Pterin-7-carboxamides as a new class of aldose reductase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2016.09.033

Aldose reductase is related to the onset and progression of diabetic complications, such as neuropathy, retinopathy, angiopathy, and so on: therefore molecules that are capable of inhibiting the enzyme are potential drugs for treatment of diabetic complications. Epalrestat is the sole aldose reductase inhibitor that is clinically used, but still has some drawbacks. Thus, the development of new aldose reductase inhibitors is still desired. We have synthesized a series of new pterin-7-carboxamides, and evaluated their in vitro inhibitory activities against human aldose reductase. All newly synthesized compounds exhibited the inhibitory activity. Among them, 1a having a glycine side chain exhibits the highest activity comparable to that of sorbinil, a highly active aldose reductase inhibitor. Molecular docking of 1a on the active site of the enzyme indicated this compound interacts with amino acid residues that are specific to the enzyme and related to suppressing side effects. Based on these results, we proved perin-7-carboxamides to be a new class of aldose reductase inhibitors, and particularly compound 1a was found to be a good candidate for further biological investigations as a drug for treatment of diabetic complications with fewer side effects.

Full text of DOI:10.1016/j.bmcl.2016.09.033

Bioorganic & Medicinal Chemistry Letters 26 (2016) 4870–4874
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Pterin-7-carboxamides as a new class of aldose reductase inhibitors
a
a
Ryota Saito a,b, , Saori Suzuki , Kaname Sasaki
⇑
a Department of Chemistry, Toho University, 2-2-1 Miyama, Funabashi, Chiba 274-8510, Japan
b Research Center for Materials with Integrated Properties, Toho University, 2-2-1 Miyama, Funabashi, Chiba 274-8510, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 20 June 2016
Revised 6 September 2016
Accepted 14 September 2016
Available online 15 September 2016
Aldose reductase is related to the onset and progression of diabetic complications, such as neuropathy,
retinopathy, angiopathy, and so on: therefore molecules that are capable of inhibiting the enzyme are
potential drugs for treatment of diabetic complications. Epalrestat is the sole aldose reductase inhibitor
that is clinically used, but still has some drawbacks. Thus, the development of new aldose reductase inhi-
bitors is still desired. We have synthesized a series of new pterin-7-carboxamides, and evaluated their
in vitro inhibitory activities against human aldose reductase. All newly synthesized compounds exhibited
the inhibitory activity. Among them, 1a having a glycine side chain exhibits the highest activity compa-
rable to that of sorbinil, a highly active aldose reductase inhibitor. Molecular docking of 1a on the active
site of the enzyme indicated this compound interacts with amino acid residues that are specific to the
enzyme and related to suppressing side effects. Based on these results, we proved perin-7-carboxamides
to be a new class of aldose reductase inhibitors, and particularly compound 1a was found to be a good
candidate for further biological investigations as a drug for treatment of diabetic complications with
fewer side effects.
Keywords:
Aldose reductase inhibitors
Pterin-7-carboxamide
Docking study
Quantitative structure–activity relationship
Ó 2016 Elsevier Ltd. All rights reserved.
Aldose reductase (ALR2 or AKR1B1) belongs to the aldo–keto
reductase (AKR) enzyme superfamily, and is the first enzyme in
the polyol pathway, which reduces glucose into sorbitol in the
presence of NADPH as a reductant.1,2 Because this enzyme is impli-
cated in the onset of diabetic complications, ALR2 inhibitors (ARIs)
have been thought to be effective in preventing diabetic complica-
tions, and thus various inhibitors, such as epalrestat, sorbinil
(Fig. 1), and others, have so far been developed.1,2 However, most
of these have failed clinical trials because of unfavorable side
effects, low efficacy, or toxicity.2 Currently, only epalrestat (Kine-
dakÒ) has been launched in the Japanese market as a chemothera-
peutic agent for treating diabetic peripheral neuropathy.3
However, during the six years after its launch, 17 cases of unex-
pected severe liver dysfunction, for which a causal relationship
to the drug cannot be ruled out, have been reported.4,5 Thus, the
development of new ALR2 inhibitors is still strongly desired.
The whole active site of ALR2 consists of three distinct pockets.2
The most important one is the catalytic site, or an ‘‘anion binding
pocket”, which is made up of Tyr48, His110, and Trp111 side chains
and commonly found in enzymes belonging to the AKR superfam-
ily.6 Adjacent to this, there is a pocket composed of the residues
Cys298, Leu300, Cys303, and Trp111. These residues are not
conserved in other AKRs, and therefore this pocket is often desig-
nated as ‘‘the specificity pocket.” The third one is a hydrophobic
pocket formed by the residues Trp20, Phe122, and Trp219.
