Phenylpropiophenone derivatives as potential anticancer agents: Synthesis, biological evaluation and quantitative structure-activity relationship study

Article abstract of DOI:10.1016/j.ejmech.2013.02.013

Series of twelve chalcone and propafenone derivatives has been synthesized and evaluated for anticancer activities against HeLa, Fem-X, PC-3, MCF-7, LS174 and K562 cell lines. The 2D-QSAR and 3D-QSAR studies were performed for all compounds with cytotoxic

Full text of DOI:10.1016/j.ejmech.2013.02.013

Accepted Manuscript
Phenylpropiophenone derivatives as potential anticancer agents: Synthesis, biological
evaluation and quantitative structure-activity relationship study
Branka M. Ivković, Katarina Nikolic, Bojana B. Ilić, Željko S. Žižak, Radmila B.
Novaković, Olivera A. Čudina, Sote M. Vladimirov
PII:
S0223-5234(13)00099-8
10.1016/j.ejmech.2013.02.013
EJMECH 6003
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 21 November 2012
Revised Date: 8 February 2013
Accepted Date: 12 February 2013
Please cite this article as: B.M. Ivković, K. Nikolic, B.B. Ilić, Ž.S. Žižak, R.B. Novaković, O.A. Čudina,
S.M. Vladimirov, Phenylpropiophenone derivatives as potential anticancer agents: Synthesis, biological
evaluation and quantitative structure-activity relationship study, European Journal of Medicinal
Chemistry (2013), doi: 10.1016/j.ejmech.2013.02.013.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Phenylpropiophenone derivatives as potential anticancer agents:
Synthesis, biological evaluation and quantitative structure-activity
relationship study
1*
1
2
3
Branka M. Ivković , Katarina Nikolic , Bojana B. Ilić , Željko S. Žižak , Radmila
4
1
1
B. Novaković , Olivera A. Čudina , and Sote M. Vladimirov
1
Department of Pharmaceutical Chemistry, Faculty of Pharmacy, University of
Belgrade, Vojvode Stepe 450, 11000 Belgrade Serbia;
2
Institute of Endocrinology, Diabetes and Metabolic Diseases, Medical School,
University of Belgrade, Dr Subotica 13, 11000 Belgrade, Serbia;
3
Institute of Oncology and Radiology of Serbia, Pasterova 14, 11000 Belgrade,
Serbia;
4
Department of Pharmacology, Clinical Pharmacology and Toxicology, Faculty of
Medicine, University of Belgrade, Dr Subotica 8, 11000 Belgrade, Serbia
*
Corresponding author:
Branka Ivkovic,
Department of Pharmaceutical Chemistry, Faculty of Pharmacy, University of
Belgrade , Vojvode Stepe 450, 11221 Belgrade-Kumodraz, P.O.Box 146, Serbia.
e-mail: blucic@pharmacy.bg.ac.rs
Phone : + 381 11 3951-335, Fax : + 381 11 3972-840
1

ACCEPTED MANUSCRIPT
Abstract
Series of twelve chalcone and propafenone derivatives has been
synthesized and evaluated for anticancer activities against HeLa, Fem-X, PC-3,
MCF-7, LS174 and K562 cell lines. The 2D-QSAR and 3D-QSAR studies were
performed for all compounds with cytotoxic activities against each cancer cell
line. Partial least squares (PLS) regression has been applied for selection of the
most relevant molecular descriptors and QSAR models building. Predictive
potentials of the created 2D-QSAR and 3D-QSAR models for each cell line were
compared, by use of leave-one-out cross-validation and external validation, and
optimal QSAR models for each cancer cell line were selected. The QSAR studies
have selected the most significant molecular descriptors and pharmacophores of
the chalcone and propafenone derivatives and proposed structures of novel
chalcone and propafenone derivatives with enhanced anticancer activity on the
HeLa, Fem-X, PC-3, MCF-7, LS174 and K562 cells.
Keywords: Propafenone derivatives, Chalcones, QSAR, anticancer agents
2

ACCEPTED MANUSCRIPT
1
. Introduction
Cancer is the second leading cause of human death . Although there are
[
1]
many therapeutic strategies, including chemotherapy and radiotherapy, high
systemic toxicity and drug resistance are the main problems in the treatment of
cancer. Therefore, scientists are focused on finding new therapeutic targets for
[
2]
cancer and discovering novel potent and selective antineoplastic drugs .
Since naturally occurring chalcones and their synthetic analogues, such as
propafenone derivatives, displayed significant cytotoxic activity against various
[
3-6]
tumor cells
the main aim of this research was to develop new propafenone
derivatives and chalcones with enhanced anticancer activity.
Apart from antineoplastic activity, chalcones displayed considerable
chemoprotective, antiangiogenic, antibacterial, antifungal, antiparasitic,
[
4]
antioxidative, and anti-inflammatory activity
.
Chemically, chalcones are 1, 3-diaryl-2-propen-1-ones with an enone
system between two aromatic rings (Figure 1). They are precursors for the
synthesis of various compounds (flavonoids, propafenone etc.) which exhibit
[
7-10].
interesting pharmacological activities
Previously performed chalcones studies have shown that its ketovinyl
moiety (-C(O)-C=C-) acts as Michel acceptor that preferential covalently binds
[
11]
reactive sulfhydryl (-SH) groups of biomolecules
in contrast to amino and
(Figure1). Therefore chalcones may
cause lower incidence of typical side effects associated with a number of
[
12-13]
hydroxy groups of the biomolecules
[
14]
alkylating agents used in chemotherapy
. One of the proposed anticancer
3

