Cytotoxicity and QSAR study of (thio)ureas derived from phenylalkylamines and pyridylalkylamines

Article abstract of DOI:10.1007/s00044-012-0402-6

Simplified 1,3-disubstituted urea derivatives (11-24) of phenylethylamines, homoveratylamines, 2-pyridylethylamines, 2-picolylamines as well as xylylenediamines were synthesized and investigated for their cytotoxic activities. The results revealed that mo

Full text of DOI:10.1007/s00044-012-0402-6

MEDICINAL
CHEMISTRY
RESEARCH
Med Chem Res (2013) 22:4016–4029
DOI 10.1007/s00044-012-0402-6
ORIGINAL RESEARCH
Cytotoxicity and QSAR study of (thio)ureas derived
from phenylalkylamines and pyridylalkylamines
•
Ratchanok Pingaew Pan Tongraung Apilak Worachartcheewan
•
•
•
•
•
Chanin Nantasenamat Supaluk Prachayasittikul Somsak Ruchirawat
Virapong Prachayasittikul
Received: 16 July 2012 / Accepted: 6 December 2012 / Published online: 19 December 2012
Ó Springer Science+Business Media New York 2012
Abstract Simplified 1,3-disubstituted urea derivatives
(11–24) of phenylethylamines, homoveratylamines, 2-pyr-
idylethylamines, 2-picolylamines as well as xylylenedi-
amines were synthesized and investigated for their
cytotoxic activities. The results revealed that most analogs
displayed cytotoxicity against HepG2 and MOLT-3 cell
lines. The bis-thiourea derivatives 23 and 24 exhibited
higher inhibitory potency against HepG2 cell than the
reference drug, etoposide. 1,10-(1,3-phenylenebis(methy-
lene))bis(3-(4-chlorophenyl)thiourea) 24 was shown to be
the most potent cytotoxic compound against MOLT-3 cell
line with an IC50 value of 1.62 lM. QSAR studies sug-
gested that compounds with high ionization potential dis-
played high cytotoxicity against HuCCA-1 cell line.
Furthermore, derivatives with dimethoxyphenyl group had
high radial distribution function with a correspondingly
high cytotoxicity against A549 cell line. Moreover, analogs
23 and 24 had low values of EHOMO (energy of the highest
occupied molecular orbital energy) as well as high cyto-
toxicity against HepG2 cell line. This study affords an
easily accessible approach for the synthesis of promising
anticancer agents. The developed QSAR models provided
pertinent information into the physicochemical properties
governing the investigated biologic properties.
Electronic supplementary material The online version of this
article (doi:10.1007/s00044-012-0402-6) contains supplementary
material, which is available to authorized users.
R. Pingaew (&) ꢀ P. Tongraung
Keywords Anticancer ꢀ Multiple linear regression ꢀ
QSAR ꢀ Urea
Department of Chemistry, Faculty of Science, Srinakharinwirot
University, Bangkok 10110, Thailand
e-mail: ratchanok@swu.ac.th
A. Worachartcheewan ꢀ C. Nantasenamat
Center of Data Mining and Biomedical Informatics,
Faculty of Medical Technology, Mahidol University,
Bangkok 10700, Thailand
Introduction
Cancer continues to be one of the major human diseases
inflicting great suffering and financial loss worldwide.
Accordingly, the design, development, and synthesis of
novel lead compounds with enhanced anticancer potency
and decreased adverse side effects are therefore of great
interest and importance.
