Naphtho[1,2-d]isothiazole acetic acid derivatives as a novel class of selective aldose reductase inhibitors

Article abstract of DOI:10.1021/jm050382p

Acetic acid derivatives of naphtho[1,2-d]isothiazole (NiT) were synthesized and tested as novel aldose reductase (ALR2) inhibitors. The parent compound 11 exhibited a fair inhibitory activity (IC₅₀ = 10 μM), which was enhanced by 2 orders of ma

Full text of DOI:10.1021/jm050382p

J. Med. Chem. 2005, 48, 6897-6907
6897
Naphtho[1,2-d]isothiazole Acetic Acid Derivatives as a Novel Class of Selective
Aldose Reductase Inhibitors
Federico Da Settimo,† Giampaolo Primofiore,† Concettina La Motta,*,† Stefania Sartini,† Sabrina Taliani,†
Francesca Simorini,† Anna Maria Marini,† Antonio Lavecchia,§ Ettore Novellino,§ and Enrico Boldrini‡
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, and Opocrin
S.p.A., Via Pacinotti 3, 41040 Corlo di Formigine, Modena, Italy
Received April 22, 2005
Acetic acid derivatives of naphtho[1,2-d]isothiazole (NiT) were synthesized and tested as novel
aldose reductase (ALR2) inhibitors. The parent compound 11 exhibited a fair inhibitory activity
(IC50 ) 10 µM), which was enhanced by 2 orders of magnitude by introducing a second carboxylic
group at position 4 (13 and 14: IC50 ) 0.55 and 0.14 µM, respectively). Substitution of the
acetic acid function with an apolar group gave inactive (29) or poorly active (25, 26, 30)
compounds, thus demonstrating that the 2-acetic group is involved in the enzyme pharma-
cophoric recognition while the 4-carboxylic moiety has only an accessory role. The potent
compounds 11, 13, 14, 26 all proved to be selective for ALR2, since none of them inhibited
aldehyde reductase, sorbitol dehydrogenase, or glutathione reductase. The isopropyl ester 31,
a prodrug of 14, was found to be effective in preventing cataract development in severely
galactosemic rats, when administered as an eyedrop solution. The theoretical binding mode of
13 and 14, obtained by docking simulations into the ALR2 crystal structure, was fully consistent
with the structure-activity relationships in the NiT series.
Introduction
okinase is greater than that of ALR2, glucose is pref-
erentially phosphorylated with ATP by this enzyme.
Under hyperglycemic conditions, hexokinase is rapidly
saturated and the polyol pathway becomes activated.
Sorbitol is formed more rapidly than it is converted to
fructose, and its polarity hinders its 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 result
in the development of long-term diabetic complications.
In addition to the osmotic imbalance, the polyol
pathway is the major contributor to the generation of
hyperglycemic oxidative stress. An increase in the ALR2
activity causes a substantial depletion of its cofactor
NADPH, which is also required for glutathione reduc-
tase (GR) to regenerate glutathione (GSH). Thus, a
significant decrease in the GSH level occurs and the
cellular antioxidant capacity is greatly impaired. Also,
the oxidation of sorbitol to fructose via SDH results in
a decrease in the cytosolic NAD+/NADH ratio, increas-
ing the amount of NADH that can be utilized by NADH
oxidase to generate reactive oxygen species (ROS).
Moreover, the increase in fructose levels accelerates the
development of diabetic complications, since fructose
and its metabolites are more potent nonenzymatic
glycation agents than glucose.10-13 Inhibition of ALR2
is therefore a useful therapeutic strategy to prevent the
onset or at least delay the progression and the severity
of these complications.
Diabetes mellitus is a debilitating disease that affects
over 200 million people worldwide. It can be successfully
controlled by the administration of insulin and/or potent
oral hypoglycemics, following newer therapeutic and
preventive modalities. Despite these advances in treat-
ment, it still remains the cause of significant morbidity
and mortality because of a progressive impairment of
the nervous, renal, vascular, and visual systems. Dam-
age of these tissues results from biochemical and
functional alterations, occurring in response to a pro-
longed exposure to hyperglycemia. Many hypotheses
have been proposed to explain the pathogenic mecha-
nism leading to these alterations, including impairment
of antioxidant defense, nonenzymatic glycation of pro-
teins, and activation of protein kinase C isoforms. The
prominent theory suggests that they are causally linked
to an increased flux of glucose through the polyol
pathway.1-9
Aldose reductase (alditol/NADP+ oxidoreductase, EC
1.1.1.21, ALR2) is the first enzyme of the polyol path-
way. It catalyzes the NADPH-dependent reduction of
glucose to sorbitol, which is then oxidized to fructose
by sorbitol dehydrogenase (L-iditol/NAD+, 5-oxidoreduc-
tase, EC 1.1.1.14, SDH). Because ALR2 has a low
substrate affinity to glucose, the conversion of glucose
to sorbitol is generally nonsignificant in normoglycemic
conditions. Actually, ALR2 must compete directly with
hexokinase of the glycolytic pathway for the utilization
of glucose, and because the substrate affinity of hex-
Many compounds have been shown to inhibit the
enzyme with various degrees of efficacy and specificity.
Currently known ALR2 inhibitors (ARIs) are usually
divided into two classes: (i) acetic acid derivatives, like
tolrestat, zopolrestat, and epalrestat; (ii) cyclic imides,
* To whom correspondence should be addressed. Phone: (+)39 050
2219 547. Fax: (+)39 050 2219 605. E-mail: lamotta@farm.unipi.it.
† Universita` di Pisa.
§ Universita` di Napoli “Federico II”.
‡ Opocrin S.p.A.
10.1021/jm050382p CCC: $30.25 © 2005 American Chemical Society
Published on Web 10/01/2005

6898 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 22
Da Settimo et al.
Chart 1
Figure 1. Overlay of inhibitor 11 (green-blue) on the experi-
mentally determined ALR2-bound conformation of tolrestat (by
atom).
phobic catalytic site, which determine not only the
inhibitory potency but also the selectivity of an inhibitor.
Thus, an increase in the overall lipophilicity of BiT
inhibitors might result in an improvement of the in vitro
effectiveness and, possibly, in ALR2 selectivity. For this
reason, we decided to insert a fused benzene ring into
the benzoisothiazole backbone to obtain the acetic acid
derivatives of 3,3-dioxide-1,2-dihydronaphtho[1,2-d]-
such as sorbinil (Chart 1). Despite being structurally
different, they all share common structural require-
ments, represented by a planar aromatic moiety and an
acidic function ionized at the ALR2 active site.14-18
Promising products during in vitro studies or in trials
with animal models often fail to proceed any further
because of a lack of efficacy or because of adverse side
effects mainly attributed to pharmacokinetic problems
and a lack of specificity for the target enzyme. To date,
only epalrestat is successfully marketed for treatment
of diabetic neuropathy, in Japan.
ALR2 is closely related to other enzymes, especially
aldehyde reductase (ALR1), SDH, and GR. ALR1 is a
detoxifying enzyme that belongs to the aldo-keto re-
ductase family, like ALR2. They both exhibit the highest
functional and structural homology in the family, show-
ing 51% identity in amino acid sequence between
human ALR2 and ALR1.19 SDH is the second enzyme
in the polyol pathway, and under diabetic conditions,
its inhibition, although useful to restore the altered
cytoplasmic redox state, increases sorbitol levels, thus
accelerating cataract formation.20,21 GR maintains physi-
ological levels of GSH, which in turn prevents glycosy-
lation of protein and protects against oxidative stress.
Furthermore, the intracellular accumulation of oxidized
glutathione (GSSG), resulting from inhibition of GR,
leads to a structural modification of ALR2, which greatly
reduces its catalytic efficiency and its sensitivity to ARI
action.22,23 The attention of medicinal chemistry is
therefore directed toward the discovery of more specific
and clinically effective inhibitors.
isothiazol-1-one (NiT) as
a novel class of ARIs
(Chart 1).
