Hydroxycoumarin derivatives: Novel and potent α-glucosidase inhibitors

Article abstract of DOI:10.1021/jm100757r

A novel class of hydroxycoumarin derivatives were found to be potent α-glucosidase inhibitors. Their syntheses were reported and the structure-activity relationship was established. Kinetic enzymatic assays indicated that compound 10 was a slow-binding and noncompetitive inhibitor with a K_i value of 589 nM, while compound 11 was a competitive inhibitor with a K_i value of 4.810 μM. Among all hydroxycoumarin derivatives studied, compounds 10 and 11 exhibited the highest activities, were specific inhibitors of α-glucosidase, and could be exploited as the lead compounds for the development of potent α-glucosidase inhibitors. Compounds 10 and 11 were also selected for further discussion for the mechanism of enzymatic inhibition.

Full text of DOI:10.1021/jm100757r

8252 J. Med. Chem. 2010, 53, 8252–8259
DOI: 10.1021/jm100757r
Hydroxycoumarin Derivatives: Novel and Potent r-Glucosidase Inhibitors
Qiong Shen,*,†,‡,§ Jialiang Shao,‡ Quan Peng,‡ Wanjin Zhang,§ Lin Ma,*,‡ Albert S. C. Chan,§ and Lianquan Gu§
†School of Pharmacy and Centre for Biomolecular Science, University of Nottingham, University Park, Nottingham NG7 2RD, U.K.,
§
‡School of Chemistry and Chemical Engineering, and School of Pharmaceutical Sciences, Sun Yat-Sen University, Guangzhou 510275, P. R. China
Received December 3, 2009
A novel class of hydroxycoumarin derivatives were found to be potent R-glucosidase inhibitors. Their syntheses
were reported and the structure-activity relationship was established. Kinetic enzymatic assays indicated that
compound 10 was a slow-binding and noncompetitive inhibitor with a Ki value of 589 nM, while compound 11
was a competitive inhibitor with a Ki value of 4.810 μM. Among all hydroxycoumarin derivatives studied,
compounds 10 and 11 exhibited the highest activities, were specific inhibitors of R-glucosidase, and could be
exploited as the lead compounds for the development of potent R-glucosidase inhibitors. Compounds 10 and
11 were also selected for further discussion for the mechanism of enzymatic inhibition.
1. Introduction
benzene-1,2,3-triol and ethylacetonacetate gave 7,8-dihydroxy-
4-methyl-2H-chromen-2-one (1) under Pechmann conditions.19
Compounds 2 and 3 were prepared through reaction of
p-quinone with acetic anhydride and subsequently treated
with either ethyl acetonacetate20 or (()-malic acid, as indi-
cated in Scheme 2. Treatment of resorcinol or 2-methylre-
sorcinol with malonic acid using a modified method21
yielded 4 and 5, which were further reacted with p-quinone
to give 10 and 11, respectively. O-Methylation of compounds
4 and 5 with methyl iodide in acetone afforded 40, 400, and 50.
Compound 6 was obtained by a method similar to that for
compounds 4 and 5. Resorcinol and ethylacetoacetate were
treated under Pechmann conditions19 to give substituted
coumarin 7. Scheme 3 illustrates the synthesis of compound
8 by Knoevenagel condensation of salicylaldehyde with
glycine methyl ester in the presence of triethylamine to give
an intermediate, which was subjected to hydrolysis with 38%
sulfuric acid to give compound 8.22 Scheme 4 illustrates the
synthesis of compound 9 by Wittig type reaction to give an
intermediate, which was treated with boron tribromide in
refluxing dichloromethane, leading to compound 9.23
R-Glucosidases are ubiquitous in nature. They are involved in
the synthesis and breakdown of R-linked di-, oligo-, and poly-
saccharides, playing a crucial role in metabolic processes such as
food storage and utilization and in cellular communication
through modification of the glycosylation state of proteins.1,2
Inhibition of R-glucosidases and related enzymes play important
roles in the treatment of diabetes, human immunodeficiency
virus (HIV/AIDS), cancer, and other degenerative diseases.3-6
Indeed several R-glucosidase inhibitors including nojirimycin,
N-butyldeoxy nojirimycin, and castanospermine have been
shown to inhibit HIV replication and HIV-mediated syncy-
tium formation.6-10
Coumarin derivatives have been reported to possess some
biological activities similar to those of R-glucosidase inhibitors
such as anti-HIV activity,11 HIV-1 protease inhibition,12 anti-
mutagenic,13 lipoxygenase and cycloxygenase inhibition,14,15 and
antitumor metastasis.16,17 Moreover, the structure similarity
between hydroxycoumarin derivatives and a potent R-glucosidase
inhibitor genistein18 prompted us to investigate the R-glucosidase
inhibitory activity of hydroxycoumarin derivatives.
Although coumarin derivatives have shown to be active
against various biological targets, its R-glucosidase inhibition
property has not been studied. In the course of looking for
novel, easily accessible, and chemically stable R-glucosidase
inhibitors, we found that hydroxycoumarin derivatives are
good candidates. We propose to examine their R-glucosidase
inhibition properties and to expand the application scope of
this type of easily accessible compound. Herein we report the
preparation and the R-glucosidase inhibition properties of a
series of hydroxycoumarin derivatives and discuss thoroughly
the mechanism of enzymatic inhibition for the first time.
