Acetic acid aldose reductase inhibitors bearing a five-membered heterocyclic core with potent topical activity in a visual impairment rat model

Article abstract of DOI:10.1021/jm701613h

A number of 1,2,4-oxadiazol-5-yl-acetic acids and oxazol-4-yl-acetic acids were synthesized and tested for their ability to inhibit aldose reductase (ALR2). The oxadiazole derivatives, 7c, 7f, 7i, and 8h, 8i, proved to be the most active compounds, exhibiting inhibitory levels in the submicromolar range. In this series, the phenyl group turned out to be the preferred substitution pattern, as its lengthening to a benzyl moiety determined a general reduction of the inhibitory potency. The lead compound, 2-[3-(4-methoxyphenyl)-1,2,4- oxadiazol-5-yl]acetic acid, 7c, showed an excellent in vivo activity, proving to prevent cataract development in severely galactosemic rats when administered as an eye-drop solution in the precorneal region of the animals. Computational studies on the ALR2 inhibitors were performed to rationalize the structure-activity relationships observed and to provide the basis for further structure-guided design of novel ALR2 inhibitors.

Full text of DOI:10.1021/jm701613h

3182
J. Med. Chem. 2008, 51, 3182–3193
Acetic Acid Aldose Reductase Inhibitors Bearing a Five-Membered Heterocyclic Core with
Potent Topical Activity in a Visual Impairment Rat Model
Concettina La Motta,*,† Stefania Sartini,† Silvia Salerno,† Francesca Simorini,† Sabrina Taliani,† Anna Maria Marini,†
Federico Da Settimo,† Luciana Marinelli,‡ Vittorio Limongelli,‡ and Ettore Novellino‡
Dipartimento di Scienze Farmaceutiche, UniVersita` di Pisa, Via Bonanno 6, 56126 Pisa, Italy, Dipartimento di Chimica Farmaceutica e
Tossicologica, UniVersita` di Napoli “Federico II”, Via D. Montesano 49, 80131 Napoli, Italy
ReceiVed December 21, 2007
A number of 1,2,4-oxadiazol-5-yl-acetic acids and oxazol-4-yl-acetic acids were synthesized and tested for
their ability to inhibit aldose reductase (ALR2). The oxadiazole derivatives, 7c, 7f, 7i, and 8h, 8i, proved
to be the most active compounds, exhibiting inhibitory levels in the submicromolar range. In this series, the
phenyl group turned out to be the preferred substitution pattern, as its lengthening to a benzyl moiety
determined a general reduction of the inhibitory potency. The lead compound, 2-[3-(4-methoxyphenyl)-
1,2,4-oxadiazol-5-yl]acetic acid, 7c, showed an excellent in vivo activity, proving to prevent cataract
development in severely galactosemic rats when administered as an eye-drop solution in the precorneal
region of the animals. Computational studies on the ALR2 inhibitors were performed to rationalize the
structure-activity relationships observed and to provide the basis for further structure-guided design of
novel ALR2 inhibitors.
Introduction
and its polarity hinders an easy penetration through membranes
and subsequent removal from tissues by diffusion. The resulting
elevated intracellular concentration of sorbitol increases cellular
osmolarity, which in turn initiates a cascade of events that lead
to the development of disabling complications, peculiar of
diabetic disease and affecting the nervous, cardiovascular, renal,
and visual systems. In addition to the osmotic imbalance, an
increase in the activity of ALR2 during hyperglycemia causes
a substantial imbalance in the free cytosolic coenzyme ratios
NADPH/NADP+ and NAD+/NADH. This alteration in the
redox state of pyridine nucleotides induces a state of pseudo-
hypoxia, which contributes to the onset of hyperglycemic
oxidative stress through the accumulation of reactive oxygen
species (ROS). ROS, in turn, trigger activation of downstream
mechanisms, namely, protein kinase C (PKC) isoforms, mito-
gen-activated protein kinases (MAPKs) and poly(ADP-ribose)-
polymerase (PARP), as well as the inflammatory cascade, which
sustain the pathogenesis of diabetic complications.5–10 Inhibition
of ALR2 is therefore a useful therapeutic strategy to prevent
the onset or, at least, delay the progression and the severity of
diabetic complications.
Aldose reductase (alditol:NADP+a 1-oxidoreductase, EC
1.1.1.2, ALR2) is a small, cytosolic, monomeric enzyme of 36
kDa. It belongs to the aldo-keto reductase (AKR) superfamily
and catalyzes the NADPH-dependent reduction of a wide variety
of medium- to long-chain aldehydes to their corresponding
alcohols, exhibiting a broad substrate specificity. ALR2 plays
a pivotal role in the development of long-term diabetic
complications, as a wealth of experimental data clearly
demonstrate.1–4 It is the key enzyme of the polyol pathway and
catalyzes the reduction of glucose to sorbitol, which is then
oxidized to fructose by sorbitol dehydrogenase (L-iditol:NAD+,
5-oxidoreductase, EC 1.1.1.14, SD). As ALR2 has a low
substrate affinity for glucose, the conversion of glucose to
sorbitol through the polyol pathway is generally nonsignificant
in normoglycemic conditions, accounting for no more than 3%
of total glucose utilization. In fact, ALR2 must compete directly
with the hexokinase of the glycolytic pathway, and as the
substrate affinity of hexokinase is greater than that of ALR2,
glucose is preferentially phosphorylated into glucose-6-phos-
phate. Conversely, under hyperglycemic conditions, hexokinase
is rapidly saturated and the polyol pathway becomes activated.
Sorbitol is formed more rapidly than it is converted to fructose,
Many compounds have been shown to inhibit ALR2 with
varying degrees of efficacy and selectivity (Chart 1).11–13
Although chemically different, they all share common structural
features, represented by a wide lipophilic, aromatic area and a
ionizable group. These moieties turn out to be essential for an
effective interaction with the binding site of the enzyme, which
appears as divided into two distinct portions: a “catalytic
subpocket” and a “specificity pocket”. The ionizable group of
the inhibitor anchors it to the “catalytic subpocket” through
hydrogen bond interactions with two polar aminoacid residues,
namely Tyr48 and His110. As a result, the lipophilic scaffold
of the molecule turns out to be oriented within the hydrophobic,
highly plastic “specificity pocket”, which exhibits a high degree
of induced-fit adaptation and can potentially host a wide range
of structures.14
* To whom all correspondence should be addressed. Phone:
(+)390502219593. Fax: (+)390502219605. E-mail: lamotta@
farm.unipi.it.
†
Dipartimento di Scienze Farmaceutiche, Universita` di Pisa.
Dipartimento di Chimica Farmaceutica e Tossicologica, Universita`
‡
“Federico II” di Napoli.
a Abbreviations: NADP+, ꢀ-nicotinamide adenine dinucleotide phosphate;
ALR2, aldose reductase; AKR, aldo-keto reductase; NAD+, ꢀ-nicotinamide
adenine dinucleotide; SD, sorbitol dehydrogenase; NADPH, ꢀ-nicotin-
amide adenine dinucleotide phosphate reduced form; NADH, ꢀ-nicotinamide
adenine dinucleotide reduced form; ROS, reactive oxygen species; PKC,
protein kinase C; MAPK, mitogen-activated protein kinase, PARP, poly-
(ADP-ribose)polymerase; ARI, aldose reductase inhibitor; HNE, hydrox-
ynonenal; AGEs, advanced glycation end products; ALR1, aldehyde
reductase; NiT, 3,3-dioxide-1,2-dihydronaphtho[1,2-d]isothiazol-1-one; SAR,
structure-activity relationship; ClogP, calculated n-octanol/water partition
coefficient; ClogD, calculated distribution coefficient; TPSA, topological
polar surface area.
Despite the impressive number of compounds described as
effective ALR2 inhibitors (ARIs), a true and widespread ARI
10.1021/jm701613h CCC: $40.75
2008 American Chemical Society
Published on Web 05/02/2008