A thorough literature survey revealed that our com-
pounds are closely related to the dihydro-1H-benzo[e]-
isoindoles developed by Wrobel and co-workers as
conformationally rigid analogues of tolrestat.28 The
author reports that a good inhibitory potency is main-
tained when the thioamide nitrogen of tolrestat is
inserted into a rigid cyclic structure such as a thiolactam
ring of five atoms. Such a conformational constraint
greatly reduces the flexibility of the molecule, with a
beneficial effect on the biological activity.
A preliminary superimposition of NiT 11 on the
ALR2-bound conformation of tolrestat shows a satisfac-
tory overlay of the common acetic acid chains and
naphthalene rings and of the sulfonyl/thiocarbonyl
groups (Figure 1). The match of the fused benzene ring
of NiT and the methoxy substituent in position 6 of
tolrestat suggests that these moieties might occupy the
same lipophilic domain in the enzyme active site. The
crystal structure of ALR2 complexed with tolrestat
clearly shows that the inhibitor binds at the catalytic
site and induces significant conformational changes in
a loop (residues 121-135) and a short segment (residues
298-303) to open and fill a contiguous hydrophobic
pocket. This so-called “specificity pocket”, which is closed
in the absence of the inhibitor, is a key determinant for
selectivity.16,29 Thus, our products, with their good
superimposition on tolrestat, could be predicted to be
more effective against ALR2 than against ALR1.
Moreover, the carbonyl group in position 1 of NiT
might theoretically form additional hydrogen bonds with
proper amino acid side chains, thus enhancing the
inhibitory activity through a better anchoring at the
ALR2 catalytic site.
Our well-documented interest in the field of carboxylic
acid inhibitors24,25 led us to develop a series of 1,1-
dioxide-benzo[d]isothiazol-3-one alkanoic acids (BiT) as
a novel class of ARIs (Chart 1). These compounds proved
to be relatively potent inhibitors, with IC50 values in
the micromolar range. The 4-NO2 acetic acid derivative
is the most active one, with an IC50 of 5.5 µM.26,27
It is well-known that the in vitro ALR2 inhibitory
efficacy of a compound is a function both of its acidic
group ionized at the anion binding site and of its
lipophilic scaffold interacting with the enzyme hydro-
These findings prompted us to synthesize and test for
their ALR2 inhibitory activity a number of derivatives
in which structural modifications were made in posi-
tions 2, 4, 6, and 9 of the naphtho[1,2-d]isothiazole

Isothiazole Acetic Acid Derivatives
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 22 6899
Scheme 1
scaffold. To define the structure-activity relationships
(SARs) of this class of inhibitors clearly, the acetic acid
derivative of linear 3,3-dioxide-1,2-dihydronaphtho[2,3-
d]isothiazol-1-one 12 was also likewise prepared and
tested.
Compound 14, the most active of the series, and its
diisopropyl ester 31 were also investigated in vivo for
their ability to prevent cataract development in galac-
tosemic rats.
Finally, docking simulations of 13 and 14 into the
human ALR2 binding site were carried out to rationalize
the SARs observed and to guide, prospectively, the
design of new analogues.
Following the same conditions as the lithiation-
carbonation reaction in our laboratory, a mixture of
three products was obtained, which were separated by
flash chromatography. The pure compounds were iden-
tified by 1H NMR spectral data, proving that the acid 2
was obtained with its regioisomer 3 and with the 1,3-
dicarboxylic compound 4 in a 30/10/60 ratio, respec-
tively. Treatment of 2 and 3 with polyphosphoric acid
resulted in the corresponding naphthoisothiazoles 5 and
6. The same reaction, conducted on 4, gave the 1,3,3-
trioxo-1,2-dihydronaphtho[1,2-d]isothiazole-4-carboxy-
lic acid 7; the carboxylic group in position 1 of compound
4 is the only acidic function involved in the reaction of
cyclization. Actually, the 1H NMR spectrum of 7 shows
the downfield doublet of the deshielded peri proton H9,
which is present in 5 but not in 6.
Chemistry
The title compounds were obtained as outlined in
Schemes 1-4. The synthesis of the naphthoisothiazole
nucleus has previously been described by J. G. Lom-
bardino.30 The author reported that ortho lithiation,
carried out with n-butyllithium, and subsequent car-
bonation with CO2 of 2-(N-tert-butyl)naphthalenesulfo-
namide, 1, gave the carboxylic acid 2, which was cyclized
and dealkylated in one step by polyphosphoric acid at
100 °C to the desired product 5.
Each compound 5-7 represents a suitable scaffold for
the design and the synthesis of novel potential ALR2
inhibitors, introducing an acetic acid chain (which is
essential for the inhibition of the enzyme) in position 2
of the NiT nucleus. Furthermore, derivative 7 possesses
an additional acid function in position 4 of the naph-
thoisothiazole backbone, which might either enhance

6900 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 22
Da Settimo et al.
Scheme 2
lyzed with hot concentrated HCl to provide the target
products 29 and 30.
Finally, to improve the ocular bioavailability of acid
14, the more liphophilic isopropyl diester32 31 was
prepared and tested.
Results and Discussion
Biological Evaluation. Table 1 lists the inhibitory
activities against ALR2 of the tested compounds, ex-
pressed as IC50 values. As stated in the Introduction,
the lead compound 11 showed an appreciable ALR2
inhibitory potency in the micromolar range (10 µM). At
first, a new analogue investigation was directed toward
the study of the influence on inhibitory potency of
substituents with different electronic properties on the
benzofused ring in the 6 or 9 positions. Both electron-
withdrawing (NO2) and electron-donating (NH2) groups
determined a detrimental effect on the inhibitory po-
tency with respect to the lead compound 11 (see
compounds 19-22); consequently, this type of substitu-
tion was no longer investigated.
Because the ALR2 catalytic site is rich in amino acids
bearing side chains able to donate or accept hydrogen
bonds,16,29 in principle a second carboxylic group on NiT
inhibitors could reinforce the enzyme/inhibitor complex
through additional H-bonds.33 Thus, compound 13 was
prepared in which an additional acid group was intro-
duced in position 4 because of its synthetic feasibility.
This modification in the molecular structure of 11 gave
an 18-fold enhancement in potency (13 IC50 ) 0.55 µM),
thus confirming our design hypothesis. This hypothesis
was further confirmed by compound 14, in which the
distance between the additional carboxylic moiety and
the molecular nucleus of the inhibitor was shortened,
as it showed a gain in inhibitory potency of 2 orders of
magnitude with respect to the lead 11 (14 IC50 ) 0.14
µM).
These results prompted us to verify which of the two
carboxylic groups of compounds 13 and 14 was primarily
involved in the recognition of the enzyme anionic
binding site. In products 25 and 26 and products 29 and
30, the two different types of carboxylic moieties of 14
and 13, respectively, were maintained in position 4,
whereas the acetic moiety in position 2 was substituted
by apolar groups with a different steric hindrance, such
as methyl and benzyl. The potencies lower than those
of 14 and 13 shown by products 25 and 26 and products
29 and 30 demonstrated that the 2-acetic group is
involved in the enzyme pharmacophoric recognition of
the inhibitor and that the 4-carboxylic moiety has only
an accessory role.
the inhibitory activity or be used to introduce appropri-
ate substituents on the nucleus.
When the sodium salts of 5-7 were heated with ethyl
bromoacetate in DMF at 100 °C, the ethyl esters 8-10
were obtained. Hydrolysis of the isomers 8 and 9 with
hot concentrated HCl provided the target acids 11 and
12. Treatment of 10 with concentrated HCl achieved
selective hydrolysis of the ethyl ester groups, and the
diacid 13 was the only product of the reaction. Under
basic hydrolytic conditions in a water/dioxane solution,31
the cleavage of all the ester functions was completed
and the acid 14 was obtained with a good yield
(Scheme 1).
Substitution of positions 6 and 9 of the NiT skeleton
with either an electron-withdrawing group, i.e., nitro,
or an electron-donating group, i.e., amino, was then
performed to verify which type of substituent the fused
benzene ring can tolerate (Scheme 2). Thus, by the
action of fuming HNO3 on the ester derivative 8, the
two regioisomers 15 and 16 were obtained, which were
separated by flash chromatography and characterized
1
by H NMR spectral data. Catalytic hydrogenation of
the nitro compounds gave the amino esters 17 and 18.
