Experimental validation and docking studies of flavone derivatives on aldose reductase involved in diabetic retinopathy, neuropathy, and nephropathy

Article abstract of DOI:10.1007/s00044-010-9412-4

The enzyme aldoreductase which plays an important role in pathogenesis of diabetic retinopathy, neuropathy, and nephropathy was purified from bovine lens, and its inhibitory activity was studied with the synthesized flavone derivatives 1-(2-hydroxyphenyl)ethanone as the starting material. Experimental study revealed that 2-chloroflavone shows less inhibitory activity of 60-70% than other flavones used in the study. To validate experimental results computationally, docking studies of new flavone derivatives synthesized were performed with the enzyme aldose reductase, and the results indicate that 3-iodo, 4-methyl, 5-chloroflavone and 2-chloroflavone bind with higher and lesser affinities. Docking studies with site directed mutagenesis of Val47Ile, Tyr48His, Pro121Phe, Trp219Tyr, Cys298Ala, Leu300Pro, Ser302Arg, and Cys303Asp of the enzyme altered the inhibition activity of aldose reductase. The regression value (R²) of 0.81 between the docking scores of the known inhibitors and the experimental logIC50 indicates the reliability of the docking studies. Biological activity and carcinogenic properties predict that 3-iodo, 4-methyl, 5-chloroflavone is the best flavone inhibitor against aldose reductase. Springer Science+Business Media, LLC 2010.

Full text of DOI:10.1007/s00044-010-9412-4

MEDICINAL
CHEMISTRY
RESEARCH
Med Chem Res (2011) 20:930–945
DOI 10.1007/s00044-010-9412-4
ORIGINAL RESEARCH
Experimental validation and docking studies of flavone derivatives
on aldose reductase involved in diabetic retinopathy, neuropathy,
and nephropathy
•
Pagadala Nataraj Sekhar P. B. Kavi Kishor P. K. Zubaidha
•
•
•
•
•
A. M. Hashmi T. A. Kadam Lakkireddy Anandareddy Marc De Maeyer
•
•
K. Praveen Kumar B. Vijaya Bhaskar T. Munichandrababu
•
•
•
G. Jayasree P. V. B. S. Narayana G. Gyananath
•
Received: 24 December 2009 / Accepted: 16 August 2010 / Published online: 16 September 2010
Ó Springer Science+Business Media, LLC 2010
Abstract The enzyme aldoreductase which plays an impor-
tant role in pathogenesis of diabetic retinopathy, neuropathy,
and nephropathy was purified from bovine lens, and its
inhibitory activity was studied with the synthesized flavone
derivatives 1-(2-hydroxyphenyl)ethanone as the starting
material. Experimental study revealed that 2-chloroflavone
shows less inhibitory activity of 60–70% than other flavones
used in the study. To validate experimental results compu-
tationally, docking studies of new flavone derivatives syn-
thesized were performed with the enzyme aldose reductase,
and the results indicate that 3-iodo, 4-methyl, 5-chloroflavone
and 2-chloroflavone bind with higher and lesser affinities.
Docking studies with site directed mutagenesis of Val47Ile,
Tyr48His, Pro121Phe, Trp219Tyr, Cys298Ala, Leu300Pro,
Ser302Arg, and Cys303Asp of the enzyme altered the inhi-
bition activity of aldose reductase. The regression value (R2)
of 0.81 between the docking scores of the known inhibitors
and the experimental logIC50 indicates the reliability of the
docking studies. Biological activity and carcinogenic prop-
erties predict that 3-iodo, 4-methyl, 5-chloroflavone is the
best flavone inhibitor against aldose reductase.
Keywords Aldose reductase ꢀ Flavones ꢀ Docking
Introduction
P. Nataraj Sekhar (&) ꢀ P. B. Kavi Kishor ꢀ L. Anandareddy
Department of Genetics, Osmania University,
Hyderabad 500 007, India
Aldose reductase is believed to be of primary importance in
the development of severe degenerative complications of
diabetes mellitus, through its ability to reduce excess D-glu-
cose into D-sorbital in non-insulin dependent tissues (Larson
etal.,1988).Aldosereductasehasbeenshowntoactonawide
range of substrates in vitro, such as aldehydes, aldoses, and
corticosteroids, but it is most effective on steroid hormones
(Wermuth and Monder, 1983), as confirmed by several
reports (Warren et al., 1993). This NADPH-dependent
enzyme reduces carbonyl oxygen to a hydroxyl in an ordered
‘bi–bi’ mechanism, in which NADPH is bound first and
NADP? is released last. Like all other NAD(P)?-dependent
enzymaticreactionsdescribed to date, this reduction isstereo-
specific with respect to the coenzyme. Upon formation of the
enzyme-coenzyme-substrate tertiary complex, a hydride is
transferred from C4 carbon atom of the nicotinamide ring
(hydrogen 4-pro-R of the A-face of the ring) of NADPH to the
carbonyl carbon of the substrate while a proton is provided by
the enzyme to the carbonyl oxygen (Hoffman et al., 1980). In
the presence of NADPH, the enzyme converts glucose to
e-mail: nattu251@rediffmail.com
P. K. Zubaidha ꢀ A. M. Hashmi
School of Chemical Sciences, S. R. T. M. University,
Nanded 431 606, India
T. A. Kadam ꢀ G. Gyananath
School of Life Sciences, S. R. T. M. University,
Nanded 431 606, India
M. De Maeyer
Laboratory of Biomolecular Modelling, Division Biochemistry,
Molecular and Structural Biology, Department of Chemistry,
Katholieke University, Leuven, Belgium
K. Praveen Kumar ꢀ B. Vijaya Bhaskar ꢀ T. Munichandrababu
Srivenkateswara University, Tirupathi 517 501, India
G. Jayasree
Srinidhi Institute of Science and Technology,
Hyderabad 500 007, India
P. V. B. S. Narayana
CARISM, SASTRA University, Thanjavur, India
123

