Design, synthesis, structure-activity relationships and X-ray structural studies of novel 1-oxopyrimido[4,5-c]quinoline-2-acetic acid derivatives as selective and potent inhibitors of human aldose reductase

Article abstract of DOI:10.1016/j.ejmech.2018.04.015

Human aldose reductase (AKR1B1, AR) is a key enzyme of the polyol pathway, catalyzing the reduction of glucose to sorbitol at high glucose concentrations, as those found in diabetic condition. Indeed, AKR1B1 overexpression is related to diabetes secondary

Full text of DOI:10.1016/j.ejmech.2018.04.015

Accepted Manuscript
Design, synthesis, structure-activity relationships and X-ray structural studies of
novel 1-oxo-pyrimido[4,5-c]quinoline-2-acetic acid derivatives as selective and potent
inhibitors of human aldose reductase
Isidro Crespo, Joan Giménez-Dejoz, Sergio Porté, Alexandra Cousido-Siah, André
Mitschler, Alberto Podjarny, Harris Pratsinis, Dimitris Kletsas, Xavier Parés, Francesc
X. Ruiz, Kamel Metwally, Jaume Farrés
PII:
S0223-5234(18)30349-0
10.1016/j.ejmech.2018.04.015
EJMECH 10364
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 4 January 2018
Revised Date: 18 March 2018
Accepted Date: 8 April 2018
Please cite this article as: I. Crespo, J. Giménez-Dejoz, S. Porté, A. Cousido-Siah, André. Mitschler,
A. Podjarny, H. Pratsinis, D. Kletsas, X. Parés, F.X. Ruiz, K. Metwally, J. Farrés, Design, synthesis,
structure-activity relationships and X-ray structural studies of novel 1-oxo-pyrimido[4,5-c]quinoline-2-
acetic acid derivatives as selective and potent inhibitors of human aldose reductase, European Journal
of Medicinal Chemistry (2018), doi: 10.1016/j.ejmech.2018.04.015.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
39
Ki app = 240 nM
68
Ki app = 73 nM
3D structure of AKR1B1-NADP+-39

ACCEPTED MANUSCRIPT
Design, synthesis, structure-activity relationships and X-ray structural studies of
novel 1-oxo-pyrimido[4,5-c]quinoline-2-acetic acid derivatives as selective and
potent inhibitors of human aldose reductase
Isidro Crespo,†,# Joan Giménez-Dejoz,† Sergio Porté,† Alexandra Cousido-Siah,‡ André
Mitschler,‡ Alberto Podjarny,‡ Harris Pratsinis,§ Dimitris Kletsas,§ Xavier Parés,†
Francesc X. Ruiz,‡, ^,* Kamel Metwally,¶,* and Jaume Farrés†,*
†Department of Biochemistry and Molecular Biology, Universitat Autònoma de
Barcelona, E-08193 Bellaterra, Barcelona, Spain
‡Department of Integrated Structural Biology, Institut de Génétique et de Biologie
Moléculaire et Cellulaire, CNRS, INSERM, UdS, 1 rue Laurent Fries 67404 CEDEX
Illkirch, France
§Laboratory of Cell Proliferation and Ageing, Institute of Biosciences and Applications,
National Centre of Scientific Research "Demokritos", Athens, Greece
¶Department of Medicinal Chemistry, Faculty of Pharmacy, Zagazig University, Zagazig,
Egypt
Author Information
*Corresponding Authors
E-mail: fxavier.ruiz@gmail.com, kametwally@hotmail.com, jaume.farres@uab.cat.
Phone: (+34)93 581 3312. Fax: (+34)93 581 1264.
Present Address
^(F.X.R.) Center for Advanced Biotechnology and Medicine and Department of
Chemistry and Chemical Biology, Rutgers University, Piscataway, New Jersey 08854–
5627, United States.
#(I.C.) ALBA Synchrotron, Carrer de la Llum 2-26, 08290 Cerdanyola del Vallès,
Barcelona, Spain.
1

ACCEPTED MANUSCRIPT
Abstract
Human aldose reductase (AKR1B1, AR) is a key enzyme of the polyol pathway,
catalyzing the reduction of glucose to sorbitol at high glucose concentrations, as those
found in diabetic condition. Indeed, AKR1B1 overexpression is related to diabetes
secondary complications and, in some cases, with cancer. For many years, research has
been focused on finding new AKR1B1 inhibitors (ARIs) to overcome these diseases.
Despite the efforts, most of the new drug candidates failed because of their poor
pharmacokinetic properties and/or unacceptable side effects. Here we report the synthesis
of a series of 1-oxo-pyrimido[4,5-c]quinoline-2-acetic acid derivatives as novel ARIs.
IC50 assays and X-ray crystallographic studies proved that these compounds are
promising hits for further drug development, with high potency and selectivity against
AKR1B1. Based on the determined X-ray structures with hit-to-lead compounds, we
designed and synthesized a second series that yielded lead compound 68 (Ki app vs.
AKR1B1 = 73 nM). These compounds are related to the previously reported 2-
aminopyrimido[4,5-c]quinolin-1(2H)-ones, which exhibit antimitotic activity. Regardless
of their similarity, the 2-amino compounds are unable to inhibit AKR1B1 while the 2-
acetic acid derivatives are not cytotoxic against fibrosarcoma HT-1080 cells. Thus, the
replacement of the amino group by an acetic acid moiety changes their biological
activity, improving their potency as ARIs.
Keywords: 1-oxo-pyrimido[4,5-c]quinoline-2-acetic acids; AKR1B1; AKR1B10;
molecular modeling; X-ray crystallography.
Abbreviations:
AKR, aldo-keto reductase; AKR1B1, aldo-keto reductase family member 1B1;
AKR1B10, aldo-keto reductase family member 1B10; AR, aldose reductase; ARI,
AKR1B1 inhibitor; DMEM, Dulbecco’s minimal essential medium; EIMS, electron
ionization mass spectroscopy; TEA, triethylamine.
2

ACCEPTED MANUSCRIPT
1. Introduction
Aldo-keto reductases (AKRs) constitute an oxidoreductase superfamily with
ubiquitous distribution in living organisms. Most AKRs are NADP(H)-dependent,
monomeric cytosolic proteins. Due to a conserved substrate-binding site (only with few
changes) and variable external loops, these enzymes can reduce carbonyl groups of a
wide variety of compounds, including carbohydrates, steroids, isoflavonoids, lipid
peroxidation products, retinoids and prostaglandins[1,2].
Human aldose reductase (AKR1B1, AR) is a potential therapeutic target against
secondary complications of diabetes because of its role in the reduction of glucose to
sorbitol. Under normoglycemic conditions, due to the high Km value of AKR1B1 for
glucose (Km = 76 mM)[3], the importance of this pathway is limited. But under
hyperglycemic conditions, the enzyme is active towards the formation of sorbitol, a
potent osmotic regulator. The high production of sorbitol along with its oxidation
product, fructose, induces cellular osmotic stress, which in turn promotes the expression
of AKR1B1. This would further accentuate the oxidative stress linked to the redox
imbalance on cofactor regeneration, accounting for secondary diabetes
complications[4,5] . Recently, AKR1B1 has also been shown as the mediator of certain
oxidative and inflammatory signaling pathways related to other pathologies, such as
cardiovascular disorders, sepsis, and cancer[6].
The use of enzyme inhibitors has been proposed as a good therapeutic strategy to
treat diseases associated to the action of AKR1B1, and the aforementioned recent
findings have renewed the interest for AKR1B1 as a target. Since AKR1B1 and other
members of the AKR superfamily, such as AKR1B10 and AKR1B15, have a very similar
active-site topology, selective inhibitors are needed to discriminate between them[4,6–
10]. Noteworthy, almost all AKR1B1 inhibitors (ARIs) possess a negatively charged
group, e.g, the carboxylic acid function of the acetic acid class of inhibitors, or the imide
group of hydantoins and their isosteres. This group interacts with the active-site region
where the catalytic residues and the positively-charged nicotinamide moiety of the
cofactor are located, the so-called anion-binding site[11]. Adjacent to the AKR1B1 active
site, an additional subpocket exists that can be opened by induced fit, the specificity
3

