Structure-based design, synthesis and biological evaluation of β-glucuronidase inhibitors

Article abstract of DOI:10.1007/s10822-014-9745-z

Using structure-based virtual screening approach, a coumarin derivative (1) was identified as β-glucuronidase inhibitor. A focused library of coumarin derivatives was synthesized by eco-benign version of chemical reaction, and all synthetic compounds were

Full text of DOI:10.1007/s10822-014-9745-z

J Comput Aided Mol Des (2014) 28:577–585
DOI 10.1007/s10822-014-9745-z
Structure-based design, synthesis and biological evaluation
of b-glucuronidase inhibitors
•
Khalid M. Khan Nida Ambreen Muhammad Taha Sobia A. Halim
•
•
•
•
•
•
Zaheer-ul-Haq Shagufta Naureen Saima Rasheed Shahnaz Perveen
•
•
Sajjad Ali Mohammad Iqbal Choudhary
Received: 21 November 2013 / Accepted: 15 April 2014 / Published online: 27 April 2014
Ó Springer International Publishing Switzerland 2014
Abstract Using structure-based virtual screening
approach, a coumarin derivative (1) was identified as b-
glucuronidase inhibitor. A focused library of coumarin
derivatives was synthesized by eco-benign version of
chemical reaction, and all synthetic compounds were
characterized by using spectroscopy. These compounds
were found to be inhibitor of b-glucuronidase with IC50
values in a micromolar range. All synthetic compounds
exhibited interesting inhibitory activity against b-glucu-
ronidase, however, their potency varied substantially from
IC50 = 9.9–352.6 lM. Of twenty-one compounds, four
exhibited a better inhibitory profile than the initial hit 1.
Interestingly, compounds 1e, 1k, 1n and 1p exhibited more
potency than the standard inhibitor with IC50 values 34.2,
21.4, 11.7, and 9.9 lM, respectively. We further studied
their dose responses and also checked our results by using
detergent Triton 9-100. We found that our results are true
and not affected by detergent.
Keywords b-Glucuronidase ꢀ Coumarin derivative ꢀ
Docking study
Introduction
The lysosomal acid hydrolase, b-glucuronidase (E.C.
3.2.1.31) has been studied extensively both as model of
enzyme sorting and transport and as an example of hor-
monally and developmentally regulated enzyme [1].
Human b-glucuronidase is reported to synthesize as an
80 kDa monomer (653 amino acids) before proteolysis
removing 18 amino acids from the C-terminal end to form
a 78 kDa monomer [2, 3]. b-glucuronidase exists as a
332 kDa homotetramer [4]. b-Glucuronidase contains
several notable structural formations, including a type of b
barrel known as a jelly roll barrel and a TIM barrel [5].
It catalyzes the stepwise degradation of b-D-glucuronic
acid residues from the non-reducing termini of glycos-
aminoglycans (Hyaluronic acid, heparan sulfate, dermatan
sulfate, and chondroitin srulfate) to yield their respective
Electronic supplementary material The online version of this
article (doi:10.1007/s10822-014-9745-z) contains supplementary
material, which is available to authorized users.
S. A. Halim ꢀ Zaheer-ul-Haq
K. M. Khan (&) ꢀ N. Ambreen ꢀ S. Naureen ꢀ S. Rasheed ꢀ
M. I. Choudhary
H. E. J. Research Institute of Chemistry, International Center for
Chemical and Biological Sciences, University of Karachi,
Karachi 75270, Pakistan
Dr. Panjwani Center for Molecular Medicine and Drug Research,
International Center for Chemical and Biological Sciences,
University of Karachi, Karachi 75270, Pakistan
S. Perveen
PCSIR Laboratories Complex, Karachi, Shahrah-e-Dr.
