WHO) indicate c.a. 50-100 million people
worldwide are annually infected with dengue and 2.5 billion people (40% of the word population) are
at risk in more than 100 countries [13]. Although many efforts have been envisaged for developing a
vaccine simultaneously for the four dengue virus serotypes (1-4), none are available at the moment
[
14]. Recently a fifth serotype was reported [15]. This complicates the situation further. Likewise, there
are no specific medicines available for treating Dengue-infected patients. The current standard medical
treatment is limited to controlling the symptoms of the disease [16,17]. DF has a strong economic
impact. For instance, those affected frequently stay out of the productive and/or educational systems for
week(s). Sectors like tourism are extremely impacted as well [18,19]. Besides this, during epidemics

ACCEPTED MANUSCRIPT
both public and privative health systems are frequently exhausted to the extreme. For instance,
according to WHO, 500,000 people are hospitalized annually with a mortality rate c.a. 2.5% [13].
To control Dengue transmission, tools for vector control are necessary, including larvicide use,
entomological monitoring, biological control as well as public information campaigns [20,21].
Currently, among the most employed larvicides are the organophosphate temephos and the
toxins of the Bacillus thuringiensis (Bti) var. israelensis [22]. Recent studies, have pointed out that the
long-term use of themephos is producing A. aegypti resistant populations [23,24]. Likewise, new
alternatives for low cost production of Bti on a large scale [26,27] and the possibility of a mosquito
resistant to the Bti toxin [28] are relevant concerns that demand further research to discover better
larvicides for A. aegypti.
An increasing number of publications have recently reported new natural and synthetic
larvicides for A. aegpti. For instance, in 2011 Neto and co-workers [29] published a review about
natural products as larvicides for A.aegypti, covering 21 different plants with LC values ranging from
5
0
0
.04 ppm to 100 ppm. Since then, other papers researching chemicals derived from natural products
have been published [30-33], but with a range of activity like that of ref. 29. On the other hand, studies
on a smaller number (in comparison with the number of natural products) of synthetic compounds
against A. aegypti have appeared in the literature. In this case both Structure-Activity Relationships
(
SAR) [34] and Quantitative Structure-Activity Relationships (QSAR) [35] based on the ligand
approach have been developed. Concerning SAR studies, Cavalcanti and co-workers [36-38] reported
on the importance of hydrophobicity for improving larvicide activity for A. aegypti for a series of
monoterpenic and benzoquinonic derivatives. Similar to this, Cantrell and co-workers observed a clear
relationship between the larvicide activity against A. aegypti and the number of methylenic units in the
linear amine substituents attached to both alantolactone and isoalantolactone [39]. In the 2000’s, Eng
and co-workers reported studies on four different classes of triorganotin complexes as larvicide activity
against A.aegypti [40-43]. In 2009, Hansch and Verma [44] revisited the Eng’s larvicide activities

ACCEPTED MANUSCRIPT
results for three of these triorganotin complexes [41-43] and proposed QSAR models with the
hydrophobic (π
X S
) and volume (MR – molar refractivity or E – Taft’s steric) parameters as the most
important for describing larvicide activity.
At this point it is important to stress that it is possible to find in the literature other SAR and
QSAR studies on the larvicide activity for other mosquitos (e.g. Culex quinquefasciatus [45-47]) or
about the repellence for adult A. aegypti [48-51]. They will not be discussed here, however, since they
are out of the scope of the present work.
In 2002, Park and co-workers [52] reported on LC50 larvicide activity against A. aegypti among
four structurally related compounds obtained from the fruits of Piper nigrum: pellitorine (0.92 ppm),
guineensine (0.89 ppm), pipercide (0.1 ppm), retrofractamide A (0.04 ppm) and the commercial
insecticidal piperine (5.10 ppm). The structural comparison of these compounds made it clear that
larvicide activity is inversely related to the distance between the 3,4-methylenedioxyphenyl and the
isobutylamide groups. Besides, the ability of the isobutylamide moiety to make hydrogen-bond
interactions, probably with some polar amino acid residue in a biological target, may explain the
expressive decrease in the larvicide activity in piperine when the isobutylamide is changed by a six-
member cyclic amide.
In an attempt to synthetize some amidic compounds, our research group discovered that the
corresponding carboxylic acid precursors had a higher larvicide activity against A. aegypti than the
corresponding amides [53]. As consequence, in 2009, our group reported for the first time on the
larvicidal activity against A. aegypti larvae of 3-(3-aryl-1,2,4-oxadiazol-5-yl) propionic acids (AOPA)
where the presence of electron-withdrawing substituents in the para position of the phenyl ring was
shown to be important for the larvicide activity [54]. Similar observations about the electronic
substituent effect on the phenyl ring were explored through QSAR equations on 1-(2,6-
halogenbenzoyl)-5-(4-halogen-phenyl) biuret compounds by Bordas and co-workers for larvicidal
activities (LC ) against A. aegypti larvae [55].
5
0

ACCEPTED MANUSCRIPT
In this work we report our research on the synthesis, characterization, and evaluation of larvicide
activity against A. aegypti of aryl-semicarbazones (1,2), aryl-thiosemicarbazones (3-10) and phenoxyl
thiosemicarbazones (12-18) (Fig. 1).
The choice of these compounds was based on a compromise among four aspects. The first was to
recover through the molecular structure two parameters previously reported in the literature as
important for larvicide activity against A. aegypti, namely hydrophobicity and the ability to form
hydrogen bonds with polar head fragments. The second was to use hydrazones, a well-established class
of compounds employed in pest control in both agriculture and horticulture [56,57]. Third was to use
(
thio)semicarbazones, which are hydrazone structural analogues with broad spectra of biological
activities [58-62], easily prepared and can be used as intermediates in the synthesis of further
heterocyclic compounds with pharmacological potential [63-66]. Fourth, the presence of the spacer unit
-
2
O-CH -CH= in the phenoxy-compounds was used to introduce some degree of rotational freedom on
the phenyl substituted moiety, improving (as was the case for other biological activities [67]) the
chances to modify the larvicide activity.
Another aspect investigated about these compounds was how safe they are for other living systems.
Therefore, in a first approach, toxicological tests were developed for a mammal model.
Electronic structure and lipophilicity calculations were used to create a QSAR model for the ligands
with a double objective: to analyze the importance of the parameters used to explain the larvicide
activity against A.aegypti for a training set of molecules; and to explore the predictability of the model.
In fact, from this QSAR study, the larvicide activities of three compounds were predicted and
evaluated.
After that, in order to improve understanding of how these compounds act on the A. aegypti, the
affinity profile of two synthetized compounds were evaluated on the A. aegypti Sterol Carrier Protein-2
(
AeSCP-2, a system beleved to be involved in the intracellular transport of cholesterol [68]). These
were then compared with the affinity profile of a previous tested inhibitor [69]. Finally, docking studies

ACCEPTED MANUSCRIPT
were performed in order to increase the evidence of this target as the potential receptor for these
compounds.
2
. Results and discussion
2
.1 Synthesis
Figure1 shows the chemical structure of all compounds investigated in this work. The synthetic
procedures employed in preparation of compounds (1-18) are shown in Scheme 1 of the experimental
section.
Please, insert Figure 1 here.
Compounds (1-10) were prepared by reacting commercially available aryl aldehydes with
semicarbazide or thiosemicarbazide. For the synthesis of phenoxymethyl thiosemicarbazones (11–17),
the substituted phenolic compound reacted with bromoacetaldehyde diethyl acetal under basic
conditions. After hydrolysis of acetal in aldehyde, the respective aldehyde was then reacted with
thiosemicarbazide and catalytic HCl in an ultrasound bath at room temperature. For the synthesis of 18,
the substituted phenolic compound reacted with 3-chloro-2-butanone. The resulting β-ketoether was
then reacted with thiosemicarbazide, to produce the compound 18.
All compounds (1–18, see Fig. 1) were purified by recrystallization and obtained at an
acceptable purity (> 95 %) in yields ranging from 40 to 97 %. The structures were determined by
NMR, infrared spectroscopy, and high-resolution mass spectrometry.
The compounds investigated were represented as (R;SU;X), where (X) stands for the thio- or
semicarbazone group, (R) the substituent at the aryl or phenoxy ring, and (SU) the spacer unit -CH= or
-
2
OCH CH= between the X and R groups.

