Synthesis and structural characterization of new benzylidene glycosides, cytotoxicity against cancer cell lines and molecular modeling studies

Article abstract of DOI:10.1016/j.molstruc.2021.130186

This work describes the synthesis, structural characterization (by combined Fourier Transform Infrared - FTIR, ¹H and ¹³C Nuclear Magnetic Resonance - NMR spectroscopy and High Resolution Mass Spectrometry - HRMS) and biological evaluation of a new series of glycosides designed from a benzylidene glucoside derived from eugenol (23) active against Candida glabrata. The mass accuracy between the calculated and found values observed in HRMS analyses were lower than 5 ppm, which are acceptable for proposing a molecular formula using this technique. We decided to keep the benzylidene group of 23, while changing either the saccharide unit (glucose or galactose) or the natural aglycone (eugenol, isoeugenol, dihydroeugenol or guaiacol) to check their influence in antifungal activity. Since the chemical modifications performed did not contribute to enhance the antifungal activity, the synthesized compounds (23–30) were further screened against four cancer cell lines (HeLa: cervix carcinoma; MDA-MB-231: breast carcinoma; T-24: urinary bladder carcinoma; and TOV-21G: ovarian carcinoma). The glucoside 27 showed promising activities (IC₅₀ 10.08–59.91 μM) against all the assayed cancer cell lines and higher values of selectivity index than doxorubicin, the control drug. The galactoside 28 demonstrated interesting results against HeLa, MDA-MB-231 and T-24 cells. This compound was active at 17.41 μM with a selectivity index greater than 13.7 against the HeLa cells, while doxorubicin was active at 10.01 μM with a selectivity index close to 1.5 considering this cell line. Further, we performed docking studies of these compounds with type II topoisomerase-DNA complex (TOP2) in order to try to explain their mechanism of action.

Full text of DOI:10.1016/j.molstruc.2021.130186

Journal of Molecular Structure 1233 (2021) 130186
Contents lists available at ScienceDirect
Journal of Molecular Structure
journal homepage: www.elsevier.com/locate/molstr
Synthesis and structural characterization of new benzylidene
glycosides, cytotoxicity against cancer cell lines and molecular
modeling studies
Vinícius Augusto Campos Péreta, Adriana Cotta Cardoso Reisa, Naiara Chaves Silvab,
Amanda Latercia Tranches Diasb, Diogo Teixeira Carvalhoc, Danielle Ferreira Diasd,
Saulo Fehelberg Pinto Bragaa, Geraldo Célio Brandãoa, Thiago Belarmino de Souzaa,∗
a School of Pharmacy - Federal University of Ouro Preto, 35400-000, Ouro Preto, MG, Brazil
b Institute of Biomedical Sciences, Federal University of Alfenas, 37130-001, Alfenas, MG, Brazil
c Pharmaceutical Sciences University, Federal University of Alfenas, 37130-001, Alfenas, MG, Brazil
d Institute of Chemistry, Federal University of Alfenas, 37130-001, Alfenas, MG, Brazil
a r t i c l e i n f o
a b s t r a c t
Article history:
This work describes the synthesis, structural characterization (by combined Fourier Transform Infrared -
FTIR, 1H and 13 C Nuclear Magnetic Resonance - NMR spectroscopy and High Resolution Mass Spectrom-
etry - HRMS) and biological evaluation of a new series of glycosides designed from a benzylidene gluco-
side derived from eugenol (23) active against Candida glabrata. The mass accuracy between the calculated
and found values observed in HRMS analyses were lower than 5 ppm, which are acceptable for propos-
ing a molecular formula using this technique. We decided to keep the benzylidene group of 23, while
changing either the saccharide unit (glucose or galactose) or the natural aglycone (eugenol, isoeugenol,
dihydroeugenol or guaiacol) to check their influence in antifungal activity. Since the chemical modifica-
tions performed did not contribute to enhance the antifungal activity, the synthesized compounds (23–
30) were further screened against four cancer cell lines (HeLa: cervix carcinoma; MDA-MB-231: breast
carcinoma; T-24: urinary bladder carcinoma; and TOV-21G: ovarian carcinoma). The glucoside 27 showed
promising activities (IC50 10.08–59.91 μM) against all the assayed cancer cell lines and higher values of
selectivity index than doxorubicin, the control drug. The galactoside 28 demonstrated interesting results
against HeLa, MDA-MB-231 and T-24 cells. This compound was active at 17.41 μM with a selectivity in-
dex greater than 13.7 against the HeLa cells, while doxorubicin was active at 10.01 μM with a selectivity
index close to 1.5 considering this cell line. Further, we performed docking studies of these compounds
with type II topoisomerase-DNA complex (TOP2) in order to try to explain their mechanism of action.
Received 14 October 2020
Revised 17 February 2021
Accepted 21 February 2021
Available online 23 February 2021
Keywords:
Eugenol analogues
Antifungal activity
Cytotoxic activity
Molecular docking
Topoisomerase II-DNA complex
© 2021 Elsevier B.V. All rights reserved.
1. Introduction
drugs provides a great treatment challenge since only three classes
of agents are currently employed to treat invasive fungal infections
It is estimated that over 1.6 million people die every year due
to serious fungal infections. This mortality rate is compared to that
of tuberculosis and is three times higher than that of malaria. To
a certain degree, this reflects an increase in the number of at-risk
patients, such as those living with HIV/AIDS, cancer or other im-
munodeficiency conditions as seen with organ transplanted indi-
viduals [1]. In addition, the scarce chemical diversity of antifungal
(polyenes, azoles and echinocandins), and resistance to these drugs
has been widely notified, as seen in the multidrug resistant strains
of Candida auris [2,3].
On the other hand, cancer is a group of diseases that can start
in almost any organ or tissue of the body due to the uncontrollable
growth of abnormal cells and represents the second leading cause
of death globally, accounting for an estimated 9.6 million deaths
(one in six deaths) in 2018. Additionally, the World Health Orga-
nization estimates that by the year 2030, 27 million cases of can-
cer can be expected. The greatest effect of this increase will be
on low- and middle-income countries [4]. Currently the available
anti-tumor treatments include surgical intervention, chemotherapy,
and radiation therapy or a combination of these options. Regarding
∗
Corresponding author at: Universidade Federal de Ouro Preto (Campus Morro
do Cruzeiro – Escola de Farmácia), 35400-000; Ouro Preto/MG; Brazil.
E-mail address: thiagobs83@yahoo.com.br (T.B. de Souza).
https://doi.org/10.1016/j.molstruc.2021.130186
0022-2860/© 2021 Elsevier B.V. All rights reserved.

V.A.C. Péret, A.C.C. Reis, N.C. Silva et al.
Journal of Molecular Structure 1233 (2021) 130186
galactose (2) were converted into the respective peracetylated
derivatives (3 and 4, respectively), from their reaction with acetic
anhydride. The peracetylated derivatives provided the correspond-
ing α-bromides (5 and 6) after reaction with hydrobromic acid in
acetic acid [11]. The reaction of the α-bromides and different phe-
nols (eugenol, isoeugenol, dihydroeugenol and guaiacol) afforded
the peracetylated glycosides 7–14, which by a methanolysis con-
dition led to the deacetylated glycosides 15–22 [10]. The glyco-
sides 7–22 were employed in the next synthetic step without char-
acterization. Finally, the reaction between the glycosides 15–22
with benzaldehyde, in the presence of zinc chloride, provided the
planned benzylidene derivatives 23–30 with yields in the range of
42 to 89% [12].
In the infrared spectra of derivatives 23–30 bands referring to
the axial deformation of the hydroxyl groups were observed be-
tween 3600 to 3300 cm−1. The bands related to the axial defor-
mation of the sp3 C H bonds were observed near 2900 cm−1. The
–
Fig. 1. Chemical structures of eugenol (I), isoeugenol (II), dihydroeugenol (III) and
guaiacol (IV).
=
axial deformations generated by aromatic C C bonds were regis-
tered between 1590 to 1370 cm−1
.
In the 1H NMR spectra of the derivatives 23–30, the signals cor-
responding to benzylidene hydrogens could be observed between
7.48–7.33 ppm as multiplets. The H-7 protons of acetal groups
were registered as singlets near 5.5 ppm for all the substances, and
were defined as a success indication for the introduction of this
group in the glycoside unit. For the eugenol derivatives (23 and
24), multiplets corresponding to the H-20 hydrogens (allyl group)
were registered near 5.9 ppm. For the isoeugenol derivatives 25
and 26, doublets corresponding to H-19 protons (propenyl group)
were recorded with coupling constants near 16 Hz, typical of trans
hydrogens. For the dihydroeugenol derivatives 27 and 28, sextets
near 1.6 ppm were observed, referring to H-20 hydrogens (propyl
group) of these derivatives. The saccharide H-1 hydrogens of com-
pounds 23–30 were registered as doublets with coupling constant
close to 8 Hz, typical of β-glycosides. The methoxyl protons of all
derivatives were observed near 3.8 ppm as singlets.
conventional chemotherapy, there are several limitations related to
the effectiveness of the anticancer drugs such as the non-selective
nature of these drugs, the high toxicity of these agents to healthy
normal tissues resulting in severe unintended and undesirable side
effects. Moreover, the development of cancer cells resistance to an-
titumor treatment is increasing every year. In order to address can-
cer, further research in the treatment of cancer is fundamental to
improving the outcomes for patients affected by this disease. These
efforts include developing more effective and less toxic treatments,
such as chemotherapy [5] and in this context, the importance of
research in the development of promising new synthetic drugs is
highlighted.
