Development of Novel Oxotriazinoindole Inhibitors of Aldose Reductase: Isosteric Sulfur/Oxygen Replacement in the Thioxotriazinoindole Cemtirestat Markedly Improved Inhibition Selectivity

Article abstract of DOI:10.1021/acs.jmedchem.9b01747

Inhibition of aldose reductase (AR), the first enzyme of the polyol pathway, is a promising approach in treatment of diabetic complications. We proceeded with optimization of the thioxotriazinoindole scaffold of the novel AR inhibitor cemtirestat by repla

Full text of DOI:10.1021/acs.jmedchem.9b01747

Subscriber access provided by BIU Pharmacie | Faculté de Pharmacie, Université Paris V
Article
Development of novel oxotriazinoindole inhibitors of aldose
reductase: isosteric sulfur/oxygen replacement in the
thioxotriazinoindole cemtirestat markedly improved inhibition selectivity
Matus Hlavac, Lucia Kovacikova, Marta Soltesova Prnova, Peter Šramel, Gabriela
Addova, Magdalena Majekova, Gilles Hanquet, Andrej Bohac, and Milan Stefek
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b01747 • Publication Date (Web): 10 Dec 2019
Downloaded from pubs.acs.org on December 10, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
Development of Novel Oxotriazinoindole Inhibitors of Aldose Reductase:
Isosteric Sulfur/Oxygen Replacement in the Thioxotriazinoindole
Cemtirestat Markedly Improved Inhibition Selectivity
7
8
9
Matúš Hlaváč1, Lucia Kováčiková2, Marta Šoltésová Prnová2, Peter Šramel1, Gabriela
Addová3, Magdaléna Májeková2, Gilles Hanquet4, Andrej Boháč1,5 and Milan Štefek2*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1 Department of Organic Chemistry, Faculty of Natural Sciences, Comenius University in Bratislava, Ilkovičova
6, 842 15 Bratislava, Slovakia
2 Institute of Experimental Pharmacology and Toxicology, CEM, SAS, Dúbravská cesta 9, 841 04 Bratislava,
Slovakia
3 Institute of Chemistry, Faculty of Natural Sciences, Comenius University in Bratislava, Ilkovičova 6, 842 15
Bratislava, Slovakia
4
Université de Strasbourg, Université de Haute-Alsace, CNRS, UMR 7042-LIMA, ECPM, 25 rue Becquerel,
67087 Strasbourg, France
5 Biomagi, Inc., Mamateyova 26, 851 04 Bratislava, Slovakia
Highlights:
• A series of novel aldose reductase inhibitors was designed and prepared.
• Isosteric replacement of sulfur by oxygen in the thioxotriazinoindole scaffold markedly
improved inhibition selectivity.
• Chemical syntheses of both the sulfur parent drug cemtirestat and oxygen congeners are
reported.
• Compound 9 displayed IC50 value of 42 nM against rat eye lens aldose reductase with
selectivity factor > 2 380 relative to rat kidney detoxification aldehyde reductase.
Abstract
Inhibition of aldose reductase (AR), the first enzyme of the polyol pathway, is a promising
approach in treatment of diabetic complications. We proceeded with optimization of the
thioxotriazinoindole scaffold of the novel aldose reductase inhibitor cemtirestat by
replacement of sulfur with oxygen. A series of 2-(3-oxo-2H-[1,2,4]triazino[5,6-b]indol-
5(3H)-yl)acetic acid derivatives (OTIs), designed by molecular modeling and docking, was
synthesized. More electronegative and less bulky oxygen of OTIs compared to the sulfur of
the original thioxotriazinoindole congeners was found to form stronger H-bond with Leu300
of AR and to render larger rotational flexibility of the carboxymethyl pharmacophore. AR
inhibitory activities of the novel compounds were characterized by the IC50 values in a
submicromolar range. Markedly enhanced inhibition selectivity relative to the structurally
related aldehyde reductase was recorded. To conclude, structure modification of the original
carboxymethylated thioxotriazinoindole cemtirestat by isosteric replacement of sulfur with
oxygen in combination with variable N(2) simple substituents provided novel analogues with
increased AR inhibition efficacy and markedly improved selectivity.
Key words: Cemtirestat; Carboxymethylated thioxotriazinoindole; Oxotriazinoindolyl acetic
acid; Aldose reductase inhibition and selectivity; Diabetic complications.
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 48
1
2
3
4
INTRODUCTION
5
6
7
8
9
Despite significant efforts that have been made in the past decade to determine the precise
physiological role of aldose reductase (AR, ALR2), our understanding of this enzyme remains
limited. AR as the first enzyme of the polyol pathway is considered a major component in the
etiology of chronic diabetic complications. Under hyperglycemic conditions, glucose
reduction is catalyzed by AR to sorbitol. Sorbitol accumulation in cells and related metabolic
derangements result in cataract, damage of peripheral nerves or kidneys and cardiovascular
complications. Therefore, efficient AR inhibitors are sought as potential remedies to prevent
diabetic complications.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In our previous study, carboxymethylated thioxotriazinoindoles were identified as AR
inhibitors with high efficacy and selectivity.1 Among the novel compounds, 2-(3-thioxo-2H-
[1,2,4]triazino[5,6-b]indol-5(3H)-yl)acetic acid (4) (cemtirestat, CMTI, Fig. 1 and 3) was
found the most promising inhibitor endowed also with antioxidant activity.2,3 Considering
excellent “lead-likeness” of cemtirestat (4), we proceeded with optimization of its
thioxotriazinoindole scaffold by replacement of sulfur with oxygen, with the aim to improve
the inhibitory efficacy and selectivity. Based on preliminary molecular modeling and docking
calculations, a series of 2-(3-oxo-2H-[1,2,4]triazino[5,6-b]indol-5(3H)-yl)acetic acid
derivatives (OTIs) 9-10, 12, 14 was proposed, synthesized and their AR inhibitory efficacy
and selectivity determined.
2
ACS Paragon Plus Environment

Page 3 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
NH
N
N
S
N
8
9
HO
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
Figure 1. 2-(3-Thioxo-2H-[1,2,4]triazino[5,6-b]indol-5(3H)-yl)acetic acid (4, cemtirestat,
CMTI)1
.
RESULTS
Drug design. We were inspired by „S“ and „O“ series of fused tricyclic aldose reductase
inhibitors patented by Toyooka et al.4 (Fig. 2). The inhibition efficacy in a low micromolar
range was reported for these compounds.
X
X = N, O, S
Y = O, S
Y
N
HOOC
Figure 2. Structure of published tricyclic aldose reductase inhibitors.
In our initial molecular modeling and docking investigation, we were interested how
isosteric replacement of S in cemtirestat (4) by O would affect interaction with AR. In our
previous study, three main tautomers of cemtirestat (4) were specified.1 Their structures along
with the analogical tautomers of an oxygen variant of cemtirestat, compound 9, are depicted
in Fig. 3.
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 48
1
2
3
4
5
6
N
NH
N
N
N
N
N
Z
Z
ZH
NH
N
N
N
N
7
8
9
O
O
O
O
O
II
O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
I
III
4
9
Z = S (compound ) or O (compound )
Figure 3. Structures of the most relevant tautomers of cemtirestat (4) and its O-analogue 9.
Energy preferences of the individual tautomers calculated by quantum-mechanical
geometry optimization are summarized in Table 1. The order of the relative energies indicates
the preference of tautomer II for both 4 and 9 in H2O or DMSO solvents. Different polarities,
aromaticity and energy of bonds of individual tautomers are most likely responsible for the
observed energy variances calculated in vacuum, water and DMSO, as shown in Table 1. All
three tautomeric forms were docked and optimized in PDB: 4QX4 model of human
recombinant AKR1B1. For compound 9, tautomeric forms I and II provided similar
positions, yet the second one was energetically more favorable. Accordingly, molecular
dynamics simulations proved that tautomer II of cemtirestat (4) gave a more stable complex
with AKR1B1.5
Fig. 4 shows superposition of cemtirestat (4) and its O-analogue 9 in the binding site of
AKR1B1. Polar carboxymethyl groups of both compounds were found to align well. Slight
distortion of the fused planar aromatic system was apparently caused by higher rotational
flexibility of the carboxymethyl moiety of 9 (in cyan) owing to less bulky oxygen compared
to the sulfur of the original thioxotriazinoindole 4 (in red, van der Waals radius of O is 1.4 Å,
while for S it is 1.9 Å). As a result, compound 9 was able to create stronger H-bond with
Leu300 (contact distance 2.8 Å vs. 3.2 Å) and gave more favorable hydrophobic interaction
(2.0 kJ/mol for 9 vs. 0.6 kJ/mol for cemtirestat 4) with NADP+ (contact distance 3.6 Å vs. 3.8
4
ACS Paragon Plus Environment

