Potent human dihydroorotate dehydrogenase inhibitory activity of new quinoline-4-carboxylic acids derived from phenolic aldehydes: Synthesis, cytotoxicity, lipophilicity and molecular docking studies

Article abstract of DOI:10.1016/j.bioorg.2020.104373

A series of novel 2-substituted quinoline-4-carboxylic acids was synthesized by Doebner reaction starting from freely available protocatechuic aldehyde and vanillin precursors. Human dihydroorotate dehydrogenase (hDHODH) was recognised as a clear molecular target for these heterocycles. All compounds were also tested for their antiproliferative potential against three cancer cells (MCF-7, A549, A375) and one normal cell line (HaCaT) to evaluate the selective cytotoxicity. Quinoline derivatives 3f and 3g were identified as potent hDHODH inhibitors while 3k and 3l demonstrated high cytotoxic activity against MCF-7 and A375 cells and good selectivity. In addition, the logD_7.4 values obtained by the experimental method were found to be in the range from ?1.15 to 1.69. The chemical structures of all compounds were confirmed by IR, NMR and elemental analysis. The compounds pharmacology on the molecular level was revealed by means of molecular docking, highlighting the structural differences that distinguish highly active from medium and low active hDHODH inhibitors.

Full text of DOI:10.1016/j.bioorg.2020.104373

Bioorganic Chemistry 105 (2020) 104373
Contents lists available at ScienceDirect
Bioorganic Chemistry
journal homepage: www.elsevier.com/locate/bioorg
Potent human dihydroorotate dehydrogenase inhibitory activity of new
quinoline-4-carboxylic acids derived from phenolic aldehydes: Synthesis,
cytotoxicity, lipophilicity and molecular docking studies
a
b
b
b
´
´
Milena M. Petrovic , Cornelia Roschger , Sidrah Chaudary , Andreas Zierer ,
a
a
a
c
´
´
Milan Mladenovic , Katarina Jakovljevic , Violeta Markovic , Bruno Botta ,
a,*
´
Milan D. Joksovic
a Faculty of Sciences, Department of Chemistry, University of Kragujevac, R. Domanovica 12, 34000 Kragujevac, Serbia
b University Clinic for Cardiac-, Vascular- and Thoracic Surgery, Medical Faculty, Johannes Kepler University Linz, Krankenhausstraße 7a, 4020 Linz, Austria
c
`
Dipartimento di Chimica e Tecnologie del Farmaco, Sapienza Universita di Roma, P.le Aldo Moro 5, 00185 Roma, Italy
A R T I C L E I N F O
A B S T R A C T
Keywords:
A series of novel 2-substituted quinoline-4-carboxylic acids was synthesized by Doebner reaction starting from
freely available protocatechuic aldehyde and vanillin precursors. Human dihydroorotate dehydrogenase
(hDHODH) was recognised as a clear molecular target for these heterocycles. All compounds were also tested for
their antiproliferative potential against three cancer cells (MCF-7, A549, A375) and one normal cell line (HaCaT)
to evaluate the selective cytotoxicity. Quinoline derivatives 3f and 3g were identified as potent hDHODH in-
hibitors while 3k and 3l demonstrated high cytotoxic activity against MCF-7 and A375 cells and good selectivity.
In addition, the logD7.4 values obtained by the experimental method were found to be in the range from ꢀ 1.15 to
1.69. The chemical structures of all compounds were confirmed by IR, NMR and elemental analysis. The com-
pounds pharmacology on the molecular level was revealed by means of molecular docking, highlighting the
structural differences that distinguish highly active from medium and low active hDHODH inhibitors.
Dihydroorotate dehydrogenase (DHODH)
hDHODH inhibitors
Quinoline-4-carboxylic acid
Cytotoxicity
Lipophilicity
Molecular docking
1. Introduction
The essential role of pyrimidine nucleotides for cell proliferation and
multiplication determines hDHODH as the main target for design and
Human dihydroorotate dehydrogenase (hDHODH), as one of the
most important enzymes in sustaining the proliferation of cancer cells,
represents a good target for chemotherapeutic drugs [1]. In the fourth
step of the de novo biosynthesis of pyrimidines, this redox enzyme ca-
talyses the oxidation of dihydroorotate to orotate mediated by flavin
mononucleotide and nicotinamide adenine dinucleotide or ubiquinone.
This essential conversion allows cells to synthesize uridine mono-
phosphate, a vital building block in the formation of ribonucleosides and
deoxyribonucleosides for RNA and DNA synthesis [2,3]. In the prolif-
eration stage, the cells depend on de novo nucleoside biosynthesis and as
a result, DHODH is frequently overexpressed in cancer cells to support
their growth. Moreover, DHODH is not only considered as a key enzyme
in the treatment of cancer, but also a potential target for malaria [4],
viral [5] and autoimmune diseases, such as rheumatoid arthritis or
multiple sclerosis [6].
synthesis of new drug candidates. In an increased demand for DNA,
RNA, glycoproteins and membrane lipids, proliferating cells depend
heavily on pyrimidine biosynthesis pathway [7]. Inhibition of hDHODH
causes an insufficient concentration of pyrimidine nucleotides required
for continued growth and in turn triggers various cytotoxic, antima-
larial, antifungal and immunosuppressive activities [8,9].
Various hDHODH inhibitors have been synthesized, out of which
brequinar, leflunomide and its active metabolite teriflunomide have
gained a lot of interest over the several past decades (Fig. 1). Lefluno-
mide, an isoxazole-based prodrug, is FDA approved for rheumatoid and
psoriatic arthritis, while teriflunomide is FDA approved for multiple
sclerosis [10]. Brequinar is a fluorinated quinoline-4-carboxylic acid
derivative with potent antineoplastic activity in numerous in vitro ex-
periments and cancer models. Despite of the impressive preclinical
evaluation, brequinar did not meet objective response in multiple phase
* Corresponding author.
´
E-mail address: mjoksovic@kg.ac.rs (M.D. Joksovic).
https://doi.org/10.1016/j.bioorg.2020.104373
Received 10 March 2020; Received in revised form 25 September 2020; Accepted 8 October 2020
Available online 12 October 2020
0045-2068/© 2020 Elsevier Inc. All rights reserved.

´
M.M. Petrovic et al.
Bioorganic Chemistry 105 (2020) 104373
known procedure [19]. The reaction was performed in N,N-dime-
thylformamide as a solvent and NaHCO3 was used to neutralize liberated
HCl. In the next step, Doebner multicomponent reaction was applied to
prepare target compounds. Aniline, pyruvic acid and various O-alky-
lated aldehydes were refluxed in ethanol for 3 h affording the final
quinoline-4-carboxylic acids in yields from 24% to 51% (Scheme 1). All
compounds easily crystalized from ethanolic solution, sometimes after
scratching the bottom or side of the flask whit a glass rod, and were
satisfactory pure. The very high purity was achieved by their dissolving
in chloroform and precipitation by hexane or by recrystallization from
ethanol. The chemical structures of all quinoline derivatives were
confirmed by elemental analysis, IR, 1H and 13C NMR data (Supporting
Information).
2.2. Biological studies
2.2.1. Inhibition of hDHODH
Fig. 1. Chemical structures of the selected hDHODH inhibitors.
The inhibition of human dihydroorotate dehydrogenase (hDHODH)
activity by the compounds was determined using the DHODH
biochemical assay. This is an indirect colorimetric method which is
based on the oxidation of L-dihydroorotic acid (L-DHO) aided by
reduction of decylubiquinone (DUQ) which is stoichiometrically
equivalent to reduction of 2,6-dichloroindophenol (DCIP) resulting in
the decrease in absorbance at 610 nm [7]. The results of enzyme
inhibitory activity are presented in Table 1. Most quinoline-4-carboxylic
acids exhibited lower IC50 values than drug leflunomide used as a
reference compound. Quinoline derivatives from protocatechuic alde-
hyde series (3d, 3e, 3f and 3g) showed better activity than compounds
synthesized from alkylated vanillin (3h, 3i, 3j, 3k and 3l).
II clinical trials for several types of solid tumors [1]. However, the recent
studies demonstrated better suitability of brequinar for non-solid tu-
mors, particularly in induced differentiation of AML cells both in vivo
and in vitro assays [11,12]. Also, synergistic action of brequinar with
oxaliplatin and gemcitabine was observed in vitro in KRAS mutant cells
[13]. Therefore, these studies highlight promising perspectives for bre-
quinar and other hDHODH inhibitors containing quinoline-4-
carboxylate scaffold. The quinoline-4-carboxylic acid derivative C44 is
one of the most potent compounds that inhibits hDHODH with IC50 of 1
nM [14]. This hindered ether fragment with the restricted rotation
across the diaryl ether bond having required hydrophobicity, is crucial
for an excellent enzyme inhibition.
The tested compounds were not so potent as brequinar, a well known
hDHODH inhibitor with an IC50 of 5–10 nM [20]. As expected, a
replacement of the conformationally flexible alkoxy groups in our series
with the planar phenyl substituent in brequinar allows stacking in-
teractions resulting in a significant potency. Activity differences be-
tween brequinar containing the terminal 2-fluorophenyl substituent and
3′,4′-disubstitution may be attributed to effects that lower the total
number of conformations of the terminal phenyl ring. The similar ob-
servations were reported for quinoline-4-carboxylates bearing terminal
2-substituted pyridine ring [16]. In addition, the fluorine at 6-position of
quinoline scaffold in brequinar molecule has also a contribution to its
high hDHODH activity.
Bearing in mind these facts, we envisioned that two neighbouring
alkyl/aryl ether units combined with quinoline scaffold may also result
in good DHODH activity and favourable pharmacological profile.
Consequently, we have prepared a series of novel 2-substituted
quinoline-4-carboxylic acids involving two alkyl/aryl ether linkages.
The simple alkylation of freely available phenolic aldehydes and then a
three-component Doebner reaction of previously alkylated aldehydes,
aniline and pyruvic acid afforded the desired compounds with potent
hDHODH activity. The results also reported antiproliferative potential
against three cancer cell lines and selective cytotoxicity. To validate
DHODH as the target of our compounds, here we report a pictorial
presentation of the docked poses for the most active compounds.
2.2.2. Cytotoxicity
Cytotoxicity of the compounds against the human breast adenocar-
cinoma cell line MCF-7, human lung adenocarcinoma cell line A549,
human melanoma cell line A375 and the non-cancerous human kerati-
nocyte cell line HaCaT was evaluated using the 3-(4,5-dimethylthiazol-
2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. The results of
antiproliferative activity in vitro are expressed as IC50, the concentration
2. Results and discussion
2.1. Chemistry
Pfizinger and Doebner reactions are two possible ways to introduce a
carboxylic acid function in the 4-position of quinoline scaffold. The
former is based on the condensation of isatin with a carbonyl compound
in the presence of suitable strong base [15]. However, this synthetic
route suffers from a long reaction time, the poor availability of isatin and
low reaction yields. Several quinoline-4-carboxylic acid derivatives with
excellent DHODH inhibitory activity were prepared in this manner in
10–47% yields and 12–48 h of reflux [14,16]. The Doebner reaction is
another method of forming 2-substituted quinoline-4-carboxylic acids
starting from anilines, benzaldehydes and pyruvic acid [17]. Unfortu-
nately, low yields, particularly for di- or trisubstituted aldehydes, and
expensive catalysts [18] have also limited this route for the synthesis of
the various quinoline heterocycles. Thus, the selection of readily avail-
able and relatively cheap starting carbonyl and amine precursors could
increase an attractability and usability of Doebner reaction.
of compound (μM) that inhibits survival of the cells by 50% as compared
to control untreated cells (Table 2). MCF-7 and A375 cancer cells were
the most sensitive to the antiproliferative action of quinoline-4-
carboxylic acids.
