Nitrophenyl derivatives as aldose reductase inhibitors

Article abstract of DOI:10.1016/S0968-0896(02)00318-8

Nitrophenyl derivatives were recently discovered as a new class of ALR2 inhibitors by means of docking and database screening of the National Cancer Institute database of organic molecules. The nitro group was predicted to bind to the Tyr48 and His110 act

Full text of DOI:10.1016/S0968-0896(02)00318-8

Bioorganic & Medicinal Chemistry 10 (2002) 3923–3931
Nitrophenyl Derivatives as Aldose Reductase Inhibitors
Luca Costantino,* Anna Maria Ferrari, Maria Cristina Gamberini and Giulio Rastelli
`
Dipartimento di Scienze Farmaceutiche, Universita di Modena e Reggio Emilia, Via Campi 183, 41100 Modena, Italy
Received 27 March 2002; accepted 18July 2002
Abstract—Nitrophenyl derivatives were recently discovered as a new class of ALR2 inhibitors by means of docking and database
screening of the National Cancer Institute database of organic molecules. The nitro group was predicted to bind to the Tyr48and
His110 active site residues of the enzyme, the site where acidic ALR2 inhibitors such as carboxylic acids bind in their anionic form.
Given the novelty of these compounds, we decided to expand their structure–activity relationships by synthesizing and testing a
series of derivatives and the corresponding compounds having a carboxylic group instead of the nitro moiety; the results obtained
were rationalized by means of docking and molecular dynamics simulations. On the whole there is an agreement between inhibitory
data and the results of molecular modeling experiments, supporting the hypothesized binding mode of these compounds.
# 2002 Elsevier Science Ltd. All rights reserved.
Introduction
Aldose reductase (ALR2) is an enzyme that catalyzes
the reduction of a variety of carbonyl compounds to the
corresponding alcohols. It is the first enzyme of the so
called ‘polyol pathway’ and it is able to reduce glucose
to sorbitol under hyperglycemic conditions. The activity
of this metabolic pathway is believed to be linked to
long-term diabetic complications such as cataract, retino-
pathy, nephropathy and neuropathy. Thus, aldose
reductase inhibitors (ARIs) represent a potential thera-
peutic strategy for preventing the onset or the progression
of these complications.1
Among the compounds with the NO2 moiety at 20 or 40
position, only the NO2 moiety at 20-position was able to
hydrogen bond efficiently with the Tyr48, His110 and
Trp 111 residues of the active site. This conclusion was
drawn from the structure of ALR2 complexed with
4-(20,40-dinitro-anilino)phenol.7 In addition, the phe-
nolic ring interacts with the hydrophobic pocket of the
enzyme, and the hydroxyl hydrogen bonds to Thr113,
contributing favorably to inhibitory activity. The most
active compound thus discovered was 4-(20-nitrophenyl-
mercapto)phenol (R=20-NO2).7
Several studies showed that the inhibitor binding site of
ALR2 possesses two subsites, able to accommodate the
two main pharmacophoric elements of ARIs. The first,
mainly hydrophilic, accommodates acidic functionalities
and is constituted by Tyr48, His110 and the oxidized
nicotinamide ring of the cofactor, NADP+; the second
binds the hydrophobic fragments of ARIs and is mainly
constituted by Trp111, Leu300 and Phe122.2ꢀ6
In this work we extend the SAR in this class of com-
pounds by firstly synthesizing the 30-nitro derivative.
Then, the nitro group at position 20, 30 and 40 was sub-
stituted with carboxylic acid moieties (Table 1). These
substitutions were made in order to make a comparison
between the inhibitory activities of our new compounds
with the activities of a well-known class of ARIs, that is,
that of the carboxylic acid-containing ARIs, in which
the carboxylate hydrogen bonds to the Tyr48and
His110 residues of the enzyme. The carboxylic acid
Recently, simple aromatic nitrocompounds of general
formula (A) were discovered as a new class of ARIs as a
result of a molecular docking and database screening
study of the National Cancer Institute database of
organic molecules into the crystal structure of ALR2.7
*Corresponding author. Tel.: +39-059-2055125; fax: +39-059-
2055131; e-mail: costantino.luca@unimo.it
0968-0896/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(02)00318-8

3924
L. Costantino et al. / Bioorg. Med. Chem. 10 (2002) 3923–3931
Table 1. Enzyme inhibition data
a
Compound
R1
R2
R3
R4
R5
R6
X
IC50
1
2
3
4
5
6
7
H
H
NO2
COOH
CH2COOH
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
OH
OH
OH
OH
OH
OH
OH
S
S
S
S
S
S
S
4.9 (3.9–6.2)
21 (18–25)
14 (11–18)
78(63–96)
35 (29–43)
18(15–21)
22 (18–27)
H
COOH
CH2COOH
H
H
H
H
H
COOH
CH2COOH
H
8
NO2
H
H
H
H
H
H
H
H
H
H
H
H
OH
OH
OH
69 (59–81)
17% at 64 mM
16 (14–18)
9
COOH
10
CH2COOH
11
12
13
14
15
16
17
18
19
NO2
NO2
NO2
NO2
NO2
NO2
NO2
NO2
NO2
H
H
H
H
H
CH3
H
H
OH
COOC2H5
CONH2
NH2
COOH
H
CH3
H
H
H
H
H
H
H
CH3
H
H
H
H
H
H
H
H
H
CH3
OH
OH
OH
OH
OH
OH
OH
OH
OH
S
S
S
S
S
S
S
S
S
3.5 (2.9–4.2)
2.30 (1.95–2.71)
2.35 (2.02–2.73)
3.48(2.91–4.16)
0.99 (0.81–1.21)
14 (11–17)
2.96 (2.52–3.48)
1.57 (1.31–1.88)
17% at 80 mM
H
H
H
20
21
22
23
NO2
H
H
H
H
H
H
H
S
S
S
0% at 15 mM
0% at 15 mM
0% at 15 mM
2.40 (1.86–3.09)
NO2
NO2
NO2
OH
NO2
NO2
H
H
H
H
OH
H
S
24
Sorbinil
Tolrestat
H
OCH3
20 (18–23)
1.68(1.44–1.96)
0.057 (0.046–0.070)
aIC50 values (mM) (95% confidence limits) or % inhibition at a given concentration. The reported concentrations are the highest possible before
precipitation of the compounds in the assay solution.