Eugenol or 4-allyl-2-methoxyphenol (compound I, Fig. 1) is an
allylphenol present in many medicinal and edible plants, being the
major component of the essential oil obtained from clove [Syzy-
gium aromaticum (L.)] [6]. It has been documented that eugenol
holds a wide range of biological activities, such as antifungal [7],
antitumor [8], antioxidant and anti-inflammatory [9]. These prop-
erties attract the attention for employing eugenol and analogues
(as isoeugenol, dihydroeugenol and guaiacol) in the development
of more active compounds or derivatives with distinct activities.
Our research group recently reported the synthesis and antimi-
crobial activity of eugenol glycoside derivatives containing differ-
ent groups in the saccharide unit. We observed that the eugenol
glucoside 23 (Fig. 2), in which there was a benzylidene group and
two hydroxyl groups in the saccharide unit, showed an interesting
fungistatic activity against the opportunistic Candida glabrata (IC50
18.1 μM), and it was 32 times more toxic to the microorganism
than to peripheral blood mononuclear cells (PBMCs) employed in
the cytotoxicity test [10].
In the 13C NMR spectra of compounds 23–30, aromatic carbons
were registered between 150 to 109 ppm, while the saccharide
carbons were observed between 104 to 65 ppm. The signs corre-
sponding to the acetal carbons were recorded near 100 ppm for all
the glycosides.
The main 1H and 13C NMR signals and FTIR bands registered for
glycosides 23–30 are summarized in Table 1 below.
The synthesized glycosides were also characterized by high res-
olution mass spectrometry (HRMS) and the molecular peaks values
are shown in Table 2. It was possible to observe that the mass ac-
curacy between the calculated and found values were lower than
5 ppm, which are acceptable for proposing a molecular formula
using this technique [13,14].
Another physicochemical properties of glycosides 23–30 were
determined, as melting point, solubility, specific optical rotation
[α]D and logP, as a way of presenting a better characterization of
the compounds. These data as well the yield of each compound are
shown in Table 3.
In view of the interesting activity and selectivity presented by
the compound 23, we decided to evaluate if further changes in
its structure could enhance these findings. Subsequently, we de-
cided to modify the aglycone subunit (eugenol, isoeugenol, di-
hydroeugenol and guaiacol) and the saccharide (D-glucose or d-
galactose). The rational with these changes was to evaluate the im-
portance of the para-substituent in the aglycone aromatic ring and
the spatial position of the benzylidene group for the antifungal ac-
tivity.
2.2. Biological evaluation
The glycosides 23–30 were evaluated against Candida albicans,
C. tropicalis, C. glabrata, C. krusei and C. parapsilosis by a mi-
crodilution method and the results are reported in the Table 4
[15,16]. Calculating the selectivity index of the active substances,
the cytotoxicity (CC50) of the derivatives 23–30 was evaluated
against healthy human lung fibroblasts cells (MRC-5) by the MTT
colorimetric method [17,18]. MTT (3-[4,5-dimethylthiazol-2-yl]−2,5-
diphenyltetrazolium bromide) is absorbed by mitochondria, where
it is transformed to formazan, a blue purple colored product and is
only formed in viable and active cells [19]. Fluconazole was used
2. Results and discussion
2.1. Chemistry
The proposed β-glycosides were obtained by the synthetic
route depicted in the Scheme 1. Initially, d-glucose (1) or d-
2

V.A.C. Péret, A.C.C. Reis, N.C. Silva et al.
Journal of Molecular Structure 1233 (2021) 130186
Fig. 2. Chemical structure of the new glycosides (24–30) designed from the hit (23).
Scheme 1. Synthetic route for obtaining the new glycosides 25–32.
Table 1
Main FTIR bands and NMR signals registered for glycosides 23–30.
Compound
FTIR (cm−1) Hydroxyl groups
Aromatic groups
1H NMR (ppm) Acetalprotons
Aromatic protons
13 C NMR (ppm) Sugar carbons
Acetal carbons
23
24
25
26
27
28
29
30
3383
3513
3642
3372
3448
3515
3386
3352
1511
1512
1510
1510
1512
1514
1504
1505
5.59
5.57
5.59
5.53
5.54
5.55
5.60
5.58
7.46–6.68
7.47–6.67
7.46–6.85
7.46–6.85
7.50–6.71
7.49–6.72
7.46–6.85
7.48–6.87
100.6–65.7
100.1–66.0
100.7–65.7
103.8–66.8
104.2–66.6
101.4–66.8
100.6–65.7
99.9–68.4
100.5
99.7
100.3
101.4
101.9
104.2
100.2
99.7
3

V.A.C. Péret, A.C.C. Reis, N.C. Silva et al.
Journal of Molecular Structure 1233 (2021) 130186
Table 2
Data obtained from HRMS analysis and molecular formula of the synthesized glycosides 23–30.
Compound
HRMS calculated [M + Na]+
HRMS found [M + H]+
Mass accuracy (ppm)
Molecular formula
23
24
25
26
27
28
29
30
437.1571
437.1571
437.1571
437.1571
439.1727
439.1727
397.1258
397.1258
437.1568
437.1567
437.1574
437.1567
439.1729
439.1728
397.1257
397.1259
0.68
0.91
0.68
0.91
0.45
0.22
0.25
0.25
C
C
C
C
C
C
C
C
23H26O7
23H26O7
23H26O7
23H26O7
23H28O7
23H28O7
20H22O7
20H22O7
Table 3
Physicochemical properties and yield of the glycosides 23–30.
Compound
Melting point ( °C)
Solubilitya
[α]D
LogP
Yield (%)
23
24
25
26
27
28
29
30
64–65
MeOH/DMSO
MeOH/DMSO
MeOH/DMSO
MeOH/DMSO
MeOH/DMSO
MeOH/DMSO
MeOH/DMSO
MeOH/DMSO
−21.5°
−67.5°
−45.5°
−74.0°
−22.0°
−67.5°
−54.25°
−45,0°
3.39
3.39
3.34
3.34
3.65
3.65
2.33
2.33
64
82
58
60
90
42
66
72
192–194
178–180
162–164
76–78
182–184
148–150
208–210
a
The solubility of the compounds was evaluated at 0.02 g.mL−1 in different solvents
(water, chloroform, methanol, ethanol, dimethyl sufoxide and acetonitrile), however they
were soluble only in methanol (MeOH) and dimethyl sulfoxide (DMSO).
Table 4
In vitro antifungal activity (IC50, μM), cytotoxicity (CC50, μM) on MRC-5 and selectivity index (SI) for the compounds 23–30.
Compound
IC50 (μM)
MRC-5 CC50 (μM)
Ca
IS
Ct
IS
Cg
IS
Ck
IS
Cp
IS
23
> 241.28
> 241.28
> 241.28
> 241.28
112.45
144.75
>1.6
–
–
18.1
–
>13
–
241.25
>1
–
–
–
–
–
–
–
–
–
2
–
24
–
–
241.25
241.25
144.75
240.11
144.07
–
>1
>1
>1.6
0.46
>1.6
–
–
–
25
–
–
–
–
–
–
–
26
–
–
–
–
–
–
–
27
–
–
–
–
–
–
–
28
> 240.11
> 267.10
> 267.10
>650.0
144.07
>1.6
–
–
–
144.07
>1.6
–
–
29
–
–
1
–
–
–
–
30
–
–
–
–
–
–
–
–
Fluconazole
>650
1
>650
16
>40
32
>20
>325
Ca: Candida albicans SC5314; Ct: Candida tropicalis ATCC 750; Cg: Candida glabrata ATCC 90,030; Ck: Candida krusei ATCC 6258; Cp:
Candida parapsilosis ATCC 22,019; IS: selectivity index, calculated as CC50/IC50
.
as a reference fungistatic drug and the results were estimated as
the minimum concentration required inhibiting 50% of the fungal
cell growth (IC50), which is related to fungistatic activity.
Although it was less active than doxorubicin, the glucoside 27
showed higher values of selectivity considering the healthy cell
line evaluated, which was used for the SI calculation (MRC-5).
While doxorubicin had a selectivity index of 3 and 3.8 for the T-24
(IC50: 5.21 μM) and TOV-21 G (IC50: 4.12 μM) lines, respectively,
the derivative 27 showed SI of 6.5 and >11 for these same tumor
cells with IC50 values of 17.11 μM (T-24) and 10.08 μM (TOV-21 G).
The discovery of anticancer compounds with high SI values rep-
resents the great challenge for the insertion of new drugs in the
market, so that they can represent safer substances with minimal
adverse effects to the patient. The galactoside 28 also showed in-
teresting results against HeLa, MDA-MB-231 and T-24 cells. This
compound was active at 17.41 μM with a SI greater than 13.7
against the HeLa cells, while doxorubicin was active at 10.01 μM
with a selectivity index close to 1.5 considering this same cell
line. This result indicated that the derivative 28 has greater safety
and selectivity for the HeLa strain in relation to the MRC-5 strain
when compared to doxorubicin. Accordingly, the galactoside 28
was also active against MDA-MB-231 and T-24 cells at 40.79 μM
and 44.61 μM with selectivity indexes greater than 5.8 and 5.3, re-
spectively. These results indicated that the presence of the propyl
group in the structures of the substances contributes to a greater
antitumor activity, since they are the only substances that have
this group in their structures. Consequently, the benzylidene group
The proposed chemical modifications on the hit 23 were not in-
teresting to increase its antifungal potential, since none of the new
derivatives showed a better activity against the yeasts evaluated
(Table 4). The most active compound was the derivative 28 (de-
rived from and d-galactose) that showed a broad action spectrum,
active against C. albicans, C. tropicalis and C. krusei at 144.07 μM.
