Rajni Sharma1,1, Ibidapo S Williams2, Linda Gatchie2, Vinay R Sonawane2, Bhabatosh Chaudhuri1,2, Sandip B Bharate1,1. 1. Natural Products Chemistry Division, Academy of Scientific & Innovative Research, and Medicinal Chemistry Division, CSIR-Indian Institute of Integrative Medicine, Canal Road, Jammu 180001, India. 2. CYP Design Ltd, Innovation Centre, 49 Oxford Street, Leicester LE1 5XY, U.K.
Abstract
Cytochrome P450 family 1 (CYP1) enzymes catalyze the metabolic activation of environmental procarcinogens such as benzo[a]pyrene, B[a]P, into carcinogens, which initiates the process of carcinogenesis. Thus, stopping the metabolic activation of procarcinogens can possibly prevent the onset of cancer. Several natural products have been reported to show unique ability in inhibiting CYP1 enzymes. We found that khellin, a naturally occurring furanochromone from Ammi visnaga, inhibits CYP1A1 enzyme with an IC50 value of 4.02 μM in CYP1A1-overexpressing human HEK293 suspension cells. To further explore this natural product for discovery of more potent and selective CYP1A1 inhibitors, two sets of semisynthetic derivatives were prepared. Treatment of khellin with alkali results in opening of a pyrone ring, yielding khellinone (2). Claisen-Schmidt condensation of khellinone (2) with various aldehydes in presence of potassium hydroxide, at room temperature, provides a series of furanochalcones 3a-v (khellinochalcones). Treatment of khellinone (2) with aryl aldehydes in the presence of piperidine, under reflux, affords the flavanone series of compounds 4a-p (khellinoflavanones). The khellinoflavanone 4l potently inhibited CYP1A1 with an IC50 value of 140 nM in live cells, with 170-fold selectivity over CYP1B1 (IC50 for CYP1B1 = 23.8 μM). Compound 4l at 3× IC50 concentration for inhibition of CYP1A1 completely protected HEK293 cells from CYP1A1-mediated B[a]P toxicity. Lung cancer cells, A549 (p53+) and Calu-1 (p53-null), blocked in growth at the S-phase by B[a]P were restored into the cell cycle by compound 4l. The results presented herein strongly indicate the potential of these khellin derivatives for further development as cancer chemopreventive agents.
Cytochrome P450 family 1 (CYP1) enzymes catalyze the metabolic activation of environmental procarcinogens such as benzo[a]pyrene, B[a]P, into carcinogens, which initiates the process of carcinogenesis. Thus, stopping the metabolic activation of procarcinogens can possibly prevent the onset of cancer. Several natural products have been reported to show unique ability in inhibiting CYP1 enzymes. We found that khellin, a naturally occurring furanochromone from Ammi visnaga, inhibits CYP1A1 enzyme with an IC50 value of 4.02 μM in CYP1A1-overexpressing humanHEK293 suspension cells. To further explore this natural product for discovery of more potent and selective CYP1A1 inhibitors, two sets of semisynthetic derivatives were prepared. Treatment of khellin with alkali results in opening of a pyrone ring, yielding khellinone (2). Claisen-Schmidt condensation of khellinone (2) with various aldehydes in presence of potassium hydroxide, at room temperature, provides a series of furanochalcones 3a-v (khellinochalcones). Treatment of khellinone (2) with aryl aldehydes in the presence of piperidine, under reflux, affords the flavanone series of compounds 4a-p (khellinoflavanones). The khellinoflavanone 4l potently inhibited CYP1A1 with an IC50 value of 140 nM in live cells, with 170-fold selectivity over CYP1B1 (IC50 for CYP1B1 = 23.8 μM). Compound 4l at 3× IC50 concentration for inhibition of CYP1A1 completely protected HEK293 cells from CYP1A1-mediated B[a]Ptoxicity. Lung cancer cells, A549 (p53+) and Calu-1 (p53-null), blocked in growth at the S-phase by B[a]P were restored into the cell cycle by compound 4l. The results presented herein strongly indicate the potential of these khellin derivatives for further development as cancer chemopreventive agents.
Ammi
visnaga (commonly known as
“Khella”; family: Apiaceae) is a widespread flowering
plant native to Europe, Asia, and North Africa and has many therapeutic
properties.[1] The khella seeds are being
used in traditional medicine for respiratory conditions including
bronchitis, asthma, and cough.[2] The seeds
of this plant contain a γ-pyrone, furanochromone khellin as
a major constituent, which is a vasodialator and antispasmodic.[1a,3] Inspired from the vasodialatory and antispasmodic property of khellin,
Roger Altounyan, an Anglo-Armenian physician discovered cromoyln sodium,
a chromone class of antiasthma drug.[4] Khellin
has also been extensively studied for its cardioactive properties
and is being used for angina pectoris therapy.[5] On the basis of the cardioactive property of khellin, the Labaz
company from Belgium has discovered an antiarrhythmic drug amiodarone[6] via medicinal chemistry on the benzofuran scaffold.