A. Worachartcheewan ꢀ C. Nantasenamat ꢀ
V. Prachayasittikul (&)
Department of Clinical Microbiology and Applied Technology,
Faculty of Medical Technology, Mahidol University,
Bangkok 10700, Thailand
e-mail: mtvpr@mahidol.ac.th
Urea derivatives comprising of N-nitrosoureas, diary-
lsulfonylureas, benzoylureas, as well as thioureas have been
well recognized for their anticancer activities (Gnewuch
and Sosnovsky, 1997; Li et al., 2009). Much effort has
therefore been devoted for studying structural variations of
urea analogs in achieving the ultimate goal of high efficacy
of treatment (Esteves-Souza et al., 2006; Fortin et al., 2010;
Gil et al., 1999; Saeed et al., 2010; Sondhi et al., 2010; Toth
S. Prachayasittikul
Center for Innovation Development and Technology Transfer,
Faculty of Medical Technology, Mahidol University,
Bangkok 10700, Thailand
S. Ruchirawat
Chulabhorn Research Institute and Chulabhorn Graduate
Institute, Bangkok 10210, Thailand
123

Med Chem Res (2013) 22:4016–4029
4017
et al., 1997). Benzoylureas such as JIMBO1 (1), BAABU
(2), and IAABU (3) were shown to be potent anticancer
agents owing to their abilities to disrupt the assembly of
mitotic spindle microtubules from free tubulin pool of
tumor cells, thus blocking cell cycle at the M-phase and
finally causing apoptotic cell death by promoting bcl-2
phosphorylation (Jiang et al., 1998a; Jiang et al., 1998b;
Jiang et al., 2002; Li et al., 2003; Song et al., 2008).
1-Aryl-3-(2-chloroethyl)ureas such as 4-tert-butyl-[3-(2-
chloroethyl)ureido]benzene (tBCEU, 4) and 4-iodo-[3-(2-
chloroethyl)ureido]benzene (ICEU, 5) (as shown in Fig. 1)
were reported to be potent antimitotics through their
covalent binding to the colchicine-binding site on intra-
cellular b-tubulin (Fortin et al., 2007).
mentioned analogs of anticancer ureas, urea and thiourea
derivatives bearing 1,3-disubstituents of phenyl, phenylalkyl
and pyridylalkyl groups were designed and synthesized.
Herein, the target ureas were prepared by the reaction of
commercially available amines (6–10) (Scheme 1) and
evaluated for their cytotoxic activities. To elucidate the
substituent influence of urea derivatives toward their cyto-
toxic activities, quantitative structure–activity relationship
(QSAR) modeling of the compounds was performed.
Materials and methods
General
The discovery of novel lead candidates as anticancer
agents is a challenging endeavor. We bear in mind that
simple and small molecular urea derivatives could be of
potential interest as potent target compounds. Of the
Melting points were determined using a Griffin melting point
apparatus and were uncorrected. Column chromatography
was carried out using silica gel 60 (70–230 mesh ASTM).
Analytical thin-layer chromatography (TLC) was performed
with silica gel 60 F254 aluminum sheets. 1H- and 13C-NMR
spectra were recorded on a Bruker AVANCE 300 NMR
spectrometer (operating at 300 MHz for 1H and 75 MHz for
13C). FTIR were obtained using a universal attenuated total
reflectance attached on a Perkin–Elmer Spectrum One
spectrometer. Mass spectra were recorded on a Bruker Dal-
tonics (microTOF). Reagents for cell culture and assay were
as follows: RPMI-1640 (Rosewell Park Memorial Institute
medium, Gibco and Hyclone laboratories, USA); HEPES
(N-2-hydroxyethylpiperazine-N0-2-ethanesulfonic
acid),
L-glutamine, penicillin, streptomycin, sodium pyruvate, and
Fig. 1 Representative urea anticancers
Scheme 1 Synthesis of (thio)urea analogs (11–24) from appropriate amines (6–10) and iso(thio)cyanates
123

4018
Med Chem Res (2013) 22:4016–4029
glucose (Sigma, USA); Ham’s/F12 (Nutrient mixture F-12);
DMEM (Dulbecco’s Modified Eagle’s Medium); FBS (fetal
bovine serum, Hyclone laboratories, USA); gentamicin
sulfate (Government Pharmaceutical Organization, Thai-
land); and MTT (3(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-
tetrazolium bromide, Sigma Aldrich, USA).