Salimuzzaman Siddiqui, Karachi 75280, Pakistan
e-mail: khalid.khan@iccs.edu
M. Taha
Atta-ur-Rahman Institute for Natural Product Discovery,
Universiti Teknologi MARA (UiTM), Puncak Alam Campus,
42300 Bandar Puncak Alam, Selangor D. E., Malaysia
S. Ali
Department of Chemistry, Karakoram International University,
Gilgit-Baltistan, Gilgit 15100, Pakistan
M. Taha
Faculty of Applied Science, UiTM, 40450 Shah Alam, Selangor,
Malaysia
123

578
J Comput Aided Mol Des (2014) 28:577–585
glycosidase and hydrolysis results in release of alcohols
and free glucuronic acid. Human b-glucuronidase is
reported to be ubiquitously present in many organs and
body fluids such as macrophages and other blood cells,
liver, spleen, kidney, intestinal secretions, lung, muscle,
bile, urine and serum [6]. To date, the only available three
dimensional structures for b-glucuronidase is the X-ray
crystal structure of human b-glucuronidase, determined by
Jain et al. [5]. Structurally the enzyme is a tetramer of
identical subunits from a single gene product. The native
enzyme, human b-glucuronidase, has a molecular weight
of 332,000 a.m.u., each monomer consisting of a single
chain of 651 amino acid residues with four Asn-linked
glycosylation sites. Each monomer is arranged into three
structural domains, comprising residues 22–223, 224–342
and 343–632. The first domain is a highly distorted barrel-
like structure with a jelly roll motif and two b-hairpin
insertions. The second domain is similar in topology to the
immunoglobulin constant domain, while the C-terminal
domain forms an a/b or TIM-barrel motif. The last portion
or the C-terminal domain is particularly important since it
contains the active site of the enzyme. Amino acids
Glu451, Glu540, and Tyr504 are proposed to be involved
in catalytic process [5]. Enhanced activity of this enzyme
have been reported in a variety of pathological conditions,
including urinary tract infection [7–10], renal diseases [11],
transplant rejection [12], epilepsy [13], and neoplasms of
bladder [14], testes, larynx, and breast [15]. Furthermore,
b-glucuronidase is reported to be released into the synovial
fluid in inflammatory joint diseases such as rheumatoid
arthritis, where these enzymes contribute to the symptoms
of these diseases [16]. The over-expression of the enzyme
also reported in some hepatic diseases and AIDS [17]. b-
Glucuronidase is also found to be involved in the etiology
of colon cancer and high intestinal level of the enzyme
leads to higher incidence of colon carcinoma [18, 19]. This
view has received support from the evidence that the
administration of a bacterial b-glucuronidase inhibitor has
resulted in a decrease in size of carcinogen-induced colonic
tumors [20]. All these reports clearly suggest that the
development of specific inhibitors of b-glucuronidase is of
significant pharmacological importance. Stemming from
our interest in the development of novel, potent b-glucu-
ronidase inhibitors [21], we report herein the effective use
of virtual screening protocol (using a combination of high-
throughput docking with multiple scoring functions) for the
identification of prospective ligands that bind to b-glucu-
ronidase. In particular, recently we have reported an in-
house molecular database, comprises of over 6,000 com-
pounds of synthetic or natural origins [22]. These struc-
turally characterized compounds fulfill the drug-like
criteria according to Lipinski’s rule of five [23], with the
exceptions of a few naturally isolated compounds. This
6,000 compounds database was used for the search of
potential candidates against b-glucuronidase. The libraries
of 6,000 compounds were docked by GOLD docking
program. Furthermore the top ranked docked poses of
GOLD were re-scored by using four different scoring
functions i.e. D-Score, G-Score, ChemScore and PMF
score embedded in CScore module of Sybyl 7.3. Subse-
quently, consensus scoring was implemented and top 5 %
(300 compounds) compounds (ranked by all the scoring
functions) were selected and their binding modes were
inspected visually. Analysis of docking results led us to
choose four compounds (Fig. 2). These compounds were
subjected to in vitro screening against b-glucuronidase
inhibition.
Materials and methods
The 3D coordinates of the compounds were generated by
Molecular Operating Environment [24] (MOE). The
‘‘Washing’’ module of MOE was applied to generate the
physiologically relevant protonation state of compounds.
The structures were minimized by MMFF94 force field and
charges were added. These compounds were only consid-
ered in one tautomeric or protonation state during docking.
Our in-house database [25, 26] comprises over 6,000 3D
structures of small molecules stored in mol format [27].
Docking protocol
˚
The 2.6 A resolution X-ray crystal structure of human b-D-
glucuronidase (PDB code 1BHG, the Protein Data Bank:
http://www.rcsb.org/pdb) was retrieved for docking stud-
ies. The protein structure was checked for missing atoms,
bonds and contacts. Hydrogen was added to enzyme
structure. The B-chain of protein and hetero-atoms
including cofactors were removed from the crystal struc-
ture. All the docking calculations were performed on Xeon
3.0 GHz Linux work station running under SUSE9.2, using
GOLD [28] as the computational software.
GOLD4.0
The configuration file was defined by the following pro-
˚
cess. The docking site was defined as all atoms within 10 A
of a specified central point (x, y, z coordinates: 80.43,
84.41, 90.48). For each independent GA run, a maximum
number of 1,000 GA operations was performed on a single
population of 50 individuals. Operator weights for cross
over, mutation and migration were set to 100, 100 and 0
respectively. To allow poor non bonded contacts at the start
of each GA run, the maximum distance between hydrogen
123

J Comput Aided Mol Des (2014) 28:577–585
579
˚
donors and fitting points were set to 5.0 A and non-bonded
˚
VdW energies were cut-off at 10 A.
compounds in the assay described below. The IC50 values
were calculated from inhibition data by using EZ-Fit
Enzyme Kinetic Program (Perrella Scientific Inc., Amherst,
USA).