ACCEPTED MANUSCRIPT
mechanisms of chalcones is based on their structure similarity with colchicine.
Actually, anticancer activity of chalcones is the result of their binding to tubulin
[
15]
and prevention of tubulin polymerization into microtubules
.
On the other side, propafenone (Figure 1) belongs to Ic class of
antiarrhythmic drugs that block the voltage-dependent cardiac sodium channels.
Furthermore, propafenone inhibits: L-type of calcium channels, potassium
[
16]
channels, and β-adrenergic receptors in the heart
; the growth of cultured
[
17]
Plasmodium falciparum
and the membrane glycoprotein (P-glycoprotein
[
5, 6, 18, 19]
(PGP)) involved in mediated drugs resistance (MDR)
.
Earlier pharmacological studies discovered that ion channels blockers
[
20-22]
selectively impair proliferation of cancer cells
. These observations were
further confirmed by comprehensive studies on the functional implications of ion
[
23]
channels in the biology of cancer cells
. Expression and activity of different
channel types mark and regulate specific stages of cancer establishment and
[
24]
progression . Since more attention is recently paid on discovery of anticancer
[25],
[20]
drugs that act as ion channels modulators
or blockers
, the propafenone
[
17, 26-27]
became good lead for design of novel antineoplastics
.
Previous SAR and QSAR studies of propafenone derivatives have
selected the ethers, carbonyl and benzyl groups as well as the substitution
pattern at the nitrogen atom (Figure 1) as determinants for pharmacological
[
18, 28]
activity of the propafenone and its derivatives
.
Because of the previous finding that ether group of the propafenone
[
18, 28]
derivatives is very important for their anticancer activity
the aim of our
4

ACCEPTED MANUSCRIPT
research work was to develop chalcones with hydroxyl group at the
corresponding position (Scheme1), as compounds with potentially enhanced
anticancer activity.
Expression and activity of different channel types mark and regulate
specific stages of cancer progression in the cancer cells. New findings indicated
that inner pore of the ion channels is formed from different aromatic and nonpolar
[
20-25].
aliphatic aminoacid residues
Since benzyl moiety of the propafenone
[
18, 28]
derivatives is also very important for their antineoplastic activity
the
hydrogen atom of the benzyl moiety of propafenone and chalcone was
substituted with methyl-, trifluoromethyl-, fluoro-, and chloro- substituents
(Scheme1). All these groups are lipophilic with different electronic and steric
effects. These new substituents case significant changes of lipophilicity and
stereoelectronic properties of the benzyl moiety. Therefore, the substitution of H-
atom by –F, -Cl atoms or –CH , -CF atomic groups greatly change the
3
3
conformational preferences of the chalcone and propafenone derivatives. These
structural modifications could result in formation of stronger intermolecular bonds
of the benzyl moiety of the chalcone/propafenone derivatives with hydrophobic
aminoacid residues in the inner pore of the ion channels and enhanced
anticancer activities. The changes in the ion channels activities cased by the
chalcone/propafenone derivatives could have beneficial effect on anticancer
activity of the other antineoplastics.
Also, these structural modifications of the chalcone molecules result in
changes of lipophilicity and electron density on the double bond in ketovinyl
5

ACCEPTED MANUSCRIPT
moiety and therefore have influence on the reaction with –SH groups of biological
activates proteins. Introduction of electron donating groups(such as -CH ) in the
3
ortho or para position in the benzyl moiety increase electron density of carbon
atoms in the double bonds that result in decrease of reactivity and activity of the
compounds. But, introduction of electron withdrawing groups (such as -F, -Cl and
–
CF ) in the ortho or para position in the benzyl moiety decrease electron
3
density carbon atoms in the double bonds that increase reactivity and activity of
the compounds.
So far, modifications of propafenone structure were performed on amino
and ether moiety of the propafenone. These changes have an influence on
[
26-28]
lipophilicity and ionization of the compounds
. Since recent studies of ion
channel blockers pointed on importance of the hydrophobic interactions between
[
24]
the ligand and the ion channel inside the pore
, our synthetic route was
focused on modification of the benzyl moiety of propafenone and its
intermediary’s chalcones.
Cytotoxic activities of twelve synthesized chalcone and propafenone
derivatives were assayed against HeLa, Fem-X, K562, MCF-7, PC-3 and LS174
cancer cell lines. The QSAR studies of the examined chalcones and propafenone
analogs have selected molecular descriptors and pharmacophores of the
compounds responsible for their anticancer activities.
2
. Results and Discussion
In this paper, the synthesis, anticancer activity of twelve 1,3-diphenyl-
propan-1-one derivates as well as 2D and 3D QSAR studies aimed to rationalize
6

ACCEPTED MANUSCRIPT
structural characteristics needed for significant potency, are reported. The set
consist of six chalcones and six propafenone derivatives and their anticancer
activities were evaluated toward six human tumor cell lines. The synthetic rout
and structures of proposed chalcones and propafenone derivatives are given in
′
Scheme1. Six simple 2 -hydroxychalcones bearing mono substituent in their ring
B (methyl-, chloro-, fluoro- and trifluoromethyl-) were synthesized by Claisen-
[
29]
Schmidt method
and evaluated for their anticancer in vitro activity. Due to the
rapid development of tumor cells resistance to antiproliferative agents, it is vital to
discover novel scaffold for the design and synthesis of the new low-molecular
antiproliferative agents to help in the battle against tumor cell lines. These
compounds, except 1a, are known in the literature, but their antiproliferative
activity has not been investigated.
Since recent studies of ion channel blockers pointed on importance of the
[
hydrophobic interactions between the ligand and the ion channel inside the pore
24]
,
our synthetic route was focused on modification of the benzyl moiety of
propafenone introducing of trifloromethyl, fluoro, chloro- or methyl groups in the
ortho or para position.
2
.1. Synthesis of chalcones and propafenone derivatives
Synthesis of the chalcones and novel propafenone derivatives is shown in
the Scheme 1. The proposed derivatives were synthesized in five steps and their
1
13
structures have been verified by IR, H-NMR, C-NMR, ESI-MS and HRMS
spectroscopy. The first step in the synthetic route consisted of the condensation
of 2-hydroxy acetophenone with ortho or para mono substituted benzaldehydes
7

ACCEPTED MANUSCRIPT
in basic conditions (10% aqueous NaOH) to give α,β-unsaturated ketones
(chalcones, Scheme 1). The use of catalytic hydrogenation with a 5% Pd/C and
careful monitoring of the H -consumption resulted in the formation of the
2
saturated ketone structures (ortho-hydroxyphenones, 2a-f). The addition of
phenolic nucleophiles to (±) epichlorohydrin in basic media led to the formation of
aryloxyepoxipropane derivatives (3a-f) which were subjected to the aminolysis
with propylamine to give the aryloxypropanolamine derivatives (amines). The
formation of the hydrochlorides is performed by dissolving the amine in dry
diethyl ether and adding 1M HCl solution in dry diethyl ether. The white
precipitate of hydrochloride was filtered off and purified via crystallization
(Scheme 1).
2
.2. Cytotoxic studies
The antiproliferative activity of all given compounds was tested against six
tumor cell lines: human cervical carcinoma (HeLa cells), human melanoma (Fem-
x), human prostate cancer (PC-3), human breast cancer (MCF-7), human colon
carcinoma (LS174 cells) and human chronic myelogenous leukemia (K562 cells).
All compounds were also screened for cytotoxic activity against normal human
peripheral blood mononuclear cells (PBMC) non-stimulated and stimulated with
phytohaemaglutinin (PHA).
All compounds exerted a dose dependent antiproliferative action at
micromolar concentrations toward investigated tumor cell lines, detected by MTT
test. The results are given in Table 1. All compounds showed activity to all six cell
lines, but HeLa, Fem-X and K562 cells were generally the most susceptible ones
8

