Nam Q H Doan1, Ngan T K Nguyen2, Vu B Duong2, Ha T T Nguyen3, Long B Vong3, Diem N Duong4, Nguyet-Thu T Nguyen4, Tuyen L T Nguyen5, Tuoi T H Do6, Tuyen N Truong2. 1. Faculty of Pharmacy, Van Lang University, Ho Chi Minh City 700000, Vietnam. 2. Department of Organic Chemistry, Faculty of Pharmacy, University of Medicine and Pharmacy at Ho Chi Minh City, Ho Chi Minh City 700000, Vietnam. 3. School of Biomedical Engineering, International University, Vietnam National University Ho Chi Minh City, Ho Chi Minh City 700000, Vietnam. 4. Immunology Lab, Vaccines and Biologicals Production Department, Pasteur Institute in Ho Chi Minh City, Ho Chi Minh City 700000, Vietnam. 5. Saigon Pharmaceutical Sciences and Technologies Center, Ho Chi Minh City 700000, Vietnam. 6. Department of Pharmacology, Faculty of Pharmacy, University of Medicine and Pharmacy at Ho Chi Minh City, Ho Chi Minh City 700000, Vietnam.
Abstract
Addressing the growing burden of cancer and the shortcomings of chemotherapy in cancer treatment are the current research goals. Research to overcome the limitations of curcumin and to improve its anticancer activity via its heterocycle-fused monocarbonyl analogues (MACs) has immense potential. In this study, 32 asymmetric MACs fused with 1-aryl-1H-pyrazole (7a-10h) were synthesized and characterized to develop new curcumin analogues. Subsequently, via initial screening for cytotoxic activity, nine compounds exhibited potential growth inhibition against MDA-MB-231 (IC50 2.43-7.84 μM) and HepG2 (IC50 4.98-14.65 μM), in which seven compounds showing higher selectivities on two cancer cell lines than the noncancerous LLC-PK1 were selected for cell-free in vitro screening for effects on microtubule assembly activity. Among those, compounds 7d, 7h, and 10c showed effective inhibitions of microtubule assembly at 20.0 μM (40.76-52.03%), indicating that they could act as microtubule-destabilizing agents. From the screening results, three most potential compounds, 7d, 7h, and 10c, were selected for further evaluation of cellular effects on breast cancer MDA-MB-231 cells. The apoptosis-inducing study indicated that these three compounds could cause morphological changes at 1.0 μM and could enhance caspase-3 activity (1.33-1.57 times) at 10.0 μM in MDA-MB-231 cells, confirming their apoptosis-inducing activities. Additionally, in cell cycle analysis, compounds 7d and 7h at 2.5 μM and 10c at 5.0 μM also arrested MDA-MB-231 cells in the G2/M phase. Finally, the results from in silico studies revealed that the predicted absorption, distribution, metabolism, excretion, and the toxicity (ADMET) profile of the most potent MACs might have several advantages in addition to potential disadvantages, and compound 7h could bind into (ΔG -10.08 kcal·mol-1) and access wider space at the colchicine-binding site (CBS) than that of colchicine or nocodazole via molecular docking studies. In conclusion, our study serves as a basis for the design of promising synthetic compounds as anticancer agents in the future.
Addressing the growing burden of cancer and the shortcomings of chemotherapy in cancer treatment are the current research goals. Research to overcome the limitations of curcumin and to improve its anticancer activity via its heterocycle-fused monocarbonyl analogues (MACs) has immense potential. In this study, 32 asymmetric MACs fused with 1-aryl-1H-pyrazole (7a-10h) were synthesized and characterized to develop new curcumin analogues. Subsequently, via initial screening for cytotoxic activity, nine compounds exhibited potential growth inhibition against MDA-MB-231 (IC50 2.43-7.84 μM) and HepG2 (IC50 4.98-14.65 μM), in which seven compounds showing higher selectivities on two cancer cell lines than the noncancerous LLC-PK1 were selected for cell-free in vitro screening for effects on microtubule assembly activity. Among those, compounds 7d, 7h, and 10c showed effective inhibitions of microtubule assembly at 20.0 μM (40.76-52.03%), indicating that they could act as microtubule-destabilizing agents. From the screening results, three most potential compounds, 7d, 7h, and 10c, were selected for further evaluation of cellular effects on breast cancer MDA-MB-231 cells. The apoptosis-inducing study indicated that these three compounds could cause morphological changes at 1.0 μM and could enhance caspase-3 activity (1.33-1.57 times) at 10.0 μM in MDA-MB-231 cells, confirming their apoptosis-inducing activities. Additionally, in cell cycle analysis, compounds 7d and 7h at 2.5 μM and 10c at 5.0 μM also arrested MDA-MB-231 cells in the G2/M phase. Finally, the results from in silico studies revealed that the predicted absorption, distribution, metabolism, excretion, and the toxicity (ADMET) profile of the most potent MACs might have several advantages in addition to potential disadvantages, and compound 7h could bind into (ΔG -10.08 kcal·mol-1) and access wider space at the colchicine-binding site (CBS) than that of colchicine or nocodazole via molecular docking studies. In conclusion, our study serves as a basis for the design of promising synthetic compounds as anticancer agents in the future.
Globally, 18,094,716 million
cases of cancer were diagnosed in
2020, according to the Global Cancer Observatory (GLOBOCAN). Among
those, breast and liver cancer were the most common ones, ranking
first (11.7%) and seventh (4.7%) in incidence rate, and ranking fifth
(6.9%) and third (8.3%) in mortality, respectively.[1] With increasing burden in almost every country, prevention
and treatment of cancer are significant public health challenges.
Concerning cancer treatment, chemotherapy, a method using chemical
drugs to kill and/or inhibit the growth and proliferation of cancer
cells—alongside surgical operation, radiotherapy, and biotherapy—is
the mainstay. Nevertheless, this treatment method has been facing
several challenges. For instance, several active pharmaceutical ingredients
might exhibit poor solubility, poor stability in the GI tract, and
poor permeability through the intestinal epithelium such as paclitaxel
(Taxol) or doxorubicin (Adriamycin). These unfavorable physicochemical
or pharmacokinetic properties could lead to their limited bioavailabilities,
making them unsuitable for oral administration and requiring chemical
modification or formulation development to be used at their therapeutic
doses. Moreover, the biggest problem of chemotherapy chemicals is
their inability to distinguish between cancer cells and normal ones,
which might cause significant toxicity and side effects and limit
the efficacy due to their low therapeutic index.[2,3]To overcome the shortcomings of chemotherapy in cancer treatment,
recent research strategies are efforts to discover new active compounds
with higher safety and efficacy profiles. Natural products have been
excellent sources of medicinal agents, either by themselves or as
a template for synthetic agents.[4] Curcumin
or diferuloylmethane (Cur) is a natural polyphenolic
and the major component in the rhizome of Curcuma longa Linn. (Turmeric) (Figure ).[5] The ability of curcumin to
modulate various signal cascades and pathways, such as p53,[6] NF-κB,[7] AP-1
family,[8] and STAT family,[9] confirms the potential of curcumin to be an effective regulator
of a diverse variety of molecular targets.[10] Therefore, it possesses a diverse range of biological activities,
such as anticarcinogenic,[11] antitumor,[12] anti-inflammatory,[13] and also being nontoxic even at high dosages.[14] Despite its great biological activities, curcumin is not
widely accepted as an ideal pharmacologically effective molecule due
to some limitations in physicochemical properties such as poor water
and plasma solubility, chemical and biological instability, and poor
oral bioavailability.[15] Hence, the development
of synthesized analogues of curcumin, with a similar safety profile,
as well as increased activity and improved oral bioavailability, has
scientific significance. Among prominent sites of structure modification
(Figure ), the β-diketone
moiety was found to be a specific substrate of aldo–keto reductases
and could be in vivo metabolized rapidly whereas
monocarbonyl analogues of curcumin (MACs) possessing a five-carbon
dienone linker (e.g., a–i) have generally
improved pharmacokinetic profiles than that of curcumin.[16,17]
Figure 1
Structure of curcumin and prominent sites of structure
modification:
aryl sidechains (I), β-diketone (II), olefin linkers (III), and an active methylene group
(IV).