Page 5 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Å). Compound 9 interacts with the residues of the specificity pocket Phe122 and Leu300
while letting the specificity pocket closed (Fig. 4 and 5), which is in line with the previously
published crystal structure of AKR1B1 complexed with cemtirestat 4 (PDB: 4QX4).1 These
preliminary in silico findings provided a promise of better fitting of the O-derivative 9 into
the binding site of AR when compared with its parent molecule of cemtirestat (4).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. Relative Energy Preferences of Individual Tautomers of Cemtirestat (4) and its O-
analogue 9
Relative energies (kJ/mol)*
Tautomer
Cemtirestat (4)
Compound 9
vacuum
0.4
water
41.0
0
DMSO
vacuum
0
water
47.0
0
DMSO
23.5
0
I
18.8
0
II
0
2.5
III
33.8
51.9
49.0
38.9
39.4
42.5
* Values of total energies, calculated as described in the experimental section, related by
subtraction to the most stable tautomer in a given environment
Chemistry. A commercially available compound, 2-(3-thioxo-2H-[1,2,4]triazino[5,6-
b]indol-5(3H)-yl)acetic acid (4) [CAS: 309283-89-4], recently identified as ALR2 inhibitor1
and reported also under designation of cemtirestat,3,6 is to our knowledge devoid of any
literature report on its chemical synthesis.7 Therefore we developed a procedure for its
preparation from isatine (1). Compound 1 was alkylated to 2 by deprotonation using calcium
hydride,8 followed by addition of an excess of ethyl chloroacetate in DMF and stirring at
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 48
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Superposition of 9 and cemtirestat (4) in the binding site of AKR1B1 (PDB:
4QX4). Cemtirestat (4) is colored in red and the corresponding amino acid residues in orange.
Compound 9 is shown in cyan and the complementary amino acid residues in dark blue. H-
bonds are indicated as yellow and hydrophobic interactions as green lines.
Fig
Figure 5. Positions of cemtirestat 4 (left, PDB: 4QX4) and compound 9 (right, predicted) in
the cavities of the AKR1B1 binding site. The yellow lines closer to an active site entrance
label contact distances with Leu300 of the specificity pocket.
6
ACS Paragon Plus Environment

Page 7 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
100°C within 5 hours. The resulting compound 2 underwent condensation with
thiosemicarbazide in DMF at 100 °C to give intermediate 3. Finally, the latter was
refluxed in an aqueous solution of potassium carbonate during two days to afford the ALR2
inhibitor 4 in overall yield 39 %. (Scheme 1)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
HS
NH
H
N
NH
N
O
O
N
S
a
c
b
O
O
O
N
N
N
N
N
O
61 %
82 %
78 %
H
O
O
HO
cemtirestat
O
O
1
2
3
4
(isatine)
(overyield 39 %)
Scheme 1. Preparation of cemtirestat (4) from isatine (1). Reagents and conditions: (a) (1)
CaH2 1.10 mol eq, DMF (abs), 100 °C, 1 h; (2) ethyl chloroacetate 3.00 mol eq, 40 °C to 100
°C, 5 h. (b) Thiosemicarbazide 1.20 mol eq, DMF (abs), 100 °C, 1 h. (c) K2CO3 15 % aq sol,
reflux, 48 h.
The synthetic route to the other computer predicted compounds 9 and 10 is summarized in
Scheme 2. Exploitation of the synthetic pathway applied for the sulfur containing compound
cemtirestat (4) was unproductive in preparation of its oxo-analogue 9, since 2-
carbamoylhydrazono-containing intermediate (a tautomeric oxo-isoster of 3 from Scheme 1,
its structure is not given) did not cyclize in the basic conditions. Therefore, in this case isatine
(1) was refluxed with semicarbazide hydrochloride in ethanol9 to give compound 5 in 96 %
yield10-12 which upon reflux with aqueous solution of sodium hydroxide (1.0 M) overnight
lead to triazinedione 6 in 69 % yield13 (Scheme 2).
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 48
1
2
3
4
5
O
H
N
NH2
O
NH
N
N
O
NH
N
N
O
6
7
8
9
NH
b
a
c
O
N
O
O
N
N
N
96 %
69 %
88 %
NH2
H
H
H
*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
(isatine)
5
6
7
NH
N
NH
NH
N
N
N
O
d
S
O
N
N
(22 %
and
20 %)
g
e
N
N
89 %
63 %
HO
HO
O
O
O
O
8a
4 (cemtirestat)
9
1
4
(overyields: 8 % from
and 35 % from via )
+
1
O
O
O
HO
N
N
O
f
N
N
N
O
N
O
N
63 %
N
O
8b
HO
O
10
(overyield 7 %)
* see Scheme 1
Scheme 2. Synthesis of predicted inhibitors 9 and 10 from isatine 1. Reagents and conditions:
(a) Semicarbazide hydrochloride 1.20 mol eq, EtOH, reflux, 30 min. (b) NaOH (aq 1.00 M),
reflux, overnight. (c) Glacial acetic acid, reflux, 3 h. (d) (1) CaH2 2.00 mol eq, DMF (abs), 60
°C, 40 min; (2) ethyl chloroacetate 2.50 mol eq, 50 °C, 3 h; separated by FLC on SiO2. (e)
NaOH 2.00 mol eq, MeOH / H2O : 4 / 1, reflux, 2 h. (f) NaOH 3.50 mol eq, MeOH / H2O : 4 /
1, reflux, 1.5 h. (g) (1) NaOH 2.00 mol eq, H2O; (2) KMnO4 2.00 mol eq, NaOH, 5.50 mol
eq, H2O, rt, 30 min.
Acidic treatment of 6 followed by alkylation14 of the resulting triazino-indole derivative 7
with ethyl chloroacetate afforded a 1 : 1 mixture of mono and dialkylated products 8a and 8b
in relatively low yields of 22 and 20 %, respectively, after FLC separation on SiO2. Finally,
hydrolysis of 8a and 8b under basic conditions delivered 9 and 10 in 63 % yield for both.
8
ACS Paragon Plus Environment