The selectivity index (SI) is the ratio of the IC50 values for the MCF-7,
A549 and A375 cancer cell lines to the IC50 value for the normal HaCaT
cells (Table 3).
The greater value of SI indicates less toxic compound to the normal
cells offering a safer potential therapy. The compounds with SI higher
than 10 could be considered as selective ones. The quinoline derivative
3k from vanillin series displayed the highest SI ranging from 22.0 (A375
cells) to 26.0 (MCF-7 cells). The SI value from 0.82 to 4.64 suggests that
the compounds 3a-n cannot successfully differentiate between normal
and A549 cancer cell lines.
The first step in our synthesis of 2-substituted quinoline-4-carboxylic
acids was based on alkylation of commercially available vanillin and
protocatechuic aldehyde by different alkyl halides according to the
2.2.3. SAR studies
Three principal regions in the molecule of quinoline-4-carboxylic
2

´
M.M. Petrovic et al.
Bioorganic Chemistry 105 (2020) 104373
Scheme 1. Reagents and conditions: a) alkyl halide (MeI, nPrBr, nBuBr, iBuBr or BnCl), K2CO3, DMF, 5 h, reflux, HCl; b) EtOH, 3 h, reflux.
Table 1
Table 2
Inhibition of human dihydroorotate dehydrogenase (hDHODH) by quinoline-4-
IC50 values (
μ
M) found for the MCF-7, A549, A375 and normal HaCaT cell lines
carboxylic acid derivatives 3a-n.
after treatment with the quinoline-4-carboxylic acids 3a-n.
Compd.
IC50 ± SD (
μ
M)
Compd.
IC50 ± SD (
μ
M)
Compd.
IC50 ± SD (μM)
3a
3b
3c
3d
3e
3f
0.268 ± 0.018
0.340 ± 0.027
2.549 ± 0.193
0.270 ± 0.021
0.384 ± 0.029
0.153 ± 0.009
0.148 ± 0.009
0.467 ± 0.036
3i
1.387 ± 0.733
0.449 ± 0.039
0.340 ± 0.033
0.438 ± 0.033
1.073 ± 0.105
0.944 ± 0.047
0.900 ± 0.068
0.010 ± 0.005
MCF-7
A549
A375
HaCaT
3j
3a
3b
3c
3d
3e
3f
3.44 ± 0.80
3.21 ± 1.08
19.88 ± 0.02
19.36 ± 0.03
30.11 ± 0.03
12.49 ± 0.07
13.26 ± 0.04
10.75 ± 0.07
13.79 ± 0.05
24.86 ± 0.05
37.90 ± 0.06
15.07 ± 0.05
24.03 ± 0.02
8.96 ± 0.05
9.37 ± 0.05
22.71 ± 0.02
5.67 ± 2.07
4.67 ± 1.55
34.08 ± 11.82
3.50 ± 0.73
7.91 ± 3.31
5.42 ± 1.14
1.52 ± 0.44
6.79 ± 2.70
23.23 ± 7.08
2.69 ± 0.58
4.88 ± 1.81
3.48 ± 1.38
14.74 ± 6.36
6.15 ± 2.34
21.19 ± 8.32
15.80 ± 6.51
54.49 ± 13.87
20.08 ± 9.97
30.90 ± 9.34
22.62 ± 7.76
11.93 ± 5.47
69.50 ± 15.45
41.62 ± 11.42
36.13 ± 10.63
107.50 ± 20.49
32.85 ± 13.97
43.44 ± 11.37
57.48 ± 13.06
3k
3l
19.54 ± 8.90
3.29 ± 0.75
6.78 ± 2.15
3.47 ± 0.91
2.52 ± 0.72
6.61 ± 2.79
12.56 ± 5.86
7.28 ± 2.91
4.14 ± 1.72
1.77 ± 0.57
12.21 ± 5.20
10.06 ± 4.70
3m
3n
3g
3h
leflunomide
brequinar
3g
3h
3i
acid derivatives are responsible for enzyme activity: (i) there is a strict
requirement for carboxylic acid at the C-4 position (see the atom
numbering of the brequinar molecule in Fig. 1); (ii) a phenyl group at the
C-2 position with hydrophobic substituents is necessary and (iii) fluorine
or strong electron-withdrawing trifluoromethyl group at the C-6 posi-
tion have a beneficial effect for inhibitory activity [21]. Although
fluorine is widely used in drug synthesis to improve activity, the novel
investigation highlights “the dark side of fluorine” [22]. Drug-
metabolizing enzymes can promote oxidative defluorination and
3j
3k
3l
3m
3n
release fluoride generating various toxic effects. Very often, fluorine-
containing chemotherapeutics are usually highly toxic against normal
cell lines and exhibit low selective cytotoxicity [23]. For these reasons,
3

´
M.M. Petrovic et al.
Bioorganic Chemistry 105 (2020) 104373
for further comparison with our compounds. The compound 3i with two
methoxy groups at the 3′ and 4′ position displayed moderate DHODH
inhibition and low selectivity index. The polar hydroxyl groups in 3m
and 3n were not able to allow the required strong hydrophobic inter-
action between these two compounds and enzyme. Evidently, cytotoxic
and enzyme inhibition data suggest that the presence of at least one
hydrophobic group at the C-3′ and C-4′ position is necessary to retain the
good biological activity.
Table 3
Selectivity of the cytotoxicity of the compounds 3a-n against tumor cells as
compared with HaCaT cells.
Compd.
Selectivity index (SI)
MCF-7
A549
A375
3a
3b
3c
3d
3e
3f
6.16
4.92
2.79
6.10
4.56
6.52
4.73
10.51
3.31
4.96
25.97
18.56
3.56
5.71
1.07
0.82
1.81
1.61
2.33
2.10
0.87
2.80
1.10
2.40
4.47
3.67
4.64
2.53
3.74
3.38
1.60
5.74
3.91
4.17
7.85
10.24
1.79
13.43
22.03
9.44
2.95
9.35
Analysing the results from Tables 1 and 2, we can postulate that
better DHODH inhibition leads to better cytotoxic activity. These data
indicate that DHODH inhibition correlates with decreased tumor cell
growth, especially in MCF-7 and A375 cell lines.
3g
3h
3i
3j
2.2.4. Lipophilicity study
3k
3l
In the drug discovery process, the optimization of solubility and
lipophilicity plays an important role due to the close association of these
properties with the absorption, distribution, metabolism, excretion, and
toxicity (ADMET) of compounds [24]. High target potency in combi-
nation with high lipophilicity may increase the risk of ADMET-related
damage by enhancing drug promiscuity and nonspecific toxicity [25].
Therefore, balanced optimization is of great importance in the synthesis
of compounds with good pharmacological profile. The main goal of
optimization is to achieve good potency without significantly increasing
lipophilicity at the same time.
3m
3n
we designed the synthesis of quinoline-4-carboxylic acids omitting
fluorine at the C-6 position in order to achieve better selective cyto-
toxicity and minimize toxic side effects.
Our initial intention was to validate how the different position of two
benzyloxy groups in the compounds 3a, 3b and 3d influences on cyto-
toxic and DHODH inhibitory activity (Table 1). The analogue 3d was the
most active against A549 and A375 cell lines with the best selectivity
index (SI) and consequently its protocatechuic and vanillin precursors
(1c and 1d) were selected for further derivatization. Interestingly, 3a
(2′,3′-disubstitution) and 3d (3′,4′-disubstitution) displayed the same
One of the most important stages in the drug discovery process is
based on the drug absorption through membrane via passive diffusion.
In order to facilitate absorption, drugs should be lipophilic enough to
penetrate the lipid cores of membranes, but not too lipophilic that they
remain there. Therefore, lipophilicity as the measure of drug affinity for
a lipid environment plays a crucial role in the pharmaceutical industry
because it indicates the relationship of drugs with their pharmacokinetic
and metabolic properties [26]. Lipophilicity can be measured by the
distribution of a compound between the nonpolar organic phase (n-
octanol) and the aqueous phase. The partition coefficient P represents
the ratio of the equilibrium concentrations of a dissolved compound in
each phase and can be determined by a variety of methods [27]. How-
ever, when an ionisable compound is equilibrated in this two-phase
system, its concentration is directly related to the distribution coeffi-
cient D which is the partition coefficient of the compound in both
neutral and ionized form at any pH. The logD at pH = 7.4 has the highest
importance because of its physiological relevance and resemblance to
real biological partitions.
DHODH activity with an IC50 = 0.27 μM. When benzyloxy group was
replaced by methoxy substituent (3c), a substantial loss of cytotoxic and
enzyme inhibitory activity was observed. Further syntheses were per-
formed to see the effects of a series of 3′,4′-disubstituted quinoline
compounds for antiproliferative potential and enzyme inhibition. The
synthesized analogues 3f and 3g containing n-butyl and isobutyl groups
at C-3′ and C-4′ position respectively, were the most potent in DHODH
assays with IC50 = 0.15
MCF-7 and IC50 = 5.42
μ
M. The derivatives 3f (IC50 = 3.47
M against A375 cells) and 3g (IC50 = 2.52
M against A375 cells) were found to
μM against
μ
μM
against MCF-7 and IC50 = 1.52
μ
show strong antiproliferative potential. Unfortunately, the selectivity
index for these two compounds was relatively low; the 3g with bulkier
hydrophobic isobutyl group was two-fold more toxic against normal
HaCaT cells than 3f with n-butyl one. A comparison between 3f and 3g
suggests that substituent size and steric effects may be a contributing
factor to degrees of conformational freedom in antiproliferative action
of these compounds. The quinoline compound 3k, prepared from n-
butylated vanillin precursor resulted in the lowest cytotoxicity against
Highly lipophilic compounds (logD7.4 > 3.5) tend to have poor
aqueous solubility, high binding to plasma proteins, and greater po-
tential to inhibit Cytochrome P450 enzymes and interactions with P-
glycoproteins increasing the risk of drug-drug interactions. The mod-
erate lipophilicity (logD7.4 = 0–3) represents an excellent balance be-
tween permeability and solubility having low metabolic liability. These
logD7.4 values are favourable for BBB (blood-brain barrier) penetration
allowing an optimal oral absorption and cell membrane permeation in
cell-based assays. The negative values of logD in the physiologically
relevant pH lead to higher aqueous solubility and poor permeability
through biological membranes limiting the transport across the blood-
brain barrier [28].
normal HaCaT cells (IC50 = 107.50
μ
M), good antiproliferative activity
M and IC50 = 4.88 M,
M). The
against MCF-7 and A375 cells with IC50 = 4.14
μ
μ
respectively and excellent DHODH inhibition (IC50 = 0.340
μ
quinoline-4-carboxylic acids 3m and 3n synthesized from non-alkylated
phenolic aldehyde precursors demonstrated a significant decrease in the
antiproliferative and DHODH inhibitory potency. The compound 3l
prepared from isobutylated vanillin demonstrated lower DHODH ac-
tivity (IC50 = 0.44
protocatechuic aldehyde (IC50 = 0.15
effective against MCF-7 (IC50 = 1.77
3.48
μ
M) compared to 3g synthesized from isobutylated
M). However, 3l was also very
M) and A375 cancer cells (IC50
Since all our synthesized quinoline-4-carboxylic acids 3a-n are at
least partly charged at physiological pH (7.4) due to presence of ionis-
able carboxylic group, logD is the accurate descriptor of compound
lipophilicity as it describes the partition of both unionized and ionized
forms. Consequently, we measured the lipophilicity of our compounds
by their distribution between phosphate buffer at 7.4 (aqueous phase)
and n-octanol as an immiscible organic solvent. In the typical shake-flask
method, after vigorous mixing and phase separation, the compound
concentration in both solvents was measured by UV–Vis spectropho-
μ
μ
=
μ
M) with significant selectivity index (18.6 and 9.4 respectively).
