Synthetic models related to methoxalen and menthofuran-cytochrome P450 (CYP) 2A6 interactions. Benzofuran and coumarin derivatives as potent and selective inhibitors of CYP2A6

Article abstract of DOI:10.1248/cpb.c12-00872

Human microsomal cytochrome P450 (CYP) 2A6 contributes extensively to nicotine detoxication but also activates tobacco-specific procarcinogens to mutagenic products. We prepared a series of benzofuran and coumarin derivatives that have inhibitory effects on the activity of human CYP2A6. The reported compounds methoxalen and menthofuran had potent inhibitory effects on the activity of CYP2A6 with IC₅₀ values of 0.47 μM and 1.27 μM, respectively. Synthetic benzofuran (4-methoxybenzofuran: IC₅₀=2.20 μM) and coumarin (5-methoxycoumarin: IC₅₀=0.13 μM and 6-methoxycoumarin: IC₅₀=0.64 μM) derivatives, which have more selective effects than those of methoxalen and menthofuran, exhibited comparable activities against CYP2A6. These compounds can be used as a lead compounds in the design of CYP2A6 inhibitor drugs to reduce smoking and tobacco-related cancers.

Full text of DOI:10.1248/cpb.c12-00872

October 2013
Regular Article
Chem. Pharm. Bull. 61(10) 997–1001 (2013)
997
Synthetic Models Related to Methoxalen and Menthofuran–Cytochrome
P450 (CYP) 2A6 Interactions. Benzofuran and Coumarin Derivatives as
Potent and Selective Inhibitors of CYP2A6
Yuki Yamaguchi, Ichie Akimoto, Kyoko Motegi, Teruki Yoshimura, Keiji Wada,
Naozumi Nishizono,* and Kazuaki Oda*
Faculty of Pharmaceutical Sciences, Health Sciences University of Hokkaido; Ishikari-Tobetsu, Hokkaido 061–0293,
Japan.
Received October 4, 2012; accepted July 19, 2013; advance publication released online August 1, 2013
Human microsomal cytochrome P450 (CYP) 2A6 contributes extensively to nicotine detoxication but
also activates tobacco-specific procarcinogens to mutagenic products. We prepared a series of benzofuran
and coumarin derivatives that have inhibitory effects on the activity of human CYP2A6. The reported com-
pounds methoxalen and menthofuran had potent inhibitory effects on the activity of CYP2A6 with IC50 val-
ues of 0.47µM and 1.27µM, respectively. Synthetic benzofuran (4-methoxybenzofuran: IC50=2.20µM) and cou-
marin (5-methoxycoumarin: IC50=0.13µM and 6-methoxycoumarin: IC50=0.64µM) derivatives, which have
more selective effects than those of methoxalen and menthofuran, exhibited comparable activities against
CYP2A6. These compounds can be used as a lead compounds in the design of CYP2A6 inhibitor drugs to
reduce smoking and tobacco-related cancers.
Key words inhibitor; cytochrome P450 2A6; benzofuran; coumarin; methoxalen; menthofuran
Cytochrome P450 (CYP) 2A6, the major coumarin 7-hy-
droxylase present in the human liver, is known to metabo-
Experimental
Preparation of Inhibitors Compounds 1, 2, and 12 were
lize a variety of compounds, including nicotine, cotinine, commercially available.
quionoline, and aflatoxin B1.1,2) CYP2A6 does not seem to
Compound 3: 3-Methyl-4,5,6,7-tetrahydrobenzofuran
have an extensive role in human drug metabolism, but it Pyrrolidine enamine of cyclohexanone was alkylated by ethyl
has been shown to be involved in the mutagenic activation 2-bromopropionate to give 3-methyl-5,6,7,7a-tetrahydrobenzo-
of promutagens such as the tobacco-specific nitrosoamines, furan-2(4H)-one. This butenolide was reduced by DIBAL-H,
4-(methylnitrosoamino)-1-(3-pyridyl)-1-butanone (NNK) and affording compound 3. Colorless oil, 1H-NMR (500MHz,
N′-nitrosonornicotine (NNN).3–5) It has also been shown that CDCl3) δ: 1.65–1.78 (4H, m), 1.85 (3H, d, J=1.2Hz), 2.26–2.29
methoxalen, a potent human CYP2A6 inhibitor, is a strong (2H, m), 2.46 (2H, t, J=6.3Hz), 7.10 (1H, d, J=1.2Hz).