To build a protein model whose active site exists in the
probable substrate-bound conformation, the known sub-
strate p-nitrophenyl b-glucuronide was first docked into the
active site of b-glucuronidase using the GOLD. From the
docking results, a high-scoring docked pose of p-nitro-
phenyl b-glucuronide having the glucuronyl moiety ori-
ented properly toward the key residues Glu451 and Glu540
involved in the enzyme catalysis reaction was chosen to
represent its most probable binding mode. The obtained
docking model was further subjected to energy minimiza-
tion for refinement. Prior to energy minimization, the
Glu451 residue in the active site was assigned a protonated
form that was consistent with its role in acting as a proton
donor during catalysis. In addition, a structural subset was
specified as the ligand and the amino acid residues within a
b-Glucuronidase assay protocol
b-Glucuronidase activity was determined by measuring the
absorbance at 405 nm of p-nitrophenol formed from the
substrate by the spectrophotometeric method. The total
reaction volume was 250 lL. The reaction mixture con-
tained 185 lL of 0.1 M acetate buffer, 5 lL of test com-
pound solution, 10 lL of enzyme solution was incubated at
°
37 C for 30 min. The plates were read on a multiplate
reader (SpectraMax plus 384) at 405 nm after the addition
of 50 lL of 0.4 mM p-nitrophenyl-b-D-glucuronide [29].
All assays were run in triplicate. Furthermore to avoid
precipitation, the concentration of the compounds was
decreased and the reaction volume was high (200 lL) so
the chance of precipitation was less hence the addition of
detergents was not needed.
˚
10 A radius of the ligand. The complex structure was then
refined through a two-step minimization, allowing (1) only
the ligand and the side chains of the subset to relax and
then (ii) only all the atoms of the subset to relax. The
energy minimization calculations were performed in
SYBYL 7.1 using the Tripos force field, the Powell
method, a distance-dependent dielectric constant of 1r, and
Assay for antiglycation
˚
a non-bonded cutoff of 8 A. The minimization was ter-
Bovine serum albumin (BSA) was purchased from
Research Organics (Cleveland, USA), while others chem-
icals {glucose anhydrous, trichloroacetic acid (TCA)
sodium azide (NaN3), dimethyl sulfoxide (DMSO), sodium
dihydrogen phosphate (NaH2PO4), sodium chloride
(NaCl), disodium hydrogen phosphate (Na2HPO4), potas-
sium chloride (KCl), potassium dihydrogen phosphate
(KH2PO4), and sodium hydroxide (NaOH) were purchased
from the Sigma-Aldrich. Sodium phosphate buffer (pH
7.4), was prepared by mixing Na2HPO4 and NaH2PO4
(67 mM) containing sodium azide (3 mM). Phosphate
buffer saline (PBS) was prepared by mixing NaCl
(137 mM), Na2HPO4 (8.1 mM), KCl (2.68 mM), and
KH2PO4 (1.47 mM) and pH 10 was adjusted with NaOH
(0.25 mM). BSA (10 mg/mL) and glucoses anhydrous
(50 mg/mL) solutions were prepared in sodium phosphate
buffer. Test samples were prepared in DMSO (1 mM/mL).
In 96-well plate assays, each well contained 60 lL
reaction mixtures (20 lL BSA (10 mg/mL), 20 lL of
glucose anhydrous (50 mg/ml) and 20 lL test sample).
Glycated control contain 20 lL BSA, 20 lL glucose and
20 lL sodium phosphate buffer, while blank control con-
tains 20 lL BSA and 40 lL sodium phosphate buffer [30].
Reaction mixture was incubated at 37 °C for 7 days. After
incubation, 6 lL TCA 100 % was added in each well and
minated when the energy gradient convergence criterion of
˚
0.01 kcal/(mol.A) was reached.
The modeled complex structure of b-glucuronidase with
the substrate p-nitrophenyl b-glucuronide was used for the
prediction of a favorable binding mode of database com-
pounds to b-glucuronidase by docking. The top ranked
GOLD docked poses of all the compounds were further re-
scored using the CScore module implemented in SYB-
YL7.2 to estimate the binding affinity of the compounds.
To support the virtual screening results dose dependent
graphs are also given here. The O.D of compounds and
their respective % inhibition at different concentration are
given in dose–response curves.
2D similarity searching
MOE [24] was employed for similarity searching. We
selected Molecular ACCess System (MACCS) structural
keys [24] (166 bit positions) with the general form of
Tanimoto coefficient [24]. MACCS consists of 166 struc-
tural substructures/patterns with 1–10 non-hydrogen atoms.
As a default parameter, 85 % of overlap was selected.
Bioassay techniques with determination of IC50 values
°
centrifuged (15,000 rpm) for 4 min at 4 C. After centri-
The concentration of test compounds that inhibited the
enzyme activity by 50 % (IC50 values) were determined by
monitoring the effect of increasing concentrations of
fugation, the pellets were rewashed with 60 lL of TCA
(5 %). The supernatant containing glucose, standard
inhibitor and interfering substance, was removed and pellet
123

580
J Comput Aided Mol Des (2014) 28:577–585
containing, AGE-BSA were dissolved in 60 lL of PBS.
Fluorescence spectrum (ex. 370 nm), and change in fluo-
rescence intensity (370–440 nm), based on AGEs forma-
tion, were monitored by using Spectrofluorimeter RF-1500
(Shimadzu, Japan).