The presence of two hydrophobic groups at the 3′ and 4′ position (3g)
led to increased cytotoxicity against the normal HaCaT cells in com-
parison with compound 3l containing only one hydrophobic substituent.
Generally, all compounds from vanillin series except 3i (3h, 3j, 3k and
3l) showed better selectivity than 3d, 3e, 3f and 3g from protocatechuic
aldehyde series indicating more favourable pharmacological profile for
quinolines derived from vanillin. Unfortunately, there is no available
cytotoxic data of other quinoline-4-carboxylic acids against normal cells
tometry and logD7.4 was calculated using the equation logD7.4 = logCoct
Cbuffer
Among all enzymes of de novo biosynthesis of pyrimidines, only
/
.
4

´
M.M. Petrovic et al.
Bioorganic Chemistry 105 (2020) 104373
DHODH is located on the outer surface of the inner mitochondrial
membrane and drugs must be able to reach to mitochondria to inhibit
this enzyme. Thus, the lipophilic properties of the potential inhibitors
will strongly influence on DHODH activity. The results presented in
Table 4 show that the least lipophilic compounds were those containing
polar hydroxyl and/or methoxy substituents (3c, 3i, 3m and 3n),
showing negative logD7.4 values from ꢀ 1.15 to ꢀ 0.37. As it was ex-
pected, the incorporation of more hydrophobic groups into the molec-
ular structure enhanced the lipophilicity, but also caused an increase in
the DHODH inhibitory activity. The derivatives 3f and 3g containing n-
butyl and isobutyl groups displayed the best DHODH activity with
logD7.4 = 1.49 and 1.44, respectively. We observed this range of logD7.4
values as optimal for an excellent DHODH inhibition. The higher value
of logD7.4 (3e, logD7.4 = 1.69) and lower one (3d, logD7.4 = 1.34) led to
decrease of DHODH inhibitory activity. It is interesting to note that 3e
with two n-propyl groups at 3′ and 4′ position (logD7.4 = 1.69) was more
lipophilic than 3f and 3g with n-butyl and isobutyl substituents. This
behaviour can be explained by conformational reasons where the most
abundant conformers in n-octanol have slightly lower hydrophobicity
due to interaction of vicinal alkyl chains in 3f and 3g [29]. The com-
pounds from vanillin series except 3i (3h, 3j, 3k, and 3l) had lower
lipophilicity with logD7.4 from 0.33 to 1.06, still retaining good IC50
Fig. 2. LogD7.4 values versus pIC50 (ꢀ logIC50) values of DHODH inhibi-
tory activity.
carboxylic acids were notable hDHODH inhibitors, their pharma-
cology was further considered on the structure-based level. In that sense,
to perform a quality structure-based activity relationship analysis, a
profound knowledge of hDHODH tertiary structure is required. Thus,
the resolved X-ray hDHODH topography [30] accentuates two domains:
a small N-terminal domain compiled of residues Met30-Leu68, and a
larger C-terminal domain formed of residues Met78-Arg396, inter-
values for enzyme inhibition from 0.34 μM to 0.47 μM. Evidently, the
lipophilicity strongly affected DHODH inhibition and this dependence
for 3′,4′-disubstituted derivatives of 2-phenylquinoline-4-carboxylic
acids is presented graphically in Fig. 2 with relatively good correlation
coefficient r = 0.76.
Experimentally determined antiproliferative activity of our com-
pounds showed certain dependence on lipophilicity versus IC50 values,
although a linear correlation was very weak (A549 cells) to moderate
(MCF-7 cells) with coefficient r < 0.55. Except 3h, in vitro cytotoxic
activity is significantly controlled by lipophilicity in the range of logD7.4
from 0.95 to 1.49 (3a, 3b, 3d, 3f, 3g and 3l). This indicates that an easier
transport over membrane barriers at cellular level can eventually lead to
cell death. The most lipophilic compound 3e with logD7.4 = 1.69
demonstrated lower activity against all types of cell lines. On the other
side, 3m with negative logD value (ꢀ 1.15) had very similar cytotoxic
activity against A549 cells compared to 3l with logD7.4 = 0.95. It means
that lipophilicity is important, but only one of factors influencing the
cytotoxic activity.
connected by an extended loop. The C-terminal domain is an
strands conglomerate with a central barrel of eight parallel β strands
surrounded by eight helices. Within the tertiary topography, two are
α helices/β
α
the sites responsible for biochemical role of hDHODH: (1) a redox site,
made of three antiparallel strands (βC, βD and βE) at the top and two
antiparallel strands (βA and βE) at the bottom, that serves as nesting area
for the dihydroorotate as substrate and flavin mononucleotide (FMN)
binding site as cofactor; (2) a narrow tunnel within the N-terminus,
through which the ubiquinone easily approaches the FMN for the redox
reaction. From the crystallography point of view, the narrow tunnel is
the very target for inhibiting the hDHODH activity [30]. By accepting
the ubiquinone as the second cofactor, the N-terminal domain fulfils its
biochemical purpose: by virtue of two joined constitutive
α helices
(labelled 1 and 2) it is a connector between the enzyme and the in-
α
α
2.2.5. Molecular docking studies
ternal mitochondrial membrane. In other words, being immersed into
the mitochondrial membrane, distinct helices form a tunnel that har-
bours the FMN binding site and accepts the ubiquinone. Thus, further
rationalization of the N-terminal domain topography provides a basis for
understanding the mode of action of potential hDHODH inhibitors. The
tunnel’s sub-site 1 is almost exclusively complied of hydrophobic resi-
dues Leu42, Met43, Leu46, Gln47, Ala55, Leu58, Phe62, Leu68, Phe98,
Having determined that structurally optimized quinoline-4-
Table 4
LogD7.4 values for 3a-n obtained experimentally by shake-flask method in two-
phase n-octanol/phosphate buffer system.
Compd.
C1 (M)
λ (nm)
C2 (M)
logD ± SD
and Leu359, which are, as α1 and α2 helices foundations, involved in
3.9675 × 10ꢀ 5
3.9675 × 10ꢀ 5
3.4484 × 10ꢀ 5
3.9675 × 10ꢀ 5
2.9190 × 10ꢀ 5
2.7109 × 10ꢀ 5
2.7109 × 10ꢀ 5
2.6440 × 10ꢀ 5
3.4484 × 10ꢀ 5
3.0014 × 10ꢀ 5
2.8187 × 10ꢀ 5
3.0355 × 10ꢀ 5
1.1377 × 10ꢀ 4
3.3093 × 10ꢀ 5
329
348
348
350
356
348
348
348
348
356
352
348
368
368
3.7866 × 10ꢀ 5
3.8030 × 10ꢀ 5
1.0314 × 10ꢀ 5
3.7931 × 10ꢀ 5
2.8608 × 10ꢀ 5
2.6267 × 10ꢀ 5
2.6162 × 10ꢀ 5
2.4323 × 10ꢀ 5
6.7710 × 10ꢀ 6
2.0401 × 10ꢀ 5
2.3835 × 10ꢀ 5
2.7265 × 10ꢀ 5
7.5245 × 10ꢀ 6
4.2299 × 10ꢀ 6
1.32 ± 0.01
1.36 ± 0.01
ꢀ 0.37 ± 0.01
1.34 ± 0.02
1.69 ± 0.01
1.49 ± 0.01
1.44 ± 0.02
1.06 ± 0.01
ꢀ 0.61 ± 0.01
0.33 ± 0.02
0.74 ± 0.01
0.95 ± 0.01
ꢀ 1.15 ± 0.02
ꢀ 0.83 ± 0.02
membrane association. In between the tunnel and the afore mentioned
redox site there are sub-sites 2 and 3, formed of Gln47, Arg136, Tyr356
and Thr360. The amino acids Val134 and Val143 create hydrophobic
sub-site 4 that caps the narrow end of the tunnel.
3a
3b
3c
3d
3e
3f
The entire molecular docking-based discussion is performed by
respecting the decreasing order of hDHODH inhibitors activity. All of the
best-clustered conformations were comparable with the previously
experimentally determined crystals structures of either clinical drug
leflunomide (PDB ID 3F1Q) [30] or co-crystalized 4-quinoline carbox-
ylic acid derivatives, brequinar (PDB ID 1D3G) [31], 43 and 46 (PDB IDs
6CJF and 6CJG, respectively) [16]. Hence, being reversibly associated
to the hDHODH, the herein most active hDHODH inhibitor 3g (Fig. 3)
occupies almost entirely the narrow tunnel. Thus, compound adopts a
biopose in which the C-4 carboxylic acid forms a strong hydrogen bond
3g
3h
3i
3j
3k
3l
3m
3n
C1 - concentration of tested compound in n-octanol phase before partitioning
(mol Lꢀ 1);
(dHB = 2.648 Å) with one of the Arg136′s guanidino
ω-nitrogens (ω-N),
λ - the wavelength of the absorption maximum after partitioning;
C2 - concentration of tested compound in n-octanol phase after partitioning (mol
Lꢀ 1).
via the carbonyl moiety. The same
ω-N is a hydrogen bond donor for
5

´
M.M. Petrovic et al.