ACCEPTED MANUSCRIPT
pocket, giving rise to a significantly enlarged active site. This opening is generally driven
by a second, usually hydrophobic, ARI moiety[12,13].
A great number of ARIs have been identified in recent years, but very few of them
showed sufficient therapeutic efficacy in clinical trials, with epalrestat as the only ARI
commercially available to date. Most of the drug candidates failed due to poor
pharmacokinetic properties, toxicity and/or unspecific in vivo action[14]. Therefore, new
ARI scaffolds with an improved profile are needed.
Previously, Metwally and coworkers[15] characterized the 2-aminopyrimido[4,5-
c]quinolin-1(2H)-one scaffold, seeking inhibition of tubulin polymerization. They have
expanded the structure-activity relationship (SAR) of this scaffold, finding that it has
anticancer potential[16–19]. Although the overall similarity of this scaffold is low when
compared to other ARIs (as shown in Supplementary data, Table S1), it shares
substructures with alrestatin[20] and zenarestat[21], two potent AKR1B1 inhibitors
(Figure 1). In fact, several alrestatin derivatives and other isosteric quinolones have been
described as good in vitro ARIs[22]. Hence, we surmised that replacement of the 2-amino
group by a carboxylic acid moiety could make the 2-aminopyrimido[4,5-c]quinolin-
1(2H)-one scaffold amenable for AKR1B1 inhibition. Here, we synthesized and tested in
vitro 1-oxo-pyrimido[4,5-c]quinoline-2-acetic acid derivatives as new ARIs. Our results
showed that some of the derivatives are promising hits for AKR1B1 drug development
due to their low IC50 values and high selectivity for AKR1B1, when compared to the
closest homologue AKR1B10. Structural information revealed that the high potency of
the compounds is not dependent on the opening of the AKR1B1 specificity pocket and
that they explore previously unoccupied active-site regions. In addition, these compounds
showed low or no cytotoxicity against lung fibrosarcoma HT-1080 cells. Therefore,
within the series of 1-oxo-pyrimido[4,5-c]quinoline-2-acetic acid derivatives, the lead
compound 68, with a low nanomolar Ki app value and complying with the Lipinski’s rule
of five (Table S2), is a promising candidate for further drug development.
4

ACCEPTED MANUSCRIPT
Acetic Acid derivatives
(A)
Amino Derivatives
(B)
Zenarestat
(C)
Alrestatin
(D)
Figure 1. Chemical structures of 1-oxo-pyrimido[4,5-c]quinoline-2-acetic acid (A) and 2-aminopyrimido[4,5-
c]quinolin-1(2H)-one (B) derivatives compared to the classical ARIs Zenarestat (C) and Alrestatin (D).
2. Results and discussion
2.1. Design rationale
The initial 2-aminopyrimido[4,5-c]quinolin-1(2H)-one scaffold (Figure 1B)
presents a heterocyclic coplanar core reminiscent of alkaloids and flavonoids, with a
central quinoline moiety. Indeed, several ARIs obtained from natural sources are related
to those two families of metabolites, and some of them exhibit a quinoline moiety[22].
Regarding ARIs, having a quinoline isostere and whose structures in complex with
AKR1B1 holoenzyme have been solved by X-ray crystallography, zenarestat and
alrestatin are partly superimposable with the initial scaffold (see Supplementary data,
Similarity Analysis). and both possess a carboxylic acid moiety that interacts with the
anion-binding site (Figure 1)[20,21] That allowed us to propose the 1-oxo-pyrimido[4,5-
c]quinoline-2-acetic acid as a putative ARI, given its molecular similarity with the
aforementioned ARI[23].
To ascertain the feasibility of our hypothesis, we performed docking analysis of
both the initial and the proposed scaffold, including several substitutions based on
previous SAR performed with the initial scaffold. Out of the two possible PDB structures
amenable to use as docking receptors, we chose the corresponding to the AKR1B1
holoenzyme complexed with zenarestat (PDB ID 1IEI), since the cognate alrestatin
complex has two mutations in the active site (PDB ID 1AZ1). Figure 2 displays the
binding modes obtained. In all cases with the 1-oxo-pyrimido[4,5-c]quinoline-2-acetic
5

ACCEPTED MANUSCRIPT
acid derivatives, the carboxylic acid was located in the anion-binding site, similarly to
zenarestat, as we were expecting. Besides, docking was also performed with the other
main conformers of AKR1B1 active site[24] and the carboxylic acid of the best poses
was also located in the anion-binding site, except in the case of the fidarestat pocket
(Figure S1). Therefore, we decided to pursue synthesis and SAR of the designed
compounds.
Figure 2. Computed binding mode of 1-oxo-pyrimido[4,5-c]quinoline-2-acetic acid (gold sticks, 2-acetic
acid group circled with red dashes) and 2-aminopyrimido[4,5-c]quinolin-1(2H)-one (cyan sticks, 2-amino
group circled with blue dashes) scaffolds compared to zenarestat (green sticks). The interacting AKR1B1
residues are displayed in white sticks and the NADP+ molecule (from PDB ID 1IEI) is displayed in orange
sticks.
2.2. First series
2.2.1. Chemistry
The first series of target acetic acid derivatives (37-48) were synthesized
according to the synthetic route outlined in Scheme 1. A modified Pfitzinger procedure
was adopted to prepare the starting 3-aminoquinoline-4-carboxylic acid derivatives (1-
12) through reaction of the appropriate isatins with the appropriate phenacylamine
hydrochlorides under strongly basic conditions as previously reported[17,25,26].
6

ACCEPTED MANUSCRIPT
Cyclization to the desired lactones (13-24) was achieved by heating with acetic
anhydride. Treatment of the lactones with glycine ethyl ester in the presence of sodium
acetate in refluxing acetic acid afforded the acetate esters (25-36) in good yields. Basic
hydrolysis using aqueous sodium hydroxide gave the desired title compounds (37-48) in
fair yields.
Scheme 1:
Reagents and conditions:
(a) 4-Substituted phenacylamine hydrochloride, NaOH, EtOH/H2O/THF, 85 ºC to reflux. (b) Acetic
anhydride, reflux. (c) Glycine ethyl ester, CH3COONa, HAOc, reflux. (d) Aq. NaOH, 1N, RT.
2.2.2. AKR1B1 and AKR1B10 inhibitor activity
Testing compounds 37 to 48 in vitro against AKR1B1 revealed that the most
potent inhibitors were compounds 37 to 41 and 44, displaying IC50 values below 1 µM,
while the IC50 values for the other compounds ranged from 1.5 to 5 µM (Table 1).
AKR1B10 IC50 was included for each compound to establish their relative selectivity for
AKR1B1, as shown in Table 1. Compounds 37 to 41 and 44, the most potent for
AKR1B1, were also the most selective.
7

ACCEPTED MANUSCRIPT
Table 1. Inhibition of AKR1B1 and AKR1B10 by compounds of the first series.
IC50 (µM)
AKR1B1
Selectivity
Code
R1
R2
AKR1B1
0.21± 0.04
0.20 ± 0.02
0.35 ± 0.06
0.57 ± 0.12
0.26 ± 0.05
1.54 ± 0.12
1.92 ± 0.29
0.76 ± 0.09
2.22 ± 0.33
1.57 ± 0.22
4.92 ± 0.68
3.14 ± 0.53
AKR1B10
7.86
−H
−H
−Cl
−Br
37
50
23
17
38
6
37
38
39
40
41
42
43
44
45
46
47
48
10.00
8.18
−F
−Cl
−F
−Br
9.61
−F
−OCH3
−Cl
10.00
9.61
−Cl
−Cl
−Cl
−Br
−Br
−CH3
−OCH3
−Br
13.81
17.03
17.78
16.32
28.46
15.00
7
−OCH3
−Cl
22
8
−Br
10
6
−Cl
−Cl
5
2.2.3. In vitro cytotoxic activity
Most of the first series derivatives are potent AKR1B1 inhibitors (Table 1). To
analyze their cytotoxic activity, we tested them against the human fibrosarcoma HT-1080
cell line, using the MTT-assay. The results of 72-h incubation with the test compounds,
as well as doxorubicin hydrochloride used as positive control, are shown in Table 2. The
2-amino analogues previously described 71 and 72[15] are also included for comparison
(structures are shown in Table 3).
8

ACCEPTED MANUSCRIPT
Table 2. Cytotoxicity of first series compounds against HT-1080 cells.
Code
37
Cellular IC50 (µM)
> 100
38
39
40
41
42
43
44
45
> 100
> 100
> 100
> 100
30.2 (± 7.5)
> 100
78.1 (± 1.1)
> 100
> 100
46
47
> 100
86.7 (± 3.9)
48
1.23 (± 0.22) [15]
1.84 (± 0.40) [15]
0.0058 (±0.0029)
71
72
Doxorubicin
The results represent the mean (± standard deviation) of two independent experiments and are expressed as
cellular IC50, the concentration that reduced by 50% the optical density of treated cells with respect to
untreated controls.
In contrast with the results of the 2-amino compounds 71 and 72[15], their acidic
counterparts displayed no or weak cytotoxic activity, as shown in Table 2. Therefore, the
substitution of the amino group with an acetic acid moiety substantially decreases
cytotoxicity.
2.2.4. X-ray crystallography and structure-activity relationship
X-ray crystallography was used to understand the rationale of the strong inhibition
of the 1-oxo-pyrimido[4,5-c]quinoline-2-acetic acid scaffold against AKR1B1. Three
structures of complexes between AKR1B1 holoenzyme and compounds 37, 39 and 41
were obtained, at 0.94, 0.96 and 0.96 Å resolution, respectively. Table S3 shows the data
collection and refinement statistics (and Figure S2 displays Fo–Fc omit maps). All the
structures folded into the AKR prototypical (α/β)8-TIM barrel, with the cofactor position
9