Page 9 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
However, overall yield of the target compound 9 was largely improved (35 % vs 8 %) by
direct oxidative-hydrolysis of 4 by potassium permanganate in an aqueous solution of sodium
hydroxide.9 Compound 8a represents an important joint intermediate for the synthesis of other
predicted inhibitors 12 and 14 which both were prepared from 1 (over six steps) in 5 and 3 %
yields according to the Scheme 2 and 3. Compound 8a was alkylated to corresponding N-
benzyl- and N-MOM- derivatives 11 and 13, respectively, which were subsequently
saponified by basic conditions to the compounds 12 and 14.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
N
O
N
N
N
NH
N
N
N
N
N
N
N
O
O
O
O
O
a
c
b
d
N
N
N
80 %
60 %
47 %
N
N
N
44 %
N
N
O
HO
O
O
HO
O
O
O
O
O
14
13
(overyield 3 %
12
8a
(overyield 5 %
11
1
)
from
1
)
from
Scheme 3. Synthesis of 12 and 14 from the joint intermediate 8a. Reagents and conditions:
(a) (1) CaH2 2.00 mol eq, DMF (abs), 50 °C, 30 min; (2) benzyl chloride 4.00 mol eq, 80 °C,
4 h. (b) NaOH 2.00 mol eq, MeOH / H2O : 3 / 1, reflux, 2 h. (c) Methoxymethylchloride 5.00
mol eq, Hünig´s base (N,N-diisopropylethylamine, DIPEA) 5.00 mol eq, acetone (abs) from rt
to 50 °C, 5h. (d) NaOH 2.00 mol eq, MeOH / H2O : 3 / 1, 80 °C, 1 h.
Finally, compounds 15 and 17 were prepared from their precursors 8a and 9, respectively.
Esters 15 and 17 were thought to be precursors of acyl-carboxylic acids 15a and 17a as
predicted inhibitors providing an additional hydrogen bond in interaction with ALR2 enzyme.
Compound 15 was obtained in 79 % yield by acetylation of 8a. The acyl-triazinone 17 was
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 48
1
2
3
4
5
6
7
8
9
prepared from 9 that underwent almost quantitative esterification to 16 followed by acylation
with pivaloyl chloride to 17 in 79 % yield. Unfortunately, both compounds 15 and 17 were
quite unstable due to their labile acyl group bounded on annulated and partially aromatic
1,2,4-triazine-3-one ring (its general mesomeric aromatic structure is depicted in Scheme 4 in
parenthesis). Thus hydrolysis of 15 and 17 to 15a and 17a was impossible and we have not
been able to screen both predicted inhibitors on ALR2.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
R
O
O
O
N
N
N
N
N
NH
N
N
N
O
O
O
O
a
N
N
N
N
N
N
N
79 %
O
HO
O
O
15a
O
O
O
O
15
8a
R: Me, tBu
O
O
N
NH
N
N
N
NH
N
N
N
N
O
O
O
O
c
b
N
N
99 %
79 %
N
O
N
O
N
O
N
HO
HO
O
O
O
17a
9
16
17
Scheme 4 Synthesis of 15 and 17 from the intermediate 8a and one of the target compounds
9. Reagents and conditions: (a) Acetyl chloride 2.00 mol eq, N,N-dimethylaminopyridine
(DMAP) 1.00 mol eq, Hünig base (N,N-diisopropylethylamine, DIPEA) 1.60 mol eq, CH2Cl2,
rt to 40 °C, 1 h. (b) Solution of diazomethane in Et2O. (c) Pivaloyl chloride 2.00 mol eq,
DMAP 1.00 mol eq, Hünig´s base 1.50 mol eq, CH2Cl2, rt to 30 °C, 15 min.
Detailed characterizations and spectra of all the products mentioned (Scheme 1- 43) are
described in a supplementary material to this paper.
10
ACS Paragon Plus Environment

Page 11 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Enzyme inhibition. The compounds 9, 10, 12 and 14 were assessed for their ability to
inhibit the in vitro reduction of D,L-glyceraldehyde by ALR2 purified from rat lens, using
cemtirestat (4) as a reference (Table 2). Higher IC50 values (1.2 - 1.9 times) were recorded in
the presence of 1 % DMSO in the solvent.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In the next step, we evaluated the inhibition selectivity in relation to the closely related
antitarget aldehyde reductase (ALR1). The inhibition data of OTI compounds 9, 10, 12 and 14
for their inhibition of the reduction of D-glucuronate substrate by ALR1 purified from rat
kidneys are shown in Table 2 together with the selectivity factors calculated [SF = IC50
(ALR1) / IC50 (ALR2)].
For comparison, inhibition data of epalrestat, well known inhibitor of ALR2,15 and
valproic acid, inhibitor of ALR1,16 obtained under the identical experimental conditions as
those for the compounds studied, are also included in Table 2.
Moreover, inhibition of human recombinant AKR1B1 was assayed. The data shown in
Table 3 were obtained under identical reaction conditions as applied for the rat ALR2 assay,
with the protein concentration of 5 μg/mL.
Inhibition of human recombinant AKR1B10 by 9 was tested with 5 mM DL-
glyceraldehyde as a substrate, 0.12 mM NADPH cofactor and 10 μg/mL protein
concentration, yielding an IC50 value shown in Table 3.
In addition, inhibition of ALR2-mediated reduction of physiologically relevant substrates,
methylglyoxal, 4-hydroxynonenal (HNE), and 4-hydroxynonenal glutathione (GS-HNE), by 9
and 4 was determined. In the first step, Michaelis-Menten constants Km were determined for
each substrate. Then the substrate concentrations were set to be minimally 10 times higher
than the corresponding Km, to saturate the active site of the enzyme, and IC50 values were
determined for both 9 and cemtirestat (4) as shown in Table 4.
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 48
1
2
3
4
Table 2. Enzyme Inhibition Data
5
6
7
8
9
O
O
N
N
HO
N
H
N
N
N
N
N
N
NH
N
N
N
S
O
O
O
O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N
N
N
N
N
N
N
HO
HO
HO
HO
HO
O
O
O
O
O
12
14
cemtirestat (4)
9
10
ALR2
IC50 (nM)
ALR1
SFH2O
IC50 (µM)/I(%,100 µM)
Compound
IC50(ALR1)/IC50(ALR2)
1 % DMSO
H2O
H2O
9
51 ± 1
233 ± 13
787 ± 8
118 ± 3
176 ± 13
n.d.
42 ± 1
> 100 / 24.0
20.2 ± 0.4 / n.d.
> 100 / 37.8
> 100 / 13.4
35.0 ± 1.6 / n.d.
40.6 ± 2.0 / n.d.
n.d.
> 2380
168
> 230
> 1177
302
418
-
10
12
14
120 ± 5
434 ± 14
85 ± 1
116 ± 8
97 ± 19
n.d.
Cemtirestat (4)
Cemtirestat*
Epalrestat**
Valproic acid*
227 ± 19
n.d.
n.d.
56.1 ± 2.7
-
*Stefek et al (2015),1 **Majekova et al. (2017),17 n.d.- not determined.
SFH2O means selectivity factor calculated from IC50 data measured in water. Results are mean
values from two measurements or mean values ± SD from at least three measurements. I, %
of enzyme inhibition observed at 100 µM inhibitor concentration. All measurements were
performed by using the same enzyme batch of either ALR1 or ALR2.
12
ACS Paragon Plus Environment