ACCEPTED MANUSCRIPT
2
.2. Larvicide activity
The compounds synthesized 1-17 were tested as larvicides against the L4 stage of A. aegypti.
Tween-80 was selected as the co-solvent because it had the best results for dissolving the compounds in
water and because of its low toxicity to larvae, according to the classification of Kramer, Schnell and
Nickerson (1983) (LC50 > 1%) [70]. The obtained LC50 values for 1-17 are shown in Table 1.
Please, insert Table 1 here.
According Chang and co-workers (2003) essential oils or vegetal extracts showing LC50
between 50 to 100 ppm in larvicide bioassays must be considered as active, while those exhibiting LC50
lower than 50 ppm may be considered as highly active [71]. If one uses this reference, compounds 1-5,
-1
1
1 and 12 should be considered inactive, compounds 6 (LC = 310.8 µmol L ) and 13 (LC = 287.2
50 50
-
1
µmol L ) should be considered active and compounds 7-10 (LC50 = 134.8, 138.3, 122.8 and 69.7 µmol
-1
-1
L ) and 14-17 (LC = 92.0, 89.9, 70.1 and 20.9 µmol L ) should be considered highly active.
5
0
2
.3. Structure-Active Relationship (SAR)
The semicarbazone compounds tested (1 and 2) showed low activity (LC > 861.8 µmol L ).
-1
5
0
When the carbonyl (C=O) was replaced by the thiocarbonyl (C=S) bond, this led to the corresponding
-1
bioisoster compounds (6 and 10) to exhibit an improved result LC50 = 310.8 µmol L and LC50 = 69.7
-1
µmol L , respectively. According to Beraldo, this pattern was previously observed for other
pharmacological activities of (thio)semicarbazones [72,73]. This change in larvicide activity indicates a
possible change in the magnitude of the intermolecular interactions (e.g. H-bond) with a biological
target owing to the chemical change C=O for C=S in the above-mentioned compound. Another
plausible hypothesis for explaining this change in the larvicide activity may be related to the higher

ACCEPTED MANUSCRIPT
stability of the thiopeptidic compared to the peptidic environment under attack by proteolytic enzymes
[
74].
Interesting results come from the comparison of the larvicide activities in compounds
containing different spacer units (SU) between the aryl-substituted and the thiosemicarbazone moieties.
For instance, the phenoxymethyl-thiosemicarbazone derivatives 15, 14 and 17 (LC50 = 89.9, 92.0 and
-1
2
0.9 µmol L , respectively) where SU = -O-CH -CH= have a higher activity than the corresponding
2
-1
aryl-thiosemicarbazone derivatives 7, 9, and 10 (LC50 = 134.8, 122.8 and 69.7 µmol L , respectively)
where SU = -CH-. These results demonstrate the importance of the spacer unit -O-CH -CH= for the
2
compounds with larvicide activity for A. aegypti.
The satisfactory results for the activity of the thiosemicarbazones 7, 9, 14, and 15 (LC50
=
-1
1
34.8, 122.8, 92.0 and 89.9 µmol L , respectively) points out the importance of polarizable halogen
substituents at the para position for the larvicide activity. On the other hand the thiosemicarbazones
-1
-1
with substituents 4-F (6, LC = 310.8 µmol L ) and 3-Cl (13, LC = 287.2 µmol L ) performed
5
0
50
slightly worse. According Hernandes and co-workers [75], bulk and polarizable halogen atoms both
may occupy available pockets and interact via a halogen bond at the binding site of biological targets.
Thus, the relative smaller activity of 6 may be explained in terms of the smaller atomic radius and
polarizability of the substituent fluorine. Introduction of bulk halogen atoms in the orto and meta
positions tends to create steric hindrances and conformational changes. Perhaps, this also may explain
the relatively low activity of 13. It is interesting to observe that the order of activity for the para-
substituents in 6, 7 and 9, i.e., F < Cl < Br, was the same observed for AOPA by Neves Filho and co-
workers [54].
The presence of two atoms of chlorine at the meta and para positions of the aromatic ring led to
-
a significant increase in the larvicide activity (compounds 10 and 17 with LC = 69.7 and 20.9 µmol L
5
0
1
,
respectively) in comparison to their corresponding mono-chlorine compounds (7 and 15 with LC =
50

ACCEPTED MANUSCRIPT
-1
1
34.8 and 89.9 µmol L , respectively). In fact, these double-chlorine compounds are the most active
thiosemicarbazone derivatives synthesized in this work. Corroborating our results, Bordas and co-
worker, studying larvicide activity in a series of biurets, also observed the highest activities for the
compound chlorine di-substituted in the meta- and -para positions [55].
On the other hand, the presence of the electron donor groups para-OH (4, LC > 1024.4 µmol
5
0
-1
-1
L ) and para-OCH (11, LC = 503.1 µmol L ) decreased larvicide activity. This effect is in
3
50
agreement with results of Simas and co-workers who reported less active phenylpropanoide containing
hydroxyl and methoxyl groups in the para position on the aromatic ring [76]. The presence of a
methoxyl group attached to the para position on the aromatic ring did not seem to contribute to the
larvicide activity in the works of Neves Filho and co-workers [54] and Bordas and co-workers [55].
It has been reported that the presence of a phenylic ring conjugated to the α and β unsaturated
carbonyl groups in a cynanic aldehyde improves larvicide activity [76]. Because of this, a closely
related system 8 was synthesized. In fact, this activity was among the most successful (LC50 = 138.3
-1
µmol L ). However, it is important to note that in 8 the aromatic ring is conjugated not to a carbonyl
group, but to an imine function through an allylic system. Since this allylic system favors electron
delocalization, an increase in the electronic density is expected to occur on the nitrogen and sulfur
atoms with implications for the ability of these atoms to bind with enzyme metallic ions or to interact
via H-bond with residues in a biological target [72]. Nitro-compounds are well established bioactive
compounds used as antimicrobial, antiparasitic and antitumor agents [77]. The presence of the nitro
2
group (-NO ), however, seems not have contributed to the mortality of A. aegypti’s larvae since
-1
compound 3 showed a LC > 659.8 µmol L , whereas compound 9 (that does not contain the nitro
5
0
-1
group) showed a high activity with LC50 = 122.8 µmol L . One possiblility that cannot be neglected is
that the nitro group could be bio-reduced to one amino group [78] which, as an electron-donar group, is
thought to lessen larvicide activity.