However, it was 8 times less active than 23.
As most of the synthesized compounds showed considerable
cytotoxicity against the MRC-5 cells, we speculated that the deriva-
tives could be cytotoxic to cancer cell lines as well. Therefore, the
compounds 23–30 were further evaluated against four cancer cell
lines: HeLa (cervix cell carcinoma), MDA-MB-231 (breast cell carci-
noma), T24 (urinary bladder cell carcinoma) and TOV-21 G (ovarian
cell carcinoma) cells by the MTT colorimetric method [17,18]. The
results obtained were expressed as IC50 values and the selectivity
indexes of all the compounds were determined considering the cy-
totoxicity on MRC-5 (Table 5).
The most promising glycosides against the cancer cell lines
were the derivatives 27 and 28, both derived from dihydroeugenol.
The derivative 27, a glucoside, was active against the four cancer
cell lines evaluated with CC50 values between 10.08–59.91 μM.
4

V.A.C. Péret, A.C.C. Reis, N.C. Silva et al.
Journal of Molecular Structure 1233 (2021) 130186
Table 5
In vitro cytotoxicity (IC50, μM) and selectivity index (SI) for the compounds 23–30 against MRC-5 (ATCCꢀR CCL-117TM), HeLa (ATCCꢀR CCL-2TM), MDA-
MB-231 (ATCCꢀR HTB-26TM), T24 (ATCCꢀR HTB-4TM) and TOV-21 G (ATCCꢀR CRL-11730TM) cells.
Compound
IC50 (μM)
MRC-5a
HeLab
IS
MDA-MB-231a
IS
T24b
IS
TOV-21Gb
IS
23
24
25
26
27
28
29
30
DXR
> 241.28
> 241.28
> 241.28
> 241.28
112.45
48.96
105.22
113.38
79.17
59.91
17.41
48.48
203.16
10.01
1.42
1.34
1.41
1.30
> 4.93
> 2.29
> 1.81
> 3.05
4.51
67.79
90.30
79.26
70.21
58.33
40.79
1.65
1.52
1.29
1.52
2.00
1.48
> 3.56
> 2.67
> 3.05
> 3.44
1.92
117.48
104.48
30.33
97.62
17.11
44.61
114.29
1,48
1.36
> 2.05
> 2.31
> 7.96
> 2.47
6.57
51.95
1.45
2.01
1.70
1.44
1.28
1.54
> 4.64
> 1.23
> 1.27
> 1.63
> 11.16
> 1.53
1.0
195.22
78.61
1.13
1.24
1.17
1.18
147.50
10.08
1.59
1.38
1.34
1.30
> 240.11
> 267.10
> 267.10
>13.79
> 5.51
> 1.31
1.58
> 5.89
1.0
> 5.38
> 1.85
1.0
156.19
> 267.10
> 267.10
> 267.28
> 267.28
1.52
1.42
1.21
1.0
> 267.10
5.21 1.25
1.0
15.77
1.33
11.23
1.56
1.40
3.03
4.12
1.33
3.83
SI (selectivity index): CC50/ IC50; DXR: Doxorubicin.
a
Reading performed 24 h after treatment.
b
Reading performed 48 h after treatment.
positioned in the equatorial position of the carbohydrate seems to
favor the activity of the substances against the T-24 and TOV-21 G
cells since the derivative 27 was more active against these two cell
lines (IC50: 17.11 μM and 10.08 μM, respectively) as compared to
the derivative 28 (IC50: 44.61 μM and 156.19 μM, respectively),
which has such a group oriented in the axial position of the sac-
charide unit. The position of the benzylidene group also seems to
interfere in the cytotoxicity against MRC-5 of these two substances.
Whereas, the derivative 27 (benzylidene group oriented in equa-
torial) was cytotoxic at 112.45 μM while the derivative 28 (ben-
zylidene group oriented in axial) did not present cytotoxicity up
to 240.1 μM against this healthy cell line. Finally, the potency of
these compounds also seems to be dependent on the position of
the benzylidene group considering He-La cells, since derivative 27
was active at 59.91 μM and derivative 28 was active at 17.41 μM
against this cell line.
cancer agents. The glycosides and aglycons were evaluated against
A549 cells (lung cancer cells) and the results showed that the pres-
ence of a sugar amine has a dramatic effect upon anticancer activ-
ity [22].
In order to improve the in vitro cytotoxicity results this assay
was realized at different time intervals (24, 48 and 72 h after treat-
ment) for almost all cell lines (Supplementary Material). The re-
sults obtained demonstrate that there was no time-addition rela-
tionship, since for tumor cell lines HeLa, TOV-21 G and T24 the
highest IC50 was observed within 48 h after treatment.
2.3. Molecular modeling
In order to try to understand the antitumor mechanism of the
active compounds, we sought to investigate their possible molec-
ular targets. For this purpose, we first used the SwissTargetPredic-
tion webtool to search for protein target candidates for the most
promising glycosides 27 and 28 [23]. This tool uses a ligand-based
approach by performing a similarity search using chemical struc-
ture and molecular shape on a large dataset of known active com-
pounds and their respective targets [24]. The results expressed the
probability of the analyzed compounds in recognizing each protein
in the set as a target and this methodology achieved a 70% success
rate on validation tests when considering the top 15 predicted tar-
gets [23]. Unfortunately, the screening of the compound 27 showed
low probability results where the best match (17% probability) was
the Adenosine A1 receptor. Regarding compound 28, no similar ac-
tive compounds were found, thus no target was predicted. Since
the results indicated low probability targets and were unrelated to
antitumor activity, we decided to try a different approach.
The compounds 29 and 30 were the least active substances
against the evaluated cancer cells (except when we considered the
activity of the derivative 29 against HeLa cells). This may be related
to the fact that the allyl, propyl or propenyl substituents (present
in the other compounds) have been replaced by a hydrogen atom,
indicating that the presence of any of the groups in this position
contributes positively to the activity of the substances.
Finally, the derivatives 23–26 synthesized from eugenol or
isoeugenol showed moderate cytotoxicity activity against the eval-
uated cells, suggesting that the presence of a para-unsaturated
chain to the glycoside linkage negatively impacts the activity of
the substances. This negative contribution caused by the unsatu-
rated groups can be attributed to a more rigidity of the unsat-
urated groups, reducing the interaction of these derivatives with
some specific region of a possible action target present in the cells.
There are different studies showing that the connection of a
saccharide unit to different aglycones can contribute to increase
its biological potential, including its potential against cancer cells.
In addition, different methodologies for chemoselective synthe-
ses that allow regioselective and stereoselective reactions involv-
ing different carbohydrates are described [20]. Li and colleagues
[21] related the synthesis of a series of tigogenin neoglycosides
and posterior in vitro antitumor evaluation of these compounds
against five human cancer cell lines. The results revealed a sugar-
dependent activity profile of their cytotoxicity, the glycoconju-
gation converted the non-active tigogenin to the most potential
product with IC50 value of 2.7 μM and 4.6 μM against HepG2
(liver cancer cells) and MCF-7 (breast cancer cells) respectively.
The 3R-tigogenin neoglycosides exhibited enhanced antitumor ac-
tivity while the 3S-tigogenin almost showed no activity. Another
study developed by Zhang and colleagues described the synthesis
of a set of digitoxigenin glucosides and xylosides as potential anti-
On the SwissSimilarity webtool, we searched for similar struc-
tures of compound 27 on the Protein Data Bank (PDB) database us-
ing a 2D molecular fingerprint approach [25]. The PDB is a valuable
resource of experimentally determined tridimensional structures
of proteins, mostly in a complex with small molecules inhibitors,
whose fingerprints are used in the screening [26]. The fingerprints
used (FP2) are formed by a bit-string of molecular fragments de-
rived from linear paths up to seven bonds. A pair of molecules can
then be compared by their Tanimoto coefficient, the quotient of
the number of bits in common between the fingerprints pair di-
vided by the union of the bits set. This coefficient ranges from 0 to
1, where 1 represents a perfect match and is vastly used in simi-
larity searches [27]. The best cancer-related match structure found
by SwissSimilarity was etoposide with a Tanimoto index of 0.557.
Etoposide is a well-known antitumoral approved drug that acts
stabilizing the type II topoisomerase-DNA complex (TOP2), which
ultimately leads to DNA breakage and the death of tumor cells
[28]. When comparing both molecules, is it clear that the simi-
5

V.A.C. Péret, A.C.C. Reis, N.C. Silva et al.
Journal of Molecular Structure 1233 (2021) 130186
Fig. 3. Structure of etoposide, teniposide and general structure of the synthesized compounds. The glycoside moiety is highlighted in red and the methoxyphenyl ring in
blue.
larity is due to the glycosidic moiety containing an acetal group.
In addition, the methoxyphenyl ring of 27 could be seen as the E
ring in etoposide (Fig. 3). Both groups are essential for etoposide
binding to TOP2 [28]. In fact, insertion of the glycosidic moiety at
7-position of its precursor podophyllotoxin, a tubulin polymeriza-
tion inhibitor, turned it into a TOP2 inhibitor [29].
protonation and flip states of selected amino acids residues, and
validation of the docking protocol [30,31]. This validation was ac-
complished by a redocking experiment, where the protocol suc-
cessfully reproduced the conformation of etoposide in the crystal-
˚
lographic complex with an RMSD of 0.48 A (Fig. 4A).