derivatives, being completely ionized at physiological
pH, confer to ARIs an in vivo activity which is generally
lower than that of less ionized compounds such as spir-
ohydantoins (Sorbinil). While this effect is probably due
to an impaired penetration of physiological membranes of
ionized compounds,8 the nitro derivatives here developed
are neutral at physiological pH.7 The same pattern of sub-
stitutions, namely, the bioisosteric substitution between
the nitro group and the carboxylic moiety, was applied to
triphenylmethane derivatives (Table 1). These last com-
pounds were selected because benzylic and diphe-
nilmethane moieties of previously studied compounds
were shown to possess similar ALR2 inhibitory activity.9
ligand which was predicted to be mainly exposed to
solvent in the ALR2-inhibitor complex.7 We also intro-
duced a methyl group at positions 30, 40, 50 and 60 in
order to investigate the steric effects of a methyl probe
with the active site residues, as well as a benzene ring
condensed at positions 50 and 60 (Table 1).
Results and Discussion
Chemistry
The (4-hydroxyphenylthio)-benzoic and -phenylacetic
acids (compounds 2–7) were prepared by the reaction
between the corresponding aryl iodide and 4-hydroxy-
thiophenolate in presence of a catalytic amount of Cu,
(Scheme 1), following the general procedure reported.10
Finally, we decided to introduce substituents of various
nature at position 40 of 4-(20-nitrophenylmercapto)phenol
in order to investigate their effects in a portion of the

L. Costantino et al. / Bioorg. Med. Chem. 10 (2002) 3923–3931
3925
(Scheme 3). Compound 15 was obtained by means of
alkaline hydrolysis of compound 12.
Aldose reductase inhibition
Starting from the lead compound 4-(20-nitrophe-
nylmercapto)phenol (general formula A, R=20-NO2)
with an IC50 of 4.8 mM,7 several new derivatives were
synthesized and tested (Table 1). We previously reported
that when the nitro substituent is placed at position 40
instead of 20, the compound is about six times less active
(IC50: 30 mM).7 This finding was predicted by the
remarkably stronger ability of the nitro group at position
20 to hydrogen bond to Tyr48, His110 and Trp111 resi-
dues of the active site.7 In order to test the effect of a
nitro moiety at position 30, compound 1 was synthesized
and found to have an inhibitory activity which is very
close to 4-(20-nitrophenylmercapto)phenol (IC50: 4.9
mM, Table 1). Docking into ALR2 and molecular
dynamics refinement of the ALR2-inhibitor complex of
4-(20-nitrophenylmercapto)phenol and 1 have been per-
formed, and the resulting structures have been com-
pared, in order to explain the activity of 1. Figure 1
reports the structures of the ALR2 complexes with the
lead inhibitor 4-(20-nitrophenylmercapto)phenol (Fig. 1a)
and with compound 1 (Fig. 1b). 4-(20-nitrophenyl-
mercapto)phenol binds ALR2 by adopting an orientation
which is almost identical to that previously described for
its congener 4-(20,40-dinitroanilino)phenol.7 The 20-nitro
group hydrogen bonds to Tyr48, His110 and Trp111,
and the phenolic ring is inserted into the additional
hydrophobic pocket of ALR2 making Van der Waals
contacts with Trp111, Leu 300, Phe122 and Cys303.
Moreover, the hydroxyl substituent hydrogen bonds to
Thr113. The orientation of the 30-nitro derivative 1 in
the active site is such that compound 1 is still able to
hydrogen bond to Tyr48, His110, Trp111 and Thr113
(Fig. 1b). The main difference between compound 1 and
4-(20-nitrophenylmercapto)phenol is in the positioning
of the two sulphur atoms with respect to the active site
residues, difference which does not prevent the binding
of the phenol ring into the hydrophobic pocket. There-
fore, structures nicely explain why the 20- and 30-nitro
derivatives are both apt to bind ALR2, and in a similar
way.
Scheme 1.
Scheme 2. (a) C6H5CH2Br, K2CO3; (b) LiAIH4; (c) PBr3; (d) NaCN;
(e) NaCN; (e) NaOH; (f) H2/Pd-C; (g) NaOH.
The 2-(4,40-dihydroxybenzidryl)phenylacetic acid 10 was
prepared starting from the methyl ester of phenolphtalin
10a (Scheme 2). After the protection of the two phenolic
groups as benzyl derivative (10b), the ester group was
reduced with LiAlH4 (10c), the alcohol was converted into
the bromoderivative by treatment with PBr3 in CH2Cl2
(10d), then reaction with NaCN in DMSO at 70 ꢁC
afforded the corresponding nitrile 10e. The subsequent
hydrolisis with refluxing NaOH 50% afforded only the
corresponding amide 10f. After deprotection of the two
phenolic hydroxy groups by means of H2/Pd-C, alkaline
hydrolysis of the amide afforded 10.