ACCEPTED MANUSCRIPT
2
.4. Cytotoxicity Analysis
The evaluation of eventual side effects on non-target organisms from a newly synthesized
compound developed for larvicide purposes is essential, even in the beginning steps of the research. In
this work the synthesized thiosemicarbazone derivatives were submitted to cytotoxicity assays using
BALB/c mice splenocytes (as described in the experimental section) as the model mammal. In the last
column of Table 1, the highest non-toxic concentrations against the tested cells are shown. Fortunately,
when compared with saponin (the positive control), the 4-Cl-, 4-Br- and 3,4-diCl-phenoxymethyl-
thiosemicarbazones (top larvicide compounds) are non-toxic.
2
.5. QSAR
B3LYP/6-311++G(d,p) calculations to acess geometric, energetic and electronic properties for
compounds (1-18) have been reported by us elsewhere [79]. Table 2 shows the data matrix used to
obtain the quantitative structure-activity larvicide relationship (QSAR).
Please, insert Table 2 here.
The QSAR model obtained (multiple linear regression) is shown in Equation-1.
2
n=10; R = 0.97; R = 0.94; F = 54.61; s = 0.12; p= 0.00005
Equation-1.
Using this equation, Fig. 2.a shows a comparison between predicted and observed larvicide
activities. The quality of the adjustment (Equation-1) can be visualized in the residue plot as shown in

ACCEPTED MANUSCRIPT
Fig. 2.b. As one can see, the predicted and the observed activity values are highly correlated. The
quality of this regression can be appreciated, considering the equation parameter deviation, the
2
statistics parameters R and F as well as the distribution of the points around the zero in the residue
plot.
Please, insert Figure 2 here.
From Equation-1, one can see that the electric dipole moment (µ) and the decimal logarithm of the
octanol-water partition coefficient (log P), with negative and positive coefficients, respectively, are the
most important descriptors related to larvicide activity. This linear model indicates that substituents that
decrease µ (since it is always ≥0) and/or increase logP will improve the larvicide activity. These
equation features match the importance, previously observed in the literature, of both hydrophobic and
electronic parameters for describing larvicide activity against A. aegypti.
In order to check the QSAR model (Equation 1) a compound exhibiting a better activity was
predicted. Besides, some compounds showing intermediate activity and also one of the worst activities
(
the model must work inside and outside the training set as well as in both directions) were analyzed.
The compound with the higher activity, 18, was purchased from our particular collection of compounds
and tested after Equation-1 was obtained. The compound with intermediate activity, 13, was
synthesized and tested at the same time as those used in the training set. However this compound was
not used to build the model because it is the only one mono-substituted at the meta position. Finally,
compound 4, showing one of the worst results among the compounds shown in Table 1, was analyzed
in light of that QSAR equation. The calculated electric dipole moment, logP, and the predicted and
observed LC values of these compounds are shown in Table 3.
5
0
Please, insert Table 3 here.

ACCEPTED MANUSCRIPT
The data in this table show that Equation-1 succeeded in predicting correctly the best larvicide
activity for 18 and the intermediate larvicide activity for 13. The cytotoxicity of 18 was evaluated in
-1
1
6.33 µmol L , leading to a ratio (Cytotocity/LC50) equal to 4.1. It is interesting to stress that we
-1
stopped determining the experimental LC for compound 4 after the concentration of 1024 µmol L .
5
0
That decision is adequately supported by Equation-1 since the LC50 for the compound 4 is predicted at
-1
6
434 µmol L .
2
.6. Action Mode
In the attempt to improve our comprehension about how the synthesized compounds lead to
larvae death in the L4 stage, the A. aegypti sterol carrier protein-2 (AeSCP-2) inhibition was tested.
Since mosquitoes depend on exogenous sources of cholesterol for biosynthesis of steroid derivatives, it
is not surprising to find the high expression of AeSCP-2 in the larvae midgut during the feeding stage
[
80]. Therefore, compounds that can inhibit this protein have a high potential for becoming useful tool
for vector control. In 2003, Lan and co-workers published a high resolution X-ray structure of palmitic
acid (CH (CH ) COOH) co-crystallized into AeSCP-2 [81]. These authors found that the polar head of
3
2 14
the palmit acid makes an H-bond interaction with the side chain of the Arg24 residue, whereas the
metilenic moiety is in a bent conformation inside a hydrophobic pocket. Taking into account the
structural features of palmitic acid for binding AeSCP-2, two aspects require attention. First, the side
chain of Arg (-(CH ) -NH(C=NH)NH ) has an imidourea group at the end portion which is functionally
2
3
2
very similar to the (thio)semicarbazone portion of compounds 1-17. Second, the twisted disposition of
the phenoxymethyl group relative to the thiosemicarbazone moiety resembles the bent conformation of
the fatty portion relative to the carboxylic head of the palmit acid co-crystallized into AeSCP-2.

ACCEPTED MANUSCRIPT
Because of this, we tested two synthetized thiosemicarbazone derivatives, 4 and 11, as AeSCP-2
inhibitors (SCPIs) and compared them to a previous tested compound, SCPI-1 [69] (see Fig. 3).
Please, insert Figure 3 here.
Figure 3 shows that the competitive binding of compounds 4 and 11 to rAeSCP-2 had a similar
dose-response curve as that of the SCPI-1 [69]. The 50% effective concentration (EC ) of inhibiting
5
0
NBD-cholesterol binding to AeSCP-2 was 5.0 (95% Confidence Interval=2.0~17.0) and 0.6 (95%
Confidence Interval=0.3~1.4) µM for compounds 4 and 11, respectively. The EC of compounds 4 and
5
0
1
1 are within the range of identified SCPIs [69]. SCPIs have been shown to suppress dietary
cholesterol uptake in both A. Aegypti [82] and in the tobacco hornworm [83]. Therefore, the likely
mode of action of compounds 4 and 11 was the suppression of dietary cholesterol uptake in treated A.
aegypti larvae. Although Fig. 3 was fitted to the inhibition of SCP-2 whereas the response function on
the QSAR model measured the capacity of whole larvae to die (therefore they are conceptually
different things), the higher hydrophobicity of 11 (logPcalc = 2.09, Table 2) compared to 4 (logPcalc
=
0
.56, Table 3) matches the higher inhibitory activity of 11 compared to 4.
2
.7. Docking studies
In order to improve our comprehension about how the (thio)semicarbazone derivatives interact
with SCP-2, we conducted a docking study using as a binding site the palmitic acid contact residues in
SCP-2. Fig. 4 shows the superimposition of the best docking solutions obtained for compounds 4, 11,
1
1
4, 15, 16, 17 on the structure of the AeSCP-2 target. The Docking score values for molecules 4, 11,
4, 15, 16 and 17 are, respectively, 63.5, 64.9, 67.0, 65.3, 69.0 and 70.8.
Please, insert Figure 4 here.

ACCEPTED MANUSCRIPT
In order to elucidate the binding mode of these molecules into AeSCP-2, an investigation of the
intermolecular interactions was conducted. One can see in Fig. 5a the important residues mainly
involved in the interactions between AeSCP-2 and molecule 11. Figure 5a provides a clear illustration
of the binding pattern.
Please, insert Figure 5 here.
It is important to emphasize an important π−π stacking interaction that was found between the
side-chain ring of the PHE105 residue and practically all the ligands investigated. Three examples can
be seen in Fig. 5.b, particularly for molecules 4, 11 and 17. The Docking protocol considered the active
flexibility of residue PHE105 and other nine residues during the calculations. Figure 5.b shows the
alternative conformations adopted by PHE105 in order to establish and stabilize the π−π stacking
interaction with the respective ligand, searching for a maximum of planarity between the two aromatic
rings (PHE105 and ligand). Every other molecule studied binds to AeSCP-2 in a very similar way, in
comparison to molecule 11, as one can see in the panoramic view of all the Docking solutions
presented in Fig. 4. Furthermore, the important residues of AeSCP-2 involved in interactions (polar and
hydrophobic) with the docked ligands are practically the same residues involved in the interaction with
palmitic acid (ARG24, GLN25, LEU102 and PHE105) in the X-Ray structure, demonstrating that the
choice of the flexible side chains for the Docking calculations was pertinent.
Finally, in order to compare the in silico results against larvicidal activities, the LC50 values
were first converted into pLC50 (equals −log LC50, at molar concentration). The plot on Fig. 6 shows
the trend observed between the in silico Docking scores and the pLC50 data. This indicates that most of
the larvicidal compounds (those with the highest values for pLC50) are usually those with the higher