Page 13 of 48
Journal of Medicinal Chemistry
1
2
3
4
Table 3. Inhibition of Human Recombinant Enzymes by 9 in Comparison with Cemtirestat
5
6
7
8
9
Human
IC50 (nM)
recombinant
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
enzyme
9
Cemtirestat (4)
102 ± 5
AKR1B1
66 ± 2
AKR1B10
56 240 ± 5 758
21 400 ± 2 000*
*Stefek et al (2015)1
The experiments were performed in the absence of DMSO. Results are mean values ± SD
from at least three measurements.
Table 4. Inhibition of ALR2-Mediated Reduction of Three Physiologically Relevant
Substrates by 9 and Cemtirestat (4) in Comparison with D,L-Glyceraldehyde Substrate
IC50 (nM)
Substrate (mM)
Km (µM)
9
Cemtirestat (4)
116 ± 8
D,L-Glyceraldehyde (5.0)
Methylglyoxal (0.5)
HNE (0.5)
239 ± 66
38.5 ± 3.3
32.0 ± 3.9
84.7± 2.5
42 ± 2
40 ± 1
146 ± 29
216 ± 14
292 ± 25
644 ± 16
GS-HNE (1.0)
1157 ± 78
Results are mean values ± SD from at least three measurements.
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 48
1
2
3
4
5
6
7
8
9
No significant inhibition of sorbitol dehydrogenase activity (the second enzyme of the
polyol pathway) by 100 μM concentration of 9 was recorded [I (%) = 0.9 ± 0.6, n = 3].
The inhibitory effect of 9 on sorbitol accumulation in isolated rat eye lenses was evaluated.
As shown in Fig. 6, significantly increased sorbitol levels were shown in the isolated rat eye
lenses incubated with glucose, C(+), in comparison with control incubations without glucose,
C(-). Sorbitol accumulation was significantly and in a concentration-dependent way inhibited
by compounds 9, starting at a concentration as low as 10 µM (Fig. 6).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
600
500
400
**
300
***
200
100
0
***
C (-)
C (+)
1 (+)
10 (+)
50 (+)
100 (+)
Compound 9 (µM)
Figure. 6 Effect of compound 9 on accumulation of sorbitol in isolated rat eye lenses
incubated in the presence of high concentrations of glucose. C(-), control incubation without
glucose; C(+), control incubation with 50 mM of glucose; 1(+), 10(+), 50(+) and 100(+)
incubations with increasing μM concentrations of compound 9 in the presence of 50 mM
glucose. Results are mean values ± SEM from at least five independent incubations. ** p ≤
0.01 vs. C(+), *** p ≤ 0.001 vs. C(+); one-way ANOVA followed by the post-hoc Tukey
multiple comparison test.
14
ACS Paragon Plus Environment

Page 15 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Antioxidant activity. The DPPH test was employed to evaluate the radical scavenging
efficacy of compound 9 in comparison with that of cemtirestat (4). The time course of the
specific absorbance decrease of the ethanolic solution of DPPH (a free radical of 2,2-
diphenyl-1-(2,4,6-trinitrophenyl)hydrazyl, CAS [1898-66-4]) at 518 nm in the presence of 9,
14, 4 and standard melatonin is plotted in Figure 7. Considerably decreased scavenging
activity of 9 was recorded in comparison with that of cemtirestat (4). Yet the initial velocity of
DPPH radical quenching by 9 was still higher than that of standard melatonin, Table 5).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0
60
120
180
240
Time (min)
Figure 7. Time course of absorbance drop of ethanolic solution of DPPH radical (50 μM) in
the presence of 200 μM concentration of the compounds evaluated at λmax = 518 nm. ( )
Blank, (×) melatonin, (○) compound 14, (●) compound 9 and (▲) cemtirestat (4). Typical
curves from measurements performed under identical conditions.
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 48
1
2
3
4
5
6
Table 5. DPPH Test
Compound
C (µM)
-
-ΔA/30 min ± SD
7
8
9
Blank
0.013 ± 0.012
0.057 ± 0.008
0.042± 0.006
0.219 ± 0.005
0.022 ± 0.006
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
9
200
200
200
200
14
Cemtirestat (4)18
Melatonin18
The stable free radical of DPPH (50 μM) dissolved in ethanol of was incubated in the absence
(blank) or presence of 9, 14, 4, or melatonin (200 μM). Absorbance decline at 518 nm was
assessed after the initial 30-minute interval. Results are mean values ± SD from at least three
measurements.
Physicochemical properties and molecular obesity. Molecular obesity indices, generally
used to assess oral bioavailability or “drug-likeness”,19-22 were calculated for the novel
compounds 9, 10, 12, 14 and for cemtirestat (4), taking into account efficacy, size, and
lipophilicity, as shown in Table 6.
DISCUSSION
Aldose reductase (AR, ALR2, AKR1B1), the first enzyme of the polyol pathway, has been
implicated in the pathophysiology of diabetes related complications.23 This enzyme has also
been shown to reduce the carbonyl functional group of lipid peroxidation products, supporting
thus the inflammation as a consequence.24 Aldose reductase inhibitors (ARIs) thus present a
16
ACS Paragon Plus Environment

Page 17 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
Table 6. Physicochemical Properties and Molecular Obesity Indices Calculated for
Compounds 9, 10, 12 and 14 in Comparison with Cemtirestat (4)
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
MW
pKa pKaa pIC50 LogPb LogDa LEc
BEId LLEe LELPf TPSAg
Compoud
ALR2
60-140h
<500h
<5h
>0.3h >14.7h >3.8i
<7.5h
244.21 3.71 4.14 7.38
1.08
-1.75 0.57 30.23 5.67
2.98 100.87
1.98 127.31
9
10
12
302.24 3.59 4.14 6.92
334.33 3.74 4.14 6.36
288.26 3.64 4.14 7.07
0.87
3.11
1.53
-6.06 0.44 22.92 6.05
-0.26 0.36 19.05 3.25
-1.89 0.47 24.55 5.54
8.73
3.25
90.01
99.2
14
Cemtirestat(4)
260.27 3.99 4.30 6.94
1.82
-1.58 0.54 26.68 5.12
3.37
83.8
a Calculated with Pallas 3.112, logD represents the logarithm of the distribution ratio in
octanol-buffer [pH 7]
b Calculated with MarvinSketch Online 2016/ChemAxon
c Ligand efficiency, LE = -1.4log(IC50)/N, N: number of heavy atoms
d Binding efficiency index, BEI = pIC50/MW
e Lipophilic ligand efficiency, LLE = pIC50 - logP
f Ligand efficiency-dependent lipophilicity, LELP = logP/LE.
g Topological polar surface area (Å2), TPSA calculated with MedChem Designer™
h Optimal drug values.19-22
i Mean value for successful lead.21
promising therapeutic approach to treat variety of diabetic complications and inflammatory
disorders including some types of cancer related to chronic inflammation.23,25 ARIs, such as
the acetic acid derivatives (epalrestat, zopolrestat), spiro hydantoins (sorbinil), or spiro
succinimides (minalrestat), have been mainly studied in relation to diabetic complications, yet
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 48
1
2
3
4
5
6
7
8
9
with poor clinical outcome.26-28 At the present time, epalrestat is the sole aldose reductase
inhibitor approved for clinical use, but only in Asia market.29 In the pursuit for more efficient
ARIs of high selectivity and good bioavailability, the interest has been oriented in recent years
towards new chemotypes.30-32
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A recently developed ARI, 2-(3-thioxo-2H-[1,2,4]triazino[5,6-b]indol-5-yl)acetic acid
(cemtirestat, 4), belongs to a group of novel carboxymethylated thioxotriazinoindole aldose
reductase inhibitors of high inhibitory efficacy characterized with an IC50 in submicromolar
scale and high selectivity in relation to congeneric aldo-keto reductases.1-3 High resolution X-
ray crystallography of the human recombinant AKR1B1 enzyme complexed with cemtirestat
(4) demonstrated a unique pattern of cemtirestat binding, with the specifity pocket closed,
contrary to the interaction of the structurally related lidorestat.1 As shown in our latest in vivo
study, cemtirestat (4) normalized symptoms of diabetic peripheral neuropathy in Zucker
Diabetic Fatty rats with high significance.6 Moreover, “lead-like” structural features of
cemtirestat (4) render the compound a favourable candidate for structure optimization.
The rationale for bioisosteric replacement of sulfur with oxygen arose from preliminary
molecular modeling and docking study which pointed to preferable fitting of the O-derivative
into the inhibitor binding site of AR. Indeed, the O-variant of cemtirestat, oxotriazinoindole
(OTI) compound 9, demonstrates significantly increased inhibition efficacy characterized by
IC50 values 2-3 times lower compared to 4, depending on the solvent used. SAR evaluation in
the series of novel OTI derivatives revealed that the presence of simple substituents at N(2)
position of 9 decreased their ALR2 inhibition efficacy. The most remarkable decrease was
recorded for compound 12 with the lipophilic benzyl substituent. Higher IC50 values were
recorded for measurements performed in the presence of DMSO (1 %) compared to assays in
water only. DMSO, which is often used to keep hydrophobic compounds in solution, was
recently reported to act as a weak aldose reductase differential inhibitor.33
18
ACS Paragon Plus Environment