The study of compounds 2–10 (Table 1) served to
investigate the differences in inhibitory activity between
nitrophenyl derivatives and the corresponding car-
boxylic acids (the acetic acid moiety was selected
because the most active ARIs possess this group).11 In
the class of diarylsulphides, the nitro derivatives 4-(20-
nitrophenylmercapto)phenol (IC50 4.8 mM7) and 4-(30-
nitrophenylmercapto)phenol 1 proved to be more active
than the corresponding benzoic and phenylacetic acid
derivatives 2–5 (Table 1) while for position 40, the car-
boxylic acids 6,7 are as active as the nitrophenyl deri-
vative (4-(40-nitrophenylmercapto)phenol, IC50: 30
mM).7 Within the triphenylmethanes 8–10, the nitro
compound 8 is still more active than the benzoic deri-
vative 9, but it is four times less active than the acetic
acid 10. Thus the activity exerted by the nitro group, in
comparison to the carboxylic acid derivatives, depends
The 20-nitro-40-methoxydiphenylsulphide (24), the
4-(nitrosubstituted arylmercapto)-phenols and -naphta-
lenes (compounds 12–14, 16–23) were obtained by the
reaction of the appropriate nitroaryl halide with the
corresponding thiophenol in alkaline medium (Scheme
3); likewise, compound 11 was obtained starting from 4-
chloronitrophenol 11a protected as acetylderivative

3926
L. Costantino et al. / Bioorg. Med. Chem. 10 (2002) 3923–3931
Scheme 3.
on the nature of the substrate (diarylsulphide or triphe-
nylmethane) and on the position in which this moiety is
inserted.
bulkier pyrazine ring which is condensed at positions 40
and 50, is as active as 4-(20-nitrophenylmercapto)phenol,
again in agreement with the proposed binding mode.
In order to test the hypothesis derived from modeling
experiments that position 40 of diarylsulphides is direc-
ted toward the opening of the active site of the enzyme,
compounds 11–15 were synthesized. As expected, the
activity of these compounds is rather close to the activity
of 4-(20-nitrophenylmercapto)phenol (Table 1). More-
over, we introduced a methyl probe at positions 30, 40 50
and 60 (compounds 16–19). While 17 and 18 turned out
to be active as 4-(20-nitrophenylmercapto)phenol, com-
pound 16 is 3-times less active and 19 is almost inactive
(Table 1). Remarkably, the trend of inhibitory activity is
in full agreement with the mode of binding predicted for
this class of compounds. The molecular surface of the
ALR2–4-(20-nitrophenylmercapto)phenol active site
drawn in Figure 1c unequivocally shows that while
positions 30 and 60 are sterically hindered and cannot
accept substituents, positions 40 and 50 are directed
toward the opening of the active site, and can therefore
be substituted without loss of activity. Based on struc-
tures, a methyl at position 30 would be in steric conflict
with the Tyr48and Val47 side chains, while a methyl at
position 60 would be in steric conflict with the Leu300 and
Trp219 side chains (Fig. 1a and c). As a further evidence
that position 60 is in steric conflict with the enzyme, com-
pounds 20–22, which have a bulkier benzene ring con-
densed at positions 60 and 50, are completely inactive
(Table 1). On the contrary, compound 23, having a
Finally, the substitution of the hydroxy group in 4-(20-
nitrophenylmercapto)phenol with a methoxy group (24)
lead to a less active compound (IC50: 20 mM, Table 1).
The structure reported in Figure 1a shows, in fact, that
this hydroxyl hydrogen bonds to Thr113, contributing
favorably to inhibitory activity.7
In conclusion, we found a qualitative agreement
between inhibitory data and molecular modeling
experiments; the results here obtained will be useful
for the synthesis of new ARIs neutral at physiological
pH.
Experimental
Chemistry
Melting points were determined on a Buchi 510 capil-
lary melting point apparatus and are uncorrected. Ele-
mental analyses for the tested compounds were within
1
+/ꢀ0.4% of the theoretical values. H NMR spectra
(200 MHz) were recorded on a Bruker AC200 spectro-
meter; chemical shifts are reported as as d (ppm) relative to
TMS.; DMSO-d6 was used as solvent unless otherwise
noted. TLC on silica gel plates was used to check product
purity. Silica gel 60 (Merck, 70–230 mesh) was used for

L. Costantino et al. / Bioorg. Med. Chem. 10 (2002) 3923–3931
3927
Figure 1. (a) Stereoview of active site residues of ALR2 interacting with the lead inhibitor 4-(20-nitrophenylmercapto)phenol after refinement with
MD. (b) Stereoview of active site residues of ALR2 interacting with the inhibitor 4-(30-nitrophenylmercapto)phenol 1 after refinement with MD. For
clarity, only the polar hydrogens are shown, and hydrogen bonds are shown as dotted lines. Water molecules are shown as dots. (c) Molecular
surface of ALR2 with the inhibitor 4-(20-nitrophenylmercapto)phenol bound at the active site. Positions substituted in the present work are num-
bered from 30 to 60.
column chromatography. Microanalyses were carried out
in the Microanalysis Laboratory of the Dipartimento di
Scienze Farmaceutiche, Modena University. 4-(30-nitro-
phenyl)mercaptophenol 1 was synthesized according to
ref 12, 4,40-(nitrobenzyliden)diphenol (8) and 2-(4,40-di-
hydroxybenzidryl)benzoic acid 9 according to ref 13
1-Chloro-2-nitronaphtalene 20a and 1-chloro-4-nitro-
naphtalene 21a were synthesized according to ref 14;
2,4-dinitro-1-chloronaphtalene 22a was synthesized
accordingto ref 15; 3-chloro-2-nitrotoluene 16a and
3-chloro-4-nitrotoluene 18a were synthesized according
to refs 16,17.

3928
L. Costantino et al. / Bioorg. Med. Chem. 10 (2002) 3923–3931
(4-Hydroxyphenylthio)-benzoic and-acetic acids (2–7).
To a stirred mixture of the appropriate iodobenzoic or
iodophenylacetic acid (4.00 mmol) and 4-hydro-
xythiophenol (4.00 mmol) in water (30 mL), KOH was
added (8.00 mmol), and Cuꢁ (0.31 mmol). The reaction
mixture was then refluxed for 12 h, cooled, filtered
and the filtrate acidified. The precipitate thus formed
was collected, washed with water and purified as
described.
reaction mixture was extracted with EtOAc (3 ꢂ 25
mL), the organic layer was dried (Na2SO4) and the sol-
vent removed under reduced pressure. The residue thus
obtained was purified as described.