ACCEPTED MANUSCRIPT
with IC values similar to cisplatin for some derivatives e.g. 5d, 1c, 5a. Cisplatin
50
was used as positive control.
In order to investigate antiproliferative activity, the synthesized compounds were
divided in two groups: 1) propafenone derivates and 2) intermediate chalkones.
Propafenone derivates show better efficacy in inhibition of cancer cell
proliferation (Table 1). According to results in Table 1, it is possible to make
comparative efficacy ranking of derivates on certain human cancer cell lines.
HeLA: 5d > 5a > 5b > 5c = 5e > 5f
Fem X: 5d > 5b > 5e > 5f > 5a > 5c
K562: 5d > 5b ~ 5a > 5f > 5e > 5c
MCF-7: 5a > 5c ~ 5d> 5b> 5f~ 5e
PC-3: 5d > 5b > 5a > 5f > 5e > 5c
LS174: 5d > 5a ~ 5b > 5f > 5e > 5c
Ortho-substituted derivates show better efficacy than para-substituted analogs
and the most potent is 5d derivate. Introduction of electronegative, lipophilic
chloro group in ortho position of benyzl moiety contributes better efficacy of this
derivate compared to other sythesized compounds on five of six cancer cell lines.
30
In our previous studies we have demonstrated that propafenone derivates with
substituted terminal benzene ring (benzyl moiety) interact with ion channels and
show better selectivity compared to lead molecule (propafenone) Those effects
.
are confirmed in presence of selective ion channel blockers. Para-substituted
derivates showed better potassium channel selectivity, while the effects of ortho-
substituted derivates were not antagonized by potassium channel blockers.
9

ACCEPTED MANUSCRIPT
Ortho-substituted derivates (5d and 5a) showed selective effect on calcium
channels. In vitro studies with 5d derivate on isolated rat aorta (unpublished
results) showed that effect of 5d at concentrations lower than 10 ꢀM could be
antagonized by channel blocker, which means that the effect is not selective.
This fact may support our hypothesis that 5d derivate in lower concentration
takes effect as a nonselective ion channel blocker. For that reason, this derivate
is more potent than others, because cancer cells expresses various sub/types of
channels depending on cell cycle phase.
Compuond 5d showed the best activity on HeLa cell culture line (3 ꢀmol/l). HeLa
cell line includes calcium permeability channels with nonselective permeability to
potassium and other monovalent ions. Those channels are not voltage-
[
20, 24-25]
dependent and show the characteristics of receptors channels
. It can be
assumed that 5d derivate exerts its antiproliferative effect by disrupting calcium
homeostasis affected by potassium and calcium channels.
[
30]
In vitro
and in silico studies (unpublished results) showed potential potassium
blocker activity of ortho-fluorinated (5c) derivate. However, 5c derivate shows
three to six times less effect than 5d derivate. There is no difference in effects of
this derivate on normal and cancer cells (Table 2). Ion channels in cancerous
cells differ from channels in normal cells. In this case, electronegativity of fluorine
adversely affects the interactions of terminal benzene ring with channel
aminoacids residues in cancerous cells. The exception is MCF-7 breast cancer
line. The efficacy of 5d derivate is less than 5a, but almost equal as 5c. Those
[
20, 24]
cells expresses calcium and BKCa channels
. In vitro studies on rat aorta
1
0

ACCEPTED MANUSCRIPT
showed better calcium channel selectivity of 5a derivate in comparison to other
five derivates. Consequently, its activity on MCF-7 cell line is expected. Activity,
as well as selectivity of investigated chalkones are less compared to propafenone
derivates, due to chalkones non specific effect. According to literature data,
activity of chalkones on HeLa, PC-3 and MCF-7 cell lines (in concentration range
[
4]
from nmol/l to mmol/l) depends on their structure . Test results show satisfactory
anticancer activity of chalkones in concentration range 9-70 ꢀmol/l. This activity
is lower or equivalent to activity described in literature, due to simple chemical
structure of synthesized chalkones. It would mean that keto-vinyl group and its
substituents are essential for anticancer activity.The efficacy of chalkones on all
cell culture lines are approximate, which confirms chalkones non specific
mechanism of action. Activity ranking follows electronegativity of substituents in
the benzyl moiety.
HeLA: 1c > 1d > 1b ~ 1a > 1f > 1e
Fem X: 1c > 1d = 1a > 1e = 1f ~ 1b
K562: 1c > 1a > 1d > 1e ~ 1f ~ 1b
MCF-7: 1c > 1a > 1e > 1d > 1f > 1b
PC-3: 1b > 1d ~ 1c > 1f > 1e > 1a
LS174: 1c > 1a > 1d > 1e > 5f > 1b
Electropositive carbon atoms in double bond are more reactive to nucleophilic -
[
4]
SH groups of aminoacid residues with which chalkones react .
The most potent chalkones compounds are 1OF (from 4.8 ± 0.6 on K562 to 21.6
±
3.5 on PC-3 cell line) and 1OCl (from 9.8 ± 1.4 on K562 to 32.2 ± 1.2 on
1
1