chemopreventive agent against NNK induction of lung tu- 13C-NMR (125MHz, CDCl3) δ: 7.2, 20.1, 22.8, 22.9, 23.0,
morigenesis.3–5) Pharmacogenomic studies have suggested that 117.5, 119.4, 136.7, 150.3. Its spectral data are in accordance
male smokers completely lacking CYP2A6 were more resis- with previously reported data.8,9)
tant to lung cancer. In addition, inhibition of CYP-mediated
Compound 4: 7-Methoxy-4,5,6,7-tetrahydrobenzofuran
metabolism would lead to slower elimination of nicotine from Cyclohexa-1,2-dione was treated with chloroacetaldehyde to
the bodies of smokers and, consequently, to a decrease in give 5,6-dihydrobenzofuran-7(4H)-none. This ketone was
the number of cigarettes needed to maintain a constant level reduced with LiAlH4 to give an appropriate alcohol (racemic
of nicotine in the body. This decrease in smoking would de- mixture). The obtained alcohol was alkylated by methyl io-
crease the exposure to carcinogenic nitrosoamine. For this dide in the presence of NaH, affording compound 4. Colorless
reason, CYP2A6 inhibitors have been proposed as a novel oil, 1H-NMR (500MHz, CDCl3) δ: 1.70–1.95 (3H, m), 2.05
targets for reducing tobacco-related cancer risk. An ideal (1H, m), 2.36 (1H, m), 2.56 (1H, m), 3.47 (3H, s), 4.29 (1H,
drug candidate for use as a CYP2A6 inhibitor and a smoking d, J=3.5Hz), 6.21 (1H, d, J=1.7Hz), 7.31 (1H, d, J=1.7Hz).
cessation agent must be highly potent and selective to avoid 13C-NMR (125MHz, CDCl3) δ: 19.2, 22.3, 29.2, 56.9, 70.8,
undesirable effects.6)
110.3, 120.8, 142.1, 150.2. Its spectral data are in accordance
Although a number of compounds, including methoxalen, with previously reported data.10)
have been used as inhibitors of CYP2A6, these compounds
Compound 5: 3-Methylbenzofuran 3-Methylbenzofuran-
generally lack selectivity for targeting CYP2A6 and many of 2-carboxylic acid (commercially available) was heated with
these compounds also inhibit activities of other drug-metabo- copper metal in quinoline to give compound 5. Colorless
lizing enzymes (such as CYP3A4 and CYP2D6) and therefore oil, 1H-NMR (500MHz, CDCl3) δ: 2.22 (3H, d, J=1.2Hz),
can result in untoward drug-drug interactions.7) Herein we re- 7.18–7.26 (2H,m), 7.38 (1H, dd, J=1.2, 8.0Hz), 7.48 (1H, m),
port the synthesis of benzofuran and coumarin derivatives and 7.51 (1H, dd, J=1.2, 8.0Hz). 13C-NMR (125MHz, CDCl3)
results of assays of their inhibitory effects on CYP2A6 activ- δ: 6.4, 110.6, 115.2, 119.0, 121.9, 123.8, 128.4, 141.4, 155.5.
ity with a view to defining the relationship between structures Its spectral data are in accordance with previously reported
and inhibitory effects on CYP2A6 activity.
data.11)
Compound 6: 3,6-Dimethylbenzofuran12) 2-Hydroxy-4-
methylacetophenone was alkylated with ethyl bromoacetate
to give ethyl 2-acetyl-5-methylphenoxyacetate. This ester was
The authors declare no conflict of interest.
© 2013 The Pharmaceutical Society of Japan
*To whom correspondence should be addressed. e-mail: nishizon@hoku-iryo-u.ac.jp;
k-oda@hoku-iryo-u.ac.jp

998
Vol. 61, No. 10
heated with KOH in dioxane, affording compound 6. Color- try (ESI-LR-MS) m/z: 148 (M+). ESI-high resolution (HR)-MS
1
less oil, H-NMR (500MHz, CDCl3) δ: 2.20 (3H, d, J=1.2Hz), m/z: 148.0520 (Calcd for C9H8O2: 148.0524). Anal. Calcd for
2.42 (3H, s), 6.58 (1H, d, J=1.2Hz), 6.77 (1H, dd, J=2.3, C9H8O2: C, 72.96; H, 5.44. Found: C, 72.92; H, 5.34.
