Ultrasounds-mediated 10-seconds synthesis of chalcones as potential farnesyltransferase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2020.127149

A broad range of chalcones and derivatives were easily and rapidly synthesized, following Claisen-Schmidt condensation of (hetero)aryl ketones and (hetero)aryl aldehydes using a ultrasound probe. A comparison was made with classical magnetic stirring experiments, and an optimization study was realized, showing lithium hydroxide to be the best basic catalyst of the studied condensations. By-products of the reactions (β-hydroxy-ketone, diketones, and cyclohexanols) were also isolated. All compounds were evaluated in vitro for their ability to inhibit human farnesyltransferase, a protein implicated in cancer and rare diseases and on the NCI-60 cancer cell lines panel. Molecules showed inhibitory activity on the target protein and cytostatic effect on different cell lines with particular activity against MCF7, breast cancer cells.

Full text of DOI:10.1016/j.bmcl.2020.127149

Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Ultrasounds-mediated 10-seconds synthesis of chalcones as potential
farnesyltransferase inhibitors
⁎
Germain Homerina, Adrian Sorin Nicaa,b, Amaury Farcec, Joëlle Duboisd, Alina Ghineta,b,e,
a Yncréa Hauts-de-France, Laboratory of Sustainable Chemistry and Health, Health & Environment Department, Team Sustainable Chemistry, Ecole des Hautes Etudes
d’Ingénieur (HEI), UCLille, 13 rue de Toul, F-59046 Lille, France
b Univ. Lille, Inserm, CHU Lille, Institut Pasteur de Lille, U1167 – RID-AGE – Facteurs de risque et déterminants moléculaires des maladies liées au vieillissement, F-59000
Lille, France
c Faculté des Sciences Pharmaceutiques et Biologiques de Lille, F-59006 Lille Cedex, France
d Institut de Chimie des Substances Naturelles, UPR2301 CNRS, Centre de Recherche de Gif, Avenue de la Terrasse, F-91198 Gif-sur-Yvette Cedex, France
e ‘Al. I. Cuza’ University of Iasi, Faculty of Chemistry, Bd. Carol I, nr. 11, 700506 Iasi, Romania
A R T I C L E I N F O
A B S T R A C T
Keywords:
A broad range of chalcones and derivatives were easily and rapidly synthesized, following Claisen-Schmidt
condensation of (hetero)aryl ketones and (hetero)aryl aldehydes using a ultrasound probe. A comparison was
made with classical magnetic stirring experiments, and an optimization study was realized, showing lithium
hydroxide to be the best basic catalyst of the studied condensations. By-products of the reactions (β-hydroxy-
ketone, diketones, and cyclohexanols) were also isolated. All compounds were evaluated in vitro for their ability
to inhibit human farnesyltransferase, a protein implicated in cancer and rare diseases and on the NCI-60 cancer
cell lines panel. Molecules showed inhibitory activity on the target protein and cytostatic effect on different cell
lines with particular activity against MCF7, breast cancer cells.
Farnesyltransferase
Inhibitor
Chalcone
Claisen-Schmidt
Ultrasounds
Chalcones are particularly easily synthesized and of great interest in
medicinal chemistry. With a broad spectrum of pharmacological ac-
tivities, chalcones were found to be mainly active against inflamma-
tion,1,2 tuberculosis,3 HIV,4 malaria5 and cancer.6–9 A literature review
dedicated to chalcones over the past ten years reveals almost 13,000
references with more than 1,500 references that describe the antitumor
activity of reported chalcones. This biological activity seems the most
investigated for such derivatives. The reported mechanisms of anti-
tumor action are numerous, chalcones acting as inhibitors of: tubulin
polymerization,10–15 HDAC,16 EGFR-tyrosine kinase (EGFR TK) phos-
phorylation,17,18 thioredoxin reductases (TrxRs),19 PI3K/Akt/mTOR
pathway,20 farnesyltransferase,9 etc. The most documented is the an-
timitotic activity of chalcones involving the inhibition of tubulin
polymerization. On the contrary, the activity of chalcones on human
farnesyltransferase was very little explored and one study has been
described by our group.9 The best farnesyltransferase inhibitor identi-
fied in the study was compound A (Fig. 1).9 Farnesyltransferase (FTase),
heterodimeric metalloenzyme, is a prenyltransferase involved in the
thiolinkage of a farnesyl (C15) isoprenoid coming from the biosynthetic
pathway of cholesterol, a critical metabolite for the prenylation of
proteins. This process is a critical mediator in several diseases. Indeed,
preventing the farnesylation process can constitute an approach in the
treatment of cancers, acquired immunodeficiency syndrome (AIDS) or
progeria.21 Progeria or Hutchinson-Gilford syndrome is an extremely
rare disease and has a prevalence of one case in five million births. The
appearance of the newborn baby is normal at birth; an accelerated
aging is noted a year after. Currently, only farnesyltransferase in-
hibitors were efficient on progeria models by blocking the progerin
prenylation. Five phases I/II and II clinical trials using Lonafarnib
(molecule B, Fig. 1), known farnesyltransferase inhibitor previously
developed for oncology with moderate results, alone or in combination
with zoledronic acid, pravastatin or with everolimus are currently on-
going in the hope of finding a treatment.22 First results showed addi-
tional bone mineral density benefit for patients but likely no added
cardiovascular benefit with the addition of pravastatin and zoledronic
acid.23 The discovery of new efficient farnesyltransferase inhibitors
remains a significant challenge in the fight against this extremely rare
disease but also other diseases.21
In this study, the investigation of chalcones bearing different het-
erocyclic units (thiophene, pyridine and ferrocene) than phenothiazine
⁎ Corresponding author at: Yncréa Hauts-de-France, Laboratory of Sustainable Chemistry and Health, Health & Environment Department, Team Sustainable
Chemistry, Ecole des Hautes Etudes d’Ingénieur (HEI), UCLille, 13 rue de Toul, F-59046 Lille, France.