under reflux for 2 h to give 13 (461.58 mg, 72 %). mp:
167–168 °C. Rf 0.53 (30 % acetone-hexane). 1H NMR
(300 MHz, DMSO-d6): d (ppm) 2.74 (t, J = 7.6 Hz, 2H,
ArCH2), 3.32 (q, J = 6.6 Hz, 2H, NHCH2), 6.16 (t, J =
5.4 Hz, 1H, NH), 7.15–7.43 (m, 9H, ArH), 8.64 (s, 1H, NH);
13C NMR (75 MHz, DMSO-d6): d (ppm) 36.2, 41.1, 119.5,
124.9, 126.6, 128.8, 128.9, 129.1, 140.0, 155.4. IR (UATR)
cm-1: 3318, 1632, 1590, 1558, 1489, 1300, 1232, 1089. TOF–
MS m/z: 275.0946 (Calcd for C15H16ClN2O: 275.0946).
General procedure for the synthesis of urea
derivatives (11–24)
A mixture of amine (2 mmol, 1 mmol for m-xylylenedi-
amine) and appropriate phenylisocyanate (2 mmol) in
dichloromethane (or acetone, 15 mL) was stirred at room
temperature (or under reflux) for 2–12 h. The solid product
formed was filtered, and the filtrate was further evaporated
under reduced pressure. The residue was purified by col-
umn chromatography and recrystallized from methanol to
give the urea derivatives (11–24).
1-(3,4-dimethoxyphenethyl)-3-(4-nitrophenyl)thiourea (14)
A mixture of homoveratylamine 7 (362.46 mg) and 4-
nitroisothiocyanate (180.18 mg) in dichloromethane was
stirred at room temperature for 3 h to afford 14 (263.84 mg,
73 %). mp: 145–146 °C. Rf 0.30 (40 % acetone-hexane).
1H NMR (300 MHz, DMSO-d6): d (ppm) 2.83 (t,
J = 7.3 Hz, 2H, ArCH2), 3.73, 3.75 (2s, 6H, 2 9 OCH3),
3.74 (br q, 2H, NHCH2), 6.78 (d, J = 8.0 Hz, 1H, C6-ArH),
6.87 (s, 1H, C2-ArH), 6.90 (d, J = 8.0 Hz, 1H, C5-ArH),
7.76 (d, J = 9.1 Hz, 2H, C20-ArH, C60-ArH), 8.15 (d,
J = 9.1 Hz, 2H, C30-ArH, C50-ArH), 8.24 (s, 1H, NH),
10.16 (s, 1H, NH); 13C NMR (75 MHz, DMSO-d6): d (ppm)
34.0, 45.8, 55.9, 56.0, 112.5, 113.0, 120.8, 121.0, 125.0,
1-(4-nitrophenyl)-3-phenethylthiourea (11)
A mixture of phenylethylamine 6 (242.36 mg) and 4-nitrois-
othiocyanate (360.36 mg) in dichloromethane was stirred
under reflux for 2 h to give 11 (421.9 mg, 70 %). mp:
141–142 °C. (145–145.5 °C (Hisaki et al., 2007)). Rf 0.40
(30 % acetone-hexane). 1H NMR (300 MHz, CDCl3): d(ppm)
3.04(t, J = 6.7 Hz, 2H, ArCH2), 3.97 (br q, 2H, NHCH2), 6.30
(s, 1H, NH), 7.09 (d, J = 8.8 Hz, 2H, C20-ArH, C60-ArH),
7.20–7.40 (m, 5H, ArH), 8.13 (d, J = 8.8 Hz, 2H, C30-ArH,
C50-ArH), 8.37 (s, 1H, NH); 13C NMR (75 MHz, CDCl3):
d (ppm) 34.4, 46.3, 122.3, 125.6, 127.1, 128.8, 129.1, 138.1,
142.6, 144.4, 179.9. IR (UATR) cm-1: 3360, 1520, 1327.
TOF–MS m/z: 302.0961 (Calcd for C15H16N3O2S: 302.0958).