Med Chem Res (2011) 20:930–945
931
sorbital, which is only slowly metabolized to fructose by
sorbital dehydrogenase, the other enzyme in the pathway,
with concurrent reduction of NAD?. It has been shown that
the oxidation of sorbitol catalyzed by sorbitol dehydrogenase
increasestheratioof NADH:NAD?, resulting in anincreased
lactate: pyruvate ratio and pseudohypoxia (Williamson et al.,
1993). Enhanced flux of glucose via polyol pathway under
hyperglycemic conditions may play an important role in the
pathogenesis of diabetic retinopathy, neuropathy, and
nephropathy (Bhatnagar and Srivastava, 1992). Thus, AR
represents an attractive pharmacological target to interrupt
the polyol pathway and hence delay in the onset/severity of
these late diabetic complications (Yabe-Nishimura, 1998). A
number of aldose reductase inhibitors (ARIs) were developed
and tested in patients for the treatment of such complications
(Gabbay et al., 1979, Judzewitsch et al., 1983, Dvornik, 1987,
Masson and Boulton, 1990). Presently, none of these drugs
are therapeutically effective due to their side effects, non-
specificity, and inhibition of other members of the aldo–keto
reductase family (Martyn et al., 1987; O’Brien et al., 1982;
Poulsom, 1986). Though a broad variety of aldose reductase
inhibitors (ARIs) were synthetically developed from lead
compounds obtained from in vitro screening studies, very few
compounds have qualified for clinical trials (Narayanan,
1993, Nakai et al., 1985). Crystallographic, modeling, and
structure–function studies of AR have yielded various classes
of ARIs with potent inhibitory activity. The drug tolrestat
which was launched in 1989 was withdrawn in 1996, prin-
cipally due to its low efficacy. Of the newer compounds,
zopolrestat and zenarestat were also withdrawn from clinical
trials (Mylari et al., 1991). At present, only epalrestat which
was developed by Ono and launched into the Japanese market
in 1992 is still available (Kawamura et al., 1997). As a group,
flavonoids are among the most potent naturally occurring
ARIs (Varma et al., 1975). As part of the ongoing project
some of the new flavones were screened for their AR inhib-
itory activity. Considering that pig AR shares 90% sequence
identity with bovine AR, this study was carried out in search
for the new potential aldose reductase inhibitors that are
useful for the treatment of diabetic complications. Hence,
docking studies were carried out with the newly derived and
known inhibitors on the crystal structure of aldoreductase and
their binding affinities calculated.
redistilled pyridine were added. The contents in the mix-
ture were shaken until the contents become warm. After
20 min, pour the reaction mixture with stirring into 1 M
HCl containing crushed ice. Filter off the product with
suction and wash it with 50 ml of ice-cold methanol and
then with 50 ml of water. Recrystallise from methanol,
cooling in ice before collecting the purified product by
filtration. Report the yield and melting point.
Preparation of O-hydroxydibenzoylmethane
Dissolve (0.2 mol) benzoyloxyacetophenone in dry pyri-
dine in a 1-liter bolt-necked flask and heat the solution to
50°C. Add with mechanical stirring (0.3 mol) of potassium
hydroxide which has been powdered rapidly in a mortar
preheated in oven at 100°C. Continue to stir by hand for
15 min, until the separation of the yellow potassium salt of
the product makes mechanical stirring impossible. Cool the
reaction mixture to room temperature and acidify it by 10%
aqueous acetic acid. Collect the pale yellow solid by suc-
tion filtration and dry it in oven at 50°C. Record the yield
of O-hydroxydibenzoylmethane and melting point.
Preparation of flavones
Dissolve 0.15 mol of O-hydroxydibenzoylmethane in a
flask containing glacial acetic acid and concentrated sul-
furic acid. Attach a reflux condenser and heat the mixture
on a boiling water bath shaking for 1 h. Pour the reaction
mixture with stirring in a crushed ice, and allow the ice to
melt. Filter off the flavone which has been separated and
wash it with water until the washings are no longer acidic
and dry in a oven at 50°C. Report the yield and melting
point at 50°C. Pathway for the synthesis of the flavone is
shown in Fig. 1.
Isolation and purification of aldose reductase
from bovine lens
Bovine lenses were obtained from a local abattoir soon
after slaughtering and were frozen until further use. Lenses
(30–60 g) were homogenized in a volume of cold buffer in
a homogenizer and centrifuged at 10,000 9 g for 20 min to
remove insoluble material. Ammonium sulfate was added
to the supernatant fluid to 40% saturation. The thick sus-
pension was allowed to stand with occasional stirring for
15 min to ensure complete precipitation. Further, this was
centrifuged and the precipitate was discarded. Additional
inert protein was removed by increasing the ammonium
sulfate concentration from 50 to 75% saturation and the
mixture was then centrifuged. The precipitated enzyme was
redissolved in 0.05 M sodium chloride and dialyzed over-
night against 0.05 M sodium chloride to increase the
Materials and methods
Synthesis of flavone
Preparation of O-benzoyloxyacetophenone
The 0.25 mol of O-hydroxy acetophenone was placed in a
flask and 0.35 mol of benzoyl chloride and 50 ml of dry,
123

932
Med Chem Res (2011) 20:930–945
O.CO.C6H5
dimethyl sulfoxide to get 1.8 mM concentration. The blank
contained all the reagents except DL-glyceraldehydes. The
reaction was initiated by adding NADPH and was followed
by UV–visible double-beam spectrophotometer at 340 nm.
A unit of activity was defined as change in absorbance of
0.001 per minute.
OH
C6H5.COCL
PYRIDINE
CO.CH3
CO.CH3
-OH
Protein–ligand docking
OH
The substrate, product, and inhibitors including all hydro-
gen atoms were built and optimized with chemsketch
software suite. Crystal structure of aldoreductase (PDB:
1AH0) from pig having 90% similarity with aldoreductase
from bovine lens was used for docking study. Extremely
Fast Rigid Exhaustive Docking (FRED) version 2.1 was
used for docking (Open Eye Scientific Software, Santa Fe,
NM). FRED docking roughly consists of two steps: shape
fitting and optimization. During shape fitting, the ligand
O
H+
CO.CH2.CO.C6H5
O
FLAVONE
Fig. 1 Pathway for the synthesis of flavone using 1-(2-hydroxy-
phenyl)ethanone as the starting material in the presence of pyridine
and benzoyl chloride
˚
was placed into a 0.5 A-resolution grid box encompassing
specific activity of the purified enzyme. Purification of
aldose reductase was carried by column chromatography
on DEAE-cellulose column (2 9 25 cm) by equilibration
with 0.01 M phosphate buffer and then the dialyzed
enzyme preparation was adsorbed on the column. Further,
it was washed with 0.01 M phosphate buffer until the
absorbance at 280 nm of the elute was\0.1. The elution of
the enzyme was accomplished with a linear gradient. The
inhibition was studied by addition of coenzyme and was
monitored on a UV–visible double-beam spectrophotome-
ter (Schimazdu make) at 340 nm.
all active site atoms (including hydrogen’s) using a smooth
Gaussian potential (McGann et al., 2003). A series of three
optimization filters were then processed, which consists of
(1) refining the position of hydroxyl hydrogen atoms of the
ligand, (2) rigid body optimization, and (3) optimization of
the ligand pose in the dihedral angle space. In the opti-
mization step, four scoring functions were available:
Gaussian shape scoring (McGann et al., 2003), Chemscore
(Eldridge et al., 1997), PLP (Gehlhaar et al., 1995), and
Screenscore (Stahl and Rarey, 2001). Preliminary docking
trials indicated us to select Chemscore for the 3-optimi-
zation filters. This program generates an ensemble of dif-
ferent rigid body orientations (poses) for each compound
conformer within the binding pocket and then passes each
molecule against a negative image of the binding site.
Poses clashing with this ‘bump map’ are eliminated. Poses
surviving the bump test are then scored and ranked with a
Gaussian shape function. We defined the binding pocket
using the ligand-free protein structure and a box enclosing
the binding site. This box was defined by extending the size
Measurement of aldose reductase activity
DL-glyceraldehydes, NADPH, and quercetin were pur-
chased from Sigma-Aldrich Chemicals (Germany),
whereas DEAE-cellulose from Bangalore Genie Pvt. Ltd.
(India). Protein concentration was determined using stan-
dard calibration curve of bovine serum albumin. Bovine
lens for the isolation and purification of aldose reductase
were obtained from freshly slaughtered animals. Melting
points were determined in open capillaries on a Mettler FP
51 apparatus and are uncorrected. IR spectra were taken on
FT-IR Schimazdu Perkin Elmer 1310 Infrared spectro-
photometer. HNMR spectra were recorded on Varian
Gemini (200 MHz) spectrophotometer, and TMS was used
as internal standard. All flavones were prepared by reported
procedure. AR activity was assayed according to the
method described by Hayman and Kinoshita (1965). The
incubation mixture contained phosphate buffer of 0.067 M
(final pH of the reaction mixture was 6.2), lithium sulfate
of 0.2 M, NADPH of 5 9 10-5 M, enzyme solution of
100 ll, DL-glyceraldehydes of 5 9 10-4 M, and distilled
water to a final volume of 3 ml with (24 lM) or without
flavone. Each flavonoid compound was dissolved in
˚
of a ligand by 4 A (add box parameter of FRED). One
unique pose for each of the best-scored compounds was
saved for the subsequent steps. The compounds used for
docking were converted in 3D with OMEGA which has
previously been shown to select a conformation similar to
that of the X-ray input when using appropriate parameters
(Bostrom et al., 2003) (same protocol as above; Open Eye
Scientific Software, Santa Fe, NM). To this set, the coen-
zyme (generation of multiconformer with Omega) corre-
sponding to the modeled protein was added. It is an
implementation of multiconformer docking, meaning that a
conformational search of the ligand was first carried out,
and all relevant low-energy conformations were then rig-
idly placed in the binding site. This two-step process allows
123