E-mail address: alina.ghinet@yncrea.fr (A. Ghinet).
https://doi.org/10.1016/j.bmcl.2020.127149
Received 21 February 2020; Received in revised form 23 March 2020; Accepted 27 March 2020
0960-894X/©2020ElsevierLtd.Allrightsreserved.
Pleasecitethisarticleas:GermainHomerin,etal.,Bioorganic&MedicinalChemistryLetters,https://doi.org/10.1016/j.bmcl.2020.127149

G. Homerin, et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxxx
Fig. 1. Structure of reported human FTase inhibitors (indolizine-chalcone (A) and Lonafarnib (B)) and of target chalcones 1.
investigated previously (e.g. compound A) as potential human farne-
syltransferase inhibitors is reported (Fig. 1). Moreover, it has been de-
scribed that chalcones were powerful chelators of zinc cation (Zn2+).24
This property is important in the inactivation of farnesyltransferase
because the enzyme needs Zn2+ to perform its functions. Chalcone
derivatives could inhibit Zn2+ metalloenzymes such as human FTase.
Chalcones of this study were prepared by Claisen-Schmidt con-
densation of corresponding (hetero)aryl ketones and (hetero)aryl al-
dehydes (Scheme 1).25 According to the literature, bis(thienyl) chal-
cone 1a can be obtained from 1-thiophen-2-yl-ethanone and thiophene-
2-carbaldehyde using strong bases such as alkali or alkaline earth hy-
droxides.2 However, this method suffers from drawbacks: it is very
sensitive to reaction conditions such as dilution of the medium or base
quantity. As an example, β-hydroxy-ketone 2a was obtained when
water was used as the solvent,26,27 and cyclohexanol 3a was isolated in
presence of a large amount of base in low diluted medium.28,29 In our
hands, using a classical magnetic stirring and sodium hydroxide in tert-
butanol, only 13% yield was achieved in the best case for the synthesis
of chalcone 1a, which was hard to separate from by-products such as
cyclohexanol 3a (see supplementary material section). Moreover, using
1-(6-chloro-3-pyridinyl)-1-ethanone and thiophene-2-carbaldehyde in
ethanol and water, the substitution of the chlorine by an ethoxy group
led to 1b′ and 3b′ in very low yield, and as the only isolated compounds
of a complex mixture.
yield after 10 s of sonication. However, for the synthesis of 1b, the use
of NaOH provided only ethoxy-pyridine 1b′ in poor yield. The same
reaction performed using LiOH afforded the desired chalcone 1b.
Consequently, the following experiments were realized using LiOH as
the base.
A screening of different solvents was next performed (Table 2).
Mixtures of water and ethanol provided only the desired chalcone 1a
(Table 2, entries 6–8), and a 10 time up-scaled experiment was realized
with success (Table 2, entry 9). Water alone does not enable the de-
hydration of 2a, ethanol alone led to the formation of by-product 3a,
while non-polar solvents are not suitable for the reaction.
A second screening was realized to determine the relative amounts
of water, ethanol and lithium hydroxide needed, and to determine the
settings of sonication like amplitude, duration, and temperature
(Table 3). The amount of lithium hydroxide, the amplitude and dura-
tion of the sonication, and the temperature of the medium do not seem
to play a role in the reaction progress. The dilution and the relative
quantity of ethanol in water, on the contrary, are of great importance.
β-Hydroxy-ketone 2a was the major product when water alone was
used as the solvent (Table 3, entries 1–7 and 16), except in diluted
conditions (Table 3, entry 9), where chalcone 1a was obtained as the
major product. The dilution factor was raised and chalcone 1a was
obtained in good yields when water was used (Table 3, entries 9–11, 13
and 15). With less than 50% of water in the mixture, the conversion was
medium (Table 3, entries 8, 12, and 14), and undesired products were
formed. Finally, conditions were chosen as follows: i) sonication of the
medium was preferred to classical magnetic stirring; ii) the solvent was
a mixture of water and ethanol (1:1); iii) one equivalent of lithium
hydroxide was optimal; iv) a concentration of 0.02 mol.l−1 was pre-
ferred. In these conditions, chalcone 1a was isolated in 91% yield
(10–15% was obtained using classical magnetic stirring).