Other known pharmacological activities of khellin and its derivatives
include melanin inhibition[1d,7] and anti-inflammatory,[8] anticancer,[9] and cancer
chemopreventive activity.[10]Many
natural compounds and their derivatives modulate the expression
and activities of cytochrome P450 family 1 (CYP1) enzymes which are
known to bioactivate procarcinogens.[11] The
cytochrome P450 proteins comprise a large family of enzymes that can
catalyze the oxidative metabolism of a variety of xenobiotic chemicals
including toxic compounds, drugs, and chemical carcinogens as well
as endogenous compounds. As a part of our continuing efforts to discover
CYP1A1 inhibitors as cancer chemopreventive agents,[12] the furanochromone khellin (1) was identified
as a selective CYP1A1 inhibitor (CYP1A1 IC50 = 4.02 μM,
CYP1B1 IC50 = 34.6 μM) via screening of institutional
natural products repository (Figure ).[10]
Figure 1
Structure of khellin
(1), its CYP1A1 and CYP1B1 inhibitory
activity, and interactions with the CYP1A1 active site. The H-bonding
is shown by the dark blue dotted lines, whereas π–π
interactions are shown by the light blue dotted line.
Structure of khellin
(1), its CYP1A1 and CYP1B1 inhibitory
activity, and interactions with the CYP1A1 active site. The H-bonding
is shown by the dark blue dotted lines, whereas π–π
interactions are shown by the light blue dotted line.To further explore this natural product for fine-tuning
its potency
and selectivity profile, semisynthetic modifications were carried
out. Literature reports indicated that semisynthetic modifications
of khellin have been performed earlier,[13] however, never been explored as cancer chemopreventive agents. Herein,
we report synthesis of two series of khellin derivatives (khellinochalcones
and khellinoflavanones) and their biological evaluation as CYP1A1
inhibitors and possible cancer chemopreventive agents.The in
vivo studies[14] have demonstrated
that CYP1A1 is involved in benzo[a]pyrene (B[a]P)-induced carcinogenesis in mice. These studies have
also shown that polycyclic aromatic hydrocarbon (PAH)-treated mice
overexpress CYP1A1 in the lungs and liver. Therefore, herein, we also
studied the ability of khellin derivatives to protect lung cancer
cells and normal cells from B[a]P-induced toxicity.
Results
and Discussion
Synthesis of Khellin Derivatives
Khellin was identified
as one of the “hit” (active) from the natural products
repository screening,[10] in the search of
new leads for cancer chemoprevention. However, the level of activity
was in a micromolar range (IC50, 4.02 μM); therefore,
it was necessary to modify the structure of khellin to modulate its
ability to inhibit CYP1A1. On the basis of the structural features
of khellin, it was decided to convert khellin to khellinone followed
by its condensation with aryl aldehydes. The presence of the acetyl
group in khellinone provides an easiest handle for chemical modifications.Khellin (1) was treated with potassium hydroxide in
ethanol, resulting in the formation of khellinone (2).[15] Two series of compounds were prepared from khellinone
(2) as follows:[16] (a) furanochalcones
by the Claisen–Schmidt condensation of khellinone with different
substituted aryl aldehydes in presence of KOH and (b) furanoflavanones
by treatment of khellinone with aryl aldehydes in the presence of
piperidine. Khellinone (2) was treated with aryl aldehydes
under an alkaline condition (KOH), at room temperature, to obtain
khellinochalcones 3a–v (Figure , path b). Interestingly, with slight modifications
in the reaction conditions (base and reaction temperature), another
series of products were formed. Khellinone (2), when
it was reacted with aryl aldehydes in ethanol with catalytic amounts
of piperidine (instead of KOH), under reflux conditions, yielded furanoflavanones 4a–p (khellinoflavanones) (Figure path c). Mechanistically, the flavanones
were formed via cyclization of chalcones. All synthesized compounds
were characterized based on their spectroscopic properties. The coupling
constant of the newly formed bond (C8–C9) in chalcones was
found to be 16 Hz, which confirmed the trans orientation of the double
bond. The 1H NMR spectrum of flavanone 4b showed
the signals of H-7 at δ 5.45 ppm and H-6 at δ 3.04 and
2.94 ppm, confirming the formation of the flavanone skeleton. Further,
the 13C NMR and DEPT135 spectrum confirmed the appearance
of extra −CH at
δ 45.0 (C-6) and −CH at δ 78.4
(C-7). Interestingly, when khellinone (2) was treated
with thiophene-2-carboxaldehyde, it underwent double aldehyde coupling,
producing product 4p. The 1H NMR spectrum
of 4p did not show the presence of signal for CH2; however, the extra proton appeared at δ 7.13 ppm,
indicating the presence of an additional olefinic proton.
