Prenylflavonoids from Glycyrrhiza uralensis and their protein tyrosine phosphatase-1B inhibitory activities

Article abstract of DOI:10.1016/j.bmcl.2010.07.110

Two new 2-arylbenzofurans, glycybenzofuran (1) and cyclolicocoumarone (2), together with 10 known flavonoids including licocoumarone (3), glycyrrhisoflavone (4), glisoflavone (5), cycloglycyrrhisoflavone (6), isoliquiritigenin (7), licoflavone A (8), apig

Full text of DOI:10.1016/j.bmcl.2010.07.110

Bioorganic & Medicinal Chemistry Letters 20 (2010) 5398–5401
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Prenylflavonoids from Glycyrrhiza uralensis and their protein tyrosine
phosphatase-1B inhibitory activities
b
c
a
Songpei Li a, Wei Li a, , Yinghua Wang , Yoshihisa Asada , Kazuo Koike
*
a Faculty of Pharmaceutical Sciences, Toho University, Miyama 2-2-1, Funabashi, Chiba 274-8510, Japan
b Ningxia Institute for Drug Control, Yinchuan 750004, People’s Republic of China
c Faculty of Pharmaceutical Sciences, Tokyo University of Science, Yamazaki 2641, Noda, Chiba 278-8510, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Two new 2-arylbenzofurans, glycybenzofuran (1) and cyclolicocoumarone (2), together with 10 known
flavonoids including licocoumarone (3), glycyrrhisoflavone (4), glisoflavone (5), cycloglycyrrhisoflavone
(6), isoliquiritigenin (7), licoflavone A (8), apigenin (9), isokaempferide (10), glycycoumarin (11), and iso-
glycycoumarin (12), were isolated from the roots of Glycyrrhiza uralensis and their structures were deter-
mined by extensive spectroscopic analyses. Compounds 1 and 5 showed significant protein tyrosine
Received 23 June 2010
Accepted 26 July 2010
Available online 3 August 2010
Keywords:
Glycyrrhiza uralensis
Flavonoids
phosphatase-1B (PTP1B) inhibitory activity in vitro with the IC50 values of 25.5 and 27.9 lM, respectively.
The structure–activity relationship indicated that the presence of prenyl group and ortho-hydroxy group
is important for exhibiting the activity. Kinetic analysis indicated that compound 1 inhibits PTP1B by a
competitive mode, whereas compound 5 by a mixed mode.
Protein tyrosine phosphatase-1B
Ó 2010 Elsevier Ltd. All rights reserved.
Type 2 diabetes mellitus (T2DM) has become one of the most
serious health problems worldwide.1 The etiology of T2DM is mul-
ti-factorial and normally characterized by b-cell dysfunction and
insulin resistance.2 On the other hand, obesity is one of the major
risk factors for developing T2DM due to insulin resistance.3,4 Pro-
tein tyrosine phosphatase-1B (PTP1B) is an intracellular protein
tyrosine phosphatase, expressing ubiquitously in the classical insu-
lin-targeted tissues such as liver, muscle, and fat. PTP1B plays a key
role as a negative regulator in both insulin and leptin signaling by
catalyzing the dephosphorylation of activated insulin receptor,
insulin receptor substrate proteins, and leptin receptor-associated
kinase, Jak2.5,6 Therefore, inhibition of PTP1B has been demon-
strated to be an effective therapeutic approach for T2DM and
obesity treatment by biochemical, genetic and pharmacological
studies.7
The roots of Glycyrrhiza uralensis (Leguminosae) is one of the
oldest and most frequently used crude drugs in traditional Chinese
medicines for its extensive pharmacological effects, such as anti-
inflammatory, antiprotozoal and antioxidative activity.8 The effec-
tiveness of flavonoids from G. uralensis in ameliorating diabetes
and abdominal obesity has been reported in diabetic KK-Ay mice
model and high fat diet-induced obese C57BL mice model.9
During our ongoing chemical studies on the bioactive natural
compounds from traditional Chinese medicines, approximately
200 Chinese medicinal plant extracts have been screened for
their PTP1B inhibitory activity. Among them, the 95% EtOH ex-
tract of G. uralensis roots showed significant PTP1B inhibitory
activity, yielding up to 68.8% inhibition at 30 lg/mL concentra-
tion. Further bioassay-guided fractionation of this extract led to
the isolation of 12 flavonoids including two new compounds.