ACCEPTED MANUSCRIPT
docking scores, demonstrating that the molecules with more stable or positive Docking scores (i.e.,
greater in silico affinity for the AeSCP-2 target) are also the most active larvicidal compounds (i.e.,
greater pLC50 values). This kind of trend between in vitro and in silico data, showing the corroboration
among experimental and theoretical results, was also found in other studies undertaken by our group
[
84-86].
Please, insert Figure 6 here.
Additionally, the in vitro values (EC50) for the inhibition of NBD-cholesterol binding to
-1
-1
rAeSCP-2 protein (see Fig. 3), available for molecules 4 and 11, are 5.0 µmolL and 0.6 µmolL ,
respectively. These results also corroborate with the Docking score values for the same two molecules
in the AeSCP-2 target, which are 63.5 e 64.9, respectively. Between these two compounds, molecule
1
1 presents the highest in vitro inhibition potency, and has also the greater in silico affinity (high
Docking score) for the AeSCP-2 target, by showing a trend among in silico and in vitro results,
indicating that the molecules with more stable or positive Docking scores (i.e., greater in silico affinity
for the AeSCP-2 target) are also the most active larvicidal compounds (i.e., greater in vitro pLC50
values).
3
. Conclusion
Seventeen compounds belonging to the two classes of aryl-(thio)semicarbazone and
phenoxymethyl-thiosemicarbazones derivatives were synthetized and tested against the L4 stage for A.
aegypti. In general thiosemicarbazone exhibited an improved larvicide activity compared to the
corresponding semicarbazone compounds. For the thiosemicarbazones, those with the phenoxymethyl

ACCEPTED MANUSCRIPT
group showed a higher larvicide activity. As desirable, the subset formed by eleven aryl- and
phenoxymethyl-thiosemicarbazone derivatives showed a large variance of larvicide activity, with the
-1
-1
LC50 ranging from 21 µmol L to 311 µmol L . A QSAR equation from the LC50 against the electric
dipole moment (µ) and the logarithm of the partition coefficient (logP) was obtained using the multiple
linear regression technique. Using this equation we were doubly successful. First, in predicting the
larvicide activity of two compounds not used to create the model, with the more active one, 18, at 4.1
µM. Second, it was possible to explain why deciding to stop the analysis on the experimental LC for
5
0
-1
the compound 4 was correct - the predicted LC50 6434 µ mol L , is too high. The dose-response profile
observed for the compounds 4 and 11 on the inhibition of AeSCP-2 are supported by the QSAR model.
The docking calculations corroborate the hypothesis of the (thio)semicarbazone derivatives acting
through the inhibition of the AeSCP-2 target.
4
. Experimental
4
.1. General Chemistry
The arylhydrazone (1-10) were prepared essentially as reported previously [87,88] from
commercially available aldehydes. Only compound (3) a nitro aryl-thiosemiccarbazone, the
corresponding start aldehyde was obtained by reaction between 4-bromobenzaldeyde and nitric acid.
Compound (5), bis-thiosemicarbazone derivative, was prepared using 2 mols of thiosemicarbazide for
the 1.3-dicarbaldeyde. These reactions proceeded well upon refluxing (3-5 h.) with ethanol as a solvent,
a rate of 65 to 97% being observed overall.
For the synthesis of phenoxymethyl-thiosemicarbazones derivatives (11-17), the aldehydes were
obtained from different phenols. The step process involved ether formation and acid hydrolysis of the
acetal intermediary with moderate yields (36-56%). To accomplish the synthesis of 3-phenoxybutan-2-
ones derivatives (18-20), start 3-phenoxybutan-2-ones intermediates were obtained by a reaction

ACCEPTED MANUSCRIPT
between 4-tert-butylphenol and 3-chloro-2-butanone using potassium carbonate and potassium iodide
(
Scheme 1).
Please, insert Scheme 1 here.
Reagents were purchased from Acros Organics, Fluka, Sigma-Aldrich or Vetec and solvents
from Vetec or Dinâmica. The deuterated solvents (DMSO-d , CDCl , D O) were supplied from CIL
6
3
2
(
Tédia Brazil). The reactions were monitored in thin layer chromatography (TLC) using silica gel 60
containing a fluorescent indicator F254. The chromatographic plates were visualized under UV light (at
dual wavelength 365 or 254 nm). Melting points were measured using a Thomas Hoover capillary
1
13
instrument and the values were not subsequently corrected. The H and C NMR were performed for all
1
compounds, DEPT analysis as well as the addition of D
2
O for locating NH signals in the H NMR were
1
13
determined. The H and C NMR spectra were obtained using Unity Plus model Varian instruments
1
13
1
13
(
400 MHz for H, C for 100 MHz) or Bruker AMX (300 MHz to 75.5 MHz for H and C), using
1
tetramethylsilane as the internal standard. The number of signals in the H NMR spectra was designated
as follows: s / singlet; / d doublet, t / triplet, dd / double doublet, q / quartet, m / multiplet. Infrared
spectroscopy was performed with a Bruker instrument (model IFS 66) using KBr pellets. The Elemental
Analysis was performed with a Carlo Erba instrument model E-1110 or Perkin Elmer 2400 seriesii.
High-resolution electrospray ionization mass spectra (HRESIMS) were acquired on a nanoUPLC-Xevo
G2 Tof (Waters) in the positive ionization mode.
4
.2. Synthesis of compounds (1-10). Example for compound (1). In a round bottom flask for 100 mL, 4-
fluorobenzaldehyde (2.5 mmol) was dissolved in ethanol (15mL), then HCl (03 drops) were added to
the reaction at room temperature. Semicarbazide hydrochloride (2.5 mmol) was added and the mixture
was maintained under magnetic stirring for 3 h at room temperature. After this time, the mixture was
o
cooled at 0 C and the precipitate was filtered in a Büchner funnel with a sintered disc filter, washed

ACCEPTED MANUSCRIPT
with cold water, n-hexane and then dried over SiO . Compounds were recrystallized in hot ethanol, to
2
provide compounds with acceptable purity.
o
4
.2.1. 4-Fluorobenzaldehyde semicarbazone (1). Colorless crystals, yield = 84 %; mp ( C): 230-232; IR
-1 1
(
KBr): 3463 and 3275 (NH
2
), 3064 (NH), 1708 (C=O), 1591 (C=N) cm . H NMR (300 MHz, DMSO-
d ): δ 6.51 (s, 2H, NH ), 7.20 (m, 2H, Ar-H), 7.77 (m, 2H, Ar-H), 7.82 (s, 1H, CH=N), 10.24 (s, 1H,
6
2
1
3
NH). C NMR (75.5 MHz, ppm, DMSO-d ): δ 164.1 (C Ar), 156.8 (C=O), 138.1 (CH=N), 131.4 (C
6
Ar), 128.6 (CH Ar), 115.5 (CH Ar). Anal. Calcd for C H FN O: C, 53.04; H, 4.45; N, 23.19; Found: C,
8
8
3
+
5
3.06; H 4.56; N 22.39. HRESIMS m/z: 180.0590 [M-H] .
o
4
.2.2. 3,4-Dichlorobenzaldehyde semicarbazone (2). Colorless crystals, yield = 88 %; mp ( C): 246-249;
-1 1
2
IR (KBr): 3465 and 3279 (NH ), 3155 (NH), 1700 (C=O), 1588 (C=N) cm . H NMR (300 MHz,
DMSO-d ): δ 6.66 (s, 2H, NH ), 7.61 (d, 1H, J = 8.3 Hz, Ar-H), 7.66 (d, 1H, J = 8.3 Hz, Ar-H), 7.78
6
2
1
3
(
(
s, 1H, Ar-H), 8.12 (s, 1H, CH=N), 10.42 (s, 1H, NH). C NMR (75.5 MHz, ppm, DMSO-d ): δ 156.6
6
C=O), 136.4 (CH=N), 135.7 (C Ar), 131.6 (C Ar), 130.9 (C Ar), 130.6 (CH Ar), 127.6 (CH Ar),
1
1
26.8 (CH Ar). Anal. Calcd for C H Cl N O: C, 41.41; H, 3.04; N, 18.11; Found: C, 41.61; H, 3.23; N,
8 7 2 3
+
7.48. HRESIMS m/z: 231.9981 [M] .
o
4
.2.3. 4-Bromo-3-nitrobenzaldehyde thiosemicarbazone (3).Yellowish crystals
,
yield = 83 %; mp ( C):
-1 1
2
38-240; IR (KBr): 3417 and 3258 (NH ), 3156 (NH), 1519 (C=N) cm . H NMR (300 MHz, DMSO-
2
d6):δ 7.91 (d, 1H, J = 8.0 Hz, Ar-H), 7.93 (d, 1H, J = 8.0 Hz, Ar-H), 8.04 (s, 1H, CH=N), 8.28 (s, 1H,
13
NH ), 8.34 (s, 1H, NH ), 8.56 (s,1H, Ar-H), 11.65 (s, 1H, NH). C NMR (75.5 MHz, ppm, DMSO-d ):
2
2
6
δ 178.3 (C=S), 150.4 (C Ar), 138.5 (CH=N), 135.6 (C Ar), 134.6 (CH Ar), 131.9 (CH Ar), 122.4 (CH
Ar), 112.9 (C Ar). Anal. Calcd for C H BrN O S: C, 31.70; H, 2.33; N, 18.48; Found: C, 30.29; H,
8
7
4
2
+
2
.30; N, 16.76. HRESIMS m/z: 303.9576 [M+H] .