Figure 2
Synthesis of
khellin derivatives 3a–v and 4a–p. Reagents and conditions: (a) KOH (1 M), ethanol,
80 °C, 12 h, 61%; (b) RCHO (R = aryl or heteroaryl), KOH (1 M),
methanol, 0 °C to rt, 12–14 h, 40–85%; and (c)
RCHO (R = aryl or heteroaryl), piperidine, ethanol, reflux at 70 °C,
12 h, 51–68%. The “R” groups of different derivatives 3a–v and 4a–p are shown in Tables and 3.
Synthesis of
khellin derivatives 3a–v and 4a–p. Reagents and conditions: (a) KOH (1 M), ethanol,
80 °C, 12 h, 61%; (b) RCHO (R = aryl or heteroaryl), KOH (1 M),
methanol, 0 °C to rt, 12–14 h, 40–85%; and (c)
RCHO (R = aryl or heteroaryl), piperidine, ethanol, reflux at 70 °C,
12 h, 51–68%. The “R” groups of different derivatives 3a–v and 4a–p are shown in Tables and 3.
Table 1
In Vitro CYP1A1 Inhibition by Furanochalcones
in Sacchrosomes and HEK293 Cells, Overexpressing CYP1A1, at 10 μM
Concentration
α-Naphthoflavone (ANF) was
used as a positive control.
The percentage inhibition values
represent mean and standard deviations (±SD) of results obtained
from two independent experiments.
Table 3
In Vitro CYP1A1 Inhibition by Furanoflavanones
in Sacchrosomes and HEK293 Suspension Cells, Overexpressing CYP1A1,
at 10 μM Concentrationsa
The IC50 values represent
average (±SD) of results obtained from three independent experiments.
Pyridyl Furanochalcone
Derivative 3g Is a Pan Inhibitor
of CYP1 Family Enzymes
A series of furanochalcones 3a–v were tested for inhibition of CYP1A1 in yeast
microsomes (Sacchrosomes) and human live HEK293 cells, overexpressing
CYP1A1, at 10 μM (Table ).[10,12] The parent natural product khellin (1) showed 88 and
81% inhibition of CYP1A1, at 10 μM, in the two assays, respectively.
Several chalcone derivatives showed higher inhibition of CYP1A1 in
both assays. The furanochalcone 3a, bearing a simple
phenyl ring (as R group) showed 92 and 97% inhibition of CYP1A1 in
the microsomal and live cells assays, respectively. This result was
expected, as chalcones have been reported to possess potent CYP1A1
inhibition activity.[12a] The most potent
derivative among the series of chalcones was the compound 3g, which bears a pyridin-3yl moiety as the R group. This compound
showed 97 and 99% inhibition of CYP1A1 in Sacchrosomes and in live
human cells, respectively. The results of in vitro CYP1A1 inhibition
by furanochalcones in Sacchrosomes and HEK293 cells, at 10 μM
concentration, are shown in Table .α-Naphthoflavone (ANF) was
used as a positive control.The percentage inhibition values
represent mean and standard deviations (±SD) of results obtained
from two independent experiments.The IC50 values were then determined for
selected furanochalcones
in a panel of enzymes belonging to cytochrome P450 subfamilies 1,
2, and 3. The results are shown in Table . Chalcones 3e (bearing 4-fluoro-phenyl
as the R group) and 3p (bearing 2,4,5-tri-OMe phenyl
as the R group) displayed highly selective inhibition of CYP1A1 with
IC50 values of 0.32 and 0.30 μM, respectively. These
two compounds do not inhibit the other two enzymes from the CYP1 subfamily,
that is, CYP1A2 or CYP1B1 (IC50 > 10 μM). Further,
they do not inhibit the CYP2 and CYP3 subfamily of enzymes commonly
involved in drug metabolism (Table ). The pyridyl furanochalcone 3g showed
inhibition of Sacchrosomes containing CYP1A1, CYP1A2, and CYP1B1 with
IC50 values of 0.34, 0.17, and 0.66 μM, respectively,
whereas no inhibition of the CYP2 and CYP3 subfamily of enzymes was
observed (IC50 > 20 μM). The compound 3g showed IC50 values of 0.47 and 0.27 μM for CYP1A1
and CYP1B1 inhibition in live human cell assays, respectively.