In this work, the sonication of the mixture was tested instead of
classical magnetic steering. Globally, the results were comparable ex-
cept in the case of chalcones 1a and 1b, but the reaction time was
decreased from hours to seconds. Different bases were next selected and
tested for the Claisen-Schmidt condensation of 1-thiophen-2-yl-etha-
none and thiophene-2-carbaldehyde (Table 1). The blank gave no re-
action. In the case of cesium carbonate, the reaction was not complete
even after 60 s of sonication, and the use of cesium fluoride led to a
complex mixture. Alkali hydroxides (NaOH and LiOH) were more sui-
table than other tested bases and led to the desired chalcone 1a in good
The above described parameters were applied to corresponding
ketones and aldehydes, and chalcones 1a-1q were obtained in medium
to high yields (see Scheme 1). Interestingly, 6-chloropyridine chalcone
2

G. Homerin, et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxxx
1
1
Table 1
O
R
O
R
LiOH
H2O/EtOH
Influence of the base on the Claisen-Schmidt condensation of 1-thiophen-2-yl-
O
O
-H2O
ethanone and thiophene-2-carbaldehyde (1 equiv of base, 0.89 mmol).a
+
2
1
OH
H
R
R
Duration (s) 2a/1a/4a/3a (1H NMR yield
determined relatively to ketone
reagent)
2
2
Entry Water/
EtOH (mL)
Base
2
1
R
R
O
1
R
1
2
3
4
5
0.4/10.0
–
60
10
10
0/0/0/0
2
0.4/10.0
0.2/10.0
2.9/10.0
1.4/10.0
NaOH
LiOH
0/ > 90/0/0
0/ > 90/0/0
0/8/36/12
__b
O
R
O
O
1
1
1
Cs2CO3 60
1
R
O
R
R
CsF
60
1
1
2
2
a
1H NMR of the crude was realized immediately after ultrasonic probe
R
R
R
R
OH
3
4
O
mediated condensation of 1-thiophen-2-yl-ethanone and thiophene-2-carbal-
dehyde.
O
O
b
S
S
S
S
Complex mixture.
1a
1b
1b'
N
EtO
N
Cl
(91%)
(13%)
(41%)
(2%)
(6%)
O
Table 2
(0%)
Influence of the solvent on the Claisen-Schmidt condensation of 1-thiophen-2-
yl-ethanone and thiophene-2-carbaldehyde (1 equiv LiOH, 0.32 mmol,
0.25 mol/L).a
O
O
Cl
Cl Cl
S
1d
1e
(22%) N
1c
N
Cl
Cl
N
N
Cl
(54%)N
Entry Solvent
Ti - Tf (°C) Energy (J) 2a/1a/3a (1H NMR yield
(55%)
O
O
Cl
O
Cl
relatively to ketone
reagent)
Cl
Cl
1f
1g
(28%)
1h
Cl
(87%) N
(35%)
Cl
Cl
Cl Cl
N
1
2
3
4
5
6
7
8
9
H2O
22–26
23–28
23–32
23–30
23–30
26–32
25–31
27–32
28–30
181
288
306
264
256
356
290
354
377
> 90/0/0
0/67/15
0/0/0
O
O
O
O
Cl
EtOH
S
C
C
C
C
C
C
C
C
DMF
C
C
C
Fe2+
C
Fe2+
C
C
Cl
N
Cl
N
THF
0/0/0
Cl
C
1j
(33%)
1k
(82%)
O
C
1i
C
C
C
(43%) C
2-Me-THF
H2O:EtOH (1/1)
H2O:EtOH (1/3)
H2O:EtOH (3/1)
Upscale of entry
6 (10X)
0/0/0
0/ > 90/0
13/77/0
0/ > 90/0
0/ > 90/0
O
Cl
Cl
Cl
Cl
Cl
Cl
Cl
S
S
1l
(86%)
1m
(99%)
1n
(89%)
Cl
(90%)
O
Cl
O
O
Cl
Cl
Ti = initial medium temperature; Tf = final medium temperature.
Cl
Cl
Cl
S
a
1H NMR of the crude was realized immediately after ultrasonic-probe
1o
(95%)
(70%)
1p
(58%)
1q
(69%)
Cl
Cl
O
Cl
Cl
mediated condensation of 1-thiophen-2-yl-ethanone and thiophene-2-carbal-
dehyde.
Cl
OH
O
Cl
Cl
S
S
If 6-chloropyridyl chalcones 1b-1j were obtained in medium yields
(31–43%) due to purification issues, other aryl and thienyl chalcones 1a
and 1 k-1r were obtained in very good yields (91–99%). Most of the
time, by-products of general formula 3 and 4 were observed in very low
quantity in the crude. Using 1-(6-chloro-pyridin-3-yl)-ethanone and 6-
chloro-pyridine-3-carbaldehyde, the isolation in very low yield (0.1%)
of compound 4d succeeded but not compound 3d. In the same light,
using 1-thiophen-2-yl-ethanone and 2,4-dichloro-benzaldehyde, com-
pound 4q (14% yield) was isolated but not compound 3q (which was
isolated from the corresponding experiment realized under magnetic
stirring). For the other examples, only the chalcone was isolated, except
for the synthesis of chalcone 1j, where enone 5 was isolated (Scheme
2). This compound was also observed as traces in the crude of the
syntheses of chalcones 1c-1f (from 6-chloro-pyridine-3-carbaldehyde),
but was not isolated despite efforts.