ACCEPTED MANUSCRIPT
LS174 cell line). The selectivity of chalkones derivates is less compared to
propafenone derivates. Selectivity index for 1OF derivate is 3.67 on MCF-7 cell
line, while is 0.81 on PC-3 cell line. 1CF3 derivate showed better selectivity (SI
values from 4.55 on K562 to 1.04 on PC-3 cell line).
2
.3. QSAR studies
The 2D-QSAR and 3D-QSAR studies were performed on twelve
synthesized chalcone and propafenone derivatives (Scheme 1) with cytotoxic
activities (Table 1) against six cancer cell lines (HeLa, Fem-X, K562, MCF-7, PC-
3
, and LS174).
The constitutional, geometrical, physico-chemical and electronic descriptors of
the optimised molecular models were used as independent variables, while
negative logarithm of cytotoxic activities (pIC = (log1/IC )) were used as
50
50
dependent variables for the QSAR studies. The PLS regression has been applied
for selection of the most relevant molecular descriptors and QSAR models
building.
The applied PLS methodology has selected sets of significant descriptors
during development of each QSAR model. Overfitting of the QSAR models is
avoided by two parallel procedures. First, during the QSAR modelling the RMSE
for training and test set were monitored. The optimal QSAR models were
selected when RMSEP of test set begins to increase while RMSEE of training set
continues do decrease. Second, the formed QSAR models with different number
of selected significant descriptors were compared and optimal one is chosen by
2
2
2
comparing R , Q , RMSEE, R Obs vs. Pred, and RMSEP.
1
2

ACCEPTED MANUSCRIPT
Optimal 2D-QSAR models for each cell line (HeLa, Fem-X, K562, MCF-7, PC-
2
3
and LS174) were selected by use of statistical parameters such as: R , F ratio,
P-value, RMSEE of the training set, the leave-one-out cross-validation parameter
2
2
(
Q ), and external validation test set (Fig. 2) (R
and RMSEP (Tables 3-
Obs vs. Pred
8
).
The 2D-QSAR (HeLa) model relating five variables (HOMO, BIC4, and
2
2
RDF145s), with one significant component (A=1), R : 0.824, Q (Y): 0.772, F ratio:
5.240, P-value: 0.001, RMSEE: 0.103, and RMSEP: 0.179 (Table 3), was
formed.
The 2D-QSAR (Fem-X) model relating two variables (Mor08u and Mor27i),
1
2
2
with one significant component (A=1), R : 0.647, Q (Y): 0.585, F ratio: 6.340, P-
value: 0.019, RMSEE: 0.144, and RMSEP: 0.622 (Table 4), was created.
The 2D-QSAR (K562) model relating two variables (Mor06e and CMC-50),
2
2
with one significant component (A=1), R : 0.639, Q (Y): 0.587, F ratio: 6.394, P-
value: 0.018, RMSEE: 0.155, and RMSEP: 0.232 (Table 5), was selected.
The 2D-QSAR (MCF-7) model relating two variables (Eig02_AEA(dm), and
2
2
H5s), with one significant component (A=1), R : 0.822, Q (Y): 0.812, F ratio:
9.4464, P-value: 0.0005, RMSEE: 0.071, and RMSEP: 0.124 (Table 6), was
obtained.
The 2D-QSAR (PC-3) model relating two variables (Mor32s and G(O..Cl)),
1
2
2
with one significant component (A=1), R : 0.851, Q (Y): 0.591, F ratio: 6.513, P-
value: 0.017, RMSEE: 0.055, and RMSEP: 0.214 (Table 7), was formed.
1
3

ACCEPTED MANUSCRIPT
The 2D-QSAR (LS174) model relating three variables (RDF04s, G1e, and
2
2
G2u), with one significant component (A=1), R : 0.609, Q (Y): 0.555, F ratio:
.608, P-value: 0.026, RMSEE: 0.133, and RMSEP: 0.266 (Table 8), was
created.
5
2
2
In order to investigate the statistical significance of the R (Y) and Q (Y) and to
test the model for overfitting due to the chance correlation the response
[
43, 45]
permutation test (Y scrambling) is applied for all 2D-QSAR models
. The
2
2
R (Y) – intercepts were less than 0.4 while the Q (Y)-intercepts were less than
[
43, 45]
0
.05
The 3D-QSAR studies of the cytotoxic activities of the chalcones and
propafenone derivatives against six cancer cell lines (HeLa, Fem-X, K562, MCF-
, PC-3 and LS174) were started from computing highly relevant MIF-3D maps of
interaction energies between the examined molecule and four chemical probes
DRY, O, N1 and TIP). These maps are encoded into GRIND and GRIND2
for the six developed 2D-QSAR models.
7
(
descriptors and compressed into new types of variables- with the intensity of the
field at a node and the mutual node−node distances between the chosen nodes
as a scoring function. The obtained variables create matrix of descriptors that
were used for development of 3D-QSAR models by use of the PLS regression.
Quality of the obtained 3D-QSAR models was examined by use of the
2
2
regression factor R , the leave-one-out cross-validation parameter (Q ), RMSEE
2
of the training set, RMSEP of the test set, and R
of the test set (Fig. 2)
Obs vs. Pred
(Tables 3-8).
1
4

ACCEPTED MANUSCRIPT
2
The 3D-QSAR (HeLa) model with three significant components (A=3), R :
2
0
.97, Q (Y): 0.75, RMSEE: 0.054, and RMSEP: 0.232 was created. The variables
v366: DRY-N1, v301: DRY-O, and v568: O-TIP are positively correlated with the
activity, while V155/v160: N1-N1 variables are negatively correlated with activity
against HeLa cells.
The propafenone derivatives with substituent at ortho position of the benzyl
moiety (5d, 5a, and 5b) were the most potent against HeLa cells. The strongest
positive influences on cytotoxic activity of propafenone derivatives against HeLa
cells were determined for O-TIP descriptors (v565, v566, v568), created between
H-bond donating hydroxyl/or amino group and topological area at positions C4
and C5 of the benzyl group (Fig. 3A,3B); O-DRY descriptor (v301) formed
between H-bond donating hydroxyl/or amino group and hydrophobic area of the
benzyl moiety (Fig 3A, 3B); and N1-DRY descriptor (v366) created between H-
bond accepting oxygen of hydroxyl group and hydrophobic area of the benzyl
moiety (Figure 3A). The strongest negative influence on cytotoxic activity of
propafenone derivatives against HeLa cells was determined for N1-N1
descriptors (v155/v160) created between H-bond accepting carbonyl and ether
groups (Fig. 3B). Finally is concluded that the interactions between hydroxyl/or
amino group with topological and hydrophobic regions of the benzyl moiety are
responsible for stronger anticancer activity of the propafenone derivatives, while
specific relations between H-bond accepting carbonyl and ether groups are able
to decrease cytotoxic activity of the propafenone derivatives against HeLa cell
lines.
1
5

