2-phenylsubstituted-3-hydroxyquinolin-4(1H)-one-carboxamides: Structure-cytotoxic activity relationship study

Article abstract of DOI:10.1021/co100013t

A structure-activity relationship of some derivatives of 2-phenylsubstituted- 3-hydroxyquinolin-4(1H)-one-7-carboxamides was systematically studied using combinatorial solid-phase synthesis and in vitro cytotoxic activity screening on representative cance

Full text of DOI:10.1021/co100013t

RESEARCH ARTICLE
pubs.acs.org/acscombsci
2-Phenylsubstituted-3-Hydroxyquinolin-4(1H)-one-Carboxamides:
Structure-Cytotoxic Activity Relationship Study
Miroslav Soural,*,† Jan Hlavꢀaꢁc,† Petr Funk,† Petr Dꢁzubꢀak,‡ and Mariꢀan Hajdꢀuch‡
†Department of Organic Chemistry, Institute of Molecular and Translational Medicine, Faculty of Science, Palackꢀy University,
771 46 Olomouc, Czech Republic
‡Laboratory of Experimental Medicine, Institute of Molecular and Translational Medicine, Faculty of Medicine and Dentistry,
Palackꢀy University and University Hospital in Olomouc, Puꢁskinova 6, 775 20 Olomouc, Czech Republic
S Supporting Information
b
ABSTRACT: A structure-activity relationship of some derivatives of 2-phenylsubstituted-
3-hydroxyquinolin-4(1H)-one-7-carboxamides was systematically studied using combi-
natorial solid-phase synthesis and in vitro cytotoxic activity screening on representative
cancer lines. The effect of substituent type in position 2 as well as of the carboxamide
group was investigated via synthesis of generic libraries constructed with respect to
polarity and bulkiness of appropriate substituents. The process of development afforded a set of compounds with significant cytotoxic
activity. Subsequently, corresponding 2-phenylsubstituted-3-hydroxyquinolin-4(1H)-one-6-carboxamides and 2-phenylsubstituted-
3-hydroxyquinolin-4(1H)-one-8-carboxamides were prepared to evaluate the influence of the carboxamide group position on the
resulting biological activity.
KEYWORDS: hydroxyquinolinones, solid-phase synthesis, combinatorial chemistry, cytotoxicity, structure-activity relationship
’ INTRODUCTION
Compounds derived from the 3-hydroxyquinolin-4(1H)-one
skeleton (hydroxyquinolinones hereafter) represent a novel class
of compounds with interesting biological properties.1 Among
their currently known effects are antiprotozoal, immunosuppres-
sive, and anticancer activities, the latter of which have been
studied the most frequently (Figure 1).
Although a relatively large number of variously substituted
hydroxyquinolinones have been synthesized using solution-
phase synthesis and introduced as biologically promising sub-
stances,2-6 such compounds have never been systematically
studied in terms of their optimal structure. Bearing this in mind
we recently focused our attention on the application of solid-
phase combinatorial chemistry which allows efficient preparation
of sets of derivatives (chemical libraries) and subsequent sys-
Figure 1. Examples of already known hydroxyquinolinones with antic-
tematic structure-activity relationship (SAR) studies. So far we
have introduced two methodologies for solid-phase combinato-
rial synthesis of hydroxyquinolinone derivatives using the split-
and-split concept.7,8 In this study, we describe the subsequent
process of biological testing and development of 2-substituted-
3-hydroxyquinolin-4(1H)-one-7-carboxamides7 (Scheme 1) focused
on determination of their in vitro cytotoxicity toward selected
cancer cell lines followed by fine-tuning the activity through a
substitution of the carboxamide group.
Our general concept for the creation of chemical libraries
is based on focused synthesis of limited sets of compounds
(so-called “generation libraries”) followed by evaluation of their
biological properties and subsequent step-by-step structure/
activity optimization (Scheme 2), rather than on routine pre-
paration of sizable libraries with multiple similar substituents
ancer activities.
with minimum chemical diversity. Such a sensitive procedure
furnishes several advantages: (i) it is timesaving; (ii) it is econom-
ical(especiallywhencommerciallypoorly available buildingblocks
are needed); (iii) the process excludes useless preparation, puri-
fication, and screening often of a large number of derivatives with
unsuitable biological properties.
Since there are only a limited number of substances to be
synthesized and tested according to the above-described con-
cept, the generation libraries are not scheduled to furnish series
of derivatives with full combination of used building blocks.
Received: September 23, 2010
Published: November 10, 2010
r
2010 American Chemical Society
39
dx.doi.org/10.1021/co100013t ACS Comb. Sci. 2011, 13, 39–44
|