8.6Hz), 6.94 (1H, d, J=2.3Hz), 7.34 (1H, d, J=8.6Hz), 7.43
(1H, d, J=1.7Hz). 13C-NMR (125MHz, CDCl3) δ: 6.5, 10.3, one Methoxalen was reduced with 10% Pd–C to give com-
111.2, 118.2, 120.0, 121.9, 123.8, 128.7, 141.4, 153.5.
pound 13. mp 159–161°C. (lit.20) 163°C). 1H-NMR (CDCl3)
Compound 13: 9-Methoxy-2H-furo[3,2-g]chromen-7(3H)-
Compound 7: 3-Methoxybenzofuran 3-Methoxybenzo- δ: 3.27 (2H, t, J=8.6Hz), 4.04 (3H, s), 4.72 (2H, t, J=8.6Hz),
furan-2-carboxylic acid (commercially available) was heated 6.21 (1H, d, J=9.5Hz), 6.98 (1H, s), 7.58 (1H, d, J=9.5Hz).
with copper metal in quinoline to give compound 7. Colorless 13C-NMR (CDCl3) δ: 29.2, 61.1, 73.3, 112.5, 113.7, 117.4, 125.8,
1
oil, H-NMR (500MHz, CDCl3) δ: 3.88 (3H, s), 7.21 (1H, m), 131.8, 144.0, 147.6, 154.7, 161.0. Electron ionization (EI)-
7.30 (1H, m), 7.40 (1H, d, J=8.6Hz), 7.48 (1H, m), 7.53–7.55 LR-MS m/z: 218 (M+). EI-HR-MS m/z: 218.0580 (Calcd for
(1H, m) 13C-NMR (125MHz, CDCl3) δ: 58.2, 111.7, 118.8, C12H10O4: 218.0579).
121.8, 122.2, 124.4, 125.1, 145.5, 153.8. Its spectral data are in
accordance with previously reported data.13,14)
Compound 14: 4-Methoxycoumarin 4-Hydroxycoumarin
(commercially available) was refluxed with methyl iodide,
Compound 8: 4-Methoxybenzofuran Cyclohexa-1,3- K2CO3 in acetone to give compound 14. mp 122–124°C. (lit.21)
1
dione was treated with 3-bromopyruvic acid in the presence 123.5–124°C) H-NMR (500MHz, CDCl3) δ: 3.99 (3H, s), 5.69
of NaOH. The obtained tetrahydrobenzofuran derivative (1H, s), 7.26 (1H, dd, J=7.5, 8.0Hz), 7.31 (1H, d, J=8.6Hz),
was refluxed with 10% palladium carbon in decalin to give 7.54 (1H, ddd, J=1.2, 7.5, 8.6Hz), 7.80 (1H, dd, J=1.2, 8.0Hz).
4-hydroxybenzofuran-2-carboxylic acid. This carboxylic acid 13C-NMR (125MHz, CDCl3) δ: 56.4, 90.2, 115.7, 116.9, 123.1,
was heated with copper metal in quinoline, affording 4-hy- 124.0, 132.5, 153.4, 163.0, 166.5. Its spectral data are in accor-
droxybenzofuran. Compound 815) was obtained by methylation dance with previously reported data.21)
1
of alcohol described as compound 4. Colorless oil, H-NMR
Compound 15: 5-Methoxycoumarin 2,6-Dimethoxy-
(500MHz, CDCl3) δ: 3.94 (3H, s), 6.66 (1H, d, J=2.3Hz), benzaldehyde was treated with methyl (triphenylphos-
6.86 (1H, d, J=2.3Hz), 7.14 (1H, d, J=8.6Hz), 7.22 (1H, t, J= phoranylidene)acetate in toluene to give (E)-methyl-3-(2,6-
8.6Hz), 7.53 (1H, d, J=2.3Hz) 13C-NMR (125MHz, CDCl3) δ: dimethoxyphenyl)acrylate.22) This ester was cyclized with
55.6, 103.6, 104.1, 104.8, 117.8, 125.0, 143.6, 153.7, 156.4.