1r
(99%)
(90%)
2a
(81%)
Cl
Cl
Cl
O
S
O
O
O
S
O
O
S
S
S
N
N
S
Cl
Cl
OH
Cl
S
OH
S
S
OH
S
N
OEt
Cl
3a
3b
(0.1%)
3q
(0%)
(0%)
(4%)
(6%)
Cl
N
Cl
Cl
O
S
O
O
O
O
O
S
S
4b
(25%)
4d
(0.1%)
4q
(14%)
Cl
N
Cl Cl
N
N
Cl
N
In order to understand the formation of enone 5, the ultra-sonica-
tion of 6-chloro-pyridine-3-carbaldehyde alone with lithium hydroxide
was performed. The Cannizzaro reaction products30 alcohol 631 and
acid 7 were obtained (Scheme 3). The formation of compound 5 is thus
difficult to explain however oxidations and/or reductions of the re-
agents and/or the solvents would be a track to follow.
Scheme 1. Claisen-Schmidt condensation of (hetero-)aryl ketones and (hetero-)
aryl aldehydes, and compounds synthesized in this study (yields are given
within brackets, relatively to aldehyde and ketone reagents; regular writing
correspond to ultrasounds mediated experiments, and italic writing correspond
to classical stirring experiments).
1b was obtained in 31% yield only under sonication while it was not
possible to isolate it when realizing the reaction under magnetic stir-
ring. Also, lower amount of by-product 6-ethoxypyridine 1b′ was
formed (6% yield under magnetic stirring and 2% yield under sonica-
tion). Consequently, sonication induced less substitution of the
chlorine, however, di-keto-aryl compound 4b and cyclohexanol 3b
were also isolated.
The geometry of cyclohexanols
3
was then investigated.
Spectroscopic studies for 3a, 3b and 3q (1H, 13C, 2D-NMR) were un-
dertaken. Results were in accordance with Gezegen’s cyclohexanol28
(see Supplementary Data for spectra), not with Shan’s product.29 It
confirmed that 3b has the same geometry as 3a and 3q, and it showed
the position of the ethyl moiety, which is located on the pyridine next to
the hydroxyl group (Fig. 2).
3

G. Homerin, et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxxx
Table 3
Screening of the optimal volumes of water and EtOH, and number of equivalents of LiOH on the Claisen-Schmidt condensation of 1-thiophen-2-yl-ethanone and
thiophene-2-carbaldehyde.a
Entry Water (mL) EtOH (mL) Quantities of reagents
(mmol)
LiOH
Amplitude Duration (s) Ti - Tf (°C) Energy (J) 2a/1a/4a/3a (1H NMR yield relatively to
(equiv)
ketone reagent)
1
12.5
12.5
12.5
12.5
12.5
12.5
12.5
0
0
3.2
1
0.3
0.3
0.3
0.3
0.3
0.2
0.4
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
10
10
10
5
25–28
23–27
24–28
21–26
25–38
23–29
23–29
23–28
22–26
26–32
28–30
25–31
27–32
23–31
27–32
23–27
350
340
362
173
699
283
279
288
181
356
377
290
354
310
342
354
> 90/0/0/0
> 90/0/0/0
> 90/0/0/0
> 90/0/0/0
> 90/0/0/0
> 90/0/0/0
> 90/0/0/0
0/67/0/15
2
0
3.2
0.1
0.01
1
3
0
3.2
4
0
3.2
5
0
3.2
1
20
10
10
10
10
10
10
10
10
10
10
10
6
0
3.2
1
7
0
3.2
1
8
12.5
0
0.32
0.32
0.32
3.2
1
9
12.5
6.25
62.5
3.1
1
0/ > 90/0/0
0/ > 90/0/0
0/ > 90/0/0
13/77/0/0
10
11
12
13
14
15
16
6.25
62.5
9.4
3.1
12.5
6.25
0
1
1
0.32
0.32
0.32
0.32
0.32
1
9.4
1
0/ > 90/0/0
0/67/0/15
0
0.1
0.1
0.1
6.25
12.5
0/ > 90/0/0
> 90/0/0/0
Ti = initial medium temperature; Tf = final medium temperature.
a
1H NMR of the crude was realized immediately after ultrasonic-probe mediated condensation of 1-thiophen-2-yl-ethanone and thiophene-2-carbaldehyde.
inactive chalcones 1c and 1m, were all from medium to highly active.
Chalcone 1l was the most active in the thienyl-containing series
(IC50 = 7.4 µM, Table 4). All molecules decorated with two chlor-
opyridine units (e. g. chalcone 1d) or a chloropyridine and a phenyl unit
(e. g. chalcones 1e-h) were completely inactive against human FTase.
By-products (cyclohexanols 3a, 3b and 3q and diketones 4b, 4d and
4q) were also unable to inhibit the protein probably due to their very
bulky structure preventing them to reach the binding site and interact
with farnesyltransferase. Symmetric tetrachloro-substituted chalcones
1o and 1p displayed modest inhibitory activity, the 2,4-dichlorophenyl
substitution being preferred for the activity over the 3,4-dichlorophenyl
ring. (1o: IC50 = 38.7 µM; 1p: IC50 = 96.6 µM, Table 4). Dissymmetric
tetrachloro-substituted chalcones 1n and 1r interacted completely dif-
ferent with FTase. The 2,4-dichlorophenylcarbonyl unit and the 3,4-
dichlorophenyl ring placed next to the ethylenic bridge in chalcone 1n
were not tolerated to inhibit the protein while the switching of the two
rings (placement of the 2,4-dichlorophenyl next to the ethylenic bridge
and the 3,4-dichlorophenyl next to the carbonyl function in chalcone
1r) resulted in favorable interaction (1r: IC50 = 8.7 µM, Table 4).