Table 2
IC50 Values of Selected
Furanochalcones in a Panel of CYP Enzymes in Sacchrosomes
IC50 in μM (Sacchrosomes)a
entry
1A1
1A2
1B1
2A6, 2B6, 2C8,
2C9, 2C18, 2C19, 2D6, 2E1, 3A4
3e
0.32 ± 0.04
>10
>10
>20
3g
0.34 ± 0.03
0.17 ± 0.02
0.66 ± 0.04
>20
3p
0.30 ± 0.05
>20
>20
>20
3r
0.28 ± 0.03
4.2 ± 0.8
0.78 ± 0.04
>20
The IC50 values represent
average (±SD) of results obtained from three independent experiments.
The IC50 values represent
average (±SD) of results obtained from three independent experiments.The in silico docking studies
have shown that the chalcone 3g displays H-bonding and
π–π interactions
with the active site residues of CYP1A1. The carbonyl oxygen of the
chalcone 3g showed three-way H-bonding interactions with
Asp 320, Thr 497, and Asn 222 residues, whereas the pyridine ring
displays π–π stacking with Phe 224 and Phe 258
residues. The furan ring of the chalcone 3g has been
oriented toward the heme of the CYP1A1 enzyme (Figure ), as it was observed in the case of khellin
(Figure ).
Figure 3
Interactions
of chalcone 3g with the CYP1A1 (PDB ID: 4I8V) active site. The
H-bonding is shown by the dark blue dotted lines, whereas π–π
interactions are shown by the light blue dotted lines.
Interactions
of chalcone 3g with the CYP1A1 (PDB ID: 4I8V) active site. The
H-bonding is shown by the dark blue dotted lines, whereas π–π
interactions are shown by the light blue dotted lines.
Furanoflavanone Derivative 4l Is a Potent and Selective
Inhibitor of the CYP1A1 Enzyme
A series of furanoflavanones 4a–p were tested for the inhibition of the CYP1A1 enzyme
borne on yeast microsomes (Sacchrosomes) and CYP1A1-overexpressing
HEK293 cells, at 10 μM concentration (Table ). Several flavanone derivatives showed >85%
inhibition of the CYP1A1 enzyme in Sacchrosomes as well as in live
human cells overexpressing CYP1A1. The compound 4a, bearing
a simple phenyl as the R group, has shown 97 and 87% inhibition of
CYP1A1 in Sacchrosomes and in live human cells, respectively. The
3-pyridyl-substituted flavanone (compound 4e) also showed
potent inhibition of CYP1A1 enzyme in both assay systems.The IC50 values represent
average (±SD) of results obtained from three independent experiments.Next, the IC50 values
were determined for three best
flavanones, in live human cell assays expressing specific CYP enzymes.
Compound 4a was found to be a dual inhibitor of CYP1A1
and CYP1B1 with slight selectivity toward CYP1B1. However, the compounds 4e and 4l were selective CYP1A1 inhibitors, showing
9.2- and 170-fold selectivity over CYP1B1 (Table ), respectively. These compounds do not inhibit
CYP1A2, CYP2D6, and CYP3A4 (IC50 > 10 μM). The
excellent
selectivity profile of flavanone 4l against CYP2D6, CYP3A4,
and also within the CYP family 1 enzymes, indicated its potential
for further exploration.
Table 4
IC50 Values
of Selected
Furanochalcones for CYP1A1, CYP1A2, CYP1B1, CYP2D6, and CYP3A4 Inhibition
in Live Human HEK293 Cells
IC50 in μM (live cells)a
entry
CYP1A1
CYP1A2
CYP1B1
CYP2D6
CYP3A4
4a
1.84 ± 0.14
>10
0.566 ± 0.08
>10
>10
4e
0.45 ± 0.06
>10
4.16 ± 0.25
>10
>10
4l
0.14 ± 0.04
>10
23.8 ± 1.2
>10
>10
The IC50 values represent
average (±SD) of results obtained from three independent experiments.