ACS Combinatorial Science
RESEARCH ARTICLE
Scheme 1. Building Blocks and General Structure of 2-Substituted-3-hydroxyquinolin-4(1H)-one-7-carboxamides with
Two Diversity Positions7
Scheme 2. Step-by-Step Procedure for Research and Development of Compounds with Cytotoxic Activity
Their goal is rather a trend analysis of each diverse substitution.
An example of this approach for development of biologically
active compounds is given in this article.
Table 1. Summary of the First-Generation Library Cytotoxi-
city Screening (Relative IC50, μM)a
’ RESULTS AND DISCUSSION
With respect to the general concept (Scheme 2) the construc-
tion of the first-generation library was focused on the trend
analysis of substituents in terms of their polar/nonpolar or
hydrophilic/lipophilic properties. For these purposes the carbox-
amide group was substituted with two hydrophilic ligands
(hydroxyethyl, hydroxypropyl), two strictly lipophilic ligands
(4-methylbenzyl, propyl), two heterocyclic ligands of different
polarity (imidazol-N-propyl, piperidine-N-ethyl) and also unsub-
stituted carboxamide was included in the set. To allow the
individual trend analysis of the 7-carboxamide N-substitution
effect, the mentioned ligands were used in combination with a
randomly selected single substituent at position 2 of the hydro-
xyquinolinone moiety (4-methylphenyl in this case). A similar
procedure was used to study the 2 position of the hydroxyqui-
nolinone moiety: variously substituted phenyls were incorporated
into the structure. As in the previous case, the ligands at position
2 were combined with randomly selected single 7-carboxamide
N-substitution (propyl in this case). The outcome of the first
generation library cytotoxicity screening is summarized in
Table 1.
a Average values of IC50 from 3 to 4 independent experiments with SD
ranging from 10 to 25% of the average values. NT: not tested in non-
malignant cells, the IC50 value in chemosensitive CEM leukemia cells
was >1 μM. For cell line characteristics and methodological details
please refer to our previous publications.9-11
The cytotoxic activity was analyzed on the human cancer cell
lines (CEM, acute lymphoblastic leukemia; K562, acute myeloid
leukemia; HCT116, colorectal carcinoma). The most potent
compounds, with IC50 e 1 μM for the most chemosensitive
40
dx.doi.org/10.1021/co100013t |ACS Comb. Sci. 2011, 13, 39–44