ACCEPTED MANUSCRIPT
The 3D-pharmacophore (HeLa) study indicated that further modification of the
most active 5d propafenone derivative should start with substitution with bulk and
hydrophobic substituents at the positions C4 or C5 of the benzyl moiety (O-TIP:
v565, v566, v568; O-DRY: v301; N1-DRY: v366).
2
The 3D-QSAR (Fem-X) model with four significant components (A=4), R :
2
0
.98, Q (Y): 0.58, RMSEE: 0.056, and RMSEP: 0.265 was developed. The
variables v252: TIP-TIP, v367: DRY-N1, and v631: N1-TYP are positively
correlated with the activity, while v164: N1-N1 variable is negatively correlated
with activity against Fem-X cells.
The 5d, 5b, 1c, and 5e derivatives were the most potent against Fem-X cells.
The strongest positive influences on cytotoxic activity of propafenone derivatives
against Fem-X cells were observed for N1-TIP descriptor (v631), created
between H-bond accepting carbonyl group and topological area of aliphatic chain
at the nitrogen atom (Fig. 4A); N1-DRY descriptor (v367) formed between H-
bond accepting carbonyl group and hydrophobic area of aliphatic chain at the
nitrogen atom (Fig. 4A); and TIP-TIP variable (v252) created between position 4
of the benzyl moiety and propylamino moiety (Fig. 4A). The strongest negative
influence on cytotoxic activity of the propafenone derivatives against Fem-X cells
was determined for N1-N1 descriptors (v164) created between H-bond accepting
nitrogen atom and fluorine substituent in the benzyl moiety (Fig. 4B). Finally is
concluded that the specific interactions between carbonyl group with topological
and hydrophobic regions of the aliphatic chain at the nitrogen atom are
responsible for stronger anticancer activity of the propafenone derivatives, while
1
6

ACCEPTED MANUSCRIPT
specific relations between H-bond accepting nitrogen atom and fluorine
substituent in the benzyl moiety can attenuate cytotoxic activity of the
propafenone derivatives against Fem-X cell lines (Fig. 4). The 3D-
pharmacophore (Fem-X) study indicated that benzyl moiety at position C4 and
aliphatic chain at the nitrogen atom of the propafenone molecule should be
modified with bulk and hydrophobic groups (N1-TIP: v631, N1-DRY: v367, TIP-
TIP: v252), Also, substituents with very strong electronegativity, such as fluorine,
should be avoided (N1-N1: v164).
2
The 3D-QSAR (K562) model with two significant components (A=2), R : 0.93,
2
Q (Y): 0.62, RMSEE: 0.052, and RMSEP: 0.288 was generated. The variables
v633: N1-TIP, v532: O-TIP, and v252: TIP-TIP are positively correlated with the
activity, while v164: N1-N1, v240: TIP-TIP, v265: DRY-O, and v425: DRY-TIP
variables are negatively correlated with activity against K562 cells.
The 1c, 5d, 5b, 5a, and 1a compounds were the most potent against K562
cells. The strongest positive influences on K562 anticancer activity of the agents
were determined for TIP-TIP descriptor (v252) created between benzyl moiety
and topological area of aliphatic chain at the nitrogen atom (Fig. 5A); O-TIP
descriptor (v532) created between H-bond donating hydroxyl group and
topological area at meta position C4’ (chalcone derivatives)/or aliphatic chain at
the nitrogen atom (propafenone derivatives) (Fig. 5A, 5B, 5C); and N1-TIP
descriptor (v633) created between H-bond accepting nitrogen atom and
topological area of the benzyl moiety of the propafenone derivatives (Fig. 5A).
The strongest negative influence on cytotoxic activity of the compounds against
1
7

ACCEPTED MANUSCRIPT
K562 cells were determined for: N1-N1 descriptor (v164) created between H-
bond accepting hydroxyl group and fluorine substituent (Fig. 5C); TIP-TIP
variable (v240) formed between two benzene rings (positions C4 and C4’) (Fig.
5
B, 5C), and DRY-TIP descriptor (v425) created between two benzene rings or
between benzene ring and aliphatic chain at the nitrogen atom (Fig. 5A and 5B).
From this results is concluded that the specific interactions of the aliphatic chain
at the nitrogen atom with benzyl moiety and with hydroxyl group are mainly
responsible for anticancer activity of the propafenone derivatives, while specific
interactions between H-bond accepting hydroxyl group and fluorine substituent
and between topological areas at positions C4 and C4’ of the two benzene rings
have negative influence on cytotoxic activity of the propafenone derivatives
against K562 cells. According to the 3D-pharmacophore (K562) study further
synthesis of the propafenone derivatives should start with substitution with a bulk
group at position-C2 of the benzyl moiety (N1-TIP: v633), while positions C4 and
C4’ of the two benzene rings should be unsubstituted (TIP-TIP: v240, DRY-TIP:
v425). Also, substituents with very strong electronegativity, such as fluorine,
should be avoided (N1-N1: v164).
The most important positive impact on anticancer activity of chalcone
derivatives is determined for interaction between H-bond donating hydroxyl group
and topological area at C4’ position of the same benzene ring (O-TIP: v532),
while relations between H-bond accepting hydroxyl group and fluorine substituent
(N1-N1: v164) together with interactions between two benzene rings (TIP-TIP:
v240, DRY-TIP: v425) have negative influence on anticancer activity of the
1
8

ACCEPTED MANUSCRIPT
chalcone derivatives against K562 cells. According to the 3D-pharmacophore
(K562) study further synthesis of the chalcone derivatives should start with
substitution at the positions C2 or C6 of the phenyl group with electronegative
substituents (O-TIP: v532), while positions C4 and C4’ of the two benzene rings
should be unsubstituted (TIP-TIP: v240, DRY-TIP: v425).
2
The 3D-QSAR (MCF-7) model with two significant components (A=2), R :
2
0
.95, Q (Y): 0.80, RMSEE: 0.030, and RMSEP: 0.314 was obtained. The
variables v548: O-TIP, v30: DRY-DRY, and v439: DRY-TIP are positively
correlated with the activity, while v143: N1-N1 variable is negatively correlated
with activity against MCF-7 cells.
The chalcone and propafenone derivatives with –F or –CF substituent at
3
ortho position (5c, 5a, 1c, and 1a) were the most potent against MCF-7 cells. The
strongest positive influences on cytotoxic activity of compounds against MCF-7
cells were determined for TIP-DRY descriptor (v439) formed between topological
area at positions 4’ and 5’ of the benzene ring and hydrophobic region of aliphatic
chain at nitrogen atom (Fig. 6A); DRY-DRY descriptor (v30) created between
hydrophobic area of the benzyl moiety and aliphatic chain at nitrogen atom (Fig.
6
A); and O-TIP descriptors (v548) formed between H-bond donating hydroxyl
group and topological area at positions C4’ of the benzene ring and between H-
bond donating hydroxyl (phenol) group and topological area of aliphatic chain at
nitrogen atom (Fig. 6A, 6B). The strongest negative influence on cytotoxic activity
of the compounds against MCF-7 cells was determined for N1-N1 descriptors
(v143) created between H-bond accepting carbonyl and hydroxyl groups of the
1
9