The IC50 values represent
average (±SD) of results obtained from three independent experiments.The in silico docking studies
of flavanone 4l with
CYP1A1 have shown that it displays interactions similar to chalcone 3g; however, the molecule was found to be flipped by 180°
inside the active site. The furan ring of chalcone 3g was oriented toward the heme, whereas it is away from heme in the
case of flavanone 4l. However, despite this flip, the
three-way H-bonding interaction of carbonyl oxygen with Asp 320, Thr
497, and Asn 222 residues was common in both the inhibitors. The flavanone 4l also showed two π–π stacking interactions
with Phe 224 and Phe 258 residues. The interactions of 4l with CYP1A1 are shown in Figure .
Figure 4
Interactions of flavanone 4l with the CYP1A1
(PDB
ID: 4I8V) active
site.
Interactions of flavanone 4l with the CYP1A1
(PDB
ID: 4I8V) active
site.On comparison of the CYP1A1 and
CYP1B1 inhibition data of naphthyl
pyridochalcones published earlier by us,[17] with the data of khellinochalcones and khellinoflavanones, an interesting
structure–activity relationship was observed. The 3-pyridyl
naphthyl chalcone[17] (structure shown in Figure ) has been reported
as potent and selective CYP1B1 inhibitor with an IC50 value
of 4 nM, showing selectivity over CYP1A1. The replacement of the naphthyl
ring with 4,7-dimethoxybenzofuran has resulted in improved CYP1A1
inhibitory activity (Figure ). The 4,7-dimethoxybenzofuran ring bearing 3-pyridyl chalcone 3g showed CYP1A1 inhibition with an IC50 value
of 470 nM. Compound 3g inhibited CYP1B1 in live human
cell assay with an IC50 value of 270 nM, indicating slight
selectivity toward CYP1B1. Furthermore, when the chalcone skeleton
was cyclized in the form of a flavanone (3g vs 4e), there was a huge gain in CYP1A1 selectivity. This indicated
that flavanones have better selectivity toward CYP1A1, compared with
chalcones. Interestingly, when the pyridyl ring of flavanone 4e was replaced with substituted phenyl, the potent and selective
CYP1A1 inhibitor 4l was identified. The compound 4l bearing a 3-bromo-4-fluorophenyl ring instead of 3-pyridyl
(of 4e) inhibited CYP1A1 with an IC50 value
of 140 nM and 170-fold selectivity over CYP1B1 (IC50 for
CYP1B1 = 23.8 μM). The CYP1A1 and CYP1B1 activities of chalcones
and flavanones, indicating selectivity within these two enzymes, are
depicted in Figure .
Figure 5
Structure–activity relationship between chalcones and flavanones.
Structure–activity relationship between chalcones and flavanones.
Furanochalcone 3g and Furanoflavanone 4l Protect CYP1A1-Overexpressing
Normal Human Cells from CYP1A1-Mediated
B[a]P Toxicity
CYP1A1 is an aryl hydroxylase
which provides metabolic activation of PAHs and polyhalogenated aromatic
hydrocarbons. It is mainly expressed in the extrahepatic tissue such
as the lung, placenta, skin, and so forth.[18] The CYP1A1 enzyme is known to metabolically activate B[a]P to its carcinogenic form.[18,19] Therefore, to investigate
the potential of identified CYP1A1 inhibitors to stop the process
of metabolic activation, thereby protecting normal cells from carcinogenesis,
the B[a]Ptoxicity in normal HEK293 cells and HEK293
cells,[20] overexpressing CYP1A1, was studied.[10,12] EC50 of B[a]P in normal HEK293 cells
was found to be 14 μM, whereas in CYP1A1-overexpressing cells,
it was highly toxic (EC50 1.4 μM), indicating that
CYP1A1 metabolizes B[a]P to its toxic form. The cotreatment
of these CYP1A1-overexpressing cells with CYP1A1 inhibitors 3g or 4l, at their 3× IC50 concentrations,
resulted in nullifying of B[a]Ptoxicity. The cells
were completely protected from B[a]Ptoxicity as
indicated by restoration of the B[a]P EC50 to normal (i.e., 14 μM) (Figure ).
Figure 6
Protection of CYP1A1-overexpressing HEK293 cells
from B[a]P toxicity. B[a]P was used
in the concentration
range of 0.05–100 μM in the presence of 3× IC50 values of the compounds 3g or 4l (as determined in the human cell assay where cells expressed CYP1A1
when they were grown in suspension). The cell viability was determined
by 3-4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide (MTT)
assay. The cell viability determination using sulforhodamine B assay[21] provided results similar to the MTT assay. EC50 values are presented in μM concentrations and represent
the average of three independent experiments (***P < 0.001).