Page 19 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
All compounds evaluated were revealed to be less efficient inhibitors of ALR1
compared to ALR2. With exception of compound 10, the ALR1 IC50 values of the OTI
derivatives (9, 12 and 14) were found to be over 100 µM. For these derivatives, the
percentage inhibition I(%) at 100 µM concentration of the inhibitor was determined.
Estimates of the particular selectivity factors calculated for the most efficient inhibitors 9 and
14 were found to be > 2380 and > 1177, respectively. Striking is an enormous increase of the
selectivity factor in the couple cemtirestat (4) vs 9 (from 302 to > 2380, Table 2). Calculations
of molecular docking into the binding site of ALR1 may offer a feasible explanation: cation-π
interaction of protonated Arg312 in ALR1 with the aromatic ring of O-derivative 9 supported
by two H-bonds keeps this molecule out of reach of NADP+ cofactor in contrast to more
tightly attached cemtirestat 4 (Fig. 52 in Supplement). In addition, a hydrophobic interaction
between the methylene residue of an acetate group of 9 and 4, respectively, and the
nicotinamide ring of NADP+ should also be taken into consideration.34
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Based on the IC50 values shown in Table 2, compound 9 inhibits ALR2 more efficiently
(IC50 = 51 nM) than the reference inhibitor epalrestat (IC50 = 227 nM) while ALR1 activity
was inhibited by 9 less efficiently (IC50 > 100 nM) in comparison with the standard inhibitor
of ALR1 valproic acid (IC50 = 56 nM).
In assaying inhibition of human recombinant AKR1B1, values closely related to those of
the rat ALR2 enzyme were obtained for 9 and cemtirestat (4) (Table 3 vs. Table 2). Markedly
less efficient inhibition of AKR1B10 by 9 compared to AKR1B1 was characterized (Table 3)
by selectivity factor above 850.
Enzyme-inhibitor interactions of selected aldose reductase inhibitors were surveyed.
Interaction diagrams and visualized binding sites of AKR1B1 complexed with cemtirestat,
tolrestat, zopolrestat, epalrestat and fidarestat in comparison with those of compound 9 are
shown in the Supporting information. Experimental values of IC50 for inhibition of AKR1B1,
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 48
1
2
3
4
5
6
7
8
9
numbers of hydrogen bonds with amino acid residues of the binding site (HBAA), numbers of
hydrogen bonds with water molecules (HBw), numbers of π-π interactions, values of the
smallest distance of the hydrophobic interaction between methylene group of the acetate
substituent and nicotine amide ring of NADP+ (dhph, if exists) and the distance of anion - π
interaction (danion-π) between the carboxylic oxygen (or hydantoin oxygen in fidarestat), and
positively charged nicotinamide ring of NADP+ are summarized in Table S1 in Supporting
information.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Aldose reductase is a low selectivity oxidoreductase which can utilize both hydrophilic
and hydrophobic aldehydes as substrates. However, this enzyme appears to be not wholly
admissible since it may discriminate between variable substrates. The ability of diverse
substrates to interact with AR with variable affinity suggests that these molecules may
approach the active site of the enzyme in multiple ways.35 Indeed diverse Km values were
obtained for three alternative substrates studied, methylglyoxal (38.5 μM), HNE (32.0 μM)
and GS-HNE (84.7 μM), in comparison with D,L-glyceraldehyde (239 μM) as shown in
Table 4. Inhibition efficacies of 9 and cemtirestat (4) were determined with AR active site
saturated with the substrates. Under these conditions, both 9 and cemtirestat (4) only poorly
discriminated in their inhibition efficacy between methylglyoxal and D,L-glyceraldehyde
substrates in spite of their markedly different Km values (Table 4). Significantly less efficient
inhibition efficacy of 9 and cemtirestat (4) was recorded for the additional two substrates
studied HNE and GS-HNE. Interestingly, compound 9 inhibits AR-mediated reduction of
methylglyoxal and HNE more efficiently (3.7 and 3.0 times, respectively) than cemtirestat
(4). On one side, this activity may hamper detoxification of these toxic products of
carbohydrate metabolism and lipid peroxidation. On the other hand, more efficient inhibition
of AR-mediated reduction of the GS-HNE adduct by 9 may attenuate the proinflammatory
potential of this metabolic intermediate more efficiently (4 times) than cemtirestat (4). It is
20
ACS Paragon Plus Environment