Acetic Acid Aldose Reductase Inhibitors
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 11 3183
Chart 1. Aldose Reductase Inhibitors
therapy is not yet an established fact, seeing that to date, only
epalrestat is successfully marketed for treatment of diabetic
neuropathy in Japan. Products that appear to be promising during
in vitro studies or in trials with animal models often fail to
proceed any further, showing uncertain results in clinical trials
with humans. This failures are generally due to either a limited
efficacy or the emergence of adverse side effects. The former
is often caused by an inappropriate dosing schedule of the
inhibitor or by an unfitting pharmacokinetic, tethered with its
absorption, distribution, metabolism, and excretion properties.
The latter are mainly a consequence of the important physi-
ological role performed by the enzyme. Actually, besides the
polyol pathway, ALR2 shows other activities, mostly linked to
oxidative defense mechanisms. This enzyme is highly efficient
in reducing toxic aldehydes such as hydroxynonenal (HNE),
methyl glyoxale, and 3-deoxyglucosone, which arise in large
quantities from pathological conditions connected with oxidative
stress and are responsible for the formation of protein cross-
links and advanced glycation end products (AGEs). ALR2 also
catalyzes the reduction of glutathione conjugates of unsaturated
aldehydes, showing in most cases an efficency higher than that
of the parent compound, free aldehyde.15–17 ALR2 therefore
ensures a complete and efficient removal of the major electro-
philic end products of lipid peroxidation, playing a significant
aldehyde removal function, which becomes highly relevant
under particular conditions of oxidative stress when other
antioxidant mechanisms are overwhelmed. ALR2 shares its
detoxification role with aldehyde reductase (EC 1.1.1.2, ALR1),
a closely related enzyme. Both ALR2 and ALR1 belong to the
AKR superfamily possessing the highest structural homology,
with 65% identity in their amino acid sequences.18 The least
conserved residues are located at the C-terminal end of the
proteins in a region lining the hydrophobic specificity pocket.
This portion of the active site is responsible for the substrate
and inhibitor specificity of the two enzymes and can therefore
be usefully exploited for the design of selective compounds,
which must be able to inhibit ALR2 without affecting the
detoxification activity of ALR1.19–21
Thus, clinically effective and safe therapies to prevent long-
term diabetic complications can be achieved through the
development of novel ARIs, which necessarily must combine
high levels of efficacy and selectivity with suitable pharmaco-
kinetic properties.
Our research group has been working in the ARI field for
many years22–26 with the aim of disclosing novel drugs that are
candidates for the topical treatment of diabetic visual impair-
ments, which are the leading cause of blindness in people aged
from 20 to 75. We have identified structurally different
inhibitors, one of the latest being acetic acid derivatives of 3,3-
dioxide-1,2-dihydronaphtho[1,2-d]isothiazol-1-one (NiT), which
proved to be relatively potent inhibitors, showing IC50 values
in the micromolar/submicromolar range (Chart 1).25 These
compounds made it possible to disclose a novel and unexpected
binding mode to the active site of the enzyme, confirming once
more the high plasticity of ALR2. X-ray analysis of the enzyme
crystal structure complexed with the most active NiT derivative
revealed the opening of a novel subpocket, never highlighted
before, which hosts the naphthyl portion of the scaffold,
extending the catalytic subpocket of the active site.27,28
The most active NiT inhibitor was also investigated in vivo
for its ability to prevent cataract development, the major visual
complication of diabetic disease. Despite its good in vitro
inhibitory potency, this compound did not show any significant
efficacy. When administered as a 3% eye-drop solution in the
precorneal region of severely galactosemic rats, it failed to
prevent lens degeneration. On the contrary, converted to a
suitable isopropyl ester prodrug, it exhibited an excellent activity,
proving to protect 55% of the animals treated from cataract
development when administered with the same dose regimen.
These results suggest that the NiT compound, as a carboxylic
acid, possesses a poor ocular absorption that prevents it from
reaching the lens, the intraocular site of action. It is well-known
that many factors compete with the ocular absorption of a
topically applied drug. They include rapid drainage from the
application site, induced lachrymation, tear turnover, and corneal
permeability, this latter being the prominent one. The cornea

3184 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 11
La Motta et al.
Scheme 1. Synthesis of 3-Aryl-1,2,4-oxadiazol-5-yl-acetic Acids 7a-i and 8a-i
Scheme 2. Synthesis of 2-(2-Phenyloxazol-4-yl)acetic Acids
11a-d
represents the major pathway for drug penetration, and in terms
of permeability it can be considered as consisting of three
primary layers: epithelium, stroma, and endothelium. Both the
epithelium and the endothelium are lipophilic and constitute the
main barriers for hydrophilic compounds. The stroma is an
aqueous layer, thus limiting permeability of lipophilic com-
pounds. Therefore, the ability of a drug to penetrate through
the cornea is strictly dependent on its hydrophilic-lipophilic
properties, which must be adequately balanced. Carboxylic-type
compounds are almost completely in their ionized form at
physiological pH values. Therefore, they generally exhibit a low
in vivo activity due to an unfavorable tissue penetration.
However, a number of literature examples, also in the ARI
field,29 demonstrate that the pharmacokinetic properties of these
compounds can be optimized in order to improve their perme-
ability profile. Moving from these observations, we planned to
realize appropriate structural modifications of the NiT scaffold
in order to design a simpler molecular skeleton possessing a
favorable pharmacokinetic profile, suitable for ocular tissue
penetration. Thus, removing the central phenyl ring from the
naphtho[1,2-d]isothiazole nucleus, we obtained five-membered
ALR2 inhibitors, possessing an acetic acid function and a
hydrophobic group. On the basis of synthetic feasibility and,
above all, on the basis of previously reported works by G. Klebe
and co-workers on similar structures,30,31 we targeted two
different five-membered rings, namely 1,2,4-oxadiazole and
oxazole, as suitable candidates for the development of novel
inhibitors.
out to be fully consistent with the SARs observed and helped
to rationalize them.
Chemistry
The synthesis of the target 1,2,4-oxadiazole inhibitors 7a-i
and 8a-i was performed as outlined in Scheme 1. The
appropriately substituted nitriles 1a-i and 2a-i were treated
with hydroxylamine hydrochloride, in the presence of potassium
carbonate, to give the N-hydroxybenzimidamide and N-hydroxy-
2-phenylacetimidamide intermediates 3a-i and 4a-i, respec-
tively. Condensation of 3a-i and 4a-i with ethyl 3-chloro-3-
oxopropanoate, in the presence of sodium hydride, provided the
ethyl esters 5a-i and 6a-i, which were then hydrolyzed with
boiling sodium hydroxide to obtain, after acidification, the
desired acids 7a-i and 8a-i.
In this work, we present the synthesis and an extensive
biological evaluation of a number of oxadiazole and oxazole
acetic acid derivatives. To clearly define the structure-activity
relationships (SARs) of the two classes of inhibitors, suitable
substituted phenyl or benzyl groups were introduced into
position 3 of the oxadiazole ring and position 2 of the oxazole
ring. Moreover, compound 7c, which proved to be the most
active of all, was also investigated in vivo in view of its ability
to prevent cataract development in severely galactosemic rats.
Molecular docking simulations of the most potent inhibitors into
the human ALR2 binding site were also carried out in order to
propose the mode of binding of these compounds, which turned
The oxazole inhibitors 11a-d were prepared as reported in
Scheme 2, in accordance with a literature procedure.32 Reaction
of the commercially available benzamides 9a-d with ethyl
4-chloro-3-oxobutanoate led to the ethyl esters 10a-d. Hy-

Acetic Acid Aldose Reductase Inhibitors
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 11 3185
Table 2. ALR2 Inhibition Data of Oxazol-4-yl-acetic Acids 11a-d
Table 1. ALR2 Inhibition Data of [1,2,4]oxadiazol-5-yl-acetic Acids
7a-i and 8a-i
N
R
ALR2 IC50 (µM)a
N
n
R
ALR2 IC50 (µM)a
ALR1 IC50 (µM)a
11a
11b
11c
11d
H
n.a.b
7a
7b
7c
7d
7e
7f
7g
7h
7i
8a
8b
8c
8d
8e
8f
8g
8h
8i
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
H
3.33 (2.83-3.83)
2.50 (2.05-2.95)
0.27 (0.23-0.31)
8.43 (6.75-10.11)
3.95 (3.16-4.74)
0.69 (0.61-0.77)
1.08 (0.88-1.28)
2.00 (1.65-2.36)
0.71 (0.63-0.79)
28.4 (22.44-33.36)
5.29 (4.24-6.34)
8.18 (6.71-9.65)
21.7 (17.8-25.6)
5.11 (4.35-5.87)
1.66 (1.43-1.89)
6.30 (5.23-7.37)
0.36 (0.33-0.39)
0.30 (0.22-0.37)
0.05 (0.03-0.06)
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
CH3
OCH3
F
n.a.b
4-CH3
4-OCH3
4-F
4-Cl
4-Br
4-CF3
2-F-4-Br
3-NO2
H
4-CH3
4-OCH3
4-F
4-Cl
4-Br
4-CF3
2-F-4-Br
3-NO2
54.8 (46.58-63.02)
n.a.b
Tolrestat
0.05 (0.03-0.06)
a IC50 (95% CL) values represent the concentration required to produce
50% enzyme inhibition. b n.a.: not active. Inhibition occurred at a concentra-
tion higher than 100 µM.
such as Ponalrestat,33 ranirestat,34 and minalrestat,35 displayed
a good activity as well, although it was less potent than the
bromo-substituted 7f. Finally, the insertion of a nitro group in
the meta position of the phenyl ring, as in 7i (IC50 0.71 µM),
determined a 5-fold enhancement in potency with respect to
the parent compound 7a, confirming the fruitful contribution
of this type of substituent to the interaction with the amino acid
residues surrounding the ALR2 active site, as previously
observed by G. Klebe and co-workers for similar inhibitors.31
To explore the effects of increasing the distance between the
pendant 3-phenyl ring and the five-membered heterocyclic core,
a methylene spacer was inserted between them to give deriva-
tives 8a-i (Table 1). This distance appeared to be a crucial
element in determining the best spatial relationship between the
pharmacophoric groups of these compounds. The lengthened
derivatives were generally less active than the shorter ones, with
the only exception of 8h,i, which exhibited a better inhibitory
activity than the parent 7h,i. However, the novel compounds
displayed parallel SARs to those of series 7: the insertion of
electron-donating substituents in the para position of the benzyl
framework, as in derivatives 8b (R ) CH3, IC50 5.29 µM) and
8c (R ) OCH3, IC50 8.18 µM), determined an appreciable
increase in the inhibitory efficacy with respect to the unsubsti-
tuted compound, 8a (IC50 28.4 µM), and the presence of a
halogen atom provided a rank order of inihibitory potency fluoro
(8d, IC50 21.7) < chloro (8e, IC50 5.11) < bromo (8f, IC50 1.66)
> trifluoromethyl (8g, IC50 6.30). In contrast to the 3-phenyl
series 7, the presence of two halogen atoms in the ortho and
para positions of the benzyl group determined a remarkable
enhancement in the inhibitory activity. Compound 8h proved
to be highly potent, with an IC50 value of 0.36 µM and a 80-
fold increase in efficacy with respect to the lead 8a. In addition,
the introduction of a nitro substituent in the meta position of
the 3-benzyl group gave 8i (IC50 0.30 µM), which proved to
be the most active of the whole series. These results confirm
once more the positive effect of this substitution pattern in a
specific and forceful interaction with the enzyme active site.
Finally, to investigate the role played by the oxadiazole
heterocyclic fragment of the inhibitor template in the interaction
with the enzyme, a number of oxazol-4-yl-acetic acids deriva-
tives, 11a-d, were likewise synthesized and tested. As sum-
marized in Table 2, none of them proved to inhibit ALR2
significantly, with the only exception of the para-methoxy
substituted 11c (IC50 54.8 µM), which, although effective,
displayed a 200-fold decrease in inhibitory potency when
compared to the corresponding oxadiazole 7c.
Tolrestat
a IC50 (95% CL) values represent the concentration required to produce
50% enzyme inhibition.
drolysis of 10a-d under alkaline medium afforded, after
acidification, the target compounds 11a-d.
Results and Discussion
Biological Evaluation. We started this research project by
synthesizing a number of 3-phenyl-1,2,4-oxadiazol-5-yl-acetic
acid derivatives, 7a-i, and testing them for their efficacy against
ALR2 and selectivity for ALR1. Within this series, different
substituents were introduced in various positions around the
phenyl ring in order to verify the influence of electron-
withdrawal, electron-release, steric bulk, and orientation on the
inhibitory potency of these molecules. As shown in Table 1,
which lists the activity against ALR2 expressed as IC50 values,
all the compounds synthesized proved to inhibit this enzyme,
exhibiting potency levels in the micromolar/submicromolar
range. The insertion of electron-donating substituents in the para
position of the 3-phenyl group determined a general increase
in the inhibitory potency with respect to the unsubstituted parent
compound 7a (IC50 3.33 µM). Actually, derivative 7b, bearing
a methyl group, showed an IC50 value of 2.50 µM and derivative
7c, carrying a methoxy substituent, proved to be extremely
potent, with an IC50 value of 0.27 µM and a gain in inhibitory
potency of 12-fold when compared to the lead 7a.
The presence of electron-withdrawing, halogen atoms in the
same position as the pendant phenyl ring gave rise to opposite
but not contrasting results. In fact, moving from the less bulky,
more electron-attracting halogen to the bulkier, less electron-
attracting one, we observed a progressive enhancement in the
IC50 values of the corresponding compounds, which moved from
8.43 µM for derivative 7d to 3.95 µM for 7e and to 0.69 µM
for 7f, bearing a fluoro, chloro, and bromo atom, respectively.
Clearly, the size and the electronic feature of the halogen atom
affected the binding affinity of these molecules considerably,
allowing for each of them a different interaction with the
enzyme. The insertion of the bulkier trifluoromethyl group broke
off this positive trend, as derivative 7g showed an IC50 value
of 1.08 µM. Compound 7h (IC50 2.00 µM), featuring the ortho-
fluoro para-bromo substitution pattern typical of potent ARIs
All the compounds synthesized, 7a-i, 8a-i, and 11c, were
assayed for their ability to inhibit ALR1, but none of them
showed any appreciable activity (IC50 > 10 µM, Table 1),
proving to be completely selective inhibitors of ALR2.