Bioorganic Chemistry 105 (2020) 104373
Fig. 3. The bioactive conformation of 3g as
hDHODH inhibitor. For the clarity of presentation,
only the N-terminal amino acids, orotate as hDHODH
substrate, and FMN as hDHODH coenzyme were
depicted, whereas the remaining of the enzyme is
illustrated with surface. Only polar hydrogens were
preserved for presentation. The established hydrogen
bonds are presented with blue lines. (For interpreta-
tion of the references to colour in this figure legend,
the reader is referred to the web version of this
article.)
most of the remaining compounds, as well, and it will be used accord-
ingly in further discussion. The remaining hydroxyl group of the C-4
carboxylic acid scaffold is involved in electrostatic interferences with
the Gln47′s amide nitrogen (another often but not an ubiquitous inter-
action). Consequently, the quinoline ring is placed orthogonally related
to FMN and seems to seal the passage towards the coenzyme. The het-
erocycle’s orientation is additionally conditioned with the T-shaped (i.e.
edge to face) hydrophobic interactions with Val134, Val143, and Tyr356
(through the phenyl moiety), eclipsed (i.e. face to face) steric hindrance
with Thr360′ side chain methyl group, and parallel-displaced hydro-
phobic interactions with His56. The quinoline’s nitrogen attracts the
Thr98′s side chain methyl group by means of the induced dipole of low
intensity. The later described positioning of the main core is preserved
with some minor alterations within the bioactive conformations of all
the remaining quinoline-4-carboxylic acids. Furthermore, the extension
of the quinoline in the form of another aromatic scaffold has a significant
impact on the activity of 3g. Thus, the second aromatic portion is
involved, on the quinoline’s nitrogen side, in the eclipsed hydrophobic
interactions with Leu359 and the parallel-displaced hydrophobic in-
teractions with Phe98, while on the C-4 carboxylic acid side the scaffold
makes the repulsive interactions with Leu46 (edge to face) and Met34
(face to face). Throughout the entire set of compounds, the second ar-
omatic scaffold more-less retains its spatial arrangement. Such constel-
lation of the second aromatic moiety influences the C-3′ isobutyl
positioning in the proximity of Leu58 and Phe62, while the C-4′ isobutyl
portion is narrow to Phe62 and Leu68. The ether oxygen linkers are in
both scenarios attracted by means of the electrostatic interactions with
Tyr38. The bio-conformation of 3f (Fig. 4) matches the one of 3g. While
retaining the already described hydrogen bond with
ω-N (this time
slightly weaker, dHB = 2.779 Å), 3f is narrow to the second medium
strength hydrogen bond with Arg136′s δ-N (dHB = 3.176 Å). Despite the
additional advantageous interaction, the slightly lower activity of 3f
related to 3g can be attributed to the double n-butyl substitution of the
second aromatic moiety (i.e. to the absence of one methyl portion in
each of the C-3′ and C-4′ substituents).
Despite embracing the comparable conformation to the ones of 3g
Fig. 4. The bioactive conformation of 3f as hDHODH
inhibitor. For the clarity of presentation, only the N-
terminal amino acids, orotate as hDHODH substrate,
and FMN as hDHODH coenzyme were depicted,
whereas the remaining of the enzyme is illustrated
with surface. Only polar hydrogens were preserved
for presentation. The established hydrogen bonds are
presented with blue lines. (For interpretation of the
references to colour in this figure legend, the reader is
referred to the web version of this article.)
6

´
M.M. Petrovic et al.
Bioorganic Chemistry 105 (2020) 104373
and 3f, 3a’s lower biopotential is primary a consequence of weaker
inhibitory activity, 3n (Supporting Information Fig. S40), likewise es-
tablishes the corresponding hydrogen bond with Gln47 (dHB = 2.908 Å),
via the C-4 carboxylic acid’s –OH, alongside with the ones with Arg136′s
hydrogen bonds with Arg136′s
ω-N and δ-N (dHB = 3.279 and 3.263 Å,
respectively, Supporting Information Fig. S29). Moreover, 3a’s 2′,3′-
benzyloxy disubstitution, i.e. the introduction of the third aromatic
portion, in general, pushes the newly incorporated phenyl scaffolds to-
wards the outskirts of the tunnel, viz. closer to Leu46 and Leu58
(eclipsed interactions of the 2′-substituent), Leu42 (T-shaped in-
teractions of both substituents), and Phe62 (edge to face interactions of
3′-substituent). The corresponding 3′,4′-disubstitution (3d, Supporting
Information Fig. S32) makes no significant differences within the nar-
row tunnel doorway, but the displacement of the C-2′-substituent to the
vicinal position (i.e. the 3a to 3d transformation) causes the spatial
lateral movement of the quinoline ring towards the redox site for 0.553
Å by average, ultimately causing the loss of the hydrogen bond with the
Arg136′s δ-N and the alleviation of bioactivity in comparison with 3g
and 3f. The importance of the proper positioning of benzyloxy scaffolds
is the most expressively depicted with the consideration of 3b’s biopose
(Supporting Information Fig. S30). Thus, the benzyloxy-based 2′,4′-
disubstitution for the distinct hDHODH inhibitor provides even three
weak hydrogen bonds via the C-4 carboxylic acid in the depth of the
narrow tunnel: both carbonyl and hydroxyl groups are engaged with
ω
-N and δ-N (dHB = 3.044 and 3.239 Å, respectively). Still, compound’s
significantly lower activity, related to the most active hits, is most
certainly a consequence of the absence of bulky groups: the C-3′-
methoxy and the C-4′-hydroxy portions have no power to engage the
hydrophobic residues at the entrance of the narrow tunnel in full. The
same conclusion can be made for 3m (Supporting Information Fig. S39),
as well: despite the numerous already elaborated hydrogen bonds via the
C-4 carboxylic acid with the sub-sites 2 and 3 (two with Arg136′s
ω-N:
dHB = 3.054, and 3.164 Å, respectively; one with Arg136′s δ-N: dHB
=
3.235 Å; one with Glu47: dHB = 2.990 Å; one with the peptide nitrogen
of Thr360: dHB = 2.056 Å), the total absence of hydrophobic C-3′/C-4′
disubstituents reduces the activity of the compound in great manner.
Further on, the C-2′/C-4′ disubstitution with methoxy portions (3c,
Supporting Information Fig. S31) slightly displaces molecule and in-
fluences the reduction in a C-4 carboxylic acid-enabled hydrogen bonds
number: only the one with Arg136′s
ω-N and the bond with Glu47
remain, dHB = 3.111 and 3.186 Å, respectively. The prolongation of the
C-4′ substituent for two methylene residues (3j, Supporting Information
Fig. S36), accompanied with the displacement of the methoxy scaffold
form C-2′ towards C-3′, influences the formation of four C-4 carboxylic
Arg136′s
ω-N (dHB = 3.046 and 3.199 Å, respectively); hydroxyl group
attracts the Arg136′s δ-N, as well (dHB = 3.283 Å). Paradoxically, there is
a decrease in the activity of 3b related to 3d, implying that benzyloxy
substituents are perhaps too bulky for the narrow tunnel, generally
speaking.
acid-provided hydrogen bonds (two with Arg136′s
ω-N: dHB = 2.713 and
3.217 Å, respectively; one with Arg136′s δ-N: dHB = 3.122; one with
Glu47: dHB = 3.032 Å). Still, the non-voluminous C-3′/C-4′ dis-
ubstituents are not enough to provide the high activity against
hDHODH. The importance of bulky substituents at positions C-3′ and C-
4′ is the most clearly seen in the bioactive conformation of 3i (Sup-
porting Information Fig. S35, C-3′/C-4′-dimethoxy substitution), the
least active hDHODH inhibitor, capable to establish three hydrogen
bonds via the C-4 carboxylic acid with a narrow tunnel (one with
Still, the substituents of the second aromatic portion should not be
spatially insignificant. Within the structure of 3k (Supporting Informa-
tion Fig. S37) there is the methoxy group at the position C-3′ and the
butyloxy portion attached to the C-4′. The C-3′ portion gives low, but
still a contribution while interacting with Leu46, Leu58, and Phe62,
whereas the C-4′ butyloxy scaffold is more tightly bound with Phe62 and
Leu68. The alleviation in bulkiness somehow increases the strength of
Arg136′s
ω
-N: dHB = 2.818 Å; one with Arg136′s δ-N: dHB = 3.190 Å; one
the hydrogen bond towards the Arg136′s
ω
-N (dHB = 2.647 Å), proving
with Thr360′s peptide nitrogen: dHB = 2.286 Å).
the later postulate. The precise hydrogen bond and the C-4′ contributors
retain the activity of 3k on a high level, although lower related to pre-
decessors. The next ranked compound by activity is 3e (Supporting In-
formation Fig. S33), differing related to 3g (Fig. 3) in the C-3′,4′-propoxy
disubstitution. While the distinct substituents generate the expected
hydrophobic interactions with Leu46, Leu58, Phe62, and Leu68, the
3e’s C-4 carboxylic acid scaffold establishes even four hydrogen bonds
3. Conclusion
In summary, a new class of 2-substituted quinoline-4-carboxylic
acids containing two alkyl/aryl ether linkages was designed, synthe-
sized and evaluated for its hDHODH inhibitory activity and anti-
proliferative potential against three human cancer cells and one normal
cell line. Protocatechuic aldehyde and vanillin were freely available
starting compounds for a three-component Doebner reaction. Quinoline
derivatives 3f and 3g from protocatechuic aldehyde series were the most
effective hDHODH inhibitors while 3k and 3l from vanillin precursor
had high cytotoxic activity against MCF-7 and A375 cells and good
selectivity index. The presence of two hydrophobic groups such as n-
butyl and isobutyl is crucial for excellent enzyme inhibition. However,
only one hydrophobic substituent (3k) led to good selectivity in cyto-
toxic action retaining potent cytotoxicity and hDHODH inhibitory ac-
tivity. Generally, good hDHODH inhibitors were also good cytotoxic
agents suggesting a correlation between hDHODH inhibition and tumor
cell growth particularly in A375 and MCF-7 cells. In addition, the lipo-
philic properties of 3a-n significantly influence on DHODH activity
because this enzyme is located in the mitochondrial membrane and
compound must be able to reach to mitochondria to inhibit DHODH. The
optimal lipophilicity of 3f and 3g with their logD7.4 = 1.49 and 1.44
respectively was responsible for an excellent DHODH inhibition. The
cytotoxic activity is mainly controlled by lipophilicity only in the narrow
range of logD7.4 values from 0.95 to 1.49. The more detailed structure-
activity relationship studies were performed on the structure-based
level, aiming to reveal the individual contribution of compounds’
functional groups while interfering with hDHODH (in particular with
the narrow tunnel within the N-terminal domain). Upon conducting the
molecular docking, a high level of correlation between ligand-based and
structure-based SAR studies was achieved. Hence, while elaborating the
with the sub-sites 2 and 3 (two with Arg136′s
ω-N: dHB = 2.740, and
3.206 Å, respectively; one with Arg136′s δ-N: dHB = 3.112 Å; one with
the peptide nitrogen of Thr360: dHB = 2.008 Å). In other words, 3e
should be the star of the entire study. However, it seems that the
reduction of C-3′,4′-disubstituents’ voluminosity diminishes the 3e’s
potency on the structure-based level. Furthermore, the C-3′-methoxy
group/C-4′-isobutyloxy portion ensemble is the structural marker of 3l
(Supporting Information Fig. S38), taking some features from 3k (Sup-
porting Information Fig. S37, the C-3′-methoxy group) and 3g (Fig. 3,
the C-4′-isobutyloxy portion). Albeit the 3l’s hydrophobic interactions
resemble the ones of 3k (Supporting Information Fig. S37) and 3g
(Fig. 3), the inhibitor’s C-4 carboxylic acid has a feature to create
hydrogen bonds both with Arg136′s δ-N (via the C-4 carboxylic acid’s
–
-C O: dHB = 2.779 Å) and Gln47 (via the C-4 carboxylic acid’s –OH: dHB
–
= 3.023 Å). Despite auspicious interactions, the opinion is that the C-3′-
methoxy group existence alleviates 3l’s biopotential. An interesting
conformation was adopted by 3h (Supporting Information Fig. S34), as
well. Within, the C-3′-methoxy group has been totally overpowered by
the C-4′-benzyloxy moiety, which is involved in the parallel displaced
interactions with Phe62, and some weaker interactions with Leu42,
Leu46, Leu58, and Leu68. The C-3′-methoxy group, as a runner up, in-
terferes only with Phe98 and Leu359. Latter interactions and only one
hydrogen bond with Arg136′s
ω-N (dHB = 2.261 Å) place this inhibitor in
the midst of the herein activity scale.