Finally, hydrolysis of 15-18 with hot concentrated HCl
gave the desired compounds 19-22.
Finally, the shift of the junction between naphthalene
and isothiazole from 1,2 to 2,3 to give the linear
derivative 12 was totally detrimental for activity.
Evidently, only the angular products feature the best
spatial relationship among the pharmacophoric groups.
The selected compounds 11, 13, 14, and 26, when
assayed for their ability to inhibit other ALR2-related
enzymes, ALR1, SDH, and GR, showed no appreciable
inhibitory property, proving them to be completely
selective inhibitors of ALR2 (Table 2).
To test the effective role of the acid residues of 13 and
14 in inhibitor binding and activity, derivatives 25-26
and 29-30 were also prepared and tested. In these
compounds, the N-acetic acid moiety was replaced with
either a methyl or a benzyl group to verify the effects
resulting from changes in residue size and lipophilicity
(Scheme 3). Thus, treating the sodium salt of 7 with the
appropriate alkyl halide gave the ester derivatives 23
and 24, which were hydrolyzed in alkaline medium to
the corresponding carboxylic acids 25 and 26. These
compounds were then alkylated with ethyl bromoac-
etate, and the resulting esters 27 and 28 were hydro-
Pharmacological Evaluation of Compounds 14
and 31. Compound 14 was administered as an eyedrop
solution in the precorneal region to investigate its in

Isothiazole Acetic Acid Derivatives
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 22 6901
Scheme 3
Scheme 4
50% galactose diet, 90% of the animals treated only with
the vehicle developed nuclear cataract. No protection
was observed in those treated with 3% ophthalmic
solution of acid 14, whereas rats treated with 3%
solution of the isopropyl ester 31 and 1% solution of
tolrestat exhibited 55% and 53% protection, respec-
tively. No nuclear cataract was developed by animals
administered 3% ophthalmic solution of tolrestat. The
potency displayed by prodrug 31 with respect to acid
14 suggests that its higher lipophilicity is essential to
permeate the corneal tissue and that its inhibitory
activity is ensured by a quick hydrolysis to the active
acid by corneal esterases.
Table 1. ALR2 Inhibition Data of Acid Derivatives 11-14,
19-22, 25, 26, 29, 30
Molecular Modeling. To better understand the
fairly good ARI inhibitory potency of both compounds
13 and 14 at the molecular level, experiments on
docking into the binding pocket of the porcine ALR2/
NADP+/tolrestat complex (PDB entry code 1AH3) were
performed. Docking experiments were carried out using
the automated docking program AutoDock,38 which
allows torsional flexibility in the ligand and incorporates
an efficient Lamarckian genetic search algorithm to-
gether with an empirical free energy function.
Although the inhibition assays on our compounds
were conducted on rat ALR2, the use of the crystal
structure of porcine ALR2 for docking studies is justified
by the following facts: (i) the crystal structure of rat
ALR2 is unknown; (ii) the porcine and rat sequences of
this enzyme are characterized by 81% identity and 87%
homology;39 (iii) all active-site residues, including those
of the specificity pocket, are largely conserved across
the ALR2 isoforms so far sequenced.
The energy-minimized structure of tolrestat was
preliminarily docked into ALR2 to examine how closely
the AutoDock algorithm can reproduce the experimen-
tally determined binding conformation of tolrestat at the
active site of ALR2. The docking test indicated that the
largest cluster of similar conformations with the lowest-
energy docked structure reproduced very closely the
crystallographic binding mode of tolrestat to ALR2. The
hydrogen bonds predicted by AutoDock were virtually
identical to those found in the crystal structure.
Docking of 13 and 14 to the ALR2 crystal structure
revealed a very clear preference for a single position in
the active site; the results with the best binding energy
(-8.9 and -8.4 kcal/mol) were found 33 and 39 times
out of 50, respectively. The ligands were found to be in
compd
R1
R2
R3 R4
IC50 a (µM)
11
CH2COOH H
H
H
10 (8.32-11.96)
nab
0.55 (0.49-0.65)
0.14 (0.102-0.155)
190 (159.8-221.3)
12
13
CH2COOH COOCH2COOH H
H
H
H
14
CH2COOH COOH
CH2COOH H
CH2COOH H
CH2COOH H
CH2COOH H
CH3
CH2C6H5 COOH
CH3
H
NO2
H
NH2
H
H
H
19
20
NO2 180 (156.9-209.9)
100 (87.5-114.7)
NH2 100 (86.7-111.8)
21
H
22
25
COOH
H
H
H
H
88.0 (72.6-101.2)
68.0 (51.5-77.6)
nab
100 (91.1-109.3)
0.05 (0.03-0.06)
26
29
COOCH2COOH H
30
CH2C6H5 COOCH2COOH H
tolrestat
a IC50 (95% CL) values represent the concentration required to
produce 50% enzyme inhibition. b na ) not active.
vivo ability to prevent cataract development in severely
galactosemic rats. Topical administration can in prin-
ciple achieve significant drug levels in the lens with
negligible effects on other tissues, thus avoiding bio-
availability and/or metabolism-related problems associ-
ated with systemic administration.34 It has been
reported35-37 that esters of acid drugs display better
corneal permeability and ocular bioavailability by virtue
of their higher lipophilicity. Moreover, the employment
of an ocular prodrug, which is not active by itself, has
the general advantage of being metabolized to its active
form only after permeation through the corneal tissue,
thus acting only where necessary. Considering that
isopropyl esters have shown a good permeability in the
corneal tissue,35 both carboxylic groups of compound 14
were transformed into their isopropyl esters to give 31,
which was tested in vivo as a prodrug. The pharmaco-
logical data are reported in Table 3. After 21 days of

6902 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 22
Da Settimo et al.
Table 2. Enzyme Inhibition Data of Selected Compounds 11, 13, 14, and 26
aldose reductase
aldehyde reductase
sorbitol dehydrogenase
glutathione reductase
compd
IC50 a (µM)
IC50 a (µM)
IC50 a (µM)
IC50 a (µM)
11
10 (8.32-11.96)
0.55 (0.49-0.65)
0.14 (0.102-0.155)
68.0 (51.5-77.6)
0.65 (0.49-0.82)
0.05 (0.03-0.06)
7.81 (5.47-10.15)
nab
nab
nab
13
nab
nab
nab
14
26
sorbinil
tolrestat
quercetin
nab
nab
nab
nab
nab
nab
0.029 (0.020-0.038)
nab
nab
nab
nab
nab
2.32 (2.05-2.78)
35.6 (31.6-40.0)
48.8 (34.7-63.4)
a IC50 values represent the concentration required to produce 50% enzyme inhibition. b na ) not active.
Table 3. Effects of Treatment with Ophthalmic Solution of 14,
31, and Tolrestat on Development of Nuclear Cataract in
Severely Galactosemic Rats
rats with nuclear cataract, %
day of
treatment
14
(3%)
31
(3%)
tolrestat
(1%)
tolrestat
(3%)
control
11
12
13
14
15
16
17
18
19
20
21
13
25
25
25
25
31
50
50
75
88
90
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
13
15
25
25
25
25
50
85
90
0
0
0
0
23
32
32
32
43
43
47
47
10
10
28
35
45
45
the same location as tolrestat in the crystal structure.
Figure 2 illustrates a selection of residues (4 Å) inter-
acting with the inhibitors in the structure of the
minimized complex.
As previously observed in crystal structures of ALR2
complexed with carboxylic acid inhibitors,15,16,39-41 both
analyzed molecules were bound in the active site with
their acetic acid moiety entering the anion binding site.
Both carboxylic oxygen atoms are within hydrogen-
bonding distance of the imidazolium Nꢀ2 atom of His110.