3186 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 11
La Motta et al.
Table 4. Effects of Treatment with Ophthalmic Solution of 7c and
Tolrestat on Development of Nuclear Cataract in Severely Galactosemic
Rats
Table 3. Physicochemical Property Predictions of Compounds 7c,f,g,i,
8f,h,i, and Tolrestat
a
N
pKa (M)
ClogPb
ClogDc
TPSAd (Å)
rats with nuclear cataract
7c
7f
7g
7i
8f
8h
8i
4.149
3.711
3.116
3.265
3.865
3.367
3.681
3.065
0.40
1.00
1.00
-0.92
0.90
0.09
-0.24
0.11
-1.92
-1.63
0.12
-3.00
-0.63
-0.26
-2.44
0.06
85.45
76.22
76.22
122.04
76.22
76.22
122.04
81.86
day of treatment
control
7c (3%)
Tolrestat (3%)
11
12
13
14
15
16
17
18
19
20
21
13
25
25
25
25
31
50
65
80
92
100
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Tolrestat
a pKa predictions refers to the carboxylic groups. b Calculated n-octanol/
water partition coefficient. c Calculated distribution coefficient. d Topological
polar surface area.
13
13
20
Pharmacological Evaluation of Compound 7c. The target
of our research program was to disclose potent and selective
ALR2 inhibitors as novel potential candidates for the topical
treatment of diabetic visual impairments. We therefore tested
the efficacy of the novel compounds in preventing nuclear
cataract after administration as an eye-drop solution in the
precorneal region of severely galactosemic rats. Topical ap-
plication is, in fact, the best way of drug administration for the
treatment of ocular diseases, as it makes it possible to achieve
significant drug levels in the site of action, thus circumventing
first-pass metabolism in the liver and avoiding undesirable side
effects typical of systemic administration.36 As stated in the
Introduction, topically administered drugs must possess ad-
equately balanced hydrophilic-lipophilic properties in order to
accomplish an exhaustive ocular absorption. Thus, focusing on
the most active in vitro inhibitors, namely 7c,f,g,i and 8f,h,i,
we analyzed their key physicochemical properties in order to
identify the best candidate for the in vivo assay. Table 3 lists
pKa, ClogP, ClogD, and TPSA data, which were calculated in
silico as useful descriptors to estimate ionization, lipophilicity,
and polarity, respectively, of the selected compounds.37 For a
proper comparison, the same values were calculated also for
tolrestat, a carboxylic-type ALR2 inhibitor, which proved to
prevent cataract development in severely galactosemic rats when
topically administered through ocular instillation.36 Derivative
7c, showing the best inhibitory potency, turned out to be less
ionized, more lipophilic, and fairly polar when compared to its
analogues. Moreover, 7c emerged also as less ionized, more
lipophilic, and almost as polar as tolrestat, thus providing, among
all compounds, the most promising data set for an adequate
ocular uptake. Therefore, 7c was administered as a 3% solution
to young rats fed a 50% galactose diet, four times daily for 21
days, in order to verify its ability to prevent or, at least, delay,
nuclear cataract development. Its effectiveness was evaluated
with respect to tolrestat. The pharmacological data are reported
in Table 4. At the end of the assay, 100% of galactosemic rats
treated only with the vehicle developed nuclear cataracts,
whereas an almost complete and significant protection was
detected in rats treated with 7c. No nuclear cataracts were
developed by rats administered a 3% ophthalmic solution of
tolrestat. Furthermore, a slit-lamp examination of rat lenses was
performed to better evaluate the progression of opacification
(Figure 1). Galactosemic animals treated with the vehicle
showed a mature cataract (Figure 1a), while up to 80% of the
animals treated with 7c developed widespread small vacuoles
in the equatorial area, attaining stage 2 of lens opacity as graded
by Kador et al.38 (Figure 1b). The same was also true for all
the animals administered with tolrestat, whose lenses, reported
in Figure 1c, showed only slight changes with respect to lenses
from animals on a normal diet, administered with the vehicle,
which remained clear throughout the assay (Figure 1d).
Thus, the novel five-membered ARI 7c proved to be not only
a potent and selective inhibitor but also an excellent drug
candidate for the topical treatment of ocular diseases.
Docking Studies. A three-dimensional comparison of the
available aldose reductase-inhibitor X-ray complexes underlines
the fact that ALR2 has a fairly flexible binding site able to
change its shape with respect to the different inhibitors bound.
In particular, the most striking difference resides in a pocket
named the “specificity pocket”, which can be closed, as in the
case of sorbinil, or open, as in the case of IDD594, tolrestat, or
zopolrestat. The matter is further complicated by the fact that
the degree and the mode of the pocket opening is inhibitor-
dependent, e.g., in the case of tolrestat, the opening of the
specificity pocket is perpendicular to that of IDD594. Very
recently, it has been demonstrated that neither classic molecular
dynamics simulations nor the currently available flexible docking
approach are really able to predict such rearrangements upon
ligand binding and that the Glide docking program, which
performs a sort of “induced-fit docking”, did not succeed in
obtaining better binding modes with respect to other docking
programs such as Autodock, which fits the inhibitor into a rigid
enzyme conformation.27 As a result, it seems that, to date, the
best approach aimed to find reliable binding modes for the ALR2
inhibitors is to consider multiple ALR2 binding conformations
in the docking process.
For this reason, compound 7c, the most active in the series,
was docked by means of the Autodock program into the three
main binding conformations of ALR2, best represented by the
complexes with sorbinil (PDB code: 1AH039), tolrestat (PDB
code: 2FDZ40), and IDD594 (PDB code: 1US041). Among the
different docking programs, Autodock was chosen for our
studies as it had already been effectively employed also in
searching for the most favorable ALR2 conformation to host a
given inhibitor27 and it had been successfully used in scoring
the different binding conformations of ARL2 inhibitors.31
Regarding the docking of 7c in 1AH0, although the
calculation converged toward one highly populated cluster
(40 times out of 50, ∆G of binding ) -7.0 kcal/mol), it did
not succeed in providing reliable binding modes, as for all
the proposed solutions, the anion-binding pocket is filled by
the phenyl ring, while the carboxylate group points to-
ward the solvent and interacts with Met301 and/or Ser302
backbone NH, in contrast with the experimentally observed
binding modes for almost all other carboxylate-type inhibi-
tors. As a consequence, 1AH0 was not considered for further
docking studies on our compounds.
On the other hand, better results, although not completely
convincing, were obtained when 2FZD was used as the possible