Protection of CYP1A1-overexpressing HEK293 cells
from B[a]Ptoxicity. B[a]P was used
in the concentration
range of 0.05–100 μM in the presence of 3× IC50 values of the compounds 3g or 4l (as determined in the human cell assay where cells expressed CYP1A1
when they were grown in suspension). The cell viability was determined
by 3-4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide (MTT)
assay. The cell viability determination using sulforhodamine B assay[21] provided results similar to the MTT assay. EC50 values are presented in μM concentrations and represent
the average of three independent experiments (***P < 0.001).
Furanoflavanone 4l Prevents B[a]P-Mediated S-Phase Block in Human
Lung Cancer Cells, A549 and Calu-1
The humanlung cancer
cell lines A549[22] and Calu-1[23] are known to express CYP1A1
enzyme. Furthermore, the in vivo studies in mice model[14] have shown that PAH-treated mice overexpress
CYP1A1 in the lung tissues; therefore, lung cancer cell lines were
chosen for this study. The effect of CYP1A1 inhibitors on the cell
cycle, in the presence of B[a]P, was investigated
by first pretreating A549 (ATCC CCL-185) and Calu-1 (ATCC HTB-54)
lung cancer cells (which express CYP1A1; Figure A) with the most potent CYP1A1 inhibitor,
compound 4l, and then exposing cells to B[a]P. The endogenous CYP1A1 is expected to metabolize B[a]P to its diol-epoxide and corresponding O-quinone,
products which are known to intercalate DNA. DNA intercalation should
block cells primarily at the S-phase of the cell division cycle. It
has been suggested that DNA adduct formation causes DNA mutations
that lead to the onset of cancer.[24]
Figure 7
Western blot
and FACS analyses of A549 and Calu-1 lung cancer cells.
(A) Expression of CYP1A1, in 5 μg of total protein, isolated
from A549 (lane 1), Calu-1cells (lane 3), normal HEK293 cells (lane
2; as negative control), and normal WI-38 cells (lane 4; as negative
control) was confirmed by western blotting; CYP1A1-bearing Sacchrosomes
(1 pmol) were taken as the positive control (lane 5). The CYP1A1-specific
antibody, used for confirming CYP1A1 expression in the upper panel,
was from ProteinTech (#13241-1-A); β-actin was used as a loading
control and the blot (lower panel) was probed with an antibody from
Sigma (# A2228). (B,E) FACS analyses of untreated A549 (B) and Calu-1
cells (E) growing asynchronously. FACS analyses of A549 cells, treated
with B[a]P, in the absence of compound 4l (C) and in the presence of compound 4l (D). FACS analyses
of Calu-1 cells, treated with B[a]P, in the absence
of compound 4l (F) and in the presence of compound 4l (G). Percentage of cells in G0/G1, S, and G2/M phases of the cell division cycle is shown
in the upper right-hand corner of each of the FACS analyses.
Western blot
and FACS analyses of A549 and Calu-1 lung cancer cells.
(A) Expression of CYP1A1, in 5 μg of total protein, isolated
from A549 (lane 1), Calu-1cells (lane 3), normal HEK293 cells (lane
2; as negative control), and normal WI-38 cells (lane 4; as negative
control) was confirmed by western blotting; CYP1A1-bearing Sacchrosomes
(1 pmol) were taken as the positive control (lane 5). The CYP1A1-specific
antibody, used for confirming CYP1A1 expression in the upper panel,
was from ProteinTech (#13241-1-A); β-actin was used as a loading
control and the blot (lower panel) was probed with an antibody from
Sigma (# A2228). (B,E) FACS analyses of untreated A549 (B) and Calu-1
cells (E) growing asynchronously. FACS analyses of A549 cells, treated
with B[a]P, in the absence of compound 4l (C) and in the presence of compound 4l (D). FACS analyses
of Calu-1 cells, treated with B[a]P, in the absence
of compound 4l (F) and in the presence of compound 4l (G). Percentage of cells in G0/G1, S, and G2/M phases of the cell division cycle is shown
in the upper right-hand corner of each of the FACS analyses.Upon exposure of asynchronous
A549 and Calu-1 lung cancer cells
(Figure B,E) to a
15 μM concentration of B[a]P for 3 h and then
releasing cells into fresh medium for 57 h, the cells did face a block
at the S-phase of the cell cycle (Figure C,F). This would be expected if the CYP1A1
protein (Figure A),
expressed within A549 and Calu-1cells, were to be a functional enzyme.