BBr3 in dichloromethane to give 5-hydroxycoumarin. 5-Hy-
Compound 9: 5-Methoxybenzofuran 4-Methoxyphenol droxycoumarin was refluxed with methyl iodide, Cs2CO3
was alkylated with 2-bromoacetaldehyde diethyl acetal in the in acetonitrile to give compound 15. mp 83°C. (lit.22) 82°C)
presence of KOH. The obtained ether was subsequently cy- 1H-NMR (500MHz, CDCl3) δ: 3.93 (3H, s), 6.33 (1H, d,
clized with PPA, affording 5-hydroxybenzofuran. Compound J=9.7Hz), 6.70 (1H, d, J=8.1Hz), 6.91 (1H, d, J=8.6Hz), 7.43
9 was obtained by methylation of alcohol described as com- (1H, m), 8.08 (1H, d, J=9.7Hz). 13C-NMR (125MHz, CDCl3)
pound 4. Compound 9 was obtained by methylation of alcohol δ: 56.1, 105.2, 109.3, 109.7, 114.7, 132.4, 138.6, 155.2, 156.2,
1
described as compound 4. Colorless oil, H-NMR (500MHz, 161.1. Its spectral data are in accordance with previously re-
CDCl3) δ: 3.85 (3H, s), 6.71 (1H, d, J=2.3Hz), 6.92 (1H, dd, ported data.23)
J=2.3, 9.2Hz), 7.06 (1H, d, J=2.3Hz), 7.41 (1H, d, J=9.2Hz),
Compound 16: 6-Methoxycoumarin 6-Hydroxycoumarin
7.60 (1H, d, J=2.3Hz). 13C-NMR (125MHz, CDCl3) δ: 56.0, (commercially available) was refluxed with methyl iodide,
103.6, 106.8, 111.9, 113.2, 128.1, 145.8, 150.0, 156.0. Its spec- K2CO3 in acetone to give compound 16. mp 99–101°C. (lit.23)
tral data are in accordance with previously reported data.16)
102–103°C) 1H-NMR (500MHz, CDCl3) δ: 3.84 (3H, s),
Compound 10: 6-Methoxybenzofuran Benzen-1,3-diol 6.42 (1H, d, J=9.2Hz), 6.91 (1H, d, J=2.9Hz), 7.10 (1H, dd,
was treated with chloroacetonitrile in the presence of HCl J=2.9, 8.6Hz), 7.26 (1H, d, J=8.6Hz), 7.65 (1H, d, J=9.2Hz).
to give 6-hydroxy-coumaran-3-one. This compound was re- 13C-NMR (125MHz, CDCl3) δ: 55.9, 110.1, 117.2, 118.0, 119.3,
duced with NaBH4 and dehydrated by acid to give 6-hydroxy- 119.5, 143.3, 148.6, 156.2, 161.1. Its spectral data are in accor-
benzofuran. Compound 1017) was obtained by methylation of dance with previously reported data.23)
alcohol described as compound 4. Compound 10 was obtained
Compound
17:
8-Methoxycoumarin 2-Hydroxy-3-
by methylation of alcohol described as compound 4. Colorless methoxybenzaldehyde was heated with methyl (triphenyl-
1
oil, H-NMR (500MHz, CDCl3) δ: 3.72 (3H, s), 6.58 (1H, d, phosphoranylidene)acetate to give compound 17. mp 81–83°C.
J=1.7Hz), 6.77 (1H, dd, J=2.3, 8.6Hz), 6.94 (1H, d, J=2.3Hz), (lit.22) 89°C) 1H-NMR (500MHz, CDCl3) δ: 3.96 (3H, s),
7.34 (1H, d, J=8.6Hz), 7.43 (1H, d, J=1.7Hz). 13C-NMR 6.43 (1H, d, J=9.8Hz), 7.05 (1H, d, J=8.0Hz), 7.07 (1H, d,
(125MHz, CDCl3) δ: 55.7, 96.0, 106.5, 112.1, 120.8, 121.3, J=8.0Hz), 7.20 (1H, t, J=8.0Hz), 7.68 (1H, d, J=9.8Hz).