Table 1 Inhibition of b-glucu-
ronidase by compounds 1–5,
calculated as (IC50 SEM)
Compound ID (IC50 SEM)
1
2
3
4
113.6 4.00
172 00
194 1.88
266 12.8
48.4 1.25
% inhibition was calculated through the following
formula:
d-Saccharic
acida
SEM standard error of the mean
a
standard drug
% Inhibition ¼ 1 ꢁ ½ðFluorescence of test sampleÞ=
ðFluorescence of glycatedÞꢂ ꢃ 100
Phosphodiesterase I inhibition assay
Additionally, the CScore module of Sybyl7.3 [25] were
utilized to re-score the top scored docked poses obtained by
GOLD, by using four scoring functions (D-score, G-score,
ChemScore, and PMF-score).
Activity against snake venom was assayed by taking
33 mM Tris-HC1 buffer pH 8.8, 30 mM Mgacetate with
0.000742 U/well final concentration of enzyme using a
microtiter plate assay and 0.33 mM bis-(p-nitropheny1)
phosphate (Sigma N-3002) as substrate. From Merck Cy-
stein and EDTA were used as positive controls
(1C50 = 748 0.015, 274 0.007 lM, respectively).
After 30 min pre-incubation of the enzyme with the test
samples, enzyme activity was monitored spectrophoto-
The activity of target protein was determined as
a
described in supporting information. D-Saccharic acid
(standard inhibitor for b-glucuronidase inhibitory activity).
The ranks and scores of compounds 1–4, obtained by all
the above mentioned scoring functions are reported in
supporting information (Table 1).
In order to evaluate the efficacy of scoring functions to
identify the known inhibitors from the random database,
five known inhibitors (Fig. 2) [37–41] were included in
6,000 compounds. All the above mentioned scoring func-
tions were able to identify the known inhibitors in the top
5 % of the screened database. After the docking/scoring
protocol, top 5 % of the ranked database (predicted by all
the above mentioned scoring functions) was selected and
the docked poses were inspected visually. Analysis of
docking results led us to choose four compounds (Fig. 2).
These compounds were subjected to in vitro screening
against b-glucuronidase inhibition. All the selected com-
pounds showed biological activities in micro-molar range.
The IC50 value of compounds 1–4 ranged between
113.6 4.00 to 266 12.8 lM (Table 1).
°
metrically at 37 C on a microtitre plate reader (Spectra-
Max, Molecular Devices) by following the rate (change in
OD/min) of release of p-nitrophenol from p-nitrophenyl
phosphate at 410 nm. All assays were conducted in tripli-
cate [31–34].
Results and discussion
Structure based virtual screening experiment was per-
formed targeting the active site amino acid residues
(Glu451, Glu540, and Tyr504) of b-glucuronidase. The
crystal structure of human b-glucuronidase was retrieved
from the Protein Data Bank (PDB code: 1BHG). To scru-
tinize the performance of our molecular modeling process,
the chromogenic derivative of known physiological sub-
strate p-nitro phenyl b-glucuronide [35] was first docked
into the active site of b-glucuronidase using GOLD [28].
The purpose of such initial modeling was to build a protein
model whose active site exists in the probable substrate-
bound conformation. The modeled substrate-bound struc-
ture of human b-glucuronidase showed that the glycosidic
bond of p-nitrophenyl D-glucuronide was properly oriented
towards the catalytic residues Glu451 and Glu540 (Figure
S1 in supporting information). It has been proposed for
human b-glucuronidase that during the catalysis, Glu451
acts as an acid/base catalyst while Glu540 serve as a
nucleophilic residue [36]. This modeled protein structure
was then used in the prediction of a favorable binding
mode. Ranking of database compounds was achieved via
GOLD docking program. GOLD was used with two scor-
ing functions GOLD-Score [21] and ASP. [22].
In addition 2D similarity searches were conducted by
using the Tanimoto coefficient with MACCS keys (166
bits) as implemented in MOE [24]. The chemical structures
of
5,7,8,30,40,50-hexahydroxyflavone-O-b-D-glucuronide
and Scopoletin were used as queries (Figure S2). Com-
pounds 1–4 were returned with the best Tanimoto values
ranging from 0.66 to 0.88. 2D Similarity profile is sum-
marized in Table-1 in supplementary information. Detailed
information about virtual screening strategy and the flow
chart of high throughput docking (Figure S3) is described
in the Supporting Information.
The binding mode of compound 1, obtained by GOLD,
is presented in Fig. 1a, while the binding mode of com-
pounds 2–4 is shown in supporting information (Figure
S4). Molecular docking investigations have clearly dem-
onstrated that all the compounds displayed a good steric
and chemical complementarities with the ligand binding
site. Compound 1 was found to be completely surrounded
123

J Comput Aided Mol Des (2014) 28:577–585
581
which contribute to stabilization of compound 1 in com-
plex with b-glucuronidase.