Page 21 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
well documented that the reduced GS-HNE conjugate takes part in the inflammation
signaling36.
7
8
9
Clinical trials in diabetic patients37-40 revealed the presence of an activated form of
human AKR1B1 with a diminished sensitivity to inhibitors. The only catalytically-relevant
structural variance observed in the activated isozyme was the conversion of residue Cys298 to
serine or alanine. Therefore we have taken the activated form of aldose reductase (PDB:
6F8O)41 to perform docking and optimization of the resulting inhibitor-enzyme complex for
compound 9 and cemtirestat (4). As shown in Figure S53 (Supporting information), both
compounds behaved in a similar way, where the original position of the carboxymethyl group
was more or less retained, but the tricyclic skeleton was turned in 180° due to the steric
hindrance of Oγ-Ser298. As a result, the hydrogen bond between the ligands and Leu300,
favorable in the native enzyme, could not be created in the activated form of AKR1B1.
Accordingly, the calculated values of score function decreased from 8.52 to 7.39 kJ/mol for
cemtirestat (4) and from 8.81 kJ/mol to 7.51 kJ/mol for compound 9. The corresponding
dissociation constants increased markedly, from 569 to 3800 nM for cemtirestat (4) and from
347 to 3140 nM for compound 9. These preliminary findings need further experimental
corroboration.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The in vitro experiments performed at the level of isolated enzymes were complemented
with the ex vivo studies performed at the organ level of isolated rat eye lenses. The
significantly elevated sorbitol accumulation in the lenses incubated in with glucose reflects
increased flux of glucose through the polyol pathway. Significant inhibition of the lens
sorbitol by 9 was recorded at a concentration as low as 10 µM (Fig. 6). At 50 µМ, the
inhibition accounted for approximately 75% for 9, while about 60 % and 25 % inhibition was
recorded for equimolar concentrations of cemtirestat (4)1 and epalrestat,39 respectively.
Apparently, the final response observed at an organ level is the consequence of interplay of
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 48
1
2
3
4
5
6
physicochemical properties (lipophilicity and ionizability) determining bioavailability and
inhibition efficacy of an inhibitor.
7
8
9
Inhibition of sorbitol dehydrogenase, the second enzyme of the polyol pathway,
concurrently with inhibition of aldose reductase, would, at least partially, eliminate the
sorbitol-decreasing effect of aldose reductase inhibition. Obviously this is not the case, since
no significant inhibition of sorbitol dehydrogenase activity by 9 at a concentration as high as
100 µМ was recorded [I (%) = 0.9 ± 0.6, n = 3]. Similar finding was reported for cemtirestat
(4).2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In our previous study, cemtirestat (4) was reported as a free radical scavenger and chain
breaking antioxidant.18 Here we compared antiradical activity of compound 9 with that of
cemtirestat (4) in a DPPH test. As shown in Fig.7, cemtirestat (4) started a considerable
absorbance diminution, corresponding to the transfer of the most labile H atom, which was
followed by a rather sluggish absorbance drop reflecting the radical scavenging activity of the
antioxidant degradation products. Differently, compound 9 affected DPPH absorbance only
marginally, pointing to its decreased radical scavenging activity compared to sulfur
containing cemtirestat (4). DPPH, a feeble hydrogen atom abstractor, is regarded as a good
kinetic analogue for peroxyl ROO• radicals.42 Yet on the basis of the kinetics of DPPH
decolorization, the free radical scavenging efficacy of 9 still exceeded that of compound 14
and also the equimolar melatonin used as standard antioxidant (Table 5).
Molecular obesity indices, taking into account efficacy, size and lipophilicity, were
calculated for the novel OTIs 9, 10, 12 and 14 (Table 6). Calculated values of the ligand
efficiency LE and the binding efficiency index BEI were found in the range of optimal
values (>0.3 and >14.7, respectively).19-22 For compounds 9, 10 and 14, high inhibition
efficacy characterized by pIC50 values close to 7 in combination with low values of logP
resulted in the lipophilic ligand efficiency values (LLE = pIC50 - logP) above the lower limit
22
ACS Paragon Plus Environment

Page 23 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
for successful lead 3.8.21 The LLE figures above 5 point to preference of specific binding to
aldose reductase.43 On the other hand, less-specific interactions are expected for compound
12. As shown in Table 6, the ligand-efficiency-dependent lipophilicity (LELP) values are in
the recommended range <7.520 with the exception of compound 12. This finding suggests
lower drug-likeness of 12 in comparison with the remaining compounds. Moreover, the
replacement of sulfur in 4 with oxygen in 9 resulted in decrease of LELP index from 3.4 to
3.0. The drop in LELP index supports the preferred role of specific component in binding of 9
to the enzyme. As a result markedly elevated selectivity relative to structurally related
AKR1A1 was recorded for 9 in comparison to 4 (see Table 2). All of the compounds shown
in Table 6 meet the threshold criteria of TPSA (60 – 140 Å2) for good oral absorbtion. 22,44
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CONCLUSIONS
Structure modification of the original carboxymethylated thioxotriazinoindole cemtirestat (4)
by isosteric replacement of sulfur with oxygen in combination with variable N-2 simple
substituents provided novel derivatives 9 and 14 with increased aldose reductase inhibition
efficacy and markedly increased selectivity, yet with decreased free radical scavenging
ability. Molecular obesity values indicate excellent drug-likeness of the novel OTI derivatives
with favorable oral bioavailability predicted. To continue this research, Biomagi Inc. has
already designed additional OTI derivatives, presently under preparation, with expected better
ALR2 affinity, solubility and also improved antioxidant action.
EXPERIMENTAL SECTION
Computational methods. The initial structures of compounds were calculated by
equilibrium conformer search procedure (MMFF94) in the program SPARTAN’08
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 48
1
2
3
4
5
6
7
8
9
(Wavefunction, Inc., Irvine, CA, 2009). The geometries of individual tautomers for compound
9 (tautomers analogical to those calculated for compound cemtirestat (4) in the paper of
Stefek et al.1) were obtained by full optimization in DFT B3LYP 6-31G* method in vacuum,
in water and in DMSO. The energies of solvated tautomers in water and in DMSO were
calculated by SM8 method of Cramer and Truhlar.45
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The structure of human recombinant enzyme AKR1B1 in complex with NADP+ and
cemtirestat (4) (PDB: 4QX4)1 was used for docking. The optimal conformers of the inhibitors
were docked into the enzyme-cofactor complex by the program Yasara46 using AMBER14
force field47 and utilizing the flexible ligands option. The first five clusters (if existing) as
ordered by AutoDock binding energy were then optimized in water. The simulation cell
overlapping the complex by 8 Å was filled by water molecules (final density 0.997 g/mL) and
Na+ and Cl- ions in the amount of 0.9% of the overall mass of water environment. The ratio of
the ions was chosen to neutralize the final charge of the complex. Re-calculation of pKa
values of amino acids was performed and pH 6.2 was maintained.48 A standard optimization
protocol, which consists of the steepest gradient optimization, molecular dynamics and
simulated annealing, was used. The rational geometry with the highest binding energy
(according to Yasara convention) was chosen for analysis of the key interactions. To validate
the protocol, cross-docking procedure was performed and confirmed a high value of accuracy
as 90.0%.17 In order to compare the complexes with compound 9 and with cemtirestat (4), the
original PDB: 4QX4 structure was relaxed with the same optimization protocol.
Chemical syntheses. Starting chemicals for syntheses were purchased from Sigma-
Aldrich (St. Louis, MO, USA), Fluorochem (Hadfield, UK), AlfaAesar (part of Thermo
Fisher Scientific, Heysham, UK) or Acros (part of Thermo Fisher Scientific, Geel, Belgium)
vendors. Other chemicals were purchased from local commercial sources and were of
analytical grade quality. Melting points were measured by Digital Melting Point Apparatus
24
ACS Paragon Plus Environment