Acetic Acid Aldose Reductase Inhibitors
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 11 3187
Figure 2. Stick representation of (a) compound 7c (green) and (b) 7f
(green) bound into the active site of ALR2 shown as light-blue Connolly
surface. The NADP+ is represented with its carbon atoms colored in
orange. ALR2 residues relevant for inhibitor binding have been labelled,
and their side chains are shown as white sticks. Nonpolar hydrogens
have been removed for clarity reasons. Hydrogen bonds are displayed
as yellow dashed lines.
solvent-exposed. Therefore, this binding mode does not fully
agree with all the SARs developed concerning the phenyl ring.
Docking calculations using the IDD594-bound conformation
(PDB code: 1US0) definitely showed the best results. Indeed,
docking calculations of compound 7c clearly converged to one
binding mode (31 out of 50 runs, ∆G ) -8.5 kcal/mol),
consistent both with the known carboxylate-type inhibitor
binding modes and the SAR studies observed for series 7.
Moreover, docking of 7c into the IDD594-bound conformation
resulted in a better energy score, which is more in line with the
experimentally tested inhibitory activity. The inhibitor is tightly
anchored in the anion-binding pocket, with its carboxylate
function involved in an electrostatic interaction with the
positively charged nicotinamide moiety of the cofactor NADP+,
and a network of H-bonds with Tyr48 OH, His110 Nꢀ2H, and
Trp111 Nꢀ1H (Figure 2a). The phenyl ring of 7c gets tightly
trapped in the hydrophobic cage of the “specificity pocket”
formed by Trp79, Trp111, Phe115, Phe122, and Leu300,
performing aromatic and hydrophobic interactions with Phe115,
Phe122, Trp79, and Leu300 and a parallel π-π interaction with
the Trp111 side chain with a distance of 3.7 Å between the
two aromatic planes.
The oxygen of the methoxy group was hydrogen-bonded to
the Thr113 OH (the distance between the two oxygens is 3.2
Å), while the methyl group formed van der Waals contacts with
the Thr113 Cγ and the Cys303 Cꢀ (distances are 3.1 and 3.2 Å,
respectively). These latter two interactions would explain the
difference in activity observed between 7b and 7c and 7a and
7b, respectively.
Regarding the oxadiazole spacer of 7c, although the oxygen
of the ring was not found to be involved in any direct H-bond
with the surrounding residues, due to its withdrawing property,
it certainly improves stacking interactions with electron-rich
rings such as the indole moiety of Trp111. However, we have
to note the close proximity of the oxygen atom to the Cys298
SH group (≈ 3 Å), which, together with the known flexibility
of the Val297-Leu300 region,31,42 would suggest a possible
interaction.
As a result, our docking studies performed on 7c indicate
that, rather than the three main ALR2 conformations, the
IDD594-bound structure is the most favorable to host our
inhibitors. A superimposition of 7c and IDD594 in their bound
conformations (1US0 as receptor) shows that both compounds
are firmly anchored in the active site through hydrogen bonds
to His110, Tyr48, and Trp111 and establish a strong electrostatic
interaction with NADP+. Moreover, both the inhibitors share
Figure 1. Slit-lamp evaluation of cataract formation in severely
galactosemic rats after 21 days of treatment with test compounds 7c
and tolrestat. Image (a) shows lenses of galactosemic control rats
administered only the vehicle. Image (b) shows lenses of galactosemic
rats treated with 7c at 3%. Image (c) shows lenses of galactosemic rats
treated with tolrestat at 3%. Image (d) shows lenses of non-galactosemic
control rats administered only the vehicle.
conformation. Indeed, the most highly populated family (∆G
of binding ) -6.6 kcal/mol found 13 times out of 50) has the
inhibitor carboxylate group anchored in the anion-binding pocket
through two hydrogen bonds, with Tyr48 OH and His110 Nꢀ2H,
and an electrostatic interaction with the nicotinamide moiety
of the cofactor NADP+. However, the para-methoxy phenyl
ring is not deeply fitted into the “specificity pocket”, as it is
supposed to be from the experimental data, but remains partially

3188 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 11
La Motta et al.
Figure 3. (a) Superimposition of the theoretical bound conformations
of (a) 7c (green), (b) 7f (yellow), and the experimental bound
conformation of IDD594 (pink) in the active site of ALR2 (PDB code:
1US0) shown as light-blue Connolly surface. The NADP+ is represented
with its carbon atoms colored in orange. ALR2 residues relevant for
inhibitor binding have been labelled, and their side chains are shown
as white sticks. Nonpolar hydrogens have been removed for clarity
reasons. Hydrogen bonds are displayed as yellow dashed lines.
Figure 4. Electrostatic potential of compound 7d and 7f computed
using B3LYP and 6-31G* basis set with a scale of -0.017 (red) to
0.017 hartree (blue).
the stacking between one of their aromatic rings and the Trp111
side chain and interact with the Thr113 side chain (Figure 3a).
The results of our docking experiments using compound 7c
as the ligand are perfectly in line with results of a virtual
screening (VS) experiment performed by G. Klebe et al.30 using
the IDD594-bound conformation as the ALR2 structure. From
this in silico screening, six new carboxylate-type leads were
found and, interestingly, two of them, namely 4-[3-(3-nitrophe-
nyl)-1,2,4-oxadiazol-5-yl]butanoic acid and 2-[5-(5-nitrofuran-
2-yl)-1,3,4-oxadiazol-2-yl-thio]acetic acid, share an overall
similarity with ours. A crystallographic study of the complex
between the 1,2,4-oxadiazole-containing inhibitor and ALR2
revealed the presence of an interstitial water molecule, mediating
a hydrogen bond contact between the Trp111 Nꢀ1H and the
carboxylic group of the inhibitor. On the basis of these results,
we cannot exclude a priori the presence of a water molecule to
mediate a similar contact in our 1,2,4-oxadiazole inhibitors,
hence 7c was also docked in the 1,2,4-oxadiazole-bound
conformation of ALR2 taking into account such a water
molecule (PDB code: 2IKG31). Docking results converged to
one highly populated cluster of solutions (37 times out of 50,
∆G ) -6.4 kcal/mol), which showed that, while the carboxylate
group of 7c H-bonds with the water molecule mediating the
contact with Trp111 Nꢀ1H, it is not close enough to interact
with the Tyr48 and His110 side chains. As a consequence of
the lack of these interactions, the ∆G of binding of the lowest
energy conformation of the cluster is fairly high (∆G ) -6.4
kcal/mol) and considerably worse than that calculated when the
docking of 7c was performed using 1US0 as the receptor (∆G
) -8.5 kcal/mol). Thus, the presence of such a water molecule
seemed unlikely in the case of our inhibitors, and consequently
the docking of 7f, 7i, 8a, and 8h was subsequently performed
using the IDD-bound ALR2 conformation (1US0).
Figure 5. Stick representation of (a) compound 7i (green) and (b) 8h
(green) bound into the active site of ALR2 shown as light-blue Connolly
surface. The NADP+ is represented with its carbon atoms colored in
orange. ALR2 residues relevant for inhibitor binding have been labelled,
and their side chains are shown as white sticks. Nonpolar hydrogens
have been removed for clarity reasons. Hydrogen bonds are displayed
as yellow dashed lines.
one with the bromine atom gave the best IC50 value (see
compounds 7d, 7e, and 7f in Table 1). With the aim of
rationalizing these SAR data and to get an insight into the
electrostatic potential profile of these compounds, we performed
some ab initio calculations mapping the electrostatic potential
of 7d and 7f onto the molecular surfaces representing the
respective electron densities (Figure 4). While the fluorine atom,
due to its small atomic radius, is entirely electronegative and
consequently unable to engage in an interaction with an oxygen
atom lone pair, the larger bromine shows the emergence of an
electropositive crown, giving rise to a charge-transfer interaction
with the electron pairs of the Thr113 Oγ. Regarding 7e,
molecular electrostatic potential calculations revealed a reduced
size and intensity of the electropositive crown for the chlorine
atom, compared with that of bromine; this is perfectly in line
with the observation that 7e has an IC50 lower than 7d but higher
than 7f. Thus, the most favorable activity of 7f, if compared
with 7d and 7e, can be attributed to the better halogen-oxygen
lone pair interaction established with Thr113 Oγ.
Regarding compound 7i, it exhibits a submicromolar IC50
(0.71 µM) and docking calculations converged to one single
cluster of solutions. The hydrogen bonds with Tyr48 OH, His110
Nꢀ2H, and Trp111 Nꢀ1H were all maintained, as well as the
electrostatic interaction with the nicotinamide moiety of the
NADP+ cofactor. The phenyl ring occupied the specificity
pocket stacked between the Tpr111 and Leu300 side chains,
with the nitro group anchored to the upper part of this pocket
through two hydrogen bonds with the Cys80 and Thr113 side
chains (Figure 5a). However, it has to be said that, according
to the previously mentioned X-ray analysis31 of the butanoic
The molecular docking calculations of 7f converged toward
one single cluster of solutions, where the inhibitor was involved
in hydrogen bonds with Tyr48 OH and His110 Nꢀ2H and
electrostatic interactions with the positively charged nicotina-
mide moiety of the cofactor (Figure 2b). The polarizable
bromine atom is lodged at the tip of the hydrophobic pocket
and proved to be quite close to Thr113 ≈ 3.2 Å. Interestingly,
the position of this deeply buried bromine atom is spatially
equivalent to that occupied by the bromine atom present in
zenarestat and in IDD594 (Figure 3b).
The SAR studies clearly indicate that, among the compounds
sharing a halogen in the para position on the phenyl ring, the