In the experiment design, we also decided to include the drug
teniposide, an etoposide analog bearing a thiophene ring in the
ketal substituent, thus most closely related to our phenyl ketal
derivatives 27 and 28. The docking was performed using a search
To certify whether our compounds could mimic the etoposide
mechanism of action and thereby suggest their antitumor activity
to TOP2 inhibition, we performed docking experiments with glyco-
side 27 and 28 on the TOP2 crystallographic structure. The tridi-
mensional structure of TOP2 was retrieved from PDB as a complex
with DNA and etoposide (PDB 3qx3). The docking was performed
on GOLD and followed the good practice established in the liter-
ature, such as protonation and structural minimization of ligands,
evaluation of water molecules within the binding site, analysis of
˚
area of 8 A within etoposide in the crystallographic complex and
used the ChemPLP fitness function. None of our ligands predicted
that the binding poses had their glycosidic rings superposed to
the etoposide one, thus indicating they would not share the same
binding mode. Interestingly, teniposide, bearing a substituent as
bulky as the phenyl ring, indeed docked superposing etoposide
Fig. 4. A) Superposition of etoposide crystallographic conformation (orange sticks) and best scoring pose on redocking experiment (green sticks). B) Superposition of etopo-
side crystallographic conformation (orange lines), best scoring pose of compound 27 in normal (cyan lines) and shape constraint (blue lines) docking experiment. Glycosidic
moiety is depicted as sticks to highlight the positioning difference in the binding site of TOP2 (shown as a transparent surface area). C) The best scoring pose of compound
27 (cyan sticks). Hydrogen bonds are represented as green dashed lines and pi-stacking interactions as pink dashed lines. Protein amino acid residues and nucleotides are
shown as white and pink sticks, respectively.
6

V.A.C. Péret, A.C.C. Reis, N.C. Silva et al.
Journal of Molecular Structure 1233 (2021) 130186
Fig. 5. Superposition of etoposide crystallographic conformation (orange sticks) and the best scoring pose of teniposide (cyan sticks). DNA is schematically shown in pink
and protein is shown as a cyan ribbon with a transparent surface area.
Table 6
pounds, as all galactoside derivatives ranked poorly, including the
Scoring of unbiased docked compounds.
in vitro active compound 28 (Table 6).
Rank
Compound
ChemPLP Score
Although molecular modeling has its inherent limitations, the
results gave us strong evidence that the TOP2 would not be the
target of the active compounds, in spite of the similarity they have
with the saccharide unit of the etoposide. Despite all the effort,
these results would help to narrow down further mechanism re-
lated assays, which in turn might help to optimize the molecules
in an iterative process.
#1
#2
#3
#4
#5
#6
#7
#8
#9
#10
teniposide
164.76
155.02
113.16
112.68
110.8
104.45
95.97
94.21
90.61
82.1
etoposide
27
25
23
29
28
24
26
30
3. Conclusion
In this research study, seven new glycosides were synthesized
from d-glucose and d-galactose as well as different natural phe-
nols, and characterized by spectroscopy and spectrometry meth-
ods. The synthesized compounds did not show greater antifungal
activity than the hit 23, however, they were active against four
cancer cell lines: HeLa: cervix cell carcinoma; MDA-MB-231: breast
cell carcinoma; T24: urinary bladder cell carcinoma; and TOV-21G:
ovarian cell carcinoma from 10.08 μM. The derivatives 27 and 28
showed the best activities against the evaluated cell lines with IC50
values close to doxorubicin and better selectivity indexes for some
cancer cell lines than this antitumoral drug. Initial docking stud-
ies showed that TOP2 might not be the molecular target of active
compounds and these results should help us focus on a different
approach for structure optimization.
crystallographic conformation (Fig. 5), indicating that the differ-
ence in the binding mode of the synthesized compounds might be
caused by the methoxyphenyl ring.
In an attempt to address this question, we intentionally biased
the docking by awarding an additional score to constrain the po-
sitioning of this moiety. Even granting a generous additional 50
points, no ligands received more than 36 points, indicating that
steric clash might prevent correct positioning (Fig. 4B). Next, we
removed the phenyl ring from the glycosidic ketal and docked ke-
tone derivatives, intending to reproduce etoposide conformation.
Once again, the glycosidic moiety did not superpose the crystal-
lographic ligand, unless the constrain award was kept. Moreover,
when compared to etoposide and teniposide, the synthesized com-
pounds lack the interactions provided by the four fused ring sys-
tem. Although these drugs are not considered intercalators, this
system still plays an important role in ligand binding, since the
A and B rings stack with a guanine base, while A, B, and D rings
establish interactions with protein residues. At the same time, they
provide a spacer to correctly position the E ring, which is responsi-
ble for many important contacts with the protein residues (Gly478,
Asp479, and Arg503) and a deoxyribose ring that was not fully ob-
served in our ligands [28] (Fig. 4C). Besides, the best scoring com-
pound in the unbiased docking, 27, scored 27% less than the etopo-
side scored in the redocking experiment, where a score above 75%
the reference ligand is desirable [32]. Also, the ranking order be-
tween the ligands does not discriminate active from inactive com-
4. Materials and methods
4.1. Physical measurements
The melting points of the compounds were determined on a Re-
ichert Austria apparatus and are uncorrected. IR spectroscopy was
performed on a Shimadzu FTIR-Affinity-1 spectrometer. 1H NMR
and 13C NMR spectra were obtained on a Bruker AC-400 spectrom-
eter (400 MHz for 1H NMR and 100 MHz for 13C NMR spectra)
in deuterated chloroform (CDCl3) or dimethylsulfoxide (DMSO–d6).
Chemical shifts (δ) were reported in parts per million (ppm) with
reference to tetramethylsilane (TMS) as the internal standard and
coupling constants (J) were reported in Hertz (Hz). The following
7

V.A.C. Péret, A.C.C. Reis, N.C. Silva et al.
Journal of Molecular Structure 1233 (2021) 130186
abbreviations were used for the 1H multiplicities: singlet (s), dou-
blet (d), double doublet (dd), triplet (t) and multiplet (m). High
resolution mass spectra (HRMS) were acquired using a LCMS-Q-
ORBITRAP mass spectrometer and the samples were solubilized in
MeOH 50%, following manual injection. The specific optical rota-
from 0.169 mmol to 1.090 mmol, under cooling at 0 °C. The mix-
ture was stirred at 0 °C for 30 min when the completion of the
reaction was noted by TLC (eluent: ethyl acetate; revealing agent:
ethanolic solution of H2SO4 at 15%). After this time IRA-120 resin
was added to the medium until the pH 5 was reached. Then, the
resin was filtered off and the solvent of the filtrate was evaporated
under reduced pressure resulting in the pure products 15–22 with
yields in the range of 75–100%. These products were employed in
the next synthetic step without characterization.
ꢀR
tion [α]D were measured on a Quimis polarimeter, at 22 °C. The
reactions development was monitored by thin-layer chromatogra-
phy (TLC) on commercial silica gel 60 plates.
4.2. General procedure for the synthesis of derivatives 3 and 4
4.6. General procedure for the synthesis of derivatives 23–30
Sulfuric acid (20 drops) was added to a solution of 1 or 2
(13.9 mmol) in acetic anhydride (25 mL). The mixture was kept
under ultrasonic irradiation for 3 h at room temperature when the
completion of the reaction was noted by TLC (eluent: ethyl ac-
etate/hexane 1:1; revealing agent: ethanolic solution of H2SO4 at
15%). After this time, the mixture was poured into an ice bath and
the solid product obtained after manual stirring was filtered off.
The crude product was purified by recrystallization from ethanol
resulting in a white amorphous solid with yields in the range of
60–70%.
To a mixture of benzaldehyde (2 mL) and ZnCl2 (2.5 eq) was
added the corresponding deacetylated glycoside (15–22; 1 eq). The
mixture was stirred at room temperature for 3 h when the com-
pletion of the reaction was noted by TLC (eluent: ethyl acetate; re-
vealing agent: ethanolic solution of H2SO4 at 15%). Then, the mix-
ture was poured into ice water and petroleum ether (1:1) allowing
the formation of the pure products as white amorphous solids in
the interface of these solvents, which were separated by filtration,
With yields in the range of 42–89%.
4-allyl-2-methoxyphenyl-(4,6-O-benzyl-β-D-glucopyranoside) (23)
´
´
This compound was prepared by 0.242 mmol of 15 and 0.60 mmol
of ZnCl2, according to the general procedure. The product was ob-
tained in 64% yield as white amorphous solid; mp 64–66 °C; [α]D
−21.5 (c 0.002, MeOH); IR (ῡ/cm–1): 3383, 2990, 2879, 1638, 1511.