4-(20-Nitro-40-hydroxyphenylmercapto)phenol (11). This
compound was prepared as described above, starting
from 4-(20-nitro-40-acetoxyphenylmercapto)phenol, which
in turn was prepared from 4-chloro-3-nitrophenol 14a,
following standard procedures. Purification by means of
column chromatography (CH2Cl2/CH3OH 97.5/2.5,
then cyclohexane/EtOAc 60/40). Yield 6%; mp 178–
181 ꢁC (diethyl ether/petroleum ether); 1H NMR: d
10.35 (1H, s), 10.00 (1H, s), 7.55 (1H, d, J 2.00), 7.38
(2H, m), 7.06 (1H, dd, J 2.00, J 8.28), 6.90 (2H, m), 6.75
(1H, d, J 2.00). Anal. calcd for C12H9NO4S: C 54.75; H,
3.45; N, 5.32; found: C, 54.96, H, 3.36; N, 5.20.
3-(40-Hydroxyphenylmercapto)benzoic acid (2). Yield
86%; mp 174–175 ꢁC (EtOAc/petroleum ether); 1H
NMR: d 12.95 (1H, s), 9.90 (1H, s), 7.70 (1H, m), 7.59
(1H, m), 7.38(4H, m), 6.8(2H, m). Anal. calcd for
C13H10O3S: C 63.40; H, 4.09; found: C, 63.01, H, 4.15.
3-(40-Hydroxyphenylmercapto)phenylacetic acid (3).
Yield 58%; mp 150–152 ꢁC (EtOAc/petroleum ether);
1H NMR: d 11.00 (2H, broad s), 7.35 (2H, m), 7.20 (1H,
m), 7.06 (2H, m), 6.96 (1H, m), 6.84 (2H, m), 3.50 (2H,
s). Anal. calcd for C14H12O3S: C 64.60; H, 4.65; found:
C, 64.32, H, 4.78.
Ethyl 3-nitro-4-(40-hydroxyphenylmercapto)benzoate (12).
Purification by means of column chromatography
(CH2Cl2/CH3OH 95/5). Yield 60%; mp 148–150 ꢁC (di-
1
ethyl ether/petroleum ether); H NMR (CDCl3): d 8.86
(1H, d, J 1.83), 7.95 (1H, dd, J 8.60, J 1.83), 7.45 (2H, m),
7.05 (2H, m), 6.95 (1H, d, J 8.78), 6.11 (1H, broad s), 4.45
(2H, q), 1.43 (3H, t). Anal. calcd for C15H13NO5S: C
56.42; H, 4.10; N, 4.39; found: C, 56.20, H, 4.18; N, 4.48.
2-(40-Hydroxyphenylmercapto)benzoic acid (4). Yield
80%; mp 238–240 ꢁC (acetone/petroleum ether) (193 ꢁC
ref. 18); IR (nujol) (cmꢀ1): 3622, 3356, 2649, 2360, 1673;
1H NMR: d 13.12 (1H, s), 9.99 (1H, s), 7.96 (1H, d, J
7.56), 7.36 (3H, m), 7.13 (1H, t), 6.92 (2H, m), 6.64 (1H,
d, J 8.00). Anal. calcd for C13H10O3S: C 63.40; H, 4.09;
found: C, 63.06, H, 4.15.
4-(40-Carboxamide-20-nitrophenylmercapto)phenol (13).
ꢁ
1
Yield 10%; mp 250–252 C; H NMR: d32;10.16 (1H,
s), 8.72 (1H, d, J 2.03), 8.15 (1H, s), 7.98 (1H, dd, J 2.00,
J 8.78), 7.55 (1H, s), 7.45 (2H, m), 6.95 (2H, m), 6.85 (1H,
d, J 8.60). Anal. calcd for C13H10N2O4S: C 53.79; H, 3.47;
N, 9.65; found: C, 53.70, H, 3.55; N, 9.60.
2-(40-Hydroxyphenylmercapto)phenylacetic acid (5).
Yield 67%; mp 172–173 ꢁC (acetone/petroleum ether);
1H NMR: d 12.00 (1H, s), 9.50 (1H, s), 7.23 (5H, m),
7.00 (1H, m), 6.81 (2H, m), 3.78 (2H, s). Anal. calcd for
C14H12O3S: C 64.60; H, 4.65; found: C, 64.51, H, 4.68.
4-(40-Amino-20-nitrophenylmercapto)phenol (14). Yield
ꢁ
1
5%; mp 172–175 C; H NMR: d 9.88 (1H, broad s),
7.30 (3H, m), 6.85 (2H, m), 6.78 (1H, d, J 2.56), 6.68
(1H, d, J 8.60), 5.73 (2H, s). Anal. calcd for
C12H10N2O3S: C 54.95; H, 3.84; N, 10.68; found: C,
54.59, H, 3.90; N, 10.88.
4-(40-Hydroxyphenylmercapto)benzoic acid (6). Yield
75%; mp 192–194 ꢁC (diethyl ether/petroleum ether); IR
(cmꢀ1) (Nujol): 3328, 2677, 2565, 1684; 1H NMR: d 7.05
(2H, m), 6.60 (2H, m), 6.30 (2H, m), 6.12 (2H, m), 4.06
(2H, s). Anal. calcd for C13H10O3S: C 63.40; H, 4.09;
found: C, 63.36, H, 4.19.
3-Nitro-4-(40-hydroxyphenylmercapto)benzoic acid (15).