Acetic Acid Aldose Reductase Inhibitors
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 11 3189
Table 5. Results of 50 Independent Docking Runs for Each Inhibitor
Using the ALR2/IDD594-Bound Conformation (1US0)
a
b
c
e
N
Ntot
focc
∆Gbinding
T-Kid
IC50
7c
7f
7i
8a
8h
7
1
1
14
6
31
50
50
18
37
-8.5
-6.9
-8.4
-8.2
-8.4
546
8500
690
1000
696
270
690
710
28400
360
a Ntot is the total number of clusters. b focc is the frequency of occurrence
and it represents the number of results in the chosen cluster (see
Experimental Section). c ∆Gbinding is the estimated free energy of binding
of the chosen cluster and is given in kcal/mol. d T-Ki is the theoretical Ki
predicted by the AutoDock program, and it is expressed in nM. e IC50 is
expressed in nM and it refers to the experimentally determined IC50 reported
in Table 1.
Figure 6. Electrostatic potential of compounds 7a, 8a, and 8h
computed using B3LYP and 6-31G* basis set with a scale of -0.017
(red) to 0.017 kcal mol-1 (blue).
by the absence of substituents on the benzyl ring capable of
establishing any favorable contact with the enzyme.
acid inhibitor containing the 1,2,4-oxadiazole core structure, the
presence of the meta-nitro substituent is able to induce a flip of
the Ala299-Leu300 backbone, thus establishing a hydrogen
bond between one of the ligand’s nitro oxygens and the Leu300
backbone NH (PDB code of the complex: 2IKG). Moreover, a
small rotation of the ꢁ1 of Tyr309 making room for the nitro
group was also observed, allowing a “nonclassical hydrogen
bond” of the nitro oxygen with the Cδ1H and Cꢀ1H of Tyr309.
These two structural variations with respect to the 1US0
structure seem to be typical of the nitro-containing inhibitors,
as compounds possessing a weaker hydrogen bond acceptor such
as a fluorine atom in the corresponding position were not able
to induce a similar enzyme rearrangement. Thus, just for
compound 7i, with the aim of obtaining further insight into the
positioning of the nitro group, docking calculations were also
performed using 2IKG as the receptor. Surprisingly, even if the
lowest energy cluster (∆G ) -7.7 kcal/mol) was not highly
populated (8 out of 50 runs), it showed a binding mode perfectly
superimposable on that obtained using 1US0 as the referring
model. By contrast, the most populated cluster (25 out of 50
runs) clearly showed the nitro group bound to the Leu300
backbone NH group, but the main interactions with Tyr48 OH,
His110 Nꢀ2H, and Trp111 Nꢀ1H were completely lost, thus
leading to a higher ∆G value (-6.5 kcal/mol). In conclusion,
the docking calculations on 7i would lead one to think that the
nitro group binds to the upper part of the specificity pocket,
although the other hypothesis cannot be excluded a priori due
to a possible rearrangement of the enzyme upon ligand binding.
On the basis of these results, in prospect it would be worth
inserting two hydrogen bond acceptors in both of the meta
positions of the phenyl ring of 7i with the aim of simultaneously
binding the upper and the lower part of the specificity pocket.
Accordingly, inhibitor 8h, bearing an ortho-fluoro, para-
bromo phenyl ring, shows a more positive electrostatic potential
on the aryl ring, compared with all the other inhibitors (Figure
6). Thus, the low IC50 observed for 8h can be ascribed to the
strengthening both of π-π interaction with the Trp111 side
chain on one hand and of the bromine short contact with Thr113
Oγ lone pair on the other. Moreover, docking of 8h into the
active site shows that the fluorine atom in the ortho position
would establish a polar interaction with the Tyr309 hydroxyl
group (Figure 5b).
It has to be said that, for all docking calculations, the new
version of the AutoDock program (version 4.0) was used. The
latest version of this software implements a novel force field,
which, using an improved thermodynamic model of the binding
process, allows inclusion of intramolecular terms in the esti-
mated free energy of binding. Moreover, the new force field
includes a full desolvation model that contains terms for all atom
types. Unlike the previous version, AutoDock 4 computes,
considering the desolvation energy term, the internal energy of
the unbound state of the ligand before docking, and includes
the difference of the internal energy between this unbound state
and the final docked state when estimating the free energy of
binding. Thus, all these changes are expected to result in an
improved prediction of the free energy of binding and thus of
the inhibition constant Ki associated with a given binding mode.
It is known that for a competitive inhibitor a linear correlation
between Ki and IC50 exists:
IC50 ) K 1 + S ⁄ K
(
)
m
i
where S and Km are the substrate concentration and the
Michaelis-Menten constant of the substrate, respectively.43
Because S and Km are both constant, seeing that the assays are
performed under the same conditions, a direct comparison
between the Ki values calculated by AutoDock and the
experimental IC50 is feasible. Looking at Table 5, it emerges
that, as a general rule, the estimated Ki reproduces reasonably
well the experimental inhibition data. In fact, the predicted Ki
of 7c, 7i, and 8h are in the nanomolar range, in accordance
with the experimental IC50. Also for compound 8a, both the
experimental IC50 and the predicted Ki are in the micromolar
range. Moreover, 7c, which is the most active inhibitor of the
series, has the lowest Ki. According to the IC50 values,
compound 8a, compared with 7c, 7i, and 8h, is considerably
less active and correspondingly it has a higher Ki. By contrast,
7f is the only compound that has a considerable discrepancy
between the predicted and the experimental activity. Indeed, 7f
was predicted to be less active than it really is and this could
be probably due to the peculiar electrostatic properties of the
Compounds 8a-i, which possess a methylene bridge con-
necting the phenyl and the oxadiazole rings, are generally less
potent inhibitors than their analogues 7a-i. Docking of 8a and
8h into the ALR2 active site shows that their binding mode
does not substantially diverge from that observed for 7c, 7f,
and 7i. However, due to the presence of a methylene group,
compounds 8a-i have a different shape and mostly different
electron properties with respect to compounds 7a-i. For
example, regarding 8a, the electron-donating property of the
methylene spacer results in a more negative value of the
electrostatic potential on the phenyl ring (Figure 6) compared
with compound 7a, and this negatively affects the interaction
with the electron-rich indole ring of Trp111. This finding can
be extended to all the other methylene-containing compounds,
resulting in a general lowering of their inhibitory activity. The
particularly low activity of 8a is likely to be determined also