Structure of curcumin and prominent sites of structure
modification:
aryl sidechains (I), β-diketone (II), olefin linkers (III), and an active methylene group
(IV).Moreover, MACs fused
with five- or six-membered heterocycles such
as quinazoline b,[18] thiazole c,[19] imidazole d,[20] indole e,[21] and imidazo[1,2-a]pyridine f,[22] as well as MACs that have the ketone moiety
in the dienone motif replaced with cyclopentanone g,[23] cyclohexanone h,[24] or 1H-piperidine-4-one i,[25] have been well studied for improving anticancer
potency than curcumin (Figure ). Thus, the general structure portrayed by potential compounds
composed of one (hetero)aryl linked with another by penta-1,4-dien-3-one
moiety could be considered an optimal scaffold for developing novel
MACs as potential anticancer agents.[22]
Figure 2
Chemical
structure of several symmetric MACs (a, c, and d) and asymmetric MACs (b, e, f, g, h,
and i) suggests that the acetone moiety in the five-carbon
dienone linker (a) could be replaced by cyclopentanone
(g), cyclohexanone (h), and 1H-piperidine-4-one (i) while the phenyl ring could also
be replaced by several heterocyclics such as quinazoline (b), thiazole (c), imidazole (d), indole
(e), imidazo[1,2-a]pyridine (f), and pyridine (i).
Chemical
structure of several symmetric MACs (a, c, and d) and asymmetric MACs (b, e, f, g, h,
and i) suggests that the acetone moiety in the five-carbon
dienone linker (a) could be replaced by cyclopentanone
(g), cyclohexanone (h), and 1H-piperidine-4-one (i) while the phenyl ring could also
be replaced by several heterocyclics such as quinazoline (b), thiazole (c), imidazole (d), indole
(e), imidazo[1,2-a]pyridine (f), and pyridine (i).On the other hand, 1H-pyrazole
is a five-membered
heterocycle constituting a class of compounds that received substantial
attention due to their synthetic and effective biological importance.[26] In particular, different bioactivity pyrazole-containing
structures were described as anticancer,[27−29] antibacterial,[30] anti-inflammatory,[31] antituberculosis,[32] and antiviral agents.[33] Regarding anticancer activities, several drugs
possessing 1H-pyrazole are commercially available
and used to treat cancer such as pazopanib (Votrient), ruxotlitinib
(Jakavi), crizotinib (Xalkori), encorafenib (Braftovi), and lorlatinib
(Lorbrena). Previous studies have shown that 1H-pyrazole
derivatives could exhibit antiproliferation activity in vitro and antitumor activity in vivo, suggesting that
anticancer agents based on 1H-pyrazole structure
are still of significant interest. Recent studies showed that compounds
containing 1H-pyrazole could inhibit the growth of
several cancer cell types such as lung cancer, brain cancer, colorectal
cancer, renal cancer, prostate cancer, pancreatic cancer, and blood
cancer.[27−29] Among them, the antiproliferation activity compounds
containing the 1-aryl-1H-pyrazole scaffold, especially
on breast cancer cells MDA-MB-231 (j) and liver cancer
cells HepG2 (k), were also reported.[34−37] The anticancer activities of
1H-pyrazole derivatives demonstrated that their inhibitory
activities of various cancer-related targets such as topoisomerase
II (l),[38] EGFR (m),[36,39] MEK,[40] VEGFR,[36,41] GGT1,[37] microtubule (n),[42,43] HDACs,[44] Pim 1–3,[45] and carbonic anhydrase IX and XII[46] could lead to the construction of promising compounds (Figure ).[47,48]
Figure 3
Chemical
structures of anticancer drugs (pazopanib, ruxolitinib,
crizotinib, and lorlatinib) and several pyrazole-containing anticancer
agents (j–n) from previous studies.
Chemical
structures of anticancer drugs (pazopanib, ruxolitinib,
crizotinib, and lorlatinib) and several pyrazole-containing anticancer
agents (j–n) from previous studies.The synthetic feasibility and biological activities[26,47,48] of 1H-pyrazole
derivatives inspired us to synthesize asymmetric MACs fused with 1H-pyrazole. In the structure of the targeted MACs, we aimed
to retain one side of the symmetric structure of curcumin, of which
the phenyl ring was substituted with hydroxy or alkoxy groups. This
phenyl ring could be connected with the 1-aryl-1H-pyrazole scaffold via a five-carbon dienone linker motif (Figure ). Regarding the
substitution on the phenyl ring at the N1 position of 1H-pyrazole, we chose the nitro group due to the anticancer
potential of nitro-containing bioactive compounds described in the
literature elsewhere,[49] some of which contained
2-, 3-, or 4-nitrophenyl moieties.[18,50] The cytotoxicities
of compounds containing this functional group have been described
through various mechanisms, such as inhibition of topoisomerase,[51] alkylation of DNA,[52] or inhibition of tubulin polymerization.[53,54]
Figure 4
Design
of new symmetric and asymmetric MACs fused with 1-aryl-1H-pyrazole as anticancer agents.
Design
of new symmetric and asymmetric MACs fused with 1-aryl-1H-pyrazole as anticancer agents.In line with previous studies in the development
of curcumin analogue-fused
heterocycle, herein, we reported MAC-fused 1-Aryl-1H-pyrazole derivatives as desired anticancer agents.
Results and Discussion
Chemistry
In this
study, the target
compounds, 32 asymmetric MACs, which were derivatives of (1E,4E)-1-phenyl-5-(1-phenyl-1H-pyrazol-4-yl)penta-1,4-dien-3-ones (7a–10h),
were synthesized by employing KOH-catalyzed Claisen–Schmidt
condensation between (i) a 1H-pyrazole-4-carbaldehyde
derivative (4a–4d) with (ii) a 4-phenylbut-3-en-2-one
derivative (6a–6h) (Scheme ). 3-Nitro group was selected
to be substituted on the phenyl ring at the N1 position
of 1H-pyrazole (8a–8h), due to its potent anticancer activity.[49−54] Among the synthetic compounds, 7a, 7c, 7g, 10a, and 10g were previously described in
the literature.[55−57] All of these compounds were simply purified by recrystallization
in ethanol, ethyl acetate, acetone, or ethyl acetate–ethanol
mixture (1:1); or by column chromatography with 25% ethyl acetate
in n-hexane as the eluent. Finally, the target compounds
were well characterized by spectroscopic techniques such as IR, high-performance
liquid chromatography (HPLC), high-resolution mass spectrometry (HR-MS), 1H-, and 13C-NMR, which were in full accordance
with the designed structures (Figures S26–S153). The 1H- and 13C-NMR (DMSO-d6) spectra of asymmetric MACs showed the disappearance
of specific signals belonging to the aldehyde (9.90–10.32 ppm
in 1H- and 184.0–186.3 ppm in 13C-NMR)
in 1H-pyrazole-4-carbaldehydes (4a–4d) (Figures S1–S9) and
the methyl ketone (2.34–2.39 ppm in 1H- and 27.0–27.3
ppm in 13C-NMR) in 4-phenylbut-3-en-2-ones (6a–6h) (Figures S10–S25). Furthermore, the 1H-NMR (DMSO-d6) spectrum of these MACs showed two pairs of doublet signals
with a J value of ∼16.0 Hz, confirming the
(E) configuration of two double bonds formed by Claisen–Schmidt
condensation.