The ethyl 3-nitro-4-(40-hydroxyphenyl)benzoate 15
(0.10 g, 0.31 mmol) was treated with NaOH 2 N (3 mL)
and EtOH (3 mL) at rt with stirring. After 15 min the
solution thus obtained was acidified and the solid thus
formed was collected and washed with water. Yield
0.040 g (44%); mp 238–240 ꢁC (acetone/petroleum
4-(40-Hydroxyphenylmercapto)phenylacetic acid (7).
Yield 81%; mp 185–187 ꢁC (EtOAc/petroleum ether);
1H NMR: d 12.25 (1H, s), 9.80 (1H, s), 7.33 (2H, m),
7.18(2H, m), 7.06 (2H, m), 6.84 (2H, m), 3.60 (2H, m).
Anal. calcd for C14H12O3S: C 64.60; H, 4.65; found: C,
64.33, H, 4.78.
1
ether); H NMR: d 13.40 (1H, s), 10.05 (1H, s), 8.66
(1H, d, J 1.80), 8.04 (1H, dd, J 1.80, J 8.49), 7.45 (2H,
m), 6.95 (3H, m). Anal. calcd for C13H9NO5S: C 53.61;
H, 3.11; N, 4.81; found: C, 53.36, H, 3.20; N, 4.99.
4-(20-Nitrosubstitutedphenylmercapto)phenols
(11–14,
15–19 and 22, 23). To a stirred solution of the appro-
priate aryl halide 11a–14a, 15a–19a and 22a, 23a (chlo-
ride or bromide) (2.36 mmol) in EtOH (10 mL), a
solution of 4-hydroxythiophenol (7.93 mmol) in EtOH
(2 mL) was added, followed by K2CO3 (7.93 mmol).
The reaction mixture was then refluxed for 6 h, (in the
case of 2,4-dinitro-1-(40-hydroxyphenylthio)naphthalene
22 the reaction was performed at 0 ꢁC for 30 min, in the
case of 7-nitro-6-(40-hydroxyphenylmercapto)quinox-
aline 23 at 40 ꢁC for 1 h), then cooled. Water was added;
the pH was brought to 6.00 with HCl 1 N. Then the
4-(30-Methyl-20-nitrophenylmercapto)phenol (16). Pur-
ification by means of column chromatography (CH2Cl2/
CH3OH 97.5/2.5 then cyclohexane/EtOAc 75/25). Yield
5%; 1H NMR (CDCl3): d7.44 (2H, m), 7.20 (1H, t), 7.10
(1H, d), 6.94 (1H, d), 6.88 (2H, m), 5.33 (1H, s), 2.40
(3H, s). Anal. calcd for C13H11NO3S: C 59.76; H, 4.24;
N, 5.36; found: C, 59.98, H, 4.21; N, 5.30.
4-(40-Methyl-20-nitrophenylmercapto)phenol (17). Pur-
ification by means of column chromatography

L. Costantino et al. / Bioorg. Med. Chem. 10 (2002) 3923–3931
3929
(CH2Cl2:CH3OH 95/5). Yield 81%; mp 124–1236 ꢁC
5.22 (1H, t), 5.06 (4H, s), 4.45 (2H, d).
(128 ꢁC ref 19); 1H NMR (CDCl3): d 8.06 (1H, d, J 0.70),
7.45 (2H, m), 7.18(1H, dd, J 0.70, J 8.74), 7.00 (2H, m),
6.77 (1H, d, J 8.74), 4.95 (1H, broad s), 2.38 (3H, s).
Anal. calcd for C13H11NO3S: C 59.76; H, 4.24; N, 5.36;
found: C, 59.58, H, 4.36; N, 5.45.
2-(4,40-Dibenzyloxybenzidril)benzyl bromide (10d). To a
solution of 2-(4,40-dibenzyloxybenzidril)benzylalcohol
10c (3.0 g, 6.17 mmol) in anhyd CH2Cl2 (20 mL) at 0 ꢁC
and under stirring, a solution of PBr3 in CH2Cl2 (1 M,
2.5 mL, 2.5 mmol) was added, then the solution was
maintained at rt for 2 h. After, the solution was cooled
at 0 ꢁC, ice was added and extracted with CH2Cl2 (3ꢂ40
mL). The organic layer was dried (Na2SO4), the solvent
removed under reduced pressure and the residue pur-
ified by column chromatography (ciclohexane/EtOAc
4-(50-Methyl-20-nitrophenylmercapto)phenol (18). Pur-
ification by means of column chromatography (cyclo-
ꢁ
1
hexane/EtOAc 80/20). Yield 24%; mp 142–145 C; H
NMR: d 7.25 (1H, d, J 8.20), 6.58 (2H, m), 6.20 (1H, dd, J
2.00, J 8.20), 6.13 (2H, m), 5.80 (1H, d, J 2.00), 4.05 (1H,
s), 1.38(3H, s). Anal. calcd for C 13H11NO3S: C 59.76; H,
4.24; N, 5.36; found: C, 59.70, H, 4.26; N, 5.55.
1
90/10). Yield 2.90 g (86%), oil; H NMR: d 7.80 (12H,
m), 7.00 (10H, s), 5.90 (1H, s), 5.05 (4H, s), 4.61 (2H, s).
4-(60-Methyl-20-nitrophenylmercapto)phenol (19). Pur-
ification by means of column chromatography (cyclo-
hexane/EtOAc 80/20). Yield 50%; mp 88–91 ꢁC; 1H
NMR: d 9.66 (1H, broad s), 7.70 (1H, m), 7.55 (2H, m),
7.02 (2H, m), 6.70 (2H, m), 2.32 (3H, s). Anal. calcd for
C13H11NO3S: C 59.76; H, 4.24; N, 5.36; found: C, 59.91,
H, 4.20; N, 5.28.