However, when cells were first preincubated for 20 h with the CYP1A1-specific
inhibitor, compound 4l, at its 3× IC50 concentrations, and then treated with a 15 μM concentration
of B[a]P for 3 h, followed by release into fresh
medium, the cells were preponderantly released from the block (Figure D,G). This would
imply that CYP1A1-mediated metabolism of B[a]P in
A549 and Calu-1 cells has been prevented by compound 4l, demonstrating the efficacy of the compound in preventing B[a]P-mediated toxicity in lung cancer cells.
Conclusions
In summary, we have identified that the semisynthetic modifications
on the natural product khellin have resulted in significant improvement
in its CYP1A1 inhibitory activity. The furanochalcone derivative 3g and furanoflavanone derivative 4l displayed
potent inhibition of CYP1A1 with IC50 values of 470 and
140 nM, respectively. These inhibitors completely protected normal
human cells from the toxicity of procarcinogen B[a]P via inhibition of its metabolic activation. Moreover, compound 4l also prevented manifestation of B[a]P-mediated
S-phase block (i.e. DNA intercalation that occurs at S-phase upon
B[a]P treatment) in lung cancer cells. Thus, the
leads identified herein have potential for further investigation as
cancer chemopreventive agents.
Experimental Section
General
The commercially
available khellin was purchased
from Sigma-Aldrich (CAS: 82-02-0, Lot≠ BCBJ3836U). NMR spectra
were recorded on Bruker-AVANCE DPX FT-NMR 500 and 400 MHz instruments.
Chemical shift values for protons/carbons are reported in parts per
million downfield from tetramethylsilane. Electrospray ionization
mass spectrometry (ESIMS) and high-resolution mass spectrometry (HRMS)
spectra were recorded on Agilent 1100 LC-Q-TOF and HRMS-6540-UHD machines.
IR spectra were recorded on a PerkinElmer IR spectrophotometer. The
melting points were recorded on a digital melting point apparatus.The purity of all derivatives was determined using high-performance
liquid chromatography (HPLC) analysis. The sample solution of compounds
of concentration 0.5 mg/mL was prepared in acetonitrile and filtered
through a 0.45 μ membrane filter injected (15 μL) in the
HPLC system (Waters) using a Chromolith C-18e column (Merck, 50 ×
4.6 mm). The binary mobile phase consisting of A (water) and B (acetonitrile)
was used. The isocratic elution was used as follows: 80 min run time
with 90% acetonitrile and 10% water. The flow rate was 1 mL/min and
the results were analyzed using a UV-PDA detector.
Synthesis of
Khellinone (2)
The mixture
of khellin (900 mg, 3.46 mmol) and potassium hydroxide (1 M) in ethanol
was refluxed at 90 °C over a period of 12–14 h. After
completion of the reaction, ethanol was evaporated over a vacuo rotavapor
and the obtained residue was partitioned in between dichloromethane
and water. The dichloromethane layer was washed with brine solution
and was dried over anhydrous sodium sulfate. The solvent was evaporated
over the vacuo rotavapor and the obtained crude product was purified
using silica gel column chromatography (5–10% ethyl acetate
in hexane) to get khellinone (2) as a yellow powder.
The purchased khellin along with khellinone were characterized using
spectral analysis.Khellin (1):
white crystals; HPLC: tR = 4.6 min (99%
purity); mp 155–156 °C; IR (CHCl3) νmax: 3436, 3160, 3137, 2989, 2931, 2960, 2830, 1619, 1586,
1471, 1444, 1424, 1364, 1380, 1300, 1265 cm–1; 1H NMR (400 MHz, CDCl3): δ (ppm) 7.51 (d, J = 2.2 Hz, 1H, CH), 6.91 (d, J = 4.0 Hz,
1H, CH), 4.15 (s, 3H, OMe), 4.05 (s, 3H, OMe), 2.73 (s, 3H, Me); 13C NMR (100 MHz, CDCl3): δ (ppm) 206.2 (C=O),
153.5 (C-7a), 152.3 (C-6), 151.6 (C-3), 143.8 (OCH=CH), 128.8 (C-7), 110.8 (C-3a), 110.5 (C-5), 106.7 (OCH=CH), 61.0 (OMe), 60.9 (OMe), 33.2 (Me); HR-ESIMS m/z: 237.0759 [M + H]+ calcd
for C12H12O5 + H+ (237.0757).[25]
Khellinone (80 mg, 0.338 mmol) was reacted with different
aromatic
aldehydes (1.2 equiv) in the presence of catalytic amount of 1 M KOH
in 50 mL of methanol at a temperature of 0–1 °C over a
period of 12–14 h. After completion of the reaction, methanol
was evaporated over the vacuo rotavapor and the obtained residue was
partitioned in between dichloromethane and water. The dichloromethane
layer was washed with brine solution and was dried over anhydrous
sodium sulfate. The solvent was evaporated over the vacuo rotavapor
and the obtained crude product was purified using silica gel column
chromatography (5–25%) to get furanochalcones 3a–v in 40–85% yield.