Med Chem Res (2011) 20:930–945
933
only the remaining six rotational and transnational degrees
of freedom for the rigid conformer to be considered. The
FRED process uses a series of shape-based filters, and the
default scoring function is based on Gaussian shape fitting
(Schulz-Gasch and Stahl, 2003).
(-376.62) (Fig. 2a) compared to sorbitol (-293.37)
(Fig. 2b). From Table 2 it is evident that enzyme–substrate
complex has large favorable total docking scores of -33.92
using Chemguass, -10.77 of Chemscore, -30.88 of PLP
score, -70.33 of Screenscore, and -230.74 of Shapeguass
score. Docking studies of the inhibitors with the active
site of aldose reductase showed that 3-iodo, 4-methyl,
5-chloroflavone binds with higher affinity (-688.46)
(Fig. 3a) compared to 3-iodo, 5-chloro, 2,4-dichloroflavone
(-612.33) (Fig. 3b), 3-iodo, 4-methyl, 5 chloro,2,4, di-
chloroflavone (-520.3) (Fig. 3c), 4-nitroflavone (-634.74)
(Fig. 3d), 3-iodo, 5-chloroflavone (-670.58) (Fig. 3e), and
2-chloroflavone (-575.68) (Fig. 3f). The experimental
results also show that 3-iodo, 4-chloro, 5-chloroflavone,
3-iodo, 5-chloro, 2,4-dichloroflavone, 3-iodo, 4-methyl, 5
chloro,2,4, dichloroflavone inhibit 80–90% of the enzyme
activity (Table 1), thus correlating exactly with our dock-
ing studies (Table 3). The only inhibitor, which is deviat-
ing with experimental results, is 4-nitroflavone, which
exhibited 60–70% of enzyme inhibition. This flavone
shows higher affinity compared to 3-iodo, 4-methyl,
5-chloro, 2,4-dichloroflavone and 3-iodo, 5-chloro, 2,4-di-
chloroflavone with the crystal structure of pig aldose
reductase. Hydrogen bonding and hydrophobic interactions
were clearly seen between O17 atom of 3-iodo, 4-methyl,
5-chloro, 2,4-dichloroflavone with nitrogen atoms of
Ala299 and Leu300. Also, interaction of O16 of 3-iodo,
5-chloro, 2,4-dichloroflavone and 2,4-dichloroflavone with
nitrogen atoms of Ala299 and Leu300 with two hydrogen
bonding in addition to seven hydrophobic interactions with
Trp20, Trp79, Trp111, Phe122, Leu124, Trp219, and
Leu300. The O17 atom of 3-iodo, 4-chloro, 5-chloroflav-
one showed hydrogen bonding interactions with nitrogen
atoms of Ala299 and Leu300, O16 atoms of 3-iodo,
5-chloroflavone and 2-chloroflavone displayed its interac-
tion with nitrogen atoms of Ala299 and Leu300 and O20 of
4-nitroflavone shows two hydrogen bonding with nitrogen
atoms of Leu300 and Met301 residues of the active site
predicting that the hydrophobic ring system of these
inhibitors gets trapped tightly in a pocket formed by
Trp111, Lue300, and Phe122 and are shared by both the
active site between Trp20 and Trp111-Phe122 and speci-
ficity pocket between Trp111, Leu300, and Phe122. These
results indicate that benzene ring of the inhibitors was
located within hydrophobic contacts with the side chains of
Trp20, Val47, Phe122, Phe121, Trp219, and Leu300.
These interactions were suggested to be responsible for the
high affinity and selectivity of inhibitors for aldose
reductase as shown in Table 3. These studies also revealed
that Ala299 and Leu300 are important anchoring residues
for the inhibitor binding (Table 3). To evaluate these
results, docking studies of the inhibitors with site-directed
mutagenesis were carried out with the crystal structure of
Bioactivity
The bioavailability of compounds was assessed using
adsorption, distribution, metabolism, elimination (ADME)
prediction methods. Compounds were also tested to the
Lipinski’s Rule of Five using MOLINSPIRATION. The
polar surface areas (PSA) were calculated since it is another
key property linked to drug absorption, including intestinal
absorption, bioavailability, Caco-2 permeability, and blood–
brain barrier penetration. Thus, passively absorbed mole-
2
˚
cules with a PSA [ 140 A are thought to have low oral
bioavailability. Drug likeliness was also calculated using
OSIRISserver, which is based on a list of about 5,300distinct
substructure fragments created by 3,300 traded drugs as well
as 15,000 commercially available chemicals yielding a
complete list of all available fragments with associated drug-
likeliness. The drug score combines drug-likeliness, cLogP,
logS, molecular weight, and toxicity risks as a total value
which may be used to judge the compound’s overall potential
to qualify for a drug.
Results
Activity of the enzyme
The flavones 3-iodo, 4-chloro, 5-chloroflavone, 2,4-di-
chloroflavone, 3-iodo, 4-methyl, 5-chloro, 2,4-dichlorof-
lavone, 3-iodo, 5-chloro, 2,4-dichloroflavone inhibited
80–90%, 3-iodo, 5-chloroflavone 70–80%; 4-nitroflavone
and 2-chloroflavone 60–70% of the activity (Table 1).
Docking studies of the substrate, product, and flavone
derivatives
Docking of the substrate, product, and flavone derivatives
with the active site of the enzyme aldoreductase (PDB:
1AH0) was performed using FRED v 2.0, which is based
on rigid body shape-fitting. To understand interactions
between the enzyme–substrate, product, and inhibitors,
enzyme–product, and enzyme–inhibitor complexes were
generated using the OPENEYE software suite (Figs. 2, 3,
4). Table 2 shows the docking scores of substrate and
product including the Chemgauss score, Chemscore, PLP
score, Screenscore, and Shapeguass score for all the resi-
dues in the active sites of enzyme. From Table 2 it was
found that the glucose is binding with higher affinity
123

934
Med Chem Res (2011) 20:930–945
Table 1 Synthesized flavone derivatives used in the study for the inhibition activity of aldose reductase
S. no.
1
Compound
Chemical structure
Enzyme inhibition
2-Chloroflavone
?
O
Cl
O
O
2
4-Nitroflavone
O
N+
?
O-
O
I
3
3-Iodo-5-chloroflavone
??
O
Cl
O
O
I
4
3-Iodo, 4-methyl, 5-chloroflavone
???
H3C
Cl
O
Cl
Cl
5
3-Iodo, 4-methyl, 5-chloro,2,4-dichloroflavone
???
I
H3C
Cl
O
O
Cl
6
3-Iodo, 5-chloro,2,4-dichloroflavone
???
I
O
Cl
Cl
O
?: 60–70% inhibition
??: 70–80% Inhibition
???: 80–90% Inhibition
123