3190 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 11
La Motta et al.
purified by recrystallization. When obtained as oily products, the
title compounds were purified by silica gel column chromatography
using petroleum ether 60-80 °C/AcOEt 9/1 as the eluent (Sup-
porting Information, Tables 3 and 4).
General Procedure for the Synthesis of 2-(3-Phenyl-1,2,4-oxadi-
azol-5-yl)acetic Acids 7a-i and 2-(3-Benzyl-1,2,4-oxadiazol-5-yl)-
acetic Acids 8a-i. A suspension of the appropriate ester derivative
5a-i or 6a-i (1.00 mmol) in 5 mL of aqueous sodium hydroxide
1N was heated under stirring at 110 °C until hydrolysis was
complete (1 h, TLC analysis). After cooling, the reaction mixture
was acidified with hydrochloric acid 1 M under ice-cooling. The
white solid separated was collected and purified by recrystallization
(Supporting Information, Tables 5 and 6).
General Procedure for the Synthesis of Ethyl 2-(2-Phenyl-
oxazol-4-yl)acetates 10a-d. A mixture of the appropriate benza-
mide (1.00 mmol) and ethyl 4-chloro-3-oxobutanoate (1.00 mmol)
was heated at 125 °C under stirring until the disappearance of the
starting material (1 h, TLC analysis). After cooling, the resulting
mass was diluted with saturated sodium bicarbonate, then extracted
with cyclohexane. The combined extracts were dried with magne-
sium sulfate, filtered, and evaporated to dryness to the desired ester,
which was purified by recrystallization. When obtained as oily
products, the title compounds were purified by silica gel column
chromatography using petroleum ether 60-80 °C/AcOEt 9/1 as
the eluent (Supporting Information, Table 7).
General Procedure for the Synthesis of 2-(2-Phenyloxa-
zol-4-yl)acetic Acids 11a-d. A suspension of the appropriate ester
10a-d (1.00 mmol) in 5 mL of aqueous sodium hydroxide 1N
was heated at reflux under stirring until hydrolysis was complete
(1 h, TLC analysis). The reaction mixture was cooled to room
temperature and then acidified with hydrochloric acid 1M under
ice-cooling. The white solid precipitated was collected, washed with
water, and purified by recrystallization (Supporting Information,
Tables 8 and 9).
Biology. Materials and Methods. Aldose reductase (ALR2) and
aldehyde reductase (ALR1) were obtained from Sprague-Dawley
albino rats, 120-140 g body weight, supplied by Harlan Nossan,
Italy. To minimize cross-contamination between ALR2 and ALR1
in the enzyme preparation, rat lens, in which ALR2 is the
predominant enzyme, and kidney, where ALR1 shows the highest
concentration, were used for the isolation of ALR2 and ALR1,
respectively.
Pyridine coenzyme, D,L-glyceraldehyde, and sodium D-glucur-
onate were from Sigma-Aldrich. Tolrestat was obtained from
Lorestat Recordati, Italy. All other chemicals were of reagent grade.
Enzyme Preparation. Aldose Reductase (ALR2). A purified
rat lens extract was prepared in accordance with the method of
Hayman and Kinoshita44 with slight modifications. Lenses were
quickly removed from rats following euthanasia and were homog-
enized (Glas-Potter) in 3 volumes of cold deionized water. The
homogenate was centrifuged at 12000 rpm at 0-4 °C for 30 min.
Saturated ammonium sulfate solution was added to the supernatant
fraction to form a 40% solution, which was stirred for 30 min at
0-4 °C and then centrifuged at 12000 rpm for 15 min. Following
this same procedure, the recovered supernatant was subsequently
fractionated with saturated ammonium sulfate solution using first
a 50% and then a 75% salt saturation. The precipitate recovered
from the 75% saturated fraction, possessing ALR2 activity, was
redissolved in 0.05 M NaCl and dialyzed overnight in 0.05 M NaCl.
The dialyzed material was used for the enzymatic assay.
Aldehyde Reductase (ALR1). Rat kidney ALR1 was prepared
in accordance with a previously reported method.33 Kidneys were
quickly removed from normal killed rats and homogenized (Glas-
Potter) in 3 volumes of 10 mM sodium phosphate buffer, pH 7.2,
containing 0.25 M sucrose, 2.0 mM EDTA dipotassium salt, and
2.5 mM ꢀ-mercaptoethanol. The homogenate was centrifuged at
12000 rpm at 0-4 °C for 30 min, and the supernatant was subjected
to a 40-75% ammonium sulfate fractionation, following the same
procedure previously described for ALR2. The precipitate obtained
from the 75% ammonium sulfate saturation, possessing ALR1
activity, was redissolved in 50 volumes of 10 mM sodium phosphate
bromine atom, which are not accurately taken into account by
the docking scoring function. However, when the Lamarckian
genetic algorithm is used as a search method, the focc (Table 5),
which for 7f is the best possible, is an important indication of
the goodness of the compound.
To sum up, through an ensemble docking approach, we
defined first the most favorable ALR2 binding conformation
for our inhibitors. The presence of an interstitial water molecule
mediating the contact with Trp111 seemed unlikely, and the
binding modes of our compounds were defined. The proposed
modes of binding of our inhibitors are perfectly in line with
the SARs here observed, and with the exception of 7f, the
predicted Ki are in line with the experimental IC50s observed.
The knowledge of the structure of the binding site together with
an understanding of the binding mode of our compounds
provides the platform for further structure-guided design
hypotheses of novel drug candidates with a higher affinity and
selectivity.
Experimental Section
Chemistry. Melting points were determined using a Reichert
Ko¨fler hot-stage apparatus and are uncorrected. Infrared spectra
were recorded with a FT-IR spectrometer Nicolet/Avatar in Nujol
mulls. Routine nuclear magnetic resonance spectra were recorded
in DMSO-d6 solution on a Varian Gemini 200 spectrometer
operating at 200 MHz. Mass spectra were obtained on a Hewlett-
Packard 5988 A spectrometer using a direct injection probe and an
electron beam energy of 70 eV. Evaporation was performed in
vacuo (rotary evaporator). Analytical TLCs were carried out on
Merck 0.2 mm precoated silica gel aluminum sheets (60 F-254).
Silica gel column chromatographies were performed with Merck
Silica gel 60 (230–400 mesh ASTM). Elemental analyses were
performed by our Analytical Laboratory and agreed with theoretical
values to within (0.4%.
Benzonitrile, 4-methylbenzonitrile, 4-methoxybenzonitrile, 4-fluo-
robenzonitrile, 4-chlorobenzonitrile, 4-bromobenzonitrile, 4-trifluo-
romethylbenzonitrile, 4-bromo-2-fluorobenzonitrile, 3-nitrobenzonitrile,
2-phenylacetonitrile, 2-p-tolylacetonitrile, 2-(4-methoxyphenyl)aceto-
nitrile, 2-(4-fluorophenyl)acetonitrile, 2-(4-bromophenyl)acetonitrile,
2-(4-chlorophenyl)acetonitrile, 2-(4-bromo-2-fluorophenyl)acetonitrile,
2-(3-nitro-phenyl)acetonitrile, benzamide, 4-methylbenzamide, 4-meth-
oxybenzamide, 4-fluorobenzamide, ethyl 3-chloro-3-oxopropanoate,
and ethyl 4-chloro-3-oxobutanoate used to obtain the target inhibitors
were from Sigma-Aldrich and Fluka.
General Procedure for the Synthesis of N-Hydroxybenzi-
midamides 3a-i and N-Hydroxy-2-phenylacetimidamides 4a-i.
A solution of the appropriate nitrile 1a-i or 2a-i (1.00 mmol),
hydroxylamine hydrochloride (1.35 mmol), and potassium carbonate
(1.00 mmol) in ethanol was left under stirring at room temperature
for 1 h, then heated under reflux until the disappearance of the
starting materials (6 h, TLC analysis). After cooling, the resulting
mixture was filtered and the solvent was evaporated to dryness under
reduced pressure to give the target compound as a white solid, which
was purified by recrystallization (Supporting Information, Tables
1 and 2).
General Procedure for the Synthesis of Ethyl 2-(3-Phenyl-1,2,4-
oxadiazol-5-yl)acetates 5a-i and Ethyl 2-(3-Benzyl-1,2,4-oxadi
azol-5-yl)acetates 6a-i. Sodium hydride (1.00 mmol, 60% disper-
sion in mineral oil) was added portionwise, under a nitrogen
atmosphere, to an ice-cooled solution of the appropriate N-
hydroxybenzimidamide 3a-i (1.00 mmol), or N-hydroxy-2-phe-
nylacetimidamide 4a-i (1.00 mmol), in dry toluene. Once addition
was complete, the mixture was left under stirring at room
temperature for 1 h. Then, ethyl 3-chloro-3-oxopropanoate (1.00
mmol) was added dropwise, under stirring and with cooling, and
the resulting suspension was heated under reflux until disappearance
of the starting materials (18-24 h, TLC analysis). After cooling,
the reaction mixture was filtered and the solvent was evaporated
to dryness under reduced pressure to the desired ester, which was

Acetic Acid Aldose Reductase Inhibitors
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 11 3191
buffer, pH 7.2, containing 2.0 mM EDTA dipotassium salt and 2.0
mM ꢀ-mercaptoethanol, and was dialyzed overnight using the same
buffer. The dialyzed material was used in the enzymatic assay.
Enzymatic Assays. The activity of the two test enzymes was
determined spectrophotometrically by monitoring the change in
absorbance at 340 nm, which is due to the oxidation of NADPH
catalyzed by ALR2 and ALR1. The change in pyridine coenzyme
concentration/min was determined using a Beckman DU-64 kinetics
software program (Solf Pack TM Module).
ALR2 activity was assayed at 30 °C in a reaction mixture
containing 0.25 mL of 10 mM D,L-glyceraldehyde, 0.25 mL of 0.104
mM NADPH, 0.25 mL of 0.1 M sodium phosphate buffer (pH 6.2),
0.1 mL of enzyme extract, and 0.15 mL of deionized water in a
total volume of 1 mL. All the above reagents, except D,L-
glyceraldehyde, were incubated at 30 °C for 10 min; the substrate
was then added to start the reaction, which was monitored for 5
min. Enzyme activity was calibrated by diluting the enzymatic
solution in order to obtain an average reaction rate of 0.011 (
0.0010 absorbance units/min for the sample.
ALR1 activity was determined at 37 °C in a reaction mixture
containing 0.25 mL of 20 mM sodium D-glucuronate, 0.25 mL of
0.12 mM NADPH, 0.25 mL of dialyzed enzymatic solution, and
0.25 mL of 0.1 M sodium phosphate buffer (pH 7.2) in a total
volume of 1 mL. The enzyme activity was calibrated by diluting
the dialyzed enzymatic solution in order to obtain an average
reaction rate of 0.015 ( 0.0010 absorbance/min for the sample.
Enzymatic Inhibition. The inhibitory activity of the newly
synthesized compounds against ALR2 and ALR1 was assayed by
adding 0.1 mL of the inhibitor solution to the reaction mixture
described above. All the inhibitors were solubilized in water and
the solubility was facilitated by adjustment to a favorable pH. After
complete solution, the pH was readjusted to 7. To correct for the
nonenzymatic oxidation of NADPH and for absorption by the
compounds tested, a reference blank containing all the above assay
components except the substrate was prepared. The inhibitory effect
of the new derivatives was routinely estimated at a concentration
of 10-4 M. Those compounds found to be active were tested at
additional concentrations between 10-5 and 10-8 M. The deter-
mination of the IC50 values was performed by linear regression
analysis of the log-dose response curve, which was generated using
at least four concentrations of the inhibitor, causing an inhibition
between 20% and 80%, with two replicates at each concentration.
The 95% confidence limits (95% CL) were calculated from t values
for n - 2, where n is the total number of determinations.
Pharmacology. Materials and Methods. Experiments were
carried out using Sprague-Dawley albino rats, 45-55 g body
weight, supplied by Harlan-Nossan, Italy. Animal care and treatment
conformed to the ARVO Resolution on the Use of Animals in
Ophthalmic and Vision Research. The galactose diet consisted of
a pulverized mixture of 50% D-galactose and 50% TRM laboratory
chow (Harlan Teckland UK), and the control diet consisted of
normal TRM. Both control and experimental rats had access to food
and water ad libitum.
opacifications in the equatorial region; stage 2, peripheral vesicles
and cortical opacities; stage 3, diffuse opacification involving the
whole lens; stage 4, matured nuclear cataract readily visible as a
white spot even to the naked eye. The number of animals that
attained each stage was recorded and the significance of differences
was analyzed using the Mann-Whitney U-test.
Computational Chemistry. Molecular Electrostatic Poten-
tial Calculations. For all compounds, the electrostatic potential
was calculated by means of GAUSSIAN 0348 and mapped onto
the electron density surface for each compound. The isovalue of
0.0004 electron/Bhor3 was chosen for the definition of the density
surface, while the electrostatic potential was computed using B3LYP
functional and 6-31G* basis set with a scale of -0.017 (red) to
0.017 hartree (blue), -10.5 and 10.5 kcal mol-1, respectively.
Docking Simulations. Molecular docking of 7c, 7f, 7g, 7i, 8a,
and 8h into the three-dimensional X-ray structures of ALR-2 (PDB
codes: 1AH0, 2FZD, 1US0, 2IKG) was carried out using the
AutoDock software package (version 4.0) as implemented through
the graphic user interface AutoDockTools (ADT 1.4.6).49,50
Ligands and Protein Setup. The structures of the inhibitors were
first generated from the standard fragment library of the SYBYL
software version 7.3.51 Geometry optimizations were achieved with
the SYBYL/MAXIMIN2 minimizer by applying the BFGS (Broy-
den, Fletcher, Goldfarb, and Shannon) algorithm52 with a conver-
gence criterion of 0.05 kcal/mol and employing the TRIPOS force
field. Then the constructed compounds (7c, 7f, 7g, 7i, 8a, and 8h)
and ALR2 X-ray structures (PDB codes: 1AH0, 2FDZ, 1US0,
2IKG) were converted to AutoDock format files using ADT. The
inhibitors were modeled in their deprotonated, negatively charged
states, and partial atomic charges on atoms were calculated with
the Gastaiger method as implemented in ADT. It has to be said
that although the inhibition assays on our compounds were
conducted on rat ALR2, we used human ALR2 crystal structure
(1US0) to explain ligand activity. This is because the three-
dimensional structure of rat ALR2 is still unknown. However,
human and rat sequences of ALR2 share 84% identity (calculated
through the Align program as implemented in the Workbench server
(http://workbench.sdsc.edu/)) and all active-site residues, including
those of the specificity pocket, are conserved between the two
enzymes. The only exception is represented by the conservative
mutation of Phe121 (human), which is a tyrosine in the rat, although
this residue was not involved in our inhibitor binding in any of the
dockings performed.
Docking Setup. The docking area was defined by a box, centered
on the CR of the Tpr111 residue. Grid points of 60 × 60 × 60
with 0.375 Å spacing were calculated around the docking area for
all the ligand atom types using AutoGrid4. For each ligand, 50
separate docking calculations were performed. Each docking
calculation consisted of 1 × 106 energy evaluations using the
Lamarckian genetic algorithm local search (GALS) method. A low-
frequency local search in accordance with the method of Solis and
Wets was applied to docking trials to ensure that the final solution
represents a local minimum. Each docking run was performed with
a population size of 150, and 300 rounds of Solis and Wets local
search were applied, with a probability of 0.06. A mutation rate of
0.02 and a crossover rate of 0.8 were used to generate new docking
trials for subsequent generations. The GALS method evaluates a
population of possible docking solutions and propagates the most
successful individuals from each generation into the next one. The
docking results from each of the 50 calculations were clustered on
the basis of root-mean square deviation (rmsd ) 1 Å) between the
Cartesian coordinates of the ligand atoms and were ranked on the
basis of the free energy of binding. For compounds 7c, 7f, and 7i,
the most highly populated family and that possessing the lowest
energy of binding coincide, and within this family, the binding mode
which presents the higher number of interactions with the enzyme
is described in the Results and Discussion Section. For compounds
8a and 8h, the family which is the most highly populated, was
considered and was described as more in agreement with SAR data
with respect to the lowest-energy family.
Prevention of Cataract Development. The assay was performed
following previously reported procedures, with slight modifica-
tions.45–47 Animals were randomly divided into four groups of equal
average body weight with 15 rats per group. The test compound
7c and the reference compound tolrestat were administered four
times daily as 3% eye-drops in both eyes. The vehicle in which
compounds were contained were administered with the same dose
regimen to the control group, which was given access to the
galactose diet, and to a group fed with a normal diet, which was
included to record the aspect of normal lenses. The groups treated
with the tested compounds were predosed one day before switching
their diet to the galactose-containing chow.
At the beginning of the experiment and after 6, 8, 10, 14, and
21 days of treatment, lenses were examined using slit-lamp
microscopy after dilating the pupils with 1% atropine, Farmigea,
Italy, to establish their integrity. The extent of lens opacity was
graded as follows, according to the classification of Kador et al.:38
stage 0, normal lens; stage 1, small vacuoles and/or tenuous radial