It has been shown previously that treatment of 1,4-benoquinone
with hydroxycoumarins afforded 2,3-disubstituted but not
2,5-disubstituted 1,4-benoquinones in a region-selective
fashion.24 The X-ray structure of 10 confirmed it to be a
2,3-disubstituted 1,4-p-dihydroxyphenol (Figures 1 and 2).25
However, a 1,4-benzoquinone structure was indicated for 11,
1
based on results from mass spectrometry and H NMR in
comparison with published data.24
2.2. Biological Assay. Both p-nitrophenyl (PNPa) glyco-
sides and R-glucosidase (from baker’s yeast) were purchased
from Sigma (St. Louis, MO). Compounds 12, 13, and 14
2. Chemistry and Biological Assay
a Abbreviations: PNP, p-nitrophenyl; CD, circular dichroism, DMF,
dimethylformamide; DMSO, dimethylsulfoxide; ESI, electrospray ioniza-
tion; HPLC, high pressure liquid chromatography; LC, liquid chromatog-
raphy; RT, room temperature; MS, mass spectrometry; NMR, nuclear
magneticresonance; THF, tetrahydrofuran; TLC, thin layer chromatogra-
phy; UV, ultraviolet; NADþ, nicotinamide adenine dinucleotide coenzyme;
TMS, tetramethylsilane; FAB, fast-atom bombardment; DNS, dinitrosa-
licyclic acid; PNPG, p-nitrophenyl-β-D-galactopyranoside; IC50, concen-
tration required for 50% of full inhibitory effect.
2.1. Synthesis of Compounds 1-11. The general synthesis is
described in Schemes 1-4. Condensation reaction between
*To whom correspondence should be addressed. For Q.S.: phone, 86-
20-84110688; fax, 86-20-84112786; e-mail,nespq@mail.sysu.edu.cn. For
L.M.: phone, 86-20-84113690; fax, 86-20-84112245; e-mail, cesmal@
mail.sysu.edu.cn..
r
pubs.acs.org/jmc
Published on Web 11/05/2010
2010 American Chemical Society

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 23 8253
Scheme 1a
a Reagents and conditions: (a) ethyl acetonacetate, H2SO4, 0-5 °C; (b) malonic acid, ZnCl2, POCl3, 60-65 °C, 12 h; (c) 4 or 5, p-quinone, 50% aq
acetone, 50 °C, 10 h; (d) 40, 400, 50, anhydrous K2CO3, anhydrous acetone, iodomethane, reflux, 8 h.
Scheme 2a
constants (Ki) were generated by the Lineweaver-Burk
plots.28 Circular spectroscopy was used to examine the com-
pounds affecting on the secondary structure of R-glucosidase
protein.29 Selected compounds were evaluated for β-
glucosidase30,31 and β-D-galactosidase inhibition.32,33
3. Results and Discussion
a Reagents and conditions: (a) acetic anhydride, concentrated sulfuric
The compounds were tested in R-glucosidase enzymatic
assay using yeast R-glucosidase as described in the Experi-
mental Section. 5,7-Dihydroxy-4-methyl-2H-chromen-2-one
(12), deoxynojirimycin (13), and genistein (14) (Figure 3) were
used for a comparative study. The results are shown in Table 1.
Compound 10 was the most potent R-glucosidase inhibitor
with an IC50 value of 0.86 μM, being ∼14 times more potent
than genistein (14, IC50 = 12.36 μM). Compound 11 also
exhibited potent inhibitory activity with an IC50 value of 2.82
μM, being more effective than genistein, a known R-glucosidase
inhibitor. Hydroxycoumarin derivatives (1-5) had signifi-
cantly reduced inhibitory activity with IC50 in the range
38-94 μM. Nevertheless 4,7-dihydroxy-2H-chromen-2-one
(4) and compound 8 gave modest activity with IC50 values of
38.85 and 35.30 μM, having a potency similar to that of the
reference inhibitor deoxynojirimycin (13). Compounds 40, 400,
and 50 showed weak inhibitory activity. Compound 7 has
potency similar to that of compounds 1 and 5. Compounds 3,
6, and 9 were not active at the highest concentrations tested.
Analysis of the structure-activity relationship indicated
that the introduction of either a methyl or hydroxyl group at
the 8-position of hydroxycoumarin (4) resulted in a decreased
R-glucosidase inhibitory activity as shown by compounds 1
acid, 40-50 °C; (b) (()-malic acid, 75% sulfuric acid, 80 °C; (c) ethyl
acetoacetate, 75% sulfuric acid, 80 °C.
Scheme 3a
a Reagents and conditions: (a) glycine methyl ester, Et3N; (b) hydro-
lysis, 38% sulfuric acid.
Scheme 4a
a Reagents and conditions: (a) Ph3PdCHCOOMe, toluene, reflux;
(b) BBr3, CH2Cl2, reflux.
(Figure 3) were purchased from Aldrich-Sigma and used for
a comparative study.
All compounds were tested for their ability to inhibit
R-glucosidase using the reported procedures.26,27 The inhibition

8254 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 23
Shen et al.
H-bonding donor/acceptor, while the methoxy group can just
act as an H-bonding acceptor. Compounds 10 and 11, posses-
sing excellent enzymatic inhibiting activity, were favorable for
binding to the R-glucosidase. 10 was more potent than 11,
confirming further deactivation of the methyl group at the
8-position of hydroxycoumarin, as shown also in 5. It is
assumed that the hydroxyl groups at C-1 and C-4 of 10 may
serve as hydrogen donors to form hydrogen bonds with the
enzyme, while 11 may not be able to form such interactions.