Scheme 1
Synthesis of 1H-Pyrazole-4-carbaldehydes
(4a–4d), 4-Phenylbut-3-en-2-ones
(6a–6i), and Asymmetric MACs Fused
with
1-Aryl-1H-pyrazole (7a–10h)
Reagents and conditions:
(i)
AcOH, reflux, 1 h (3a) or EtOH, 65–70 °C,
2 h (3b), 81–91%. (ii) AcOH, EtOH, reflux, 5 h,
91–92%. (iii) POCl3, DMF, 0–5 to 70–80
°C, 5–6 h, 55–95%. (iv, for 4c only)
KOH, EtOH–H2O (3:2), reflux, 30 min, 85%. (v) KOH,
Me2CO–H2O (2:3), 0–5 °C–r.t.
(or 60 °C), 2–48 h, 57–87%. (vi) 4% (w/v) KOH in
EtOH or saturated KOH in t-BuOH, 0–5 °C–r.t.
(or 60 °C), 4–24 h, 35–54%. Rings A and B were used to denote that the phenyl rings belonged
to phenyl hydrazine and benzaldehyde derivatives, respectively.
Synthesis of 1H-Pyrazole-4-carbaldehydes
(4a–4d), 4-Phenylbut-3-en-2-ones
(6a–6i), and Asymmetric MACs Fused
with
1-Aryl-1H-pyrazole (7a–10h)
Reagents and conditions:
(i)
AcOH, reflux, 1 h (3a) or EtOH, 65–70 °C,
2 h (3b), 81–91%. (ii) AcOH, EtOH, reflux, 5 h,
91–92%. (iii) POCl3, DMF, 0–5 to 70–80
°C, 5–6 h, 55–95%. (iv, for 4c only)
KOH, EtOH–H2O (3:2), reflux, 30 min, 85%. (v) KOH,
Me2CO–H2O (2:3), 0–5 °C–r.t.
(or 60 °C), 2–48 h, 57–87%. (vi) 4% (w/v) KOH in
EtOH or saturated KOH in t-BuOH, 0–5 °C–r.t.
(or 60 °C), 4–24 h, 35–54%. Rings A and B were used to denote that the phenyl rings belonged
to phenyl hydrazine and benzaldehyde derivatives, respectively.
Biological Activities
In Vitro Cytotoxicity
All synthetic
MACs (7a–10h) were screened for
their in vitro cytotoxicity against MDA-MB-231 and
HepG2 cancer cell lines using the MTT assay.[58] Overall, the IC50 values (μM) of synthesized MACs
suggested that the synthesized compounds exhibited stronger cytotoxicity
on MDA-MB-231 than on HepG2 cells (Table ). When compared to those of curcumin and
paclitaxel, some compounds exhibited potential cytotoxicity against
cancer cells. In particular, on MDA-MB-231 cells, stronger cytotoxicity
than curcumin was found in 21 MACs in which seven of them were significantly
stronger than paclitaxel. The respective figures for HepG2 cells were
16 MACs compared to curcumin in which six were superior to paclitaxel.
Regarding different substitutions on 1H-pyrazole
of asymmetric MACs (7a–10h), compounds 9a–9h possessing 3-carboxy-1H-pyrazole
were found to be less cytotoxic than the others on MDA-MB-231 and
appeared to be inactive on HepG2 cells. Meanwhile, compounds 7a–7h, 8a–8h, and 10a–10h possessing 3-methyl
or 3-phenyl-1H-pyrazole were found to be more potent
on two cell lines. Concerning different substitutions on ring B of asymmetric MACs, nonphenolic-substituted MACs generally
exhibited less cell growth inhibition activity than the phenolic-substituted,
except 3,4-dimethoxy- and 3,4,5-trimethoxy-substituted MACs. Among
weakly cytotoxic/inactive compounds, which were possibly attributed
to the poor absorption and permeability of these compounds through
the cell membrane, one can find too hydrophobic structures such as
nonphenolic-substituted ones carrying 1,3-diphenyl-1H-pyrazole or too hydrophilic phenolic-substituted compounds carrying
3-carboxy-1H-pyrazole. For this reason, the hydrophobic/hydrophilic
nature of the compounds might be a major factor in their cytotoxicity.
Table 1
Chemical Structure of Asymmetric MACs
Fused with 1-Aryl-1H-pyrazole-Synthesized 7a–10h, Their Cytotoxic Activity (IC50, μM), and Selectivity Index (SI) of MACs Demonstrating the
Most Potent Cytotoxicities
substitution
IC50 (μM)a
SIf
compound
R1
R2
R3
R4
MDA-MB-231b
HepG2c
LLC-PK1d
MDA-MB-231b
HepG2c
7a
H
Me
Cl
3,4-diOMe
6.60 ± 0.16
11.55 ± 1.20
NDe
ND
ND
7b
H
Me
Cl
2,4-diOMe
13.29 ± 0.39
>25
ND
ND
ND
7c
H
Me
Cl
3-OMe-4-OH
7.71 ± 0.32
8.00 ± 0.87*
9.56 ± 0.15
1.24
1.19
7d
H
Me
Cl
3-OH-4-OMe
5.61 ± 0.24*
14.20 ± 0.77
16.13 ± 0.22
2.88
1.14
7e
H
Me
Cl
3-OEt-4-OH
13.49 ± 1.43
13.18 ± 1.41
ND
ND
ND
7f
H
Me
Cl
3,4-methylendioxy
16.67 ± 0.07
>25
ND
ND
ND
7g
H
Me
Cl
3,4,5-triOMe
7.84 ± 0.07
13.03 ± 0.12
ND
ND
ND
7h
H
Me
Cl
3,5-diOMe-4-OH
3.64 ± 0.02*
6.03 ± 0.06*
12.91 ± 0.16
3.55
2.14
8a
NO2
Me
Cl
3,4-diOMe
6.89 ± 0.31
7.19 ± 0.37*
21.04 ± 0.47
3.05
2.93
8b
NO2
Me
Cl
2,4-diOMe
12.26 ± 1.29
18.32 ± 3.14
ND
ND
ND
8c
NO2
Me
Cl
3-OMe-4-OH
7.00 ± 0.02
16.60 ± 0.45
ND
ND
ND
8d
NO2
Me
Cl
3-OH-4-OMe
7.25 ± 0.08
25.31 ± 0.30
ND
ND
ND
8e
NO2
Me
Cl
3-OEt-4-OH
15.11 ± 0.20
11.18 ± 0.04
ND
ND
ND
8f
NO2
Me
Cl
3,4-methylendioxy
>25
>25
ND
ND
ND
8g
NO2
Me
Cl
3,4,5-triOMe
2.43 ± 0.04*
14.65 ± 0.10
5.93 ± 0.04
2.44
0.40
8h
NO2
Me
Cl
3,5-diOMe-4-OH
2.56 ± 0.01*
6.69 ± 0.04*
15.38 ± 0.05
6.01
2.30
9a
H
COOH
H
3,4-diOMe
19.32 ± 2.61
>25
ND
ND
ND
9b
H
COOH
H
2,4-diOMe
19.90 ± 6.77
>25
ND
ND
ND
9c
H
COOH
H
3-OMe-4-OH
22.61 ± 4.74
>25
ND
ND
ND
9d
H
COOH
H
3-OH-4-OMe
>25
>25
ND
ND
ND
9e
H
COOH
H
3-OEt-4-OH
22.63 ± 2.17
>25
ND
ND
ND
9f
H
COOH
H
3,4-methylendioxy
15.75 ± 0.88
>25
ND
ND
ND
9g
H
COOH
H
3,4,5-triOMe
>25
>25
ND
ND
ND
9h
H
COOH
H
3,5-diOMe-4-OH
>25
>25
ND
ND
ND
10a
H
Ph
H
3,4-diOMe
23.32 ± 2.37
>25
ND
ND
ND
10b
H
Ph
H
2,4-diOMe
>25
>25
ND
ND
ND
10c
H
Ph
H
3-OMe-4-OH
5.50 ± 0.22*
10.59 ± 1.38
15.79 ± 0.28
2.87
1.49
10d
H
Ph
H
3-OH-4-OMe
3.29 ± 0.18*
6.74 ± 0.26*
8.39 ± 0.12
2.55
1.25
10e
H
Ph
H
3-OEt-4-OH
6.14 ± 0.33
13.47 ± 1.53
ND
ND
ND
10f
H
Ph
H
3,4-methylendioxy
22.03 ± 3.24
>25
ND
ND
ND
10g
H
Ph
H
3,4,5-triOMe
14.29 ± 0.37
>25
ND
ND
ND
10h
H
Ph
H
3,5-diOMe-4-OH
4.06 ± 0.02*
4.98 ± 0.09*
10.19 ± 0.20
2.51
2.05
curcumin
20.65 ± 0.80
23.45 ± 7.27
>25
1.72
1.51
paclitaxel
8.79 ± 0.96
12.00 ± 0.56
0.73 ± 0.04
0.08
0.06
Half-maximal inhibitory concentrations,
after being treated with compounds for 72 h, were represented in mean
± standard deviation (SD) of three individual experiments with p < 0.05 (*) when compared to paclitaxel.