Finally, the most active compound in this study was not a chalcone
but the β-hydroxy-ketone 2a (IC50 = 564 nM, Table 4) obtained as
intermediate in the synthesis of chalcone 1a. Compared to structurally
closed chalcone 1a, β-hydroxy-ketone 2a was 17.8 times more active
than its congener (1a: IC50 = 12 µM, Table 4).
O
O
C
C
C
C
C
H
C
C
+
Fe2+
C
N
Cl
C
C
LiOH
H2O/EtOH
O
O
C
C
C
+
C
C
Fe2+
C
5
1j
(13%)
C
(45%)
N
Cl
C
C
C
Scheme 2. Synthesis of enone 5 as a by-product of the Claisen-Schmidt con-
densation of 1-(ferrocenyl)-ethanone and 6-chloro-pyridine-3-carbaldehyde.
Fourteen molecules 1b-h, 1l-m, 1p, 1r, 3a, 3q and 4q have been
selected by the National Cancer Institute (NCI), Germantown, for eva-
luation of their antiproliferative potential against the NCI-60 cancer cell
lines, including multidrug-resistant (MDR) tumor cell lines (HCT-15:
colorectal adenocarcinoma; NCI/ADR-RES: human ovary adenocarci-
noma; RXF 393: human kidney poorly differentiated hypernephroma;
MCF7: human breast adenocarcinoma and SF-539: human CNS glio-
blastoma). Resumed representative biological efficacy is described in
Table 5. All selected molecules underwent the NCI 60 cell one-dose
screen and were tested at a 10 µM concentration.
Scheme 3. Cannizzaro reaction of 6-chloro-pyridine-3-carbaldehyde.
It is to be noted that a degradation of compound 1o was observed
over time (the efforts to isolate the degradation product did not suc-
ceed). Finally, desired chalcones were all obtained in better yields using
ultra-sonication and lithium hydroxide compared to the classical pro-
cedure.
Chalcones 1a-r, β-hydroxy-ketone 2a, cyclohexanol 3 and diketones
4 were evaluated in vitro, according to a previously described pro-
tocol,32 for their ability to inhibit human FTase (Table 4). DMSO and
chaetomellic acid A were used as negative and positive controls, re-
spectively. First, it is to be noted that the determination of chalcone 1b
activity was not possible due to intrinsic fluorescence at the test wa-
velength. Ferrocenyl compounds 1i and 1j displayed a similar activity
with IC50 values in the low micromolar range and were the most active
chalcones in the current study. Chalcones bearing a thienyl unit, except
Farnesyltransferase inhibitors generally display a cytostatic effect
rather than cytotoxic activity. This behavior was also registered in this
study. Unfortunately, the most active chalcones against FTase 1i and 1j
and β-hydroxy-ketone 2a were not selected by the NCI for biological
evaluation. Nevertheless, the other selected molecules displayed se-
lective cytostatic effect. It is to be noted that no molecule was active
against any SNC, renal and prostate cancer cells from the NCI-60 panel
(see Supplementary information section for full one dose graphs for
tested molecules). The most sensitive cancer cell line to chalcones was
4

G. Homerin, et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxxx
Cl
Cl
S
OH
H
S
O
N
OH
H
S
OH
H
S
H
H
S
S
H
H
H
S
O
O
H
H
H
S
Cl
Cl
O
O
Cl
S
H
N
N
O
O
S
Cl
3a
3b'
3q
O
O
OH
H
OH
H
H
H
H
H
H
H
O
O
O
O
O
O
Gezegen's product
Shan's product
Fig. 2. Structure of cyclohexanols 3a, 3b′ and 3q compared to literature equivalents.
Table 4
chemical agent induced excellent cell growth inhibition on K-562 and
CCRF-CEM leukemia cells (92.3% and 93% inhibition, respectively), on
LOX IMVI melanoma cells (93.5% inhibition) and on MCF7 breast
cancer cells (82.9% inhibition). A cytotoxic effect on SW-620 colon
cancer cells was also registered for chalcone 1f (33.2% reduction of pre-
existing tumor cells). Chalcones 1c, 1g and 1 m were less active than
chalcone 1f on the NCI-60 panel but have demonstrated growth in-
hibition greater than 50% against four different cell lines. It is inter-
esting to note that chalcones 1f and 1g have similar structures the only
difference being the positioning of the chloropyridine and 2,4-di-
chlorophenyl units. The positioning of the chloropyridine next to the
carbonyl group in chalcone 1f is more favorable for antiproliferative
activity while the positioning of the same motif next to the ethylenic
bridge in chalcone 1g greatly decreased the activity. The same ob-
servation applies to chalcones 1l and 1m having the left and the right
units switched. When the heterocyclic unit (in this case 2-thienyl) was
placed next to the carbonyl in chalcone 1m, the antiproliferative ac-
tivity was stronger than that of the analogue chalcone 1l where the
thienyl was placed next to the ethylenic linkage (Table 5). Chalcone 1l
has thus completely lost the activity against the leukemia cells and
induced 72.5% inhibition only on MCF7 cells.