Med Chem Res (2011) 20:930–945
935
Fig. 2 Binding of glucose (a) and sorbitol (b) in the active site of aldoreductase. Substrate, product and residue labels are represented in blue
color and protein is represented in red color. Distances in angstroms are represented in black lines (Color figure online)
pig aldoreductase. Based on the molecular structure and
reaction mechanisms, active site residues in the hydro-
phobic pocket were hypothesized to be important for the
reaction. Thus, we selected for mutagenesis those amino
acids that are thought to be important for the inhibitor
binding and hydrophobic pocket formation. Mutation of
Val47 to Ile (PDB: 2PD5) which caps the catalytic site and
accommodates between the aromatic moieties of Tyr48 and
Phe122 to isoleucine pushes the neighboring Phe121 side
chain slightly out of place compared to wild type. This
resulted in the loss of activity of 2-chloroflavone; 3-iodo,
5-chloro, 2,4-dichloroflavone and 4-nitroflavone with the
docking scores of -569.91, -606.22, and -630.99 but
exhibited good affinity against 2,4-dichloroflavone, 3-iodo,
4-chloro, 5-chloroflavone, 3-iodo, 4-methyl, 5-chloro,
2,4-dichloroflavone, 3-iodo, 5-chloroflavone compared to
wild type (Steuber et al., 2008) (Table 3). Mutation of
Tyr48 to His (PDB: 2ACU) involved in proton donor,
functioning as a general acid catalyst to facilitate the
hydride transfer from NADPH to the carbonyl carbon of
the substrate to His altered both the hydrophobic properties
and steric bulk of the residue. This resulted in partially
retained hydrogen bonding of the hydroxyl moiety in an
almost complete loss of activity of 2-chloroflavone and
3-iodo, 4-chloro, 5-chloroflavone compared to wild type
protein (Bohren et al., 1994). Phe121, located close to
Val47 and involved in integrity, dynamics and water
accessibility of the sub-pocket between Phe122 and Tyr48,
when mutated to Pro121 (PDB: 2PDB), showed no loss of
activity of these inhibitors compared to wild type. This
shows that ligands occupy both the specificity pocket and
sub-pocket near Val47-Tyr48 and exhibit no greater
decrease in affinity with respect to F121P (Steuber et al.,
2008). Consistent with our hypothesis, mutation of
Trp219Tyr (PDB: 2IPW) resulted in complete loss of
activity of these inhibitors due to loss of hydrophobic
contributions (Brownlee et al., 2006). Mutation of
Cys298Ala (PDB: 2IPW) also resulted in the loss of
activity of the inhibitors which create a new pocket for the
chlorine moieties to bind due to lack of cysteine sulfhydryl
groups. Replacing Cys298 with a smaller amino acid such
as glycine may change the reactivity due to the size of
hydrophobic pocket. Leu300 in C-terminal loop responsi-
ble for open and closed states of the protein when mutated
to proline changed the dynamic properties of the loop
region and disqualified the backbone nitrogen for hydrogen
bonding, but did not change the activity of flavones expect
3-iodo, 4-chloro, 5-chloroflavone and 3-iodo, 5-chlorof-
lavone compared to wild type (Brownlee et al., 2006).
Mutation of Ser302Arg (PDB: 2PDM) resulted in shifting
the backbone regions of Leu301 and Arg302 away from
Val130. This avoids close contacts with the arginine side
chain and reduces the activity of 2-chloroflavone, 3-iodo,
5-chloroflavone and 4-nitroflavone with docking scores of
-574.06, -667.85, and -632.76, respectively (Steuber
et al., 2008). Finally, mutation of Cys303Asp (PDB: 2PDQ)
resulted in loss of activity of these inhibitors compared to
wild type and other mutants (Steuber et al., 2008) (Table 3).
Based on the results from the mutagenesis, it appears that
these active site residues have different roles (Table 4).
123

936
Med Chem Res (2011) 20:930–945
Fig. 3 Binding of 3-iodo,
4-methl, 5-chloroflavone (a),
3-iodo, 5-chloro, 2,4-
dichloroflavone (b), 3-iodo,
4-methyl, 5 chloro,2,4, dichloro
flavone (c), 4-nitroflavone (d),
3-iodo, 5-chloroflavone (e), and
2-chloroflavone (f) in the active
site of aldoreductase enzyme in
the presence of NADPH.
Inhibitors and labels are
represented in blue color and
residues represented in red
color. Distances in angstroms
are represented in black lines
(Color figure online)
123

Med Chem Res (2011) 20:930–945
937
Fig. 4 Binding of isoorientin (a), orientin (b), vitexin (c), cis-
coumaric acid (d), p-coumaric acid (e), luteolin 6-C-(60-O-trans-
caffeoylglucoside) (f), and vittariflavone (g), in the active site of the
aldose reductase (1AH0) in the presence of NADPH. Inhibitors and
labels are represented in blue color and residues represented in red
color. Distances in angstroms are represented in black lines (Color
figure online)
Docking studies with known inhibitors
atoms of Ala299, Met301, and Ser302 with the total
docking score of -794.85 (Fig. 4a), orientin showed three
hydrogen bonding interactions with nitrogen, oxygen, and
O7 groups of Leu300, Val47, and NADPH with the total
docking score of -712.72 (Fig. 4b). While, vitexin showed
two hydrogen bonding interactions with nitrogen atoms of
Leu300 and Met301 (Fig. 4c), cis and para-coumaric acids
exhibited two hydrogen bonding interactions with Ne2 and
To evaluate the results of docking for their accuracy,
known inhibitors isoorientin, orientin, vitexin, cis-couma-
ric acid, p-coumaric acid, luteolin 6-C-(60-O-trans-caf-
feoylglucoside), vittariflavone, tricin were docked against
wild type of aldose reductase (PDB: 1AH0). Isoorientin
showed three hydrogen bonding interactions with N and Oc
123

938
Med Chem Res (2011) 20:930–945
Fig. 4 continued
nitrogen atoms of His110 (Fig. 4d, e), luteolin 6-C-(60-O-
trans-caffeoylglucoside) displayed eight hydrogen bonding
interactions with nitrogen, NH2, Ne, and Ne1 atoms of
Ser303, Lys221, Arg217, and Trp297 (Fig. 4f). Vittarif-
lavone showed four hydrogen bonding interactions with
nitrogen, Sc Ne1 and Ne2 atoms of Met301, Cys298, Trp20,
and His110 (Fig. 4g), but tricin displayed only two hydro-
gen bonding interactions with nitrogen atoms of Met301
and Ala299 residues, respectively (Fig. 4h; Table 5). Total
docking scores calculated are shown in Table 6, and it is
123