Human breast cancer cell line.
Human liver cancer cell line.
Normal porcine kidney cell line.
ND—not determined.
The SI of each MAC was calculated
by dividing the corresponding IC50 average value of the
MAC on the normal cell line by that on each cancer cell line.
Half-maximal inhibitory concentrations,
after being treated with compounds for 72 h, were represented in mean
± standard deviation (SD) of three individual experiments with p < 0.05 (*) when compared to paclitaxel.Human breast cancer cell line.Human liver cancer cell line.Normal porcine kidney cell line.ND—not determined.The SI of each MAC was calculated
by dividing the corresponding IC50 average value of the
MAC on the normal cell line by that on each cancer cell line.From cytotoxicity analysis of the
synthesized MACs carrying diversified
substituents on the 1H-pyrazole and ring B, the impact of substitution was intriguing, and the following conclusions
had been drawn from the structure–activity relationship analysis
of these compounds (Figure ).
Figure 5
Structure–activity
relationship of MACs fused with 1-aryl-1H-pyrazole
for increased anticancer potency based on cytotoxic
studies.
The number of electron-donating substituents
on ring B should be as high as possible, as could be
seen in potential 3,4,5-trisubstituted compounds such as 7g, 7h, 8g, 8h, and 10h. If there are two substituents on ring B, the 3,4-disubstitution
would be more effective than the 2,4-disubstitution, except for 3,4-methylenedioxy.
In most cases, compounds replacing an alkoxy with a hydroxy group
on ring B showed more cytotoxicity, and when comparing
different alkoxy groups, an ethoxy group showed less effectiveness
on cytotoxicity than that of a methoxy one.The presence of hydrophobic groups
(1-phenyl and 3-methyl/phenyl) on 1H-pyrazole was
attributed to superior cytotoxicity. If the substituents on 1H-pyrazole are 1,3-diphenyl, the substituents on ring B should include hydroxy groups such as 3-hydroxy (e.g., 7d and 10d) or 4-hydroxy (e.g., 7c, 7h, 8h, 10c, or 10h) to ensure the hydrophilic–hydrophobic balance.The effects of the 3-nitro group on
ring A and 5-chloro group on 1H-pyrazole
to cytotoxicities were unclear.Structure–activity
relationship of MACs fused with 1-aryl-1H-pyrazole
for increased anticancer potency based on cytotoxic
studies.From the careful analysis above,
the combined result indicated
that nine compounds, including 7c, 7d, 7h, 8a, 8g, 8h, 10c, 10d, and 10h, exhibited the
strongest cell growth inhibition against two selected cell lines.
For further evaluation of their selectivity toward cancer cells, these
compounds were also tested on a normal porcine kidney cell line (LLC-PK1),
and the selective index (SI) on each cancer cell line showed that
compound 8h was the most selective on MDA-MB-231 and
HepG2, followed by compounds 7h, 8a, and 10h (Table ). In addition, other compounds such as 7d, 10c, and 10d also showed higher selectivity on MDA-MB-231
and HepG2 than the noncancerous ones. Importantly, the highly selective
compounds showed more selectivity than paclitaxel on two cancer cell
lines compared to the normal one, which suggested that these compounds
could have lower toxicity on normal cells, thereby probably safer
than paclitaxel. Consequently, compounds 7d, 7h, 8a, 8h, 10c, 10d, and 10h were the most potent compounds in this study
due to their promising cytotoxicity and were selected to further investigate
their effects at the cellular and molecular levels.
In Vitro Tubulin Polymerization
Inhibitory Activity
The microtubule cytoskeleton plays pivotal
roles in several biological functions, ranging from intracellular
trafficking and positioning of cellular components in interphase and
the formation of the mitotic spindle during cell division to the establishment
and maintenance of cell morphology and cell motility.[59,60] Previously, several MACs fused with heterocyclics were reported
for inhibitory activities of microtubule polymerization,[21,22,61] while others possessing enone
moieties such as 4′-methoxy-2-styrylchromone[62] and 5-(indol-3-yl)-1-phenyl penta-2,4-dien-1-one[63] could stabilize microtubules. In addition, chalcones,
possessing enone moieties connecting two aryl or heteroaryl units
that were structurally similar to the MACs, could induce microtubule
polymerization[64,65] or reversibly bind in the CBS
to inhibit this process,[66,67] which depended on the
different structures of chalcones. Therefore, evaluating the effects
of synthesized MACs on tubulin polymerization in a free-cell in vitro assay was crucial. The screening results of the
selected compounds, including 7d, 7h, 8a, 8h, 10c, 10d, and 10h, for their effects on tubulin polymerization via a free-cell in vitro assay (Table , Figure ) showed that compounds 7d, 7h, and 10c at 20.0 μM exhibited the highest microtubule-destabilizing
activities, with the respective decreasing percentages in the control’s Vmax being 52.03 ± 0.65, 45.29 ± 0.81,
and 40.76 ± 0.33%, respectively, while that of colchicine at
3.0 μM was 94.65 ± 0.97%. The effective inhibition of compounds 7d, 7h, and 10c against tubulin
polymerization indicated that these MACs might target specific binding
sites of tubulin, such as the CBS and act as microtubule-destabilizing
agents to affect the microtubule dynamics.
Table 2
In Vitro Effects
of Selected Compounds on Inhibitions of Tubulin Polymerization
compound
% inhibitiona
compound
% inhibitiona
7d
52.03 ± 0.65
10c
40.76 ± 0.33
7h
45.29 ± 0.81
10d
35.48 ± 0.51
8a
25.90 ± 0.33
10h
25.73 ± 0.15
8h
22.11 ± 0.84
colchicineb
94.65 ± 0.97
The final concentration of test
compounds and DMSO in each reaction mixture were 20 μM and 1%
(v/v), respectively. Data were reduced to Vmax, which is the maximum slope of the growth phase, then the Vmax data were converted into a change in percentage
of the control’s Vmax and were
represented as the mean ± SD of two individual experiments. The
final concentration of DMSO in the reaction mixture of the control
sample was 1% (v/v).
The
final concentration of colchicine
in the reaction mixture of the negative control sample was 3 μM,
whereas that of DMSO was 0.138% (v/v).
Figure 6
Effects of the selected
compounds on microtubule dynamics. Polymerization
of tubulin at 37 °C in the presence of colchicine (3 μM),
the selected compounds (20 μM), and the control sample containing
an equivalent amount of 1% (v/v) DMSO were monitored continuously
by recording fluorescence, with the respective value of excitation
and emission wavelength being 360 and 410 nm, over 60 min. The reaction
was initiated by the addition of tubulin to a final concentration
of 2.0 mg·mL–1.