Inhibitory activity of studied compounds on human FTase.a
Entry Compound
% Inhibition of FTase at
IC50 (µM)b R2
100 µM
1
1a
1b
1b′
1c
1d
1e
1f
100
__c
86
28
26
54
53
39
22
64
97
70
98
47
53
69
60
84
74
86
2
12.0
__c
24.9
–
0.9946
2
__c
3
0.9228
4
–
5
–
–
6
–
–
7
–
–
8
1g
1h
1i
–
–
9
–
–
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
1.0
4.6
20.8
7.4
–
0.9464
1j
0.9818
1k
1l
0.9823
0.9493
1m
1n
1o
1p
1q
1r
–
–
–
38.7
96.6
25.8
8.7
0.6
–
0.9815
0.9090
0.9415
0.9867
2a
3a
3b′
3q
4b
4d
4q
0.8356
Differences between IC50 on human FTase in the (sub)micromolar
range and IC50 on cancer cell growth in the micromolar or ten micro-
molar range is characteristic to potent FTIs and was reported previously
on different series of compounds.33
–
24
11
19
4
–
–
–
–
–
–
–
–
In the case of the presence of two different heterocyclic units in the
structure of chalcones (e.g. chalcones 1b and 1c), the best cell growth
inhibition was induced by chalcone 1c bearing the thienyl unit next to
the carbonyl instead of chloropyridine in the structure of chalcone 1b
(Table 5). The presence of two identical chloropyridines in chalcone 1d
completely abolished the biological activity.
0
–
–
Chaetomellic acid A 100
0.18
0.9898
a
Inhibition of human farnesyltransferase in vitro at 100 µM concentration of
tested compound.
b
IC50 values are indicated as the mean of two independent experiments each
realized in duplicate.
The absence of a heterocyclic ring in chalcones 1p and 1r bearing
two chloro-substituted phenyl units was detrimental for the anti-
proliferative activity. Moreover, bulkier derivatives 3a, 3q and 4q ob-
tained as by-products in the synthesis of target chalcones 1a and 1q,
respectively, were not active on the NCI-60 panel (Table 5).
c
Intrinsic fluorescence at the test wavelength (data not reliable).
MCF7 (breast cancer). Seven chalcones (1c, 1e-h, 1l and 1 m) inhibited
the growth of multi-drug resistant MCF7 breast cancer cells.
Chalcones 1c and 1m were the most active compounds on MCF7
cells displaying cell growth inhibition of 84.3% and 87.1% at 10 µM
concentration. Leukemia K-562, colon cancer HCT-116 and melanoma
LOX IMVI cells were also sensitive after treatment with chalcones.
Chalcone 1f bearing a 2,4-dichlorophenyl and a chloropyridine rings
was the most cytostatic agent in the current study (Table 5). The
Chalcones 1i-l, 1p and 1q and β-hydroxy-ketone 2a have been
docked in the binding site of human farnesyltransferase in order to
understand the binding mode and highlight the structural elements
necessary for interaction with the protein (Fig. 3).
Chalcone 1i has two-thirds of the solutions that superimposed with
the highest score, as represented in Fig. 3(a). The chloropyridine unit of
5

G. Homerin, et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxxx
Table 5
Results of the in vitro human cancer cell growth inhibition for selected compounds 1b-h, 1l-m, 1p-r, 3a, 3q and 4qa,b
Compound
1b
1c
1d
1e
1f
1g
1h
1l
1m
1p
1r
3a
3q
4q
Cell type
Leukemia
Cell line
GI%a,b
SR
17.5
2.9
n.t.d
65.8
38.6
n.t.
n.t.
6.1
3.1
n.t.
5.3
1.0
–
n.t.
n.t.
55.7
18.9
6.5
n.t.
–
78.0
71.6
31.3
66.9
24.7
15.4
29.6
39.8
45.7
6.8
6.1
9.7
3.7
7.2
8.1
–
2.8
2.4
–
7.9
14.5
5.0
17.7
15.3
3.4
–
10.0
16.5
22.3
26.8
18.5
4.7
9.7
K-562
48.3
26.6
n.t.
92.3
93.0
n.t.
25.3
15.2
n.t.
1.7
–
25.0
7.9
CCRF-CEM
RPMI-8226
NCI-H522
KM12
8.6
95.8
18.0
8.9
–
5.4
11.4
0.3
–
17.3
17.3
4.8
Non-Small Cell Lung Cancer
Colon Cancer
37.2
57.8
47.6
35.5
85.0
45.8
84.3
45.8
44.4
40.1
55.1
37.6
43.5
66.4
61.3
23.5
100.0e
63.3
93.5
21.5
82.9
24.2
3.6
17.0
32.4
17.6
19.3
21.9
11.2
69.8
8.6
2.8
5.5
2.2
–
c
–
SW-620
HCT-116
LOX IMVI
OVCAR-3
MCF7
–
7.9
–
12.7
21.2
3.1
4.4
8.9
–
–
56.6
65.1
–
6.1
–
13.9
–
6.1
7.8
–
34.7
10.9
–
Melanoma
–
Ovarian Cancer
Breast cancer
–
–
–
–
–
11.5
18.3
32.3
8.8
63.4
72.5
87.1
14.2
28.3
12.2
10.2
a
b
c
d
e
Data obtained from NCI’s in vitro 60-cell one dose screen (experiments conducted with 10 µM concentration of tested molecule).
GI% is the percentage of growth inhibition of tumor cells.
No inhibition effect.
Not tested.
Cytotoxic effect: growth percent of −33.2% on SW-620 cells induced by chalcone 1f.