ACCEPTED MANUSCRIPT
o
4
2
6
8
1
.2.4. 4-Hydroxybenzaldehyde thiosemicarbazone (4). Brownish crystal, yield = 84 %; mp ( C): 221-
-1 1
23; IR (KBr): 3467 and 3359 (NH
2
), 3129 (NH), 1509 (C=N) cm . H NMR (300 MHz, DMSO-d
), 7.94 (s, 1H, CH=N),
), 9.90 (s, 1H, OH), 11.25 (s, 1H, NH). C NMR (75.5 MHz, ppm, DMSO-d ): δ
77.4 (C=S), 159.2 (C Ar), 142.7 (CH=N), 129.0 (CH Ar), 125.1 (C Ar), 115.5 (CH Ar). Anal. Calcd
6
): δ
.77 (d, 2H, J = 8.3 Hz, Ar-H), 7.60 (d, 2H, J = 8.3 Hz, Ar-H), 7.84 (s, 1H, NH
2
1
3
.07 (s, 1H, NH
2
6
for C H N OS: C, 49.22; H, 4.65; N, 21.52; Found: C, 49.20; H, 4.68; N, 20.61. HRESIMS m/z:
8
9
3
+
1
94.0405 [M-H] .
4
.2.5. 2,2'-(1,3-Phenylenebis(methanylylidene))bis(thiosemicarbazide) (5). Colorless crystals, yield = 65
o
-1 1
%
; mp ( C): 255-258; IR (KBr): 3423 and 3235 (NH
2
), 3148 (NH), 1524 (C=N) cm . H NMR (300
MHz, DMSO-d
6
): δ 7.42 (t, 1H, J = 7.7 Hz, Ar-H), 7.79 (d, 2H, J = 7.7 Hz, Ar-H), 8.05 (s, 2H, CH=N),
1
3
8
.10 (s broad, 2H, NH ), 8.22 (s, 1H, Ar-H), 8.26 (s, 2H, NH ), 11.55 (s, 2H, NH). C NMR (75.5
2
2
MHz, ppm, DMSO-d ): δ 178.0 (C=S), 141.6 (CH=N), 134.7 (C Ar), 128.9 (CH Ar), 128.6 (CH Ar),
6
1
2
25.4 (CH Ar). Anal. Calcd for C H N S : C, 42.84; H, 4.31; N, 29.98; Found: C, 36.38; H, 5.11; N,
10 12 6 2
+
3.68. HRESIMS m/z: 281.0589 [M+H] .
o
4
.2.6. 4-Fluorobenzaldehyde thiosemicarbazone (6). Beige crystals, yield = 97 %; mp ( C): 189-191;
-
1 1
IR (KBr): 3391 and 3235 (NH
2
), 3156 (NH), 1533 (C=N) cm . H NMR (300 MHz, DMSO-d
), 8.20 (s d, 1H, NH
s, 1H, NH). C NMR (75.5 MHz, ppm, DMSO-d ): δ 177.9 (C=S), 164.6 (C Ar), 141.0 (CH=N),
6
): δ 7.23
(
(
m, 2H, Ar-H), 7.87 (m, 2H, Ar-H), 8.02 (s, 1H, CH=N), 8.03 (s, 1H, NH
2
2
), 11.43
1
3
6
1
30.8 (C Ar), 129.4 (CH Ar), 115.7 (CH Ar). Anal. Calcd for C H FN S: C, 48.72; H, 4.09; N, 21.31;
8 8 3
+
Found: C, 47.90; H, 4.21; N, 20.08. HRESIMS m/z: 198.0450 [M+H] .

ACCEPTED MANUSCRIPT
o
4
.2.7. 4-Chlorobenzaldehyde thiosemicarbazone (7). Colorless crystals, yield: 77%; mp ( C): 217-220;
-1 1
IR (KBr): 3435 and 3279 (NH ), 3164 (NH), 1523 (C=N) cm . H NMR (300 MHz, DMSO-d ): δ 7.45
2
6
(
(
(
d, 2H, J = 8.3 Hz, Ar-H), 7.83 (d, 2H, J = 8.3 Hz, Ar-H), 8.01 (s, 1H, CH=N), 8.08 (s, 1H, NH
2
), 8.25
1
3
s, 1H, NH
2
), 11.49 (s, 1H, NH). C NMR (75.5 MHz, ppm, DMSO-d ): δ 178.0 (C=S), 140.8
6
CH=N), 134.2 (C Ar), 133.1 (C Ar), 128.9 (CH Ar), 128.6 (CH Ar). Anal. Calcd for C
8
H
8
3
ClN S: C,
+
4
4.97; H, 3.77; N, 19.67; Found: C, 45.30; H, 3.96; N, 19.12. HRESIMS m/z: 214.0140 [M+H] .
o
4
.2.8. 2-(3-Phenylallylidene)thiosemicarbazide (8). Yellowish crystals, yield = 90%; mp ( C):110-113.
-1 1
IR (KBr): 3418 and 3260 (NH ), 3155 (NH), 1537 (C=N)cm . H NMR (300 MHz, DMSO-d ): δ 6.86
2
6
(
dd, 1H, J = 8.9 Hz; J = 15.8 Hz, CH=CH), 7.02 (d, 1H, J = 15.89 Hz, CH=CH), 7.43-7.24 (m, 3H, Ar-
H), 7.55 (d, 2H, J = 7.79 Hz, Ar-H), 7.61 (s, 1H, NH ), 7.89 (d, 1H, J = 8.9 Hz, CH=N), 8.17 (s, 1H,
), 11.40 (s, 1H, NH). C NMR (75.5 MHz, ppm, DMSO-d ): δ 177.6 (C=S), 144.7 (CH=N), 138.8
CH=N), 135.8 (C Ar), 128.8 (CH Ar), 126.9 (CH Ar), 125.0 (CH Ar). Anal. Calcd for C H N S: C,
2
13
NH
2
6
(
1
0
11
3
+
5
8.51; H, 5.40; N, 20.47; Found: C, 53.93; H, 6.20; N, 17.65. HRESIMS m/z: 206.0749 [M+H] .
o
4
.2.9. 4-Bromobenzaldehyde thiosemicarbazone (9). Beige crystals, yield = 82 %; mp ( C): 209-211, IR
-1 1
(
KBr): 3436 and 3287 (NH ), 3165 (NH), 1522 (C=N) cm . H NMR (300 MHz, DMSO-d ): δ 7.58 (d,
2 6
2
H, J = 8.3 Hz, Ar-H), 7.76 (d, 2H, J = 8.3 Hz, Ar-H), 8.00 (s, 1H, CH=N), 8.08 (s, 1H, NH
2
), 8.24 (s
1
3
broad, 1H, NH
2
), 11.49 (s, 1H, NH). C NMR (75.5 MHz, ppm, DMSO-d
6
): δ 178.1 (C=S), 140.9
(
CH=N), 134.5 (C Ar), 131.5 (CH Ar), 129.1 (CH Ar), 122.9 (C Ar). Anal. Calcd for C
8
H
8
3
BrN S: C,
+
3
7.22; H, 3.12; N, 16.28; Found: C, 57.32; H, 3.24; N, 15.78. HRESIMS m/z: 259.9630 [M+H] .
o
4
2
7
.2.10. 3,4-Dichlorobenzaldehyde thiosemicarbazone (10). White crystals, yield = 62 %; mp ( C): 212-
-1 1
15. IR (KBr): 3396 and 3255 (NH ), 3154 (NH), 1539 (C=N)cm . H NMR (300 MHz, DMSO-d ): δ
2
6
.63 (d, 1H, J = 8.3 Hz, Ar-H), 7.71 (dd, 1H, J = 1.7 Hz, J = 8.3 Hz, Ar-H), 7.98 (s, 1H, CH=N), 8.24