Effects of the selected
compounds on microtubule dynamics. Polymerization
of tubulin at 37 °C in the presence of colchicine (3 μM),
the selected compounds (20 μM), and the control sample containing
an equivalent amount of 1% (v/v) DMSO were monitored continuously
by recording fluorescence, with the respective value of excitation
and emission wavelength being 360 and 410 nm, over 60 min. The reaction
was initiated by the addition of tubulin to a final concentration
of 2.0 mg·mL–1.The final concentration of test
compounds and DMSO in each reaction mixture were 20 μM and 1%
(v/v), respectively. Data were reduced to Vmax, which is the maximum slope of the growth phase, then the Vmax data were converted into a change in percentage
of the control’s Vmax and were
represented as the mean ± SD of two individual experiments. The
final concentration of DMSO in the reaction mixture of the control
sample was 1% (v/v).The
final concentration of colchicine
in the reaction mixture of the negative control sample was 3 μM,
whereas that of DMSO was 0.138% (v/v).Overall, the cell-free in vitro study
indicated
that the selected MACs showed significant inhibitions on microtubule
assembly. On the other hand, regarding cytotoxic studies against breast
cancer cells MDA-MB-231 and liver cancer cells HepG2, these MACs exhibited
stronger cell growth inhibitions and higher selectivities against
the former. In comparison to paclitaxel—a microtubule-targeting
agent (MTA) widely employed for breast cancer treatment—among
the initial nine candidates, compounds 7d, 7h, and 10c demonstrated strong cytotoxicity (IC50 3.64–5.61 μM) and roughly three times the selectivity
(SI 2.87–3.55) toward MDA-MB-231, while exhibiting strong inhibition
on tubulin polymerization (40.76–52.03%). Therefore, these
compounds were chosen for further examination regarding their cellular
effects on this line of breast cancer cells.
In Vitro Apoptosis-Inducing
Activity
Many cytotoxic agents act by inducing apoptosis
as their common mechanism. Therefore, it was interesting to examine
the apoptosis-inducing effect of compounds 7d, 7h, and 10c on MDA-MB-231 cells by the AO/EB
staining and the caspase-3 activity assay method.
Acridine Orange/Ethidium Bromide (AO/EB)
Staining
The morphological changes induced by the compounds
in MDA-MB-231 cells were further studied by employing the AO/EB staining
technique.[68] It could be observed that
the control cells showed normal morphology with intact nucleus architecture
and appeared green, whereas treated MDA-MB-231 cells demonstrated
morphological changes, which were the characteristic features of apoptotic
cells such as cytoplasmic shrinkage, nucleus condensation, and cell
membrane blebbing at 1.0 μM of the selected potential compounds
(Figure A). This confirmed
that these potential compounds could induce programmed cell death
in MDA-MB-231 cells.
Figure 7
(A) Morphological changes in MDA-MB-231 cells treated
with and
without compounds 7d, 7h, or 10c for 24 h. Red arrows indicated compromised cells with early signs
of apoptosis such as condensed chromatins, horseshoe-shaped nucleus,
and nucleus fragmentation. (B) Enhancements of caspase-3 activity
of 7d, 7h, and 10c in MDA-MB-231
cells for 24 and 48 h. Data were represented as the mean ± SD
of three individual experiments with p < 0.05
(*) and p < 0.01 (**) when compared to control.
(A) Morphological changes in MDA-MB-231 cells treated
with and
without compounds 7d, 7h, or 10c for 24 h. Red arrows indicated compromised cells with early signs
of apoptosis such as condensed chromatins, horseshoe-shaped nucleus,
and nucleus fragmentation. (B) Enhancements of caspase-3 activity
of 7d, 7h, and 10c in MDA-MB-231
cells for 24 and 48 h. Data were represented as the mean ± SD
of three individual experiments with p < 0.05
(*) and p < 0.01 (**) when compared to control.
Caspase-3 Activity
Assay
Caspase-3,
along with caspase-6 and 7—highly conserved cysteine proteases—carry
out the mass proteolysis that leads to apoptosis.[69] To quantify the apoptosis-inducing ability in MDA-MB-231
cells, the colorimetric-based caspase-3 assay was employed to determine
the increase in caspase-3 activity between the apoptotic sample and
the untreated control. After treating the selected compounds for 24
and 48 h, results from analyzing cell lysates (Figure B) exhibited that, while camptothecin activated
caspase-3 at 2.48 and 2.58 folds at 1.0 μM for two respective
times, only compound 10c at 10.0 μM exhibited a
1.33-fold increase in caspase-3 activity in MDA-MB-231 cells after
24 h. Meanwhile, after 48 h, compounds 7d and 7h at 10.0 μM could induce caspase-3 activity that was 1.49 times
higher than the control cells. This observation suggested that these
compounds could induce apoptosis in MDA-MB-231 cells via the increase
in caspase-3 activity.
Cell
Cycle Analysis
Cytotoxic agents
usually alter the regulation of the cell cycle, resulting in the arrest
of cell division in various phases, thereby preventing the growth
and proliferation of cancer cells. The screening results revealed
that 7d, 7h, and 10c could
exhibit significant antiproliferative activity on MDA-MB-231 cells
and could inhibit tubulin polymerization. Therefore, the influences
of selected MACs on the MDA-MB-231 cell cycle were determined by flow
cytometry analysis of a population of cells stained with propidium
iodide (PI).[70] Regarding the G2/M population, while there was a slight increase to 37.07% in the
case of 7d at 2.5 μM, the treatment with 7h at 2.5 μM and 10c at 5.0 μM displayed
higher values at 45.05 and 40.08%, respectively, compared to the control
(33.20%), pointing out that compounds 7d, 7h, and 10c could result in G2/M cell cycle
arrest in MDA-MB-231 cells (Figure ). This observation could be attributed to their inhibitory
effects on the assembly of microtubules to form the mitotic spindle,
which could prevent cells at the end of the G2 phase from
entering the beginning of the M phase, hence increasing the proportion
of the G2/M cell population.
Figure 8
Effect of compounds 7d, 7h, and 10c on cell cycle of
MDA-MB-231 cells. Cells were treated
with these selected compounds for 24 h followed by the analysis of
cell cycle distribution using the PI staining method. All assays were
conducted in triplicate.
Effect of compounds 7d, 7h, and 10c on cell cycle of
MDA-MB-231 cells. Cells were treated
with these selected compounds for 24 h followed by the analysis of
cell cycle distribution using the PI staining method. All assays were
conducted in triplicate.
In Silico Molecular Modeling
Prediction
of Pharmacokinetic and Toxicity
Profiles
Pharmacokinetics including the absorption, distribution,
metabolism, excretion, and the toxicity (ADMET) profile of compounds 7d, 7h, and 10c were predicted by
ADMETlab2.0 (https://admetmesh.scbdd.com/) and SwissADME (http://www.swissadme.ch/) online platforms,[71,72] in comparison with curcumin,
colchicine, and paclitaxel (Table , Figure , and Tables S1 and S2).
Table 3
ADMET Profile
of Three Potent MACs
Including Compounds 7d, 7h, and 10c, in Comparison with Curcumin, Colchicine, and Paclitaxela
Data shown in [ ] represented optimal
values or defined output values. Abbreviations: MW: molecular weight;
TPSA: topological polar surface area; MDCK: Madin–Darby Canine
kidney cells; Pgp: P-glycoprotein; BBB: blood–brain
barrier; FFAMDD: the maximum recommended daily dose provides an estimate
of the toxic dose threshold of chemicals in humans. The output value
is within the range of 0–1. For the classification endpoints,
the prediction probability values are transformed into six symbols
that were divided into three empirical-based decision states visually
represented with different colors, including (1) excellent/green:
0–0.1 (− – −) and 0.1–0.3 (−
−), (2) medium/yellow: 0.3–0.5 (−) and 0.5–0.7
(+), and (3) red/poor: 0.7–0.9 (+ +) and 0.9–1.0 (+
+ +). Full details of ADMET profiles predicted are shown in Tables S1 and S2.
Figure 9
Physicochemical properties
of compounds 7d, 7h, and 10c, in comparison with those of curcumin,
colchicine, and paclitaxel. Data were retrieved from (A) SwissADME
(http://www.swissadme.ch/) and (B) ADMETlab2.0 (https://admetmesh.scbdd.com/) platforms. The colored zone (pink) was the suitable physicochemical
space for oral bioavailability. Lines and dots representing each compound
are shown in the figure. Details of physicochemical properties are
shown in Tables S1 and S2. TPSA: topological
polar surface area.