131.9, 142.2, 146.8, 147.9, 149.2, 180.4. IR (UATR) cm-1
:
3333, 1596, 1513, 1328, 1301, 1260, 1233, 1026. TOF–MS
m/z: 362.1179 (Calcd for C17H20N3O4S: 362.1169).
1-(3,4-dimethoxyphenethyl)-3-(4-chlorophenyl)thiourea (15)
A mixture of homoveratylamine 7 (362.46 mg) and
4-chloroisothiocyanate (360.36 mg) in dichloromethane
was stirred at room temperature for 3 h to give 15
(491.2 mg, 70 %). mp: 117–118 °C (114 °C (Dwivedi
1
1-(4-chlorophenyl)-3-phenethylthiourea (12)
(Zhang et al., 2000)
et al., 1972)). Rf 0.42 (30 % acetone-hexane). H NMR
(300 MHz, DMSO-d6): d (ppm) 2.80 (t, J = 7.4 Hz, 2H,
ArCH2), 3.69 (br q, 2H, NHCH2), 3.72, 3.74 (2s, 6H,
2 9 OCH3), 6.75 (dd, J = 8.2, 1.5 Hz, 1H, C6-ArH), 6.84
(d, J = 1.5 Hz, 1H, C2-ArH), 6.89 (d, J = 8.2 Hz, 1H, C5-
ArH), 7.33 (d, J = 8.9 Hz, 2H, C20-ArH, C60-ArH), 7.39 (d,
J = 8.9 Hz, 2H, C30-ArH, C50-ArH), 7.76 (s, 1H, NH), 9.62
(s, 1H, NH); 13C NMR (75 MHz, DMSO-d6): d (ppm) 34.3,
45.8, 55.9, 56.0, 112.4, 113.0, 121.0, 124.9, 128.8, 132.1,
138.8, 147.8, 149.2, 180.8. IR (UATR) cm-1: 3350, 3304,
1591, 1508, 1489, 1327, 1260, 1229, 1136, 1014. TOF–MS
m/z: 351.0939 (Calcd for C17H20ClN2O2S: 351.0928).
A mixture of phenylethylamine 6 (242.36 mg) and 4-chlo-
roisothiocyanate (339.26 mg) in dichloromethane was stir-
red under reflux for 2 h to provide 12 (418.76 mg, 72 %).
mp: 120–121 °C. Rf 0.50 (30 % acetone-hexane). 1H NMR
(300 MHz, DMSO-d6): d (ppm) 2.87 (t, J = 7.6 Hz, 2H,
ArCH2), 3.69 (br q, 2H, NHCH2), 7.19–7.42 (m, 9H, ArH),
7.83 (s, 1H, NH), 9.63 (s, 1H, NH); 13C NMR (75 MHz,
DMSO-d6): d (ppm) 34.8, 45.7, 125.0, 126.7, 128.3, 128.9,
129.1, 138.7, 139.7, 180.8. IR (UATR) cm-1: 3375, 1584,
1527, 1486, 1299, 1286, 1245, 1089. TOF–MS m/z:
291.0720 (Calcd for C15H16ClN2S: 291.0717).
1-(3,4-dimethoxyphenethyl)-3-(4-chlorophenyl)urea (16)
1-(4-chlorophenyl)-3-phenethylurea (13)
A mixture of homoveratylamine 7 (362.46 mg) and
4-chloroisocyanate (307.14 mg) in dichloromethane was
stirred under reflux for 2 h to afford 16 (502.2 mg, 75 %).