Aldehyde reductase (ALR1 or AKR1A1), another member of the
AKR superfamily, has 65% sequence homology with ALR2, as well
as structural homology.7,8 This enzyme is often found in the kid-
neys and detoxifies or metabolizes toxic aldehydes; it is not asso-
ciated with the onset of diabetic complications in spite of its
similarity to ALR2. Therefore, it is important for ARIs to specifically
interact with amino acid residues characteristic of ALR2, because
the selectivity towards ALR2 and not ALR1 plays a role in suppress-
ing the side effects.7,8 With respect to this, it is known that Leu300
in the specificity pocket of ALR2 is replaced by Pro in ALR1, while
the anion binding pocket of ALR1 is composed of the same amino
acid residues as those of ALR2. Thus, the specific interaction with
Leu300 as well as that with the aromatic residues composing the
specificity pocket in ALR2 is considered essential for inhibitors to
accomplish the enzyme-selective inhibition.9 In this context, we
take notice of pterin-7-carboxamides, reported as ricin toxin A-
chain (RTA) inhibitors, because the pterin ring in these compounds
is capable of interacting with an aromatic residue through an aro-
matic-aromatic interaction as well as hydrogen bonding with the
protein backbone and amino acid residues adjacent to the aromatic
residue.10 These features of the pterin ring motivated us to utilize
this unit as a key structure interacting with both an aromatic
⇑
Corresponding author.
http://dx.doi.org/10.1016/j.bmcl.2016.09.033
0960-894X/Ó 2016 Elsevier Ltd. All rights reserved.

R. Saito et al. / Bioorg. Med. Chem. Lett. 26 (2016) 4870–4874
4871
A careful examination of the observed h-ALR2 inhibitory data
for 1a–o reveals some structure–activity trends. First, compounds
having
activity than the
jugated with -amino acids are well suited for the polypeptide
environment composed of -amino acids.
L-amino acid residues showed markedly higher inhibitory
D
-counterparts (1b–j), indicating that pterins con-
L
L
Evaluation of the inhibitory activity of compounds having
hydrophobic residues (1bL–eL, 1gL) showed that the activity appar-
ently decreased with increasing size of the residues. To assess this
observation quantitatively, the correlation between the pIC50 val-
ues and the physicochemical parameters presented in Table 1
was examined by means of the multiple linear regression analysis.
Figure 1. Structures of epalrestat and sorbinil.
residue and Leu300 in the specificity pocket of ALR2. In addition,
peptide-conjugated pterin carboxamides were proven to be readily
accessible,11 indicating that the carboxylate functional group,
which is required to attain high ALR2 inhibitory activity, can be
readily introduced. However, no attempt to develop ARIs based
on pterin-7-carboxamides has been made so far. Furthermore,
pterin rings are found in some biologically profitable compounds,
such as biopterin and folic acid (Fig. 2),12 suggesting that newly
designed ARIs synthesized from pterin are potential low-toxicity
agents. Thus, in this study, several amino-acid-conjugated pterin-
7-carboxamides have been synthesized and evaluated for their
in vitro inhibitory activity against recombinant human ALR2 (h-
ALR2) for the first time.
Synthesis of the pterin-7-carboxamides was achieved, as shown
in Scheme 1, based on a previously reported method that utilizes
DBU as a key additive to dissolve pterin in organic solvents and
accelerate the reaction.10,11 With this method, the pterin-amino
acid conjugates (1a–o) were readily obtained by simply mixing
pterin-7-carboxymethyl (7MCP) and unprotected amino acids in
the presence of DBU in methanol.
Here,
acid residues,
p
represents the hydrophobicity parameter for the amino
is the upsilon steric parameter, and L, B1, and B5
m
are the STERIMOL length and maximum and minimum width
parameters, respectively.17 The following equation was obtained
from the analysis as a statistically significant model:
pIC50 ¼ 7:492 ꢁ 1:409B1 ꢁ 0:232B5 ꢁ 0:053L þ 0:223
p
ðn
¼ 10; r2 ¼ 0:951; s ¼ 0:163Þ
ð1Þ
Despite the limited number of samples (n = 10), the high coefficient
of determination (r2 = 0.951) and low standard error (s = 0.163)
revealed that a good correlation was obtained. The Fischer test
proved this equation to be significant with 99% confidence interval
(p <0.01). It is obvious that the inhibitory activity can be explained
predominantly with the steric parameters, B1 and B5. The negative
sign of the coefficients for these parameters denotes the decrease
in activity with an increase in the bulkiness of the amino acid side
chain. A plot of the experimental versus predicted inhibitory activ-
ity using this model proved the predictive potential of this model
(Fig. 3). When the IC50 values of 1fL and 1jL are compared with those
of 1bL and 1gL, respectively, it was found that replacing one hydro-
gen in the residues of Ala and Phe with a hydroxy group, giving Ser
and Tyr, respectively, markedly decreases the activity. Perhaps the
hydroxy groups allow additional hydrogen bonding that negatively
influences the fitting of the substrate to the active site.