ACCEPTED MANUSCRIPT
conserved, and provided a clear understanding of the relevant interactions that make up
the AKR1B1-inhibitor complexes.
Figure 3. A) Top view of inhibitor 39 (yellow sticks) complexed to AKR1B1 holoenzyme (protein in grey
surface and NADP+ in orange sticks). B) Atomic representation of the active site of AKR1B1 holoenzyme
complexed with 39 (coloring as in A, except for protein residues in white sticks), with distances in Å,
hydrogen bonds displayed with smudge green dashed lines and hydrophobic contacts in cyan dashed lines.
C) Superimposition of AKR1B1 (surface corresponds to PDB ID 1IEI) complexes with 39 (yellow sticks
and surface), alrestatin (violet sticks and surface, two molecules bound in a stacked arrangement) and
zenarestat (green sticks and surface).
10

ACCEPTED MANUSCRIPT
The inhibitors are oriented in the active site of AKR1B1 laying between loops B
and C (Figures 3A and C) with the carboxylate head close to the positively charged
nicotinamide moiety of the cofactor and forming hydrogen bonds with the OH of Tyr48,
the Nε2 of His110, and the Nε1 of Trp111 (Figure 3B), in close agreement with our
docking predictions (Figure 2). Besides, in all the structures there is an interstitial water
molecule at 2.9 Å distance from the nitrogen atom of the quinolone ring of the inhibitor
(Figures 3B and S4). The central pyrimido[4,5-c]quinolin-1(2H)-one moiety is
displaying hydrophobic interactions with Trp20, Phe122 and Trp219, while the 5-phenyl
moiety is stacked with Trp219. All these interactions anchor the inhibitor within the
enzyme active site (Figure 2B).
To note also that, while the acetic acid moiety is located roughly identically in the three
structures, the pyrimido[4,5-c]quinolin-1(2H)-one moiety can relatively swing between
Trp20 and Phe122, driven by the distinct interactions of their 5-phenyl moiety (Figure
S3). Nevertheless, in all three structures, the specificity pocket (localized between Trp111
and Leu300) is not open, remaining in the holoenzyme conformation characterized by a
closed loop C (with a closed Leu300, the gatekeeper residue, Figure 3A)[4,13].
The inhibition analysis indicates that substituents at the R1 position are the main
determinants of the inhibitory potency of this series (Table 1). Consistently, the obtained
structures show that substituents at the R1 position are important for the inhibitor
binding. Thus, R1 groups are located in a solvent exposed region (Figures 2A and B). In
this sense, compounds 37 to 41, bearing the smaller and more polar −H or −F at R1,
display IC50 values below 1 µM. The rest of the compounds, bearing more apolar
substituents (−Cl, −Br, −CH3 or −OCH3), display IC50 values above 1 µM (with the
exception of 44). The SAR for the R2 substituents is not that clear-cut. However, the
AKR1B1 holoenzyme complex with 41, which bears a −OCH3 substituent at R2, is the
only one where an R2 group is interacting with the protein (with the main-chain amino
groups of Leu301 and Ser302, Figure S3B), consistent with a lower IC50 value. Thus, the
−OCH3 substituent in 44 would similarly account for the slightly stronger inhibition of
this compound, though other possibilities could not be ruled out. Regarding the AKR1B1
selectivity of the current scaffold over AKR1B10, while no structural information was
11

ACCEPTED MANUSCRIPT
obtained with the latter, the most likely explanation might be the differences in loop C
residue composition, as loop B is highly conserved in between the two enzymes.
The docking poses obtained were similar to the co-crystallized ligands, but they
present some differences (apart from revealing the closed state of the specificity pocket),
the most notable being: i) the −CH2− of the 2-acetic acid group is flipped in the docking,
while in the X-ray structures the whole 2-acetic acid group was superimposable to the
cognate group in the AKR1B1-zenarestat complex (Figure 1); ii) the R2-substituted aryl
moiety is not stacked against Trp219.
We also compared our structures with the previously solved structures of
AKR1B1 complexed with alrestatin and zenarestat (PDB IDs 1AZ1 and 1IEI,
respectively, Figure 3C). Their superimposition allows visualizing that the current
compounds and alrestatin bind in a similar manner to the enzyme. On the other hand,
zenarestat induces the opening of the specificity pocket (Figure 3C). Therefore, we
strived to probe the R3 substituent, which appears oriented towards the area where the
specificity pocket is located (Figure 3C). Interestingly, the parent pyrimido[4,5-
c]quinolin-1(2H)-one scaffold[15] was modified by the addition of a halobenzyl group in
R3. Such a moiety is known to open the specificity pocket[27,28]. However, previous
studies by us and others[13,29,30] have shown that opening the specificity pocket of
AKR1B1 by a compound does not necessarily increase its inhibitory potency. In this
regard, we decided to design and synthesize a second series of R3 derivatives
contemplating both scenarios, either opening or not the specificity pocket, to assess how
this might affect their inhibitory potency.
2.3. Second series
2.3.1. Chemistry
The synthetic route was modified in order to obtain the 3-substituted benzyl
derivatives of the second series of target compounds (64-66) as depicted in Scheme 2.
Treatment of 3-aminoquinoline-4-carboxylic acid derivatives (1, 3) with phenylacetyl
chloride or 4-chlorophenylacetyl chloride in the presence of triethylamine in refluxing
THF yielded the N-acyl intermediates which were directly cyclized with acetic anhydride
to give the corresponding lactones (50-52). The acetate esters (57-59) were prepared
12

ACCEPTED MANUSCRIPT
following the same procedure used to synthesize their counterparts (25-36) as previously
mentioned.
Scheme 2:
Reagents and conditions:
(a) Phenylacetyl chloride or 4-chlorophenylacetyl chloride, TEA, THF, reflux. (b) The appropriate acid
anhydride, reflux. (c) Glycine ethyl ester, CH3COONa, HAOc, reflux. (d) HCl/HAOc, Reflux. (e) Aq.
NaOH, 1N, RT.
Attempts to hydrolyze the esters using aqueous sodium hydroxide were
unsuccessful and gave dark colored reaction mixtures probably due to high reactivity of
the benzylic side chain under basic conditions. Finally, the desired target acids (64-66)
13

ACCEPTED MANUSCRIPT
were obtained by hydrolysis of the esters under acidic conditions using a hydrochloric
and acetic acid mixture under reflux conditions. The 3-ethyl and 3-propyl substituted
acetic acids (68-70) were prepared following the same procedure used to prepare their
first series analogues with replacing acetic anhydride by propionic and butyric
anhydrides, respectively. Compound 67 was obtained in a similar fashion but starting
with the unsubstituted 3-aminoquinoline-4-carboxylic acid (49). Structures of all the
target compounds (37-48) and (64-70) were characterized by means of mass spectrometry
1
and, H and 13C NMR spectroscopy. Their purity was satisfactorily confirmed by
elemental analysis.
2.3.2. In vitro AKR1B1 and AKR1B10 inhibitor activity
In the second series of compounds, we used 37 as scaffold to probe the R3
substituent, and the inhibitory potency was measured by IC50 analysis (Table 3). The
most potent compounds are 68 and 69, without an aromatic group in R3. In fact, the most
potent compounds of this series, 67-70, contain an alkyl chain instead. The potency is
approximately 10 and 5-fold lower for 65, with benzyl, than for 68 and 69, respectively,
with an alkyl chain. On the other hand, it can be observed that the addition of fluorine to
R1 in 70 decreases the potency by a factor of 4 as compared to 68, with H in R1.
Regarding the selectivity to AKR1B1 in respect to AKR1B10, this second series also
seems to improve AKR1B10 inhibitory potency to the point that the selectivity for
AKR1B1 is lower. Thus, the SAR behind it will be a subject of further study, given the
added difficulty posed by the fact that the active site of AKR1B10 is even more flexible
than that of AKR1B1 and presents an extra subpocket at the base of loop A.
We also tested the amino derivatives (compounds 71 and 72)[15] to determine, by
comparison, the effect of the lack of the acidic group. As it can be seen in Table 3, no
inhibition was observed up to 10 µM level, indicating that despite sharing a structure
similar to the 2-amino derivatives, the acetic acid group is essential for AKR1B1
inhibition. This, and the fact that the 2-amino compounds are cytotoxic by tubulin
polymerization inhibition[15], make the amino or acidic motif as the selector between the
inhibition of tubulin polymerization and AKR1B1.
14