Further, the Oη hydrogen of Tyr48 and the N ꢀ1 hydrogen
of Trp111 form two hydrogen bonds with alternate
carboxylic oxygen atoms. The acetic acid substituent is
also held in place by numerous van der Waals interac-
tions with the nicotinamide ring of NADP+. Thus, the
carboxylate group in position 2 is firmly anchored in the
active site, suggesting that it is an essential requirement
for the inhibitory activity of these series of compounds.
Intriguingly, the couples of compounds 25/26 and 29/
30, in which the acetic substituent is replaced by a
methyl and a benzyl group, respectively, are completely
inactive evidently because they lack the acidic substitu-
ent able to interact with Tyr48, His110, and Trp111,
which are three key residues in binding and cataly-
sis.42,43
Figure 2. Docked structure of (a) inhibitor 13 (green-blue)
and (b) inhibitor 14 (cyan) in the ALR2 active site. Only amino
acids located within 4 Å distance of the bound ligand are
displayed and labeled. Nonpolar hydrogens have been removed
for clarity. NADP+ is shown in magenta. Hydrogen bonds are
represented by dashed yellow lines.
ing the slight difference in activity between 13 (IC50
0.55 µM) and 14 (IC50 ) 0.14 µM).
)
Moreover, inspection of the 14/ALR2 complex indi-
In addition, the carboxylate group and the carbonyl
oxygen of the ester function in the 4 position of 13 are
engaged in hydrogen-bonding interactions with both
Trp20 and Trp219 Nꢀ1 hydrogens, respectively. Interest-
ingly, the carboxylate group in the 4 position of inhibitor
cates that the carbonyl oxygen in the 1 position of the
ꢀ1
isothiazole ring forms a hydrogen bond with the N
hydrogen of Trp111, while one of the two sulfonyl
oxygens is positioned to form a weak hydrogen bond
ꢀ1
ꢀ1
14 is also hydrogen-bonded with the N hydrogen of
with the N of Trp20, thus contributing to a further
Trp219, thus providing further evidence of the impor-
tance of these additional acidic groups for inhibitor
binding to ALR2.
The hydrophobic naphthalene system of both inhibi-
tors gets trapped tightly in a hydrophobic pocket formed
stabilization of the complex. By contrast, in the case of
inhibitor 13, it is noted that neither the carbonyl oxygen
in the 1 position nor the two sulfonyl oxygens are
involved in hydrogen-bonding interactions with Trp111
N
ꢀ1 or Trp20 Nꢀ1 hydrogens, respectively, thus explain-

Isothiazole Acetic Acid Derivatives
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 22 6903
8.63 (m, 6H, ArH), 13.20 (bs 1H, COOH, exc). MS m/e (%):
by Trp79, Trp111, Phe115, Phe122, Cys303, and Trp219,
making a T-shaped interaction with the Phe122 side
chain. The finding that 13 and 14 fit into this pocket,
designed as a specificity pocket, has important conse-
quences for the selectivity of inhibitors for ALR2 with
respect to ALR1. In fact, these two enzymes differ
mainly in the residues composing the hydrophobic
pocket, the rest of the enzyme being largely con-
served.16,44 Inhibitors able to bind in this pocket have
been reported to be selective for ALR2.16 The finding
that 13 and 14 bind in the hydrophobic pocket accounts
for the selectivity observed for these inhibitors.
From the docking model illustrated in Figure 3, it may
also be deduced that the low ALR2 inhibitory activity
of the 6- or 9-substituted derivatives (compounds 19-
22) is a consequence of steric clashes with the walls of
the enzyme hydrophobic cleft.
To understand the ALR2 inactivity of 12, this com-
pound was docked into the ALR2 catalytic site. Au-
toDock predicted a binding mode associated with a
scarcely populated cluster (8/50) and a high estimated
∆Gbind (-7,32 kcal/mol). From a visual inspection of the
12/ALR2 complex, it seems clear that the linear struc-
ture of the ligand prevents it from adapting properly
into the ALR2 active site, thus not allowing the acetic
residue in the 2 position of the ligand to interact
favorably with the oxyanion hole made up of His110,
Trp111, and Tyr48 residues. This may be the reason
for the loss of ALR2 inhibitory activity of 12.
292 (100). Anal. (C15H17NO4S) C, H, N.
4: mp >300 °C (AcOEt); Rf ) 0.096; yield 60%. IR ν cm-1
:
1
3446, 3208, 3160, 1711. H NMR δ ppm: 1.18 (s, 9H, CH3),
6.81-8.43 (m, 5H, ArH), 14.00 (bs, 1H, COOH, exc). MS m/e
(%): [M+] 351 (14), 278 (100). Anal. (C16H17NO6S) C, H, N.
3,3-Dioxidenaphtho[1,2-d]isothiazole-1(2H)-one 5, 1,1-
Dioxidenaphtho[2,3-d]isothiazole-3(2H)-one 6, and 1,3,3-
Trioxo-1,2-dihydronaphtho[1,2-d]isothiazole-4-carbox-
ylic Acid 7. A suspension of carboxylic acid 2, 3, or 4 (1.00
mmol) in polyphosphoric acid (2.00 g) was heated at 100 °C
for 15 min while being stirred with a glass stick. The resulting
thick syrup was slowly poured into crushed ice, and the
resulting white solid was collected and purified by recrystal-
lization from the appropriate solvent.
5: mp 245-248 °C (AcOEt); yield 30%. IR ν cm-1: 3303,
1736, 1705. 1H NMR δ ppm: 7.83 (t, 1H, H7), 7.89 (t, 1H, H8),
8.12 (d, 1H, H5), 8.24 (d, 1H, H6), 8.57 (d, 1H, H4), 9.10 (d, 1H,
H9). MS m/e (%): [M+] 233 (63), 126 (100). Anal. (C11H7NO3S)
C, H, N.
6: mp >300 °C (iPrOH); yield 30%. IR ν cm-1: 3207, 1733,
1
1700. H NMR δ ppm: 7.84 (t, 2H, H6, H7), 8.28 (t, 1H, H8),
8.31 (s, 1H, NH, exc), 8.33 (t, 1H,H5), 8.70 (s, 1H, H9), 8.83 (s,
1H, H4). MS m/e (%): [M+] 233 (82), 126 (100). Anal. (C11H7-
NO3S) C, H, N.
7: mp >300 °C (AcOEt); yield 85%. IR ν cm-1: 3407, 3088,
1
1717, 1697. H NMR δ ppm: 7.90-7.99 (m, 2H, H7, H8), 7.95
(s, 1H, NH, exc), 8.42 (d, 1H, H6), 9.12 (s, 1H, H5), 9.28 (d, 1H,
H9), 14.02 (bs, 1H, COOH, exc). MS m/e (%): [M+] 277 (89),
114 (100). Anal. (C12H7NO5S) C, H, N.
(1,3,3-Trioxo-1,2-dihydronaphtho[1,2-d]isothiazol-2-
yl)acetic Acid Ethyl Ester 8, (1,3,3-Trioxo-2,3-dihy-
dronaphtho[2,3-d]isothiazol-2-yl)acetic Acid Ethyl Ester
9, and 2-Ethoxycarbonylmethyl-1,3,3-trioxo-1,2-dihy-
dronaphtho[1,2-d]isothiazole-4-carboxylic Acid Ethoxy-
carbonylmethyl Ester 10. The naphthoisothiazole derivative
5, 6, or 7 (1.00 mmol) was treated with sodium ethoxide (1.00
mmol for 5 and 6; 2.00 mmol for 7) in ethanol solution and
then evaporated to dryness. The sodium salt obtained was
suspended in 5 mL of DMF, ethyl bromoacetate (1.00 mmol
for 5 and 6; 2.00 mmol for 7) was added, and the mixture was
heated at 120 °C until the disappearance of the starting
material (4-8 h, TLC analysis). The reaction mixture was then
poured into crushed ice, and the separated solid was collected
and purified by recrystallization (Supporting Information,
Tables 1 and 3).