The next ranked compound by means of the decreased hDHODH
7

´
M.M. Petrovic et al.
Bioorganic Chemistry 105 (2020) 104373
interactions of either quinoline’s C-4 carboxylic acid with narrow tun-
nel’s sub-sites 2 and 3, or quinoline’s C-2 Ph-C2′/C3′/C4′-disubstituents
with tunnel’s sub-site 1, structural characteristics responsible for
providing the hDHODH inhibitory activity for any of targeted com-
pounds were determined alongside with the guidelines for further
rational design of innovative hDHODH inhibitors. The obtained bio-
conformations of 3a-n were in high agreement with the previously
experimentally determined crystals structures of either clinical drug
leflunomide or co-crystalized 4-quinoline carboxylic acid derivatives.
DMSO‑d6): 4.93 (s, 2H, CH2), 5.27 (s, 2H, CH2), 7.14 (s, 5H, Ar-H), 7.25
(d, 1H, J = 8.4 Hz, Ar-H), 7.31–7.47 (m, 5H, Ar-H), 7.54–7.57 (m, 2H,
Ar-H), 7.69–7.76 (m, 1H, Ar-H), 7.81–7.88 (m, 1H, Ar-H), 8.13 (d, 1H, J
= 8.4 Hz, Ar-H), 8.39 (s, 1H, Ar-H), 8.73 (d, 1H, J = 8.4 Hz, Ar-H), 13.89
(bs, 1H, COOH); 13C NMR (50 MHz, DMSO‑d6): 70.45, 74.92, 115.55,
122.67, 123.44, 123.84, 124.53, 125.46, 127.88 (4C), 127.98 (4C),
128.24 (2C), 128.49 (2C), 129.84 (2C), 134.07, 135.88, 136.91, 146.03,
148.46, 152.09, 155.83, 167.58; Anal. Calcd. For C30H23NO4 (461.51 g/
mol): C, 78.07; H, 5.02; N, 3.04; Found: C, 77.85; H, 4.99; N, 3.05.
4. Experimental section
4.1.2.2. 2-(2,4-Bis(benzyloxy)phenyl)quinoline-4-carboxylic acid (3b).
Yellow powder; yield: 0.24 g (51%); mp 231–232 C; IR (KBr, cmꢀ 1):
◦
3432, 1702, 1604, 1507, 1187, 1002, 698; 1H NMR (200 MHz,
DMSO‑d6): 5.22 (s, 2H, CH2), 5.27 (s, 2H, CH2), 6.84 (d, 1H, J = 8.6 Hz,
Ar-H), 7.00 (s, 1H, Ar-H), 7.33–7.52 (m, 10H, Ar-H), 7.63–7.70 (m, 1H,
Ar-H), 7.76–7.84 (m, 1H, Ar-H), 7.90 (d, 1H, J = 8.6 Hz, Ar-H), 8.11 (d,
1H, J = 8.2 Hz, Ar-H), 8.55 (s, 1H, Ar-H), 8.68 (d, 1H, J = 8.2 Hz, Ar-H),
13.78 (bs, 1H, COOH); 13C NMR (50 MHz, DMSO‑d6): 69.72, 70.11,
100.97, 107.30, 121.12, 123.10, 124.05, 125.37, 127.38, 127.49 (2C),
127.82, 127.88 (2C), 128.00, 128.43 (2C), 128.53 (2C), 129.62, 129.71,
132.16, 135.35, 136.64, 136.86, 148.58, 155.64, 157.40, 160.92,
167.73; Anal. Calcd. For C30H23NO4 (461.51 g/mol): C, 78.07; H, 5.02;
N, 3.04; Found: C, 77.92; H, 5.00; N, 3.05.
4.1. Chemistry
4.1.1. Physical measurements and methods
Melting points were determined on a Mel-Temp capillary melting
points apparatus, model 1001, and are uncorrected. Elemental (C, H, N,
S) analysis of the samples was carried out in the Center for Instrumental
Analysis, Faculty of Chemistry, Belgrade. IR spectra were obtained on a
Perkin Elmer Spectrum One FT-IR spectrometer with a KBr disc. 1H and
13C NMR spectra were recorded on a Varian Gemini 200 MHz spec-
trometer. UV–Vis spectra were recorded on a double beam UV–Vis
spectrophotometer model Cary 300 (Agilent Technologies, Santa Clara,
USA) with 1.0 cm quartz cells.
4.1.2.3. 2-(2,4-Dimethoxyphenyl)quinoline-4-carboxylic acid (3c). Light
yellow powder; yield: 0.11 g (36%); mp 181–182 C; IR (KBr, cmꢀ 1):
◦
4.1.2. General procedure for synthesis quinoline-4-carboxylic acids 3a-n
Procedure for the synthesis of 2a-l [19]. To the solution of aldehyde
1a-d (2.2 mmol) in DMF (5 mL), K2CO3 (1.216 g, 8.8 mmol for 1a-c;
0.608 g, 4.4 mmol for 1d) was added, followed by the corresponding
alkyl halide (MeI, nPrBr, nBuBr, iBuBr or BnCl) (for 1a-c 5.07 mmol; for
1d 2.53 mmol). The resulting mixture was then refluxed for 5 h. After
completion, the suspension was filtrated, the precipitate was discarded,
and 25 mL of cold water was added to the filtrate. The pH of the solution
was adjusted to 4 with HCl aqueous solution (2 M), upon which a pre-
cipitate was formed in the case of compounds 2a-d, 2g, and 2h. The
formed suspension was then stirred for 1 h at the room temperature,
filtrated, washed with a small amount of cold water, and dried over
CaCl2. In the case of compounds 2e, 2f, 2i and 2j-l, the solution was
extracted with ethyl acetate (2 × 30 mL) after acidification, and the
organic portion was dried over Na2SO4. Afterward, the solvent was
evaporated under reduced pressure, and the obtained compounds 2e, 2f,
2i, and 2j-l were dried over CaCl2. The aldehydes 2f-h, 2j-l were purified
by column chromatography (stationary phase: silica gel, eluent: chlo-
roform), while the other compounds 2a-e and 2i were obtained with
satisfactory purity and were used for further synthesis without
purification.
3431, 1705, 1601, 1309, 1202, 1035, 821; 1H NMR (200 MHz,
DMSO‑d6): 3.86 (s, 3H, CH3), 3.90 (s, 3H, CH3), 6.73 (d, 1H, J = 8.2 Hz,
Ar-H), 6.75 (s, 1H, Ar-H), 7.63–7.71 (m, 1H, Ar-H), 7.77–7.90 (m, 2H,
Ar-H), 8.11 (d, 1H, J = 8.2 Hz, Ar-H), 8.39 (s, 1H, Ar-H), 8.67 (d, 1H, J =
8.2 Hz, Ar-H), 13.84 (bs, 1H, COOH); 13C NMR (50 MHz, DMSO‑d6):
55.58, 56.03, 98.89, 106.17, 120.52, 123.05, 123.78, 125.38, 127.33,
129.62, 129.69, 132.09, 135.48, 148.54, 155.72, 158.51, 161.96,
167.82; Anal. Calcd. For C18H15NO4 (309.32 g/mol): C, 69.89; H, 4.89;
N, 4.53; Found: C, 69.61; H, 4.91; N, 4.52.
4.1.2.4. 2-(3,4-Bis(benzyloxy)phenyl)quinoline-4-carboxylic acid (3d).
Yellow powder; yield: 0.18 g (38%); mp 174–175 C; IR (KBr, cmꢀ 1):
◦
3432, 1713, 1599, 1267, 1141, 1023, 734; 1H NMR (200 MHz,
DMSO‑d6): 5.24 (s, 2H, CH2), 5.30 (s, 2H, CH2), 7.24 (d, 1H, J = 8.6 Hz,
Ar-H), 7.30–7.56 (m, 10H, Ar-H), 7.62–7.70 (m, 1H, Ar-H), 7.79–7.89
(m, 2H, Ar-H), 8.04 (s, 1H, Ar-H), 8.12 (d, 1H, J = 8.4 Hz, Ar-H), 8.41 (s,
1H, Ar-H), 8.59 (d, 1H, J = 8.4 Hz, Ar-H), 13.99 (bs, 1H, COOH); 13
C
NMR (50 MHz, DMSO‑d6): 70.16, 70.55, 113.26, 114.32, 118.86,
120.84, 123.22, 125.44, 127.40, 127.62 (2C), 127.79 (2C), 127.90 (2C),
128.47 (4C), 129.66, 130.15, 131.10, 137.11, 137.35, 137.68, 148.36,
148.65, 150.25, 155.42, 167.79; Anal. Calcd. For C30H23NO4 (461.51 g/
mol): C, 78.07; H, 5.02; N, 3.04; Found: C, 78.13; H, 5.01; N, 3.03.
Procedure for the synthesis of 3a-n. A mixture of the corresponding
aldehyde 1a, 1d, and 2a-l (1 mmol) and freshly distilled pyruvic acid
(0.132 g, 1.5 mmol) in absolute ethanol (2 mL) was refluxed for 15 min.
After cooling the flask, the solution of aniline (0.093 g, 1 mmol) in ab-
solute ethanol (1 mL) was added, and the mixture was refluxed for 3 h.
4.1.2.5. 2-(3,4-Dipropoxyphenyl)quinoline-4-carboxylic acid (3e). Yel-
low powder; yield: 0.11 g (30%); mp 134–135 ◦C; IR (KBr, cmꢀ 1): 3431,
2964, 1710, 1599, 1272, 1143, 977; 1H NMR (200 MHz, DMSO‑d6): 1.01
(t, 3H, J = 7.4 Hz, CH3), 1.03 (t, 3H, J = 7.4 Hz, CH3), 1.69–1.88 (m, 4H,
CH2CH3), 4.02 (t, 2H, J = 6.4 Hz, OCH2), 4.08 (t, 2H, J = 6.4 Hz, OCH2),
7.12 (d, 1H, J = 8.4 Hz, Ar-H), 7.62–7.70 (m, 1H, Ar-H), 7.78–7.90 (m,
3H, Ar-H), 8.13 (d, 1H, J = 8.2 Hz, Ar-H), 8.42 (s, 1H, Ar-H), 8.59 (d, 1H,
J = 8.2 Hz, Ar-H), 13.97 (bs, 1H, COOH); 13C NMR (50 MHz, DMSO‑d6):
10.54, 10.63, 22.30, 22.45, 69.94, 70.29, 112.51, 113.54, 118.80,
120.58, 123.15, 125.38, 127.28, 129.62, 130.07, 130.66, 137.54,
148.36, 148.88, 150.63, 155.54, 167.74; Anal. Calcd. For C22H23NO4
(365.42 g/mol): C, 72.31; H, 6.34; N, 3.83; Found: C, 72.18; H, 6.36; N,
3.82.