2-(4,40-Dibenzyloxybenzidril)benzyl cyanide (10e). To a
solution of NaCN (0.39 g, 8.00 mmol) in DMSO (2 mL)
at 70 ꢁC
a
solution of 2-(4,40-dibenzyloxybenzi-
dril)benzyl bromide 10d (2.90 g, 5.28mmol) in DMSO
was added and the mixture was stirred for 90 min. Then
water was added, the mixture was extracted with diethyl
ether (3ꢂ50 mL), the organic layer was dried (Na2SO4)
and the solvent removed under reduced pressure. The
residue thus obtained was purified by means of column
chromatography. (cyclohexane/EtOAc 90/10). Yield 2.2
2,4-Dinitro-1-(40-hydroxyphenylthio)naphthalene
Purification by means of column chromatography
(22).
(cyclohexane/EtOAc 75/25). Yield 78%; mp 148–150 ꢁC
g (84%), oil; IR (cmꢀ1): 2248; H NMR: d 7.40 (12H,
1
1
(EtOAC/cyclohexane); H NMR: d 9.78(1H, s), 8.84
m), 7.00 (10H, s), 5.72 (1H, s), 5.05 (4H, s), 3.86 (2H, s).
(1H, s), 8.74 (1H, m), 8.39 (1H, m), 7.94 (2H, m), 7.20
(2H, m), 6.71 (2H, m). Anal. calcd for C16H10N2O5S: C
56.14; H, 2.94; N, 8.18; found: C, 56.10, H, 2.98; N,
8.11.
2-(4,40-Dibenzyloxybenzidril)phenylacetamide (10f).
A
suspension of 2-(4,40-dibenzyloxybenzidril)benzyl cya-
nide 10e (1.0 g, 1.78mmol) in NaOH 50% (5.0 mL,
37.50 mmol) was refluxed for 12 h, then cooled. The
aqueous layer was discarted, and the organic layer was
extracted with CH2Cl2 (3ꢂ50 mL), the organic layer
was dried (Na2SO4) and the solvent removed under
reduced pressure. The residue thus obtained was pur-
ified by column chromatography (CH2Cl2/CH3OH
87.5/12.5). Yield 0.60g (58%); IR (Nujol) (cmꢀ1): 1665.
7-Nitro-6-(40-hydroxyphenylmercapto)quinoxaline (23).
Purification by means of column chromatography
(CH2Cl2/CH3OH 95/5). Yield 15%; mp 260–262 ꢁC; H
1
NMR: d 10.20 (1H, s), 8.95 (3H, m), 7.51 (2H, m), 7.28
(1H, s), 7.02 (2H, m). Anal. calcd for C14H9N3O3S: C
56.18; H, 3.03; N, 14.04; found: C, 55.96, H, 3.15; N,
14.20.
2-(4,40-Dihydroxybenzidril)phenylacetamide (10g).
solution of
2-(4,40-dibenzyloxybenzidryl)phenyl-
A
Methyl 2-(4,40-dibenzyloxybenzidril)benzoate (10b). To a
solution of methyl 2-(4,40-dihydroxybenzidril)benzoate
(10a) (2.83 g, 8.47 mmol) in anhyd acetone (50 mL)
benzyl bromide was added (1.38mL, 16.95 mmol) and
K2CO3 (2.34 g, 16.93 mmol) at rt and under stirring; the
mixture was refluxed for 18h, then cooled and filtered;
the solvent was removed under reduced pressure and the
residue was purified by column chromatography (cyclo-
hexane/EtOAc 87.5/12.5). Yield 3.69 g (85%); 1H NMR
(CDCl3): d 7.82 (1H, m), 7.40 (12H, m), 7.00 (9H, m),
6.55 (1H, s), 5.02 (4H, s), 3.78(3H, s).
acetamide 10f (0.60 g, 1.17 mmol) in methanol (30 mL)
was hydrogenated at 3 atm for 3 h in presence of Pd/C
10% (50 mg). At the end of the reaction, the mixture
was filtered and the solvent was removed under reduced
1
pressure. Yield 0.30 g (77%); H NMR: d 9.30 (2H, s),
7.40 (1H, s), 7.10 (3H, m), 6.90 (1H, s), 6.82 (5H, m),
6.65 (4H, m), 5.72 (1H, s), 3.22 (2H, s).
2-(4,40-Dihydroxybenzidryl)phenylacetic acid (10).
suspension of
2-(4,40-dihydroxybenzidryl)phenyl-
A
acetamide 10g (0.30 g, 0.90 mmol) in NaOH 30% was
refluxed for 4 h; afer cooling, the solution was acidified
(HCl 5 N) and the precipitate thus formed was collected
and washed with water. Yield 0.20 g (67%); mp 180–
181 ꢁC; IR (Nujol) (cmꢀ1): 1689; 1H NMR: d 12.15 (1H,
s), 9.05 (2H, s), 7.15 (3H, m), 6.78(9H, m), 5.56 (1H, s),
3.48(2H, s). Anal. calcd for C 21H18O4: C, 75.43; H,
5.43; found: C, 75.63, H, 5.40.
2-(4,40-Dibenzyloxyibenzidril)benzylalcohol (10c).
A
solution of methyl 2-(4,40-dibenzyloxybenzidril)benzoate
10b (3.69 g,7.18mmol) in anhyd diethyl ether (30 mL)
was added to a suspension of LiAlH4 (0.20 g, 5.38mmol)
in anhyd diethyl ether (70 mL) under stirring. The mix-
ture was stirred at 30 ꢁC for 900, then ethyl acetate was
added, followed by water (30 mL) and H2SO4 (10%, 50
mL). The mixture was then extracted with diethyl ether,
the organic layer was dried (Na2SO4) and the solvent
removed under reduced pressure. Yield 3.00 g (86%), oil;
1H NMR: d 7.40 (12H, m), 6.92 (10H, s), 5.72 (1H, s),
2-(20) and 4-Nitro-1-phenylthionaphtalene (21). 2-nitro-
40-methoxydiphenylsulphide (24). A solution of thiophenol
(for 20 and 21) or 4-methoxythiophenol (for 24) (5.63

3930
L. Costantino et al. / Bioorg. Med. Chem. 10 (2002) 3923–3931
mmol) in EtOH (2 mL) was added to a solution of the
appropriate aryl halide (5.00 mmol) in EtOH (10 mL),
then pulverized KOH (5.63 mmol) was added. The
solution was warmed at 70ꢁC for 4 h, cooled and filtered.