ACS Combinatorial Science
RESEARCH ARTICLE
CEM cell line, were also tested on the normal human cells
(BJ line of normal cycling fibroblasts and PBMC - human
peripheral blood mononuclear cells) to analyze the therapeutic
index (TI) which is based on the ratio between the IC50 for
normal human cells and cancer cell lines. The SAR ought to be
evaluated for individual cell lines independently because it may
vary due to different tissue origin and potential molecular targets
in particular cell lines. However, in our case, the trend of activity
was very similar for all tested cell lines indicating similar
or identical target(s), and thus we will discuss SAR generally.
The results from the first-generation library screening show the
following dependence:
(b) Derivatives with lipophilic carboxamide N-substi-
tution (such as propyl, 4-methylbenzyl) exhibit promis-
ing cytotoxicity (relative IC50 = 2.2-3.1 μM); (c) If
the carboxamide is substituted with hydrophilic ligands
(hydroxypropyl, hydroxyethyl) there is no activity (except
very low cytotoxicity of compound 8 against K562: IC50
=
55.1); (d) If the carboxamide is substituted with the two
named heterocyclic ligands the activity is medium or very
low (relative IC50 = 11.9-168.9 μM).
(ii) Trend in the position 2 substitution versus cytotoxic effect
(entries 6, 12-17): When variously substituted aromatic
ligands are incorporated, the cytotoxicity test furnishes
approximately similar results, that is, no significant trend
among the substituents used is observed. However, in a
more detailed comparison the derivatives 16 and 17
(bearing aminogroup or dipropylaminogroup in position 4)
exhibit lower activity against K562 and HCT116. In con-
(i) Trend in the N-carboxamide substitution versus cytotoxic effect
(entries 5-11): (a) If the carboxamide is N-unsubstituted
the activity is rather low (relative IC50 = 9.1-12.9 μM);
Table 2. Summary of the Second-Generation Library
Cytotoxicity Screening (Relative IC50, μM)a
Table 3. Summary of the Third-Generation Library
Cytotoxicity Screening (Relative IC50, μM)a
Cmp.
R1
CEM
K562
HCT116
BJ
PBMC
18
19
20
21
22
23
24
25
26
27
28
29
cyclopropyl
cyclobutyl
cyclopentyl
cyclohexyl
cycloheptyl
methyl
3.7
16.9
13.6
15.3
12.6
9.5
15.1
10.6
12.9
9.4
NT
NT
NT
NT
NT
NT
NT
NT
9.6
NT
NT
NT
NT
NT
NT
NT
NT
96.9
75.9
43.2
57.6
Cmp.
R1
CEM
K562
HCT116
BJ
PBMC
1.2
1.1
30
31
32
33
34
35
36
37
38
39
octyl
decyl
1.0
4.6
4.4
3. 9
4.8
2.1
6.8
6.1
3.2
2.4
2.5
9.2
NT
NT
7.6
NT
NT
10.6
21.1
74.9
NT
NT
94.6
19.1
2.5
1.1
1.1
11.4
10.4
7.0
1.0
10.9
32.1
8.1
dodecyl
0.70
0.98
0.45
1.13
7.5
3.9
25.8
11.6
7.5
cyclooctyl
9.9
isobutyl
1.8
cyclododecyl
2-norbornyl
2-ethylBn
4.1
71.3
NT
NT
9.3
pentyl
1.0
5.6
9.1
benzyl
0.94
0.96
0.75
0.87
6.5
6.6
10.5
6.9
2-Me-Bn
3-Me-Bn
4-Me-Bn
7.6
7.9
16.7
5.0
indan-2-yl
0.93
0.69
0.74
3.1
4.1
naphthyl-1-Me
2.4
52.2
6.9
3.4
7.4
32.7
2,2-diphenylEt
5.3
a Average values of IC50 from 3 to 4 independent experiments with SD
ranging from 10 to 25% of the average values. NT: not tested in non-
malignant cells, the IC50 value in chemosensitive CEM leukemia cells
was >1 μM. For cell line characteristics and methodological details
please refer to our previous publications.9-11
a Average values of IC50 from 3 to 4 independent experiments with SD
ranging from 10 to 25% of the average values. NT: not tested in non-
malignant cells, the IC50 value in chemosensitive CEM leukemia cells
was >1 μM. For cell line characteristics and methodological details
please refer to our previous publications.9-11
Scheme 3. Preparation of 6- and 8-Carboxamidesa
a Reagents: (i) amine, 10% HAc/DMF, overnight, NaBH(OAc)3, 5 h; (ii) 4-amino-3-(methoxycarbonyl)benzoic acid or 2-amino-3-(methoxycarbonyl)-
benzoic acid, DIC, HOBt, DCM, DMF, rt, overnight; (iii) potassium trimethylsilanolate, THF, rt, 7 h; haloketone, TEA, DMF, rt, 2 h; (v) 50% TFA, DCM,
30 min; (vi) TFA, 80 °C, 2 h.
41
dx.doi.org/10.1021/co100013t |ACS Comb. Sci. 2011, 13, 39–44