FPP
FPP
FPP
Zn
Zn
Zn
Tyr166
Tyr166
Tyr166
(a)
(b)
(c)
Zn
FPP
Zn
FPP
Zn
FPP
Tyr166
Tyr166
Tyr166
(d)
(e)
(f)
FPP
FPP
FPP
Zn
Zn
Zn
Tyr861
Tyr861
Tyr166
Tyr166
Tyr166
Arg702
(g)
(h)
(i)
Fig. 3. Docking of synthesized molecules in the active site of human farnesyltransferase: chalcone 1i (a), chalcone 1j (b), chalcone 1k (c), chalcone 1l (d), chalcone
1p (e-f), chalcone 1q (g) and β-hydroxy-ketone 2a (enantiomer R (h); enantiomer S (i)). Yellow dashed lines represent hydrogen bonds between investigated
molecules and aminoacids present in the active site of human farnesyltransferase.
6

G. Homerin, et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxxx
chalcone 1i is positioned in front of the zinc of the protein while the
ferrocene unit occupies the pocket formed by amino acids Tyr861,
Arg702 and Tyr166, also known as the A2 pocket.34 Chalcone 1j, re-
verse analogue of chalcone 1i, also interacts with the zinc via the
chloropyridine unit while the ferrocene occupies the same pocket as in
chalcone 1i (Fig. 3(b)). The positioning of the ferrocene from chalcones
1i and 1j agrees with our previous studies on ketones bearing bulky
lipophilic groups such as paracyclophane, adamantane, noradamantane
or ferrocene located in the A2 binding site of farnesyltransferase.34
Chalcone 1k has twelve solutions superimposed with the best score
(Fig. 3(c)). In this configuration, it attaches to the Tyr166 via a hy-
drogen bond between the carbonyl of the chalcone and the phenol of
the amino acid. The other solutions for chalcone 1k are placed in the
opposite direction with the thienyl unit in place of the 2,4-di-
chlorophenyl ring and do not make any single interaction. This may
explain the moderate IC50 value of 20.8 µM obtained for this compound
on FTase.
v) placement of the 2,4-dichlorophenyl next to the ethylenic bridge and
the 3,4-dichlorophenyl next to the carbonyl function in chalcone 1r
resulted in favorable interaction; vi) bulky by-products 3 and 4 were
not active on FTase; vii) the β-hydroxy-ketone 2a was the most active
compound in the current study.
Compounds were also tested for their ability to prevent cancer cell
growth and underwent the one-dose screen at the National Cancer
Institute (NCI) at 10 µM concentration. Molecules showed a cytostatic
effect on different cell lines with particular activity against MCF7,
breast cancer cells.
The β-hydroxy-ketone 2a deserves further biological investigation
on progeria cells. Additional pharmacomodulations are expected on this
scaffold and could lead to a new generation of potent farnesyl-
transferase inhibitors.
Declaration of Competing Interest
Chalcone 1l, structurally very closed to chalcone 1k, interacts in
similar way (Fig. 3(d)). The hydrogen bond is conserved between the
carbonyl of the chalcone linkage and the phenol of the Tyr166 and the
chlorine atoms are slightly closer for interaction with the zinc atom
compared to chalcone 1k.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgments
Tetrachloro-substituted chalcone 1p places less well, with two so-
lutions (Fig. 3(e) and (f)) and at least two other smaller groups that are
less well defined. The conformation with the highest score does not
form hydrogen bond but is correctly placed to interact with the zinc
cation (Fig. 3(e)). The second conformation is placed conversely and
forms a hydrogen bond with Tyr166 (Fig. 3(f)). However, the very small
number of solutions can justify the modest activity of this molecule on
the FTase (IC50 = 96.6 µM).
The authors gratefully acknowledge Interreg NWE for financial
support of this work which is part of the RIVER project.
Acknowledgements are also addressed to Yncréa Hauts-de-France, HEI,
for financial support of this work, especially for the investments allo-
cated to our team.
The authors also acknowledge the National Cancer Institute (NCI)
for the biological evaluation of compounds on their 60-cell panel: the
testing was performed by the Developmental Therapeutics Program,
Division of Cancer Treatment and Diagnosis (the URL to the Program’s
website: http://dtp.cancer.gov).
Chalcone 1q behaves differently and forms a hydrogen bond via its
carbonyl group but with the Tyr861 (Fig. 3(g)) and is fixed in the active
site by stacking with the farnesyl moiety.
The best farnesyltransferase inhibitor identified in the current study
was β-hydroxy-ketone 2a synthesized as racemate. The both en-
antiomers R and S of 2a were investigated in the binding site of the
protein (Fig. 3(h) and (i)). The enantiomer R has two conformations
which are positioned in the same place and which maintain a hydrogen
bond with Tyr861 via the carbonyl unit (Fig. 3(h)). The difference
between the two conformations retrieved for (R)-2a is the change of
side of the 2-thienyl rings. The enantiomer S is even better positioned
and has a single solution for 26 of the 30 conformations. The hydrogen
bond with Tyr166 is still present and there is a second interaction with
Arg702.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bmcl.2020.127149.
References
1. Bano S, Javed K, Ahmad S, et al. Eur J Med Chem. 2013;65:51.
2. Ko H-H, Tsao L-T, Yu K-L, Liu C-T, Wang J-P, Lin C-N. Bioorg Med Chem. 2003;11:105.
3. Bhat ZS, Lah HU, Rather MA, et al. Med Chem Comm. 2018;9:165.
4. Sharma H, Patil S, Sanchez TW, Neamati N, Schinazi RF, Buolamwini JK. Bioorg Med
Chem. 2011;19:2030.