Page 25 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Barnstead Electrothermal IA9200 and are uncorrected. H- and 13C-NMR spectra were
recorded on Varian Gemini (300/600 MHz and 75/150 MHz, resp.), chemical shifts are given
in parts per million (ppm), tetramethylsilane was used as internal standard and DMSO-d6 as
the solvent, unless otherwise specified. IR spectra were acquired on FT-IR-ATR REACT IR
1000 (ASI Applied Systems) with diamond probe and MTS detector. Mass spectra were
performed on LC–MS (Agilent Technologies 1200 Series equipped with Mass spectrometer
Agilent Technologies 6100 Quadrupole LC–MS). The course of the reactions was followed
by TLC analysis (Merck Silica gel 60 F254). UV lamp (254 nm) and iodine vapors were used
for the visualization of TLC spots. All tested compounds 4, 9, 10, 12 and 14 possess purity
over 95 %. Their purity was determined by melting point and combustion analysis (elemental
analysis found values for carbon, hydrogen, and nitrogen within 0.4 % of the calculated
values for the proposed formula).
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Ethyl 2-(2,3-dioxoindolin-1-yl)acetate (2). A mixture of 2.00 g (13.6 mmol, 1.00 mol
eq) isatine 1 and 630 mg (15.0 mmol, 1.10 mol eq) of CaH2 in 15 mL of DMF (abs) was
stirred at 100 °C for 1 h under Ar and then cooled to 40 °C. At this temperature 4.37 mL (40.8
mmol, 3.00 mol eq) of ethyl chloroacetate was added into the reaction mixture within 10 min.
The resulting mixture was heated to 100 °C for 5 h and then allowed to cool to rt. The
reaction was poured into a vigorously stirred aqueous 0.5 M HCl solution (50 mL), which
resulted in a red precipitate. The obtained solid material was filtered off and washed with
H2O. The crude product was further purified by trituration in a mixture of hexane / ethyl
acetate to afford 1.93 g (8.28 mmol, 61 %) of ethyl 2-(2,3-dioxoindolin-1-yl)acetate (2) as a
red solid material.
Novelty: The synthesis of compound 2 was previously described in the literature with
quantitative yield and characterized by its M.P., 1H-,13C-NMR, IR, MS and HRMS spectrum.8
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 48
1
2
3
4
5
6
7
8
9
M.p.: 125.8 - 127.5 °C [DMF / H2O] (lit. 132 - 133 °C).1H-NMR (300 MHz, CDCl3): δ 7.58
(dd, 1H, J(4,5) = 7.6 Hz, J(4,6) = 1.3 Hz, H-C(4)), 7.52 (ddd, 1H, J(5,6) = 9.1 Hz, J(6,7) = 7.8
Hz, J(4,6) = 1.3 Hz, H-C(6)), 7.09 (ddd, 1H, J(5,6) = 9.1 Hz, J(4,5) = 7.6 Hz, J(5,7) = 1.3 Hz,
H-C(5)), 6.72 (dd, 1H, J(6,7) = 7.8 Hz, J(5,7) = 1.3 Hz, H-C(7)), 4.42 (s, 2H, NCH2COOEt),
4.18 (q, 2H, J(CH2,CH3) = 7.2 Hz, -OCH2CH3), 1.22 (t, 3H, J(CH2,CH3) = 7.2 Hz, -
OCH2CH3).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2-(1-(2-Ethoxy-2-oxoethyl)-2-oxoindolin-3-ylidene)hydrazinecarbimidothioic acid (3).
Ethyl 2-(2,3-dioxoindolin-1-yl)acetate (2) 500 mg (2.15 mmol, 1.00 mol eq) and 235 mg
(2.58 mmol, 1.20 mol eq) of thiosemicarbazide were dissolved in 10 mL of DMF (abs) and
stirred at 100 °C for 1 h. Then the reaction mixture was cooled to rt and 20 mL of water
added. The separated aqueous solution was extracted with EA (5 x 10 mL) and the combined
organic layer washed with brine (3 x 10 mL), dried over Na2SO4, filtered and concentrated
under reduced pressure. The crude product was purified by trituration in a mixture of hexane /
ethyl acetate to yield 539 mg (1.76 mmol, 82 %) of 2-(1-(2-ethoxy-2-oxoethyl)-2-oxoindolin-
3-ylidene)hydrazinecarbimidothioic acid (3) as a red - orange solid material. Novelty: 2-(1-
(2-ethoxy-2-oxoethyl)-2-oxoindolin-3-ylidene)hydrazinecarbimidothioic acid (3) was not yet
1
described in the literature. M.p.: 209.3 - 214.0 °C [EA]. H-NMR (300 MHz, CDCl3): δ
12.25, 9.14 and 8.79 (3 x br s, 3 x 1H, -NH-, =NH and -SH), 7.74 (dd, 1H, J(4,5) = 7.6 Hz,
J(4,6) = 1.3 Hz, H-C(4)), 7.43 (ddd, 1H, J(5,6) = 8.9 Hz, J(6,7) = 7.8 Hz, J(4,6) = 1.3 Hz, H-
C(6)), 7.18 (ddd, 1H, J(5,6) = 8.9 Hz, J(4,5) = 7.6 Hz, J(5,7) = 1.3 Hz, H-C(5)), 7.17 (dd, 1H,
J(6,7) = 7.8 Hz, J(5,7) = 1.3 Hz, H-C(7)), 4.69 (s, 2H, NCH2COOEt), 4.18 (q, 2H,
J(CH2,CH3) = 7.1 Hz, -OCH2CH3), 1.22 (t, 3H, J(CH2,CH3) = 7.1 Hz, -OCH2CH3). 13C-NMR
(75 MHz, CDCl3): δ 179.2, 167.9, 161.2, 143.0, 131.6, 130.8, 123.7, 121.2, 119.6, 110.6, 61.8
(-OCH2CH3), 41.4 (NCH2COOEt), 14.4 (-OCH2CH3). FT-IR (solid, cm-1): 3411 (m), 3240
(m), 3145 (m), 2985 (w), 2941 (w), 1736 (s), 1688 (s), 1612 (s), 1471 (s), 1452 (s), 1418 (m),
26
ACS Paragon Plus Environment