General Procedure for the Synthesis of (1,3,3-Trioxo-
1,2-dihydronaphtho[1,2-d]isothiazol-2-yl)acetic Acid 11,
(1,3,3-Trioxo-2,3-dihydronaphtho[2,3-d]isothiazol-2-yl)-
acetic Acid 12, 2-Carboxymethyl-1,3,3-trioxo-1,2-dihy-
dronaphtho[1,2-d]isothiazole-4-carboxylic Acid Car-
boxymethyl Ester 13, (9-Nitro-1,3,3-trioxo-1,2-dihydro-
naphtho[1,2-d]isothiazol-2-yl)acetic Acid 19, (6-Nitro-
1,3,3-trioxo-1,2-dihydronaphtho[1,2-d]isothiazol-2-yl)-
acetic Acid 20, (9-Amino-1,3,3-trioxo-1,2-dihydronaphtho-
[1,2-d]isothiazol-2-yl)acetic Acid 21, (6-Amino-1,3,3-trioxo-
1,2-dihydronaphtho[1,2-d]isothiazol-2-yl)acetic Acid 22,
2-Methyl-1,3,3-trioxo-1,2-dihydronaphtho[1,2-d]isothia-
zole-4-carboxylic Acid 29, and 2-Benzyl-1,3,3-trioxo-1,2-
dihydronaphtho[1,2-d]isothiazole-4-carboxylic Acid 30.
The appropriate ethyl ester 8-10, 15-18, 27, 28 (1.00 mmol)
was suspended in 3 mL of concentrated hydrochloric acid and
heated under stirring at 100 °C until hydrolysis was complete
(3-12 h, TLC analysis). After cooling at room temperature,
the reaction mixture was diluted with water and the precipi-
tated target acid 11-13, 19-22, 29, or 30 was collected and
purified by recrystallization (Supporting Information, Tables
2 and 4).
Experimental Section
Chemistry. Melting points were determined using a Re-
ichert Ko¨fler hot-stage apparatus and are uncorrected. Infra-
red 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 evapora-
tor). Analytical TLC was carried out on Merck 0.2 mm
precoated silica gel aluminum sheets (60 F-254). Silica gel 60
(230-400 mesh) was used for flash chromatography. Elemen-
tal analyses were performed by our Analytical Laboratory and
agreed with theoretical values to within (0.4%. The 2-(N-tert-
butyl)naphthalenesulfonamide 1 was prepared in accordance
with a reported procedure.30
2-tert-Butylsulfamoylnaphthalene-1-carboxylic Acid
2, 3-tert-Butylsulfamoylnaphthalene-2-carboxylic Acid
3, and 2-tert-Butylsulfamoylnaphthalene-1,3-dicarbox-
ylic Acid, 4. n-Butyllithium (1.6 M solution in hexane, 7.12
mL, 11.40 mmol) was added dropwise to an ice-cooled solution
of 2-(N-tert-butyl)naphthalenesulfonamide 1 (1.00 g, 3.80
mmol) in 20 mL of anhydrous THF under a dry nitrogen
atmosphere. Once addition was complete, the reaction mixture
was left under stirring at room temperature for 2 h, then
poured into a suspension of solid carbon dioxide in anhydrous
ether. The resulting mixture was diluted with water and
acidified with concentrated hydrochloric acid under ice-cooling,
then extracted with ether. The combined ether extracts were
dried, filtered, and evaporated. The resulting crude product
was flash-chromatographed (eluting system: dichloromethane/
acetic acid ) 9/1) to provide the target compounds 2-4, which
were purified by recrystallization.
(9-Nitro-1,3,3-trioxo-1,2-dihydronaphtho[1,2-d]isothia-
zol-2-yl)acetic Acid Ethyl Ester 15 and (6-Nitro-1,3,3-
trioxo-1,2-dihydronaphtho[1,2-d]isothiazol-2-yl)acetic
Acid Ethyl Ester 16. Fuming nitric acid (0.40 mL, 10 mmol)
was added dropwise to the ice-cooled ethyl ester 8 (0.319 g,
1.00 mmol). Once addition was complete, the reaction mixture
was left under stirring at room temperature for 15 min, then
2: mp 200-203 °C (AcOEt); Rf ) 0.42; yield 30%. IR ν cm-1
:
1
3201, 1736, 1716. H NMR δ ppm: 1.10 (s, 9H, CH3), 7.69-
8.17 (m, 6H, ArH), 13.70 (bs, 1H, COOH, exc). MS m/e (%):
292 (100). Anal. (C15H17NO4S) C, H, N.
3: mp 186-190 °C (CHCl3); Rf ) 0.65; yield 10%. IR ν cm-1
:
1
3257, 3073, 1721. H NMR δ ppm: 1.16 (s, 9H, CH3), 7.00-

6904 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 22
Da Settimo et al.
poured onto crushed ice. The separated yellow solid was
collected, dried, and flash-chromatographed (eluting system,
ethyl acetate/petroleum ether 60-80 °C ) 5/5) to obtain the
target isomers 15 and 16, which were purified by recrystalli-
zation (Supporting Information, Tables 1 and 3).
(9-Amino-1,3,3-trioxo-1,2-dihydronaphtho[1,2-d]isothi-
azol-2-yl)acetic Acid Ethyl Ester 17 and (6-Amino-1,3,3-
trioxo-1,2-dihydronaphtho[1,2-d]isothiazol-2-yl)acetic
Acid Ethyl Ester 18. A suspension of the nitro derivative 15
or 16 (0.364 g, 1.00 mmol) and 10% palladium on carbon (0.10
mmol) in 100 mL of absolute ethanol was hydrogenated at
atmospheric pressure and room temperature until the uptake
of hydrogen was achieved. After filtration of the catalyst, the
solvent was evaporated to dryness to give the target compound
17 or 18 as a yellow solid, which was purified by recrystalli-
zation (Supporting Information, Tables 1 and 3).
2-Methyl-1,3,3-trioxo-1,2-dihydronaphtho[1,2-d]isothi-
azole-4-carboxylic Acid Methyl Ester 23 and 2-Benzyl-
1,3,3-trioxo-1,2-dihydronaphtho[1,2-d]isothiazole-4-car-
boxylic Acid Benzyl Ester 24. The naphthoisothiazole
derivative 7 (0.277 g, 1.00 mmol) was treated with 2 equiv of
sodium ethoxide in ethanol solution and then evaporated to
dryness. The sodium salt obtained was suspended in 5 mL of
DMF. Methyl iodide or benzyl bromide (1.20 mmol) was added,
and the mixture was heated at 120 °C until the disappearance
of the starting material (18-24 h, TLC analysis). The reaction
mixture was then poured into crushed ice, and the separated
solid was collected and purified by recrystallization (Support-
ing Information, Tables 1 and 3).
General Procedure for the Synthesis of 2-Carboxy-
methyl-1,3,3-trioxo-1,2-dihydronaphtho[1,2-d]isothi-
azole-4-carboxylic Acid 14, 2-Methyl-1,3,3-trioxo-1,2-di-
hydronaphtho[1,2-d]isothiazole-4-carboxylic Acid 25,
and 2-Benzyl-1,3,3-trioxo-1,2-dihydronaphtho[1,2-d]isothi-
azole-4-carboxylic Acid 26. A suspension of the ester
derivative 10, 23, or 24 (1.00 mmol) in a mixture of dioxane
(20 mL) and 5% NaOH solution (20 mL) was left under stirring
at room temperature until the hydrolysis was complete (16-
18 h, TLC analysis). The reaction mixture was concentrated
in vacuo to half its volume and was then acidified with
concentrated hydrochloric acid under ice-cooling. The white
solid separated was collected and purified by recrystallization
(Supporting Information, Tables 2 and 4).
(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.
Glutathione reductase (GR) type IV from baker’s yeast
(100-300 U/mg), sorbitol dehydrogenase (SDH) from sheep
liver (10 U/mg of protein), pyridine coenzymes, D,L-glyceral-
dehyde, glutathione disulfide, sodium D-glucuronate, sorbitol,
and quercetin were from Sigma Chemical Co. Sorbinil was a
gift from Pfizer, Groton, CT. 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 Kinoshita45 with slight modifications.