The binding mode analysis of compound 2 (Figure S4a
in the Supporting Information) revealed that the com-
pound is stabilized by hydrogen bonds with important
active site amino acid residues. The hydrogen bonding
was observed between the carbonyl oxygen of the com-
pound 2 and the binding site residues Arg600 and
Lys606. The carbonyl oxygen of 2 accepts hydrogen
bonds with the side chain side chain guanidinium moiety
of Arg600 and the side chain amino group of Lys606 with
˚
the distance of 2.29 and 1.79 A, respectively. The phenyl
ring of the compound 2 is stabilized by the p–p interac-
tions provided by the side chain phenol ring of Tyr 504
and Tyr508 (Fig. 2).
Similar to the binding mode of compound 2, the car-
bonyl moiety of compound 3 also mediates hydrogen
bonding interactions with the side chains of Arg600 and
Lys606. The observed hydrogen bond distance between the
˚
side chain guanidinium (Ne2) of Arg600 was 2.84 A, while
the hydrogen bond distance between the carbonyl moiety
˚
and the side chain amino nitrogen (Nf) was 1.41 A. In
addition, compound 3 also mediates hydrogen bonding
˚
with side chain of Asp207 (1.85 A).
The predicted docked pose of compound 4 revealed that
it is hydrogen bonded to the imidazole ring of His385. The
Ne2 of imidazole ring of His385 served as hydrogen bond
acceptor and accepted the phenolic hydroxyl hydrogen of
˚
compound 4 with the distance of 1.94 A. Furthermore the
Fig. 1 Binding mode analysis of compound 1 (a) and compound 1p
(b) in the catalytic binding site of b-glucuronidase. Hydrogen bonds
are indicated as dashed lines
phenol group of Tyr508 provides p-p interaction to the
benzene of chromen moiety of compound 4. The predicted
binding mode of compound 4 is presented in Figure S4b in
the Supporting Information.
by the catalytic amino acid residues in the active site of the
enzyme and mediating strong hydrogen bond with OE2 of
˚
Glu451 (2.73 A). The carbonyl oxygen of the compound 1
created a hydrogen bond with hydroxyl group of Tyr504
˚
(2.45 A) and the side chain amino nitrogen of Lys606
˚
(3.06 A). Additionally, the benzopyrone moiety of com-
Subsequently, we decided to scrutinize the inhibitory
potential of these compounds against other available
enzymes. Compound 1 was randomly screened against
phosphodiesterase inhibition and for anti-glycation poten-
tial. Compounds 2 and 4 were tested for urease inhibitory
activity (screening protocols are described in the Support-
ing Information). Compound 1 found to be inactive against
urease and phosphodiesterase. Similarly, no anti-glycation
activity was observed. However, interestingly compounds
2 and 4 showed anti-urease activities in micromolar range.
pound 1 depicted p–p interaction with the side chain
phenol ring of Tyr504. In addition, several hydrophobic
interactions with Tyr205, Asp207, Asn484, Asn450,
Asn502, Tyr504, Tyr508, Glu540, Ag600 and Lys606
H
O
O
O
H
O
H
O
O
O
O
O
OH
O
O
OH
O
4
1
2
3
Fig. 2 The structures of compounds 1-4, identified by virtual screening
123

582
J Comput Aided Mol Des (2014) 28:577–585
Scheme 1 b-Glucuronidase
inhibitory activities of coumarin
diones (1a–1u). SEM standard
error of the mean, N/D not
determined
According to this preliminary study, compound 1 was
identified as a specific inhibitor of b-glucuronidase; Results
are summarized in scheme.
These preliminary studies suggested that compound 1
possess a significant inhibitory activity and can serve as
useful starting point for further studies. We therefore
123

J Comput Aided Mol Des (2014) 28:577–585
583
Scheme 1 continued
decided to make a small library of derivatives of compound
1. We assume that different derivatives of compound 1 (4-
hydroxy coumarin) can further enhance/improve the
potency and selectivity of the ligand against b-glucuroni-
dase. In particular, coumarins are a large class of com-
pounds, which are reported to have a large number of
biological and therapeutically important properties,
including antiviral, antifungal, anticoagulant, antithrom-
botic, tuberculostatic, antioxidant and antitumor activities
[29]. Due to the physiological importance of coumarins, we
synthesized a small focused library of compound 1 deriv-
ative, with the aim of increasing the potency and selectivity
against the b-glucuronidase enzyme.
of organic reaction. Eco-benign versions of organic reac-
tions are potential candidates for the synthesis of biologi-
cally active compounds because of low cost, non-
inflammability and non toxicity [22]. However, so far as
we know, there are no reports concerning the use of eco-
benign synthesis of coumarin analogues. Scheme 1
The synthesis was achieved by a single step chemical
reaction. A mixture of aryl aldehyde (1.0 g, 10 mmol) and
4-hydroxy coumarin (1.6 g, 10 mmol), in water (10 mL)
was stirred at 100 °C. Heating and stirring was continued
for 1–2 h at 100 °C. The reaction was monitored by TLC.
After the completion of reaction, the solid was filtered and
washed subsequently with boiling water and hexane. After
drying in vacuum, the coumarin diones were obtained as
solid products.