We evaluated the selectivity of potent several R-gluco-
sidase inhibitors for two sugar hydrolases (β-glucosidase
and β-D-galactosidase) inhibitions (Tables 2 and 3). As indi-
cated, compound 8 neither inhibited β-glucosidase nor β-D-
galactosidase. Compounds 10 and 11 showed weak inhibition on
β-glucosidase and β-D-galactosidase (IC50 > 200 μM). Com-
pound 12 showed modest inhibitory activity against β-gluco-
sidase and β-D-galactosidase. Compounds 4 and 5 did not
inhibit the β-D-galactosidase. Compound 5 gave some inhibi-
tion against the β-glucosidase (IC50 = 87.6 μM), whereas
compound 4 had potency similar to that of compounds 10
and 11. Therefore, compounds 4, 8, 10, and 11 could belong to
specific R-glucosidase inhibitors.
Figure 1. X-ray crystal structure of 10:25 asymmetric unit, with
displacement ellipsoids drawn at the 50% probability level.
To further investigate the inhibitory effect of com-
pounds 10 and 11 on the activity of R-glucosidase, we also
performed dose-response experiments. These results (not
shown) showed that the activity of R-glucosidase was de-
creased by compounds 10 and 11 in a dose-responsive
manner, indicating that these two compounds have strong
affinity toward R-glucosidase.
Slow binding, or slow onset of inhibition, is a widespread
phenomenon among potent glycosidase inhibitors. In order to
support our point, we investigated whether the IC50 values
vary by preincubation of the inhibitor 10 with the enzyme.
These results show that the IC50 value for the inhibition of
R-glucosidase vary depending on the time of pretreatment
with compound 10. When the substrate and compound 10
were added simultaneously, the IC50 was 88.36 μM. This value
decreased almost 100-fold when R-glucosidase was treated
with compound 10 at 37 °C for 1 h before the initiation of
enzyme reaction (not shown). The data indicate that the
inhibition mode seems to be a slow-binding inhibitory mode.
The mechanism of the representative inhibitors binding to
the R-glucosidase was further studied by analysis of the Line-
weave-Burk plots.28 The Ki value was calculated using the
values of Vmax obtained at 1.00 and 3.00 μM for compound 10
and the values of Vmax obtained at 2.00 and 5.00 μM for
compound 11, respectively. It revealed that compound 10 was
a noncompetitive inhibitor with a Ki value of 0.587 μM
(Figure 4). In contrast, compound 11 was a competitive
inhibitor with a Ki value of 4.81 μM.
Figure 2. Compound 1025 crystallizes from dimethylformamide
(DMF) as a trisolvate.
Figure 3. Structures of 12-14.
and 5 in comparison with 4. An additional hydroxyl or a
methyl substituted at 6-position was not tolerated; thus, com-
pounds 2, 3, 6, and 9 almost completely abolished the activity.
In contrast, additional hydroxyl substituted at 3-position
played a major role in the inhibitory activity, so compound
8 showed modest activity. Introduction of a hydroxyl group at
the 4-position of hydroxycoumarin (4) played a more major
role in the inhibitory activity than that of methyl group at the
4-position of hydroxycoumarin (7). O-Methylation of 4-OH
or/and 7-OH led to a large decrease in the activity as shown by
compounds 40, 400, and 50 in comparison with 4 and 5. This
indicates that 4-OH and 7-OH in compounds 4 and 5 were
crucial to the activity, primarily as an H-bonding donor to
interact with R-glucosidase, since the hydroxyl group is an
The percentage of R-helix and loop (β-turn and random)
and the special ratio of R/β within the secondary structure of
the enzyme molecular of R/β structure relate to the native
conformation.34,35 The special functional domain may be
important for preservation of enzymatic activity. To further
elucidate the inhibitory mechanism, the effects of inhibitors 4,
10, 11 on the secondary structure of R-glucosidase were
studied with circular dichroism spectra.29,36 The results
showed that the inhibitors caused a decline in the percent of
R-helix in a dose dependent manner and changes in the ratios
of R/β, as shown in Table 4. Generally, the helices are the most
stable elements of secondary structure of protein.37 This sug-
gested that inhibitors caused some disturbance in the second-
ary structure and affected the folding of R-glucosidase.29,34

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 23 8255
Table 1. Inhibition of R-Glucosidasea,26,27
a Values are the mean of at least three independent determinations. NA means no inhibition (less than 20% inhibition at 200 μM).
Table 2. Selection of Compounds for Inhibition of β-Glucosidasea,30,31
test sample content ( μmol/L) maximum inhibition (%) IC50 ( μM)
conformational changes may have influence on the potency of
enzyme inhibition.37 The inhibitor induced conformational
changes of specific elements of the secondary structure in the
enzyme molecule may be part of the mechanism of the inhi-
bition required to prevent the hydration of the substrate
binding site and also required to induce the cleft closure to
avoid substrate entrance.34,35,37 This may result in the failure
of formation the active center, a decrease of the substrate
binding, a misfolding of the polypeptide chain, and finally in-
activation of the enzyme. These may have been caused by the
cluster effect, and these results may contribute to a more
complete understanding of the enzyme-inhibitor interaction.
Taken together, the reasons why the compounds could
inhibit the activity of R-glucosidase and why the specific ele-
ments of the secondarystructureinthe enzymemoleculecould
have been changed, even why these two compounds could
have strong affinity toward R-glucosidase, were mainly as-
cribed to the interaction (cluster effect) between the compounds
and enzyme. Compounds 10 and 11 were selected for further
detailed evaluation for the mechanism of enzymatic inhibition.