144.2, 156.1, 158.2. Its spectral data are in accordance with 13C-NMR (125MHz, CDCl3) δ: 56.3, 113.8, 117.1, 119.3, 119.6,
previously reported data.18)
124.4, 143.7, 143.8, 147.4, 160.3. Its spectral data are in accor-
Compound 11: 7-Methoxybenzofuran19) 2-Hydroxy-3- dance with previously reported data.22)
methoxybenzaldehyde was treated with ethyl chloroacetate in
Assay. Assay of Inhibition Assays of inhibition of
the presence of K2CO3 to give 6-methoxybenzofuran-2-car- CYP2A6 activity (GENTEST Co., Human CYP2A6 + cyto-
boxylic acid. This carboxylic acid was heated with copper chrome b5+P450 reductase (Baculovirus)) by benzofuran and
metal in quinoline, affording compound 11. Colorless oil, coumarin derivatives were based on microsomal coumarin
1H-NMR (500MHz, CDCl3) δ: 4.00 (3H, s), 6.76 (1H, d, J= 7-hydroxylation. The 7-hydroxylation of coumarin was used
2.3Hz), 6.80 (1H, d, J=8.0Hz), 7.16 (1H, t, J=8.0Hz), 7.20 as the index of CYP2A6 activity. The rate of coumarin 7-hy-
(1H, d, J=8.0Hz), 7.62 (1H, d, J=2.3Hz). 13C-NMR (125MHz, droxylation was determined by the method of Yamazaki et
CDCl3) δ: 56.1, 106.4, 107.0, 113.6, 123.6, 129.2, 144.4, 145.0, al.24) Expressed human CYP2A6 (0.5pmol) was incubated
145.7. Electrospray ionization-low resolution-mass spectrome- with coumarin (final concentration of 1µM) and an inhibi-

October 2013
999
Chart 1
tor (0.01–10 µM) at 37°C in the presence of phosphate buffer (2pmol) was incubated with dextromethorphan (final concen-
(100 mM, pH 7.4), nicotinamide adenine dinucleotide phos- tration of 4µM) and an inhibitor (50µM) at 37°C in the pres-
phate (NADP+) (0.8mM), glucose-6-phosphate (7.9mM), and ence of phosphate buffer (500mM, pH 7.4), NADP+ (0.8 mM),
glucose-6-phosphate dehydrogenase (1unit/mL) in a total glucose-6-phosphate (7.9mM), and glucose-6-phosphate dehy-
volume of 400µL. The reaction mixture was preincubated for drogenase (1unit/mL) in a total volume of 500µL. The reac-
2min at 37°C, and the reaction was started by the addition tion was started by addition of an NADPH-generating system
of a reduced nicotinamide adenine dinucleotide phosphate and was terminated after incubation for 10min by addition
(NADPH)-generating system of the microsomal solution. The of 50µL of 60% HClO4. After further mixing, the reaction
reaction was stopped after 6-min incubation with 50µL of mixture was centrifuged at 10000rpm for 10min at 4°C. The
60% HClO4. After further mixing, the reaction mixture was supernatant was analyzed by the HPLC-fluorescence method
centrifuged at 10000rpm for 5min at 4°C. The supernatant (Ex 280nm, Em 310nm). The HPLC system consisted of a
was analyzed by the HPLC-fluorescence method (Ex 338nm, Hitachi model L7100 pump, Hitachi D-7500 detector, and Hi-
Em 458nm). The HPLC system consisted of a Hitachi model tachi F-1050 fluorescence spectrophotometer.
L7100 pump, Hitachi D-7500 detector, and Hitachi F-1050
fluorescence spectrophotometer.
Assays of inhibition of CYP3A4 activity (GENTEST Co.,
Results and Discussion
We selected menthofuran26) (1) and methoxalen (2) as lead
Human CYP3A4 + cytochrome b5+P450 reductase (Baculo- compounds and prepared various benzofuran and couma-
virus)) by benzofuran and coumarin derivatives were based on rin derivatives (Chart 1). A series of benzofuran derivatives
microsomal testosterone 6β-hydroxylation. The rate of testos- (3,8,9) 4,10,11) 5,12) 6,13) 7,14,15) 8,14,15) 9,16) 10,17,18) and 1119)) and
terone 6β-hydroxylation was determined by the method of Guo coumarin derivatives (13,20) 14,21) 15,22) 16,23) and 1722)) were
et al.7) Expressed human CYP3A4 (2.5pmol) was incubated prepared by the reported procedure. Assays of inhibition of
with testosterone (final concentration of 0.2mM) and an inhibi- CYP2A6 activity by benzofuran and coumarin derivatives
tor (0.01–10 µM) at 37°C in the presence of phosphate buffer were based on microsomal coumarin 7-hydroxylation.24) The
(100 mM, pH 7.4), NADP+ (0.8 mM), glucose-6-phosphate results are shown in Table 1.