Physicochemical properties
of compounds 7d, 7h, and 10c, in comparison with those of curcumin,
colchicine, and paclitaxel. Data were retrieved from (A) SwissADME
(http://www.swissadme.ch/) and (B) ADMETlab2.0 (https://admetmesh.scbdd.com/) platforms. The colored zone (pink) was the suitable physicochemical
space for oral bioavailability. Lines and dots representing each compound
are shown in the figure. Details of physicochemical properties are
shown in Tables S1 and S2. TPSA: topological
polar surface area.Data shown in [ ] represented optimal
values or defined output values. Abbreviations: MW: molecular weight;
TPSA: topological polar surface area; MDCK: Madin–Darby Canine
kidney cells; Pgp: P-glycoprotein; BBB: blood–brain
barrier; FFAMDD: the maximum recommended daily dose provides an estimate
of the toxic dose threshold of chemicals in humans. The output value
is within the range of 0–1. For the classification endpoints,
the prediction probability values are transformed into six symbols
that were divided into three empirical-based decision states visually
represented with different colors, including (1) excellent/green:
0–0.1 (− – −) and 0.1–0.3 (−
−), (2) medium/yellow: 0.3–0.5 (−) and 0.5–0.7
(+), and (3) red/poor: 0.7–0.9 (+ +) and 0.9–1.0 (+
+ +). Full details of ADMET profiles predicted are shown in Tables S1 and S2.Regarding the physicochemical descriptors, the results
estimated
from two platforms showed that compounds 7d, 7h, and 10c exhibited proper physical and chemical properties.
All three compounds met the requirements of 5/6 properties of the
SwissADME and 10/13 properties of the ADMETlab2.0 model, being comparable
to curcumin—with 5/6 and 13/13 properties—and colchicine—with
6/6 and 13/13 properties—respectively. Paclitaxel exhibited
fewer suitable physicochemical properties, meeting only 2/6 and 5/13
properties in optimal ranges according to the SwissADME and ADMETlab2.0
models, respectively. With relatively promising physicochemical properties,
three potent MACs satisfied the Lipinski rule and the GoldenTriangle
rule to attain favorable oral pharmacokinetic profiles.[73,74] Moreover, the structure of these compounds did not contain the alerting
substructures in PAINS and also showed a high possibility of easy
synthesis of drug-like molecules.[75,76] Nevertheless,
the TPSA value of these MACs below 75 Å2 made them
unsatisfactory to the Pfizer rule,[77] while
only compound 7d conformed to the GSK rule due to its
molecular weight being below 400.[78] These
observations suggested that compounds 7d, 7h, and 10c have potential molecular structures with appropriate
physicochemical properties; however, further developments based on
their structures would be necessary to find out optimal lead compounds.As for the prediction of pharmacokinetics, during the absorption
phase, three potent MACs exhibited good passive permeabilities in
both Caco-2 and MDCK cell line models. These permeabilities could
be compared with those of curcumin, colchicine, and paclitaxel, and
even higher than that of paclitaxel in the case of the Caco-2 model.
Additionally, both online platforms predicted that paclitaxel could
be a substrate of P-glycoprotein (Pgp), while the
others could not. Therefore, the results showed an agreement between
SwissADME and ADMETlab2.0 predictions in which potent MACs, curcumin,
and colchicine demonstrated higher absorption through the gastrointestinal
(GI) tract, and also, higher oral bioavailabilities (F20% and F30%) than those of
paclitaxel.Next, in the distribution phase, only colchicine
showed favorable
percentages of plasma protein binding (PPB) form (≤90%) and
free form (≥5%), contrary to the cases of curcumin and the
analogues. The high percentage of PPB form over 90% of potent MACs,
curcumin, and paclitaxel suggested that they might have a low therapeutic
index. Interestingly, all compounds had proper volumes of distribution
(0.04–20 L·kg–1) and most of them could
not be able to penetrate the blood–brain barrier (BBB), which
is advantageous as it could avoid side effects on the central nervous
system (CNS).Concerning the metabolism processes, all five
popular human cytochrome
P450 enzymes (CYPs) were predicted to involve in the metabolism of
compound 7h and colchicine, especially enzymes CYP2C9
and CYP1A2. Similar to compound 7h, compounds 7d, 10c, and curcumin were also predicted to be the substrates
and could be metabolized by CYP2D6, CYP3A4, and mainly CYP2C9. Meanwhile,
CYP3A4 is the only enzyme among those mentioned having the ability
to metabolize paclitaxel at a moderate level. On the other hand, three
potent MACs exhibited considerable abilities to inhibit CYP1A2, CYP2C19,
and CYP2C9, while CYP2C9 and CYP3A4 could be inhibited by paclitaxel.In the elimination phase, curcumin and three potent analogues displayed
medium-to-high clearances that were higher than those of colchicine
and paclitaxel. When combining the results of clearance and volume
of distribution, curcumin showed the lowest elimination half-life
due to the highest clearance, whereas the figures for potent MACs
were higher since they showed lower clearances but similar volumes
of distribution to curcumin. Although there were improvements in retention
time in the body compared to curcumin, all potent MACs were generally
predicted to be eliminated more rapidly than colchicine and paclitaxel,
suggesting the need for higher doses and/or shorter dosing intervals
to maintain their therapeutic concentrations.Finally, prediction
of possible toxicities of three potent MACs
compared with curcumin, colchicine, and paclitaxel was also conducted.
The predicting results generally showed that all compounds might have
low risks of causing hERG blockade, eye corrosion, and also mutagenicity
in the AMES toxicity test, whereas their FDAMDD results could be positive,
meaning that they could be toxic with a maximum daily dose below 0.011
mmol·(kg-bw)−1·day–1. Higher potentials for causing skin sensitization, carcinogenicity,
and eye irritation could be seen in curcumin and three potent analogues
than those of colchicine and paclitaxel. Additionally, except for
colchicine, the remaining might have high respiratory toxicities and
high risks of induction of liver injuries. Interestingly, three potent
analogues were predicted to have fewer adverse hepatic effects than
colchicine and paclitaxel, while possibly being less toxic in the
oral acute toxicity study in rats than those of curcumin and paclitaxel,
with LD50 values of over 500 mg·kg–1.
Molecular Docking
Molecular docking
study was performed on the CBS of tubulin (PDB ID 4O2B)[79] (Figure A) to gain a better understanding of the potency of the selected
compounds (7d, 7h, and 10c)
and to support the aforementioned biological results. The binding
ability of these compounds with tubulin was representatively investigated
through the binding mode of 7h. Compound 7h was selected for its potential cytotoxicity with higher selectivity
against cancer cells, its effective inhibition of microtubule assembly,
together with its promising cellular effects on breast cancer cells.
As a preselection step for validating the docking protocols and parameters,
the redocking step of the native ligand—colchicine—was
carried out before compound 7h was docked. From the redocking
result, the top docked pose of colchicine exhibited binding energy
and root-mean-square deviation (RMSD) value, respectively, being at
−9.46 kcal·mol–1 and 0.0487 nm (Figures S154, S156, and S158), signifying that
the docking protocols and parameters were reliable for predicting
and evaluating the binding mode of compound 7h.
Figure 10
Results of
molecular docking of compound 7h into the
CBS. (A) Overall view of α,β-tubulin (shown in black and
white cartoon representation) bound with compound 7h (shown
in sticks inside the ligand surface) inside the CBS established by
secondary structures (shown in green–cyan cartoon) at the α,β-tubulin
interface. (B) Close-up view of interactions between compound 7h (shown in pink sticks) with several labeled residues (shown
in green–cyan sticks) in the CBS. (C) Molecular structures
of colchicine and nocodazole. (D) Superimposition of colchicine and
nocodazole (shown in white and black sticks, respectively) with the
top pose of compound 7h (shown in pink sticks). The αGTP
molecule (shown in limon sticks) and ion αMg++ (shown
in the lime–green sphere) were also labeled. Hydrogen bonds
are highlighted as dashed yellow lines. Heteroatoms are colored blue
(nitrogen), red (oxygen), orange (phosphorus), yellow (sulfur), or
green (chlorine).