ACCEPTED MANUSCRIPT
chalcone derivatives (Fig. 6B). Finally is concluded that the interactions between
aliphatic chain of the amino group with topological and hydrophobic regions of
the benzene ring (especially C4’ and C5’ positions) are responsible for stronger
anticancer activity of the propafenone derivatives against MCF-7 cell lines.
The 3D-pharmacophore (MCF-7) study indicated that further modification of
the most active 5a/5c propafenone derivatives should start with substitution with
bulk and hydrophobic substituents at the position C4’ or C5’ of the benzene ring
(TIP-DRY: v439, DRY-DRY: v30, O-TIP: v548). The 3D-pharmacophore (MCF-7)
study indicated that modification of the most active 1a/1c chalcone derivatives
should start with substitution with bulk and hydrophobic groups at the position 4
or 5 of the B-benzene ring (O-TIP).
2
The 3D-QSAR (PC-3) model with two significant components (A=2), R : 0.96,
2
Q (Y): 0.75, RMSEE: 0.018, and RMSEP: 0.219 was formed. The variables v36:
DRY-DRY, v294: DRY-O, and v252: TIP-TIP are positively correlated with the
activity, while v157: N1-N1 variable is negatively correlated with activity against
PC-3 cells.
The chalcone and propafenone derivatives with –Cl or –CH substituent at
3
ortho position (5d, 5b, 1d, and 1b) were the most potent against PC-3 cells. The
strongest positive influences on cytotoxic activity of the propafenone and
chalcone derivatives against PC-3 cells were determined for DRY-O descriptor
(v294) formed between hydrophobic region of the benzyl moiety (B-benzene ring)
and H-bond donating hydroxyl (phenol) group (Fig. 7A, 7B, 7C); TIP-TIP
descriptor (v252) created between topological area at positions C4’, C4, and C5
2
0

ACCEPTED MANUSCRIPT
of the benzene rings and aliphatic chain at nitrogen atom (Fig. 7A, 7B); and DRY-
DRY descriptor (v36) created between hydrophobic area of the benzyl moiety
and aliphatic chain at nitrogen atom (Fig. 7A, 7B). The strongest negative
influence on cytotoxic activity of the compounds against PC-3 cells was
determined for N1-N1 descriptors (v157) created between H-bond accepting
carbonyl and amino groups of the propafenone derivatives (Fig. 7A, 7B). Finally
is concluded that the interactions between aliphatic chain of the amino group with
topological and hydrophobic regions of the benzene ring (especially C4’, C4 and
C5 positions) are responsible for stronger anticancer activity of the propafenone
derivatives against PC-3 cell lines.
The 3D-pharmacophore (PC-3) study indicated that further modification of the
most active 5d/5b propafenone derivatives should start with substitution with bulk
and hydrophobic groups at the position C4’, C4 or C5 of the benzene rings (TIP-
TIP: v252, DRY-DRY: v36), together with substitution with hydrophobic and
electronegative group in aliphatic chain at nitrogen atom (TIP-TIP: v252, DRY-
DRY: v36, DRY-O: v294). The 3D-pharmacophore (PC-3) study indicated that
modification of the most active 1b/1d chalcone derivatives should begin with
substitution with hydrophobic group in the B-benzene ring (DRY-O: v294).
2
The 3D-QSAR (LS174) model with three significant components (A=3), R :
2
0
.96, Q (Y): 0.76, RMSEE: 0.057, and RMSEP: 0.177 was created. The variables
v366: DRY-N1, and v632: N1-TIP are positively correlated with the activity, while
v226: TIP-TIP variable is negatively correlated with activity against LS174 cells.
2
1

ACCEPTED MANUSCRIPT
The 5d, 1c, 5a, 5b, and 5f derivatives were the most active against LS174
cells. The strongest positive influences on cytotoxic activity of the propafenone
derivatives against LS174 cells were determined for DRY-N1 descriptor (v366)
formed between hydrophobic region of the benzyl moiety and H-bond accepting
hydroxyl group (Fig. 8A, 8B); and N1-TIP descriptors (v632) created between H-
bond accepting carbonyl/or hydroxyl group and topological area at positions
C5’/or C4 of the benzene rings (Fig. 8A, 8B). The strongest negative influence on
cytotoxic activity of the compounds against LS174 cells was determined for TIP-
TIP descriptors (v226) created between topological areas around hydroxyl group
and positions C4’ and C5’ of the benzene ring of the propafenone derivatives
(Fig. 8A, 8B). Finally is concluded that the interactions carbonyl/or hydroxyl group
with topological and hydrophobic regions of the benzene rings (especially C5’
and C4 positions) have influence on stronger anticancer activity, while topological
relations between hydroxyl group and positions C4’/or C5’ of the benzene ring of
the propafenone derivatives have negative impact on their activity against LS174
cells.
The 3D-pharmacophore (LS174) study indicated that further modification of
the most active 5d/5a propafenone derivatives should start with substitution with
bulk and hydrophobic groups at the position C4 of the benzyl moiety (N1-TIP:
v602, DRY-N1: v366, TIP-TIP: v226). Substitution of the positions C4’ and C5’ of
the benzene ring of the propafenone derivatives should be avoided because of
the unfavourable interaction (TIP-TIP: v226).
2
2