(7.9 mM), and glucose-6-phosphate dehydrogenase (1unit/mL)
At first, the inhibitory activities of menthofuran analogues
in a total volume of 500µL. The reaction mixture was prein- (3–6) were compared. None of them exhibited activity
cubated for 2min at 37°C, and the reaction was started by the against CYP2A6 comparable to the activity of menthofuran.
addition of an NADPH-generating system of the microsomal Modification of the cyclohexane ring of menthofuran 1 to af-
solution. The reaction was stopped after 15-min incubation ford aromatic ring analogues (5: 3-methylbenzofuran and 6:
with 3mL of AcOEt. After 11α-hydroxy progesterone (internal 3,6-dimethylbenzofuran) decreased the potency of CYP2A6
standard) had been added, the reaction mixture was centri- inhibition. It was found that aromatization (5, 6) of the cyclo-
fuged for 5min at 2800rpm. The supernatant was analyzed by hexane ring of menthofuran 1 influenced the inhibitory effects
the HPLC-UV method (240nm). The HPLC system consisted on CYP2A6. The inhibitory potency of compound 3, without
of a Shimadzu model LC-10ATvp pump, Shimadzu C-R8A a methyl group at the 6-position on the ring of menthofuran,
detector, and Shimadzu SPD-10Avp UV spectrophotometer.
was decreased compared with that of menthofuran, indicat-
Assays of inhibition of CYP2D6 activity (GENTEST Co., ing that the existence of a methyl group at the 6-position of
Human CYP3A4 + cytochrome b5+P450 reductase (Baculo- menthofuran is also important for potency of CYP2A6 inhi-
virus)) by benzofuran and coumarin derivatives were based bition. 7-Methoxy-4,5,6,7-tetrahydrobenzofuran (4) showed
on microsomal dextromethorphan 3-hydroxylation. The rate poor inhibitory activity against CYP2A6. Methoxalen 2 has a
of dextromethorphan 3-hydroxylation was determined by methoxy group at the 9-position on the furanocoumarin ring.
the method of Madeira et al.25) Expressed human CYP2D6 Thus, the influence of the location of the methoxy group on

1000
Vol. 61, No. 10
Table 1. IC50 Values of Benzofuran and Coumarin Derivatives on Coumarin 7-Hydroxylation
IC50 values (µM)
IC50 values (µM)
Compound
Compound
CYP2A6
CYP2A6
Menthofuran 1
1.27
0.47
4.56
>10
7.35
5.90
4.53
2.20
4.12
10
11
12
13
14
15
16
17
9.62
>10
>10
>10
>10
0.13
Methoxalen 2
3
4
5
6
7
8
9
0.64
4.50
Table 2. IC50 Values of Compounds 8, 15, and 16 for CYP2A6, CYP3A4, and CYP2D6
IC50 values (µM)
Compound
CYP2A6
CYP3A4
CYP2D6
Menthofuran 1
Methoxalen 2
1.27
0.47
>50
>50
>50
5.46
2.20
>50
>50
0.13
0.64
>50
>50
>50
>50
the benzofuran ring (7–11) was studied. 4-Methoxybenzofuran tures and inhibitory effects on CYP2A6 activity. Compound
8 had the strongest inhibitory effect on CYP2A6 among the 6, aromatized analogue of menthofuran 1, showed a reduced
benzofuran derivatives. Interestingly, bergapten 12 (methoxy inhibitory effect on CYP2A6. Although the structure of
group at the 4-position on the furanocoumarin ring) showed CYP2A6-menthofuran has not yet been determined by X-ray
a much weaker inhibitory effect on CYP2A6 than that of crystallography, formation of furan epoxide was detected in
methoxalen 2 (methoxy group at the 9-position on the furano- menthofuran.29) This means that the furan oxygen is posi-
coumarin ring). A reductive furanocoumarin derivative (13) tioned closer than other parts of menthofuran 1 to the heme
showed an IC50 value for CYP2A6 of >10 µM. Apparently, a iron in the active site of CYP2A6. Changes in molecular size
double bond of the furan ring was essential for a potent in- and shape caused by aromatization probably prevent access of
hibitory effect on CYP2A6. Finally, the influence of the loca- the furan part of 6 to the heme iron. Substitution of a methoxy
tion of the methoxy group on the coumarin ring (14–17) was substituent for a methyl substituent on the benzofuran ring re-
studied. It was found that the potency of inhibition was also sulted in increased inhibitory potency toward CYP2A6. It was
very sensitive to location of the methoxy group on the couma- also found that the potency of inhibition was very sensitive to
rin ring. Introduction of a methoxy group at the 5 or 6 posi- location of the methoxy group on the benzofuran ring, which
tion on the coumarin ring resulted in significantly increased is likely to reflect hydrogen-bonding interaction with Asn297
inhibitory potency, whereas the same substituent in the 4 or 8 in the active site of CYP2A6.
position gave rise to low potency of CYP2A6 inhibition.