We next considered the effect of alkyl chain length on the inhi-
bitory activity. As can be seen from Table 1, it is obvious that there
is no proportional relationship between the methylene chain
length and activity. The activity decreased when the chain length
increased from 1 to 3 and increased with elongation of the chain
length until 5. The cause of this behavior is unclear, but it is likely
that the flexible unbranched alkyl chains fold themselves to fit the
whole structure appropriate to the enzyme’s active site, especially
when the chain length is 5. We also examined the pterin conju-
gated with Gly-Phe (1o), which has a chain length close to that
of 1m, and found it to be ineffective, probably because of the
hydrogen-bondable amide functionality and the branched large
benzyl group.
The ALR2 inhibitory activity of the synthesized pterin-7-carbox-
amides (1a–o) was evaluated in vitro by measuring their inhibitory
effects on the reduction of D,L-glyceraldehyde with recombinant h-
ALR2 in the presence of coenzyme NADPH as a reductant.13 When
h-ALR2 reduces glyceraldehyde, coenzyme NADPH is oxidized to
NADP+. Therefore, the progress of the reduction reaction can be
monitored spectrophotometrically by measuring the absorbance
of NADPH at 340 nm. This method has been proven reliable.14
The respective h-ALR2 inhibitory activities of 1a–o were reported
as their IC50 values, which express the 50% inhibition concentration
of the compounds in the bioreduction. We also measured the inhi-
bitory activity of pterincarboxylic acids (2 and 3) and folic acid for
comparison with that of epalrestat used as a positive standard. All
of the IC50 values are compiled in Table 1. It was found that 1a
exhibited the highest inhibitory activity (IC50 = 1.97
value was almost comparable to that of sorbinil, a highly active
M). Compounds 215 and 3,16 which have a carboxy
lM). This
ARI (IC50 ꢀ 2
l
group directly attached to the pterin ring, yielded poor results.
Compound 2 was found to be insoluble in the assay medium. Com-
pound 3 showed some inhibition, with an IC50 = 26.5 lM, suggest-
ing that a linker between the pterin ring and the carboxyl group is
Overall, the simplest 1a was the most potent ALR2 inhibitor
among the compounds tested.
required. Folic acid exhibited weak but explicit inhibitory activity
Next, in order to obtain a better insight into the interaction of
the pterins with the active site, docking simulations were con-
ducted. In these experiments, the protein structure coded as
2IKG19 in the Protein Data Bank was employed because the struc-
(IC50 = 52.5 lM). Owing to the synthetic difficulty, we have not
constructed a library of 6-substitited pterin, but this result left
the door open for further investigation of ALR2 inhibitory activity
of 6-substituted pterins.
Figure 2. Structures of biopterin and folic acid.

4872
R. Saito et al. / Bioorg. Med. Chem. Lett. 26 (2016) 4870–4874
Scheme 1. Synthesis of pterin-7-carboxamides 1a–o.
Table 1
In vitro h-ALR2 inhibitory activity of pterin-7-carboxamides (1a–o), 2, 3, folic acid, and epalrestat, compiled with some physicochemical parameters
R1
R2
IC50
-form
(l
M)
pa
vb
Lc
B1
B5
d
e
Compound
Entry
L
D-form
1a
1b
1c
1d
1e
1f
1g
1h
1i
1j
1k
1l
1m
1n
1o
2
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
CO2H
H
CO-Gly-OH
CO-Ala-OH
CO-Val-OH
CO-Leu-OH
CO-Ile-OH
1.97
12.4
0.00
0.31
0
0.52
2.06
2.87
1.00
1.52
1.00
2.04
72.0
194
97.2
278
130
68.1
42.1f
56.6
580
83.4
61.7
69.0
29.2
46.0
25.1
42.6
216
1.21
1.70
1.80
–0.04
1.79
2.10
1.96
0.96
0.76
0.98
1.02
0.53
0.70
0.70
0.57
0.70
4.11
4.92
4.92
3.97
4.62
8.33
6.28
4.73
1.90
1.52
1.90
1.52
1.52
1.52
1.71
1.52
3.17
4.45
3.49
2.70
6.02
3.58
3.11
6.72
CO-Ser-OH
CO-Phe-OH
CO-Hphi-OH
CO-Phgj-OH
CO-Tyr-OH
COꢁb-Ala-OH
CONH(CH2)3CO2H
CONH(CH2)4CO2H
CONH(CH2)5CO2H
CO-Gly-Phe-OH
H
17.5
25.5
11.7
4.02
9% (10.5 mM)g
h
–
3
CO2H
26.5
Folic acid
Epalrestat
52.5
0.0665
a
Hydrophobicity parameters of amino acid residues.