ACCEPTED MANUSCRIPT
Table 3. Inhibition of AKR1B1 and AKR1B10 by compounds of the second series.
IC50 (µM)
AKR1B1
Selectivity
Code
R1
R2
R3
AKR1B1
AKR1B10
−H
−H
−H
−H
−H
−H
−H
−F
−Cl
−Cl
−Cl
−H
−CH3
−CH2-C6H5
−CH2-4-Cl-C6H4
−CH2-4-Cl-C6H4
−CH3
0.21± 0.04
2.65 ± 0.41
1.01 ± 0.12
0.66 ± 0.10
0.43 ± 0.03
0.12 ± 0.01
0.19 ± 0.03
0.41 ± 0.04
NI
7.86
37
1
37
2.35
2.99
5.87
6.39
4.71
3.33
3.89
ND
64
3
65
9
66
−H
14
39
18
9
67
−Cl
−Cl
−Cl
−Cl
−Br
−CH2CH3
68
−CH2CH2CH3
−CH2CH3
69
70
−Cl
−Cl
4-Cl-C6H4
−
71*
72*
4-Br-C6H4
NI
ND
−
NI: No inhibition observed up to 10 µM. ND: Not determined. *Compounds from ref.[15], where the 2-
acetic acid group is replaced by a 2-amino group.
The results show that the most potent compounds display IC50 values close to the
enzyme concentration used in the assay. It is generally believed that in this case,
reversible inhibitors can be considered as tight-binding inhibitors[31]. Adjusting our data
to Morrison equation[31] [32], we were able to obtain the value of apparent Ki for the
most potent compounds (Table 4). Remarkably, the compound 68 showed the lowest
apparent Ki value, 73 nM, which ranges within the values for the best ARIs reported[32]
and complies with the Lipinski’s rule of five (Table S2).
15

ACCEPTED MANUSCRIPT
Table 4. Apparent Ki values for AKR1B1 using Morrison equation fitting for a tight-
binding inhibitor (IC50 < 0.5 µM).
Code
37
IC50 (µM)
0.21 ± 0.04
0.20 ± 0.02
0.35 ± 0.06
0.26 ± 0.05
0.46 ± 0.03
0.12 ± 0.004
0.19 ± 0.03
0.41 ± 0.04
Ki app (µM)
0.13 ± 0.02
0.15 ± 0.01
0.24 ± 0.05
0.24 ± 0.05
0.44 ± 0.04
0.073 ± 0.010
0.13 ± 0.02
0.50 ± 0.06
38
39
41
67
68
69
70
2.3.3. Structure-activity relationship
The SAR studies of the second series of compounds were conducted using the
structure of the AKR1B1 holoenzyme complexed to compound 39 as a receptor for
molecular docking with Autodock 4. In this complex, the specificity pocked remains in
the closed position (Figures 3A and B). As expected, the compounds containing an alkyl
moiety at R3 could fit well into the receptor, binding in a subpocket at the vicinity of
Leu300 (Figure 4A), while the compounds bearing an aryl moiety adopted unfavorable
poses with the aryl moiety exposed to the solvent (data not shown). The ethyl groups of
68 and 70 exhibit the optimal interaction with Leu300 (Figure 4B). As aforementioned,
compound 70 displays a 4-fold lower inhibition than 68. The docking simulations
suggest, as a possible reason, that the F atom at R1 might be too close to the carbonyl
oxygen of Val47 (Figure 4C), which would require an energetically unfavorable
rearrangement.
Therefore, the SAR studies of the second series seem to indicate that compounds
with a presumably favorable structure for opening the specificity pocket (benzyl group at
R3) do not provide interactions that offset the energetic barrier for this process (which is
most likely energetically unfavorable[13]), resulting into worse IC50 values for AKR1B1.
On the other hand, the R3 alkyl compounds can be accommodated in the closed
16

ACCEPTED MANUSCRIPT
holoenzyme and provide, in some cases e.g. 68 and 69, additional interactions that
slightly increase the inhibitory potency.
Figure 4. A) Top view of the superimposition of the inhibitor 39 (yellow sticks) complexed to AKR1B1
holoenzyme (protein in grey surface) with the following derivatives of the second series (in sticks): 67
(green), 68 (purple), 69 (purpleblue), 70 (magenta). B) Atomic view of the superimposition displayed in A,
including contacts with Leu300 in dashed lines: red (close contact), green (adequate contact), gold (too
distant contact). NADP+ in orange sticks. C) Atomic view of the superimposition displayed in A, including
detail of the F atom contact (39: yellow sticks, 70: magenta sticks) with Val47 of AKR1B1, with distances
indicated as in B.
17

ACCEPTED MANUSCRIPT
3. Conclusions
In this work, we have been influenced by the structure-based bioisosterism
approach to develop novel and pharmacologically-improved ARIs. The term was first
utilized by Sheng and coworkers[33] to describe the combination of bioisosterism and
molecular modeling in order to design novel triazole antifungal derivatives. As reviewed
in ref.[34], bioisosterism is a very useful drug design method to improve pharmacological
activity, gain selectivity and optimize the pharmacokinetics of lead compounds. In the
current case, however, the replacement of the 2-amino group by the 2-acetic acid has led
to achieving a different biological function compared to the original, which is the
opposite effect prompted by bioisosterim. To note that we deliberately used the acetic
acid moiety for our study, given that this provided a very high likelihood of at least
anchoring this novel scaffold to AKR1B1, as proven by experimental and virtual
approaches[35].
We initially synthesized the first series of target compounds (37-48) with a 3-
methyl side chain using acetic anhydride to form the lactones and basic hydrolysis of the
intermediate esters to the target acids. This series of compounds have proved to be potent
inhibitors of AKR1B1 with appreciable selectivity for it, compared to AKR1B10.
Interestingly, they also show low cytotoxicity in cellular experiments. Subsequently, we
were interested in improving the AKR1B1 inhibitory profile through a structure-based
drug design approach, using both X-ray crystallography and molecular docking. Indeed,
these structural studies have provided information that allowed the synthesis of better
fitting compounds, and support that their action does not require the opening of the
specificity pocket of the enzyme. The structures obtained also show that 1-oxo-
pyrimido[4,5-c]quinoline-2-acetic acid derivatives occupy a region of the active site, the
loop B (featuring Trp219, Figure 3), which previously was only filled by four ARIs out
of the 125 deposited in the PDB in complex with AR (csv file and Fig. S5 in
Supplementary data).
Next, the synthesis of the second series of target compounds (64-70) has probed
alkyl as well as benzyl substituents at position 3. In particular, the synthetic route to the
3-benzyl analogues (64-66) was modified in order to avoid basic hydrolysis in the final
step. From these series, compound 68 has emerged, bearing an alkyl group that
18

ACCEPTED MANUSCRIPT
presumably interacts with AKR1B1 having the specificity pocket closed. While the
benzyl substituents might open the specificity pocket of AKR1B1, they display worse
inhibitory potency than the best compounds of the first series. We hypothesize that this
decrease is due to the penalty of opening this pocket without novel interactions strong
enough to offset the energetic cost. Supporting our hypothesis, Klebe’s group[13] has
shown recently in another series of ARIs that there is an energetic barrier (~5 kJ/mol) for
opening the specificity pocket of AKR1B1. In conclusion, an important takeaway from
this work is that, at least for ARI design, seeking only compounds provoking induced-fit
can result in worsening of a series performance. In contrast, a proper space filling of the
holoenzyme pocket may be in some instances a better strategy. Investigating the
selectivity of the synthesized compounds against AKR1A1 is currently underway.
4. Materials and methods
4.1. Chemistry
4.1.1. General considerations
Melting points are uncorrected and were measured on a Gallenkamp melting point
1
apparatus. H and 13C NMR spectra were recorded on Bruker 400-MHz, JEOL
RESONANCE 500-MHz, and Varian-Mercury 300-MHz spectrometers. Chemical shifts
were expressed in parts per million (ppm) downfield from tetramethylsilane (TMS) and
coupling constants (J) were reported in Hertz. Mass spectra were recorded on a Shimadzu
QP2010 mass spectrometer. Elemental analyses (C, H, and N) were used to determine the
purity of all synthesized target compounds, and the results were within ± 0.4% of the
calculated values. All compounds were routinely checked by thin-layer chromatography
(TLC) on aluminum-backed silica gel plates. All solvents used in this study were dried by
standard methods. Intermediate 3-amino-2-(4-substitutedphenyl)-6-substitutedquinoline-
4-carboxylic acids (1-4,6,7,9-12,49)[17,25,26] and 3-methyl-5-substitutedphenyl-1H-
[1,3]oxazino[4,5-c]quinolin-1-ones (13,14,18)[15,17] were synthesized following
reported procedures.
19