Herein, we report the isolation and structural determination of
these flavonoids, and the evaluation of their PTP1B inhibitory
activities.
The roots of G. uralensis were collected from Yanchi, Ningxia Hui
Autonomous Region, People’s Republic of China. The 95% EtOH ex-
tract (200 g) from the roots of G. uralensis (1 kg) was loaded onto a
Diaion HP-20 open column and sequentially eluted with 70% and
90% aqueous MeOH. The 90% MeOH fraction (33.4 g) was further
separated by a silica gel open column with a gradient of CHCl3–
MeOH. The CHCl3–MeOH (95:5) eluting fraction (4.5 g) was sepa-
rated by repeated preparative HPLC with MeOH–H2O (0.06% TFA)
(70:30) to afford two new flavonoids, 1 (4 mg) and 2 (4 mg), to-
gether with ten known compounds, namely, licocoumarone (3,
4 mg),10 glycyrrhisoflavone (4, 6 mg),11 glisoflavone (5, 4 mg),12
cycloglycyrrhisoflavone (6, 5 mg),11 isoliquiritigenin (7, 24 mg),13
licoflavone A (8, 2 mg)14, apigenin (9, 2 mg),15 isokaempferide
(10, 7 mg),16 glycycoumarin (11, 74 mg),17 and isoglycycoumarin
(12, 8 mg).12 The structures of the known compounds 3–12 were
identified by comparison of their spectroscopic data with those
in the literature.
* Corresponding author. Tel.: +81 47 4721161; fax: +81 47 4721404.
E-mail address: liwei@phar.toho-u.ac.jp (W. Li).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.07.110

S. Li et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5398–5401
5399
RO
7
4
1
O
6' 5'
O
O
RO
6
RO
8
9
O
1'
2
4'
OH
OR
OR
5
3
3'
2'
1'' 2''
HO
CH3
RO
OCH3
OCH3
OCH3
10
5''
3''
4''
3
14
R=H
R=CH3
1
R=H
2
13 R=CH3
HO
O
OH
O
HO
HO
OH
O
OR
O
O
OH
OH
OH
O
OH
R=H
4
6
7
5
R=CH3
OH
OH
OH
HO
O
HO
O
O
HO
O
O
OCH3
OH
O
OH
8
9
10
RO
O
O
O
O
O
OCH3
RO
OR
OCH3
HO
OH
11 R=H
12
15
R=CH3
Compound 1 was obtained as a pale brown solid.18 Its molecular
formula was determined as C21H22O5 by the positive-ion HRESIMS
data (m/z 377.1362, [M+Na]+). The 1H NMR spectrum (Table 1)
showed a set of ABX-type aromatic proton resonances at d 7.29
(1H, d, J = 8.2 Hz), 6.73 (1H, dd, J = 8.2, 1.8 Hz) and 6.84 (1H, d,
3.27 (2H, d, J = 6.4 Hz), 5.21 (1H, m), 1.75 (3H, s) and 1.67 (3H, s),
together with three separated singlet resonances for a methoxy
HO
HO
O
OH
J = 1.8 Hz), and a set of
c,c-dimethylallyl proton resonances at d
CH3
OCH3
Table 1
1H and 13C NMR spectroscopic data of 1 in CD3OD
Figure 1. Key HMBC correlations of 1.
Position
dC, mult.
dH (J in Hz)
2
3
4
5
6
7
8
9
146.7, qC
115.1, qC
120.0, CH
112.1, CH
156.1, qC
98.6, CH
157.1, qC
124.5, qC
8.9, CH3
104.6, qC
160.5, qC
114.6, qC
159.1, qC
99.6, CH
156.7, qC
23.5, CH2
125.4, CH
131.0, qC
18.0, CH3
26.0, CH3
61.3, CH3
HO
HO
O
7.29, d (8.2)
6.73, dd (8.2, 1.8)
OH
6.84, d (1.8)
2.05, s
CH3
CH
OCH3
H
10
Figure 2. Key NOESY correlations of 1.