mp: 147–148 °C. Rf 0.33 (30 % acetone-hexane). 1H NMR
A mixture of phenylethylamine 6 (242.36 mg) and 4-chlo-
roisocyanate (307.14 mg) in dichloromethane was stirred
123

Med Chem Res (2013) 22:4016–4029
4019
(300 MHz, DMSO-d6): d (ppm) 2.67 (t, J = 7.0 Hz, 2H,
ArCH2), 3.30 (q, J = 6.7 Hz, 2H, NHCH2), 3.72, 3.74 (2s,
6H, 2 9 OCH3), 6.10 (s, 1H, NH), 6.73 (dd, J = 8.1,
1.3 Hz, 1H, C6-ArH), 6.81 (d, J = 1.3 Hz, 1H, C2-ArH),
6.87 (d, J = 8.1 Hz, 1H, C5-ArH), 7.25 (d, J = 8.9 Hz,
2H, C20-ArH, C60-ArH), 7.40 (d, J = 8.9 Hz, 2H, C30-
ArH, C50-ArH), 8.63 (s, 1H, NH); 13C NMR (75 MHz,
DMSO-d6): d (ppm) 35.8, 41.2, 55.9, 56.0, 112.4, 113.0,
119.5, 120.9, 124.9, 128.9, 129.1, 132.4, 140.0, 147.7,
149.1, 155.4. IR (UATR) cm-1: 3333, 3292, 1627, 1590,
1559, 1518, 1240, 1031. TOF–MS m/z: 335.1159 (Calcd
for C17H20ClN2O3: 335.1157).
under reflux for 2 h to furnish 19 (430.14 mg, 78 %). mp:
138–139 °C. Rf 0.48 (30 % acetone-hexane). 1H NMR
(300 MHz, DMSO-d6): d (ppm) 2.91 (t, J = 6.8 Hz, 2H,
PyCH2), 3.47 (q, J = 6.8 Hz, 2H, NHCH2), 6.23 (t,
J = 5.6 Hz, 1H, NH), 7.20–7.50 (m, 6H, C3-PyH, C5-PyH,
C20-ArH, C30-ArH, C50-ArH, C60-ArH), 7.71 (dt, J = 7.6,
1.7 Hz, 1H, C4-PyH), 8.51 (d, J = 4.8 Hz, 1H, C6-PyH),
8.66 (s, 1H, NH); 13C NMR (75 MHz, DMSO-d6): d (ppm)
38.2, 39.2, 119.5, 120.3, 121.9, 123.7, 128.9, 129.1, 137.0,
139.0, 140.0, 149.5, 155.4, 159.7. IR (UATR) cm-1: 3299,
1634, 1589, 1558, 1490, 1397, 1299, 1237, 1085. TOF–MS
m/z: 276.0898 (Calcd for C14H15ClN3O: 276.0898).
1-(4-nitrophenyl)-3-(2-(pyridin-2-yl)ethyl)thiourea (17)
1-(4-nitrophenyl)-3-(pyridin-2-ylmethyl)thiourea (20)
A mixture of 2-pyridylethylamine 9 (244.34 mg) and
4-nitroisothiocyanate (360.36 mg) in acetone was stirred
under reflux for 2 h to give 17 (483.76 mg, 80 %). mp:
157–158 °C (159–160 °C (Valdes-Martinez et al., 2000)).
Rf 0.33 (40 % acetone-hexane). 1H NMR (300 MHz,
DMSO-d6): d (ppm) 3.07 (t, J = 6.9 Hz, 2H, PyCH2), 3.89
(br q, 2H, NHCH2), 7.26 (dd, J = 7.4, 5.0 Hz, 1H, C5-
PyH), 7.33 (d, J = 7.6 Hz, 1H, C3-PyH), 7.74 (dt, J = 7.6,
1.7 Hz, 1H, C4-PyH), 7.78 (d, J = 9.1 Hz, 2H, C20-ArH,
C60-ArH), 8.16 (d, J = 9.1 Hz, 2H, C30-ArH, C50-ArH),
8.36 (s, 1H, NH), 8.53 (d, J = 5.0 Hz, 1H, C6-PyH), 10.21
(s, 1H, NH); 13C NMR (75 MHz, DMSO-d6): d (ppm)
36.4, 43.9, 120.9, 122.2, 123.8, 125.0, 137.2, 142.3, 146.8,
149.5, 159.3, 180.4. IR (UATR) cm-1: 3300, 3185, 1594,
1523, 1508, 1340, 1300, 1250, 1183, 1112. TOF–MS m/z:
303.0898 (Calcd for C14H15N4O2S: 303.0910).