Next, we examined coumarin derivatives. Furanocoumarin
To better understand substrate (coumarin) and inhibitor derivative having a dihydrofuran ring 13 showed poor inhibi-
(methoxalen) binding to CYP2A6, Johnson and colleagues tory activity. This result was reasonable because it is believed
determined the structure of the enzyme by X-ray crystal- that interaction of CYPs with furanocoumarin occurs at the
lography.27) The CYP2A6 structure shows a compact, hydro- part of the unsaturated furan ring.30) In this experiment, 15
phobic active site with one hydrogen bond donor, Asn297. candidate molecules were selected and tested for inhibition
In the crystal structure, Asn297 acts as a hydrogen donor to potency. 5-Methoxycoumarin 15 was most potent of all tested
the carbonyl oxygen of coumarin, and the 7-position of cou- compounds as a CYP2A6 inhibitor with an IC50 value of 0.13.
marin is closest to the heme iron in active site of CYP2A6 Rahnasto et al. reported that structural characteristics such as
for reigioselective oxidation.28) These results and our findings located hydrophobic and hydrogen donor features are impor-
provide a clue for elucidating the relationship between struc- tant for inhibition potency.31) We speculated that the presence

October 2013
1001
Chem., 69, 9100–9108 (1981).
9) Gingerich S. B., Campbell W. H., Bricca C. E., Jennings P. W., J.
Org. Chem., 46, 2589–2591 (1989).
10) Kuroda C., Nishio E., Nat. Prod. Commun., 2, 581–585 (2007).
11) Xie X., Chen B., Lu J., Han J., She X., Pan X., Tetrahedron Lett.,
45, 6235–6237 (2004).
12) Suzuki T., Tanemura K., Horaguchi T., Shimizu T., Sakakibara T.,
J. Heterocycl. Chem., 29, 423–429 (1992).
13) Adam W., Albrecht O., Feineis E., Reuther I., Saha-Möller C. R.,
Seufert-Baumbach P., Wild D., Liebigs Ann. Chem., 33–40 (1991).
14) Einhorn J., Demerseman P., Royer R., J. Heterocycl. Chem., 22,
1243–1247 (1985).
15) Kneen G., Maddocks P. J., Synth. Commun., 16, 1635–1640 (1986).
16) Barker P., Finke P., Thompson K., Synth. Commun., 19, 257–265
(1989).
17) Coleman P. J., Brashear K. M., Askew B. C., Hutchinson J. H.,
McVean C. A., Duong L. T., Feuston B. P., Fernandez-Metzler C.,
Gentile M. A., Hartman G. D., Kimmel D. B., Leu C.-T., Lipfert
L., Merkle K., Pennypacker B., Prueksaritanont T., Rodan G. A.,
Wesolowski G. A., Rodan S. B., Duggan M. E., J. Med. Chem., 47,
4829–4837 (2004).
of a hydrogen bond between a 5-methoxy group of couma-
rin and Asn297 in the active site of CYP2A6 has an effect
on inhibition potency. As previously described, coumarin is
known as a substrate of CYP2A6.32) We found that coumarin
derivatives work not only as substrates of CYP2A6 but also as
inhibitors.
In order to further explore the selectivity of CYP2A6 in-
hibition, compounds 8, 15, and 16—a selection of the most
potent inhibitors of CYP2A6 among compounds examined
in this study—were tested for their inhibitory potency for the
most prominent drug-metabolizing enzymes CYP3A4 and
CYP2D6.7) The results are shown in Table 2.
As stated above, a number of compounds, including
methoxalen, have been reported to be inhibitors of CYP2A6.
However, these compounds generally lack selectivity for
targeting CYP2A6, and many also inhibit other drug-metab-
olizing enzymes such as CYP3A4A and 2D6.6) Menthofuran,
an excellent selective inhibitor of CYP2A6, also has disadvan-
tages in toxicity and stability.33) Fortunately, of the compounds
examined, compounds 8, 15, and 16 were observed to have 18) Martín-Santamaría S., Rodríguez J. J., de Pascual-Teresa S., Gordon
S., Bengtsson M., Garrido-Laguna I., Rubio-Viqueira B., López-Ca-
sas P. P., Hidalgo M., de Pascual-Teresa B., Ramos A., Org. Biomol.
Chem., 6, 3486–3496 (2008).
highly selective inhibitory potency for CYP2A6. These com-
pounds may serve as lead structures to develop even more po-
tent and selective inhibitors of CYP2A6 that are more acces-
sible than menthofuran and methoxalen and have comparable
activity against CYP2A6.
The dose–response relationship between the number of cig-
19) Musser J. H., Chakraborty U., Bailey K., Sciortino S., Whyzmuzis
C., Amin D., Stutherland C. A., J. Med. Chem., 30, 62–67 (1987).