Med Chem Res (2011) 20:930–945
939
Table 2 The total scores of Chemguass score, Chemscore score, PLP score, Screenscore, and Shapeguass scores of the best-docked confor-
mations of substrate and product in the active site of aldose reductase
Ligand
Chemguass
Chemscore
PLP score
Screenscore
Shapeguass
Total score
Glucose
Sorbitol
-33.92
-36.34
-10.77
-6.4
-30.86
-11.62
-70.33
-40.59
-230.74
-198.42
-376.62
-293.37
Table 3 Hydrogen bonding
and hydrophobic interactions of
the best-docked conformations
of the flavone derivatives in the
active site of aldose reducatse
Compound
Hydrogen bonding
Hydrophobic
Protein Atom–Atom Lig Protein Atom–Atom Lig
3-Iodo, 4-methyl, 5-chloro, 2,4-dichloroflavone Ala299 N399–O1
Leu300 N400–O1
Lig Trp20
C98–C7
Lig
Lig
Lig Trp79 C399–I1
Trp111 C568–C10 Lig
Phe122 C626–C1 Lig
Leu124 C637–CL1 Lig
Trp219 1117–CL3 Lig
Leu300 1539–C6
Lig
Lig
Lig
Lig
Lig
Lig
Lig
3-Iodo, 4-methyl, 5-chloroflavone
Leu300 N400–O1
Lig Trp20
C98–C2
Val47 C228–c3
Trp79 C399–C5
Phe122 C627–I1
Trp219 1118–C8
Leu300 1535–C6
Met301 1543–CL1 LIG
Lig Trp20 C98–C1 Lig
3-Iodo, 5-chloro, 2,4-dichloroflavone
Ala299 N399–O1
Leu300 N400–O1
Lig Val47 C228–CL1 Lig
Trp79 C399–CL3 Lig
Trp111 C566–CL3 Lig
Phe122 C627–CL3 Lig
Trp219 1117–CL2 Lig
Leu300 1539–CL3 Lig
4-Nitroflavone
Leu300 N400–O4
Met301 N401–O4
Leu300 N400–O1
Lig Leu212 1091–C7
Lig Leu228 1170–C8
Lig
Lig
Lig
Lig
Lig
Lig
Lig
Lig
3-Iodo, 5-chloroflavone
Lig Trp20
C95–C2
Val47 C228–C3
Trp79 C398–C3
Trp219 1118–C9
Leu300 1535–C7
2-Chloroflavone
Trp20
C95–C6
Val47 C227–CL1 Lig
Trp79 C402–C5
Phe121 C626–C9
Phe122 C635–C1
Trp219 1127–C3
Leu300 1552–C1
Lig
Lig
Lig
Lig
Lig
clear from the table that flavones such as isoorientin, vit-
exin, luteolin 6-C-(60-O-trans-caffeoylglucoside) and tricin
exhibit much stronger aldose reductase inhibition than other
compounds exactly correlating with the experimental data
of 1.91, 2.03, 0.0134 and 2.03 lM. Of these compounds,
luteolin 6-C-(60-O-trans-caffeoylglucoside) exhibited
stronger affinity against aldose reductase activity with the
docking score of -829.58 and the inhibitory potency as
123

940
Med Chem Res (2011) 20:930–945
Table 4 Site-directed mutagenesis of the best-docked conformations of flavone derivatives in the active site of aldose reductase
Compound
Wild
type
Val47Ile Tyr48His Pro121Phe Trp219tyr Cys298Ala Leu300Pro Ser303Arg Cys303Asp
3-Iodo, 4-methyl, 5-chloroflavone
-688.46 -687.42 -685.46 -688.48
-360.31 -679.43
-373.13 -601.72
-385.66 -615.01
-651.75
-644.98
-644.04
-688.1
-688.16
-612.11
-619.53
3-Iodo, 5-chloro,2,4-dichloroflavone -612.33 -606.22 -638.14 -612.3
-612.38
-641.56
3-Iodo, 4-methyl, 5-chloro, 2,4-
dichloroflavone
-620.3 -629.97 -637.89 -620.4
3-Iodo, 5-chloroflavone
4-Nitroflavone
-670.58 -669 -678.35 -670.59
-331.71 -660.92
-567.34 -624.91
-342.73 -565.1
-622.89
-700.31
-657.23
-667.85
-632.76
-574.06
-670.33
-634.59
-575.46
-634.74 -630.99 -665.6 -634.88
-575.68 -569.91 -568.81 -575.69
2-Chloroflavone
Total scores are predicted based on Chemguass score, Chemscore score, PLP score, Screenscore, and Shapeguass scores using OPENEYE
software
Table 5 Hydrogen bonding
˚
Distance (A)
Compound
Isoorientin
Lig
Atom–Atom
Protein
interactions of the known
flavonoids predicted using
SPDBV software along with
their distances in Angstroms
Lig
Lig
Lig
Lig
Lig
Lig
Lig
Lig
Lig
Lig
Lig
Lig
Lig
Lig
Lig
Lig
Lig
Lig
Lig
Lig
Lig
Lig
Lig
Lig
O28–N
Ala299
Met301
Ser302
Leu300
Val47
2.17
2.94
3.29
2.91
3.10
3.35
2.23
3.28
2.40
3.35
2.40
3.35
2.20
3.34
2.58
3.03
2.85
3.03
2.32
3.23
3.25
2.97
2.91
2.83
O28–N
H32–Oc
O22–N
Orientin
H31–O
H27–O7
O22–N
NADPH
Leu300
Met301
His110
Met301
His110
Met301
Ser302
Ser302
Lys221
Arg217
Arg217
Arg217
Arg217
Trp297
Met301
Cys298
Trp20
Vitexin
O22–N
Cis-Coumaric acid
O11–Ne2
O12–N
p-Coumaric scid
O11–Ne2
O12–N
luteolin 6-C-(60-O-trans- caffeoylglucoside)
O39–N
O34–N
O38–N
O38–NH2
O38–Ne
O36–NH2
O36–Ne
H36–Ne1
O27–N
Vittariflavone
O35–SG
H34–Ne1
H33–Ne2
His110
expressed by IC50 value was 0.0134 lM. A regression
analysis of docking scores and logIC50 for known flavo-
noids was carried out and the scatter plot was drawn for wild
type (Fig. 5). It was found that the R2 value for wild type
was 0.81 exactly correlate well with the results obtained
from experimental data (Jung et al., 2007). Hence, docking
results of these inhibitors perhaps can be used as a new
pharmacophore for lead generation and optimization of
novel aldose reductase inhibitors. The bioactivity of the
molecules predicted using Molsoft server shows (Table 7)
1–2 rotatable bonds, molecular weight in between 200 and
500 with a logP value between -0.2 and 3.2. The drug
likeliness of the molecules generated was calculated based
on the comparison of the 5,000 marketed drugs (positives)
and 10,000 carefully selected non-drug compounds (nega-
tives). It appears from the results that all of the molecules
contain positive scores within the range from 0 to 4
(Table 8).
123

Med Chem Res (2011) 20:930–945
941
Table 6 The total scores of Chemguass, Chemscore, PLP, Screenscore, and Shapeguass score of the best-docked conformations of known
inhibitors in the active site of aldose reductase
Compound
Chemguass Chemscore PLP score Screenscore Shapeguass Total score IC50 (lM)
Isoorietin
-75.99
-69.89
-64.76
-38.86
-38.86
-22.52
-15
-67.2
-151.95
-130.86
-139.8
-96.02
-96.02
-128.93
-123.2
-120.52
-477.19
-447.64
-430.47
-250.43
-250.43
-556.3
-794.85
-712.72
-713.15
-448.9
1.91
–
Orientin
-49.33
-51.41
-46.49
-46.49
-32.69
-45.95
-58.42
Vitexin
-26.71
-17.1
-17.1
-28.55
-11.53
-20.41
2.03
–
Cis-coumaric acid
p-Coumaric acid
Luteolin 6-C-(60-O-trans-caffeoylglucoside) -83.11
-448.9
0.14
0.0134
–
-829.58
-790.25
-682.52
Vittariflavone
Tricin
-79.5
-530.07
-412.14
-71.03
2.03
Discussion
Aldoreductase inhibitors have been shown to prevent or
delay significantly diabetic complications. Many synthetic
inhibitors of the enzyme were tested for their clinical use,
albeit with limited success (Raskin and Rosenstock, 1987).
Synthetic compounds with diverse structures such as
sorbinil, epalrestat, other hydantoin derivatives, and
flavonoids from natural origin have been extensively
studied and reported to inhibit ALR2 (Beyer-Mears and
Cruz, 1985, Terashima et al., 1984, Inagaki et al., 1982).
Docking studies of 2-chloroflavone and 4-nitroflavone
revealed that chlorine in the second position has higher
affinity compared to fourth due to conformational changes
Fig. 5 Correlation of the docking scores of the known inhibitors with
their experimental logIC50 values
Table 7 Biological activity values of flavone analogs calculated using MOLINSPIRATION server
Name
miLogP TPSA NATOMS MW
NON Nohnh N violation N rotb Volume
2-Chloroflavone
0.565 28.241 18
257.696
2
5
2
2
2
2
0
0
0
0
0
0
0
0
0
0
0
0
1
2
1
1
1
1
216.624
226.423
240.615
240.615
284.247
267.680
4-Nitroflavone
-0.106 74.065 20
268.248
383.592
383.592
466.509
452.482
3-Iodo, 5-chloroflavone
1.6
1.6
28.241 19
28.241 19
3-Iodo, 4-chloro, 5-chloroflavone
3-Iodo, 4-methyl, 5-chloro, 2,4-dichloroflavone
3-Iodo, 5-chloro, 2,4-dichloroflavone
3.261 28.241 22
2.884 28.241 21
Table 8 Carcinogenic property values of flavone analogs calculated using OSIRIS server
Compounds
Muta
genic
Tumaro
genic
Irritant Reproductive
effect
cLogP Solubility MW Drug
likeness
Drug
score
2-Chloroflavone
-
?
-
-
-
-
?
-
-
-
-
-
-
-
-
-
-
-
-
-
4.21
3.33
5.14
5.45
6.68
-4.48
-4.38
-5.5
256.0 2.12
0.63
0.25
0.42
0.36
0.25
4-Nitroflavone
269 2.03
382.0 1.71
396.0 1.36
464.0 3.22
3-Iodo, 5-chloroflavone
3-Iodo, 4-chloro, 5-chloroflavone
-5.84
-7.31
3-Iodo, 4-methyl, 5-chloro, 2,4-
dichloroflavone
3-Iodo, 5-chloro, 2,4-dichloroflavone
-
-
-
-
6.36
-6.97
450.0 3.43
0.27
123