A mixture of 2-picolylamine 8 (216.28 mg) and 4-nitrois-
othiocyanate (360.36 mg) in acetone was stirred under
reflux for 2 h to provide 20 (490.14 mg, 85 %). mp:
167–168 °C (159–160 °C (Suksai et al., 2009)). Rf 0.45
(30 % acetone-hexane). 1H NMR (300 MHz, DMSO-d6): d
(ppm) 4.85 (d, J = 2.8 Hz, 2H, CH2), 7.32 (dd, J = 7.0,
5.1 Hz, 1H, C5-PyH), 7.40 (d, J = 7.8 Hz, 1H, C3-PyH),
7.80 (dt, J = 7.8, 1.7 Hz, 1H, C4-PyH), 7.94 (d,
J = 9.2 Hz, 2H, C20-ArH, C60-ArH), 8.20 (d, J = 9.2 Hz,
2H, C30-ArH, C50-ArH), 8.56 (d, J = 4.5 Hz, 1H, C6-PyH),
8.79 (s, 1H, NH), 10.54 (s, 1H, NH); 13C NMR (75 MHz,
DMSO-d6): d (ppm) 49.3, 121.0, 122.1, 122.9, 125.0, 137.3,
142.4, 146.8, 149.3, 157.2, 180.8. IR (UATR) cm-1: 3207,
1644, 1593, 1496, 1328, 1287, 1205, 1108, 858. TOF–MS
m/z: 289.0749 (Calcd for C13H13N4O2S: 289.0754).
1-(4-chlorophenyl)-3-(pyridin-2-ylmethyl)thiourea (21)
(Bourdais and Omar, 1980)
1-(4-chlorophenyl)-3-(2-(pyridin-2-yl)ethyl)thiourea (18)
A mixture of 2-pyridylethylamine 9 (244.34 mg) and
4-chloroisothiocyanate (339.26 mg) in acetone was stirred
under reflux for 2 h to give 18 (408.52 mg, 70 %). mp:
120–121 °C. Rf 0.30 (40 % acetone-hexane). 1H NMR
(300 MHz, DMSO-d6): d (ppm) 3.02 (t, J = 6.9 Hz, 2H,
PyCH2), 3.83 (br q, 2H, NHCH2), 7.21–7.44 (m, 6H, C3-
PyH, C5-PyH, C20-ArH, C30-ArH, C50-ArH, C60-ArH),
7.73 (dt, J = 7.6, 1.6 Hz, 1H, C4-PyH), 7.94 (s, 1H, NH),
8.49 (d, J = 4.9 Hz, 1H, C6-PyH), 9.66 (s, 1H, NH); 13C
NMR (75 MHz, DMSO-d6): d (ppm) 36.7, 43.9, 122.1,
123.8, 125.1, 128.4, 128.9, 137.1, 138.7, 149.4, 159.5,
180.7. IR (UATR) cm-1: 3300, 3256, 1664, 1609, 1595,
1542, 1488, 1312, 1239, 1088, 830. TOF–MS m/z:
292.0674 (Calcd for C14H15ClN3S: 292.0670).