Several methods are reported for the synthesis of cou-
marin analogues [42–45]. In these reactions, different types
of chemicals are used that can be hazardous for environ-
ment. To reduce the use and generation of hazardous
chemicals, we have utilized environmental friendly version
The inhibitory potential of the coumarin derivatives
were assessed against b-glucuronidase activity by mea-
suring the absorbance at 405 nm of p-nitrophenol released
123

584
J Comput Aided Mol Des (2014) 28:577–585
from the substrate by the spectrophotometeric method. The
total reaction volume was 250 lL. The reaction mixture
contained 185 lL of 0.1 M acetate buffer, 5 lL of test
compound solution and 10 lL of enzyme solution which
features for future designing of the selective b-glucuroni-
dase inhibitors.
In order to elucidate the b-glucuronidase binding
mechanism and orientation of coumarin derivatives in
more detail, molecular dynamic simulation study is in
process and the results will be published later.
°
was incubated at 37 C for 30 min. The plates were read on
a multiplate reader (SpectraMax plus 384) at 405 nm after
the addition of 50 lL of 0.4 mM p-nitrophenyl-b-D-glu-
curonide (26). All assays were run in triplicate.
All synthetic compounds exhibited interesting inhibi-
tory activity against b-glucuronidase, however, their
potency varied substantially from IC50 = 9.9–352.6 lM.
Of the twenty-one compounds, four exhibited a better
inhibitory profile than the initial hit 1. Interestingly,
compounds 1e, 1 k, 1n and 1p exhibited more potency
than the standard inhibitor with IC50 values 34.2, 21.4,
11.7, and 9.9 lM, respectively. The results are summa-
rized in Scheme 1. We repeat our experiment with Triton
9-100 as detergent as procedure described in [46] and
we got same results.
Conclusion
In summary, we have employed a computer-aided drug
design strategy, involving structure based virtual screening
and in vitro enzyme assay in a consecutive manner to
identify potent b-glucuronidase inhibitors. Compound 1
was identified as a starting point for the development of
more potent and selective inhibitors. Subsequently, we
synthesized and tested a focused library of coumarin
derivatives based on the structure of 1 which exhibited a
significant potency with IC50 values ranging between 9.9
and 352.6 lM. The proposed scaffold of b-glucuronidase
inhibitors offers the possibility of convenient further
modifications that could give rise to structures with
improved inhibitory activity and selectivity toward the
enzyme.
The binding mode analysis revealed that the coumarin
moiety of all these derivatives oriented similar to com-
pound 1, in the binding site of b-glucuronidase. Com-
pounds 1p, the most potent analogue of the series with an
IC50 of 9.9 lM, showed that electron rich five-membered
heterocyclic ring substitutions is the primary determinant
of the inhibitory activity. The docking analysis revealed the
electron rich heterocyclic ring interact with the electro-
positive side chain of His385. Furthermore, methyl sub-
stituent of 1p exhibited additional Van der Waals
interactions with His385, and thus contribute in the sta-
bilization of the ligand. Additionally, the side chain indole
ring of Trp587 is involved in p–p stacking with the furan
moiety of compound 1p. Moreover, the first carbonyl
oxygen of compound 1p form a hydrogen bond with the
Acknowledgments The authors are thankful to Higher Education
Commission (HEC) Pakistan for providing financial support under
‘‘National Research Program for Universities’’ to Project No.
20-1394.
References
1. Paigen
K (1989) Mammalian beta-glucuronidase: genetics,
molecular biology, and cell biology. Prog Nucl Acid Res Mol
Biol 37:155–205
˚
side chain hydroxyl hydrogen of Tyr504 (3.01 A), while
another carbonyl oxygen of compound 1p is involved in
hydrogen bonding with the side chain hydrogen of Glu451
˚
(2.03 A). The hydrogen bond distance between the ring
2. Islam MR, Grubb JH, Sly WS (1993) C-terminal processing of
human beta glucuronidase. The propeptide is required for full
expression of catalytic activity, intracellular retention, and proper
phosphorylation. J Biol Chem 268:22627–22633
3. Shipley JM, Grubb JH, Sly WS (1993) The role of glycosylation
and phosphorylation in the expression of active human b-glucu-
ronidase. J Biol Chem 268:12193–12198
4. Kim HW, Mino K, Ishikawa K (2008) Crystallization and pre-
liminary X-ray analysis of endoglucanase from Pyrococcus hor-
ikoshii’’. Acta Crystallogr Sect F Struct Biol Cryst Commun
64:1169–1171
5. Jain S, Drendel WB, Chen ZW, Mathews FS, Sly WS, Grubb JH
(1996) Structure of human beta-glucuronidase reveals candidate
lysosomal targeting and active-site motif. Nat Struct Biol
3:375–381
oxygen of the coumarin moiety and the side chain amidic
˚
nitrogen of Asn484 was found to be 2.29 A. While the
hydrogen bond distance between the side chain hydroxyl of
˚
Tyr405 and the furan oxygen was about 2.45 A. As shown
in Fig. 1b, compound 1p is also stabilized by several
hydrophobic interactions with Asp207, Asn502, Tyr508,
His509, Ile517, Ile521, Trp587, Arg600, and Lys606 in the
active site of b-glucuronidase. Therefore, it is observed in a
detailed binding mode analysis that the binding of the
inhibitors in the active site of b-glucuronidase can be
facilitated by the introduction of multiple hydrogen bonds
with the side chain and backbone groups. Furthermore, the
hydrophobic interactions with the residues near the active
site can play a significant role in stabilizing the inhibitors
in the active site of b-glucuronidase. This may serve as key
6. Sperker B, Mu¨rdter TE, Schick M, Eckhardt K, Bosslet K, Kro-
emer HK (1997) Interindividual variability in expression and
activity of human beta-glucuronidase in liver and kidney: con-
sequences for drug metabolism.