ACCEPTED MANUSCRIPT
4.1.2. General procedure for the synthesis of 3-amino-2-(4-substitutedphenyl)-6-
substitutedquinoline-4-carboxylic acids (1-12).
A solution of the appropriate phenacylamine hydrochloride (10 mmol) in a
mixture of water (18 mL), ethanol (18 mL), and THF (8 mL), was added to a solution of
the appropriate isatin (8 mmol) and sodium hydroxide (57 mmol) in water (8 mL) at
85ºC, in a dropwise manner over 2 h. The resulting mixture was heated under reflux for
additional 30 min. The reaction mixture was concentrated under reduced pressure and
filtered through Celite®. The filtrate was acidified with acetic acid and the precipitate
obtained was filtered, washed with water, and dried. The crude product was purified by
recrystallization from ethanol/ethyl acetate.
4.1.2.1. 2-(4-Methoxyphenyl)-6-fluoro-3-aminoquinoline-4-carboxylic acid (5).
1
Yield: 51%, mp 224-226 ºC; H NMR (DMSO-d6):
δ 3.83 (s, 3H, OCH3), 7.08-7.11 (m,
2H, Ar-H), 7.24-7.28 (m, 1H, Ar-H), 7.59-7.62 (m, 2H, Ar-H), 7.82-7.87 (m, 1H, Ar-H),
8.16-8.19 (m, 1H, Ar-H). EIMS: m/z 312 [M]+.
4.1.2.2. 2-(4-Methoxyphenyl)-6-chloro-3-aminoquinoline-4-carboxylic acid (
8).
Yield: 54%, mp 242-244 ºC; 1H NMR (DMSO-d6):
δ
3.82 (s, 3H, OCH3), 7.08-7.10 (m,
2H, Ar-H), 7.36-7.39 (dd, J1= 2.28, J2 = 8.68 Hz, 1H, Ar-H), 7.60-7.62 (m, 2H, Ar-H),
7.78-7.80 (d, J = 8.72 Hz, 1H, Ar-H), 8.48-8.49 (d, J = 2.28 Hz, 1H, Ar-H). EIMS: m/z
328 [M]+.
4.1.3. General procedure for the synthesis of 3,9-disubstituted-5-un/substitutedphenyl-
1H-[1,3]oxazino[4,5-c]quinolin-1-ones (13-24, 53-56).
A mixture of the appropriate 3-amino-2-substitutedphenyl-4-quinolinecarboxylic acid (10
mmol) and the appropriate acid anhydride (10 mL) was heated at reflux for 5 h. After
cooling, the precipitate was filtered, washed with cold ethanol, and dried. The crude
product was recrystallized from either DMF/EtOH or DMF/H2O.
4.1.3.1. 5-(4-Chlorophenyl)-9-fluoro-3-methyl-1H-[1,3]oxazino[4,5-c]quinolin-1-one
(
15).
Yield: 67%, mp 218-220 ºC (DMF/EtOH); H NMR (DMSO-d6):
7.59-7.62 (m, 2H, Ar-H), 7.82-7.87 (m, 1H, Ar-H), 8.01-8.05 (m, 2H, Ar-H), 8.26-8.30
1
δ
2.48 (s, 3H, CH3),
20

ACCEPTED MANUSCRIPT
(dd, J1= 7.32, J2 = 9.16 Hz, 1H, Ar-H), 8.86-8.89 (dd, J1 = 2.76, J2 = 11.00 Hz, 1H, Ar-
H). EIMS: m/z 340 [M]+.
4.1.3.2. 5-(4-Bromophenyl)-9-fluoro-3-methyl-1H-[1,3]oxazino[4,5-c]quinolin-1-one
(16).
1
Yield: 67%, mp 204-206 ºC (DMF/EtOH); H NMR (DMSO-d6):
δ
2.48 (s, 3H, CH3),
7.72-7.76 (m, 2H, Ar-H), 7.80-7.85 (m, 1H, Ar-H), 7.93-7.96 (m, 2H, Ar-H), 8.24-8.28
(dd, J1 = 5.48, J2 = 9.16 Hz, 1H, Ar-H), 8.83-8.86 (dd, J1 = 2.72, J2 = 11.00 Hz, 1H, Ar-
H). EIMS: m/z 384 [M]+.
4.1.3.3. 5-(4-Methoxyphenyl)-9-fluoro-3-methyl-1H-[1,3]oxazino[4,5-c]quinolin-1-one
(17).
Yield: 53%, mp 188-189 ºC (DMF/H2O); 1H NMR (DMSO-d6):
δ
2.48 (s, 3H, CH3), 3.84
(s, 3H, OCH3), 7.07-7.10 (m, 2H, Ar-H), 7.78-7.83 (m, 1H, Ar-H), 7.93-7.96 (m, 1H, Ar-
H), 8.02-8.05 (m, 2H, Ar-H), 8.21-8.25 (dd, J1 = 5.52, J2= 9.16 Hz, 1H, Ar-H), 8.83-8.87
(dd, J1 =3.20, J2 = 11.44 Hz, 1H, Ar-H). EIMS: m/z 336 [M]+.
4.1.3.4. 5-(4-Bromophenyl)-9-chloro-3-methyl-1H-[1,3]oxazino[4,5-c]quinolin-1-one
(19).
1
Yield: 69%, mp 249-251 ºC (DMF/EtOH); H NMR (DMSO-d6):
δ
2.48 (s, 3H, CH3),
7.74-7.76 (m, 2H, Ar-H), 7.92-7.97 (m, 3H, Ar-H), 8.19-8.22 (d, J = 9.16 Hz, 1H, Ar-H),
9.20-9.21 (d, J = 2.28 Hz, 1H, Ar-H). EIMS: m/z 402 [M+2]+.
4.1.3.5. 5-(4-Methoxyphenyl)-9-chloro-3-methyl-1H-[1,3]oxazino[4,5-c]quinolin-1-one
(20).
Yield: 66%, mp 232-234 ºC (DMF/H2O); 1H NMR (DMSO-d6):
δ
2.48 (s, 3H, CH3), 3.85
(s, 3H, OCH3), 7.08-7.10 (dd, J1 = 2.28, J2 = 9.16 Hz, 1H, Ar-H), 7.89-7.92 (dt, J = 2.32,
8.68 Hz, 1H, Ar-H), 8.05-8.07 (dd, J1 = 2.28, J2 = 9.08 Hz, 1H, Ar-H), 8.16-8.18 (dd, J1 =
1.84, J2 = 9.16 Hz, 1H, Ar-H), 9.16-9.20 (t, J = 1.84 Hz, 1H, Ar-H). EIMS: m/z 352 [M]+.
4.1.3.6. 5-(4-Chlorophenyl)-9-bromo-3-methyl-1H-[1,3]oxazino[4,5-c]quinolin-1-one
(21).
1
Yield: 63%, mp 241-244 ºC (DMF/EtOH); H NMR (DMSO-d6):
δ
2.48 (s, 3H, CH3),
7.60-7.63 (m, 2H, Ar-H), 8.03-8.15 (m, 4H, Ar-H), 9.37 (s, 1H, Ar-H). EIMS: m/z 402
[M+2]+.
21

ACCEPTED MANUSCRIPT
4.1.3.7. 5-(4-Bromophenyl)-9-bromo-3-methyl-1H-[1,3]oxazino[4,5-c]quinolin-1-one
(22).
1
Yield: 70%, mp 238-241 ºC (DMF/EtOH); H NMR (DMSO-d6):
δ
2.48 (s, 3H, CH3),
7.074-7.77 (m, 2H, Ar-H), 7.95-7.98 (m, 2H, Ar-H), 8.03-8.06 (dd, J1 = 2.28, J2 = 9.16
Hz, 1H, Ar-H), 8.12-8.14 (d, J = 9.16 Hz, 1H, Ar-H), 9.37-9.38 (d, J = 1.84 Hz, 1H, Ar-
H). EIMS: m/z 446 [M+2]+.
4.1.3.8. 5-(4-Chlorophenyl)-3,9-dimethyl-1H-[1,3]oxazino[4,5-c]quinolin-1-one (23).
1
Yield: 65%, mp 220-222 ºC (DMF/EtOH); H NMR (DMSO-d6):
δ 2.36 (s, 3H, CH3),
2.39 (s, 3H, CH3), 7.48-7.50 (m, 2H, Ar-H), 7.62-7.65 (dd, J1 = 1.84, J2 = 8.72 Hz, 1H,
Ar-H), 7.91-7.98 (m, 3H, Ar-H), 8.89 (s, 1H, Ar-H). EIMS: m/z 336 [M]+.
4.1.3.9. 5-(4-Chlorophenyl)-9-methoxy-3-methyl-1H-[1,3]oxazino[4,5-c]quinolin-1-one
(24).
1
Yield: 67%, mp 245-246 ºC (DMF/EtOH); H NMR (DMSO-d6):
δ
2.46 (s, 3H, CH3),
3.96 (s, 3H, OCH3), 7.54-7.60 (m, 3H, Ar-H), 8.01-8.03 (d, J = 8.24 Hz, 2H, Ar-H), 8.09-
8.12 (d, J = 9.40 Hz, 1H, Ar-H), 8.62-8.63 (d, J = 2.76 Hz, 1H, Ar-H). EIMS: m/z 352
[M]+.
4.1.3.10. 5-Phenyl-3-methyl-1H-[1,3]oxazino[4,5-c]quinolin-1-one (53).
Yield: 73%, mp 196-198 ºC (DMF/EtOH); 1H NMR (CDCl3):
δ 2.57 (s, 3H, CH3), 7.52-
7.59 (m, 3H, Ar-H), 7.78-7.88 (m, 2H, Ar-H), 8.07-8.09 (m, 2H, Ar-H), 8.27-8.29 (d, J =
8.00 Hz, 1H, Ar-H), 9.35-9.37 (d, J = 8.40 Hz, 1H, Ar-H).
4.1.3.11. 5-(4-Chlorophenyl)-3-ethyl-1H-[1,3]oxazino[4,5-c]quinolin-1-one (54).
Yield: 71%, mp 199-200 ºC (DMF/EtOH); 1H NMR (CDCl3):
δ 1.38-1.42 (t, J = 7.60 Hz,
3H, CH3), 2.83-2.88 (q, J = 7.60 Hz, 2H, CH2), 7.52-7.54 (d, J = 8.40 Hz, 2H, Ar-H),
7.79-7.88 (tt, J = 7.20, 15.60, 31.60 Hz, 2H, Ar-H), 8.12-8.14 (m, 2H, Ar-H), 8.25-8.27
(d, J = 8.40 Hz, 1H, Ar-H), 9.35-9.37 (d, J = 8.40 Hz, 1H, Ar-H).
4.1.3.12. 5-(4-Chlorophenyl)-3-(1-propyl)-1H-[1,3]oxazino[4,5-c]quinolin-1-one (55).
Yield: 62%, mp 158-160 ºC (DMF/H2O); 1H NMR (CDCl3):
δ 1.08-1.11 (t, J = 7.60 Hz,
3H, CH3), 1.86-1.96 (m, 2H, CH2), 2.71-2.81 (t, J = 7.60 Hz, 3H, CH2), 7.52-7.54 (d, J =
8.40 Hz, 2H, Ar-H), 7.78-7.88 (m, 2H, Ar-H), 8.10-8.12 (d, J = 8.40 Hz, 2H, Ar-H),
8.24-8.26 (d, J = 8.00 Hz, 1H, Ar-H), 9.35-9.37 (d, J = 8.00 Hz, 1H, Ar-H).
22