3192 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 11
La Motta et al.
(23) Da Settimo, F.; Primofiore, G.; Da Settimo, A.; La Motta, C.; Simorini,
F.; Novellino, E.; Greco, G.; Lavecchia, A.; Boldrini, E. Novel, Highly
Potent Aldose Reductase Inhibitors: Cyano-(2-oxo-2,3-dihydroindol-
3-yl)-acetic Acid Derivatives. J. Med. Chem. 2003, 46, 1419–1428.
(24) Da Settimo, F.; Primofiore, G.; La Motta, C.; Salerno, S.; Novellino,
E.; Greco, G.; Lavecchia, A.; Laneri, S.; Boldrini, E. Spirohydantoin
Derivatives of Thiopyrano[2,3-b]pyridin-4(4H)-one as Potent in Vitro
and in Vivo Aldose Reductase Inhibitors. Bioorg. Med. Chem. 2005,
13, 491–499.
Energy Refinement of Ligand/ALR2 Complexes. Refinement
of the predicted ligand/ALR2 Complexes was achieved through
energy minimizations using the TRIPOS force field as implemented
in the SYBYL package. These geometric optimizations included
3000 steps of a steepest descent minimization, followed by 2000
steps of conjugate gradient minimization, keeping the backbone
atoms fixed and protein side chains and the ligand free to move.
(25) Da Settimo, F.; Primofiore, G.; La Motta, C.; Sartini, S.; Taliani, S.;
Simorini, F.; Marini, A. M.; Lavecchia, A.; Novellino, E.; Boldrini,
E. Naphtho[1,2-d]isothiazole Acetic Acid Derivatives as a Novel Class
of Selective Aldose Reductase Inhibitors. J. Med. Chem. 2005, 48,
6897–6907.
Supporting Information Available: Tables 1-10 including
physical, spectral, and analytical data of compounds described. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(26) La Motta, C.; Sartini, S.; Mugnaini, L.; Simorini, F.; Taliani, S.;
Salerno, S.; Marini, A. M.; Da Settimo, F.; Lavecchia, A.; Novellino,
E.; Cantore, M.; Failli, P.; Ciuffi, M. Pyrido[1,2-a]pyrimidin-4-one
Derivatives as a Novel Class of Selective Aldose Reductase Inhibitors
Exhibiting Antioxidant Activity. J. Med. Chem. 2007, 50, 4917–4927.
(27) Zentgraf, M.; Steuber, H.; Koch, C.; La Motta, C.; Sartini, S.; Sotriffer,
C. A.; Klebe, G. How Reliable are Current Docking Approaches for
Structure-Based Drug Design. Angew. Chem., Int. Ed. 2007, 46, 3575–
3578.
(28) Steuber, H.; Zentgraf, M.; La Motta, C.; Sartini, S.; Heine, A.; Klebe,
G. Evidence for a Novel Binding Site Conformer of Aldose Reductase
in Ligand Bound State. J. Mol. Biol. 2007, 369, 186–197.
(29) Van Zandt, M. C.; Jones, M. L.; Gunn, D. E.; Geraci, L. S.; Jones,
J. H.; Sawicki, D. R.; Sredy, J.; Jacot, J. L.; Dicioccio, A. T.; Petrova,
T.; Mitschler, A.; Podjarny, A. D. Discovery of 3-[(4,5,7-trifluoroben-
zothiazol-2-yl)methyl]indole-N-acetic Acid (Lidorestat) and Congeners
as Highly Potent and Selective Inhibitors of Aldose Reductase for
Treatment of Chronic Diabetic Complications. J. Med. Chem. 2005,
48, 3141–3152.
References
(1) Kador, P. F. The Role of Aldose Reductase in the Development of
Diabetic Complications. Med. Res. ReV. 1998, 8, 325–352.
(2) Yabe-Nishimura, C. Aldose Reductase in Glucose Toxicity: a Potential
Target for the Prevention of Diabetic Complications. Pharmacol. ReV.
1998, 50, 21–33.
(3) Brownlee, M. Biochemistry and Molecular Cell Biology of Diabetic
Complications. Nature 2001, 414, 813–820.
(4) Altan, V. M. The Pharmacology of Diabetic Complications. Curr. Med.
Chem. 2003, 10, 1317–1327.
(5) Chung, S. S.; Chung, S. K. Genetic Analysis of Aldose Reductase in
Diabetic Complications. Curr. Med. Chem. 2003, 10, 1375–1387.
(6) Williamson, J. R.; Chang, K.; Frangos, M.; Hasan, K. S.; Ido, Y.;
Kawamura, T.; Nyengaard, J. R.; van der Enden, M.; Kilo, C.; Tilton,
R. G. Hyperglycemic Pseudohypoxia and Diabetic Complications.
Diabetes 1993, 42, 801–813.
(7) Ido, Y.; Williamson, J. R. Hyperglycemic Cytosolic Reductive Stress
“Pseudohypoxia”: Implications for Diabetic Retinopathy. InVest.
Ophthalmol. Vis. Sci. 1997, 38, 1467–1470.
(8) Purves, T.; Middlemas, A.; Agthon, S.; Jude, E. B.; Boulton, A. J.;
Fernyhough, P.; Tomlinson, D. R. A Role for Mitogen-Activated
Protein Kinases in the Etiology of Diabetic Neuropathy. FASEB J.
2001, 15, 2508–2514.
(9) Obrosova, I. G.; Li, F.; Abatan, O. I.; Forsell, M. A.; Komjati, K.;
Pacher, P.; Szabo, C.; Stevens, M. J. Role of Poly(ADP-ribose)poly-
merase Activation in Diabetic Neuropathy. Diabetes 2004, 53, 711–
720.
(10) Joussen, A. M.; Poulaki, V.; Le, M. L.; Koizumi, K.; Esser, C.; Janicki,
H.; Scharaermeyer, U.; Kociok, N.; Fauser, S.; Kirchhof, B.; Kern,
T. S.; Adamis, A. P. A Central Role for Inflammation in the
Pathogenesis of Diabetic Retinopathy. FASEB J. 2004, 18, 1450–1452.
(11) Sarges, R.; Oates, P. J. Aldose Reductase Inhibitors: Recent Develop-
ments. Prog. Drug Res. 1993, 40, 99–161.
(12) Suzen, S.; Buyukbingol, E. Recent Studies of Aldose Reductase
Enzyme Inhibition for Diabetic Complications. Curr. Med. Chem.
2003, 10, 1329–1352.
(30) Kraemer, O.; Hazemann, I.; Podjarny, A. D.; Klebe, G. Virtual
Screening for Inhibitors of Human Aldose Reductase. Proteins: Struct.,
Funct., Bioinf. 2004, 55, 814–823.
(31) Steuber, H.; Heine, A.; Klebe, G. Structural and Thermodynamic Study
on Aldose Reductase: Nitro-substituted Inhibitors with Strong En-
thalpic Binding Contribution. J. Mol. Biol. 2007, 368, 618–638.
(32) Faul, M. M.; Winneroski, L. L.; York, J. S.; Reinhard, M. R.; Hoying,
R. C.; Gritton, W. H.; Dominianni, S. J. Synthesis of 2-Phenyloxazole
Derivatives Containing Amino Acids as Insulin Sensitivity Enhancers
for Treatment of Type II Diabetes. Heterocycles 2001, 55, 689–704.
(33) Ward, W. H. J.; Sennitt, C. M.; Ross, H.; Dingle, A.; Timms, D.;
Mirrlees, D. J.; Tuffin, D. P. Ponalrestat: A Potent and Specific
Inhibitor of Aldose Reductase. Biochem. Pharmacol. 1990, 39, 337–
346.
(34) Negro, T.; Murata, M.; Ueda, S.; Fujitani, B.; Ono, Y.; Kuromiya,
A.; Komiya, M.; Suzuki, K.; Matsumoto, J. Novel, Highly Potent
Aldose Reductase Inhibitors: (R)-(-)-2-(4-Bromo-2-fluorobenzyl)-
1,2,3,4-tetrahydropyrrolo[1,2-a]pyrazine-4-spiro-3′-pyrrolidine-1,2′,3,5′-
tetrone (AS-3201) and its Congeners. J. Med. Chem. 1998, 41, 4118–
4129.
(35) Malamas, M. S.; Hohman, T. C.; Millen, J. Novel Spirosuccinimide
Aldose Reductase Inhibitors Derived from Isoquinoline-1,3-diones:
2-[(4-Bromo-2-fluorophenyl)methyl]-6-fluorospiro[isoquinoline-4(1H),3′-
pyrrolidine]-1,2′,3,5′(2H)-tetrone and Congeners. 1. J. Med. Chem.
1994, 37, 2043–2058.
(36) Banditelli, S.; Boldrini, E.; Vilardo, G. P.; Cecconi, I.; Cappiello, M.;
Dal Monte, M.; Marini, I.; Del Corso, A.; Mura, U. A New Approach
Against Sugar Cataract Through Aldose Reductase Inhibitors. Exp.
Eye Res. 1999, 69, 533–538.
(37) The pKa, CLogP, CLogD calculations were performed with the
chemsilico server on the web site: http: //chemsilico.com. The TPSA
calculation was performed with Marvin Sketch and Calculator Plugins
on the web-Site: http://www.chemaxon.com/demosite/marvin/index.
html.
(38) Kador, P. F.; Akagi, Y.; Kinoshita, J. H. The Effect of Aldose
Reductase and its Inhibition on Sugar Cataract Formation. Metabolism
1986, 35, 15–19.
(39) Urzhumtsev, A.; Tete-Favier, F.; Mitschler, A.; Barbanton, J.; Barth,
P.; Urzhumtseva, L.; Biellmann, J. F.; Podjarny, A.; Moras, D. A
“Specificity” Pocket Inferred from the Crystal Structures of the
Complexes of Aldose Reductase with the Pharmaceutically Important
Inhibitors Tolrestat and Sorbinil. Structure 1997, 5, 601–612.
(40) Steuber, H.; Zentgraf, M.; Gerlach, C.; Sotriffer, C. A.; Heine, A.;
Klebe, G. Expect the Unexpected or Caveat for Drug Designers:
Multiple Structure Determinations Using Aldose Reductase Crystals
Treated under Varying Soaking and Co-crystallisation Conditions. J.
Mol. Biol. 2006, 363, 174–187.
(13) Kawanishi, K.; Ueda, H.; Moriyasu, M. Aldose Reductase Inhibitors
from Nature. Curr. Med. Chem. 2003, 10, 1329–1352.
(14) Klebe, G.; Kraemer, O.; Sotriffer, C. Strategies for the Design of
Inhibitors of Aldose Reductase, an Enzyme Showing Pronounced
Induced Fit Adaptations. Cell. Mol. Life Sci. 2004, 61, 783–793.
(15) Dixit, B. L.; Balendiran, G. K.; Watowich, S. J.; Srivastava, S.;
Ramana, K. V.; Petrash, J. M.; Bhatnagar, A.; Srivastava, S. K. Kinetic
and Structural Characterization of the Glutathione-Binding Site of
Aldose Reductase. J. Biol. Chem. 2000, 275, 21587–21595.
(16) Ramana, K. V.; Dixit, B. L.; Srivastava, S.; Balendiran, G. K.;
Srivastava, S. K.; Bhatnagar, A. Selective Recognition of Glutathi-
olated Aldehydes by Aldose Reductase. Biochemistry 2000, 39, 12172–
12180.
(17) Srivastava, S. K.; Ramana, K. V.; Bhatnagar, A. Role of Aldose
Reductase and Oxidative Damage in Diabetes and the Consequent
Potential for Therapeutic Options. Endocr. ReV. 2005, 26, 380–392.
(18) Bohren, K. M.; Bullock, B.; Wermuth, B.; Gabbay, K. H. The Aldo-
Keto Reductase Superfamily. J. Biol. Chem. 1989, 264, 9547–9551.
(19) Jez, J. M.; Bennett, M. J.; Schlegel, B. P.; Lewis, M.; Penning, T. M.
Comparative Anatomy of the Aldo-Keto Reductase Superfamily.
Biochem. J. 1997, 326, 625–636.
(20) El-Kabbani, O.; Wilson, D. K.; Petrash, J. M.; Quiocho, F. A. Structural
Features of the Aldose Reductase and Aldehyde Reductase Inhibitor-
Binding Sites. Mol. Vis. 1998, 4, 19–25.
(21) Petrash, J. M. All in the Family: Aldose Reductase and Closely Related
Aldo-Keto Reductases. Cell. Mol. Life Sci. 2004, 61, 737–749.
(22) Da Settimo, F.; Primofiore, G.; Da Settimo, A.; La Motta, C.; Taliani,
S.; Simorini, F.; Novellino, E.; Greco, G.; Lavecchia, A.; Boldrini, E.
[1,2,4]Triazino[4,3-a]benzimidazole Acetic Acid Derivatives: A New
Series of Selective Aldose Reductase Inhibitors. J. Med. Chem. 2001,
44, 4359–4369.
(41) Howard, E. I.; Sanishvili, R.; Cachau, R. E.; Mitschler, A.; Chevrier,
B.; Barth, P.; Lamour, V.; Van Zandt, M.; Sibley, E.; Bon, C.; Moras,