We know the crystal structure of 10 (Figures 3 and 4) and
note that the 1-hydroxy group serves as a hydrogen-bond
donor to a DMF molecule, whereas the substituent at the
3-position functions as a hydrogen-bond donor to two DMF
molecules,25 indicating that compound 10 could serve as a
hydrogen-bond donor easily. Therefore, we proposed that the
4
400
400
400
200
200
400
36.7
>400
87.6
NA
5
8
0
10
11
12
23.0
15.6
>200
>200
45.4
a Values are the mean of at least three independent determinations.
NA means no inhibition (less than 20% inhibition at 400 μM).
Table 3. Selection of Compounds for Inhibition of β-D-Galactosidasea,32,33
test sample content ( μmol/L) maximum inhibition (%) IC50 ( μM)
4
400
400
400
200
200
400
0
NA
NA
5
0
8
0
NA
10
11
12
2.0
31.7
53.9
>200
>200
>200
a Values are the mean of at least three independent determinations.
NA means no inhibition (less than 20% inhibition at 400 μM).
Apparently, the enzyme undergoes a series of inhibitor-
induced conformational changes as shown in Figure 5 (confor-
mational changes of the of R-glucosidase is most obvious
when compound 11 interacted with the enzyme), but minor

8256 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 23
Shen et al.
Figure 4. Double-reciprocal plots of the inhibition kinetics of yeast
R-glucosidase by compounds 10 (up) and 11 (down). R-Glucosidase
(50 μL, 10 U/mL) was treated with 10 or 11 at 37 °C for 1 h, followed
by varying concentrations of PNP.
Table 4. Effects of Hydroxycoumarin Inhibitors on the Secondary
Structure of R-Glucosidasea
compd
R-helix (%) β-sheet (%) β-turn (%) random (%)
without inhibitor
10 (0.5 μM)
10 (0.3 μM)
without inhibitor
11 (5.0 μM)
11 (8.0 μM)
without inhibitor
4 (30.0 μM)
4 (15.0 μM)
35.00
32.70
33.40
32.60
26.00
25.40
50.10
37.10
47.10
33.30
30.00
23.80
27.60
33.60
34.30
26.70
7.40
6.60
9.20
25.10
28.10
29.40
28.90
31.10
32.00
20.00
34.10
29.40
13.40
10.90
9.30
8.20
3.20
21.40
18.00
5.30
a Values are the mean of at least three independent determinations.
1-hydroxy group and the substituent at the 3-position func-
tions may receive and donate the hydrogen bond from the side
chain of R-glucosidase protein residues because residues are
believed to play an important role in the catalytic mechanism
as the corresponding residues in oligo-1,6-glucosidase.38 The
phenol moiety seems to be an effective chemical group for
binding its catalytic activity.38
The multiple hydroxyl functions of 10 and 11 are critical for
their inhibitory activity toward R-glucosidase. The X-ray
crystal structure34 of maltose in complex with R-glucosidase
indicates that one of its glucose rings is bound by multiple
hydrogen bonds to Asn-153, His-203, Asp-260, Asp-116, and
Arg-263 while hydrophobically stacked between Val-117 and
the β-nicotinamide ring of NADþ. Another ring of maltose
Figure 5. CD (circular dichroism) spectra of enzyme after enzyme
interaction with compounds 4, 10, 11. The CD spectra were recor-
ded with 5 times accumulation.
formed a hydrophobic interaction with Ph-238 to stabilize the
interactions. The structures of compounds 10 and 11 have
some similarity to that of maltose (many hydroxy groups and
six-member cyclic structure), indicating that a similar interac-
tion might be formed between compounds and R-glucosidase.
Serving as a hydrogen-bond, the hydroxy groups at the
1C-/4C-position of 10 donors may contribute to its greater
potency in comparison to 11. p-Dihydroxyphenol has been

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 23 8257
displaced by p-benzoquinone,25,29 and the number to form the
hydrogen bond may decrease in the absence of some phenol
moiety and then result in weakening of the hydrogen bond
interaction. A strong hydrogen bond between the phenolic
oxygen of chalcone derivatives and R-glucosidase protein resi-
dues in the active binding site has been suggested previously,38
supporting the potential role of the phenol group as a surro-
gate for the terminal monosaccharide of the substrate.38,39
Furthermore, the bis-chromen-2-one moieties (extended
π-conjugated system) of 10 or 11 may facilitate the extensive
hydrophobic interactions (π-stacking) within the active bind-
ing site.38,40 The two interactions (mainly hydrogen bond and
hydrophobic interactions) between the inhibitors and enzyme
form a cluster effect to influence the conformation and then
change the activity of the enzyme. The aforementioned results
also strongly suggest that the hydroxyl functions make sub-
stantial contributions to the remarkably enhanced inhibitory
activities as shown by compounds 4 and 5 in comparison with
40, 400, and 50. However, the detailed structural characteristics
of the inhibitor in complex with R-glucosidase can only be
fully elucidated by X-ray crystallography.
In summary we have prepared a class of novel R-glucosi-
dase inhibitors bearing a coumarin skeleton. The lead com-
pounds 10 and 11 have demonstrated excellent inhibitory
activities against R-glucosidase with IC50 values in a range
from low micromolar to subnanomolar and specific inhibitors
of R-glucosidase. The mechanism studies indicated that the
hydrogen bond and extensive hydrophobic interactions be-
tween the inhibitors and enzyme are in a cooperative fashion
(cluster effect) to have influence on the conformation of the
enzyme, to change the specific elements of the secondary
structure, and finally to cause inactivation of the enzyme.
Knowledge gained in this study could help future develop-
ments for searching for new R-glucosidase inhibitors.
temperature for 11/2 h. It was then cooled to room temperature,
filtered, and the precipitate was washed to give 2.
6,7-Dihydroxynaphthalen-2(1H)-one (3). A similar procedure
to that described above was employed.