The cyclohexanol derivatives (e.g. compound 3q) narrowly enter in
the pocket of farnesyltransferase. There is no solution that stands out
from the different loose conformations (data not shown) which may
explain the absence of the biological activity of these bulky derivatives.
To conclude, an ultra-fast green Claisen-Schmidt synthesis of a
series of chalcones has been developed under sonication in a mixture of
ethanol/water and using lithium hydroxide as base. The chemical
transformations were compared with corresponding classical magnetic
stirring reactions and allowed to reduce the reaction duration from
hours to 10 s with similar to superior results. In some cases, by-products
were detected in the medium and were purified and identified as cy-
clohexanol or bisketone derivatives.
5. Insuasty B, Ramírez J, Becerra D, et al. Eur J Med Chem. 2015;93:401.
6. Mahapatra DK, Umar Bharti SK, Asati V. Eur J Med Chem. 2015:98.
7. Khanapure S, Jagadale M, Bansode P, Choudhari P, Rashinkar G. J Mol Struct.
2018;1173:142.
8. Duddukuri NK, Thatikonda S, Godugu C, Kumar RA, Doijad N. Chem Select.
2018;3:6859.
9. Moise I-M, Ghinet A, Belei D, Dubois J, Farce A, Bicu E. Bioorg Med Chem Lett.
2016;26:3730–3734 and references cited therein.
10. Wang G, Liu W, Gong Z, Huang Y, Li Y, Peng Z. Bioorg Chem. 2020. https://doi.org/
10.1016/j.bioorg.2019.103565 Ahead of print.
11. Wang G, Peng Z, Li Y. Chem Pharm Bull. 2019;67:725–728.
12. Wang G, Liu W, Gong Z, Huang Y, Li Y, Peng ZJ. Enz Inhib Med Chem.
2020;35:139–144.
13. Yan W, Chen X, Ya L, Yu W, Feng X. Invest New Drugs. 2019;37:784–790.
14. Xu F, Li W, Shuai W, et al. Eur J Med Chem. 2019;173:1–14.
15. Li W, Xu F, Shuai W, et al. J Med Chem. 2019;62:993–1013.
16. Wang B, Chen X, Gao J, et al. Bioorg Med Chem Lett. 2019;29:2638–2645.
17. Alswah M, Bayoumi AH, Elgamal K, Elmorsy A, Ihmaid S, Ahmed HEA. Molecules.
2018;23:48–63.
Synthesized molecules 1–4 have been evaluated on human farne-
syltransferase inhibitors in vitro. Chalcones 1 were globally able to in-
hibit the protein and allowed to establish additional structure–activity
relationships in the series: i) the ferrocenyl moiety was tolerated and
provided the most active chalcones in this study with IC50 in the low
micromolar range (e.g. 1i and 1j); ii) the presence of two pyridine units
(e. g. chalcone 1d) or a chloropyridine and a phenyl unit (e. g. chalcones
1e-h) in the structure of chalcones abolished the activity; iii) thienyl-
substituted chalcones conserved an inhibitory potential; iv) symmetric
tetrachloro-substituted chalcones 1o and 1p displayed modest activity;
18. Mphahlele MJ, Maluleka MM, Parbhoo N, Malindisa ST. Int J Mol Sci. 2018:19 2552/
1-2552/18.
19. Ng H-L, Ma X, Chew E-H, Chui W-K. J Med Chem. 2017;60:1734–1745.
20. Zhou X, Zhou R, Li Q, et al. Anticancer Drugs. 2019;30:241–250.
21. Manaswiyoungkul P, de Araujo ED, Gunning PT. RSC Adv. 2020;11:51–71.
22. http://clinicaltrials.gov/, consulted on January 2020.
23. Gordon LB, Kleinman ME, Massaro J, et al. Circulation. 2016;134:114–125.
24. Fosso MY, LeVine H, Green KD, Tsodikov OV, Garneau-Tsodikova S. Org Biomol
7

G. Homerin, et al.
Chem. 2015;13:9418–9426.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxxx
30. Mojtahedi MM, Akbarzadeh E, Sharifi R, Abaee MS. Org Lett. 2007;9:2791.
25. Nielsen AT, Houlihan WJ. Org React. 1968;16:1.
26. Merad J, Borkar P, Bouyon Y, et al. Org Lett. 2015;17:2118.
27. Tabatabaeian K, Mamaghani M, Mahmoodi NO, Keshavarz E. Int J Inorg Chem. 2010
No pp. given.
31. Lee C-LK, Loh T-P. Org Lett. 2005;7:2965–2967.
32. Coudray L, de Figueiredo RM, Duez S, Cortial S, Dubois JJ. Enz Inhib Med Chem.
2009;24:972.
33. Wlodarczyk N, Le Broc-Ryckewaert D, Gilleron P, et al. J Med Chem. 2011;54:1178.
34. Ghinet A, Rigo B, Dubois J, Farce A, Hénichart J-P, Gautret P. Med Chem Commun.
2012;3:1147–1154.
28. Gezegen H, Ceylan M. Synth Commun. 2015;45:2344.
29. Shan Z, Hu X, Hu L, Peng X. Helv Chim Acta. 2009;92:1102.
8