1H NMR (400 MHz, DMSO–d6) δ: 7.46 – 7.43 (m; 2H; H-9), 7.39
– 7.36 (m; 3H; H-11 and H-10), 7.04 (d; 1H; J3 = 8.4 Hz; H-17),
6.81 (d; 1H; J4 = 2 Hz; H-14), 6.68 (dd; 1H; J3 = 8 Hz; J4 = 2 Hz;
H-16), 5.95 – 5.91 (m; 1H; H-20), 5.59 (s; 1H; H-7), 5.55 (d; 1H;
J3 = 5.6 Hz; OH), 5.44 (d; 1H; J3 = 5.2 Hz; OH), 5.11 – 5.01 (m;
3H; H-21 and H-1), 4.16 (dd; 1H; J3 = 10 Hz; J3 = 4.8 Hz; H-6),
3.74 (s; 3H; H-18), 3.67 (t; 1H; J3 = 10 Hz; H-4), 3.67 – 3.51 (m;
4.3. General procedure for the synthesis of derivatives 5 and 6
To a flask containing acetic anhydride (20 mL) cooled to 0 °C
was added dropwise 48% hydrobromic acid (5 mL). This first so-
lution was added dropwise to a second solution consisted of 5 or
6 (5.12 mmol) in dichloromethane (10 mL) at 0 °C. After the ad-
dition, the mixture was stirred at room temperature for 5 h when
the completion of the reaction was noted by TLC (eluent: ethyl ac-
etate/hexane 1:1; revealing agent: ethanolic solution of H2SO4 at
15%). Then, cold water (100 mL) was added to the mixture, which
was extracted with dichloromethane (3 × 30 mL). The obtained
organic layer was washed with saturated sodium bicarbonate solu-
tion (3 × 30 mL) followed by water (2 × 100 mL). The resulting
organic layer was dried by anhydrous sodium sulfate, filtered and
the solvent was removed under reduced pressure resulting in the
crude products 5 or 6 with yields in the range of 90–95%. These
products were employed in the next step without further purifica-
tion due to their labile nature.
ꢁ
2H; H-5 and H-6 ), 3.46 (t; 1H; J3 = 9.2 Hz; H-3), 3.38 (m; H-
2), 3.03 (d; 2H; J3 = 6.8 Hz; H-19). 13C NMR (100 MHz, DMSO–
d6) δ: 149.13 (1C; C13); 144.37 (1C; C12); 137.90 (1C; C20); 137.78
(1C; C8); 134.04 (1C; C15); 128.89 (1C; C11); 128.08 (2C; C10);
126.37 (2C; C9); 120.28 (1C; C16); 115.95 (1C; C17); 115.66 (1C;
C21); 113.00 (1C; C14); 100.67 (1C; C1); 100.51 (1C; C7); 80.35
(1C; C4); 74.15 (1C; C5) 73.01 (1C; C2); 67.85 (1C; C6); 65.76 (1C;
C3); 55.62 (1C; C18); 39.56 (1C; C19). HRMS-ESI: m/z calcd. for
C23H26O7 (M+Na)+: 437.1571; found: 437.1568.
4.4. General procedure for synthesis of derivatives 7–14
4-allyl-2-methoxyphenyl-(4,6-O-benzyl-β-D-galactopyranoside
´
´
(24) This compound was prepared by 0.257 mmol of 16 and
0.73 mmol of ZnCl2, according to the general procedure. The
product was obtained in 82% yield as white amorphous solid;
mp 192–194 °C; [α]D −67.5 (c 0.002, MeOH); IR (ῡ/cm–1): 3513,
3339, 2938, 2892, 1512. 1H NMR (400 MHz, DMSO–d6) δ: 7.47 –
7.45 (m; 2H; H-9), 7.39 – 7.35 (m; 3H; H-11 and H-10), 7.04 (d;
1H; J3 = 8.4 Hz; H-17), 6.81 (d; 1H; J3 = 1.6 Hz; H-14), 6.67 (dd;
1H; J3 = 8.2; Hz; J4 = 1.6 Hz; H-16), 5.98 – 5.91 (m; 1H; H-20),
5.57 (s; 1H; H-7), 5.19 (d; 1H; J3 = 5.2 Hz; OH), 5.09 −5.02 (m;
3H; OH and H-21), 4.98 (d; 1H; J3 = 7.28 Hz; H-1), 4.12 (d; 1H;
J3 = 3.2 Hz; H-4), 4.02 (dd; 2H; J2 = 21.8 Hz; J3 = 11 Hz; H-6),
3.74 (s; 3H; H-18), 3.70 (s; 1H; H-3), 3.65 – 3.55 (m; 2H; H-2
and H-5), 3.30 (d; 2H; J3 = 6.8 Hz; H-19). 13C NMR (100 MHz,
DMSO–d6) δ: 148.94 (1C; C13); 144.60 (1C; C12); 138.58 (1C; C8);
137.94 (1C; C20); 133.49 (1C; C15); 128.64 (1C; C11); 127.93 (2C;
C10); 126.28 (2C; C9); 120.17 (1C; C18); 115.58 (1C; C21); 115.42
(1C; C17); 112.86 (1C; C14); 100.18 (1C; C1); 99.76 (1C; C7); 75.88
(1C; C4); 71.95 (1C; C5); 69.65 (1C; C2); 68.47 (1C; C6); 66.03
(1C; C3); 55.54 (1C; C18); 38.5 (1C; C19). HRMS-ESI: m/z calcd. for
C23H26O7 (M+Na)+: 437.1571; found: 437.1567.
To a flask containing the corresponding phenol (3 eq) was
added a solution of lithium hydroxide (2.7 eq) in water (5 mL).
The mixture was stirred at room temperature until the complete
solubilization of the phenol. Then a solution of 5 or 6 (1 eq) in
acetone (15 mL) was added. The mixture was stirred for 6 h at
room temperature when the completion of the reaction was noted
by TLC (eluent: ethyl acetate/hexane 7:3; revealing agent: ethano-
lic solution of H2SO4 at 15%). The acetone was removed under
an air flux and the crude product extracted with dichloromethane
(3 × 30 mL). The resulting organic layer was washed with aqueous
10% NaOH (3 × 30 mL) and water (2 × 100 mL), dried by anhy-
drous sodium sulfate, filtered and the solvent was removed under
reduce pressure resulting in crude products as amorphous solids
or oils. All these products were purified by recrystallization from
isopropanol resulting in white amorphous solids with yields in the
range of 12–72%. These products were employed in the next syn-
thetic step without characterization.
4.5. General procedure for the synthesis of derivatives 15–22
4-propenyl-2-methoxyphenyl-(4,6-O-benzyl-β-D-glycopyranoside)
´
´
To a solution of KOH (50 mg) in methanol (10 mL) were
(25) This compound was prepared by 0.281 mmol of 17 and
added the peracetylated glycosides 7–16, in quantities ranging
0.69 mmol of ZnCl2, according to the general procedure. The
8

V.A.C. Péret, A.C.C. Reis, N.C. Silva et al.
Journal of Molecular Structure 1233 (2021) 130186
product was obtained in 58% yield as white amorphous solid; mp
178–180 °C; [α]D −45.5 (c 0.002, MeOH); IR (ῡ/cm–1): 3642, 3474,
3385, 2979, 2878, 1510. 1H NMR (400 MHz, DMSO–d6) δ: 7.46
– 7.43 (m; 2H; H-9), 7.38 – 7.37 (m; 3H; H-11 and H-10), 7.05 –
7.01 (m; 2H; H-16 and H-14), 6.85 (dd; 1H; J3 = 8.4 J4 =1.6 Hz;
H-17), 6.34 (dd; 1H; J3 = 16 Hz; J4 = 1.2 Hz; H-19), 6.24 – 6.15
(m; 1H; H-20), 5.59 (s; 1H; H-7), 5.57 (d; 1H; J3 = 5.6 Hz; OH),
5.45 (d; 1H; J3 = 5.2 Hz; OH), 5.13 (d; 1H; J3 = 7.6 Hz; H-1), 4.17
(dd; 1H; J2 = 9.8 Hz J3 = 4.8 Hz; H-6), 3.72 (s; 3H; H-18), 3.69
(m; 2H; H-9), 7.35 – 7.34 (m; 3H; H-10 and H-11), 7.19 (d; 1H;
J3 = 8 Hz; H-17), 6.72 (d; 2H; J3 = 7.2 Hz; H-14 and H-16), 5.55 (s;
1H; H-7), 4.67 (d; 1H; J3 = 7.6 Hz; H-1), 4.37 (d; 1H; J3 = 11.6 Hz;
H-6), 4.21 (d; 1H; J3 = 3.2 Hz; H-4), 4.08 (dd; 1H; J2 = 12.8 Hz
ꢁ
J3 = 1.6 Hz; H-6 ), 3.99 (t; 1H; J3 = 8 Hz; H-2), 3.85 (s; 3H; H-18),
3.75 (dd; 1H; J3 = 9.8 Hz J3 = 3.6 Hz; H-5), 3.53 (s; 1H; H-3), 2.55
(t; 2H; J3 = 7.2 Hz; H-19), 1.62 (sext; 2H; J3 = 7.2 Hz; H-20), 0.94
(t; 3H; J3 = 7.6 Hz; H-21).13C NMR (100 MHz, CDCl3) δ: 150.19
(1C; C13); 143.73 (1C; C12); 139.57 (1C; C8); 137.47 (1C; C15);
129.22 (1C; C11); 128.18 (2C; C10); 126.45 (2C; C9); 121.13 (1C;
C17); 120.80 (1C; C14); 112.24 (1C; C16); 104.21 (1C; C7); 101.47
(1C; C1); 75.06 (1C; C4); 72.54 (1C; C2); 71.09 (1C; C5); 69.06
(1C; C6); 66.88 (1C; C3); 55.83 (1C; C18); 37.82 (1C; C19); 24.61
(1C; C20); 13.77 (1C; C21). HRMS-ESI: m/z calcd. for C23H28O7
(M+Na)+: 439.1727; found: 439.1728.