ACCEPTED MANUSCRIPT
4.1.3.13. 5-(4-Chlorophenyl)-9-fluoro-3-ethyl-1H-[1,3]oxazino[4,5-c]quinolin-1-one
(56).
Yield: 64%, mp 185-186 ºC (DMF/EtOH); 1H NMR (CDCl3):
δ
1.39-1.42 (t, J = 7.60 Hz,
3H, CH3), 2.83-2.89 (q, J = 7.60 Hz, 2H, CH2), 7.21-7.26 (m, 2H, Ar-H), 7.57-7.62 (m,
1H, Ar-H), 8.15-8.25 (m, 3H, Ar-H), 8.99-9.02 (dd, J1 = 6.80, J2 = 10.40 Hz, 1H, Ar-H).
4.1.4. General procedure for the synthesis of 3-un/substitutedbenzyl-5-
un/substitutedphenyl-1H-[1,3]oxazino[4,5-c]quinolin-1-ones (50-52).
The appropriate phenylacetyl chloride (12 mmol) was added to a stirred mixture
of the appropriate 3-aminoquinoline-4-carboxylic acid (10 mmol) and TEA (12 mmol) in
THF (10 mL), in a dropwise manner at 0ºC. After addition was complete, the mixture
was heated at reflux for 8 h. The reaction mixture was evaporated under vacuum and the
residue was washed several times with n-hexane. Acetic anhydride was added and the
mixture was heated at reflux for 3 h. After cooling to room temperature, precipitate was
filtered, washed with n-hexane and dried. The crude product was recrystallized from
either DMF/EtOH or DMF/H2O.
4.1.4.1. 5-(4-Chlorophenyl)-3-benzyl-1H-[1,3]oxazino[4,5-c]quinolin-1-one (50).
1
Yield: 57%, mp 161-162 ºC (DMF/H2O); H NMR (CDCl3):
δ 4.09 (s, 2H, CH2), 7.40-
7.46 (m, 7H, Ar-H), 7.77-7.87 (tt, J = 7.20, 14.80, 31.60 Hz, 2H, Ar-H), 8.02-8.04 (d, J =
8.40 Hz, 2H, Ar-H), 8.23-8.25 (d, J = 8.00 Hz, 1H, Ar-H), 9.32-9.34 (d, J = 8.40 Hz, 1H,
Ar-H).
4.1.4.2. 5-(4-Chlorophenyl)-3-(4-chlorobenzyl)-1H-[1,3]oxazino[4,5-c]quinolin-1-one
(51).
Yield: 53%, mp 177-179 ºC (DMF/EtOH); 1H NMR (CDCl3):
δ
4.06 (s, 2H, CH2), 7.33-
7.40 (m, 4H, Ar-H), 7.45-7.47 (d, J = 8.40 Hz, 2H, Ar-H), 7.78-7.88 (m, 2H, Ar-H), 7.97-
7.99 (d, J = 8.80 Hz, 2H, Ar-H), 8.23-8.25 (d, J = 7.60 Hz, 1H, Ar-H), 9.31-9.33 (d, J =
7.60 Hz, 1H, Ar-H).
4.1.4.3. 5-Phenyl-3-(4-chlorobenzyl)-1H-[1,3]oxazino[4,5-c]quinolin-1-one (52).
Yield: 61%, mp 141-142 ºC (DMF/EtOH); 1H NMR (CDCl3):
δ 4.06 (s, 2H, CH2), 7.37-
7.52 (m, 7H, Ar-H), 7.80-7.86 (m, 2H, Ar-H), 8.01 (s, 2H, Ar-H), 8.27-8.29 (d, J = 6.40
Hz, 1H, Ar-H), 9.33-9.35 (d, J = 7.60 Hz, 1H, Ar-H).
23

ACCEPTED MANUSCRIPT
4.1.5. General procedure for the synthesis of ethyl 2-(5-(4-un/substitutedphenyl)-3,9-
disubstituted-1-oxopyrimido[4,5-c]quinolin-2(1H)yl)acetates (25-36, 57-63).
A mixture of the appropriate lactone (10 mmol), glycine ethyl ester hydrochloride
(11 mmol), and anhydrous sodium acetate (20 mmol) in glacial acetic acid (10 mL), was
heated at reflux for 24 h. After cooling, the reaction mixture was poured into ice water
and the precipitate was filtered, washed with aqueous sodium bicarbonate then washed
with water and ethanol and dried. The crude product was recrystallized from either DMF,
DMF/EtOH or DMF/H2O
4.1.5.1. Ethyl
2-(5-(4-chlorophenyl)-3-methyl-1-oxopyrimido[4,5-c]quinolin-
2(1H)yl)acetate (25).
1
Yield: 62%, mp 212-214 ºC (DMF/H2O); H NMR (DMSO-d6):
δ 1.09-1.13 (t, J = 6.88
Hz, 3H, CH3), 2.37 (s, 3H, CH3), 4.06-4.11 (q, J = 6.88 Hz, 2H, CH2), 4.93 (s, 2H, CH2),
7.45-7.49 (m, 2H, Ar-H), 7.68-7.74 (m, 2H, Ar-H), 7.98-8.00 (m, 2H, Ar-H), 8.04-8.07
(m, 1H, Ar-H), 9.43-9.45 (m, 1H, Ar-H). EIMS: m/z 407 [M]+.
4.1.5.2. Ethyl
2-(5-(4-bromophenyl)-3-methyl-1-oxopyrimido[4,5-c]quinolin-
2(1H)yl)acetate (26).
Yield: 64%, mp 216-219 ºC (DMF/EtOH); 1H NMR (DMSO-d6):
δ 1.22-1.25 (t, J = 6.88
Hz, 3H, CH3), 2.60 (s, 3H, CH3), 4.18-4.22 (q, J = 6.88 Hz, 2H, CH2), 5.04 (s, 2H, CH2),
7.71-7.73 (d, J = 8.24 Hz, 2H, Ar-H), 7.77-7.86 (m, 2H, Ar-H), 8.02-8.04 (m, 2H, Ar-H),
8.15-8.17 (d, J = 8.24 Hz, 1H, Ar-H), 9.53-9.55 (d, J = 8.24 Hz, 1H, Ar-H). EIMS: m/z
453 [M+2]+.
4.1.5.3. Ethyl 2-(5-(4-chlorophenyl)-9-fluoro-3-methyl-1-oxopyrimido[4,5-c]quinolin-
2(1H)yl)acetate (27).
Yield: 58%, mp 255-257 ºC (DMF/EtOH); 1H NMR (DMSO-d6):
δ 1.09-1.12 (t, J = 7.08
Hz, 3H, CH3), 2.36 (s, 3H, CH3), 4.07-4.11 (q, J = 6.92 Hz, 2H, CH2), 4.92 (s, 2H, CH2),
7.45-7.47 (d, J = 8.24 Hz, 2H, Ar-H), 7.60-7.65 (ddd, J1 = 3.20, J2 = 8.68, J3 = 17.40 Hz,
1H, Ar-H), 7.95-7.97 (m, 2H, Ar-H), 8.08-8.12 (dd, J1 = 5.96, J2 = 9.20 Hz, 1H, Ar-H),
9.07-9.11 (dd, J1 = 2.72, J2 = 11.44 Hz, 1H, Ar-H). EIMS: m/z 425 [M]+.
4.1.5.4. Ethyl 2-(5-(4-bromophenyl)-9-fluoro-3-methyl-1-oxopyrimido[4,5-c]quinolin-
2(1H)yl)acetate (28).
24