Results of
molecular docking of compound 7h into the
CBS. (A) Overall view of α,β-tubulin (shown in black and
white cartoon representation) bound with compound 7h (shown
in sticks inside the ligand surface) inside the CBS established by
secondary structures (shown in green–cyan cartoon) at the α,β-tubulin
interface. (B) Close-up view of interactions between compound 7h (shown in pink sticks) with several labeled residues (shown
in green–cyan sticks) in the CBS. (C) Molecular structures
of colchicine and nocodazole. (D) Superimposition of colchicine and
nocodazole (shown in white and black sticks, respectively) with the
top pose of compound 7h (shown in pink sticks). The αGTP
molecule (shown in limon sticks) and ion αMg++ (shown
in the lime–green sphere) were also labeled. Hydrogen bonds
are highlighted as dashed yellow lines. Heteroatoms are colored blue
(nitrogen), red (oxygen), orange (phosphorus), yellow (sulfur), or
green (chlorine).Regarding compound 7h, the best-docked
pose of this
compound showed that its ring B could enter the hydrophobic
pocket inside β-tubulin shaped by H7 helix, S6, and S8 strands,
hence forming several hydrophobic contacts with GLY237-THR239, CYS241
(H7 helix), ALA316-ILE318 (S8 strand), and TYR202 (S6 strand), together
with three canonical hydrogen bonds between hydroxy/methoxy groups
of ring B and backbone atoms of VAL238 (0.211 nm), THR240
(0.276 nm), and CYS241 (0.206 nm). Meanwhile, the 1H-pyrazole of the top pose is located at the interface and navigated
ring A to the space between αT5 and T7 loops, while
its carbonyl group projected toward the S8 strand to form a canonical
hydrogen bond with ALA317 (0.210 nm) (Figure B). These interactions increased the stability
of compound 7h in the CBS, resulting in the significant
binding energy of this compound, at −10.08 kcal·mol–1 (Figures S155, S157, and Table S3). With the insights into the binding mode of compound 7h, it could be observed that this compound could bind well
at the CBS and could occupy the same position as several common colchicine-binding
site inhibitors (CBSIs). Interestingly, from the best of our knowledge,
the CBS could be subdivided into the main zone in the center (zone
2) and two additional zones located on either sides of the main zone
(zones 1 and 3),[80] and none of CBSIs occupying
three zones simultaneously have been described so far.[81] The superimposition of the top docked poses
of compound 7h with two well-known CBSIs, colchicine[82,83] and nocodazole (PDB ID 5CA1)[79,84] (Figure C,D), demonstrated that the binding ability
of these MACs could spread in a wider space in the CBS, which were
more accessible to the α-tubulin surface (zone 1) than nocodazole
and were concurrently buried deeper inside β-tubulin (zone 3)
than colchicine.
Conclusions
In conclusion,
MACs fused with 1-aryl-1H-pyrazole,
especially asymmetric MACs, described in our present study demonstrated
the potential anticancer activities and indicated exciting possibilities
of developing new cancer therapeutics via their structural modifications.
Although the most potent MACs had several benefits in addition to
potential drawbacks in terms of biological activities and ADMET properties
when compared with well-known drugs, we expect that our study could
serve as a basis for the design of synthetic compounds that would
appear promising for other in vitro or in
vivo anticancer activity studies as tubulin polymerization
inhibitors in the future.
Experimental Section
General
All of the chemicals, reagents,
starting materials, and solvents were procured from commercial suppliers
such as Acros, Fisher Scientific (Hampton, NH), AK Scientific (Union
City, CA), Sigma-Aldrich (St. Louis, MO), and Gibco-Invitrogen (Waltham,
MA) and were used without further purification. Analytical thin-layer
chromatography (TLC) was performed using Merck precoated silica gel
60 F254 (8.0–12.0 μm) aluminum plates (0.2
mm), while column chromatography was conducted using silica gel 60
(40.0–63.0 μm) (Kenilworth, NJ). Visualization of spots
on TLC plates was achieved by UV light (254 and 365 nm). The purity
of all compounds was detected by a Shimadzu Prominence-i LC2030C 3D
(Kyoto, Japan) (column: Gemini C18 5 μm × 4.6 mm ×
250 mm, flow rate: 1.0 mL·min–1, UV wavelength:
430 nm, eluent/methanol water from 70:30 to 90:10). Melting points
were checked by the open capillary tube method using a Sanyo–Gallenkamp
melting point apparatus (Hampton, NH) and they are uncorrected. IR
spectra for all of the compounds were recorded on an IRAffinity-1S
Shimadzu instrument (Kyoto, Japan) using the ATR technique. 1H- and 13C-NMR spectra were recorded on a Bruker Advance
II Instrument (Billerica, MA) operated at 500 and 125 MHz, respectively.
HR-MS spectra were performed by a Shimadzu LC-20A system (Kyoto, Japan)
with the LCMS-IT-TOF detector, or by the Agilent LC 6545 system (Santa
Clara, CA) with the LCMS Q-TOF detector. The human breast (MDA-MB-231),
human liver (HepG2), and porcine kidney cell line (LLC-PK1) were purchased
from American Type Culture Collection (ATCC, Manassas, VA).
Synthetic Procedure
Initially, 1H-pyrazole-4-carbaldehyde
derivatives (4a–4d) were synthesized
through 2–3 consecutive steps,
which were described before,[85−87] starting from (i) phenyl hydrazine
derivatives, and (ii) methyl acetoacetate, ethyl pyruvate, or acetophenone,
to afford 1H-pyrazol-5(4H)-ones
(3a and 3b)[86] or hydrazones (3c and 3d).[85] Subsequently, via the Vilsmeier–Haack
reaction, 3a and 3b underwent a formylation–chlorination
process to obtain intermediates 4a and 4b,[86] whereas 3c and 3d underwent a cyclization–formylation process to obtain
intermediates 4c′ and 4d,[85] respectively. The intermediate 3c was synthesized by hydrolysis of ester 4c′ in
the presence of aqueous alkali, followed by acidification with aqueous
acid.[87] The other core structures, 4-phenylbut-3-en-2-one
derivatives (6a–6h), were prepared by KOH-catalyzed
condensation of benzaldehyde derivatives with acetone. Finally, KOH-catalyzed
condensation between 4a–4d and 6a–6h furnished the target asymmetric MACs 7a–10h (Scheme ).[88] Synthetic compounds were characterized by spectroscopic
techniques such as IR, HR-MS, 1H-, and 13C-NMR.
General Procedure for the Synthesis of Asymmetric
MACs 7a–10h
A mixture of a 4-phenylbut-3-en-2-one
derivative (6a–6h, 1.05 mmol) and
4% KOH solution in ethanol or saturated KOH solution in tert-butanol (5 mL) was stirred at 0–5 °C for 10–15
min. To this mixture, 1H-pyrazole-4-carbaldehyde
(4a–4d, 1.00 mmol) was added in small
portions. The reaction mixture was carried out at room temperature,
or 60 °C if 6c or 6e was used, for
4–24 h. Upon completion, the mixture was added with cold water
(10–15 mL), then the excess amount of KOH was either neutralized
or acidified to pH = 2–4 if 4c was used, by a
concentrated solution of HCl. The precipitate was filtered, then dried,
and purified by recrystallization using solvents such as ethanol,
ethyl acetate, acetone, and the ethyl acetate–ethanol (1:1)
mixture, or by column chromatography using the n-hexane–ethyl
acetate (3:1) mixture to obtain the corresponding asymmetric MACs.