ACS Combinatorial Science
RESEARCH ARTICLE
Table 4. Completion of the 7-Carboxamides Library
of the substitution of phenyl in position 2, the 4-amino-3,
5-dichloro substitution, which is submicromolar for CEM cells,
was fixed for a detailed study of the carboxamide substitution.
Instead of the propyl, the carboxamide group was substituted
with a set of various lipophilic ligands of aliphatic, cycloaliphatic,
or aromatic structure of different shape/size. The results of the
second-generation library cytotoxicity screening are summarized
in Table 2.
(Relative IC50, μM)a
Whereas the first-generation library results proved the exis-
tence of a carboxamide ligands polarity-activity relationship, the
outcome of the second-generation library screening indicates
also the existence of the shape/size-activity relationship in terms
of the carboxamide lipophilic substitution. For instance, elonga-
tion of the aliphatic chain slightly increases activity against
all lines (methylfpentyl, compare entries 23-25) and expand-
ing the carbocycle moderately increases activity against CEM
(cyclopropylfcyclopentyl, compare entries 18-20). To further
explore this possible relationship we decided to synthesize and
screen a third-generation library focused on the survey of larger
cyclic hydrocarbons (such as cyclooctane, cyclododecane, nor-
bornan, indan, naphthalene, or 2,2-diphenylethan) and longer
aliphatic chains. The outcome of the third-generation library
cytotoxicity screening is summarized in Table 3. The results
indicate that the presence of longer carbon chains as well as larger
carbocycles does not further significantly increase the cytotoxi-
city of the substrates.
a Average values of IC50 from 3 to 4 independent experiments with SD
ranging from 10 to 25% of the average values. For cell line characteristics
and methodological details please refer to our previous publications.9,10
Table 5. Summaryof the 6-CarboxamidesLibrary Cytotoxicity
Screening (Relative IC50, μM)a
For the comparative study of carboxamide structural isomers
described in the second part of this paper, five more derivatives
(entries 40-44) were prepared and screened (Table 4). The results
obtained independently confirm the basic trend in N-carboxamide
substitution versus cytotoxicity effect discussed earlier in this paper
(the first-generation library results).
Carboxamide Group Location versus Cytotoxic Activity
Relationship. Our next goal was to investigate the effect of the
carboxamide group location on the resulting cytotoxicity. For
this purpose we synthesized a set of analogous 2-phenylsubsti-
tuted-3-hydroxyquinolin-4(1H)-one-6-carboxamides and 2-phenyl-
substituted-3-hydroxyquinolin-4(1H)-one-8-carboxamides. Synthe-
sis was performed according to Scheme 3 using the method
previously described.7
As in the case of 7-carboxamides,7 both 6-carboxamides and
8-carboxamides were obtained in an excellent yield (85% on
average) and crude purity (above 95% in each case, LC/MS
traces) without need for further purification. Unfortunately,
the method could not be applied for the preparation of 5-carbox-
amides: When 3-amino-2-(methoxycarbonyl)benzoic acid was
used to acylate immobilized amines (step ii), surprisingly no
reaction was observed. For this reason 2-substituted-3-hydro-
xyquinolin-4(1H)-one-5-carboxamides are not discussed in
this paper. To carry out the SAR study (in terms of the
carboxamide location) as detailed as possible we did not use
the step-by-step development strategy but complete sets of
6- and 8-carboxamides were prepared and screened. The
results obtained from MTT cytotoxicity tests are summar-
ized in Tables 5 and 6.
From the whole library screening results it is evident that a
lot of common features exist for 6-, 7- and 8-carboxamides when
appropriate structural isomers are compared. For instance,
unsubstituted carboxamides (compare entries 40, 45, and 76)
or carboxamides with hydrophilic ligands (compare entries 41,
42 to 47, 48 and 78, 79) exhibit rather medium or weak cyto-
toxicity, whereas generally the best results are again obtained for
a Average values of IC50 from 3 to 4 independent experiments with SD
ranging from 10 to 25% of the average values. NT: not tested in non-
malignant cells, the IC50 value in chemosensitive CEM leukemia cells
was >1 μM. For cell line characteristics and methodological details
please refer to our previous publications.9-11
trast, the derivative 16 shows the best cytotoxicity against
CEM among all the first-generation library members.
Therapeutic index for the most active compounds was
indentified within the range of 2-70.
Our results predetermined derivatives with lipophilic substit-
uents bearing a carboxamide group (and almost any aromatic
ligand at position 2 of the hydroxyquinolinone skeleton) for a
more detailed survey. Because the activity is almost independent
42
dx.doi.org/10.1021/co100013t |ACS Comb. Sci. 2011, 13, 39–44