10
20
30
40
50
6.23, s
H3CO
H
60
O
H3CO
100
200
300
400
500
OCH3
3.27, d (6.4)
5.21, m
OCH3
CH3
OCH3
1.75, s
1.67, s
3.36, s
Figure 3. Key NOESY correlations of 13.

5400
S. Li et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5398–5401
HO
O
O
O
at d 3.36, a methyl at d 2.05 and an aromatic proton at d 6.23. The
13C NMR spectrum showed 21 carbon resonances. All protonated
carbons were assigned to their corresponding protons by the
HMQC correlations. The characteristic chemical shift of the methy-
11
CH3O
OH
O
C
lene carbon in the
both ortho-positions to the
c
,
c
-dimethylallyl moiety at d 23.5 indicated that
HOO
R
c,c-dimethylallyl moiety were occu-
peroxidase
pied by the oxygenated substituents.19 In the HMBC spectrum
(Fig. 1), the correlations from dH 3.27 (H2-100) to the oxygenated
aromatic carbons dC 159.1 and 160.5, dH 6.23 to dC 156.7 and
159.1, and dH 3.36 to dC 160.5, established the structure of 1-meth-
H
HO
O
OH
H
O
C
O
O
OH
R
HO
O
CH3O
O
OH
oxyl-2-(c,c-dimethylallyl)-3,5-dihydroxyphenyl moiety. Two qua-
ternary olefinic carbon resonances at d 115.1 and 146.7 suggested
the 2,3-disubstituted benzofuran skeleton of 1. The observed
HMBC correlations and the ABX spin system of the aromatic pro-
tons in benzofuran skeleton indicated the presence of the 6-hydro-
xyl moiety. Finally, the 3-methyl moiety in benzofuran skeleton,
and the aryl moiety at C-2 were further confirmed by the observa-
tion of the HMBC correlation from dH 2.05 to dC 124.0 (C-9) and the
NOESY correlation (Fig. 2) between dH 2.05 (H3-10) and 7.29 (H-4),
respectively. Permethylation of 1 by TMSCHN2 afforded compound
13,20 which showed three additional methoxy resonances than 1.
All methoxy proton resonances were assigned by the NOESY
correlations as shown in Figure 3. In the NOESY spectrum of 13,
the presence of correlations between H3-10/20-OCH3 and H3-10/
60-OCH3, as well as the absence of correlation between H3-10/6-
OCH3, completely eliminated the possibility of 1 being 2-(2,4-dihy-
OCH3
HO
OH
HO
OH
CHO
CH3O
O
OH
3
HO
OH
CH3
1
CH3O
O
OH
Figure 4. Possible biosynthesis of 1 and 3.
Table 2
droxyphenyl)-5-(
ylbenzofuran. Therefore, the structure of 1 was unambiguously
determined as 2-[4,6-dihydroxy-3-( -dimethylallyl)-2-methoxy-
phenyl]-6-hydroxy-3-methyl-benzofuran, and given a trivial name
glycybenzofuran.
Although several 3-carbonated, 3-formylated or 3-hydrox-
ymethylated 2-arylbezofurans have been reported from Oryza
sativa21 and Andira inermis,22 2-aryl-3-methylbezofurans is
found rarely as natural products. To our knowledge, glyc-
ybenzofuran (1) is the first example of 2-aryl-3-methylbezofu-
rans from natural sources. Its possible biosynthesis was
proposed in Figure 4.
c,c-dimethylallyl)-6-hydroxy-4-methoxy-3-meth-
The IC50 values of the active compounds
against PTP1B
c,c
a
Compound
IC50
(lM)
1
3
4
5
25.5 2.2
71.2 2.5
58.7 1.4
27.9 1.4
54.5 2.1
183.9 4.5
4.4 0.1
8
11
RK-682
a
IC50 values were determined by regres-
sion analyses and expressed as mean SD of
three replicates.