Among asymmetric MACs synthesized, 7a, 7c, 7g, 10a, and 10g were described before
in the literature.[55,56]
The in vitro anticancer activity of synthetic compounds was
determined using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium
bromide (MTT) assay. About 2.5–5 × 103 cells
per well were seeded in 100 μL of DMEM or RPMI 1640, supplemented
with 10% FBS, 4 mM of l-glutamine, 100 IU·mL–1 penicillin, and 100 μM streptomycin in each well of 96-well
plates and incubated for 24 h at 37 °C in a 5% CO2 incubator. Compounds at designed concentrations were added to wells
with respective vehicle control. Paclitaxel and equivalent amounts
of solvent used to dissolve the compounds were used as the positive
and negative controls, respectively, whereas curcumin was used for
comparison. After 72 h of incubation period, the medium was removed,
then 100 μL of MTT (0.5 mg·mL–1) containing
serum-free medium was added to each well, and the plates were incubated
for 3 h. The supernatant from each well was removed carefully; formazan
crystals were dissolved in 200 μL of acidified isopropanol,
and the absorbance was recorded at 570 nm wavelength with a Multiskan
FC Microplate Photometer (Thermo Fisher Scientific, Waltham, MA).
Then, cytotoxic activities and half-maximal inhibitory concentrations
(IC50) were calculated.[58]
Tubulin Polymerization Inhibitory Assay
To assess the effect of potential compounds on tubulin polymerization,
a fluorescence-based in vitro tubulin polymerization
assay was carried out according to the manufacturer’s protocol
(BK011P, Cytoskeleton, Denver, CO). The reaction mixture consisting
of 2.0 mg·mL–1 porcine brain tubulin in 80
mM PIPES at pH 6.9, 2.0 mM MgCl2, 0.5 mM EGTA, 1.0 mM GTP,
and 15% glycerol in the presence of sample compounds at a final concentration
of 20.0 μM or the equivalent amount of DMSO at a final concentration
of 1% (v/v) was prepared and added to each well of a black 96-well
plate. Paclitaxel and colchicine at a final concentration of 3.0 μM
were used as the positive and negative controls, respectively. Tubulin
polymerization was followed by a time-dependent increase in fluorescence
intensity due to the incorporation of a fluorescence reporter into
microtubules as polymerization proceeds. Tubulin assembly was determined
via measuring fluorescence variation at 410 nm (excitation wavelength
is 360 nm) and 37 °C every 60 s for 60 min was recorded using
a Varioskan LUX Multimode Microplate Reader (Thermo Fisher Scientific,
Waltham, MA). Data obtained were reduced to the maximum slope of the
growth phase (Vmax) of microtubule polymerization
using GraphPad Prism 8 (GraphPad Software, San Diego, CA), and then
the Vmax data were converted into a percentage
change in the Vmax of control.
Acridine Orange/Ethidium Bromide (AO/EB)
Staining
The morphological changes of treated and control
cells were examined by acridine orange/ethidium bromide (AO/EB) staining.
Cells were grown in 96-well plates at a density of 1.5 × 104 cells per well for 24 h and then were treated with 1.0, 2.5,
5.0, or 10 μM concentrations of compounds for 24 h. After the
incubation, cells were centrifuged at 1000 rpm within 10 min, and
then the medium was removed before the cells were stained with a 10
μL dyes mixture including 100 μg·mL–1 acridine orange and 100 μg·mL–1 ethidium
bromide. Morphological features were observed, and photographs were
taken under an Axio Lab.A1 Fluorescence Microscope (Carl Zeiss Microscopy
GmbH, Jena, Germany) at 200× magnification.[68]
Caspase-3 Activity Assay
Caspase-3
activity was assessed using the caspase-3 colorimetric assay kit (K106,
BioVision, Milpitas, CA) according to the manufacturer’s protocol.
Briefly, 1–5 × 106 cells per well were treated
with 2.5, 5.0, or 10 μM concentrations of the potential compounds
and the equivalent amounts of DMSO for 24 and 48 h. Cells treated
with 1 μM camptothecin for 24 and 48 h were used as a positive
control. After the period of treatment, the medium was removed, and
then cell pellets were resuspended and lysed with 50 μL of lysis
buffer and incubated on ice for 10 min. Cell lysates were then centrifuged
at 10 000g for 1 min at 4 °C. The concentration
of proteins was measured by Bradford assay. Then, 50 μL of 2×
reaction buffer containing 10 mM DTT and 5 μL of 4 mM caspase-3
substrate (DEVD-pNA) were added to 50 μg of
protein in 50 μL of each sample and incubated at 37 °C
for 60 min. The pNA light emission was quantified
using BioTek ELx808 Absorbance Microplate Reader at 405 nm (BioTek,
Winooski, VT). Comparison of the absorbance of pNA
from an apoptotic sample with an uninduced control allowed determination
of the fold increase in caspase-3 activity.[69]
Cell Cycle Analysis
To examine
the effect of potential compounds on the cell cycle, cells were seeded
in 6-well plates at a density of 1.5 × 105 cells/mL
and allowed to attach for 24 h. Cells were treated with 2.5, 5.0,
or 10.0 μM concentrations of potential compounds and incubated
further for 24 h. The cells then were collected, washed, and fixed
in 70% ethanol in PBS at −20 °C. After 2 h of the fixing
step, the fixed cells were pelleted and stained with propidium iodide
(2.5 μg·mL–1) in the presence of RNase
A (12.5 μg·mL–1) for 30 min at 37 °C
in dark. Finally, the cells were analyzed using an ACEA NovoCyte Flow
Cytometer (ACEA Biosciences, San Diego, CA).[70]
Molecular Modeling
Prediction
of Pharmacokinetic Profiles
ADMETlab2.0 (https://admetmesh.scbdd.com) and SwissADME (http://www.swissadme.ch) were two online platforms implemented to calculate physicochemical
descriptors and predict the pharmacokinetics and toxicity profile
of each compound.[71,72]
Molecular
Docking
The 3D structure
of α,β-tubulin (PDB ID 4O2B) was retrieved from RCSB Protein Data
Bank (http://www.rcsb.org).[79] The colchicine-binding site (CBS) at the interface
between α- and β-subunits, αGTP molecule, and αMg++ ion were involved in molecular docking studies.[82,83] Missing residues were added, whereas all solvents and cocrystallized
colchicine were removed using the receptor preparation tool in M.O.E.
2015 (Chemical Computing Group Inc., Montreal, Canada).[89] Structures of the potential compounds were sketched
by ChemBioDraw 16.0 (PerkinElmer Informatics, Waltham, MA), their
geometry optimized, and partial charges calculated with the Gasteiger–Huckel
method using Avogadro 1.2.0 (Avogadro Chemistry).[90] In AutoDock Tools 1.5.6 (Scripps Research, La Jolla, CA),
protein, colchicine, and potential ligands were loaded, hydrogens
added, and saved in *.pdbqt format. A 2.0 × 2.0 × 2.0 nm
grid box covering the binding site was created. Parameter files, *.gpf
and *.dpf, of macromolecule and ligands for grid and dock run, were
also prepared. Molecular docking was performed using AutoDock 4.2
(Scripps Research, La Jolla, CA) using the Genetic Algorithm—Local
Search methodology with default parameters.[91] The conformations obtained were analyzed and clustered using a root
mean square (RMS) cut-off of 0.2 nm, then the best binding pose was
selected based on the highest binding energy from the top first cluster.
Finally, the complex of the best pose of potential ligands with α,β-tubulin
was used to analyze the ligand–protein interactions using M.O.E.
2015 and PyMOL 2.3.4 (Schrödinger LLC, New York, NY).[92]
Authors: M Jane Cox Rosemond; Lisa St John-Williams; Toshiro Yamaguchi; Toshio Fujishita; John S Walsh Journal: Chem Biol Interact Date: 2004-03-15 Impact factor: 5.192
Figure 1
Figure 2
Figure 3
Figure 4
Scheme 1
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10