281:914–920
J Pharmacol Exp Ther
7. Bank N, Bailine SH (1965) Urinary beta-glucuronidase activity in
patients with urinary-tract infection. N Engl J Med 272:70–75
123

J Comput Aided Mol Des (2014) 28:577–585
585
8. Roberts AP, Frampton J, Karim SM, Beard RW (1967) Estima-
tion of beta-glucuronidase activity in urinary-tract infection.
N Engl J Med 276:1468–1470
31. Mamillapalli R, Haimovitz R, Ohad M, Shinitzky M (1998)
Enhancement and inhibition of snake venom phosphodiesterase
activity by lysophospholipids. FEBS Lett 436:256–258
´
32. Lopez-Gomez J, Costas MJ, Meireles RJ, Fernandez A, Romero
A, Avalos M, Cameselle JC (1998) Glycine-enhanced inhibition
´
´
9. Ronald AR, Silverblatt F, Clark H, Cutler RE, Turck M (1971)
Failure of urinary beta-glucuronidase activity to localize the site
of urinary tract infection. Appl Microbiol 21:990–992
10. Kallet HA, Lapco L (1967) Urine beta glucuronidase activity in
urinary tract disease. J Urol 97:352–356
11. Gonick HC, Kramer HJ, Schapiro AE (1973) Urinary b-glucu-
ronidase activity in renal disease. Arch Intern Med 132:63–69
12. Schapiro A, Paul W, Gonick H (1968) Urinary beta-glucuroni-
dase in urologic diseases of the kidneys. J Urol 100:146–157
13. Plum CM (1967) Beta-glucuronidase activity in serum, cerebro-
spinal fluid and urine in normal subjects and in neurological and
mental patients. Enzymol Biol Clin 8:97–112
´
of rat liver nucleotide pyrophosphatase/phosphodiesterase-I by
EDTA: a full account of the reported inhibition by commercial
preparations of acidic fibroblast growth factor (FGF-1). FEBS
Lett 421:77–79
33. Razzell WE, Khorana HG (1959) Studies on polynucleotides III.
Enzymic degradation. Substrate specificity and properties of
snake venom phosphodiesterase. J Biol Chem 234:2105–2113
34. Nakabayashi T, Matsuoka Y, Ikezawa H, Kimura Y (1994)
Alkaline phosphodiesterase I release from eucaryotic plasma
membranes by phosphatidylinositol-specific phospholipase C–IV.
The release from Cavia porcellus organs. Int J Biochem
26:171–179
35. Kato K, Yoshida K, Tsukamoto H, Nobunaga M, Masuya T,
Sawada T (1960) Synthesis of p-nitro phenyl beta-D-glucopyr-
anosiduronic acid and its utilisation as substrate for assay of beta-
glucuronidase activity. Chem Pharm Bull 8:239–242
36. Islam MR, Tomatsu S, Shah GN, Grubb JH, Jain S, Sly WS
(1999) Active site residues of human beta-glucuronidase. Evi-
dence for Glu(540) as the nucleophile and Glu(451) as the acid-
base residue. J Biol Chem 274:23451–23455
37. Kawasaki M, Hayashi T, Arisawa M, Morita N, Berganza L
(1988) H.8-hydroxytricetin 7-glucuronide. A b-glucuronidase
inhibitor from Scoparia dulcis. Photochemistry 27:3709–3711
38. Schmeda-Hirschmann G, Loyola JI, Reyes S, Hubert E, Rodri-
guez M, Rodriguez J, Dutra-Behrens M (1994) b-Glucuronidase
inhibition and diuretic activity of Fabiana imbricata R. & P.