◦
Afterward, the flask was allowed to stand at 4 C overnight, and the
formed precipitate of the corresponding compounds 3a-n was then
filtered, washed with a small amount of cold ethanol and dried over
CaCl2. The final compounds were obtained with satisfactory purity.
However, in order to obtain compounds with very high purity, com-
pounds 3a-d, 3f, 3g, 3i, and 3j-l can be subjected to further purification
by dissolving in a small amount of chloroform with mild heating and
reprecipitation with hexane, 3n by dissolving in acetone and repreci-
pitation with hexane, while 3h can be purified by recrystallization from
70% aqueous solution of EtOH.
4.1.2.1. 2-(2,3-Bis(benzyloxy)phenyl)quinoline-4-carboxylic acid (3a).
Beige powder; yield: 0.11 g (24%); mp 191–192 ◦C; IR (KBr, cmꢀ 1):
3431, 1703, 1463, 1263, 1209, 1039, 754; 1H NMR (200 MHz,
4.1.2.6. 2-(3,4-Dibutoxyphenyl)quinoline-4-carboxylic acid (3f). Yellow
powder; yield: 0.15 g (38%); mp 146–147 ◦C; IR (KBr, cmꢀ 1): 3431,
8

´
M.M. Petrovic et al.
Bioorganic Chemistry 105 (2020) 104373
2957, 1704, 1599, 1272, 1143, 856; 1H NMR (200 MHz, DMSO‑d6): 0.95
(t, 3H, J = 7,2 Hz, CH3), 0.96 (t, 3H, J = 7.4 Hz, CH3), 1.38–1.59 (m, 4H,
CH2CH3), 1.67–1.82 (m, 4H, CH2 CH2CH3), 4.05 (t, 2H, J = 6.4 Hz,
OCH2), 4.11 (t, 2H, J = 6.4 Hz, OCH2), 7.12 (d, 1H, J = 8.6 Hz, Ar-H),
7.62–7.70 (m, 1H, Ar-H), 7.78–7.90 (m, 3H, Ar-H), 8.13 (d, 1H, J = 8.4
Hz, Ar-H), 8.42 (s, 1H, Ar-H), 8.59 (d, 1H, J = 8.4 Hz, Ar-H), 13.97 (bs,
1H, COOH); 13C NMR (50 MHz, DMSO‑d6): 13.84 (2C), 18.90 (2C),
30.97, 31.10, 68.21, 68.54, 112.58, 113.58, 118.78, 120.58, 123.14,
125.36, 127.26, 129.61, 130.04, 130.68, 137.53, 148.35, 148.90,
150.67, 155.54, 167.71; Anal. Calcd. For C24H27NO4 (393.48 g/mol): C,
73.26; H, 6.92; N, 3.56; Found: C, 72.99; H, 6.94; N, 3.55.
4.1.2.11. 2-(4-Butoxy-3-methoxyphenyl)quinoline-4-carboxylic acid
1.5H2O (3k). Light orange powder; yield: 0.14
×
g
(36%); mp
180–181 ◦C; IR (KBr, cmꢀ 1): 3468, 2940, 1703, 1598, 1371, 1278, 1025;
1H NMR (200 MHz, DMSO‑d6):): 0.94 (t, 3H, J = 7.2 Hz, CH3), 1.36–1.55
(m, 2H, CH2CH3), 1.67–1.80 (m, 2H, CH2 CH2CH3), 3.91 (s, 3H,
CH3),4.04 (t, 2H, J = 6.4 Hz, OCH2), 7.11 (d, 1H, J = 8.4 Hz, Ar-H),
7.62–7.69 (m, 1H, Ar-H), 7.78–7.90 (m, 3H, Ar-H), 8.13 (d, 1H, J =
8.2 Hz, Ar-H), 8.43 (s, 1H, Ar-H), 8.59 (d, 1H, J = 8.2 Hz, Ar-H), 13.93
(bs, 1H, COOH); 13C NMR (50 MHz, DMSO‑d6): 13.84, 18.94, 30.98,
55.88, 68.08, 110.68, 112.89, 118.84, 120.41, 123.21, 125.42, 127.28,
129.64, 130.07, 130.56, 137.55, 148.40, 149.38, 150.25, 155.57,
167.78; Anal. Calcd. For C21H21NO4×1.5H2O (378.43 g/mol): C, 66.65;
H, 6.39; N, 3.70; Found: C, 66.84; H, 6.37; N, 3.71.
4.1.2.7. 2-(3,4-Diisobutoxyphenyl)quinoline-4-carboxylic acid (3g). Yel-
low powder; yield: 0.17 g (42%); mp 160–161 ◦C; IR (KBr, cmꢀ 1): 3439,
2958, 1716, 1598, 1272, 1025, 801; 1H NMR (200 MHz, DMSO‑d6): 1.02
(d, 6H, J = 6.6 Hz, CH3), 1,04 (d, 6H, J = 6.6 Hz, CH3), 1.97–2.18 (m,
2H, CHCH3), 3.83 (d, 2H, J = 6.4 Hz, OCH2), 3.90 (d, 2H, J = 6.4 Hz,
OCH2), 7.11 (d, 1H, J = 8.4 Hz, Ar-H), 7.62–7.70 (m, 1H, Ar-H),
7.78–7.88 (m, 3H, Ar-H), 8.13 (d, 1H, J = 8.2 Hz, Ar-H), 8.42 (s, 1H,
Ar-H), 8.59 (d, 1H, J = 8.2 Hz, Ar-H), 13.97 (bs, 1H, COOH); 13C NMR
(50 MHz, DMSO‑d6): 19.14 (2C), 19.22 (2C), 28.09, 28.25, 74.63, 75.00,
112.44, 113.50, 118.78, 120.58, 123.15, 125.38, 127.28, 129.62,
130.08, 130.67, 137.60, 148.36, 149.05, 150.78, 155.56, 167.74; Anal.
Calcd. For C24H27NO4 (393.48 g/mol): C, 73.26; H, 6.92; N, 3.56; Found:
C, 73.53; H, 6.94; N, 3.57.
4.1.2.12. 2-(4-Isobutoxy-3-methoxyphenyl)quinoline-4-carboxylic
acid
(3l). Yellow powder; yield: 0.10 g (29%); mp 184–185 ◦C; IR (KBr,
cmꢀ 1): 3439, 2959, 1722, 1600, 1255, 1022, 774; 1H NMR (200 MHz,
DMSO‑d6): 1.00 (d, 6H, J = 7.2 Hz, CH3), 1.96–2.16 (m, 1H, CHCH3),
3.81 (d, 2H, J = 6.6 Hz, OCH2), 3.92 (s, 3H, CH3), 7.10 (d, 1H, J = 8.4
Hz, Ar-H), 7.62–7.69 (m, 1H, Ar-H), 7.78–7.91 (m, 3H, Ar-H), 8.13 (d,
1H, J = 8.2 Hz, Ar-H), 8.43 (s, 1H, Ar-H), 8.60 (d, 1H, J = 8.2 Hz, Ar-H),
13.96 (bs, 1H, COOH); 13C NMR (50 MHz, DMSO‑d6): 19.23 (2C), 27.96,
56.00, 74.66, 110.85, 113.04, 118.83, 120.45, 123.20, 125.42, 127.27,
129.64, 130.06, 130.58, 137.54, 148.39, 150.38, 155.55, 149.42,
167.77; Anal. Calcd. For C21H21NO4 (351.40 g/mol): C, 71.78; H, 6.02;
N, 3.99; Found: C, 71.72; H, 6.03; N, 3.98.
4.1.2.8. 2-(4-(Benzyloxy)-3-methoxyphenyl)quinoline-4-carboxylic acid
× H2O (3h). Yellow powder; yield: 0.10 g (25%); mp 182–183 ◦C; IR
(KBr, cmꢀ 1): 3438, 1708, 1600, 1373, 1272, 1024, 741; 1H NMR (200
MHz, DMSO‑d6): 3.93 (s, 3H, CH3), 5.20 (s, 2H, CH2), 7.22 (d, 1H, J =
8.4 Hz, Ar-H), 7.32–7.52 (m, 5H, Ar-H), 7.62–7.70 (m, 1H, Ar-H),
7.79–7.93 (m, 3H, Ar-H), 8.14 (d, 1H, J = 8.4 Hz, Ar-H), 8.44 (s, 1H,
Ar-H), 8.59 (d, 1H, J = 8.4 Hz, Ar-H), 13.98 (bs, 1H, COOH); 13C NMR
(50 MHz, DMSO‑d6): 55.90, 70.09, 110.72, 113.58, 118.83, 120.30,
123.19, 125.41, 127.33, 127.89 (2C), 127.95, 128.48 (2C), 129.64,
130.10, 131.01, 136.94, 137.65, 148.35, 149.54, 149.78, 155.51,
167.76; Anal. Calcd. For C24H19NO4 × H2O (403.43 g/mol): C, 71.45; H,
5.25; N, 3.47; Found: C, 71.19; H, 5.23; N, 3.48.
4.1.2.13. 2-(3,4-Dihydroxyphenyl)quinoline-4-carboxylic acid (3m). Or-
ange powder; yield: 0.10 g (34%); mp > 250 ◦C; IR (KBr, cmꢀ 1): 3289,
2620, 1644, 1586, 1339, 1220, 876; 1H NMR (200 MHz, DMSO‑d6): 6.90
(d, 1H, J = 8.2 Hz, Ar-H), 7.57–7.67 (m, 2H, Ar-H), 7.76–7.84 (m, 2H,
Ar-H), 8.07 (d, 1H, J = 8.0 Hz, Ar-H), 8.32 (s, 1H, Ar-H), 8.62 (d, 1H, J =
8.0 Hz, Ar-H), 9.30 (bs, 1H, OH), 9.43 (bs, 1H, OH); 13C NMR (50 MHz,
DMSO‑d6): 114.53, 116.22, 118.97, 119.28, 123.29, 125.63, 127.24,
129.56, 129.66, 130.21, 137.30, 148.10, 146.05, 148.71, 156.09,
167.95; Anal. Calcd. For C16H11NO4 (281.26 g/mol): C, 68.32; H, 3.94;
N, 4.98; Found: C, 68.17; H, 3.95; N, 4.96.
4.1.2.14. 2-(4-Hydroxy-3-methoxyphenyl)quinoline-4-carboxylic acid ×
1.5H2O (3n). Brown powder; yield: 0.11 g (35%); mp 233–234 ◦C; IR
(KBr, cmꢀ 1): 3438, 2624, 1715, 1564, 1390, 1031, 759; 1H NMR (200
MHz, DMSO‑d6): 3.93 (s, 3H, CH3), 6.95 (d, 1H, J = 8.2 Hz, Ar-H),
7.60–7.68 (m, 1H, Ar-H), 7.74–7.89 (m, 3H, Ar-H), 8.11 (d, 1H, J =
8.4 Hz, Ar-H), 8.40 (s, 1H, Ar-H), 8.58 (d, 1H, J = 8.4 Hz, Ar-H), 9.57 (s,
1H, OH), 13.71 (bs, 1H, COOH); 13C NMR (50 MHz, DMSO‑d6): 55.97,
111.01, 115.90, 118.74, 120.76, 123.08, 125.41, 127.12, 129.39,
129.56, 130.05, 137.56, 148.16, 148.40, 149.01, 155.82, 167.83; Anal.