4,7-Dihydroxy-2H-chromen-2-one (4). A mixture of resorcin-
ol (11 g), malonic acid (12.3 g), anhydrous zinc chloride (44 g),
and phosphorus oxychloride (33 mL) was heated with stirring at
60-65 °C for 12 h. The mixture was cooled and quenched with
ice-water. The crude product was collected and dissolved in 5%
sodium carbonate. An oily byproduct was removed and acid-
ification of the remaining solution gave the residues which were
recrystallized from EtOAc to yield 4. Yield 62%, mp 244-
246 °C. MS: m/z 179 (M þ H)þ. 1H NMR (acetone-d6) δ: 7.68
(d, 1H, J=9.0 Hz, 5-H), 6.82 (d d, 1H, J=2.4 Hz, 6-H), 6.71 (d,
1H, J=2.4 Hz, 8-H), 5.55 (d, 1H, J=3.0 Hz, 3-H). Anal. Calcd
for C9 H6O4: C, 60.68; H, 3.39. Found: C, 60.53; H, 3.48.
4-Methoxy-7-hydroxy-2H-chromen-2-one (40).To a 25 mL round-
bottom flask, compound 4 (4.4 mmol), iodomethane (4.8 mmol),
K2CO3 (6.6 mmol), and anhydrous acetone (50 mL) were added.
The mixture was refluxed for 8 h. After the mixture was cooled to
RT, solvent was removed in vacuo, and the residue was dissolved in a
1:1 EtOAc/H2O mixture. The product was extracted with EtOAc,
washed with H2O and brine, then dried over anhydrous Na2SO4.
Solvents were evaporated under reduced pressure. The residue was
purified by column chromatography through silica gel (DCM/
EtOAc) to get compound 40. Yield 25%, ESI-MS: 191 (M - H)-.
1H NMR (300 MHz, DMSO) δ: 3.80 (s, 3H), 5.99 (s,1H), 6.67 (dd,
J=9.39, J=2.28, 1H), 6.75(dd, J=9.39, J=2.28, 1H), 7.95(d, J=
2.28,1H). Anal. Calcd for C10H8O4: C, 62.50; H, 4.20. Found: C,
62.41; H, 4.26.
4,7-Dimethoxyy-2H-chromen-2-one (400). 400 was obtained by
a similar method described above. Yield 32%. ESI-MS: m/z 205
(M - H)-. 1H NMR (300 MHz, CDCl3) δ: 3.87 (s, 3H), 3.97 (s,
3H), 5.56 (s, 1H), 6.79 (d, J=2.4 Hz, 1H), 6.83 (dd, J=8.4 and
2.4 Hz, 1H), 7.69 (d, J=8.4 Hz, 1H). Anal. Calcd for C11H10O4:
C, 64.07; H, 4.89. Found: C, 64.00; H, 4.92.
4,7-Dihydroxy-8-methyl-2H-chromen-2-one (5). 5 was obtai-
ned by the method described to get 4 above. Yield 75%, mp
269-271 °C. MS: m/z 205 (M - H)-. 1H NMR (acetone-d6) δ:
7.54 (d, 1H, J=9.0 Hz, 5-H), 6.88 (d, 1H, J=9.0 Hz, 6-H), 5.51
(s, 1H, 3-H), 2.22 (s, 3H, 8-CH3). Anal. Calcd for C11H10O4: C,
64.07; H, 4.89. Found: C, 64.01; H, 4.95.
4. Experimental Section
Chemistry. Melting points were measured using an Electro-
thermal 8103 apparatus and were uncorrected. IR spectra were
recorded with Perkin-Elmer 398 and FT spectrophotometers.
1H NMR spectra were recorded on a Varian Unity INOVA
500 MHz spectrometer with TMS as internal standard. The
values of the chemical shift (δ) are given in ppm, and the coupling
constants (J) are in Hz. All chemicals used were of analytical grade.
Progress of the reactions was monitored by TLC on precoated 60
F254 silica gel plates (0.25 mm, Merck). ESI-MS spectra were
performed on a Thermo Finigan LCQ DECA XP ion trap mass
spectrometer. Elemental analysis was carried out on an Elementar
Vario EL CHNS elemental analyzer. All compounds tested in vitro
were determined to be of >95% purity by HPLC and EA. The
purity of compound 10 is determined to be almost 100%, as it is
a single crystal.
7,8-Dihydroxy-4-methyl-2H-chromen-2-one (1). A mixture of
pyrogallol (12.6 g, 0.1 mol) and concentrated H2SO4 (100 mL)
was stirred at -5 to 5 °C. Ethyl acetoacetate was added
cautiously to the reaction mixture, and the solution was kept
at -5 to 5 °C for 2 h with stirring. The mixture was then poured
into cold water (1 L), and the precipitate was collected, washed
with cold water, and crystallized from EtOH.
4-Methoxy-7-hydroxy-8-methyl-2H-chromen-2-one (50). 50 was
obtained by a similar method described to get the compound 40.
1
Yield 22%. ESI-MS: m/z 219 (M - H)-. H NMR (300 MHz,
CDCl3) δ: 2.29 (s, 3H), 3.91 (s, 3H), 3.96 (s, 3H), 5.56 (s, 1H), 6.81
(d, J = 8.7 Hz, 1H), 7.62 (d, J = 8.7 Hz, 1H). Anal. Calcd for
C12H12O4: C, 65.45; H, 5.49. Found: C, 65.42; H, 5.51
4-Hydroxy-6-methyl-2H-chromen-2-one (6). 6 was obtained
by the method described to get compound 4.