Acetic Acid Aldose Reductase Inhibitors
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 11 3193
D.; Schneider, T. R.; Joachimiak, A.; Podjarny, A. Ultrahigh Resolu-
tion Drug Design I: Details of Interactions in Human Aldose
Reductase-Inhibitor Complex at 0.66 A. Proteins 2004, 55, 792–
804.
J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene,
M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev,
O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala,
P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.;
Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas,
O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.;
Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revision
C.02; Gaussian, Inc.: Wallingford, CT, 2004.
(42) Zentgraf, M.; Fokkens, J.; Sotriffer, C. A. Addressing Protein
Flexibility and Ligand Selectivity by “in Situ Cross-Docking”.
ChemMedChem 2006, 1, 1355–1359.
(43) Cheng, Y.; Prusoff, W. H. Relationship Between the Inhibition
Constant (Ki) and the Concentration of Inhibitor which Causes 50%
Inhibition (IC50) of an Enzymatic Reaction. Biochem. Pharmacol.
1973, 22, 3099–3108.
(44) Hayman, S.; Kinoshita, J. H. Isolation and Properties of Lens Aldose
Reductase. J. Biol. Chem. 1965, 240, 877–882.
(45) Datiles, M.; Fukui, H.; Kuwabara, T.; Kinoshita, J. H. Galactose
Cataract Prevention with Sorbinil, an Aldose Reductase Inhibitor: A
Light Microscopic Study. InVest. Ophthalm. Vis. Sci. 1982, 22, 174–
179.
(46) Simard-Duquesne, N.; Greselin, E.; Gonzales, R.; Dvornik, D.
Prevention of Cataract Development in Severely Galactosemic Rats
by the Aldose Reductase Inhibitor, Tolrestat (42048). Proc. Soc. Exp.
Biol. Med. 1985, 178, 599–605.
(47) Shizuo, A.; Chie, K.; Takaharu, O.; Yoshitada, N. Effect of Instillation
of Aldose Reductase Inhibitor FR74366 on Diabetic Cataract. InVest.
Ophthalm. Vis. Sci. 1991, 32, 3078–3083.
(48) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin,
K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone,
V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa,
(49) Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart, W. E.;
Belew, R. K.; Olson, A. J. Automated Docking Using a Lamarckian
Genetic Algorithm and Empirical Binding Free Energy Function.
J. Comput. Chem. 1998, 19, 1639–1662.
(50) Huey, R.; Morris, G. M.; Olson, A. J.; Goodsell, D. S. A Semiempirical
Free Energy Force Field with Charge-Based Desolvation. J. Comput.
Chem. 2007, 28, 1145–1152.
(51) SYBYL 7.3, Molecular Modeling System; Tripos Inc.: St. Louis, MO,
2006.
(52) Head, J.; Zerner, M. C. A Broyden-Fletcher-Goldfarb-Shannon.
Optimization Procedure for Molecular Geometries. Chem. Phys. Lett.
1985, 122, 264–274.
JM701613H