ꢁ
(t; 1H; J3 = 10 Hz; H-4), 3.62 – 3.53 (m; 2H; H-5 and H-6 ), 3.47
(t; 1H; J3 = 10.4 Hz; H-3), 3.39 – 3.35 (m; 1H; H-2), 1.82 (dd;
3H; J3 = 6.4 J4 = 1.2 Hz; H-21). 13C NMR (100 MHz, DMSO–d6)
δ: 149.22 (1C; C13); 145.18 (1C; C12); 137.76 (1C; C8); 132.08
(1C; C15); 130.50 (1C; C19); 128.89 (1C; C11); 128.88 (2C; C10);
128.07 (2C; C9); 123.97 (1C; C20); 118.38 (1C; C16); 115.68 (1C;
C17); 109.63 (1C; C14); 100.67 (1C; C1); 100.30 (1C; C7); 80.32
(1C; C4); 74.11 (1C; C5); 73.00 (1C; C2); 67.84 (1C; C6); 65.76 (1C;
C3); 55.60 (1C; C18); 18.23 (1C; C21). (1C; C6); 66.03 (1C; C3);
55.54 (1C; C18); 38.5 (1C; C19). HRMS-ESI: m/z calcd. for C23H26O7
(M+Na)+: 437.1571; found: 437.1574.
2-methoxyphenyl-(4,6-O-benzyl-β-D-glycopyranoside) (29) This
´
´
compound was prepared by 0.350 mmol of 21 and 0.87 mmol
of ZnCl2, according to the general procedure. The product was
obtained in 66% yield as white amorphous solid; mp 148–150 °C;
[α]D −54.25 (c 0.002, MeOH); IR (ῡ/cm–1): 3386, 2883, 1504. 1H
NMR (400 MHz, DMSO–d6) δ: 7.46 – 7.44 (m; 2H; H-9), 7.39 –
7.37 (m; 3H; H-10 and H-11), 7.12 (d; 1H; J3 = 8.4 Hz; H-17), 7.01
– 6.94 (m; 2H; H-14 and H-16), 6.89 – 6.85 (m; 1H; H-15), 5.60 (s;
1H; H-7), 5.57 (d; 1H; J3 = 5.6 Hz; OH), 5.44 (d; 1H; J3 = 4.8 Hz;
OH), 5.15 (d; 1H; J3 = 7.6 Hz; H-1), 4.13 (dd; 1H; J2 = 9.8 Hz;
J3 = 4.8 Hz; H-6), 3.75 (s; 1H; H-18), 3.68 (t; 1H; J3 = 9.6 Hz;
4-propenyl-2-methoxyphenyl-(4,6-O-benzyl-β-D-
´
´
galactopyranoside) (26) This compound was prepared by
0.367 mmol of 18 and 0.91 mmol of ZnCl2, according to the
general procedure. The product was obtained in 60% yield as
white amorphous solid; mp 162–164 °C; [α]D −74.0 (c 0.002,
MeOH); IR (ῡ/cm–1): 3372, 2936, 2876, 1510. 1H NMR (400 MHz,
CDCl3) δ: 7.46 (br s; 2H; H-9), 7.33 (br s; 3H; H-11 and H-10), 7.17
(d; 1H; J3 = 8 Hz; H-17), 6.85 (br s; 2H; H-16 and H-14), 6.33 (d;
1H; J3 = 15.6 Hz; H-19), 6.16 – 6.10 (m; 1H; H-20), 5.53 (s; 1H;
H-7), 4.68 (d; 1H; J3 = 8 Hz; H-1), 4.35 (d; 1H; J3 = 12.4 Hz; H-6),
ꢁ
H-4), 3.62 – 3.53 (m; 2H; H-6 and H-5), 3.47 (t; 1H; J3 = 5.2 Hz;
H-3), 3.41 – 3.35 (m; 1H; H-2).13C NMR (100 MHz, DMSO–d6) δ:
149.21 (1C; C13); 146.13 (1C; C12); 137.77 (1C; C8); 128.89 (1C;
C11); 128.07 (2C; C10); 126.36 (2C; C9); 122.41 (1C; C16); 120.65
(1C; C15); 115.76 (1C; C17); 112.75 (1C; C14); 100.67 (1C; C1);
100.27 (1C; C7); 80.33 (1C; C4); 74.12 (1C; C5); 73.01 (1C; C2);
67.85 (1C; C6); 65.76 (1C; C3); 55.61 (1C; C18). HRMS-ESI: m/z
calcd. for C20H22O7 (M+Na)+: 397.1258; found: 397.1257.
ꢁ
4.22 (s; 1H; H-4), 4.07 (d; 1H; J3 = 12.8 Hz; H-6 ), 3.98 (t; 1H;
J3 = 9.2 Hz; H-2), 3.85 (s; 3H; H-18), 3.75 (d; 1H; J3 = 7.6 Hz;
H-5), 3.52 (s; 1H; H-3), 1.86 (d; 3H; J3 = 6.4 Hz; H-21).13C NMR
(100 MHz, CDCl3) δ: 150.40 (1C; C13); 144.76 (1C; C12); 137.46
(1C; C8); 134.81 (1C; C15); 130.34 (1C; C19); 129.23 (1C; C11);
128.20 (2C; C10); 126.46 (2C; C9); 125.41 (1C; C20); 120.93 (1C;
C16); 118.73 (1C; C17); 109.20 (1C; C14); 103.88 (1C; C1); 101.44
(1C; C7); 75.04 (1C; C4); 72.48 (1C; C5); 71.09 (1C; C2); 69.03 (1C;
C6); 66.89 (1C; C3); 55.85 (1C; C18); 18.40 (1C; C21). HRMS-ESI:
m/z calcd. for C23H26O7 (M+Na)+: 437.1571; found: 437.1567.
2-methoxyphenyl-(4,6-O-benzyl-β-D-galactopyranoside) (30) This
´
´
compound was prepared by 0.350 mmol of 22 and 0.87 mmol of
ZnCl2, according to the general procedure. The product was ob-
tained in 72% yield as white amorphous solid; mp 208–210 °C;
[α]D −45 (c 0.002, MeOH); IR (ῡ/cm–1): 3352, 3205, 2969, 2928,
2875, 1505. 1H NMR (400 MHz, DMSO–d6) δ: 7.48 – 7.45 (m;
2H; H-9), 7.40 – 7.35 (m; 3H; H-10 and H-11), 7.12 (d; 1H;
J3 = 7.2 Hz; H-17), 7.00 – 6.92 (m; 2H; H-14 and H-16), 6.87 (sext;
1H; J3 = 7.6 Hz J4 = 1.6 Hz; H-15), 5.58 (s; 1H; H-7), 5.21 (d; 1H;
4-propyl-2-methoxyphenyl-(4,6-O-benzyl-β-D-glycopyranoside)
´
´
(27) This compound was prepared by 0.016 mmol of 19 and
0.73 mmol of ZnCl2, according to the general procedure. The
product was obtained in 90% yield as white amorphous solid; mp
76–78 °C; [α]D −22 (c 0.001, MeOH); IR (ῡ/cm–1): 3448, 2953,
2926, 2869, 1512. 1H NMR (400 MHz, CDCl3) δ: 7.50 – 7.48 (m; 2H;
H-9), 7.37 – 7.35 (m; 3H; H-11 and H-10), 7.05 (d; 1H; J3 = 8 Hz;
H-17), 6.74 – 6.71 (m; 2H; H-14 and H-16), 5.54 (s; 1H; H-7), 4.75
(d; 1H; J3 = 8 Hz; H-1), (dd; 1H; J2 = 10.4; J3 Hz = 4.8 Hz; H-6),
J3 = 5.2 Hz; OH), 5.08 (d; 1H; J3 = 6 Hz; OH), 5.03 (d; 1H; J3
=
7.6 Hz; H-1), 4.13 (d; 1H; J3 = 3.2 Hz; H-4), 4.03 (dd; J2 = 22 Hz;
ꢁ
J3 = 11 Hz; 2H; H-6 and H-6 ), 3.76 (s; 3H; H-18), 3.73 (s; 1H; H-3),
3.66 – 3.58 (m; 2H; H-5 and H-2).13C NMR (100 MHz, DMSO–d6)
δ: 149.02 (1C; C13); 146.33 (1C; C12); 138.57 (1C; C8); 128.63 (1C;
C11); 127.93 (2C; C10); 126.27 (2C; C9); 121.94 (1C; C16); 120.54
(1C; C15); 115.26 (1C; C17); 112.57 (1C; C14); 99.94 (1C; C1); 99.76
(1C; C7); 75.86 (1C; C4); 71.95 (1C; C5); 69.61 (1C; C2); 68.44
(1C; C6); 68.44 (1C; C3); 55.51 (1C; C18). HRMS-ESI: m/z calcd. for
C20H22O7 (M+Na)+: 397.1258; found: 397.1259.
ꢁ
3.88 (br s; 1H; H-6 ), 3.85 (s; 3H; H-18), 3.83 (br s; 1H; H-3), 3.75
(t; 1H; J3 = 8.5 Hz; H-2), 3.62 (t; 1H; J3 = 9.2 Hz; H-4), 3.50 (td;
1H; J3 = 9.7 Hz J3 = 4.8 Hz; H-5), 2.55 (t; 2H; J3 = 7.4 Hz; H-19),
1.63 (sext; 2H; J3 = 7.6 Hz; H-20), 0.94 (t; 3H; J3 = 7.3 Hz; H-21).