Figure 5. Inhibition of PTP1B-catalyzed hydrolysis of p-NPP by compounds 1 and 5. (A) The Lineweaver–Burk plot of the effect of compound 1 on PTP1B. Final concentrations
of 1 were as follows: (d) 3% DMSO, (j) 10 M, (s) 30 M. (B) The Lineweaver–Burk plot of the effect of compound 5 on PTP1B. Final concentrations of 5 were as
M. (C) The second plot of (A). (D) The second plot of (B). The inset shows the second-order plot of (B) with [I]2 as the
l
M, (N) 20
l
l
follows: (d) 3% DMSO, (j) 10
l
M, (?) 20
lM, (s) 30 l
horizontal axis.

S. Li et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5398–5401
5401
Compound 2 was obtained as a pale brown solid.23 Its 1H and 13
C
9. Mae, T.; Kishida, H.; Nishiyama, T.; Tsukagawa, M.; Konishi, E.; Kuroda, M.;
Mimaki, Y.; Sashida, Y.; Takahashi, K.; Kawada, T.; Nakagawa, K.; Kitahara, M. J.
Nutr. 2003, 133, 3369.
10. Watanabe, M.; Hayakawa, S.; Isemura, M.; Kumazawa, S.; Nakayama, T.; Mori,
C.; Kawakami, T. Biol. Pharm. Bull. 2002, 25, 1388.
11. Hatano, T.; Kagawa, H.; Yasuhara, T.; Okuda, T. Chem. Pharm. Bull. 1988, 36,
2090.
12. Hatano, T.; Yasuhara, T.; Fukuda, T.; Noro, T.; Okuda, T. Chem. Pharm. Bull. 1989,
37, 3005.
13. Salem, M. M.; Werbovetz, A. K. J. Nat. Prod. 2006, 69, 43.
14. Kajiyama, K.; Demizu, S.; Hiraga, Y. J. Nat. Prod. 1992, 55, 1197.
15. Li, W.; Dai, R. J.; Yu, Y. H.; Li, L.; Wu, C. M.; Luan, W. W.; Meng, W. W.; Zhang, X.
S.; Deng, Y. L. Biol. Pharm. Bull. 2007, 30, 1123.
16. Gohari, R. A.; Saeidnia, S.; Matsuo, K.; Uchiyama, N.; Yagura, T.; Ito, M.; Kiuchi,
F.; Honda, G. Nat. Med. 2003, 57, 250.
NMR data showed very similar resonances to that of licocoumarone
(3), except for some variations at dY 2.76 (2H, m), 1.71 (2H, m), 1.27
(6H, s), and dC 19.9, 29.1, 44.8 and 71.9, indicating the presence of
the dihydrodimethylpyran ring. Thus, compound 2 was determined
as the cyclization artifact of licocoumarone (3) between the prenyl
group and the hydroxyl group, and is given
cyclolicocoumarone.
a trivial name
All isolated compounds were assayed for PTP1B inhibitory
M concentration with the known PTP1B inhib-
activity first at 100
l
itor RK-682 as positive control.24 For the compounds with more
than 30% inhibition, the inhibitory activities were further mea-
sured at three different concentrations to obtain the IC50 values
by regression analyses (Table 2). Glycybenzofuran (1) and glisof-
lavone (5) showed the strongest inhibitory activity with the IC50
17. Sato, Y.; Akao, T.; He, J. X.; Nojima, H.; Kuraishi, Y.; Morota, T.; Asano, T.; Tani, T.
J. Ethnopharmacol. 2006, 105, 409.
18. Glycybenzofuran (1): pale brown solid; UV (MeOH) kmax (log
e) 205 (4.66), 262
(4.10), 302 (4.16) nm; IR (KBr) mmax 3435, 1629, 1459, 1384, 1146, 1120,
1070 cmꢀ1 1H NMR (CD3OD, 500 MHz) and 13C NMR (CD3OD, 125 MHz) see
;
Table 1; HRESIMS m/z 377.1362 [M+Na]+ (calcd for C21H22O5Na, 377.3862).