Lenses were quickly removed from rats following euthanasia
and were homogenized (Glas-Potter) in 3 volumes of cold
deionized water. The homogenate was centrifuged at 12 000
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 12 000 rpm for 15 min. Following this same
procedure, the recovered supernatant was subsequently frac-
tionated with saturated ammonium sulfate solution using first
a 50% and then a 75% salt saturation. The precipitate that
was 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.46
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 â-mercaptoethanol. The
homogenate was centrifuged at 12 000 rpm at 0-4 °C for 30
min, and the supernatant was subjected to a 40-75% am-
monium 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 buffer, pH 7.2, containing 2.0 mM EDTA dipotas-
sium salt and 2.0 mM â-mercaptoethanol and was dialyzed
overnight using the same buffer. The dialyzed material was
used in the enzymatic assay.
Enzymatic Assays. The activity of the four test enzymes
was determined spectrophotometrically by monitoring the
change in absorbance at 340 nm, which is due to the oxidation
of NADPH or the reduction of NAD+ catalyzed by ALR2, ALR1,
and GR or SDH, respectively. The change in pyridine coenzyme
concentration per minute 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.
2-Methyl-1,3,3-trioxo-1,2-dihydronaphtho[1,2-d]isothi-
azole-4-carboxylic Acid Ethoxycarbonyl Methyl Ester 27
and 2-Benzyl-1,3,3-trioxo-1,2-dihydronaphtho[1,2-d]isothi-
azole-4-carboxylic Acid Ethoxycarbonyl Methyl Ester
28. A suspension of the appropriate carboxylic acid 25 or 26
(1.00 mmol), ethyl bromoacetate (0.11 mL, 1.00 mmol), and
potassium fluoride (0.058 g, 1.00 mmol) in 5 mL of DMF was
heated at 120 °C until the disappearance of the starting
material (1-2 h, TLC analysis). The reaction mixture was then
poured into crushed ice, and the solid separated was collected
and purified by recrystallization (Supporting Information,
Tables 1 and 3).
2-Isopropoxycarbonylmethyl-1,3,3-trioxo-1,2-dihy-
dronaphtho[1,2-d]isothiazole-4-carboxylic Acid Isopro-
pyl Ester 31. The carboxylic acid 14 (0.335 g, 1.00 mmol) was
added to an ice-cooled solution of thionyl chloride (1.18 mL,
2.50 mmol) in 10 mL of isopropyl alcohol, and the reaction
mixture was heated under reflux for 12 h. After cooling, the
volatiles were removed in vacuo and the crude product thus
obtained was purified by recrystallization from ethanol. Mp
132-134 °C; yield 84%. IR ν cm-1: 1726, 1608, 1495, 1220,
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 units/min
for the sample.
1
1120. H NMR δ ppm: 1.26 (d, 6H, CH3), 1.32 (d, 6H, CH3),
3.78 (s, 2H, CH2), 5.02 (m, 1H, CH), 5.20 (m, 1H, CH), 7.60-
7.67 (m, 2H, H7, H8), 8.03 (d, 1H, H6), 7.93 (s, 1H, H5), 8.14 (d,
1H, H9). MS m/e (%): [M+] 419 (14), 143 (100). Anal. (C20H21-
NO7S) C, H, N.
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 be-
tween ALR2 and ALR1 in the enzyme preparation, rat lens
SDH activity47 was determined at 37 °C in a reaction
mixture containing 0.25 mL of 10 mM sorbitol, 0.25 mL of 0.47
mM NAD+, 0.25 mL of 3.75 mU/mL enzymatic solution, and
0.25 mL of 100 mM Tris-HCl buffer (pH 8) in a total volume

Isothiazole Acetic Acid Derivatives
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 22 6905
of 1 mL. All the reagents were incubated at 37 °C for 1 min,
after which the reaction was monitored for 3 min.
Docking to ALR2 was carried out using the empirical free
energy function and the Lamarckian genetic algorithm, ap-
plying a standard protocol with an initial population of 50
randomly placed individuals, a maximum number of 1.5 × 106
energy evaluations, a mutation rate of 0.02, a crossover rate
of 0.80, and an elitism value of 1. Proportional selection was
used, where the average of the worst energy was calculated
over a window of the previous 10 generations. For the local
search, the so-called pseudo Solis and Wets algorithm was
applied using a maximum of 300 iterations per local search.
The probability of performing local search on an individual in
the population was 0.06, and the maximum number of
consecutive successes or failures before doubling or halving
the local search step size was 4.
Fifty independent docking runs were carried out for each
ligand. Results differing by less than 1.5 Å in positional root
mean-square deviation (rmsd) were clustered together and
represented by the result with the most favorable free energy
of binding.
Ligand Setup. Molecular models of compounds 11-14 were
constructed using standard bond lengths and bond angles of
the SYBYL fragment library. The carboxylate group was taken
as dissociated. Geometry optimizations were carried out with
the SYBYL/MAXIMIN2 minimizer by applying the BFGS
(Broyden, Fletcher, Goldfarb, and Shannon) algorithm53 and
setting an rms gradient of the forces acting on each atom of
0.05 kcal mol-1 Å-1 as the convergence criterion.
Atomic charges were assigned using the Gasteiger-Marsili
formalism,54 which is the type of atomic charge used in
calibrating the AutoDock empirical free energy function.
Finally, the compounds were set up for docking with the help
of AutoTors, the main purpose of which is to define the
torsional degrees of freedom to be considered during the
docking process. The number of flexible torsions defined for
each ligand is as follows: 2 in 11, 2 in 12, 5 in 13, and 3 in 14.
Protein Setup. The crystal structure of ALR2 complexed
with tolrestat (entry code: 1AH3),16 recovered from Brookhaven
Protein Database,55 was used. The structure was set up for
docking as follows: polar hydrogens were added using the
PROTONATE utility.38 To optimize the hydrogen positions,
the structure was subjected to a short energy minimization
using the Discover module of InsightII, in accordance with the
type of force field and protein charges of the AutoDock
empirical free energy function. Solvation parameters were
added to the final protein file using the ADDSOL utility of
AutoDock3.5.
The grid maps representing the proteins in the actual
docking process were calculated with AutoGrid. The grids (one
for each atom type in the ligand plus one for electrostatic
interactions) were chosen to be sufficiently large to include
not only the active site but also significant portions of the
surrounding surface. The dimensions of the grids were thus
30 Å × 30 Å × 30 Å, with a spacing of 0.375 Å between the
grid points.
Refinement of the inhibitor/enzyme complexes was achieved
by energy minimization with the CVFF force field, permitting
only the ligand and the side chain atoms of the protein to relax.
Calculations were performed by using 3000 steps of steepest
descents and 2000 steps of conjugate gradients (down to a
maximal atomic root-mean-square derivative of 10.0 and 0.01
kcal/Å, respectively).
GR activity48 was determined at 37 °C in a mixture
containing 0.25 mL of 1 mM glutathione disulfide, 0.25 mL of
0.36 mM NADPH, 0.25 mL of 4.5 mU/mL enzymatic solution,
and 0.25 mL of 0.125 sodium phosphate buffer (pH 7.4)
supplemented with 6.3 mM EDTA potassium salt, in a total
volume of 1 mL.
Enzymatic Inhibition. The inhibitory activity of the newly
synthesized compounds against ALR2, ALR1, SDH, and GR
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 solubilization was com-
plete, the pH was readjusted to 7. To correct for the nonen-
zymatic oxidation of NADPH or reduction of NAD+ and for
the 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 10-7 M. The determination
of the IC50 values was performed by using 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 U.K.), and the
control diet consisted of normal TRM. Both control and
experimental rats had access to food and water ad libitum.
Prevention of Cataract Development. Animals were
randomly divided into groups of equal average body weight
with 15 rats per group. The test compounds 14, 31, and
tolrestat were administered four times daily as eye drops of
appropriate concentrations. The vehicle in which ARIs were
contained was administered with the same dose regimen to
the control group, which was given access to the galactose diet,
and to the group fed with a normal diet, which was included
to record the aspect of normal lenses. Groups treated with the
tested compounds were predosed 1 day before switching their
diet to galactose-containing chow.