4-Methyl-7-hydroxycoumarin (7). Compound 7 was prepared
based on a literature method.19 Yield 80%. FAB-MS m/z: 177
([M þ H]), 154 ([M - OH]). These data are in agreement with the
literature results.41
3-Hydroxy-2H-chromen-2-one (8). 8 was prepared based on a
literature procedure.22
3,30-(3,6-Dihydroxycyclohexa-1,4-diene-1,4-diyl)bis(4,7-dihy-
droxy-2H-chromen-2-one) (10). A solution of compound 4 and
p-quinone in 50% aqueous acetone was heated at 50 °C for 10 h
with stirring. Solvents were evaporated under reduced pressure.
The residue was purified by column chromatography through
silica gel (cyclohexane/acetone 1:4 v/v) to yield pure compound
6,7-Dihydroxy-4-methyl-2H-chromen-2-one (2).21 Hydroxy-
droquinone triacetate was prepared according to known methods.
A mixture of ethyl acetoacetate (0.045 mol) and hydroxydroqui-
none triacetate (0.045 mol) was stirred for 20 min, and 75%
sulfuric acid (45 mL) was added. The mixture became a homo-
geneous and gave a deep red solution. The red solution was heated
on a water bath with occasional shaking and stirring until the
temperature reached 80 °C. The mixture was maintained at this
10. Yield: 15%, mp 140-142 °C. ESI-MS: 463 (M þ H)þ. H
1
NMR (acetone-d6) δ: 7.74-7.71 (dd, 2H, 50-H), 6.91 (d, 2H,
2-H, 3-H), 6.84-6.80 (dd, 2H, 60-H), 6.63-6.62 (m, 2H, 70-OH).
Anal. Calcd for C24H14O10: C, 62.34; H, 3.04. Found: C, 62.21;
H, 3.15.
2,5-Bis(4,7-dihydroxy-8-methyl-2-oxo-2H-chromen-3-yl)cyclo-
hexa-2,5-diene-1,4-dione (11). 11 was prepared in a similar
manner. Yield: 12%, mp 152-154 °C. ESI-MS: 487 (M - H)-.

8258 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 23
Shen et al.
1H NMR (DMSO-d6) δ: 7.58, 7.40 (dd, 2H, J=6.0 Hz, 50-H),
6.83-6.72 (m, 4H, 2-H, 3-H, 60-H), 2.08 (d, 6H, 80-CH3). Anal.
Calcd for C26H16O10: C, 63.94; H, 3.30. Found: C, 63.82; H, 3.37.
Compounds 12, 13, and 14 were purchased from Aldrich-
Sigma Chemical Co. Ltd. and used for the comparative study.
Enzyme Assays. p-Nitrophenyl (PNP) glycosides and R-gluco-
sidase were purchased from Sigma (St. Louis, MO). R-Glucosidase
enzymatic assay was performed as described previously.26,27
β-Glucosidase activity was assayed30,31at 37 °C by following the
increasing absorbance at 540 nm, accompanying the hydrolysis of
the substrate (salicin). Enzyme (10 mL) was added to 1 mL of the
activity assay system containing inhibitor, 5.0 mM salicin, 0.1 M
NaAc-HAc buffer, pH 5.0. After reaction for 10 min at 40 °C,
1 mL of 3,5-dinitrosalicyclic acid (DNS) reagent1 was added to the
reaction mixture to stop the reaction. Then the mixture was heated
for 5 min in boiling water for color development. The enzyme
activity was calculated by the increased absorption of the reaction
mixture at 540 nm, using glucose as a standard. The assay was
performed in triplicate with five different concentrations around
the IC50 values that were roughly estimated in the first round of
experiments, and the mean values were adopted
(6) Dwek, R. A.; Butters, T. D.; Platt, F. M.; Zitzmann, N. Targeting
glycosylation as a therapeutic approach. Nat. Rev. Drug Discovery
2002, 1, 65–75.
(7) Winchester, B.; Fleet, G. W. Amino-sugar glycosidase inhibitors:
versatile tools for glycobiologists. Glycobiology 1992, 2, 199–210.
(8) Fischer, P. B.; Karlsson, G. B.; Dwek, R. A.; Platt, F. M. N-
Butyldeoxynojirimycin-mediated inhibition of human immunode-
ficiency virus entry correlates with impaired gp120 shedding and
gp41 exposure. J. Virol. 1996, 70, 7153–7160.
(9) Fischer, P. B.; Karlsson, G. B.; Butters, T. D.; Dwek, R. A.; Platt,
F. M. N-Butyldeoxynojirimycin-mediated inhibition of human
immunodeficiency virus entry correlates with changes in antibody
recognition of the V1/V2 region of gp120. J. Virol. 1996, 70, 7143–
7152.
(10) Walker, B. D.; Kowalski, M.; Goh, W. C.; Kozarsky, K.; Krieger,
M.; Rosen, C.; Rohrschneider, L.; Haseltine, W. A.; Sodroski, J.
Inhibition of human immunodeficiency virus syncytium formation
and virus replication by castanospermine. Proc. Natl. Acad. Sci.
U.S.A. 1987, 84, 8120–8124.
(11) Trenor, S. R.; Shultz, A. R.; Love, B. J.; Long, T. E. Coumarins in
polymers: from light harvesting to photo-cross-linkable tissue
scaffolds. Chem. Rev. 2004, 104, 3059–3077.
(12) Kirkiacharian, S.; Thuy, D. T.; Sicsic, S.; Bakhchinian, R.; Kurkjian, R.;
Tonnaire, T. Structure-activity relationships of some 3-substituted-
4-hydroxycoumarins as HIV-1 protease inhibitors. Farmaco 2002, 57,
703–708.