ACCEPTED MANUSCRIPT
Predictive potentials of the developed 2D-QSAR models for each cell line
HeLa, Fem-X, K562, MCF-7, PC-3 and LS174) were compared with the
(
2
2
2
corresponding 3D-QSAR models by use of R , Q , RMSEE, RMSEP and R
Obs vs.
of the test set (Fig. 2) (Tables 3-8).
Pred
The obtained 3D-QSAR models for the anticancer activities on HeLa, Fem-X,
K562, MCF-7, PC-3, and LS174 cell lines have expressed better statistical
2
2
2
parameters (R , Q , RMSEE, RMSEP and R
of the test set) than the
Obs vs. Pred
corresponding 2D-QSAR models (Tables 3-8).
The N1-TIP descriptors (v631-v633), created between H-bond accepting
carbonyl/or hydroxyl group and topological area at positions C5’/or C4 of the
benzene rings, have positive influence on activity against Fem-X, K562, and
LS174 cell lines (Fig. 4, 5, 8). The TIP-TIP descriptor (v252), created between
topological area at positions C4’, C4, and C5 of the benzene rings and aliphatic
chain at nitrogen atom, exert positive impact on anticancer activity on Fem-X,
K562, and PC-3 cells (Fig. 4, 5, 7). The DRY-N1 descriptor (v366/v367), formed
between hydrophobic region of the benzyl moiety and H-bond accepting hydroxyl
group, have positive influence on activity against HeLa, Fem-X, and LS174 cell
lines (Fig. 3, 4, 8). The O-TIP descriptors (v532, v548, v568), formed between H-
bond donating hydroxyl group and topological area at positions C4’ of the
benzene ring and between H-bond donating hydroxyl (phenol) group and
topological area of aliphatic chain at nitrogen atom, exert positive impact on
anticancer activity on HeLa, K562, and MCF-7 cells (Fig. 3, 5, 6). The N1-N1
descriptors (v143-v164), created between two H-bond accepting groups, have
2
3

ACCEPTED MANUSCRIPT
negative influence on activity against HeLa, Fem-X, K562, MCF-7, and PC-3 cell
lines (Fig. 3-7). Generally, high level of similarity was observed for 3D-
pharmacophores of the examined compounds responsible for HeLa, Fem-X,
K562, and PC-3 anticancer activities.
3
. Conclusion
Series of twelve chalcone and propafenone derivatives has been
synthesized and evaluated for anticancer activities against HeLa, Fem-X, PC-3,
MCF-7, LS174 and K562 cell lines.
The relationship between chemical structures of the synthesized
compounds and their cytotoxic activities against the six cancer cell lines, were
examined by 2D-QSAR and 3D-QSAR studies. The developed QSAR models
have expressed good quality in activity prediction of the cross-validation and
external validation test set compounds.
The performed QSAR studies of the chalcone and propafenone
derivatives have selected molecular descriptors and pharmacophore features
responsible for their antineoplastic activities against six cancer cell lines. Results
of the 3D-QSAR studies were used for design of novel chalcones and
propafenone derivatives with enhanced anticancer activity on the HeLa, Fem-X,
PC-3, MCF-7, LS174 and K562 cells.
4
. Material and methods
4
.1. Chemistry
2
4

ACCEPTED MANUSCRIPT
All solvents and reagents used are analytical grade having >98% purity.
For the chemical synthesis and analysis of chalcones and propafenone
derivatives the following compounds were used: 2-fluorobenzaldehyde, 4-
fluorobenzaldehyde, 2- methylbenzaldehyde and 4-methylbenzaldehyde (Merck,
Darmstadt, Germany), 2-trifluoromethylbenzaldehyde, 2-hydroxyacetophenone,
2
-chlorobenzaldehyde, epichlorhydrine, propylamine (Sigma-Aldrich Inc., St.
Louis, MO, USA), methanol, toluene, n-hexane, hydrochloride acid, sodium
hydroxyde, ethanol (Sigma-Aldrich Inc., St. Louis, MO, USA). Melting points were
determined on the Boetius PHMK 05 apparatus (VEB Wagetechnik Rapido,
Radebeul, Germany) and are uncorrected. TLC was carried out using silica gel
plates (TLC Silica gel 60 F254 aluminium sheets, Merck Darmstadt, Germany) and
the spots were visualized under UV light at 254 nm. Flash chromatography was
performed for purification of the compoumds on silica gel 60 (70-230 mesh
ASTM) (Merck Darmstadt, Germany). The structure of each compound was
confirmed by IR, NMR and LC-MS spectroscopy. The IR spectra were recorded
on the Thermo Nicolet 6700 FT-IR spectrophotometer using ATR technique in
-1
the range 4000-600 cm . The NMR spectra were recorded on the Varian Gemini
1
13
2
00 ( H NMR spectra at 200 MHz and C NMR spectra at 50 MHz) for samples
in deuterated chloroform. Chemical shifts are expressed in ppm using
tetramethylsilane as the internal standard; coupling constants (J) are in Hz.
Spliting patterns are described as singlet (s), doublet (d), triplet (t) multiplet (m)
and broad signal (bs). The mass spectra of compounds were conducted on a
TSQ Quantum Access MAX triple quadripole spectrometer (Thermo Fisher
2
5

ACCEPTED MANUSCRIPT
Scientific) equipped with electrospray ionization (ESI) source operating in positive
ion mode. Xcalibur software v 2.1.0.1139 (Thermo Fisher Scientific) was used for
data acquisition and processing. The purity of the compounds was checked using
HPLC (Agilent 1200 Series HPLC system, Agilent Technologies, Palo Alto, CA
USA) and electrospray ionization-high resolution mass spectrometry (ESI/HRSM)
were carried out on the Agilent 6210 time-of-flight LC/MS system (G1969A,
Agilent Technology, Palo Alto, CA, United States). Agilent MassHunter
Workstation software was used for data acquisition. Purity of the compounds was
>
95%.
4
.1.1. General procedure for the synthesis of chalcones
General methods for the preparation of chalcones involve Claisen-Schimdt
condensation of appropriate 2-hydroxyacetophenones and benzylaldehydes in
presence of base (Scheme 1).
2
-hydroxyacetophenone (0,012 mol) and substituted benzaldehyde (0,01 mol)
are dissolved in 96 % ethanol (10 ml) with stirring. Solution of sodium hydroxide
10%, 20 g) in water was added in portions to give a blood-red solution that was
(
stirred overnight during which chalcones, precipitated as the sodium salts. The
solution/suspension was kept 24 h at 0 ºC with occasional shaking, diluted with
ice water and acidified with cold 1M HCl (pH about 3). The resulting yellow
precipitate was filtered, washed with water (to neutral reaction to lacmus paper)
and crude mixtures were purified by flash column chromatography with toluene
as eluent. After removal of toluene under vacuum the crude products
2
6