(Solanaceae). Phytother Res 8:485–487
39. Hayashi T, Kawasaki M, Okamura K, Tamada Y, Morita N,
Tezuka Y, Kikuchi T, Miwa Y, Taga T (1992) Scoparic acid A, A
b-glucuronidase inhibitor from Scoparia dulcis. J Nat Prod
55:1748–1755
´
´
14. Hradec E, Petrık R, Pezlarova J (1965) The activity of beta-
glucuronidase in cases of bladder neoplasms. J Urol 94:430–435
15. Boyland E, Gasson JE, Williams DC (1957) Enzyme activity in
relation to cancer; the urinary beta-glucuronidase activity of
patients suffering from malignant disease. Br J Cancer 11:120–129
16. Caygill JC, Pitkeathly DA (1966) BTS/BLF/BALR early career
investigator of the year: a study of b-acetylglucosaminase and
acid phosphatase in pathological joint fluids. Ann Rheum Dis
25:137–144
17. Weissmann G, Zurier RB, Spieler PJ, Goldstein IM (1971)
Mechanisms of lysosomal enzyme release from leukocytes
exposed to immune complexes and other particles. J Exp Med
134:149S–165S
18. Reddy BS (1976) Dietary factors and cancer of the large bowel.
Semin Oncol 3:351–359
19. Goldin BR, Gorbach SL (1976) The relationship between diet and
rat fecal bacterial enzymes implicated in colon cancer. J Nat
Cancer Inst 57:371–375
20. Takada H, Hirooka T, Hiramatsu Y, Yamamoto M (1982) Effect
of b-glucuronidase inhibitor on azoxymethane-induced colonic
carcinogenesis in rats1. Cancer Res 42:331–334
21. Khan KM, Shujaat S, Rahat S, Hayat S, Choudhary MI (2002)
Beta-N-cyanoethyl acyl hydrazide derivatives: a new class of
beta-glucuronidase inhibitors. Chem Pharm Bull 50:1443–1446
22. Khan KM, Rahim F, Wadood A, Taha M, Khan M, Naureen S,
Ambreen N, Hussain S, Perveen S, Choudhary MI (2014) Eval-
uation of bisindole as potent b-glucuronidase inhibitors: synthesis
and in silico based studies. Bioorg Med Chem Lett 24:1825–1829
23. Lipinski CA, Lombardo F, Dominy BW, Feeney PJ (1997)
Experimental and computational approaches to estimate solubil-
ity and permeability in drug discovery and development settings.
Adv Drug Del Rev 23:3–25
40. Cenci di Bello I, Dorling P, Fellows L, Winchester B (1984)
Specific inhibition of human b-D-glucuronidase and a-L-idu-
ronidase by a trihydroxy pipecolic acid of plant origin. FEBS Lett
176:61–64
41. Levy G, Conchi J (1966) b-Glucuronidase and the hydrolysis of
glucuronides. In: Dutton GJ (ed) Glucuronic Acid. Academic
Press, Free and Combined New York, pp 301–364
42. Borges F, Roleira F, Milhazes N, Santana L, Uriarte E (2005)
Simple coumarins and analogues in medicinal chemistry: occur-
rence, synthesis and biological activity. Curr Med Chem
12:887–916
24. MOE (The Molecular Operating Environment), Version 2006.08
(2007) software, Chemical Computing Group Inc., Montreal,
Canada.http://www.chemcomp.com
25. SYBYL Molecular Modeling Software Packages, Version 6.9
(2004) Tripos Associates, St Louis, MO
26. http://www.iccs.edu/pcmd/compound_bank.php
27. Abid O, Babar TM, Ali FI, Ahmed S, Wadood A, Rama NH,
Uddin R, Ul-Haq Zaheer, Khan A, Choudhary MI (2010) Iden-
tification of novel urease inhibitors by high-throughput virtual
and in vitro screening. Med Chem Lett 1:145–149
28. GOLD Version 4.0. (2001) Astex Technology, Cambridge
29. Collins RA, Ng TB, Fong WP, Wan CC, Yeung HW (1997)
Inhibition of glycohydrolase enzymes by aqueous extracts of
Chinese medicinal herbs in a microplate format. Biochem Mol
Biol Int 42:1163–1169
43. Nicolaides DN, Fylaktakidou KC, Litinas KE, Hadjipavlou-Litina
D (1996) Synthesis and biological evaluation of some 4-(isox-
azolinyl or 1,2,4-oxadiazolyl) coumarins. J Heterocycl Chem
33:967–971
44. Burgart YV, Shcherbakov KV, Saloutin VI, Chupakhin ON
(2004) 5,6,7,8-Tetrafluoro-4-hydroxycoumarin derivatives in
reactions with o-phenylenediamin. Russ Chem Bul 53:1237–1239
45. Lu J, Li Y, Bai Y, Tian M (2004) Green chemistry approaches to
the synthesis of 5-arylidenethiobarbituric acids by a condensation
reactions between aromatic aldehydes and thiobarbituric acid:
comparison of water, microwave irradiation, and grinding. Het-
erocycles 63:583–589
46. Ferreira RS, Bryant C, Ang KKH, McKerrow JH, Shoichet BK,
Renslo AR (2009) Divergent modes of enzyme inhibition in a
homologous structure–activity series.
52:5005–5008
J
Med Chem lett
30. Yamaguchi F, Ariga T, Yoshimura Y, Nakazawa H (2000)
Antioxidative and anti-glycation activity of garcinol from
Garcinia indica fruit rind. J Agric Food Chem 48:180–185
123