Upsilon steric parameters.
b
c,d,e
STERIMOL length and maximum and minimum width, respectively.17
f
g
h
i
D,L-form.
% inhibition at a given concentration.
Not determined, owing to the poor water solubility of the substrate.
Hph stands for NHCH(CH2CH2Ph)CO (homophenylalanine).18
Phg stands for NHCH(Ph)CO (phenylglycine).18
j
tural features of our pterins resemble the original ligand of 2IKG.
All the docking experiments were carried out with the GOLD Suite
software package.20,21 ChemPLP scoring function was used for
docking pose prediction, and the obtained poses were rescored
with GOLDscore.22 The best pose for each ligand was determined
by considering the results obtained with both of these scoring
functions.
First, we performed the docking simulation for 1a. Figure 4
shows the predicted pose of 1a in 2IKG structure. Obviously, 1a
tightly binds to the active site of ALR2 through hydrogen-bonding
networks. The carboxylic acid moiety was well inserted in the
anion-binding pocket to hydrogen bond with His110 and Tyr48.
It should be noted that a water molecule assists the interaction
of the amide NH of the inhibitor with Trp111. The pterin ring
makes a
p p interaction with Trp111 and a hydrogen bond with
ꢁ
Thr113 in the specificity pocket. The most noteworthy thing is that
the carbonyl oxygen of the pterin makes a hydrogen bond with
Leu300, which is significant for selective inhibition toward ALR2.
Figure 3. Plots of predictive pIC50 values for the model described by Eq. (1) versus
the actual pIC50 values.

R. Saito et al. / Bioorg. Med. Chem. Lett. 26 (2016) 4870–4874
4873
His110 to misdirect the carboxylate group to the opposite side of
the anion-binding site, while the pterin ring is accommodated in
the specificity pocket by making the same interactions with
Trp111, Thr113, and Leu300 as those observed for 1a. This type
of misdirection possibly comes about with 1jL, which has a pheno-
lic hydroxy group, and is likely to give rise to decrease the activity
of these inhibitors.
In conclusion, we explored the synthesis and h-ALR2 inhibitory
activity of pterin-7-carboxamides. From the results of the aldose
reductase inhibitory activity evaluation, it was found that 1a is
the most effective inhibitor (IC50 = 1.97
lM). This value is nearly
the same as that of sorbinil, a highly potent ARI (IC50 ꢀ 2
l
M).
From the docking study for 1a, it was suggested that 1a tightly
interacts with amino acid residues composing the active site of
ALR2. Notably, the specific interaction with Leu300 that is impor-
tant for ALR2 selective inhibition was observed with 1a. Thus, 1a
is a potential drug for treating diabetic complications that exhibit
fewer side effects by only inhibiting the relevant h-ALR2.
ChemPLP: 125.9092
GOLDscore: 130.0813
Figure 4. Docking of 1a into the active site of h-ALR2 complied with the docking
scores.
Conflicts of interest
The predicted docking poses for 1bL and 1bD are shown in Figure 5
(a) and (b). It can be seen from these figures that the carboxylic
acid moiety of 1bL makes a hydrogen bond with Tyr48, while
1bD does not. In addition, 1bL had higher docking score than 1bD.
The authors declare no conflicts of interest.
Acknowledgements
Conceivably, this explains why compounds that have
L-amino acid
This research was supported by the Japan Society for the Pro-
motion of Science (grant number 25410179); and the MEXT-Sup-
ported Program for the Strategic Research Foundation at Private
Universities (2012–2016). The authors would like to thank Editage
(www.editage.jp) for English language editing.
residues have higher inhibitory activity than the -counterparts.