ACCEPTED MANUSCRIPT
Yield: 49%, mp 244-247 ºC (DMF/EtOH); 1H NMR (DMSO-d6):
δ 1.09-1.12 (t, J = 6.88
Hz, 3H, CH3), 2.39 (s, 3H, CH3), 4.05-4.09 (q, J = 6.84 Hz, 2H, CH2), 4.92 (s, 2H, CH2),
7.59-7.64 (m, 3H, Ar-H), 7.88-7.91 (m, 2H, Ar-H), 8.09-8.13 (dd, J1 = 5.96, J2= 9.16 Hz,
2H, Ar-H), 9.08-9.12 (dd, J1 = 2.76, J2 = 11.44 Hz, 1H, Ar-H). EIMS: m/z 471 [M+2]+.
4.1.5.5. Ethyl 2-(5-(4-methoxyphenyl)-9-fluoro-3-methyl-1-oxopyrimido[4,5-c]quinolin-
2(1H)yl)acetate (29).
Yield: 50%, mp 215-217 ºC (DMF/EtOH); 1H NMR (DMSO-d6):
δ 1.22-1.26 (t, J = 6.88
Hz, 3H, CH3), 2.62 (s, 3H, CH3), 3.85 (s, 3H, OCH3), 4.20-4.24 (q, J = 6.88 Hz, 2H,
CH2), 5.05 (s, 2H, CH2), 7.06-7.09 (m, 2H, Ar-H), 7.71-7.76 (ddd, J1 = 6.40, J2 = 8.24, J3
= 14.20 Hz, 1H, Ar-H), 8.09-8.13 (m, 2H, Ar-H), 8.18-8.22 (dd, J1 = 5.52, J2 = 9.16 Hz,
1H, Ar-H), 9.20-9.24 (dd, J1 = 2.76, J2= 11.48 Hz, 1H, Ar-H). EIMS: m/z 421 [M]+.
4.1.5.6. Ethyl 2-(5-(4-chlorophenyl)-9-chloro-3-methyl-1-oxopyrimido[4,5-c]quinolin-
2(1H)yl)acetate (30).
Yield: 52%, mp 253-255 ºC (DMF/EtOH); 1H NMR (DMSO-d6):
δ 1.19-1.22 (t, J = 6.84
Hz, 3H, CH3), 2.46 (s, 3H, CH3), 4.15-4.20 (q, J = 6.84 Hz, 2H, CH2), 5.02 (s, 2H, CH2),
7.56-7.58 (d, J = 8.24 Hz, 2H, Ar-H), 7.84-7.87 (dd, J1 = 6.32, J2 = 8.72 Hz, 1H, Ar-H),
8.07-8.09 (d, J = 6.68 Hz, 2H, Ar-H), 8.15-8.17 (d, J = 8.72 Hz, 1H, Ar-H), 9.55-9.56 (d,
J = 2.28, 1H, Ar-H). EIMS: m/z 441 [M]+.
4.1.5.7. Ethyl 2-(5-(4-bromophenyl)-9-chloro-3-methyl-1-oxopyrimido[4,5-c]quinolin-
2(1H)yl)acetate (31).
Yield: 60%, mp 253-256 ºC (DMF/EtOH);1H NMR (DMSO-d6):
δ 1.22-1.24 (t, J = 5.84
Hz, 3H, CH3), 2.62 (s, 3H, CH3), 4.18-4.23 (q, J = 6.84 Hz, 2H, CH2), 5.05 (s, 2H, CH2),
7.72-7.74 (m, 2H, Ar-H), 7.86-7.89 (m, 1H, Ar-H), 8.01-8.04 (m, 2H, Ar-H), 8.16-8.19
(m, 1H, Ar-H), 9.57 (s, 1H, Ar-H). EIMS: m/z 487 [M+2]+.
4.1.5.8. Ethyl 2-(5-(4-methoxyphenyl)-9-chloro-3-methyl-1-oxopyrimido[4,5-c]quinolin-
2(1H)yl)acetate (32).
1
Yield: 52%, mp 235-237 ºC (DMF/H2O); H NMR (DMSO-d6):
δ 1.22-1.25 (t, J = 6.88
Hz, 3H, CH3), 2.62 (s, 3H, CH3), 3.84 (s, 3H, OCH3), 4.18-4.23 (q, J = 6.88 Hz, 2H,
CH2), 5.03 (s, 2H, CH2), 7.06-7.08 (m, 2H, Ar-H), 7.81-7.84 (dd, J1 = 2.28, J2 = 8.72 Hz,
1H, Ar-H), 8.10-8.13 (dd, J1 = 2.28, J2 = 9.16 Hz, 2H, Ar-H), 9.52-9.53 (d, J = 2.28, 1H,
Ar-H). EIMS: m/z 437 [M]+.
25

ACCEPTED MANUSCRIPT
4.1.5.9. Ethyl 2-(5-(4-chlorophenyl)-9-bromo-3-methyl-1-oxopyrimido[4,5-c]quinolin-
2(1H)yl)acetate (33).
Yield: 56%, mp 251-253 ºC (DMF/EtOH); 1H NMR (DMSO-d6):
δ 1.22-1.26 (t, J = 6.84
Hz, 3H, CH3), 2.62 (s, 3H, CH3), 4.18-4.24 (q, J = 6.88 Hz, 2H, CH2), 5.05 (s, 2H, CH2),
7.58-7.60 (d, J = 8.72 Hz, 2H, Ar-H), 7.96-7.97 (dd, J1 = 2.28, J2 = 9.16 Hz, 1H, Ar-H),
8.07-8.11 (m, 3H, Ar-H), 9.72 (s, 1H, Ar-H). EIMS: m/z 487 [M+2]+.
4.1.5.10. Ethyl 2-(5-(4-bromophenyl)-9-bromo-3-methyl-1-oxopyrimido[4,5-c]quinolin-
2(1H)yl)acetate (34).
1
Yield: 57%, mp 256-259 ºC (DMF); H NMR (DMSO-d6):
δ 1.22-1.25 (t, J = 6.88 Hz,
3H, CH3), 2.61 (s, 3H, CH3), 4.18-4.23 (q, J = 6.88 Hz, 2H, CH2), 5.04 (s, 2H, CH2),
7.72-7.74 (m, 2H, Ar-H), 7.97-78.10 (Complex m, 4H, Ar-H), 9.72-9.74 (m, 1H, Ar-H).
4.1.5.11. Ethyl
2-(5-(4-chlorophenyl)-3,9-dimethyl-1-oxopyrimido[4,5-c]quinolin-
2(1H)yl)acetate (35).
Yield: 61%, mp 255-257 ºC (DMF/EtOH); 1H NMR (DMSO-d6):
δ 1.12-1.14 (t, J = 6.88
Hz, 3H, CH3), 2.38 (s, 3H, CH3), 2.46 (s, 3H, CH3), 4.07-4.12 (q, J = 6.88 Hz, 2H, CH2),
4.92 (s, 2H, CH2), 7.46-7.48 (m, 2H, Ar-H), 7.56-7.58 (d, J = 8.72 Hz, 1H, Ar-H), 7.93-
7.99 (m, 3H, Ar-H), 9.25 (s, 1H, Ar-H). EIMS: m/z 421 [M]+.
4.1.5.12. Ethyl
2-(5-(4-chlorophenyl)-9-methoxy-3-methyl-1-oxopyrimido[4,5-
c]quinolin-2(1H)yl)acetate (36).
Yield: 59%, mp 246-248 ºC (DMF/EtOH); 1H NMR (DMSO-d6):
δ 1.19-1.22 (t, J = 7.32
Hz, 3H, CH3), 2.56 (s, 3H, CH3), 3.90 (s, 3H, OCH3), 4.15-4.20 (q, J = 7.32 Hz, 2H,
CH2), 5.01 (s, 2H, CH2), 7.44-7.47 (dd, J1 = 2.76, J2 = 8.72 Hz, 1H, Ar-H), 7.52-7.54 (m,
2H, Ar-H), 8.04-8.06 (m, 3H, Ar-H), 9.00-9.01 (d, J = 2.76, 1H, Ar-H). EIMS: m/z 437
[M]+.
4.1.5.13. Ethyl
2-(5-(4-chlorophenyl)-3-benzyl-1-oxopyrimido[4,5-c]quinolin-
2(1H)yl)acetate (57).
Yield: 57%, mp 218-220 ºC (DMF/EtOH); 1H NMR (CDCl3):
δ 1.30-1.34 (t, J = 7.20 Hz,
3H, CH3), 4.22 (s, 2H, CH2), 4.23-4.29 (q, J = 7.20 Hz, 2H, CH2), 4.95 (s, 2H, CH2),
7.26-7.29 (m, 3H, Ar-H), 7.35-7.42 (m, 4H, Ar-H), 7.75-7.84 (m, 2H, Ar-H), 8.06-8.08
(d, J = 8.80 Hz, 2H, Ar-H), 8.26-8.28 (d, J = 7.60 Hz, 1H, Ar-H), 9.67-9.69 (d, J = 8.00
Hz, 1H, Ar-H).
26