Khellinone
(100 mg, 0.423 mmol) was treated with different aldehydes
(1.2 equiv) in the presence of catalytic amount of piperidine in ethanol
(2 mL) at 70 °C for 12 h. The solvent was evaporated on the rotary
evaporator and the remaining mixture was diluted with water and dichloromethane.
The organic layer was separated, dried over anhydrous sodium sulfate,
and concentrated to dryness on the rotary evaporator. The obtained
crude product was purified using silica gel column chromatography
(5–40%) to yield furanoflavonones 4a–p in
51–68% yield.
The activity
of test compounds for inhibition of CYP enzymes was
investigated using Sacchrosomes and recombinant humanHEK293 cells
using protocols as described in our earlier publications.[10,12] The protection from B[a]Ptoxicity in CYP1A1-expressing
normal adherent HEK293human cells was studied as per protocols described
in our earlier publications.[10,12]For IC50 determination for CYP inhibition, the enzyme inhibition was plotted
using a sigmoidal curve (four-parameter variable slope equation) and
half-inhibitory concentration (IC50) values were analyzed
statistically using GraphPad Prism Software (Version 6.0). The IC50 values of CYP inhibitors were determined by fitting a nonlinear
curve of uninhibited fraction with incubation concentration of the
inhibitors, using the Hill equation. The EC50 values for
B[a]Ptoxicity were determined by plotting the dose–response
curve of cell viability versus log concentration of B[a]P or B[a]P + test compound. EC50 values
were obtained using GraphPad Prism Software (Version 6.0), indicating
the concentration of B[a]P required for 50% reduction
in cell viability.
Profiling of Khellinochalcones on a Panel
of 12 Humanized CYP450
Enzymes
IC50 values of furanochalcone 3g were obtained by kinetic microplate assay using 15 min reaction
time. Recombinant human P450s from a Saccharomyces
cerevisiae microsome expression system supplied from
CYP Design were used. Each 100 μL reaction was performed in
100 mM potassium phosphate buffer, pH 7.4, and with a reduced nicotinamide
adenine dinucleotide phosphate (NADPH) regenerating system (final
well concentrations: 0.04 units of glucose-6-phosphate dehydrogenase,
1.3 mM NADP+, 3.3 mM glucose-6-phosphate, 3.3 mM MgCl2, and 50 μM tribasic sodium citrate). The test samples
were serially diluted in 10% dimethyl sulfoxide (DMSO) in decreasing
final well concentrations from 20 μM downward/0.5% DMSO. CYP2E1
is sensitive to solvents, and so test samples were diluted in 1% DMSO
in decreasing final well concentrations from 10 μM downward/0.5%
DMSO. The assay was initiated by adding the substrate and the NADPH
regenerating system. P450 content and substrate concentrations were
preoptimized.[12]
In Vitro Cell Proliferation
Assays and Flow Cytometric Analyses
In vitro cell proliferation
assays and flow cytometric analyses
were performed as described earlier.[28] Control
cells (i.e., in the absence of any compound) and cells (i.e., treated
with compound) were harvested after treatment with trypsin. The cells
were fixed in chilled (−20 °C) 70% ethanol for an hour,
after washing once with phosphate-buffered saline (PBS). The fixed
cells were centrifuged at room temperature. The pellets were resuspended
in PBS, in the presence of DNase-free ribonuclease (0.5 mg/mL; Sigma-Aldrich
# R-5503) before staining with a propidium iodide solution (50 μg/mL;
Sigma-Aldrich, # P-4170) for an hour (in the dark) at 4 °C. Cell
cycle analysis was performed on the Beckman Coulter (Epics Altra)
fluorescence-activated cell sorter (Beckman Coulter UK Ltd). To exclude
cell doublets or cell clumps, all events that represent single cells
were gated. Cytograms of the fluorescence peak of propidium iodide
were plotted against the integrated fluorescence/linear signal. Data
points on a straight line, within a single gate, were isolated and
the gated data were employed for plotting a histogram that represents
a complete cell cycle. The total number of events did not exceed 200
events per second. Data acquisition was stopped after collection of
around 10 000 events.
In Silico Docking with CYP1A1
The
docking of identified
CYP1A1 inhibitors with the crystal structure of CYP1A1 (PDB ID: 4I8V)[18] was performed using the protocol as described earlier.[10,12]