Then water was added to the filtrate and the mixture
was extracted with CH2Cl2 (2 ꢂ 20 mL); the organic
layer was washed with KOH 2 N, dried (Na2SO4) and
the solvent removed under reduced pressure. The residue
was then purified by means of column chromatography
(cyclohexane/EtoAc 87.5/12.5).
avoid evaporation. Prior to energy-minimization of the
complexes, only the water molecules were energy-mini-
mized and then subjected to 20ps of molecular dynamics
at 300 K, in order to let the solvent equilibrate around
the enzyme-inhibitor structure. Then, 5000 steps of
conjugate gradient minimization were performed, with
all protein residues within 10 A from the inhibitor and
all the water molecules allowed to move during mini-
mization. A 10 A non-bonded cutoff was adopted in the
simulations. Molecular dynamics of the ALR2-inhibitor
complexes was then performed at 300 K for over 400 ps.
The complexes were gradually heated to 300 K during
the first 20 ps, in order to avoid abrupt changes of
structure. SHAKE27 was turned on during molecular
dynamics, and a time-step of 2.0 fs was used. Coor-
dinates were collected and averaged every 10 ps,
using CARNAL. The last 10ps were re-optimized
with 5000 steps of conjugate gradient minimization.
Structures were visualized with the computer graphics
software MidasPlus.28 Calculations were performed
on a IBM-SP3 computer, and graphic display was
performed on SGI O2 workstations. Other details can
be found in ref 7.
2-Nitro-1-phenylthionaphtalene (20). Yield 39%; mp 85–
ꢁ
1
88 C; H NMR (CDCl3): d 8.58 (1H, m), 8.02 (2H, m),
7.66 (3H, m), 7.17 (5H, m). Anal. calcd for
C16H11NO2S: C 68.31; H, 3.94; N, 4.98; found: C, 68.00,
H, 3.98; N, 5.00.
4-Nitro-1-phenylthionaphtalene (21). Yield 29%; mp
ꢁ
1
100–103 C; H NMR (CDCl3): d 8.69 (1H, m), 8.46
(1H, m), 8.09 (1H, d, J 8.26), 7.74 (2H, m), 7.53 (5H,
m), 7.10 (1H, d, J 8.26). Anal. calcd for C16H11NO2S: C
68.31; H, 3.94; N, 4.98; found: C, 68.05, H, 3.81; N,
4.90.
2-Nitro-40-methoxydiphenylsulphide (24). Yield 78%; mp
Enzyme section
ꢁ
1
95–96 C; H NMR (CDCl3): d 8.24 (1H, m), 7.54 (2H,
m), 7.35 (1H, m), 7.21 (1H, m), 7.04 (2H, m), 6.87 (1H,
m), 3.95 (3H, s). Anal. calcd for C13H11NO3S: C 59.76;
H, 4.24; N, 5.36; found: C, 59.81, H, 4.22; N, 5.33.
Sorbinil was a gift from Pfizer; Tolrestat was synthe-
sized according to a published procedure.29 IC50 value
for Sorbinil was determined several times (n=12): IC50
(+/ꢀS.D.): 1.45 mM (+/ꢀ0.33). Calf lenses for the
purification of ALR2 were obtained from freshly-
slaughtered animals and kept to ꢀ20 ꢁC until needed.
The capsule was incised and the frozen lenses were sus-
pended in sodium phosphate buffer pH 7.00 containing
5 mM DTT (1 g tissue/3.5 mL) and stirred in an ice-cold
bath for 1 h. The suspension was then centrifuged at
22,000 g at 4 ꢁC for 40 min and the supernatant was
subjected to ion exchange chromatography (DE52).9
Enzyme activity was measured by monitoring the
change in absorbance at 340 nm which accompanies the
oxidation of NADPH catalyzed by ALR2. The assay
was performed at 37 ꢁC as previously described,9 using
4.7 mM d,l-glyceraldehyde as substrate in 0.25 M
sodium phosphate buffer pH 6.80, containing 0.38 M
ammonium sulphate and 0.11 mM NADPH. The sensi-
tivity of ALR2 inhibition by different ARIs was tested in
the above assay conditions by adding a DMSO solution
of the inhibitors in the reaction mixture. Mother solutions
of inhibitors were prepared in DMSO at such concentra-
tion that, after the addition into the assay mixture, DMSO
is at constant concentration of 1%. A reference blank
containing all the above reagents except the substrate
was used to correct for the non-enzymatic oxidation of
NADPH. IC50 values (the concentration of the inhibitor
required to produce 50% inhibition of the enzyme cat-
alyzed reaction) were determined from least squares
analyses of the linear portion of the log dose-inhibition
curves. Each curve was generated using at least three
concentrations of inhibitor causing an inhibition
between 20 and 80% with two replicates at each con-
centration. The 95% confidence limits (95% C.L.) were
calculated from T values for nꢀ2, where n is the total
number of determinations.30
Computational procedure
The aldose reductase structure used for molecular
modelling analysis is the one previously obtained after
docking and molecular dynamics simulations of a com-
plex between ALR2 and the lead compound 4-(20,40-dini-
tro-anilino)phenol previously described.7 The structures
of the ALR2 complexes with 4-(20-nitrophenyl-
mercapto)phenol and 4-(30-nitrophenylmercapto)phenol
bound at the active site were firstly energy-minimized
with the AMBER 520 program using the Cornell21 et al.
force field, and then equilibrated with molecular
dynamics. To calculate the partial atomic charges of
inhibitors, the geometries of the molecules were com-
pletely optimized using the AM1Hamiltonian, and
charges were calculated with electrostatic potential fits
to a 6-31G* ab-initio wave function using Gaussian98,
followed by a standard RESP22,23 fit. Van der Waals
parameters of the inhibitors were assigned to be con-
sistent with the Cornell force field. Parameters for
NADP+ were taken from our previous simula-
tions.7,24,25 Dihedral parameters consistent with the
Cornell et al. parameterization were assigned and, in
some cases, derived from a conformational analysis
performed with AM1. For each molecule, energy mini-
mization was performed to make sure that the opti-
mized conformation was in agreement with the
AM1-optimized conformation.