ACCEPTED MANUSCRIPT
4.1.5.14. Ethyl
2-(5-(4-chlorophenyl)-3-(4-chlorobenzyl)-1-oxopyrimido[4,5-
c]quinolin-2(1H)yl)acetate (58).
Yield: 51%, mp 245-246 ºC (DMF/EtOH); 1H NMR (CDCl3):
δ 1.33-1.36 (t, J = 6.40 Hz,
3H, CH3), 4.16 (s, 2H, CH2), 4.26-4.32 (q, J = 6.40 Hz, 2H, CH2), 4.96 (s, 2H, CH2),
7.18-7.20 (m, 2H, Ar-H), 7.35-7.40 (m, 4H, Ar-H), 7.77-7.83 (m, 2H, Ar-H), 7.95-7.98
(d, J = 7.60 Hz, 2H, Ar-H), 8.26-8.28 (d, J = 7.60 Hz, 1H, Ar-H), 9.66-9.68 (d, J = 8.00
Hz, 1H, Ar-H).
4.1.5.15. Ethyl
2-(5-phenyl-3-(4-chlorobenzyl)-1-oxopyrimido[4,5-c]quinolin-
2(1H)yl)acetate (59).
Yield: 61%, mp 212-214 ºC (DMF/EtOH); 1H NMR (CDCl3):
δ 1.31-1.35 (t, J = 6.80 Hz,
3H, CH3), 4.16 (s, 2H, CH2), 4.25-4.30 (q, J = 6.84 Hz, 2H, CH2), 4.95 (s, 2H, CH2),
7.19-7.20 (d, J = 7.60 Hz, 2H, Ar-H), 7.32-7.34 (d, J = 8.00 Hz, 2H, Ar-H), 7.43-7.45 (d,
J = 7.20 Hz, 2H, Ar-H), 7.74-7.78 (t, J = 7.60 Hz, 1H, Ar-H), 7.81-7.84 (t, J = 7.60 Hz,
1H, Ar-H), 7.99-8.00 (d, J = 6.80 Hz, 2H, Ar-H), 8.01-8.04 (m, 2H, Ar-H), 8.28-8.30 (d,
J = 7.60 Hz, 1H, Ar-H).
4.1.5.16. Ethyl
60).
Yield: 55%, mp 236-238 ºC (DMF/EtOH); 1H NMR (CDCl3):
2-(5-phenyl-3-methy-1-oxopyrimido[4,5-c]quinolin-2(1H)yl)acetate
(
δ
1.34-1.37 (t, J = 7.20 Hz,
3H, CH3), 2.65 (s, 3H, CH3), 4.30-4.36 (q, J = 7.20 Hz, 2H, CH2), 4.98 (s, 2H, CH2),
7.50-7.57 (m, 3H, Ar-H), 7.74-7.84 (m, 2H, Ar-H), 8.13-8.15 (d, J = 7.20 Hz, 2H, Ar-H),
8.28-8.30 (d, J = 8.00 Hz, 1H, Ar-H), 9.68-9.70 (d, J = 8.00 Hz, 1H, Ar-H).
4.1.5.17. Ethyl
2-(5-(4-chlorophenyl)-3-ethyl-1-oxopyrimido[4,5-c]quinolin-
2(1H)yl)acetate (61).
Yield: 58%, mp 214-216 ºC (DMF/H2O); 1H NMR (CDCl3):
δ 1.36-1.40 (t, J = 8.00 Hz,
6H, 2CH3), 2.81-2.86 (q, J = 6.40 Hz, 2H, CH2), ), 4.30-4.35 (q, J = 6.40 Hz, 2H, CH2),
4.99 (s, 2H, CH2), 7.51-7.53 (d, J = 7.60 Hz, 2H, Ar-H), 7.76-7.82 (m, 2H, Ar-H), 8.18-
8.20 (d, J = 7.20 Hz, 2H, Ar-H), 8.25-8.27 (d, J = 7.20 Hz, 1H, Ar-H), 9.67-9.69 (d, J =
7.20 Hz, 1H, Ar-H).
4.1.5.18. Ethyl
2-(5-(4-chlorophenyl)-3-(1-propyl)-1-oxopyrimido[4,5-c]quinolin-
2(1H)yl)acetate (62).
27

ACCEPTED MANUSCRIPT
Yield: 59%, mp 210-213 ºC (DMF/EtOH); 1H NMR (CDCl3):
δ 1.06-1.10 (t, J = 7.20 Hz,
3H, CH3), 1.34-1.38 (t, J = 7.20 Hz, 3H, CH3), 1.88-1.97 (m, 2H, CH2), 2.75-2.79 (t, J =
7.20 Hz, 2H, CH2), 4.30-4.36 (q, J = 7.20 Hz, 2H, CH2), 4.99 (s, 2H, CH2), 7.51-7.53 (d,
J = 8.40 Hz, 2H, Ar-H), 7.75-7.84 (m, 2H, Ar-H), 8.14-8.16 (d, J = 8.40 Hz, 2H, Ar-H),
8.26-8.28 (d, J = 78.00 Hz, 1H, Ar-H), 9.68-9.70 (d, J = 8.00 Hz, 1H, Ar-H).
4.1.5.19. Ethyl 2-(5-(4-chlorophenyl)-9-fluoro-3-ethyl-1-oxopyrimido[4,5-c]quinolin-
2(1H)yl)acetate (63).
Yield: 57%, mp 218-220 ºC (DMF/EtOH); 1H NMR (CDCl3):
δ 1.35-1.42 (m, 6H, 2CH3),
2.81-2.87 (q, J = 6.80 Hz, 2H, CH2), 4.31-4.36 (q, J = 6.40 Hz, 2H, CH2), 4.99 (s, 2H,
CH2), 7.21-7.25 (t, J = 8.00 Hz, 2H, Ar-H), 7.53-7.57 (t, J = 6.80 Hz, 1H, Ar-H), 8.22 (s,
3H, Ar-H), 9.35-9.38 (d, J = 6.80 Hz, 1H, Ar-H).
4.1.6. General procedure for the synthesis of 2-(5-(4-un/substitutedphenyl)-3,9-
disubstituted-1-oxopyrimido[4,5-c]quinolin-2(1H)yl)acetic acids (37-48, 67-70).
The appropriate ester (10 mmol) was suspended in water (10 mL) and THF was
added in portions until the entire solid dissolved. Aqueous sodium hydroxide (5%, 10
mL) was added and the reaction mixture was allowed to stir at room temperature for 10 h.
The aqueous layer was separated and acidified with acetic acid and the precipitate was
filtered, washed with water and dried. The crude product was recrystallized from either
DMF/EtOH or DMF/H2O.
4.1.6.1. 2-(5-(4-Chlorophenyl)-3-methyl-1-oxopyrimido[4,5-c]quinolin-2(1H)yl)acetic
acid (37).
1
Yield: 52%, mp 292-294 ºC (DMF/EtOH); H NMR (DMSO-d6):
δ 2.59 (s, 3H, CH3),
4.96 (s, 2H, CH2), 7.58-7.59 (d, J = 9.00 Hz, 2H, Ar-H), 7.79-7.86 (m, 2H, Ar-H), 8.09-
8.14 (d, J = 8.50 Hz, 2H, Ar-H), 8.15-8.17 (d, J = 8.00 Hz, 1H, Ar-H), 9.56-9.57 (d, J =
8.50 Hz, 1H, Ar-H). 13C NMR (DMSO-d6):
δ 23.73 (CH3), 46.59 (CH2), 118.33, 123.35,
126.01, 128.05 (2C), 129.44, 129.70, 130.14, 133.09 (2C), 134.34, 137.09, 141.01,
144.59, 156.36, 157.50, 161.40, 169.46 (COOH). Anal. calcd for C20H14ClN3O3: C,
63.25; H, 3.72; N, 11.06. Found: C, 63.07; H, 3.81; N, 11.31.
4.1.6.2. 2-(5-(4-Bromophenyl)-3-methyl-1-oxopyrimido[4,5-c]quinolin-2(1H)yl)acetic
acid (38).
28