13C NMR (100 MHz, CDCl3) δ: 149.98 (1C; C13); 143.74 (1C; C12);
139.83 (1C; C8); 136.87 (1C; C15); 129.25 (1C; C11); 128.31 (2C;
C10); 126.25 (2C; C9); 120.85 (1C; C16); 120.34 (1C; C17); 112.35
(1C; C14); 104.22 (1C; C1); 101.91 (1C; C7); 80.26 (1C; C4); 74.35
(1C; C5); 73.04 (1C; C2); 68.60 (1C; C6); 66.66 (1C; C3); 55.84 (1C;
C18); 37.82 (1C; C19); 24.61 (1C; C20); 13.78 (1C; C21). HRMS-ESI:
m/z calcd. for C23H28O7 (M+Na)+: 439.1727; found: 439.1729.
4-propyl-2-methoxyphenyl-(4,6-O-benzyl-β-D-galactopyranoside)
4.7. Antifungal activity evaluation
The glycosides 23–30 were evaluated in vitro for their antifun-
gal activity against C. albicans, C. tropicalis, C. glabrata, C. krusei
and C. parapsilosis through a broth microdilution method proposed
by documents M27A3 and M27-S4 with modifications [15,16]. The
stock solutions of all the compounds were prepared in DMSO 1%
at final concentration and tested at the following concentrations
(μg.mL−1): 100; 60; 30; 15; 7.5; 3.75; 1.875; 0.468; 0.23; and
0.06. The standard drug fluconazole was applied as control of an-
tifungal action. Results were visualized and analyzed at 530 nm
in an Anthos Zenyth 200rt Microplate Reader (Anthos Labtec In-
´
´
(28) This compound was prepared by 0.305 mmol of 20 and
0.76 mmol of ZnCl2, according to the general procedure. The
product was obtained in 42% yield as white amorphous solid; mp
182–184 °C; [α]D −67.5 (c 0.001, MeOH); IR (ῡ/cm–1): 3515, 3351,
2964, 2932, 2858, 1514. 1H NMR (400 MHz, CDCl3) δ: 7.49 – 7.47
9

V.A.C. Péret, A.C.C. Reis, N.C. Silva et al.
Journal of Molecular Structure 1233 (2021) 130186
of 10−7 [33]. The docking experiments were done using the TOP2
crystallographic structure retrieved from PDB (PDB 3qx3) and pre-
pared in the Discovery Studio Visualizer (version 17.2 - Dassault
Systèmes Biovia Corp) as follows. Water molecules and ions farther
struments, Salzburg, Austria). The inhibitory concentrations of mi-
crobial growth were determined at 50% (IC50) and 90% (IC90) in
μmol.mL−1
.
˚
4.8. Cytotoxic activity evaluation
than 5 A from the ligand as well as the unnecessary protein chain
B were removed. The flip of Asn-and Gln-and protonation states
of binding site residues were also checked. On the Hermes inter-
face (version 1.10.5 - Cambridge Crystallographic Data Centre), hy-
drogens were added to the complex, the crystallographic etoposide
molecule was removed and its coordinates were used to define the
Cytotoxic evaluation assay was performed against a panel of
human cancer cell lines and a normal cell line to determine the
50% inhibitory concentration (IC50) and the selectivity index (SI).
ꢀR
ꢀR
ATCC cell lines including cervix cell carcinoma HeLa (ATCC CCL-
ꢀR
2TM), breast cell carcinoma MDA-MB-231 (ATCC HTB-26TM), uri-
˚
binding site as the residues 8 A from it. The presence of interacting
nary bladder transitional cell carcinoma T24 (ATCC HTB-4TM),
ovarian cell carcinoma TOV-21 G (ATCC CRL-11,730TM) and nor-
mal human lung fibroblast cell MRC-5 (ATCC CCL-117TM) were
ꢀR
water molecules (63, 723, 927, 1142, 1265, 1416, and 1461) were
assigned by the GOLD algorithm, as well as their hydrogen position
(spin) in order to maximize hydrogen bonds. For each ligand, 30
independent docking runs were performed on GOLD (version 5.8.1
- Cambridge Crystallographic Data Centre) using the ChemPLP scor-
ing function and default GA settings with a minimum of 10,000
operations. An early termination rule was defined to stop the cal-
ꢀR
ꢀR
used in the assays. Cells were cultivated in complete cell medium
consisting of Dulbecco’s modified Eagle’s medium (DMEM, Cultilab,
Campinas, SP, Brazil), supplemented with 5% fetal bovine serum,
100 U/mL penicillin and 5 μg.mL−1 amphotericin B, at 37 °C, in 5%
CO2 atmosphere and harvested in log-phase for experimental use.
Cells lines were seeded on 96 well plates. Compounds were
dissolved in DMSO, and serial dilutions in DMEM with concentra-
tions ranging from 100.0 to 1.5625 μg.mL−1 were prepared, each
sample was assayed in three replicates (n = 3). Doxorubicin, an
antitumor drug, and DMSO, solvent used to dilute samples, were
used as positive and negative control, respectively. Cell lines were
exposed to different concentrations of compounds for 24 or 48
and 72 h. After incubation, cell viability was assessed by the 3-
(4,5-dimethylthiazol-2-yl)−2,5-diphenyltetrazolium bromide (MTT,
Sigma-Aldrich) assay at a concentration of 2 mg.mL−1 in PBS. MTT
solution was added in each well and incubated 120 min at 37 °C,
then 490 nm light absorbance readings were taken [17,18]. The cy-
totoxicity of each sample was expressed as IC50, i.e. the concentra-
tion of sample that inhibited cell growth by 50% [17], for the de-
termination of IC50 value, nonlinear regressions of concentration–
response curves were used, and statistical analysis of the data was
carried out by the GraphPad Prism 5.0 program (GraphPad Soft-
ware Inc., San Diego, CA). Results were expressed as the mean
S.D. of three independent experiments.
˚
culations when the top 3 solutions had an RMSD of less than 1 A.
The ligands were considered flexible, including the flip of ring cor-
ners. Constrained docking was achieved by using the shape over-
lapping constraint function, with a constraint weight of 50 and the
glycoside moiety of etoposide crystallographic conformation as a
template. Docked conformations were analyzed and figures pre-
pared on Discovery Studio Visualizer.
Author statement
Vinícius Augusto Campos Péret: synthesis and characterization
of the compounds.
Adriana Cotta Cardoso Reis: evaluation of the cytotoxicity of
the compounds.
Naiara Chaves Silva: evaluation of the antifungal activity of the
compounds.
Amanda Latercia Tranches Dias: supervision of antifungal
tests.
Diogo Teixeira Carvalho: collaboration with the characteriza-
tion of the compounds and writing.
4.9. Selectivity indexes
Danielle Ferreira Dias: collaboration with the characterization
of the compounds and writing.
To the antifungal activity the selectivity index (SI) was calcu-
Saulo Fehelberg Pinto Braga: molecular modeling studies.
Geraldo Célio Brandão: supervision of cytotoxicity tests.
Thiago Belarmino de Souza: general supervision of the work
and writing.
lated as the ratio between the 50% inhibitory concentration (CC50
)
in MRC-5 (normal cell line) of a compound and the 50% inhibitory
concentrations (IC50) of microbial growth test. To the cytotoxic ac-
tivity the SI was calculated as the ratio between the 50% inhibitory
concentration (CC50) in MRC-5 (normal cell line) of a compound
and the 50% inhibitory concentration (IC50) in each cancer cell line.
These indexes reflect the potency and possible selectivity for future
drug development.
Declaration of Competing Interest
The authors declare no conflict of interest.
4.10. Molecular modeling
Acknowledgments
Target prediction was performed from 2D structures on the
SwissTargetPrediction webtool (http://www.swisstargetprediction.
ch/) using the Homo sapiens macromolecular targets database
[23]. The similarity search was conducted on the SwissSimilarity
webtool (http://www.swisssimilarity.ch/) on the Protein Data Bank
(PDB) library using the 2D molecular fingerprint FP2 of compound
27 as the search method [25].
The authors acknowledge the PROPP/UFOP, CNPq and FAPEMIG
(APQ-00820-19) agencies for the scholarships and financial support
provided. The authors acknowledge the “Laboratório Multiusuário
de Caracterização de Moléculas (Escola de Farmácia/ Universidade
Federal de Ouro Preto/MG)” for nuclear magnetic resonance anal-
ysis. The authors acknowledge the “Base de Estruturas Cristali-
nas/CAPES” for GOLD software licensing. The authors would also
like to acknowledge the “Laboratório Multiusuário de Proteômica e
Biomoléculas (LMU-ProtBio), from Núcleo de Pesquisas em Ciên-
cias Biológicas”, Universidade Federal de Ouro Preto, MG, Brazil,
for providing the required equipments and technical expertise for
sample processing and analyses.
Ligand preparation was accomplished on Avogadro 1.2.0 by con-
verting 2D structures of etoposide, teniposide and our compounds
into 3D structures, which lowest energy conformers found by sys-
tematic rotor search were further minimized by applying 500 steps
of steepest descend followed by 5000 steps of gradient descent
algorithm using the MMFF94 forcefield with convergence criteria
10

V.A.C. Péret, A.C.C. Reis, N.C. Silva et al.