D
Figure 5(c) and (d) show the predicted docking poses for 1fL and
1fD. As can be seen from these figures, hydroxy groups of the serine
residues in both of the compounds form hydrogen bonds with
ChemPLP: 119.4125
GOLDscore: 122.1643
ChemPLP: 113.9094
GOLDscore: 115.0396
(a)
(b)
ChemPLP: 111.0406
GOLDscore: 109.6966
(c)
(d)
ChemPLP: 108.1672
GOLDscore: 105.7916
Figure 5. Docking poses for 1bL (a), 1bD (b), 1fL (c), and 1fD (d) in the active site of h-ALR2 complied with the docking scores.

4874
R. Saito et al. / Bioorg. Med. Chem. Lett. 26 (2016) 4870–4874
9. Petrova, T.; Steuber, H.; Hazemann, I.; Cousido-Siah, A.; Mitschler, A.; Chung,
Supplementary data
R.; Oka, M.; Klebe, G.; El-Kabbani, O.; Joachimiak, A.; Podjarny, A. J. Med. Chem.
2005, 48, 5659.
Supplementary data (experimental procedures, characteriza-
tion, and in vitro methods) associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/j.bmcl.
2016.09.033. These data include MOL files and InChiKeys of the
most important compounds described in this article.
10. Pruet, J. M.; Saito, R.; Manzano, L. A.; Jasheway, K. R.; Wiget, P. A.; Kamat, I.;
Anslyn, E. V.; Robertus, J. D. ACS Med. Chem. Lett. 2012, 3, 588.
11. Saito, R.; Pruet, J. M.; Manzano, L. A.; Jasheway, K.; Monzingo, A. F.; Wiget, P. A.;
Kamat, I.; Anslyn, E. V.; Robertus, J. D. J. Med. Chem. 2013, 56, 320.
12. Basu, P.; Burgmayer, S. J. N. Coord. Chem. Rev. 2011, 255, 1016.
13. Saito, R.; Tokita, M.; Uda, K.; Ishikawa, C.; Satoh, M. Tetrahedron 2009, 65, 3019.
14. Del Corso, A.; Costantino, L.; Rastelli, G.; Buono, F.; Mura, U. Exp. Eye Res. 2000,
71, 515.
References and notes
15. Thomas, A. H.; Lorente, C.; Capparelli, A. L.; Pokhrel, M. R.; Braun, A. M.;
Oliveros, E. Photochem. Photobiol. Sci. 2002, 1, 421.
1. Brownlee, M. Nature 2001, 414, 813.
2. Changjin, Z. In Diabetes Mellitus – Insights and Perspectives; Oguntibeju, O., Ed.;
InTech, 2013; p 17.
3. Ramirez, M. A.; Borja, N. L. Pharmacotherapy 2008, 28, 646.
4. Yoshizaki, T.; Ichida, K.; Ohashi, M.; Nomura, M.; Kamata, T. J. Japan Diab. Soc.
2001, 44, 39.
5. Hotta, N.; Sakamoto, N.; Shigeta, Y.; Kikkawa, R.; Goto, Y. J. Diabetes Complicat.
1996, 10, 168.
16. Pruet, J. M.; Jasheway, K. R.; Manzano, L. A.; Bai, Y.; Anslyn, E. V.; Robertus, J. D.
Eur. J. Med. Chem. 2011, 46, 3608.
17. Fauchere, J. L.; Charton, M.; Kier, L. B.; Verloop, A.; Pliska, V. Int. J. Pept. Prot. Res.
1988, 32, 269.
18. Jones, J. H. In Amino Acids, Peptides and Proteins; Ryadnov, M., Farkas, E., Eds.;
The Royal Society of Chemistry:, 2014; Vol. 39, p P009.
19. Steuber, H.; Zentgraf, M.; La Motta, C.; Sartini, S.; Heine, A.; Klebe, G. J. Mol. Biol.
2007, 369, 186.
6. Sanli, G.; Dudley, J. I.; Blaber, M. Cell Biochem. Biophys. 2003, 38, 79.
7. Rees-Milton, K. J.; Jia, Z. C.; Green, N. C.; Bhatia, M.; El-Kabbani, O.; Flynn, T. G.
Arch. Biochem. Biophys. 1998, 355, 137.
20. Jones, G. GOLD, 5.2; The Cambridge Crystallographic Data Centre (CCDC).
21. Verdonk, M. L.; Cole, J. C.; Hartshorn, M. J.; Murray, C. W.; Taylor, R. D. Proteins
2003, 52, 609.
8. El-Kabbani, O.; Carbone, V.; Darmanin, C.; Oka, M.; Mitschler, A.; Podjarny, A.;
Schulze-Briese, C.; Chung, R. P. T. J. Med. Chem. 2005, 48, 5536.
22. Korb, O.; Stützle, T.; Exner, T. E. J. Chem. Inf. Model. 2009, 49, 84.