942
Med Chem Res (2011) 20:930–945
of the molecule and its hydrophobic interactions with
Trp20, Val47, Trp79, Phe121, Phe122, Trp219, and
Leu300. These interactions are not seen in the case of 4-
nitroflavone that binds with lower affinity. Docking studies
of the substrate and its product showed that H9 and H11 of
glucose interact with Ne2 of His110 and O7 of NADPH
Cys298Ala, Leu300Pro, Ser302Arg, Cys303Asp, and
double mutation of Ser302Arg and Cys303Asp alters the
activity of these inhibitors compared to wild type due to
conformational changes and their binding modes. These
studies show that these derivatives preferably docked
within the specificity pocket of the active site were heavily
depend on the degree of halogenation. However, the cal-
culated docking scores were not very similar in all the
cases and, rather strikingly, could not be used to tell apart
those molecules that were bound in the active site from
those bound elsewhere. The preferred orientation found for
3-iodo, 4-methyl, 5-chloroflavone was chosen as the most
plausible one for the remaining inhibitors. The different
orientations of these inhibitors were thought to arise due to
increasing the number of chlorine atoms in the third, fourth
and fifth positions in the naphthalene, and bulky size NO2
group relative to chlorine in the side chain phenyl group,
show variations in the docking scores into the rigidly fixed
binding site. The docking study of 3-iodo, 4-methyl,
5-chloro, 2,4-dichloroflavone and 3-iodo, 5-chloro, 2,4-di-
chloroflavone shows that the more heavily halogenated
phenyl ring is sandwiched between the indole ring of Trp-
111 on one side and the side chains of Leu-300 on the other
side. Studies also show that the side chain of Leu300 makes
a direct contact with the aromatic ring of Trp111 in the
native human holoenzyme (PDB code: 1ADS). These
hydrophobic interactions are further stabilized by the
phenyl ring of Phe-122 whose edge contacts halogenated
phenyl ring which are similar to that described for the
larger benzothiazole ring of zopolrestat and zenarestat
(Kinoshita et al., 2002). The oxygen atom allows a nearly
orthogonal orientation of the side chain phenyl ring such
that the electronegative chlorine atom in the second phenyl
ring makes contact with both Trp-20 and the pyridinium
group of NADP?. Incorporation of methyl group in the
fourth position of 3-iodo, 4-methyl, 5-chloroflavone will
likewise result in increase of potency by orientating the
molecule in such a way that both chlorine and iodine are in
very close contact with protein atoms and engaged in
crucial hydrophobic interactions with Phe122 and Met301.
The change of binding mode of these flavones derivatives
is probably caused by the substitution of the chlorine atoms
with NO2 substituents in the side chain phenyl group in the
second position which bring severe steric hindrance
because of its big atomic volume. The structural charac-
teristic correlates well with the experimental inhibition
activities of inhibitors, which vary a lot under these two
binding modes. This is probably because the inhibitors
3-iodo, 4-methyl, 5-chloroflavone, 3-iodo, 4-methyl,
5-chloro,2,4-dichloroflavone and 3-iodo, 5-chloro,2,4-di-
chloroflavone that bear more negative charges because of
chlorine atoms than 4-nitroflavone can form more stable
hydrogen bonds with active residues, which make the
˚
with distances of 2.66 and 2.41 A. This indicates that
His110 at the bottom of the cavity is an important
anchoring residue for the substrate binding during cataly-
sis. Docking analysis also shows that O8 and H12 of sor-
bitol interact with O7 of NADPH and Ne2 of His110. This
preliminary data along with the list of hydrogen bond
interactions between the substrate, product, and residues in
the binding cavity clearly indicate that His110 residue is
more preferred in binding of the substrate, product, and
hydride transfer mechanism. Molecular modeling study
performed by Itzstein et al. showed that tyr48, His110, and
Trp111 are involved in the hydrogen bonding interactions
with three substrates, D-xylose, L-xylose, and D-lylose when
His110 was modeled with its Ne2 atom in the protonated
form and Nd1 atom in the unprotonated form (De Winter
et al., 1995). It was also inferred that reduction of
D-glyceraldehyde to glycerol catalyzed by His110 was
favored over the reduction process catalyzed by Tyr48.
Docking studies with 3-iodo-5-chloroflavone, 3-iodo,
4-methyl, 5-chloroflavone; 3-iodo, 4-methyl, 5-chloro,
2,4-dichloroflavone and 3-iodo, 5-chloro, 2,4-dichlorof-
lavone indicate that methyl in the fourth and chlorine
groups in the second and fourth positions of side chain
benzene ring are responsible for conformational changes in
the active site interacting with higher affinity compared to
3-iodo, 5-chloroflavone. The polar head groups of the
inhibitors interacted with Trp20, Val47, Phe122, Phe121,
Trp219, and Leu300 adding evidence for the generalized
mode of aldose reductase binding. The similarities in the
binding modes for the side chain and main chain chlorine,
iodine, and carboxyl group flavone derivatives in the active
site of aldose reductase are directed by both the hydro-
phobic ring system and polar head groups of the inhibitors.
The binding of these inhibitors is accompanied by a con-
formational change to the side chain of Lue300 and Phe122
that opens the specificity pocket lined by residues Trp111,
Phe113, Phe122, Ala299, and Leu300. These studies show
that Ala299 and Leu300 loop regions exhibit a high level of
plasticity throughout the known crystal structures, partic-
ularly provoked by the less restrained dynamic properties
of Phe122, Ala299, and Leu300. These residues populate
different rotameric states and thereby produce differently
shaped sub-pockets (Urzhumtsev et al., 1996). The inhib-
itor-binding site is a positively charged anion well formed
by Asp43, Lys77, Tyr48, His110, and the nicotinamide ring
at the bottom of the active site pocket (Harrison et al.,
1994). Replacement of Val47Ile, Tyr48His, Phe121Pro,
123