Journal of Molecular Structure 1233 (2021) 130186
Supplementary materials
[17] G.C. Brandão, E.G. Kroon, D.E.R. Souza, J.D.S. Filho, A.B Oliveira, Chemistry and
antiviral activity of Arrabidaea pulchra (Bignoniaceae), Molecules 18 (8) (2013)
9919–9932, doi:10.3390/molecules18089919.
[18] P.R. Twentyman, M. Luscombe, A study of some variables in a tetrazolium dye
(MTT) based assay for cell growth and chemosensitivity, Br. J. Cancer 56 (3)
(1987) 279–285, doi:10.1038/bjc.1987.190.
Supplementary material associated with this article can be
found, in the online version, at doi:10.1016/j.molstruc.2021.130186.
[19] P. Ghosh, A.R. Chowdhury, S.K. Saha, M. Glosh, M. Pal, N.C. Murmu, P.S. Baner-
jee, Synthesis and characterization of redox non-innocent cobalt(III) com-
plexes of a O,N,O donor ligand: radical generation, semi-conductivity, antibac-
terial and anticancer activities, Inorganica Chim. Acta 429 (2015) 99–108 doi:
http://dx.doi.org/10.1016/j.ica.2015.01.029.
References
[1] F. Bongomin, S. Gago, R.O. Oladele, D.W. Denning, Global and multi-national
prevalence of fungal diseases-estimate precision, J. Fungi 3 (57) (2017) 1–29,
doi:10.3390/jof3040057.
[20] R.D. Goff, J.S. Thorson, Neoglycosylation and neoglycorandomization: enabling
tools for the discovery of novel glycosylated bioactive probes and early stage
leads, MedChemCom, 5 (2014) 1036–1047, doi:10.1039/c4md00117f.
[21] G. Li, H. Xu, S. Xu, W. Wang, B. Yu, J. Zhang, Synthesis of tigogenin MeON-
Neoglycosides and their antitumor activity, Fitoterapia 125 (2018) 33–40,
doi:10.1016/j.fitote.2017.12.014.
[22] J. Shang, L.V. Ponomareva, N.S. Nandurkar, Y. Yuan, L. Fang, C. Zhan, J.S. Thor-
son, Influence of sugar amine regiochemistry on digitoxigenin neoglycoside an-
ticancer activity, ACS Mecicinal Chem. Lett. 6 (2015) 1053–1058, doi:10.1021/
acsmedchemlett.5b00120.
[23] A. Daina, O. Michielin, V. Zoete, SwissTargetPrediction: updated data and new
features for efficient prediction of protein targets of small molecules, Nucleic
Acid. Res. 47 (W1) (2019) 357–364, doi:10.1093/nar/gkz382.
[24] D. Gfeller, O. Michielin, V. Zoete, Shaping the interaction landscape of
bioactive molecules, Bioinformatics 29 (23) (2013) 3073–3079, doi:10.1093/
bioinformatics/btt540.
[2] J. Berman, D.J. Krysan, Drug resistance and tolerance in fungi, Nat. Rev. Micro-
biol. 18 (2020) 319–331, doi:10.1038/s41579-019-0322-2.
[3] K. Forsberg, K. Woodworth, M. Walters, E.L. Berkow, B. Jackson, T. Chiller,
S. Vallabhaneni, Candida auris: the recent emergence of a multidrug-resistant
fungal pathogen, Med. Mycol. 57 (1) (2019) 1–12, doi:10.1093/mmy/myy054.
[4] WHO, World Health Organization. (2020). Cancer. Retrieved from
https://www.who.int/health-topics/cancer#tab=tab_1.
[5] NCI, National Cancer Institute. (2020). Cancer Treatment. Retrieved from https:
//www.cancer.gov/about-cancer/treatment.
[6] R.S. Bendre, J.D. Rajput, S.D. Bagul, P.S. Karandikar, Outlooks on medicinal
properties of eugenol and its synthetic derivatives, Nat. Prod. Chem. Res. 04
(03) (2016) 1–6, doi:10.4172/2329-6836.1000212.
[7] R. Cai, M. Hu, Y. Zhang, C. Niu, T. Yue, Y. Yuan, Z. Wang, Antifungal activity and
mechanism of citral, limonene and eugenol against Zygosaccharomyces rouxii,
Lwt 106 (2019) 50–56, doi:10.1016/j.lwt.2019.02.059.
[8] P. Binu, N. Priya, S. Abhilash, R.C. Vineetha, H. Nair, Protective effects of
eugenol against hepatotoxicity induced by arsenic trioxide: an antileukemic
drug, Iran J. Med. Sci. 43 (3) (2018) 305–312.
[25] V. Zoete, A. Daina, C. Bovigny, O. Michielin, SwissSimilarity: a web tool for low
to ultra high throughput ligand-based virtual screening, J. Chem. Inf. Model.
56 (8) (2016) 1399–1404, doi:10.1021/acs.jcim.6b00174.
[9] S. Mateen, M.T. Rehman, S. Shahzad, S.S. Naeem, A.F. Faizy, A.Q. Khan,
M.S. Khan, F.M. Husain, S. Moin, Anti-oxidant and anti-inflammatory effects of
cinnamaldehyde and eugenol on mononuclear cells of rheumatoid arthritis pa-
tients, Eur. J. Pharmacol. 852 (5) (2019) 14–24, doi:10.1016/j.ejphar.2019.02.031.
[10] T.B. Souza, K.M.O. Brito, N.C. Silva, R.P. Rocha, G.F. Sousa, L.P. Duarte,
L.F.L. Coelho, A.L.T. Dias, M.P. Veloso, D.T. Carvalho, D.F Dias, New eugenol
glucoside-based derivative shows fungistatic and fungicidal activity against op-
portunistic Candida glabrata, Chem. Biol. Drug Des. 87 (2016) 83–90, doi:10.
1111/cbdd.12625.
[26] H.M. Berman, T. Battistuz, T.N. Bhat, W.F. Bluhm, P.E. Bourne, K. Burkhardt,
Z. Feng, et al., The protein data bank, Acta Crystallogr. Sect. D 58 (2002) 899–
907, doi:10.1107/S0907444902003451.
[27] D. Bajusz, A. Rácz, K. Héberger, Why Is tanimoto index an appropriate choice
for fingerprint-based similarity calculations? J. Cheminform 7 (1) (2015) 1–13,
doi:10.1186/s13321-015-0069-3.
[28] C.C. Wu, T.K. Li, L. Farh, L.Y. Lin, T.S. Lin, Y.J. Yu, T.J. Yen, C.W. Chiang, N.L. Chan,
Structural basis of type II topoisomerase inhibition by the anticancer drug
etoposide, Science 333 (6041) (2011) 459–462, doi:10.1126/science.1204117.
[29] M. Gordaliza, P.A. García, J.M.M. Corral, M.A. Castro, M.A Gómez-Zurita,
Podophyllotoxin: distribution, sources, applications and new cytotoxic deriva-
tives, Toxicon 44 (4) (2004) 441–459, doi:10.1016/j.toxicon.2004.05.008.
[11] M. Hunsen, D.A. Long, C.R. D’ardenne, A.L Smith, Mild one-pot preparation of
glycosyl bromides, Carbohydr. Res. 340 (2005) 2670–2674, doi:10.1016/j.carres.
2005.09.016.
[12] D.M. Hall,
A practical synthesis of methyl 4,6-O-benzylidene-α and β-
[30] S. Forli, Charting
a path to success in virtual screening, Molecules 20 (10)
d-glucopyranoside, Carbohydr. Res. 86 (1980) 158–160, doi:10.1016/
S0008-6215(00)84593-0.
(2015) 18732–18758, doi:10.3390/molecules201018732.
[31] J. Gareth, P. Willett, R.C. Glen, A.R. Leach, R. Taylor, Development and validation
of a genetic algorithm for flexible docking, J. Mol. Biol. 267 (1997) 727–748,
doi:10.1006/jmbi.1996.0897.
[13] Balogh, M.P. (2004) Debating resolution and mass accuracy in mass spectrom-
etry, 19, 34–40.
[14] M.W. Lago, M.L. Friedrich, G.D. Iop, T.B. Souza, P.A. Mello, A.I.H Adams, Capil-
lary zone electrophoresis method to assay tipranavir capsules and identifica-
tion of oxidation product and organic impurity by quadrupole-time of flight
mass spectrometry, Talanta 181 (2018) 182–189, doi:10.1016/j.talanta.2018.01.
012.
[32] K.T. Schomburg, S. Bietz, H. Briem, A.M. Henzler, S. Urbaczek, M. Rarey, Facing
the challenges of structure-based target prediction by inverse virtual screening,
J. Chem. Inf. Model. 54 (6) (2014) 1676–1686, doi:10.1021/ci500130e.
[33] M.D. Hanwell, D.E. Curtis, D.C. Lonie, T. Vandermeersch, E. Zurek, G.R. Hutchi-
son, Avogadro: an advanced semantic chemical editor, visualization, and anal-
ysis platform, J. Cheminform. 4 (1) (2012) 1–17, doi:10.1186/1758-2946-4-17.
[15] CLINICAL AND LABORATORY STANDARDS INSTITUTE, Reference method for
broth dilution antifungal susceptibility testing of yeastsApproved Standard-
-Third Edition. M27-S4, CLSI, Wayne, PA, USA, 2012 2012.
[16] CLINICAL AND LABORATORY STANDARDS INSTITUTEReference Method for
Broth Dilution Antifungal Susceptibility Testing of Yeasts; Approved Standard -
Third edtion. CLSI Document M27-A3, Clinical Laboratory Standards Institute,
Wayne, USA, 2008.
11