19. Fukai, T.; Nomura, T. Heterocycles 1989, 29, 2379.
values of 25.5 lM and 27.9 lM, respectively. It is interesting to
note that all of the active compounds were substituted by a prenyl
and ortho-hydroxy group, suggesting this partial structure maybe
important for the inhibitory activity. The cyclization artifacts 2, 6,
and 12, and the permethylated products23 13, 14,10 and 1510 of
the active compounds 1, 3 and 11, turned out to be inactive, sup-
porting the aforementioned conclusion.
To elucidate the inhibition mode, the inhibition kinetics of glyc-
ybenzofuran (1) and glisoflavone (5) were analyzed by the Linewe-
aver–Burk method with various substrate concentrations of p-NPP
(1, 2, 4, 8, 16 mM). The initial reaction velocities were measured
with (10, 20, 30 lM) and without the inhibitor. The Lineweaver–
Burk plot was shown in Figure 5. Glycybenzofuran (1) showed
the same Vmax value of 6.3 0.2 mM/h, and its Km values were
2.9, 3.5, 4.2, and 5.5 lM, respectively. The Km values increased in
20. Each compound of 1 (1.8 mg), 3 (2.7 mg), and 11 (2.2 mg) was dissolved in
MeOH 0.1 mL. After 30-fold TMSCHN2 (2 M in Et2O) was added, the solutions
were stood at room temperature for 24 h. One drop of acetic acid was added
and then the reaction solutions were evaporated under reduced pressure. The
residues were purified by preparative TLC with CHCl3–MeOH (97:3) to give
compounds 13 (1.5 mg, 81%), 14 (1.5 mg, 56%) and 15 (1.7 mg, 80%).
Trimethylether of compound 1 (13): 1H NMR (CD3OD, 500 MHz) d 1.66 (CH3,
s, 500), 1.75 (CH3, s, 400), 2.03 (2H, d, J = 6.8 Hz, H2-100), 3.37 (3H, s, 20-OCH3), 3.78
(3H, s, 60-OCH3), 3.84 (3H, s, 6-OCH3), 3.90 (3H, s, 40-OCH3), 5.15 (1H, t,
J = 6.9 Hz, H-200), 6.56 (1H, s, H-30), 6.57 (1H, d, 5.1, 2.4, H-50), 6.65 (1H, s, H-8),
7.34 (1H, dd, J = 7.5 Hz, 1.1, H-60), 7.89 (1H, d, J = 0.7 Hz, H-4); 13C NMR
(CD3OD, 125 MHz) d 161.5 (C, C-60), 160.8 (C, C-20), 159.8 (C, C-40), 159.3 (C, C-
6), 155.3 (C, C-8), 146.3 (C, C-2), 131.0 (C, C-300), 125.0 (CH, C-200), 124.2 (C, C-9)
120.1 (CH, C-4), 116.9 (C, C-30), 115.1 (C, C-3), 111.9 (CH, C-5), 106.9 (C, C-10),
96.7 (CH, C-7), 92.8 (CH, C-50), 61.6 (CH3, 20-OCH3), 56.2 (CH3, 6-OCH3), 56.3
(CH3, 40-OCH3), 56.5 (CH3, 60-OCH3), 26.0 (CH3, C-400), 23.5 (CH2, C-100), 18.0
(CH3, C-500), 8.8 (CH3, C-10). The 1H NMR data of licocoumarone trimethylether
(14) and glycycoumarin trimethylether (15) was coincident to the previous
literature values.
dose-dependent manner without changing the Vmax value, indicat-
ing that a competitive inhibition was induced. The second plot of 1
showed a linear relationship and the Ki value was calculated for
21. Han, S. J.; Ryu, S. N.; Kang, S. S. Chem. Pharm. Bull. 2004, 52, 1365.
22. Kraft, C.; Jenett Siems, K.; Kohler, I.; Siems, K.; Abbiw, D.; Bienzle, U.; Eich, E. Z.
Naturforsch. 2002, 57c, 785.