ACCEPTED MANUSCRIPT
1
3
(
d, 1H, J = 1.7 Hz, Ar-H), 8.27 (s, 1H, NH ), 8.30 (s, 1H, NH ), 11.57 (s, 1H, NH). C NMR (75.5
2 2
MHz, ppm, DMSO-d ): δ 178.2 (C=S), 139.3 (CH=N), 135.0 (C Ar), 131.8 (C Ar), 131.7 (C Ar),
6
1
1
30.7 (CH Ar), 128.1 (CH Ar), 127.7 (CH Ar). Anal. Calcd for C
8 7 2 3
H Cl N S: C, 38.73; H, 2.84; N,
6.94; Found: C, 39.09; H, 3.03; N, 16.41. HRESIMS m/z: 247.9636 [M-H]+.
4
.3. Synthesis of compounds (11-17). Example for compound (11). In a round bottom flask, phenol (3.1
mmol), 5 mL DMF and K CO (7.8 mmol) were added together. The reaction mixture was maintained
2
3
under magnetic stirring at room temperature for 30 min. Then, bromoacetaldehyde diethylacetal (9.3
mmol) was added in portions and the reaction mixture was heated under reflux for 72 h. After that, the
product was extracted with dichloromethane and the solvent was removed under reduced pressure and
2 2 4
dried in SiO . The hydrolysis of acetal in aldehyde was achieved by adding acetone (5mL), H SO (7
drops) and 10 mL water. The reaction mixture was stirred under reflux heating for 4 h. The product was
extracted with ethyl acetate and the solvent was removed under reduced pressure and then dried in
SiO . The aldehyde obtained was reacted with thiosemicarbazide as described above. Products were
2
purified by recrystallization using ethanol/water (1:1).
4
.3.1. 2-(4-Methoxyphenoxy)acetaldehyde thiosemicarbazone (11). Brownish crystals, yield = 56 %;
o
-1 1
mp ( C): 145-147; IR (KBr): 3372 and 3279 (NH
DMSO-d ): δ 3.68 (s, 3H, OCH ), 4.59 (s, 2H, CH
H, NH ), 8.17 (s, 1H, NH ), 11.33 (s, 1H, NH). C NMR (100 MHz, DMSO-d ): δ 55.3 (CH ), 67.3
2
), 3174 (NH), 1509 (C=N) cm . H NMR (400 MHz,
6
3
2
), 6.88 (m, 4H, Ar-H), 7.51 (s, 1H, CH=N), 7.68 (s,
1
3
1
2
2
6
3
(
CH ), 114.6 (CH Ar), 115.7 (CH Ar), 141.4 (CH=N), 151.8 (C Ar), 153.73 (C Ar), 178.3 (C=S). Anal.
2
Calcd. For C H N O S: C, 50.19; H, 5.48; N, 17.56; Found: C, 50.03; H, 5.28; N, 17.34. HRESIMS
1
0
13
3
2
+
m/z: 240.3020 [M+H] .

ACCEPTED MANUSCRIPT
o
4
.3.2. 2-Phenoxyacetaldehyde thiosemicarbazone (12). White crystals, yield = 40 %; mp ( C): 142-143;
-1 1
IR (KBr): 3449 and 3323 (NH ), 3158 (NH), 1536 (C=N) cm . H NMR (400 MHz, DMSO-d ): δ 4.67
2
6
(
d, 2H, J = 5.2 Hz, CH
2
), 6.95 (t, 1H, J = 7.5 Hz, Ar-H), 6.98 (d, 2H, J = 8.4 Hz, Ar-H), 7.29 (dd, 2H, J
), 8.18 (s broad,
H, NH ), 11.35 (s, 1H, NH). C NMR (100 MHz, DMSO-d ): δ 66.7 (CH ),114.6 (CH Ar), 121.0
=
7.5 Hz, J = 8.4 Hz, Ar-H), 7.53 (t, 1H, J = 5.2 Hz CH=N), 7.68 (s broad, 1H, NH
2
1
3
1
2
6
2
(
CH Ar), 129.5 (CH Ar), 141.0 (CH=N), 157.8 (C Ar), 178.3 (C=S). Anal. Calcd. For C H N OS: C,
9 11 3
+
5
1.66; H, 5.30; N, 20.08; Found: C, 51.58; H, 5.08; N, 19.76. HRESIMS m/z: 210.0621 [M+H] .
4
.3.3. 2-(3-Chlorophenoxy)acetaldehyde thiosemicarbazone (13). Beige crystals, yield = 40 %; mp
o
-1 1
(
C): 147-149; IR (KBr): 3406 and 3239 (NH
DMSO-d ): δ 4.69 (s, 2H, CH ), 7.08-6.85 (m, 3H, Ar-H), 7.31 (s, 1H, Ar-H), 7.50 (s, 1H, CH=N),
.69 (s, 1H, NH ), 8.20 (s, 1H, NH
CH ), 113.7 (CH Ar), 114.8 (CH Ar), 121.0 (CH Ar), 130.9 (CH Ar), 133.7 (C Ar), 140.4 (CH=N),
2
), 3156 (NH), 1513 (C=N) cm . H NMR (400 MHz,
6
2
1
3
7
2
2 6
), 11.37 (s, 1H, NH). C NMR (100 MHz, DMSO-d ): δ 67.1
(
2
1
4
58.8 (C Ar), 178.4 (C=S). Anal. Calcd. For C H ClN OS: C, 44.36; H, 4.14; N, 17.24; Found: C,
9 10 3
+
3.99; H, 4.04; N, 16.82. HRESIMS m/z: 244.0229 [M+H] .
o
4
1
d
8
1
.3.4. 2-(4-Bromophenoxy)acetaldehyde thiosemicarbazone (14). White crystals, yield = 45 %; mp ( C):
-1 1
66-168; IR (KBr): 3388 and 3261 (NH
): δ4.66 (s, 2H, CH ), 6.97 (s, 2H, Ar-H), 7.46 (s, 2H, Ar-H), 7.50 (s, 1H, CH=N), 7.69 (s, 1H, NH
.21 (s, 1H, NH ), 11.37 (s, 1H, NH). C NMR (100 MHz, DMSO-d ): δ 67.0 (CH ), 112.4 (C Ar),
2
), 3154 (NH), 1536 (C=N) cm . H NMR (400 MHz, DMSO-
6
2
2
),
1
3
2
6
2
17.0 (CH Ar), 132.1 (CH Ar), 140.4 (CH=N), 157.1 (C Ar), 178.3 (C=S). Anal. Calcd. For
C H BrN OS: C, 37.51; H, 3.50; N, 14.58; Found: C, 37.22; H, 3.76; N, 14.23. HRESIMS m/z:
9
10
3
+
2
88.2792 [M] .