(13) Pillai, S. P.; Menon, S. R.; Mitscher, L. A.; Pillai, C. A.; Shankel,
D. M. Umbelliferone analogues and their potential to inhibit
benzo(a)pyrene- and hydrogen peroxide-induced mutations. J. Nat.
Prod. 1999, 62, 1358–1362.
(14) Kimura, Y.; Okuda, H.; Arichi, S.; Baba, K.; Kozawa, M. Inhibi-
tion of the formation of 5-hydroxy-6,8,11,14-eicosatetraenoic acid
from arachidonic acid in polymorphonuclear leukocytes by various
coumarins. Biochim. Biophys. Acta 1985, 834, 224–229.
(15) Hofmanova, J.; Kozubik, A.; Dusek, L.; Pachernik, J. Inhibitors of
lipoxygenase metabolism exert synergistic effects with retinoic acid
on differentiation of human leukemia HL-60 cells. Eur. J. Pharma-
col. 1998, 350, 273–284.
Assay of β-Galactosidase. The activity of the β-galactosidase
(9.4 unit/mg solid, from Aspergillus oryzae, Sigma Co.) was
measured by the modified method32,33 described.
Enzymatic Kinetics. The reaction was performed according to
the above reaction conditions with inhibitors of various con-
centrations. Inhibition types for the inhibitors were determined
by double-replot of slope versus the reciprocal of the substrate
concentration. Ki values for each inhibitor were determined by
measuring the rate PNP hydrolysis by R-glucosidase at varying
inhibitor concentrations. Data were plotted in Lineweaver-
Burk plots (1/rate vs 1/[substrate]), and Ki values were deter-
mined through a Dixon plot. The Ki reported was estimated by
averaging the Ki values obtained from each of the different inhi-
bitor concentrations.
(16) Sharma, S.; Stutzman, J. D.; Kelloff, G. J.; Steele, V. E. Screening
of potential chemopreventive agents using biochemical markers of
carcinogenesis. Cancer Res. 1994, 54, 5848–5855.
Circular Dichroism Assays. Circular dichroism (CD) was
carried out with a J-810 spectropolarimeter (Jasco) in the UV
range (186-240 nm) at protein concentrations of 0.53-0.66 mg/
mL, using 0.1 cm quartz cuvettes. The CD spectra were recorded
with 5 times accumulation. The data of the secondary structure
were dealt with by using the professional software Secondary
Structure Estimation and Origin 7.0.
(17) Hayes, J. D.; Pulford, D. J.; Ellis, E. M.; McLeod, R.; James, R. F.;
Seidegard, J.; Mosialou, E.; Jernstrom, B.; Neal, G. E. Regulation
of rat glutathione S-transferase A5 by cancer chemopreventive
agents: mechanisms of inducible resistance to aflatoxin B1. Chem.-
Biol. Interact. 1998, 111-112, 51–67.
(18) Lee, D. S.; Lee, S. H. Genistein, a soy isoflavone, is a potent alpha-
glucosidase inhibitor. FEBS Lett. 2001, 501, 84–86.
(19) Sugimoto, H.; Yamanishi, Y.; Iimura, Y.; Kawakami, Y. Done-
pezil hydrochloride (E2020) and other acetylcholinesterase inhibi-
tors. Curr. Med. Chem. 2000, 7, 303–339.
(20) Vliet, E. B. 4-Methylesculetin. Organic Syntheses; Wiley & Sons:
New York, 1941; Collect. Vol. 1, p 360; Org. Synth. 1925, 4, 45.
(21) Shah, V. R.; B., J. L; Shah, R. C. New synthesis of 4-hydro-
xycoumarins. J. Org. Chem. 1960, 25, 677–678.
(22) Kokotos, G.; Tzougraki, C. Synthesis and study of substituted
coumarins;a facile preparation of D,L-o-tyrosine. J. Heterocycl.
Chem. 1986, 23, 87–92.
(23) Dubuffet, T.; Loutz, A.; Lavielle, G. An efficient large scale
synthesis of coumarins by a dealkylative boron-mediated ring
closure of 3-(orthomethoxyaryl)propenoic esters. Synth. Commun.
1999, 29, 929–936.
Acknowledgment. Q.S. expresses thanks for a Marie Curie
Fellowship (Grant PIIF-GA-2008-221486) and The Hong
Kong Polytechnic University ASD fund. The authors thank
Prof. Kevin Shakesheff, Prof. Clive Roberts, Dr. French
Elizabeth, Dr. Anthony Wilson, Dr. Wang Lei Lei, Dr. Zhang
Lei, Dr. Chen Bin, and Bin Huang for their helpful assistance.
The authors thank the anonymous reviewers for their con-
structive comments that have enriched our work.
Supporting Information Available: Experimental and spectro-
scopic details of compounds 1-3 and 6-9 and enzyme assay.
This material is available free of charge via the Internet at http://
pubs.acs.org.
(24) Zhang, S.-L.; Huang, Z.-S.; S., Y.-D.; L., Y.-M.; Y., J.-H.; H., M.;
C., A. S. C.; G., L.-Q. The unique regioselectivity in the formation
of disubstituted-1,4-benzoquinones generated from the reaction of
4-hydroxycoumarins with 1,4-benzoquinone. Tetrahedron Lett.
2006, 47, 6757–6760.
(25) Shen, Q. 3,30-(3,6-Dihydroxy-p-phenylene)bis(4,7-dihydroxy-2H-
chromen-2-one)dimethylformamide trisolvate. Acta Crystallogr.,
Sect. E 2006, 62, o4732–o4733.