Lenses were examined using slit-lamp microscopy after
dilating the pupils with 1% atropine, Farmigea, Italy, to
establish their integrity.
Nuclear cataracts, which appeared as a pronounced central
opacity readily visible as a white spot, were evaluated. The
number of animals that attained this stage was recorded, and
the ability of the test compounds to prevent cataract develop-
ment was assessed on the basis of comparison with galac-
tosemic rats treated only with the vehicle.
Computational Chemistry. Molecular modeling and graph-
ics manipulations were performed using the SYBYL software
package49 running on a Silicon Graphics R12000 workstation.
Model building of inhibitors 11-14 was accomplished with the
Tripos force field50 available within SYBYL. Energy minimiza-
tions of the inhibitor/ALR2 complexes were carried out by
employing the Insight II/Discover program,51 selecting the
CVFF force field.52
Supporting Information Available: Tables 1-4 includ-
ing physical and spectral data of compounds 8-30 and
analytical data of compounds 2-31. This material is available
free of charge via the Internet at http://pubs.acs.org.
Docking Simulations. Docking was performed with ver-
sion 3.05 of the program AutoDock.38 It combines a rapid
energy evaluation through precalculated grids of affinity
potentials with a variety of search algorithms to find suitable
binding positions for a ligand on a given protein. While the
protein is required to be rigid, the program allows torsional
flexibility in the ligand.
References
(1) King, H.; Aubert, R. E.; Herman, W. H. Global Burden of
Diabetes, 1995-2025. Prevalence, Numeric Estimates, and
Projections. Diabetes Care 1998, 21, 1414-1431.
(2) Bressler, R.; Johnson, D. New Pharmacological Approaches to
Therapy of NIDDM. Diabetes Care 1992, 15, 792-805.

6906 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 22
Da Settimo et al.
(3) Mahon, J. L.; Dupre, J.; Stiller, C. R. Lessons Learned from Use
of Cyclosporine for Insulin-Dependent Diabetes Mellitus. The
Case of Immunotherapy for Insulin-Dependent Diabetics Having
Residual Insulin Secretion. Ann. N.Y. Acad. Sci. 1993, 696,
3651-3663.
(4) Brownlee, M. Biochemistry and Molecular Cell Biology of
Diabetic Complications. Nature 2001, 414, 813-820.
(5) Scott, J. A.; King, G. L. Oxidative Stress and Antioxidant
Treatment in Diabetes. Ann. N.Y. Acad. Sci. 2004, 1031, 204-
213.
(6) Whiteside, C. I. Cellular Mechanisms and Treatment of Diabetes
Vascular Complications Converge on Reactive Oxygen Species.
Curr. Hypertens. Rep. 2005, 7, 148-154.
(7) Brownlee, M. The Pathological Implications of Protein Glycation.
Clin. Invest. Med. 1995, 18, 275-281.
(8) Tomlinson, D. R.; Stevens, E. J.; Diemel, L. T. Aldose Reductase
Inhibitors and Their Potential for the Treatment of Diabetic
Complications. Trends Pharmacol. Sci. 1994, 15, 293-297.
(9) Kador, P. F. The Role of Aldose Reductase in the Development
of Diabetic Complications. Med. Res. Rev. 1988, 8, 325-352.
(10) Bravi, M. C.; Pietrangeli, P.; Laurenti, O.; Basili, S.; Cassone-
Faldetta, M.; Ferri, C.; De Mattia, C. Polyol Pathway Activation
and Glutathione Redox Status in Non-Insulin-Dependent Dia-
betic Patients. Metabolism 1997, 46, 1194-1198.
(11) Cheng, H. M.; Gonzales, R. G. The Effect of High Glucose and
Oxidative Stress on Lens Metabolism, Aldose Reductase, and
Senile Cataractogenesis. Metabolism 1986, 35, 10-14.
(12) Chung, S. S. M.; Ho, E. C. M.; Lam, K. S. L.; Chung, S. K.
Contribution of Polyol Pathway to Diabetes-Induced Oxidative
Stress. J. Am. Soc. Nephrol. 2003, 14, S234-S236.
(13) Morre, D. M.; Lenaz, G.; Morre, D. J. Surface Oxidase and
Oxidative Stress Propagation in Aging. J. Exp. Biol. 2000, 203,
1513-1521.
(14) Sarges, R.; Oates, P. J. Aldose Reductase Inhibitors: Recent
Developments. Prog. Drug Res. 1993, 40, 99-161.
(15) Wilson, D. K.; Tarle, I.; Petrash, J. M.; Quiocho, F. A. Refined
1.8 Å Structure of Human Aldose Reductase Complexed with
the Potent Inhibitor Zopolrestat. Proc. Natl. Acad. Sci. U.S.A.
1993, 90, 9847-9851.
(16) Urzhumtsev, A.; Teˆte-Favier, F.; Mitschler, A.; Barbanton, J.;
Barth, P.; Urzhumtseva, L.; Biellmann, J.-F.; Podjarny, A. D.;
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.
(17) Harrison, D. H. T.; Bohren, K. M.; Petsko, G. A.; Ringe, D.;
Gabbay, K. H. The Alrestatin Double-Decker: Binding of Two
Inhibitor Molecules to Human Aldose Reductase Reveals a New
Specificity Determinant. Biochemistry 1997, 36, 16134-16140.
(18) Harrison, D. H.; Bohren, K. M.; Ringe, D.; Petsko, G. A.; Gabbay,
K. H. An Anionic Binding Site in Human Aldose Reductase:
Mechanistic Implications for the Binding of Citrate, Cacodylate,
and Glucose-6-phosphate. Biochemistry 1994, 33, 2011-2020.
(19) Carper, D.; Wistow, G.; Nishimura, C.; Graham, C.; Watanabe,
K.; Fuji, Y.; Hayaishi, H.; Hayaishi, O. A Superfamily of
NADPH-Dependent Reductases in Eukaryotes and Prokaryotes.
Exp. Eye Res. 1988, 49, 377-388.
(26) Da Settimo, A.; Primofiore, G.; La Motta, C.; Da Settimo, F.;
Simorini, F.; Boldrini, E.; Bianchini, P. Acid Derivatives of
Benzisothiazole-1,1-dioxide as Inhibitors of Rat Lens Aldose
Reductase. Farmaco 1996, 51, 261-267.
(27) Primofiore, G.; Da Settimo, F.; La Motta, C.; Simorini, F.;
Minutolo, A.; Boldrini, E. Benzisothiazole-1,1-dioxide Alkanoic
Acid Derivatives as Inhibitors of Rat Lens Aldose Reductase.
Farmaco 1997, 52, 583-588.
(28) Wrobel, J.; Dietrich, A.; Gorham, B. J.; Sestanj, K. Conforma-
tionally Rigid Analogues of Aldose Reductase Inhibitors, Tolr-
estat. Novel Syntheses of Naphthalene-Fused γ-, δ-, and ꢀ-Lac-
tams. J. Org. Chem. 1990, 55, 2694-2702.
(29) 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. Vision 1998, 4, 19-25.
(30) Lombardino, J. G. Preparation of Substituted 1,2-Benzoisothia-
zolin-3-one-1,1-dioxides (o-Benzoic Sulfimides). J. Org. Chem.
1971 36, 1843-1845.
(31) Nicolaou, I.; Demopoulos, V. J. Substituted Pyrrol-1-ylacetic
Acids That Combine Aldose Reductase Enzyme Inhibitory Activ-
ity and Ability To Prevent the Nonenzymatic Irreversible
Modification of Proteins from Monosaccharides. J. Med. Chem.
2003, 46, 417-426.
(32) Boissonnas, R. A.; Guttmann, St.; Jaquenoud, P.-A.; Waller, J.-
P. Synthe`se d’Analogues Structuraux de L’oxytocine (Synthesis
of Structural Analogues of Oxytocin). Helv. Chim. Acta 1956,
39, 1421-1427.