Med Chem Res (2011) 20:930–945
943
chlorine binding mode prevail at the binding strength.
Furthermore, compared to the 2-chloroflavone which bears
chlorine in the fourth position, 3-iodo, 4-methyl, 5-
chloro,2,4-dichloroflavone and 3-iodo, 5-chloro,2,4-di-
chloroflavone cause less severe steric effect and bind to the
aldose reductase active pocket effectively due to free
rotation of the branched phenyl group. To further under-
stand the inhibition mechanism of flavone derivatives, the
substrate (glucose) was docked primarily in the active site
of aldose reductase and forms hydrogen bonds with resi-
dues TYR48 and HIS110 with carbonyl O atom which is
much similar with the inhibitors’ binding mode. But glu-
cose has much lower docking score compared with the
flavone derivatives, predicting that there is a competitive
inhibition of flavone derivatives with more affinity than its
substrate in the active site of aldose reductase forming
stable hydrogen bonds with Ala299 or Leu300 or both and
effectively hinder the proton transfer process to further stop
the reduction reaction. Also based on experimental and
theoretical studies, it has been established that the reduc-
tion of an aldehyde substrate to an alcohol by aldose
reductase follows a stepwise mechanism, which consists of
a hydride transfer from the nicotinamide ring of coenzyme
NADPH and a proton transfer from one of the amino acid
residues in the active site (Varnai et al., 1999; Bohren
et al., 1994; Grimshaw et al., 1995a, b; Lee et al., 1998;
Varnai and Warshel, 2000). According to the residue
mutation experiments, the most possible proton donors are
TYR48 and protonated HIS110 (Varnai et al., 1999; Lee
et al., 1998; Varnai and Warshel, 2000). To further eval-
uate these docking studies, known flavones were docked
such as luteolin 6-C-(60-O-trans-caffeoylglucoside) was
found to exhibit strong aldose reductase activity with the
inhibitory potencies as expressed by IC50 value was
0.0134 mM (Varma and Kinoshita, 1976). This indicated
some possible relationships of the structure to the inhibit-
ing potencies of flavones. The hydroxylation in the 4’-
position has beneficial effects, and the abolition of the
double bond between C-2 and C-3 leads to a decreased
inhibition. In support of these findings, the 40-hydroxyl
group of tested compounds was important for showing the
inhibition of enzyme activity. Studies indicate that quer-
cetin-like structure of luteolin 6-C-(60-O-trans-caffeoylg-
lucoside) has stronger inhibition than other flavones
(Robison et al., 1989). Higher inhibition of ALR2 activity
was reported by Matsuda et al. (2002) with the compounds
having a catechol moiety at the B ring compared to those of
monohydroxyl and pyrogallol moiety (the 30,40,50-trihydr-
oxyl moiety), while opposite examples were found in fla-
vone 8-glucoside, vitexin [(3, IC50 = 2.03 mM) and
[orientin (2, [10 mM)]. A model with the correlation
coefficient (r2) of 0.81 was obtained for five compounds
(isoorientin, vitexin, p-coumaric acid, luteolin 6-C-(60-O-
trans-caffeoylglucoside), tricin) using the equation Y =
- 0.0029 9 -3.6458. This good correlation between
known inhibitors with the crystal structure of pig aldore-
ductase demonstrates that the binding conformations are
reasonable. These studies show that polar head groups of
these inhibitors bind in the active site with three hydro-
philic centers near the coenzyme and hydrophobic groups
bind with different conformational changes in the active
site cleft. Docking studies with known and unknown
inhibitors along with site directed mutagenesis show that
these residues are highly conserved suggesting that the
specificity pocket binds to the substrates that are unique to
AR. Therefore, the study of these AR–inhibitor complexes
and the conformational changes induced by the inhibitors
explain the remarkable adaptability of the active site, as
well as its specificity for certain inhibitors. Drug likeliness
of these molecules was predicted using the OSIRIS server.
The data show that molecules contain positive scores
within the range from 0 to 0.5 which indicated that the
above-mentioned molecules act as drugs and inhibit the
activity of aldose reductase. The results also show that
3-iodo, 4-methyl, 5-chloroflavone binds with aldose reduc-
tase with high affinity and good drug scores. This indicates
that this molecule may be the best inhibitor for aldose
reductase. On the basis of docking results, bioavailability
scores and carcinogenic properties, it is predicted that the
compound 3-iodo, 4-methyl, 5-chloroflavone is the best
antidiabetic. Thus, it is hoped that these newly designed
molecules hold promise for anti-diabetic activity if tested in
animal models.
Conclusion
Molecular docking studies of flavone derivatives based on
the experimental data with aldose reductase reveal that they
can form hydrogen bond networks with active residues by
using the iodine and chlorine groups. Chlorine in the sec-
ond position of side chain phenyl group is replaced with
NO2 substituent or without a substituent; an inhibitor
shows less binding affinity otherwise the inhibitor shows
higher binding affinity. Inhibitors exhibit their anti-aldose
reductase activities mainly by occupying the active site.
Under these two binding modes, flavone derivatives have
different experimental and theoretical inhibition activities
to aldose reductase, and the binding mode alteration can be
conveniently controlled by introducing methyl, chlorine,
and iodine groups or eliminating NO2 substituent on cer-
tain position of inhibitor molecules. Correlation of the
docking studies with inhibitory concentration values of
the known inhibitors against aldose reductase shows the
regression value of 0.81 predicting that these docking
studies are reliable. These results show that the synthesized
123