32.0 lM. As for glisoflavone (5), the Lineweaver–Burk plot in Fig-
23. Cyclolicocoumarone (2): UV (MeOH) kmax (log
(sh) (4.19) nm; IR (KBr) mmax 3433, 1623, 1509, 1459, 1385, 1312, 1255, 1212,
1111 cmꢀ1 1H NMR (CD3OD, 500 MHz): d 1.27 (6H, s, H3-400, H3-500), 1.71 (2H,
e) 294 (sh) (4.06), 320 (4.26), 334
ure 5B suggested that 5 inhibited PTP1B by a mixed manner. The
second plot of 5 showed a quadratic-like curve and a good linear
plot was gained with [I]2 as horizontal axis, indicating that 5 might
inhibit PTP1B activity by binding of two molecules to the enzyme.
;
m, H2-200), 2.76 (H, m, H2-100), 4.03 (3H, s, OCH3), 6.40 (1H, dd, J = 2.4, 8.5 Hz, H-
50), 6.41 (1H, d, J = 2.4 Hz, H-30), 6.66 (1H, d, J = 1.0 Hz, H-7), 7.21 (1H, d,
J = 1.0 Hz, H-3), 7.63 (1H, d, J = 8.5 Hz, H-60); 13C NMR (CD3OD, 125 MHz): d
19.9 (C-100), 29.1 (C-400, 500), 44.8 (C-200), 60.8 (4-OCH3), 71.9 (C-300), 93.5 (C-7)
101.9 (C-3), 103.9 (C-30), 108.1 (C-50), 111.6 (C-10), 115.0 (C-9), 117.0 (C-5),
128.1 (C-60), 152.0 (C-2), 152.1 (C-4), 154.6 (C-6), 155.3 (C-8), 156.8 (C-20),
159.3 (C-40); ESIMS (positive) m/z 341 [M+1]+.
References and notes
1. Saltiel, R. A.; Ronald Kahn, C. Nature 2001, 414, 799.
2. Hull, L. R.; Andrikopoulos, S.; Bruce Verchere, C.; Vidal, J.; Wang, F.; Cnop, M.;
Prigeon, L. R.; Kahn, E. S. Diabetes 2003, 52, 372.
3. Gallagher, J. E.; LeRoith, D.; Karnieli, E. Endocrinol. Metab. Clin. North Am. 2008,
37, 559.
4. Hays, P. N.; Galassetti, R. P.; Coker, H. R. Pharmacol. Ther. 2008, 118, 181.
5. Byon, C. H. J.; Kusari, B. A.; Kusari, J. Mol. Cell. Biochem. 1998, 182, 101.
6. Zabolotny, M. J.; Bence Hanulec, K. K.; Stricker Krongrad, A.; Haj, F.; Wang, Y. P.;
Minokoshi, Y.; Kim, Y. B.; Elmquist, K. J.; Tartaglia, A. L.; Kahn, B. B.; Neel, G. B.
Dev. Cell 2002, 2, 489.
24. PTP1B (human, recombinant) was purchased from Enzo Life Sciences, Inc. A
mixture consisting of 2 mM p-NPP and 0.05 lg PTP1B in a buffer containing
0.06 M citrate (pH 6.0), 0.1 M NaCl, 1 mM EDTA, and 1 mM dithiothreitol (DTT)
with or without a tested compound solution (prepared in the above buffer
solution containing 3% DMSO), was incubated at 37 °C for 30 min. The reaction
was terminated by adding 20 lL of 10 M NaOH. The reaction mixture was
blended by a microplate mixer for 5 min and the amount of produced p-
nitrophenol was tested by measuring the absorbance at 405 nm. The blank was
measured in the same way except adding buffer solution instead of the
enzyme.
7. Zhang, S.; Zhang, Z. Y. Drug Discovery Today 2007, 12, 373.
8. Asl, N. M.; Hosseinzadeh, H. Phytother. Res. 2008, 22, 709.