(33) Coudert, P.; Albuisson, E.; Boire, J. Y.; Durox, E.; Bastide, P.;
Couquelet, J. Synthesis of Pyridazine Acetic Acid Derivatives
Possessing Aldose Reductase Inhibitory Activity and Antioxidant
Properties. Eur. J. Med. Chem. 1994, 29, 471-477.
(34) 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.
(35) Camber, O.; Edman, P. Factors Influencing the Corneal Perme-
ability of Prostaglandin F2R and Its Isopropyl Ester in Vitro. Int.
J. Pharm. 1987, 37, 27-32.
(36) Suhonen, P.; Ja¨rvinen, T.; Peura, P.; Urtti, A. Permeability of
Pilocarpic Acid Diesters across Albino Rabbit Cornea in Vitro.
Int. J. Pharm. 1991, 74, 221-228.
(37) Chien, D.-S.; Tang-Liu, D. D.-S.; Woodward, D. F. Ocular
Penetration and Bioconversion of Prostaglandin F2R Prodrugs
in Rabbit Cornea and Conjunctiva. J. Pharm. Sci. 1997, 86,
1180-1186.
(38) 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 an Empirical Binding Free
Energy Function. J. Comput. Chem. 1998, 19, 1639-1662.
(39) Database searching (SWISS-PROT), sequence alignment, and
analysis of rat and pig ALR2 sequences were carried out using
FASTA (Pearson, W. R. Proc. Natl. Acad. Sci. U.S.A. 1988, 85,
2444-2448) and BLAST programs (Wang, S.; Pak, Y. J. Phys.
Chem. B 2000, 104, 354-359). The SWISS-PROT accession
numbers for rat and pig ALR2 types are P07943 and P80276,
respectively.
(40) Harrison, D. H. T.; Bohren, K. M.; Petsko, G. A.; Ringe, D.;
Gabbay, K. H. The Alrestatin Double-Decker: Binding of Two
Inhibitor Molecules to Human Aldose Reductase Reveals a New
Specificity Determinant. Biochemistry 1997, 36, 16134-16140.
(41) Rogniaux, H.; Van Dorsselaer, A.; Barth, P.; Biellmann, J.-F.;
Barbanton, J.; Van Zandt, M.; Chevrier, B.; Howard, E.; Mitschler,
A.; Potier, N.; Urzhumtseva, L.; Moras, D.; Podjarny, A. Binding
of Aldose Reductase Inhibitors: Correlation of Crystallographic
and Mass Spectrometric Studies. J. Am. Soc. Mass Spectrom.
1999, 10, 635-647.
(20) Costantino, L.; Rastelli, G.; Vescovini, K.; Cignarella, G.; Vi-
anello, P.; Del Corso, A.; Cappiello, M.; Mura, U.; Barlocco, D.
Synthesis, Activity, and Molecular Modeling of a New Series of
Tricyclic Pyridazinones as Selective Aldose Reductase Inhibitors.
J. Med. Chem. 1996, 39, 4396-4405.
(21) Kador, P. F.; Inoue, J.; Secchi, E. F.; Lizak, M. J.; Rodriguez,
L.; Mori, K.; Greentree, W.; Blessing, K.; Lackner, P. A.; Sato,
S. Effect of Sorbitol Dehydrogenase Inhibition on Sugar Cataract
Formation in Galactose-Fed and Diabetic Rats. Exp. Eye Res.
1998, 67, 203-208.
(42) Bohren, K. M.; Grimshaw, C. E.; Lai, C. J.; Harrison, D. H.;
Ringe, D.; Petsko, G. A.; Gabbay, K. H. Tyrosine-48 Is the Proton
Donor and Histidine-110 Directs Substrate Stereochemical
Selectivity in the Reduction Reaction of Human Aldose Reduc-
tase: Enzyme Kinetics and Crystal Structure of the Y48H
Mutant Enzyme. Biochemistry 1994, 33, 2021-2032.
(43) Grimshaw, C. E.; Bohren, K. M.; Lai, C. J.; Gabbay, K. H. Human
Aldose Reductase: pK of Tyrosine 48 Reveals the Preferred
Ionization State for Catalysis and Inhibition. Biochemistry 1995,
34, 14374-14384
(44) El-Kabbani, O.; Carper, D. A.; McGowan, M. H.; Devedjiev, Y.;
Milton, K. J.; Flynn, T. G. Studies on the Inhibitor-Binding Site
of Porcine Aldehyde Reductase: Crystal Structure of the Ho-
loenzyme-Inhibitor Ternary Complex. Proteins 1997, 28, 186-
192.
(22) Giblin, F. J.; McCready, J. P.; Schrimscher, L.; Reddy, V. N.
Peroxide-Induced Effects on Lens Cation Transport Following
Inhibition of Glutathione Reductase Activity in Vitro. Exp. Eye
Res. 1987, 45, 77-91.
(23) Reddan, J. R.; Giblin, F. J.; Dziedzic, D. C.; McCready, J. P.;
Schrimscher, L.; Reddy, V. N. Influence of the Activity of
Glutathione Reductase on the Response of Cultured Lens
Epithelial Cells from Young and Old Rabbits to Hydrogen
Peroxide. Exp. Eye Res. 1998, 46, 209-221.
(24) 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 De-
rivatives: A New Class of Selective Aldose Reductase Inhibitors.
J. Med. Chem. 2001, 44, 4359-4369.
(25) 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.
(45) Hayman, S.; Kinoshita, J. H. Isolation and Properties of Lens
Aldose Reductase. J. Biol. Chem. 1965, 240, 877-882.
(46) Ward, W. H. J.; Sennitt, C. M.; Ross, H.; Dingle, A.; Timmus,
D.; Mirrless, D. J.; Tuffin, D. P. Ponalrestat, a Potent and

Isothiazole Acetic Acid Derivatives
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 22 6907
Specific Inhibitor of Aldose Reductase. Biochem. Pharmacol.
1990, 39, 337-346.
(52) Hagler, A. F.; Lifson, S.; Dauber, P. Consistent Force Field
Studies of Intermolecular Forces in Hydrogen-Bonded Crystals.
J. Am. Chem. Soc. 1979, 101, 5122-5130.
(53) Head, J.; Zerner, M. C. A Broyden-Fletcher-Goldfarb-Shannon
Optimization Procedure for Molecular Geometries. Chem. Phys.
Lett. 1985, 122, 264-274.
(54) Gasteiger, J.; Marsili, M. Iterative Partial Equalization of Orbital
Electronegativity. A Rapid Access to Atomic Charges. Tetrahe-
dron 1980, 36, 3219-3228.
(55) Bernstein, F. C.; Koetzle, T. F.; Williams, G. J. B.; Meyer, E. F.,
Jr.; Brice, M. D.; Rodgers, J. R.; Kennard, O.; Shimanouchi, T.;
Tasumi, T. The Protein Data Bank: A Computer Based Archival
File for Macromolecular Structures. J. Mol. Biol. 1977, 112,
535-542.
(47) Lindstad, R. I.; Hermansen, L. F.; McKinley-McKee, J. S.
Inhibition and activation studies on sheep liver sorbitol dehy-
drogenase. Eur. J. Biochem. 1994, 221, 847-854.
(48) Fitzgerald, G. B.; Bauman, C.; Sajjat Hussoin, Md.; Wick, M.
M. 2,4-Dihydroxy benzylamine: a specific inhibitor of glu-
tathione reductase. Biochem. Pharmacol. 1991, 41, 185-190.
(49) SYBYL Molecular Modelling System, version 6.9.1; Tripos Inc.:
St. Louis, MO, 2003.
(50) Vinter, J. G.; Davis, A.; Saunders, M. R. Strategic Approaches
to Drug Design. 1. An Integrated Software Framework for
Molecular Modelling. J. Comput.-Aided Mol. Des. 1987, 1, 31-
55.
(51) Insight II Molecular Modelling Package and Discover 2. 2000
Simulation Package; MSI Inc.: San Diego, CA, 2000.
JM050382P