20) Row E. C., Brown S. A., Stachulski A. V., Lennard M. S., Org. Bio-
mol. Chem., 4, 1604–1610 (2006).
21) Nickisch K., Klose W., Nordhoff E., Bohlmann F., Chem. Ber., 113,
arettes smoked and development of related cancer was estab-
lished many years ago by epidemiological studies.34) Reducing
3086–3088 (1980).
the number of cigarettes smoked and decreasing the toxic-
ity and carcinogenicity of tobacco will result in less tobacco-
related cancer risk. One approach to decrease smoking is to
develop an agent that replaces nicotine and that inhibits the
formation of nicotine-related carcinogens. Preparation of po-
tent and selective inhibitors of human CYP2A6 may provide
an insight into small molecules that can be used as smoking
cessation agents.
22) Harayama T., Nakatsuka K., Nishioka H., Murakami K., Hayashida
N., Ishii H., Chem. Pharm. Bull., 42, 2170–2173 (1994).
23) Narasimhan N. S., Mali R. S., Barve M. V., Synthesis, 906–909
(1979).
24) Yamazaki H., Tanaka M., Shimada T., J. Chromatogr. B. Biomed.
Appl., 729, 13–19 (1999).
25) Madeira M., Levine M., Chang T. K. H., Mirfazaelian A., Bellward
G. D., Drug Metab. Dispos., 32, 460–467 (2004).
26) Khojasteh-Bakht S. C., Koenigs L. L., Peter R. M., Trager W. F.,
Nelson S. D., Drug Metab. Dispos., 26, 701–704 (1998).
27) Yano J. K., Hsu M.-H., Griffin K. J., Stout C. D., Johnson E. F., Nat.
Struct. Mol. Biol., 12, 822–823 (2005).
Acknowledgment This study was partially supported by
the Hikari Foundation for Medicinal Research (to KO).
28) They reported that CYP2A6 has a more compact structure and a
smaller, less pliant active site cavity and that its volume is less than
25% of the volumes calculated for structures of the human drug-
metabolizing CYPs 2C8, 2C9 and 3A4.
References and Note
1) Shimada T., Yamazaki H., Mimura M., Inui Y., Guengerichi F. P., J.
Pharmacol. Exp. Ther., 270, 414–423 (1994).
29) Khojasteh-Bakht S. C., Chen W., Koenigs L. L., Peter R. M., Nelson
S. D., Drug Metab. Dispos., 27, 574–580 (1999).
30) Koenigs L. L., Trager W. F., Biochemistry, 37, 10047–10061 (1998).
31) Rahnasto M., Wittekindt C., Juvonen R. O., Turpeinen M., Petsalo
A., Pelkonen O., Poso A., Stahl G., Höltje H.-D., Raunio H., Phar-
macogenetics J., 2007, 1–11.
2) Rendic S., Di Carlo F. J., Drug Metab. Rev., 29, 413–580 (1997).
3) Kamataki T., Fujieda M., Kiyotani K., Iwano S., Kunitoh H., Bio-
chem. Biophys. Res. Commun., 338, 306–310 (2005).
4) Takeuchi H., Saoo K., Yokohira M., Ikeda M., Maeta H., Miyazaki
M., Yamazaki H., Kamataki T., Imaida K., Cancer Res., 63, 7581–
7583 (2003).
32) Cashman J. R., Park S. B., Yang Z.-C., Wrighton S. A., Jacob P. 3rd,
Benowitz N. L., Chem. Res. Toxicol., 5, 639–646 (1992).
33) Thomassen D., Slattery J. T., Nelson S. D., J. Pharmacol. Exp.
Ther., 244, 825–829 (1988).
34) Hoffmann D., Hoffmann I., El-Bayoumy K., Chem. Res. Toxicol.,
14, 767–790 (2001).
5) Guengerich F. P., Shimada T., Yun C. H., Yamazaki H., Raney K.
D., Their R., Coles B., Harris T. M., Environ. Health Perspect., 102
(Suppl. 9), 49–53 (1994).
6) Denton T. T., Zhang X., Cashman J. R., J. Med. Chem., 48, 224–239
(2005).
7) Guo L.-Q., Taniguchi M., Xiao Y.-Q., Baba K., Ohta K., Yamazoe
Y., Jpn. J. Pharmacol., 82, 122–129 (2000).
8) Bagal S. K., Adlington R. M., Baldwin J. E., Marquez R., J. Org.