References
(1) Dove, A. Seeking sweet relief for diabetes. Nat. Biotechnol. 2002,
20, 977–981.
(2) Davies, G. J.; Wilson, K. S. Trapped in the act of catalysis. Nat.
(26) Tsujii, E.; Muroi, M.; Shiragami, N.; Takatsuki, A. Nectrisine is a
potent inhibitor of alpha-glucosidases, demonstrating activities
similarly at enzyme and cellular levels. Biochem. Biophys. Res.
Commun. 1996, 220, 459–466.
(27) Muroi, M.; Ando, O.; Bols, M.; Takatsuki, A. Processing glucosi-
dase inhibition by 1-azafagomine. Biosci., Biotechnol., Biochem.
2000, 64, 1103–1105.
(28) Segel, I. H. Enzyme Kinetics: Behavior and Analysis of Rapid
Equilibrium and Steady-State Enzyme Systems; Wiley: New York,
1975.
Struct. Biol. 1999, 6, 406–408.
(3) Braun, C.; Brayer, G. D.; Withers, S. G. Mechanism-based inhibi-
tion of yeast alpha-glucosidase and human pancreatic alpha-
amylase by a new class of inhibitors. 2-Deoxy-2,2-difluoro-alpha-
glycosides. J. Biol. Chem. 1995, 270, 26778–26781.
(4) Jacob, G. S. Glycosylation inhibitors in biology and medicine.
Curr. Opin. Struct. Biol. 1995, 5, 605–611.
(5) Mehta, A.; Zitzmann, N.; Rudd, P. M.; Block, T. M.; Dwek, R. A.
Alpha-glucosidase inhibitors as potential broad based anti-viral
agents. FEBS Lett. 1998, 430, 17–22.
(29) Raasch, C.; Armbrecht, M.; Streit, W.; Hocker, B.; Stþrater, N.;
Liebl, W. Identification of residues important for NAD binding

Article
by the Thermotoga maritima alpha-glucosidase AglA, a member of
glycoside hydrolase family 4. FEBS Lett. 2002, 517, 267–271.
(30) Chen, Q. X.; Zhang, Z.; Zhou, X. W.; Zhuang, Z. L. Kinetics of
inhibition of beta-glucosidase from Ampullarium crossean by bro-
moacetic acid. Int. J. Biochem. Cell Biol. 2000, 32, 717–723.
(31) Bernfeld, P. Methods in Enzymology; Academic Press Inc.: New York,
1955; Vol. 1.
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 23 8259
(37) Marques, M. R.; Vaso, A.; Neto, J. R.; Fossey, M. A.; Oliveira,
J. S.; Basso, L. A.; dos Santos, D. S.; de Azevedo Junior, W. F.;
Palma, M. S. Dynamics of glyphosate-induced conformational
changes of Mycobacterium tuberculosis 5-enolpyruvylshikimate-
3-phosphate synthase (EC 2.5.1.19) determined by hydro-
gen-deuterium exchange and electrospray mass spectrometry.
Biochemistry 2008, 47, 7509–7522.
(32) Biswas, T. K. Cationic form of beta-galactosidase in the germinat-
ing seeds of Vigna sinensis (Linn) Savi. Arch. Biochem. Biophys.
1986, 251, 379–384.
(33) Biswas, T. K. Beta-galactosidase activity in the germinating seeds
of Vigna sinensis. Phytochemistry 1985, 24, 2831–2833.
(34) Lodge, J. A.; Maier, T.; Liebl, W.; Hoffmann, V.; Strater, N.
Crystal structure of Thermotoga maritima alpha-glucosidase AglA
defines a new clan of NADþ-dependent glycosidases. J. Biol. Chem.
2003, 278, 19151–19158.
(35) Lovering, A. L.; Lee, S. S.; Kim, Y. W.; Withers, S. G.; Strynadka,
N. C. Mechanistic and structural analysis of a family 31 alpha-
glycosidase and its glycosyl-enzyme intermediate. J. Biol. Chem.
2005, 280, 2105–2115.
(36) Greenfield, N.; Davidson, B.; Fasman, G. D. The use of computed
optical rotatory dispersion curves for the evaluation of protein
conformation. Biochemistry 1967, 6, 1630–1637.
(38) Park, H.; Hwang, K. Y.; Oh, K. H.; Kim, Y. H.; Lee, J. Y.; Kim, K.
Discovery of novel alpha-glucosidase inhibitors based on the
virtual screening with the homology-modeled protein structure.
Bioorg. Med. Chem. 2008, 16, 284–292.
(39) Seo, W. D.; Kim, J. H.; Kang, J. E.; Ryu, H. W.; Curtis-Long,
M. J.; Lee, H. S.; Yang, M. S.; Park, K. H. Sulfonamide chalcone as
a new class of R-glucosidase inhibitors. Bioorg. Med. Chem. Lett.
2005, 15, 5514–5516.
(40) Lo Piparo, E.; Scheib, H.; Frei, N.; Williamson, G.; Grigorov, M.;
Chou, C. J. Flavonoids for controlling starch digestion: structural
requirements for inhibiting human alpha-amylase. J. Med. Chem.
2008, 51, 3555–3561.
(41) Shen, Q.; Peng, Q.; Shao, J.; Liu, X.; Huang, Z.; Pu, X.; Ma, L.; Li,
Y. M.; Chan, A. S.; Gu, L. Synthesis and biological evaluation of
functionalized coumarins as acetylcholinesterase inhibitors. Eur. J.
Med. Chem. 2005, 40, 1307–1315.