Page 27 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1375 (s), 1351 (s), 1295 (w), 1237 (s), 1154 (s), 1103 (s), 1062 (s), 1011 (s), 885 (m), 868
(m), 835 (m), 786 (m), 738 (m), 641 (m), 619 (m), 567 (w), 491 (s), 456 (s). MS (ESI m/z):
307.0 [M+H]+. Anal. calcd for C13H14N4O3S (306.34): C, 50.97; H, 4.61; N, 18.29 Found: C,
50.84; H, 4.49; N, 18.05.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2-(3-Thioxo-2H-[1,2,4]triazino[5,6-b]indol-5(3H)-yl)acetic acid (4, cemtirestat, CMTI).
Solution of hydrazinecarbimidothioic acid 3 600 mg (1.96 mmol, 1.00 mol eq) was refluxed
in 5 % aq solution of K2CO3 (50 mL) for 2 days. After cooling to rt the mixture was treated
with HCl (1 M, aq) to pH = 2. The obtained solid material was filtered off, washed with water
and dried under reduced pressure to yield 398 mg (1.53 mmol, 78 %) of acid 4 (CMTI) as an
orange solid product. Novelty: synthesis and characterization of 2-(3-thioxo-2H-
[1,2,4]triazino[5,6-b]indol-5(3H)-yl)acetic acid (4) has not yet been described in the literature.
M.p.: 340 - 370 °C [H2O] (dec). 1H-NMR (300 MHz, DMSO-d6): δ 14.67 (br s, 1H, -NH- or
-SH), 13.25 (very br, 1H, -COOH) 8.06 (dd, 1H, J(8,9) = 7.6 Hz, J(7,9) = 1.4 Hz, H-C(9)),
7.68 - 7.65 (m, 2H, H-C(6) and H-C(7)), 7.41 (ddd, 1H, J(7,8) = 8.2 Hz, J(8,9) = 7.6 Hz,
13
J(6,8) = 1.4 Hz, H-C(8)), 5.00 (s, 2H, NCH2COOH). C-NMR (150 MHz, CDCl3): δ 179.5
(C=S), 169.0 (-COOH), 148.5 (C4a), 144.4 (C5a), 135.5 (C1a), 132.3 (C7), 124.1 (C8), 122.2
(C9), 117.5 (C9a), 112.3 (C6), 43.4 (NCH2COOH). FT-IR (solid, cm-1): 3500 (w), 3099 (w),
3020 (w), 2978 (w), 2892 (w), 1736 (m), 1601 (m), 1575 (s), 1466 (m), 1438 (w), 1370 (m),
1258 (m), 1167 (s), 1147 (s), 1073 (w), 1018 (m), 945 (w), 811 (w), 784 (m), 754 (m), 639
(w), 526 (m). MS (ESI m/z): 259.0 (10 %) [M-H+]-, 214.9 (100 %) [M-CO2-H+]-. Anal. calcd
for C11H8N4O2S (260.27): C, 50.76; H, 3.10; N, 21.53 Found: C, 50.92; H, 3.32 ; N, 21.84.
2-(2-Oxoindolin-3-ylidene)hydrazinecarboxamide (5). To a solution of 3.00 g (20.4
mmol, 1.00 mol eq) isatine 1 in 40 mL of EtOH, 2.73 g (24.5 mmol, 1.20 mol eq)
semicarbazide hydrochloride was added and the mixture refluxed for 30 min. Then, 40 mL of
water was added and the mixture cooled down and stirred for further 30 min. The resulting
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 48
1
2
3
4
5
6
7
8
yellow precipitate was filtered off, washed with water and dried under IR lamp and reduced
pressure yielding 4.01 g (19.6 mmol, 96 %) of semicarbazone 5 as a yellow solid product.
Novelty: 2-(2-oxoindolin-3-ylidene)hydrazinecarboxamide (5) was described in the literature
9
1
10
(90.5 % yield) together with its M.p., H-NMR, bioactivity, UV / VIS, IR11 and HRMS
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
spectral analyses.12 M.p.: 248 - 270 °C [EtOH] (dec), (lit. 239 °C). H-NMR (300 MHz,
DMSO-d6): δ 11.73 and 11.10 (2 x br s, 1H, -NH- and –NHCONH2), 7.60 (dd, 1H, J(4,5) =
7.6 Hz, J(4,6) = 1.3 Hz, H-C(4)), 7.32 (ddd, 1H, J(5,6) = 8.9 Hz, J(6,7) = 7.8 Hz, J(4,6) = 1.3
Hz, H-C(6)), 7.13 (br s, 2H, -NH2), 7.08 (ddd, 1H, J(5,6) = 8.9 Hz, (4,5) = 7.6 Hz, J(5,7) =
J
13
1.3 Hz, H-C(5)), 6.93 (dd, 1H, J(6,7) = 7.8 Hz, J(5,7) = 1.3 Hz, H-C(7)). C-NMR (150
MHz, DMSO-d6): δ 163.2 (C2), 155.5 (s, -NHCONH2), 141.9, 131.9, 130.8, 122.7, 120.8,
120.6, 111.3.
6-(2-Aminophenyl)-1,2,4-triazine-3,5(2H,4H)-dione (6). A solution of 4.00 g (19.6
mmol, 1.00 mol eq) semicarbazone 5 was refluxed in 1 M aq solution of NaOH (75 mL)
overnight. After cooling the solution was acidified with concentrated acetic acid to pH = 5
and stirred 30 min. The resulting precipitate was filtered off, washed with water and dried
under IR lamp and reduced pressure to yield 2.77 g (13.5 mmol, 69 %) of 6-(2-aminophenyl)-
1,2,4-triazine-3,5(2H,4H)-dione (6) as a yellow solid product. Novelty: Compound 6 was
described in the literature together with its M.p. and UV / VIS spectra.13 M.p.: 270 - 320 °C
[H2O] (dec), yellow solid compound (lit: 350 °C). 1H-NMR (300 MHz, DMSO-d6): δ 12.27
and 11.93 (2 x br s, 1H, -CONHCO- and =N-NHCO-), 7.23 (dd, 1H, J(A5,A6) = 7.7 Hz,
J(A4,A6) = 1.4 Hz, H-CA(6)), 7.07 (ddd, 1H, J(A4,A5) = 8.3 Hz, J(A3,A4) = 7.9 Hz,
J(A4,A6) = 1.4 Hz, H-CA(4)), 6.68 (dd, 1H, J(A3,A4) = 7.9 Hz, J(A3,A5) = 1.4 Hz, H-
CA(3)), 6.53 (ddd, 1H, J(A4,A5) = 8.3 Hz, J(A5,A6) = 7.7 Hz, J(A3,A5) = 1.4 Hz, H-CA(5)),
5.35 (br s, 2H, -NH2). 13C-NMR (150 MHz, DMSO-d6): δ 157.4, 149.8, 147.6, 143.4; 131.0,
130.2; 116.3, 115.8, 115.4. FT-IR (solid, cm-1): 3491 (w), 3377 (w, NH), 3018 (m, NH2),
28
ACS Paragon Plus Environment

Page 29 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1699 (s, C=O), 1609 (s, C=O), 1582 (m), 1542 (m), 1465 (m), 1311 (m), 1246 (m), 1154 (m),
1040 (w), 848 (m), 751 (s), 639 (m), 552 (m), 431 (s). MS (ESI m/z): 202.9 (100 %) [M-H+]-.
2H-[1,2,4]Triazino[5,6-b]indol-3(5H)-one (7). A suspension of 2.77 g (13.5 mmol, 1.00
mol eq) 6 in 50 mL of conc. acetic acid was refluxed for 3 h. After cooling to 0 °C 20 mL,
water was added and the resulting suspension was stirred for another 30 min in an ice bath.
The obtained solid was filtered off, washed with water, dried under IR lamp and reduced
pressure to yield 2.23 g (12.0 mmol, 88 %) of 2H-[1,2,4]triazino[5,6-b]indol-3(5H)-one (7) as
a yellow solid product. Novelty: 2H-[1,2,4]triazino[5,6-b]indol-3(5H)-one (7) was described
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
in the literature together with its M.p., H-NMR, IR and MS spectrum.14 M.p.: 320 - 360 °C
[AcOH] (dec), (lit: 280 °C). 1H-NMR (300 MHz, DMSO-d6): δ 13.01 and 11.91 (2 x br s, 1H,
-NHCO- and ArNH-), 7.89 (dd, 1H, J(8,9) = 7.7 Hz, J(7,9) = 1.3 Hz, H-C (9)), 7.53 (ddd, 1H,
J(7,8) = 8.0 Hz, J(6,7) = 7.9 Hz, J(7,9) = 1.3 Hz, H-C(7)), 7.31 (dd, 1H, J(6,7) = 7.9 Hz,
J(6,8) = 1.3 Hz, H-C (6)), 7.26 (ddd, 1H, J(7,8) = 8.0 Hz, J(8,9) = 7.7 Hz, J(6,8) = 1.3 Hz, H-
13
C(8)). C-NMR (75 MHz, DMSO-d6): δ 155.5, 155.2; 146.7, 133.2; 131.1, 122.7; 121.4,
118.5, 112.7. MS (ESI m/z): 184.9 (100 %) [M-H]-. Anal. calcd for C9H6N4O (186.17): C,
58.06; H, 3.25; N, 30.09; Found: C, 58.22; H, 3.43; N, 30.19.
Ethyl 2-(3-oxo-2H-[1,2,4]triazino[5,6-b]indol-5(3H)-yl)acetate (8a) and diethyl 2,2'-(3-
oxo-2H-[1,2,4]triazino[5,6-b]indole-2,5(3H)-diyl) diacetate (8b). A mixture of 1.40 g (7.52
mmol, 1.00 mol eq) 7 and 474 mg (11.30 mmol, 1.50 mol eq) of CaH2 in 15 mL of DMF
(abs) was stirred at 60 °C for 40 min under Ar and then cooled to rt. Then 1.61 mL (15.04
mmol, 2.00 mol eq) of ethyl chloroacetate was added dropwise via a syringe into the reaction
mixture within 10 min and afterwards the mixture was heated to 50 °C. The experiment with
1.1 mol ekv of ethyl chloroacetate was not performing well, only a low amount of the two
products was formed. After 3 h a TLC analysis confirmed formation of products 8a, 8b and
the starting material 7 in a ratio: 8a/8b / 7 = 2.0 /2.5 /1.0. Continuation of reaction mixture
29
ACS Paragon Plus Environment