ACCEPTED MANUSCRIPT
4
.3.5. 2-(4-Chlorophenoxy)acetaldehyde thiosemicarbazone (15). Beige crystals, yield = 52 %; mp
o
-1 1
(
C): 178-181; IR (KBr): 3402 and 3273 (NH ), 3152 (NH), 1532 (C=N) cm . H NMR (400 MHz,
2
DMSO-d
.50 (s, 1H, CH=N), 7.69 (s, 1H, NH
DMSO-d ): δ67.1 (CH ), 116.5 (CH Ar), 124.8 (C Ar), 129.3 (CH Ar), 140.5 (CH=N), 156.7 (C Ar),
6 2
):δ4.59 (s broad, 2H, CH ), 7.01 (d, 2H, J = 8.7 Hz, Ar-H), 7.33 (d, 2H, J = 8.7 Hz, Ar-H),
1
3
7
2 2
), 8.19 (s, 1H, NH ), 11.36 (s, 1H, NH). C NMR (100 MHz,
6
2
1
1
78.4 (C=S). Anal. Calcd. For C H ClN OS: C, 44.36; H, 4.14; N, 17.24; Found: C, 44.26; H, 3.98; N,
9
10
3
+
6.71. HRESIMS m/z: 244.0230[M + H] .
4
.3.6. 2-(2,3-Dichlorophenoxy)acetaldehyde thiosemicarbazone (16). Beige crystals, yield = 40 %; mp
o
-1 1
(
C): 189-192; IR (KBr): 3430 and 3251 (NH
DMSO-d ): δ 4.80 (s, 2H, CH ), 7.45-7.00 (m, 3H, Ar-H), 7.53 (s, 1H, CH=N), 7.71 (s, 1H, NH
s, 1H, NH ), 11.41 (s, 1H, NH). C NMR (100 MHz, DMSO-d
C Ar), 122.6 (CH Ar), 128.5 (CH Ar), 132.4 (C Ar), 139.7 (CH=N), 154.7 (C Ar), 178.5 (C=S). Anal.
2
), 3156 (NH), 1545 (C=N) cm . H NMR (400 MHz,
), 8.23
), 112.8 (CH Ar), 120.1
6
2
2
1
3
(
(
2
6 2
): δ 68.3 (CH
Calcd. For C H Cl N OS: C, 38.86; H, 3.26; N, 15.11; Found: C, 38.49; H, 3.44; N, 14.83. HRESIMS
9
9
2
3
+
m/z: 277.9840 [M] .
4
.3.7. 2-(3,4-Dichlorophenoxy)acetaldehyde thiosemicarbazone (17). White crystals, yield = 56 %; mp
o
-1 1
(
C): 169-172; IR (KBr): 3408 and 3264 (NH
DMSO-d ): δ 4.70 (d, 1H, J = 4.7 Hz, CH ), 7.02 (dd, 1H, J = 2.3 Hz, Ar-H, J = 9.1 Hz, Ar-H), 7.31 (d,
H, J = 2.3 Hz, Ar-H), 7.49 (t, 1H, J = 4.7 Hz, CH=N), 7.52 (d, 1H, J = 9.1 Hz, Ar-H), 7.68 (s, 1H,
2
), 3155 (NH), 1534 (C=N) cm . H NMR (400 MHz,
6
2
1
1
3
NH ), 8.22 (s, 1H, NH ), 11.38 (s, 1H, NH). C NMR (100 MHz, DMSO-d ): δ 67.4 (CH ), 115.6 (CH
2
2
6
2
Ar), 116.7 (CH Ar), 122.9 (C Ar), 131.0 (CH Ar), 131.6 (C Ar), 139.9 (CH=N), 157.3 (C Ar), 178.4
(
C=S). Anal. Calcd. For C H Cl N OS: C, 38.86; H, 3.26; N, 15.11; Found: C, 38.55; H, 3.52; N,
9 9 2 3
+
1
4.91. HRESIMS m/z: 277.9822 [M] .

ACCEPTED MANUSCRIPT
4
.4. Synthesis of compounds (18). 3-(3,4-dichloro)phenoxy)butan-2-one was obtained by reacting 3,4-
dichlorophenol (6.4 mmol, 1.0 g) with 3-chloro-2-butanone (6.66 mmol, 0.71 g) in potassium carbonate
9.98 mmol, 1.38 g), potassium iodide (0.66 mmol, 0.11 g) and 15 mL of acetone. This mixture was
(
maintained under magnetic stirring at room temperature for 12h. The precipitate was filtered in a
Büchner funnel with a sintered disc filter and discarded. The solvent was completely evaporated and
then was extracted first into diethyl ether and water and subsequently in diethyl ether and sodium
hydroxide to 0.1 M. The compound was then dried in a SiO glass dissector under vacuum. The 3-(3,4-
2
dichloro)phenoxy-butan-2-one (6.16 mmol, 1.36g) obtained was reacted with thiosemicarbazide (6.16
mmol, 0.56 g), 4 drops of hydrochloric acid and 10 mL of ethanol in a 150 mL round bottom flask
under magnetic stirring for 2 hours. A yellowish solid was obtained, filtered in Büchner funnel with a
2
sintered disc filter, washed with cold water, and then dried in SiO . The products were purified by
crystallization using ethanol as solvent.
4
7
1
5
.4.1. 3-(3,4-Dichlorophenoxy)butan-2-one thiosemicarbazone (18). Yellowish crystals yield= 1.37 g,
o
4.41 %. mp ( C): 154-156. IR (KBr): 3420 (N–H), 3259 and 3155 (NH
2
), 1593 (C=N), 1282 (C–O),
): δ 1.42 (d, J = 6.6 Hz, 3H, CH ), 1.82 (s, 3H, CH ),
.00 (q, 1H, J = 6.6 Hz, H–C), 7.00 (dd, 1H, J = 3.0 Hz, J = 9.0 Hz, Ar-H), 7.27 (d, 1H, J = 3.0 Hz, Ar-
-1 1
084 (C=S) cm . H NMR (300 MHz, DMSO-d
6
3
3
4
H), 7.47 (d, 1H, J = 9.0 Hz, Ar-H), 7.85 (s largo, 1H, NH ), 8.24 (s largo, 1H, NH ), 10.19 (s, 1H, NH).
2
2
1
3
C NMR (75.5 MHz, DMSO-d ): δ 11.08 (CH –C=N), 18.48 (CH –C–O), 77.12 (CH–O), 116.52 (CH,
6
3
3
Ar), 117.75 (CH, Ar), 122.94 (C–Cl, Ar), 130.99 (CH, Ar), 131.55 (C–Cl, Ar), 150.58 (C=N), 156.71
(
C–O, Ar), 179.30 (S=C–NH ). Anal. Calcd. for C H N OSCl : C, 43.15; H, 4.28; N, 13.72; Found: C,
2 11 13 3 2
+
4
3.39; H, 4.09; N, 13.69. HRESIMS m/z: 306.0320 [M+H] .