ACCEPTED MANUSCRIPT
precrystallized from ethanol. Trans isomers of propane moiety of chalcones were
confirmed by the coupling constants of vinyl hydrogens (15-16 Hz).
1
a (C H F O ): (E)-1-(2-hydroxyphenyl)-3-(2-(trifluoromethyl) phenyl) prop-2-
16 11 3 2
en-1-one. The yellow solid was recrystallized from ethanol. Yield: 88.6%; Melting
º
-1
point 111-114 C. IR (ATR) (cm ): 3438.9 (-OH), 1645.1(-C=O), 1586.1 (ArC-C-
1
)
,1205.6 (ArC-OH), 1154.7(ArC-F); H NMR data (δ, ppm): 12.66 (s, 1H, -OH),
'
'
8
.31 (d, 1H, J=15.72, b), 7.93-7.73 (m, 4H, ArH; 6, 4, 3, 5), 7.63-7.37 (m, 5H, a,
'
'
13
ArH: 3,5,4,6); C NMR data (δ, ppm):193.29; 163.70; 140.71; 136.76; 133.70;
1
1
2
32.15; 130.04; 129.78; 128.04; 126.43; 126.63; 126.32; 124.45; 119.77; 118.73;
+
18.97; LC-MS: m/z = 291.00 (M -1), 145.08, 251.14, 271.02, 263.08, 289.12,
-
31.10; HRMS: Measured mass for [M-H] is 291.06321, calculated mass for
C H F O is 291.06384.
16
10
3
2
1
b (C H O ): (E)-1-(2-hydroxyphenyl)-3-o-tolylprop-2-en-1-one. The yellow
16 14 2
º
solid was recrystallized from ethanol. Yield: 90.8%; Melting point 153-156 C. IR
-1
(
ATR) (cm ): 3015.3 (-OH), 1636.5 (-C=O), 1575.5 (ArC-C-), 1258.2(ArC-OH);
1
H NMR data (δ, ppm): 12.84 (s, 1H, -OH), 8.27 (d, 1H, J=15.60, b), 7.94 (dd,
'
'
1
H, J =8.20, J =1.40, ArH 6) , 7.73-7.56 (m, 3H, a, ArH:4, 6), 7.38-7.23 (m, 3H,
1 2
'
'
13
ArH: 3, 4, 5), 7.05-7.690 (m, 2H, ArH: 3, 5), 2.50 (s, 3H, -CH ); C NMR data (δ,
3
ppm):193.80; 163.65; 143.07; 138.65; 136.39; 133.63; 131.04; 130.66; 129.69;
1
26.54; 126.42; 121.19; 120.02; 118.86; 118.62; 19.81; HRMS: Measured mass
-
for [M-H] is 237.09267, calculated mass for C H O is 237.09210.
16
13
2
1
c (C H FO ): (E)-1-(2-hydroxyphenyl)-3-(2-fluorophenyl) prop-2-en-1-one.
15 11 2
The yellow solid was recrystallized from ethanol. Yield: 83.3%; Melting point 82-
2
7

ACCEPTED MANUSCRIPT
º
-1
8
3 C. IR (ATR) (cm ): 3438.0 (-OH), 1642.9 (-C=O), 1576(ArC-C-), 1203.2
1
(
ArC-OH), 1155.2 (ArC-F); H NMR data (δ, ppm):12.77 (s, 1H, -OH), 8.04 (d,
'
1
H, J=15.72, b), 7.94 (dd, 1H, J =8.14, J =1.68, ArH: 6), 7.82 (d, 1H, J=15.73,
1
2
'
'
a), 7.70 (td, 1H, J =7.58, J =1.69, ArH 4), 7.55-7.36 (m, 2H, ArH: 5, 6), 7.26-6.91
1
2
'
13
(
m, 4H, ArH; 3, 4, 5, 3).; C NMR data (δ, ppm):193.84; 164.45; 163.65;
1
1
59.37; 138.25; 136.54; 132.31; 132.15; 130.20; 130.15; 129.75; 124.63;
24.58; 122.94; 122.79; 119.97; 118.91; 118.64; 116.62; 116.18; HRMS:
-
Measured mass for [M-H] is 241.06615, calculated mass for C H FO is
15
10
2
2
41.06703.
d (C H ClO ): (E)-1-(2-hydroxyphenyl)-3-(2-chlorophenyl) prop-2-en-1-one.
1
15
11
2
The yellow solid was recrystallized from ethanol. Yield: 82.5%; Melting point 50-
º
-1
5
3 C. IR (ATR) (cm ): 1637.8 (-C=O), 1574.3(ArC-C-), 1203.4(ArC-OH), 1019.9
1
(
ArC-Cl); H NMR data (δ, ppm): 12.75 (s, 1H, -OH), 8.34 (d, 1H, J=15.17, b),
'
'
7
.93 (dd, 1H, J =7.76, J =1.68, ArH 6), 7.79-7.74 (m, 1H, ArH 4), 7.67 (d, 1H,
1 2
J=15.73, a), 7.55-7.42 (m, 2H, ArH:3, 6), 7.39-7.26 (m, 2H, ArH; 4, 5), 7.05-6.90
'
'
13
(
m, 2H, ArH: 3, 5); C NMR data (δ, ppm):193.49; 163.63; 141.14; 136.59;
1
1
35.72; 132.88; 131.53; 130.40; 129.71; 127.89; 127.13; 122.65; 119.84; 118.91;
-
18.66; HRMS: Measured mass for [M-H] is 257.03784, calculated mass for
C H ClO is 257.03748.
15
10
2
1
e (C H FO ): (E)-1-(2-hydroxyphenyl)-3-(4-fluorophenyl) prop-2-en-1-one.
15 11 2
The yellow solid was recrystallized from ethanol. Yield: 56.0%; Melting point 84-
º
-1
8
1
8 C. IR (ATR) (cm ): 1635.7 (-C=O), 1570.6 (ArC-C-), 1203.2 (ArC-OH),
1
155.3(ArC-F); H NMR data (δ, ppm):12.79 (s, 1H, -OH), 7.94 (dd, 1H, J =7.80,
1
2
8