ACS Combinatorial Science
RESEARCH ARTICLE
Table 6. Summary of the 8-Carboxamides Library Cytotoxi-
K562 and HCT116 in comparison to 7- and 8-isomers
(compare entries 22, 30-34 and 90, 98-102 to 59,
67-71). Corresponding 7-carboxamides exhibit compar-
able activity only for CEM cells.
city Screening (Relative IC50, μM)a
(ii) Brominated derivatives 13 and 83 exhibit similar (medium)
cytotoxicity whereas the corresponding 6-isomer 52 is
totally inactive.
In conclusion, we have synthesized and screened a library
of 2-substituted-3-hydroxyquinolin-4(1H)-one-carboxamides
and investigated the effect of both carboxamide substitution and
its position on the resulting in vitro cytotoxicity. The results show
that the activity depends mainly on the substitution of the carbox-
amide, but in specific cases also the position of the carboxamide on
the quinolinone skeleton substantially changes the biological prop-
erties of the corresponding substrates. Fine-tuning the carboxamide
substitution led to derivatives with submicromolar cytotoxic activity
(relative IC50 about 250 nmol) and some derivatives suitable for
further biological research were identified with highly favorable TI
ranging approximately from 17 to 400 (Table 7).
’ ASSOCIATED CONTENT
S
Supporting Information. Supporting Information con-
b
tains details of the experimental synthetic and screening proce-
dures and spectroscopic data for synthesized compounds. This
material is available free of charge via the Internet at http://
pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*Phone: þ420 585634418. Fax: þ420 585634465. E-mail:
soural@orgchem.upol.cz.
a Average values of IC50 from 3 to 4 independent experiments with SD
ranging from 10 to 25% of the average values. NT: not tested in non-
malignant cells, the IC50 value in chemosensitive CEM leukemia cells
was >1 μM. For cell line characteristics and methodological details
please refer to our previous publications.9-11
’ ACKNOWLEDGMENT
This project was supported in part by the Ministry of
Education, Youth and Sport of the Czech Republic (Grants
MSM6198959216 and ME09057) and FM EEA/Norska
(A/CZ0046/1/0022 and CZ0099). The infrastructural part of
this project (Institute of Molecular and Translational Medicine)
was supported from the Operational Program Research and
Development for Innovations (project CZ.1.05/2.1.00/01.0030).
The human peripheral blood lymphocytes from healthy donors
were obtained based on informed consent form the Transfusion
Department, Faculty Hospital in Olomouc, Czech Republic.
Table 7. Best Compounds Resulting from the SAR Study
’ REFERENCES
(1) Hradil, P.; Hlavꢀaꢁc, J.; Soural, M.; Hajdꢀuch, M.; Kolꢀaꢁr, M.;
Veꢁceꢁrovꢀa, R. Mini-Rev. Med. Chem. 2009, 9, 696–702.
(2) Hradil, P.; Krejꢁcí, P.; Hlavꢀaꢁc, J.; Wiedermannovꢀa, I.; Lyꢁcka, A.;
Bertolasi, V. J. Heterocyclic Chem. 2004, 41, 375–379.
(3) Krejꢁcí, P.; Hradil, P.; Hlavꢀaꢁc, J.; Hajdꢀuch, M. Patent WO
2008028427.
(4) Sui, Z. H.; Nguyen, V. N.; Altom, J.; Fernandez, J.; Hilliard, J. J.;
Bernstein, J. I.; Barrett, J. F.; Ohemeng, K. A. Eur. J. Med. Chem. 1999, 34,
381–387.
(5) Sui, Z.; Jason, A.; Nguyen, V. N.; Fernandez, J.; Bernstein, J. I.;
Hilliard, J. J.; Barrett, J. F.; Podlogar, B. L.; Ohemeng, K. A. Bioorg. Med.
Chem. 1998, 6, 735–742.
(6) Iwanowicz, J. E.; Watterson, S. H.; Murali, T. G.; Pitts, W. J.;
Gu, H. H. Patent WO 2001081340
lipophilic carboxamide substitution. Also substitution of 6- and
7-isomers with heterocyclic ligands (compare entries 43, 44 to 49,
50) gives similar results. Although it may appear that the activity
does not depend on the position of the carboxamide group but
rather on the overall polarity of compounds resulting from the
carboxamide substitution, some specific exceptions can be found:
(i) For longer alkyl chains and larger cycloalkyls (C7-C12),
6-isomers give significantly better results for cell lines
43
dx.doi.org/10.1021/co100013t |ACS Comb. Sci. 2011, 13, 39–44

ACS Combinatorial Science
RESEARCH ARTICLE
(7) Soural, M.; Krchꢁnꢀak, V. J. Comb. Chem. 2007, 9, 793–796.
(8) Krupkova, S.; Soural, M.; Hlavac, J.; Hradil, P. J. Comb. Chem.
2009, 11, 951–955.
(9) Noskova, V.; Dzubak, P.; Kuzmina, G.; Ludkova, A.; Stehlik, D.;
Trojanec, R.; Janostakova, A.; Korinkova, G.; Mihal, V.; Hajduch, M.
Neoplasma 2002, 49, 418–425.
(10) Dzubak, P.; Hajduch, M.; Gazak, R.; Svobodova, A.; Psotova, J.;
Walterova, D.; Sedmera, P.; Kren, V. Bioorg. Med. Chem. 2006, 14, 3793–
3810.
(11) Sarek, J.; Klinot, J.; Dzubak, P.; Klinotovꢀa, E.; Noskovꢀa, V.;
Krecek, V.; Korinkova, G.; Thomson, J. O.; Janostakova, A.; Wang, S.;
Parsons, S.; Fisher, P. M.; Zhelev, N. Z.; Hajduch, M. J. Med. Chem. 2003,
46, 5402–5415.
44
dx.doi.org/10.1021/co100013t |ACS Comb. Sci. 2011, 13, 39–44