ACCEPTED MANUSCRIPT
4
.5. Synthesis of diethyl acetal Intermediate. In a round bottom flask, phenol (3.12 mmol) was added
to dry DMF and K CO (7.8 mmol). The mixture was kept under magnetic stirring at room temperature
2
3
for 30 min. Then bromoacetaldehyde diethyl acetal (9.36 mmol) was added and the brownish mixture
was kept under reflux heating for 72 hours. The reactions were monitored by thin-layer
chromatographic plate (TLC). After that, the product was extracted with dichloromethane and solvent
was removed under reduced pressure and then dried in SiO2.
4
.6. Synthesis of aldheydes. In a round bottom flask, acetal intermediate was mixed with acetone
(
5mL), after an acid solution (H SO 7 drops and water 10mL) had been slowly added. The mixture
2
4
o
was kept under reflux heating (100 C) for 4 hours. The reaction was monitored using a thin-layer
chromatographic plate (TLC). At the end, the product was extracted with ethyl acetate and the solvent
2
was removed under reduced pressure and then dried in SiO .
4
1
3
.7. Synthesis of intermediated compounds 3-phenoxybutan-2-ones intermediates to afford compound
8.
-(4-tertbutyl)phenoxy)butan-2-one:4-tert-butylphenol (6.66 mmol, 1.0 g), 3-chloro-2-butanone (6.66
mmol, 0.71 g), potassium carbonate (9.98 mmol, 1.38 g), potassium iodide (0.66 mmol, 0.11 g) and 15
mL of acetone were mixed in a 150 mL round bottom flask and placed under magnetic stirring for 12
hours. The precipitate was filtered in a Büchner funnel with a sintered disc filter and discarded. The
solvent was completely evaporated and then was extracted first into diethyl ether and water and
subsequently in diethyl ether and sodium hydroxide to 0.1 M. Compound was then dried in SiO
dissector under vacuum.
2
glass

ACCEPTED MANUSCRIPT
4
.8. QSAR
In this work we opted to employ a set of QSAR descriptors, named electronic and hydrophobic
descriptors, that our review of the literature (see introduction section) had pointed out as being
important for representing the larvicide activity of A. aegypti. In particular, we selected descriptors of
easy chemical interpretation and at the same time belonging to the ligand, such as: i) atomic charge or
sum of atomic charge, ii) the electric dipole moment, iii) HOMO, LUMO and HOMO-LUMO energy
2
difference, iv) Log P and its square value, LogP .
In order to obtain the QSAR model, Multiple Linear Regressions (MLR) [89] between the
ligand descriptors and the larvicide activity for A. aegypti were used due to the simplicity of their
interpretation To avoid collinearity problems between descriptors in MLR [90], before obtaining the
QSAR equation, the correlation coefficient matrix (related to the information contained in Table 2) was
determined. Only those descriptors that correlated to the larvicide activity above 0.7 and at the same
time did not correlate with each other over 0.5 were used in the MLR. In order to check the quality of
the MLR the cross-validation method was employed [91]. In all cases the Statistica program [92] was
employed.
4
.9. Docking studies for A. aegypti sterol carrier protein-2 (AeSCP-2)
The in vitro inhibition of AeSCP-2 was measured for molecules 4 and 11, and the other
molecules (14-17), shown in Table 1 and discussed above as highly active in larvicide bioassays,
provided a base for the selection of compounds (ligands) 4, 11, 14, 15, 16 and 17 for docking
calculations. The optimized structures of all the ligands were obtained by application of the RM1
method [93], available as part of the SPARTAN 08’ program [94], using internal default settings for
convergence criteria. The target structure for docking calculations and analysis was taken from Protein
Data Bank (http://www.pdb.org) under the PDB code 1PZ4 for A. aegypti sterol carrier protein-2
(
AeSCP-2) [81]. The active site was defined as all atoms within a radius of 6.0 Å from the co-

ACCEPTED MANUSCRIPT
crystallized ligand (palmitic acid, labeled as PLM in PDB). The concern to take into account the
Induced Fit effects led us to treat the side chains of ten residues as flexible during the docking
calculations, following current trends in this area. Residues ARG15, LEU16, ILE19, ASP20, ASN23,
ARG24, GLN25, LEU48, LEU102 and PHE105 were selected for the AeSCP-2 target. The CHEMPLP
score function [95] of the GOLD 5.1 program [96] was used for docking calculations, followed by the
Binana program [97], which was used to analyze the molecular interactions present in the best docking
solutions, using a default setting, except for H-bond distance, which was changed to a maximum of 3.5
Å. The figures were generated with Pymol [98].
4
.10. Larvicidal bioassay
The larvicidal activity of the thiosemicarbazones and semicarbazones was evaluated using an
adaptation [54,32] of the method recommended by the World Health Organization [99]. Stock solutions
were prepared by solubilising 5 mg of the compounds with the appropriated co-solvents (Tween80); the
resulting solution was then dissolved in 50 mL of distilled water. Dilution of the stock solutions
allowed the preparation of suitable concentrations to be tested. Fourth larvae stage A. aegypti were
added to beakers (20 larvae per beaker) containing these solutions (20 mL). Four replicate assays were
carried out for every sample concentration, and for each assay a negative control was included and
prepared as described without the active compounds. Mortality of the larvae was determined after 48 h
incubation at 28 ± 2°C, 70 ± 10 relative humidity. Larvae were considered dead when they did not
respond to stimulus or did not rise to the surface of the solution. The lethal concentration value LC50
was calculated by probit analysis using StatusPlus2006 software [32,33].
4
.11. Cytotoxicity to mouse splenocytes
BALB/c mouse splenocytes were placed into 96-well plates at a cell density of 6x10 cells/well
5
-1
in an RPMI-1640 medium supplemented with 10% of FBS and 50 µg mL of gentamycin. Each test

ACCEPTED MANUSCRIPT
-1
inhibitor was evaluated in six concentrations (1, 5, 10, 25, 50 and 100 µg mL ) in triplicate. To each
well, an aliquot of test inhibitor suspended in DMSO was added. Controls included wells only
containing either solvent (untreated cells) or saponin (positive control). The plate was incubated for 24
3
h at 37º C and 5% CO
2
. After incubation, 1.0 µCi of H-thymidine (Perkin Elmer, Waltham, USA) was
added to each well, and the plate was returned to the incubator. The plate was then transferred to a
3
beta-radiation counter (Multilabel Reader, Finland), and the percent of H-thymidine was determined.
3
Cell viability was measured as the percent of H-thymidine incorporation for treated-cells in
comparison to untreated cells. Highest non-toxic concentration for each compound was estimated.
4
.12. Inhibitory Cholesterol binding Test
Recombinant AeSCP-2 protein was purified and an NBD-cholesterol (Molecular probes,
Eugene, OR, USA)/rAeSCp-2/compound competition assay was performed as described [69]. A
separate set of tests were performed using NBD cholesterol with increasing concentration of a SCPI to
assess whether the SCPI interfered with NBD cholesterol fluorescence. If a compound interfered with
NBD cholesterol fluorescence, the background control was NBD cholesterol along with the SCPI. The
net change in NBD cholesterol fluorescence intensity was calculated by subtracting the fluorescence of
background controls from the NDB cholesterol/AeSCP-2 complex in the presence of a compound. The
data were plotted with the relative NBD cholesterol intensity (bound NBD cholesterol) as the Y-axis
and molarity of inhibitor as the X-axis using GraphPad Prism 4.0 (GraphPad Software Inc., San Diego,
CA).
AUTHOR INFORMATION
Corresponding Authors
*
8
e-mails: paraiso@ufpe.br and dmafn@ufpe.br. Tel.:(+55) 81 2126-7419 and 2126-7468. Fax: (+55)
1 2126-8442.