944
Med Chem Res (2011) 20:930–945
Inagaki K, Miwa I, Yashiro T, Okuda J (1982) Inhibition of aldose
reductases from rat and bovine lenses by hydantoin derivatives.
Chem Pharm Bull 30:3244–3254
Judzewitsch RG, Jaspan JB, Polonsky KS, Weinberg CR, Halter JB,
Halar E, Pfeifer MA, Vukadinovich C, Bernstein L, Schneider
M, Liang KL, Gabbay KH, Rubenstein AH, Porte D Jr (1983)
Aldose reductase inhibition improves nerve conduction velocity
in diabetic patients. N Engl J Med 308:119–125
flavone derivatives used in this study contains some
interesting characteristics which make them new kind of
promising aldose reductase inhibitors.
Acknowledgments The authors are thankful to the Director, Global
institute of Biotechnology, Himayatnagar, Hyderabad for his kind
support for providing the software necessary to carry out this study.
Jung SH, Lee JM, Lee HJ, Kim CY, Lee EH, Um BH (2007) Aldose
reductase and advanced glycation endproducts inhibitory effect
of Phyllostachys nigra. Biol Pharm Bull 30:1569–1572
Kinoshita T, Miyake H, Fujii T, Takakura S, Goto T (2002) The
structure of human recombinant aldose reductase complexed
with the potent inhibitor zenarestat. Acta Crystallogr D 58:
622–626
Kawamura M, Hamanaka N (1997) Development of epariestat
(Kinedak), aldose reductase inhibitor. J Synth Org Chem Japan
37:651–657
References
Beyer-Mears A, Cruz E (1985) Reversal of diabetic cataract by
sorbinil, an aldose reductase inhibitor. Diabetes 34:15–21
Bhatnagar A, Srivastava SK (1992) Aldose reductase: congenial and
injurious profiles of an enigmatic enzyme. Biochem Med Metab
Biol 48:91–121
Bohren KM, Grimshaw CE, Lai CJ, Harrison DH, Ringe D, Petsko
GA, Gabbay KH (1994) Tyrosine-48 is the proton donor and
histidine-110 directs substrate stereochemical selectivity in the
reduction reaction of human aldose reductase: enzyme kinetics
and crystal structure of the Y48H mutant enzyme. Biochemistry
33:2021–2032
Larson ER, Lipinski CA, Sarges R (1988) Medicinal chemistry of
aldose reductase inhibitors. Med Res Rev 8:159–186
Lee YS, Hodoscek M, Brooks BR, Kador PF (1998) Catalytic
mechanism of aldose reductase studied by the combined
potentials of quantum mechanics and molecular mechanics.
Biophys Chem 70:203–216
Bostrom J, Greenwood JR, Gottfries J (2003) Assessing the perfor-
mance of OMEGA with respect to retrieving bioactive confor-
mations. J Mol Graph Model 21:449–462
Brownlee JM, Carlson E, Milne AC, Pape E, Harrison DH (2006)
Structural and thermodynamic studies of simple aldose reductase
inhibitor complexes. Bioorg Chem 34:424–444
Martyn CN, Reid W, Young RJ, Ewing DJ, Clark BF (1987) Six-month
treatment with sorbinil in asymptomatic diabetic neuropathy.
Failure to improve abnormal nerve function. Diabetes 36:987–990
Masson EA, Boulton AJ (1990) Aldose reductase inhibitors in the
treatment of diabetic neuropathy: a review of the rationale and
clinical evidence. Drugs 39:190–202
Matsuda H, Morikawa T, Toguchida I, Yoshikawa M (2002)
Structural requirements of flavonoids and related compounds
for aldose reductase inhibitory activity. Chem Pharm Bull
50:788–795
De Winter HL, von Itzstein M (1995) Aldose reductase as a target for
drug design: molecular modeling calculations on the binding of
acyclic sugar substrates to the enzyme. Biochemistry 34:8299–
8308
McGann MR, Almond HR, Nicholls A, Grant JA, Brown FK (2003)
Gaussian docking functions. Biopolymers 68:76–90
Dvornik D (1987) Hyperglycaemia in the pathogenesis of diabetic
complications. In: Porte D (ed) Aldose reductase inhibition: an
approach to the prevention of diabetic complications. McGraw
Hill, New York
Eldridge MD, Murray CW, Auton TR, Paolini GV, Mee RP (1997)
Empirical scoring functions: I. The development of a fast
empirical scoring function to estimate the binding affinity of
ligands in receptor complexes. J Comput Aided Mol Des
11:425–445
Mylari BL, Larson ER, Beyer TA, Zembrowski WJ, Aldinger CE,
Dee MF, Siegel TW, Singleton DH (1991) Novel, potent aldose
reductase inhibitors: 3,4-dihydro-4-oxo-3-[[5-(trifluoromethyl)-
2-benzothiazolyl] methyl]-1-phthalazineacetic acid (zopolrestat)
and congeners. J Med Chem 34:108–122
Nakai N, Fujii Y, Kobashi K, Nomura K (1985) Aldose reductase
inhibitors: flavonoids alkaloids acetophenones benzophenones
and spirohydantoins of chroman. Arch Biochem Biophys 239:
491–496
Narayanan S (1993) Aldose reductase and its inhibition in the control
of diabetic complications. Ann Clin Lab Sci 23:148–158
O’Brien MM, Schofield PJ, Edwards MR (1982) Inhibition of human
brain aldose reductase and hexonate dehydrogenase by alrestatin
and sorbinil. J Neurochem 39:810–814
Poulsom R (1986) Inhibition of hexonate dehydrogenase and aldose
reductase from bovine retina by sorbinil statil M79175 and
valproate. Biochem Pharmacol 35:2955–2959
Raskin P, Rosenstock J (1987) Aldose reductase inhibitors and
diabetic complications: a review. Am J Med 83:298–306
Robison WG Jr, Nagata M, Laver N, Hohman TC, Kinoshita JH
(1989) Diabetic-like retinopathy in rats prevented with an aldose
reductase inhibitor. Invest Ophthalmol Vis Sci 30:2285–2292
Schulz-Gasch T, Stahl M (2003) Binding site characteristics in
structure based virtual screening: evaluation of current docking
tools. J Mol Model 9:47–57
Gabbay KH, Spack N, Loo S, Hirsch HJ, Ackil A (1979) Aldose
reductase inhibition: studies with alrestatin. Metabolism
28:471–476
Gehlhaar DK, Verkhivker GM, Rejto PA, Sherman CJ, Fogel DB,
Fogel LJ, Freer ST (1995) Molecular recognition of the inhibitor
AG-1343 by HIV-1 protease: conformationally flexible docking
by evolutionary programming. Chem Biol 2:317–324
Grimshaw CE, Bohren KM, Lai CJ, Gabbay KH (1995a) Human
aldose reductase: subtle effects revealed by rapid kinetic studies
of the C298A mutant enzyme. Biochemistry 34:14366–14373
Grimshaw CE, Bohren KM, Lai CJ, Gabbay KH (1995b) Human
aldose reductase: pK of tyrosine 48 reveals the preferred
ionization state for catalysis and inhibition. Biochemistry
34:14374–14384
Harrison DH, Bohren KM, Ringe D, Petsko GA, Gabbay KH (1994)
An anion binding site in human aldose reductase: mechanistic
implications for the binding of citrate cacodylate and glucose 6-
phosphate. Biochemistry 33:2011–2020
Hayman S, Kinoshita H (1965) Isolation and properties of lens aldose
reductase. J Biol Chem 240:877–882
Stahl M, Rarey M (2001) Detailed analysis of scoring functions for
virtual screening. J Med Chem 44:1035–1042
Hoffman PL, Wermuth B, Von Wartburg JP (1980) Human brain
aldehyde reductases: relationship to succinil semi aldehyde
reductase and aldose reductase. J Neurochem 35:354–366
Steuber H, Heine A, Podjarny A, Klebe G (2008) Merging the binding
sites of aldose and aldehyde reductase for detection of inhibitor
selectivity-determining features. J Mol Biol 379:991–1016
123

Med Chem Res (2011) 20:930–945
945
Terashima H, Hama K, Yamamoto R, Tsuboshima M, Kikkawa R,
Hatanaka I, Shigeta YJ (1984) Effects of a new aldose reductase
inhibitor on various tissues in vitro. Pharmacol Exp Ther
229:226–230
Urzhumtsev AG, Skovoroda TP, Vyu Lunin (1996) A procedure
compatible with X-PLOR for the calculation of electron-density
maps weighted using an R-free-likelihood based approach.
J Appl Crystallogr 29:741–744
Varma SD, Kinoshita JH (1976) Inhibition of lens aldose reductase by
flavonoids—their possible role in the prevention of diabetic
cataracts. Biochem Pharmacol 25:2505–2513
Varma SD, Mikuni I, Kinoshita JH (1975) Flavonoids as inhibitors of
lens aldose reductase. Science 188:1215–1216
Varnai P, Richards WG, Lyne PD (1999) Modelling the catalytic
reaction in human aldose reductase. Proteins 37:218–227
Warren JC, Murdock GM, Ma Y, Goodman SR, Zimmer WE (1993)
Molecular cloning of testicular 20-hydroxysteroid dehydroge-
nase. Biochemistry 32:1401–1406
Wermuth B, Monder C (1983) Aldose and aldehyde reductase exhibit
isocorticosteroid reductase activity. Eur J Biochem 127:279–284
Williamson JR, Chang K, Frangos M et al (1993) Hyperglycemic
pseudohypoxia and diabetic complications. Diabetes 42:801–813
Yabe-Nishimura C (1998) Aldose reductase in glucose toxicity: a
potential target for the prevention of diabetic complications.
Pharmacol Rev 50:21–33
Varnai P, Warshel A (2000) Computer simulation studies of the
catalytic mechanism of human aldose reductase. J Am Chem Soc
122:3849–3860
123