Calcd. For C17H13NO4×1.5H2O (322.32 g/mol): C, 63.35; H, 5.00; N,
4.35; Found: C, 63.01; H, 5.01; N, 4.36.
4.1.2.9. 2-(3,4-Dimethoxyphenyl)quinoline-4-carboxylic acid (3i). Yel-
low powder; yield: 0.08 g (26%); mp 235–236 ◦C; IR (KBr, cmꢀ 1): 3438,
2936, 1704, 1594, 1254, 1019, 772; 1H NMR (200 MHz, DMSO‑d6): 3.85
(s, 3H, CH3), 3.92 (s, 3H, CH3), 7.13 (d, 1H, J = 8.4 Hz), 7.62–7.70 (m,
1H, Ar-H), 7.78–7.91 (m, 3H, Ar-H), 8.13 (d, 1H, J = 8.2 Hz, Ar-H), 8.44
(s, 1H, Ar-H), 8.59 (d, 1H, J = 8.2 Hz, Ar-H), 13.97 (bs, 1H, COOH); 13
C
NMR (50 MHz, DMSO‑d6): 55.76, 55.79, 110.39, 111.87, 118.84,
120.39, 123.19, 125.42, 127.31, 129.64, 130.10, 130.64, 137.59,
148.37, 149.20, 150.79, 155.56, 167.78; Anal. Calcd. For C18H15NO4
(309.32 g/mol g/mol): C, 69.89; H, 4.89; N, 4.53; Found: C, 69.71; H,
4.88; N, 4.55.
4.1.3. Determination of lipophilicity
The logD7.4 values were experimentally obtained by the shake-flask
method. A calibration graph was plotted using five different concen-
trations of compound in n-octanol. Then, 0.5 mg (for 3a, 3b, 3d) or 0.8
mg (for 3c, 3e, 3f, 3g, 3h, 3i, 3j, 3k, 3l, 3n) of tested compounds was
4.1.2.10. 2-(3-Methoxy-4-propoxyphenyl)quinoline-4-carboxylic acid
×
H2O (3j). Yellow powder; yield: 0.16 g (44%); mp 175–176 ◦C; IR (KBr,
cmꢀ 1): 3439, 2965, 1715, 1600, 1270, 1144, 1026; 1H NMR (200 MHz,
DMSO‑d6): 1.00 (t, 3H, J = 7.4 Hz, CH3), 1.69–1.87 (m, 2H, CH2CH3),
3.92 (s, 1H, CH3), 4.01 (t, 2H, J = 6.6 Hz, OCH2), 7.11 (d, 1H, J = 8.4 Hz,
Ar-H), 7.62–7.69 (m, 1H, Ar-H), 7.78–7.91 (m, 3H, Ar-H), 8.13 (d, 1H, J
= 8.2 Hz, Ar-H), 8.43 (s, 1H, Ar-H), 8.59 (d, 1H, J = 8.2 Hz, Ar-H), 13.65
(bs, 1H, COOH); 13C NMR (50 MHz, DMSO‑d6): 10.56, 22.24, 55.87,
69.86, 110.67, 112.94, 118.73, 120.40, 123.14, 125.39, 127.25, 129.60,
130.06, 130.54, 137.70, 148.33, 149.35, 150.19, 155.54, 167.75; Anal.
Calcd. For C20H19NO4 × H2O (355.38 g/mol): C, 67.59; H, 5.96; N, 3.94;
Found: C, 67.35; H, 5.95; N, 3.95.
dissolved in 100 μl of DMSO and diluted with n-octanol to 25 mL in a
volumetric flask. From this solution 9.154 mL (for 3a, 3b, 3d) or 3.333
mL (for 3c, 3e, 3f, 3g, 3h, 3i, 3j, 3k, 3l, 3n) was diluted with n-octanol
to 10 mL. The solution for 3m was prepared by dissolving 0.8 mg in 100
μ
l of DMSO and diluted with n-octanol to 25 mL in a volumetric flask.
The absorbance of the compounds in these n-octanol solutions was
measured by UV–Vis spectrophotometry. The biphasic system contain-
ing 4 mL of previously prepared n-octanol solutions and 8 mL of phos-
phate buffer (pH = 7.4) was shaken on a mechanical shaker for 30 min.
After complete phase separation, n-octanol layer was dried over
9

´
M.M. Petrovic et al.
Bioorganic Chemistry 105 (2020) 104373
anhydrous sodium sulphate and absorbance was measured. The con-
centration was calculated from the calibration graph and logD7.4 value
was determined using equation: logD7.4 = log(y/x ꢀ y) where x repre-
sents concentration of compound in n-octanol phase before shaking and
y represents concentration of compound in n-octanol phase after the
treatment on mechanical shaker. For each compound, five independent
measurements were performed.
experimental purposes, the inhibitors were extracted from the complex,
added hydrogens, and Amber parameters were calculated by means of
Antechamber using semiempirical QM method. The protein parts of the
saved monomers were improved by adding hydrogen atoms using the
embedded leap module of Amber 12 suite [35] upon which the correct
hydrogen atoms, appropriate for pH 7.4, were assigned to each amino
acid residue. Upon preparation, proteins were merged with corre-
sponding ligands and complexes were energy minimized as follows:
through the leap module were solvated with water molecules (TIP3P
model, SOLVATEOCT Chimera command) in a box extending 10 Å in all
directions, neutralized with either Na+ or Clꢀ ions, and refined by a
single point minimization using the Sander module of Amber suite with
maximum 1000 steps of the steepest-descent energy minimization and
maximum 4000 steps of conjugate-gradient energy minimization, with a
non-bonded cutoff of 5 Å. From minimized complexes, both ligands and
proteins were extracted to be used for structure-based alignment
assessment and molecular cross-docking experiments. The separated
proteins were used as cross-docking targets, whereas the separated in-
hibitors were utilized to define the cross-docking grid spacing.
4.2. Biology
4.2.1. hDHODH inhibition assay
Recombinant human Dihydroorotate dehydrogenase (hDHODH)
(Fisher Scientific, Austria) was diluted in assay buffer (50 mM Tris, 150
mM KCl and 0.8% Triton® X-100, pH 8.0, Sigma, Austria) to a final
concentration of 2.5 µg/ml. The mixture was transferred into a 96-well
plate. Various amounts of compounds were added and incubated for 30
min at room temperature to allow the inhibitor to react with the protein.
After pre-incubation, substrate mixture (20 mM of L-DHO, 2 mM of DuQ
and 1 mM of DCIP (Fisher Scientific, Austria)) was added to activate the
reaction which was monitored by measuring the absorption at 610 nm
using a GloMax® Multimode Microplate Reader. Leflunomide (Fisher
Scientific, Austria) was used as a positive control. Each inhibitor con-
centration point was tested in triplicate. Finally, IC50 values were
determined using GraphPad Prism 6.0.
The cross-docking on all of the available hDHODH proteins crystals
was performed by applying the cuboid docking grid coordinates pro-
vided from hDHODH inhibitors as follows: the xyz coordinates (in
¨
Ångstroms) for the computation were Xmin/Xmax = ꢀ 48.696/ꢀ 17.176,
Ymin/Ymax = 15.078/ꢀ 31.862, Zmin/Zmax = 19.960/ꢀ 14.897; the
coordinates setup was performed in a manner to embrace the minimized
inhibitor spanning 10 Å in all three dimensions. Upon preparing the
optimal grid, the following setting was used: an energy range of 10 kcal/
mol and exhaustiveness of 100 with RMS Cluster Tolerance of 0.5 Å. The
outputs comprised 20 different conformations. From each set of cross-
docked ligands, the bioactive conformation (i.e. the lowest energy
conformation) of an individual compound was selected by means of
clustering the binding affinity values. The clinical drug leflunomide
(PDB ID 3F1Q) [30] and co-crystalized 4-quinoline carboxylic acid de-
rivatives, brequinar (PDB ID 1D3G) [31], 43 and 46 (PDB IDs 6CJF and
6CJG, respectively) [16] were used as reference bioactive
conformations.
4.2.2. Cell culture and cytotoxic tests
MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bro-
mide, was purchased from Sigma-Aldrich, Austria. All cell lines except
A549 (ATCC: CLL-185) were a gift from Prof. Dr. Barbara Krammer,
University of Salzburg (Austria). MCF-7, A375 and HaCaT cell lines were
cultured in Dulbecco’s modified Eagle’s medium (DMEM)–high glucose
supplemented with 10% fetal bovine serum, 1% penicillin–streptomycin
and 1% L-glutamine. A549 cell line was maintained in Roswell Park
Memorial Institute (RPMI) 1640 medium supplemented with 10% fetal
bovine serum, 1% penicillin–streptomycin and 1% L-glutamine. All
media and supplements were purchased from Sigma-Aldrich, Austria.
Cells were grown in a humidified atmosphere at 37 ◦C in 5% CO2. For all
experiments, 70–80% confluent cells were used. Cells (5000 cells/well)
in complete growth medium were seeded into 96-well culture plates.
The day after, the medium was changed to a serum free medium with
various amounts of the compounds. After an additional incubation
All the herein examined compounds were modelled by applying the
Chemaxon’s msketch module [36] by means of molecular mechanics’
optimization upon which the hydrogen atoms appropriate to pH 7.4,
were assigned. Upon structures’ generation, compounds were uploaded
into previously Vina-based molecular docking protocol to obtain the
bioactive conformations.
period of 72 h, cell viability was determined by adding 10 μl MTT so-
lution (5 mg/ml in PBS) to the compound treated and non-treated
◦
control cells for 2 h at 37 C in the dark. Then the medium was aspi-
rated, cells were lysed with 100 µl DMSO (VWR, Austria) and the
absorbance of the resulting product formazan in viable cells was
measured at 550 nm with a GloMax® Multimode Microplate Reader.
Three independent experiments (with each sample in triplicate) were
performed. Cell viability was normalized to the untreated control and
the IC50 values were determined using GraphPad Prism 6.0.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Acknowledgements
4.3. Molecular docking
This work was supported by the Serbian Ministry of Education, Sci-
ence and Technological Development (Agreement No. 451-03-68/2020-
14/200122).
The molecular docking studies were conducted by means of using the
human dihydroorotate dehydrogenases co-crystalized within the Protein
Data Bank (Supporting Information Table S1). The docking simulations
were carried out by virtue of AutoDock Vina [32] that was selected as
the best performing tool for the structure-based alignment assessment of
co-crystalized hDHODH inhibitors, that has been performed by using the
validated protocols [33]: Experimental Conformation Re-Docking (ECRD),
Randomized Conformation Re-Docking (RCRD), Experimental Conforma-
tion Cross-Docking (ECCD), and Randomized Conformation Cross-Docking
(RCCD) (Supporting Information Experimental section, Structure-based
alignment assessment sub-section).
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.bioorg.2020.104373.
References
[1] J.T. Madak, A. Bankhead, C.R. Cuthbertson, H.D. Showalter, N. Neamati, Revisiting
the role of dihydroorotate dehydrogenase as a therapeutic target for cancer,
Pharmacol. Ther. 195 (2019) 111–131.
The complexes were loaded into UCSF Chimera v1.10.1 software
[34] for Linux 64-bit architecture and visually inspected. For the
10

´
M.M. Petrovic et al.