The ALR2-inhibitor complexes were solvated with
spherical caps of TIP3P26 water molecules within 24A of
the inhibitors, resulting in 460 water molecules. Har-
monic radial forces (1.5 kcal/mol A2) were applied to

L. Costantino et al. / Bioorg. Med. Chem. 10 (2002) 3923–3931
3931
References and Notes
18. Beilstein 10, II, 71 h.
19. Beilstein 6 I 420 j.
1. Yabe-Nishimura, C. Pharmacol. Rev. 1998, 50, 21.
2. Wilson, D. K.; Tarle, I.; Petrash, J. M.; Quiocho, F. A.
Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 9847.
3. Urzhumtsev, F.; Tete-Favier, A.; Mitscler, J.; Barbanton,
P.; Barth, L.; Urzhumgtseva, J. F.; Biellmann, A. D.; Pod-
jarny, D.; Moras, D. A. Structure 1997, 5, 601.
20. Case, D.A.; Pearlman, D. A.; Caldwell, J. W.; Cheatham
III, T.E.; Ross, W.S.; Simmerling, C. L.; Darden, T.A.; Merz,
K.M.; Stanton, R.V.; Cheng, A.L.; Vincent, J.J.; Crowley, M.;
Ferguson, D.M.; Radmer, R.J.; Seibel, G.L.; Singh, U.C.;
Weiner, P.K.; Kollman, P.A. Amber 5. University of Cali-
fornia, San Francisco, 1997.
4. El-Kabbani, O.; Wilson, D. K.; Petrash, J. M.; Quiocho,
F. A. Mol. Vision 1998, 4, 19.
21. Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.;
Merz, K. M.; Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.;
Caldwell, J. W.; Kollman, P. A. J. Am. Chem. Soc. 1995, 117,
5179.
22. Bayly, C. I.; Cieplak, P.; Cornell, W. D.; Kollman, P. A.
J. Phys. Chem. 1993, 97, 10269.
23. Cieplak, P.; Bayly, C. I.; Cornell, W. D.; Kollman, P. A.
J. Comput. Chem. 1995, 16, 1357.
24. Rastelli, G.; Vianello, P.; Barlocco, D.; Costantino, L.;
Del Corso, A.; Mura, U. Bioorg. Med. Chem. Lett. 1997, 7,
1897.
25. Costantino, L.; Rastelli, G.; Vescovini, K.; Cignarella, G.;
Vianello, P.; Del Corso, A.; Cappiello, M.; Mura, U. J. Med.
Chem. 1996, 39, 4396.
26. Jorgensen, W. L.; Chanddrasekhar, J.; Madura, J. P.;
Impey, R. W.; Klein, M. L. J. Chem. Phys 1983, 79, 926.
27. Van Gusteren, W. F.; Berendsen, H. J. C. Mol. Phys.
1977, 34, 1311.
5. 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.; Pod-
jarni, A. J. Am. Soc. Mass Spectrom. 1999, 10, 635.
6. Singh, S. B.; Malamas, M. S.; Hohman, T. C.; Nilakantan,
R.; Carper, D. A.; Kitchen, D. J. Med. Chem. 2000, 43, 1062.
7. Rastelli, G.; Ferrari, A. M.; Costantino, L.; Gamberini,
M. C. Bioorg. Med. Chem. 2002, 10, 1437.
8. Eggler, J.F.; Larson, E.R.; Lipinski, C.A.; Mylari, B.L.;
Urban, F.J. In Advances in Medicinal Chemistry; Jai Press:
Greenwich, CT, 1993; Vol. 2, pp 197–246.
9. Costantino, L.; Rastelli, G.; Gamberini, M. C.; Vinson, J. A.;
Bose, P.; Iannone, A.; Staffieri, M.; Antolini, L.; Del Corso, A.;
Mura, U.; Albasini, A. J. Med. Chem. 1999, 42, 1881.
10. Livant, P.; Martin, J. C. J. Am. Chem. Soc. 1977, 99, 5761.
11. Humber, L. G. Prog. Med. Chem. 1987, 24, 299.
12. Foss, N. E.; Stehle, J. J.; Shusett, H. M.; Hadburg, D. J.
Am. Chem. Soc. 1938, 60, 2729.
28. Ferrin, T. E.; Huang, C. C.; Jarvis, L. E.; Langridge, R. J.
Mol. Graphics 1988, 6, 13.
13. Schultz, O. E.; Geller, L. Arch. Pharm 1955, 234.
14. Hodgson, H. H.; Kilner, E. J. Chem. Soc 1925, 7.
15. Ullman, F.; Bruck, W. Berichte 1908, 3932.
29. Sestanj, K.; Bellini, F.; Fung, S.; Abraham, N.; Treasur-
ywala, A.; Humber, L.; Duquesne, N. S.; Dvornick, D. J.
Med. Chem. 1984, 27, 255.
16. Singer, H.; Shive, W. J. Am. Chem. Soc. 1955, 77, 5700.
17. Elson, L. A.; Gibson, C. S.; Johnson, J. D. A. J. Chem.
Soc 1929, 2735.
30. Tallarida, R. J.; Murray, R. B. Manual of Charmacologi-
cal Calculations with Computer Programs, 2nd ed; Springer
Verlag: New York, 1987.