Bioorganic Chemistry 105 (2020) 104373
[2] R.A.G. Reis, F.A. Calil, P.R. Feliciano, M.P. Pinheiro, M.C. Nonato, The
dihydroorotate dehydrogenases: past and present, Arch. Biochem. Biophys. 632
(2017) 175–191.
[18] L.-M. Wang, L. Hu, H.-J. Chen, Y.-Y. Sui, W. Shen, One-pot synthesis of quinoline-4-
carboxylic acid derivatives in water: ytterbium perfluorooctanoate catalyzed
Doebner reaction, J. Fluor. Chem. 130 (2009) 406–409.
[3] D.B. Sykes, The emergence of dihydroorotate dehydrogenase (DHODH) as a
therapeutic target in acute myeloid leukemia, Expert Opin. Ther. Targets 22 (2018)
893–898.
[19] N. Milhazes, T. Cunha-Oliveira, P. Martins, J. Garrido, C. Oliveira, C. Rego,
F. Borges, Synthesis and cytotoxic profile of 3,4-methylenedioxymethamphetamine
(“Ecstasy”) and its metabolites on undifferentiated PC12 cells: a putative structure-
toxicity relationship, Chem. Res. Toxicol. 19 (2006) 1294–1304.
[4] A. Singh, M. Maqbool, M. Mobashir, N. Hoda, Dihydroorotate dehydrogenase: a
drug for the development of antimalarials, Eur. J. Med. Chem. 125 (2017)
640–651.
[20] S.-F. Chen, F.W. Perrella, D.L. Behrens, L.M. Papp, Inhibition of dihydroorotate
dehydrogenase activity by brequinar sodium, Cancer Res. 52 (1992) 3521–3527.
[21] S.-F. Chen, L.M. Papp, R.J. Ardecky, G.V. Rao, D.P. Hesson, M. Forbes, D.L. Dexter,
Structure-activity relationship of quinoline carboxylic acids. A new class of
inhibitors of dihydroorotate dehydrogenase, Biochem. Pharmacol. 40 (1990)
709–714.
[5] D. Boschi, A.C. Pippione, S. Sainas, M.L. Lolli, Dihydroorotate dehydrogenase
inhibitors in anti-infective drug research, Eur. J. Med. Chem. 183 (2019) 111681.
[6] M.L. Lolli, S. Sainas, A.C. Pippione, M. Giorgis, D. Boschi, F. Dosio, Use of human
dihydroorotate dehydrogenase (hDHODH) inhibitors in autoimmune diseases and
new perspectives in cancer therapy, Recent Pat. Anticancer Drug Discov. 13 (2018)
86–105.
[22] Y. Pan, The dark side of fluorine, ACS Med. Chem. Lett. 10 (2019) 1016–1019.
´
[23] S. Flis, J. Spłwinski, Inhibitory effects of 5-fluorouracil and oxaliplatin on human
[7] M.S. Dorasamy, B. Choudhary, K. Nellore, H. Subramanya, P.F. Wong,
Dihydroorotate dehydrogenase inhibitors target c-myc and arrest melanoma,
myeloma and lymphoma cells at S-phase, J. Cancer 8 (2017) 3086–3098.
[8] V.K. Vyas, M. Ghate, Recent developments in the medicinal chemistry and
therapeutic potential of dihydroorotate dehydrogenase (DHODH) inhibitors, Mini
Rev. Med. Chem. 11 (2011) 1039–1055.
colorectal cancer cell survival are synergistically enhanced by sulindac sulfide,
Anticancer Res. 29 (2009) 435–441.
[24] J.A. Arnott, S.L. Planey, The influence of lipophilicity in drug discovery and design,
Expert Opin. Drug Discov. 7 (2012) 863–875.
[25] D.A. Smith, B.C. Jones, D.K. Walker, Design of drugs involving the concepts and
theories of drug metabolism and pharmacokinetics, Med. Res. Rev. 16 (1996)
243–266.
[9] H. Munier-Lehmann, P.O. Vidalain, F. Tangy, Y.L. Janin, On dihydroorotate
dehydrogenases and their inhibitors and uses, J. Med. Chem. 56 (2013)
3148–3167.
[26] J.E. Comer, High-throughput measurement of logD and pKa, Chapter 2, in: H. van
¨
de Waterbeemd, H. Lennernas, P. Artursson (Eds.), Drug bioavailability: estimation
[10] Y.D. Fragoso, J.B. Brooks, Leflunomide and teriflunomide: altering the metabolism
of pyrimidines for the treatment of autoimmune diseases, Expert Rev. Clin.
Pharmacol. 8 (2015) 315–320.
of solubility, permeability, absorption and bioavailability, Wiley-VCH, Weinheim,
2003.
[27] L. Hitzel, A.P. Watt, K.L. Locker, An increased throughput method for the
determination of partition coefficients, Pharm. Res. 17 (2000) 1389–1395.
[28] L. Di, E.H. Kerns, Profiling drug-like properties in discovery research, Curr. Opin.
Chem. Biol. 7 (2003) 402–408.
[11] D.B. Sykes, Y.S. Kfoury, F.E. Mercier, M.J. Wawer, J.M. Law, M.K. Haynes, T.
A. Lewis, A. Schajnovitz, E. Jain, D. Lee, H. Meyer, K.A. Pierce, N.J. Tolliday,
A. Waller, S.J. Ferrara, A.L. Eheim, D. Stoeckigt, K.L. Maxcy, J.M. Cobert,
J. Bachand, B.A. Szekely, S. Mukherjee, L.A. Sklar, J.D. Kotz, C.B. Clish, R.
I. Sadreyev, P.A. Clemons, A. Janzer, S.L. Schreiber, D.T. Scadden, Inhibition of
dihydroorotate dehydrogenase overcomes differentiation blockade in acute
myeloid leukemia, Cell 167 (2016) 171–186.
[29] B. Jeffries, Z. Wang, J. Graton, S.D. Holland, T. Brind, R.D.R. Greenwood, J.-Y. Le
Questel, J.S. Scott, E. Chiarparin, B. Linclau, Reducing the lipophilicity of
perfluoroalkyl groups by CF2ꢀ F/CF2ꢀ Me or CF3/CH3 exchange, J. Med. Chem. 61
(2018) 10602–10618.
¨
[12] S. Sainas, A.C. Pippione, E. Lupino, M. Giorgis, P. Circosta, V. Gaidano, P. Goyal,
[30] M. Davies, T. Heikkila, G.A. McConkey, C.W.G. Fishwick, M.R. Parsons, A.
¨
D. Bonanni, B. Rolando, A. Cignetti, A. Ducime, M. Andersson, M. Jarvå,
P. Johnson, Structure-based design, synthesis, and characterization of inhibitors of
human and plasmodium falciparum dihydroorotate dehydrogenases, J. Med.
Chem. 52 (2009) 2683–2693.
`
R. Friemann, M. Piccinini, C. Ramondetti, B. Buccinna, S. Al-Karadaghi, D. Boschi,
G. Saglio, M.L. Lolli, Targeting myeloid differentiation using potent 2-
hydroxypyrazolo[1,5-a]pyridine scaffold-based human dihydroorotate
dehydrogenase inhibitors, J. Med. Chem. 61 (2018) 6034–6055.
[13] M. Koundinya, J. Sudhalter, A. Courjaud, B. Lionne, G. Touyer, L. Bonnet,
I. Menguy, I. Schreiber, C. Perrault, S. Vougier, B. Benhamou, B. Zhang, T. He,
Q. Gao, P. Gee, D. Simard, M.P. Castaldi, R. Tomlinson, S. Reiling, M. Barrague,
R. Newcombe, H. Cao, Y. Wang, F. Sun, J. Murtie, M. Munson, E. Yang, D. Harper,
M. Bouaboula, J. Pollard, C. Grepin, C. Garcia-Echeverria, H. Cheng, F. Adrian,
C. Winter, S. Licht, I. Cornella-Taracido, R. Arrebola, A. Morris, Dependence on the
pyrimidine biosynthetic enzyme DHODH is a synthetic lethal vulnerability in
mutant KRAS-driven cancers, Cell Chem. Biol. 25 (2018) 705–717.
[14] P. Das, X. Deng, L. Zhang, M.G. Roth, B.M.A. Fontoura, M.A. Phillips, J.K. De
Brabander, SAR-based optimization of a 4-quinoline carboxylic acid analogue with
potent antiviral activity, ACS Med. Chem. Lett. 4 (2013) 517–521.
[15] M.-G.-A. Shvekhgeimer, The Pfitzinger reaction, Chem. Heterocycl. Compd. 40
(2004) 257–294.
[31] S. Liu, E.A. Neidhardt, T.H. Grossman, T. Ocain, J. Clardy, Structures of human
dihydroorotate dehydrogenase in complex with antiproliferative agents, Structure
8 (2000) 25–33.
[32] O. Trott, A.J. Olson, AutoDock Vina: Improving the speed and accuracy of docking
with a new scoring function, efficient optimization and multithreading, J. Comput.
Chem. 31 (2010) 455–461.
´
[33] M. Mladenovic, A. Patsilinakos, A. Pirolli, M. Sabatino, R. Ragno, Understanding
the molecular determinant of reversible human monoamine oxidase B inhibitors
containing 2H-chromen-2-one core: structure-based and ligand-based derived
three-dimensional quantitative structure-activity relationships predictive models,
J. Chem. Inf. Model. 57 (2017) 787–814.
[34] E.F. Pettersen, T.D. Goddard, C.C. Huang, G.S. Couch, D.M. Greenblatt, E.C. Meng,
T.E. Ferrin, UCSF Chimera - a visualization system for exploratory research and
analysis, J. Comput. Chem. 25 (2004) 1605–1612.
[35] D.A. Case, T.A. Darden, T.E. Cheatham III, C.L. Simmerling, J. Wang, R.E. Duke,
R. Luo, R.C. Walker, W. Zhang, K.M. Merz, B. Roberts, S. Hayik, A. Roitberg,
[16] J.T. Madak, C.R. Cuthbertson, Y. Miyata, S. Tamura, E.M. Petrunak, J.A. Stuckey,
Y. Han, M. He, D. Sun, H.D. Showalter, N. Neamati, Design, synthesis, and
biological evaluation of 4-quinoline carboxylic acids as inhibitors of
dihydroorotate dehydrogenase, J. Med. Chem. 61 (2018) 5162–5186.
¨
¨
´
G. Seabra, J. Swails, A.W. Gotz, I. Kolossvary, K.F. Wong, F. Paesani, J. Vanicek, R.
M. Wolf, J. Liu, X. Wu, S.R. Brozell, T. Steinbrecher, H. Gohlke, Q. Cai, X. Ye,
J. Wang, M.-J. Hsieh, G. Cui, D.R. Roe, D.H. Mathews, M.G. Seetin, R. Salomon-
Ferrer, C. Sagui, V. Babin, T. Luchko, S. Gusarov, A. Kovalenko, P.A. Kollman,
AMBER 12, University of California, San Francisco, 2012.
[17] O. Doebner, Ueber α-alkylcinchoninsauren und α-alkylchinoline, Justus Liebigs
Ann. Chem. 242 (1887) 265–289.
[36] Marvin Beans 15.4.27.0, 2015, ChemAxon. http://www.chemaxon.com (accessed
January 2015).
11