A mixture of 2-picolylamine 8 (216.28 mg) and 4-chlorois-
othiocyanate (339.26 mg) in acetone was stirred under reflux
for 2 h to give 21 (444.44 mg, 80 %). mp: 176–177 °C. Rf
1
0.50 (30 % acetone-hexane). H NMR (300 MHz, DMSO-
d6):d (ppm) 4.81(d, J = 3.3 Hz, 2H, CH2), 7.29(dd, J = 7.3,
5.7 Hz,1H,C5-PyH),7.36(d, J = 7.8 Hz,1H, C3-PyH),7.38
(d, J = 8.8 Hz, 2H, C20-ArH, C60-ArH), 7.54 (d, J = 8.8 Hz,
2H, C30-ArH, C50-ArH), 7.79 (dt, J = 7.8, 1.6 Hz, 1H,
C4-PyH), 8.38 (s, 1H, NH), 8.53 (d, J = 4.3 Hz, 1H, C6-
PyH), 9.97 (s, 1H, NH); 13C NMR (75 MHz, DMSO-d6): d
(ppm) 49.3, 121.9, 122.7, 125.2, 128.4, 128.9, 137.2, 138.8,
149.2, 157.9, 181.3. IR (UATR) cm-1: 3202, 1595, 1522,
1490, 1338, 1305, 1090. TOF–MS m/z: 278.0503 (Calcd for
C13H13ClN3S: 278.0513).
1-(4-chlorophenyl)-3-(2-(pyridin-2-yl)ethyl)urea (19)
1-(4-chlorophenyl)-3-(pyridin-2-ylmethyl)urea (22)
A mixture of 2-pyridylethylamine 9 (244.34 mg) and
4-chloroisocyanate (307.14 mg) in acetone was stirred
A mixture of 2-picolylamine 8 (216.28 mg) and 4-chlo-
roisocyanate (307.14 mg) in acetone was stirred under
123

4020
Med Chem Res (2013) 22:4016–4029
reflux for 2 h to afford 22 (340.22 mg, 65 %). mp:
175–176 °C. Rf 0.33 (30 % acetone-hexane). 1H NMR
(300 MHz, DMSO-d6): d (ppm) 4.41 (d, J = 5.6 Hz, 2H,
CH2), 6.81 (d, J = 5.6 Hz, 2H, NH), 7.23–7.30 (m, 3H,
C5-PyH, C20-ArH, C60-ArH), 7.34 (d, J = 7.7 Hz, 1H, C3-
PyH), 7.44 (d, J = 8.9 Hz, 2H, C30-ArH, C50-ArH), 7.77
(dt, J = 7.7, 1.7 Hz, 1H, C4-PyH), 8.52 (d, J = 4.3 Hz,
1H, C6-PyH), 8.93 (s, 1H, NH); 13C NMR (75 MHz,
DMSO-d6): d (ppm) 45.0, 119.6, 121.6, 122.6, 125.0,
10 % FBS were seeded at 1 9 104 cells (100 lL) per well in
a 96-well plate and incubated in humidified atmosphere,
95 % air, and 5 % CO2 at 37 °C. After 24 h, additional
medium (100 lL) containing the test compound and vehicle
was added to a final concentration of 50 lg/mL, 0.2 %
DMSO and further incubated for 3 days. After that, the cells
were fixed with EtOH–H2O (95:5, v/v), stained with crystal
violet solution, and lysed with a solution of 0.1 N HCl in
MeOH. A549 and HepG2 cells were stained with MTT. The
absorbance was measured at 550 nm, and the number of
surviving cells was determined from the absorbance. IC50
values were deduced as the drug and sample concentrations
at 50 % inhibition of cell growth. Etoposide and/or doxo-
rubicin were used as reference drugs.
128.9, 137.2, 139.9, 149.2, 155.6, 159.2. IR (UATR) cm-1
:
3300, 3256, 1665, 1595, 1542, 1489, 1313, 1239, 1088,
830. TOF–MS m/z: 262.0743 (Calcd for C13H13ClN3O:
262.0742).
1,10-(1,3-phenylenebis(methylene))bis(3-(4-
nitrophenyl)thiourea) (23) (Nishizawa et al., 2003)
Computational analysis
A mixture of xylylenediamine 10 (136.19 mg) and 4-nit-
roisothiocyanate (360.36 mg) in dichloromethane was
stirred at room temperature for 12 h to give 23 (397.25 mg,
80 %). mp: 198–199 °C. Rf 0.38 (40 % acetone-hexane).
1H NMR (300 MHz, DMSO-d6): d (ppm) 4.76 (s, 4H,
CH2), 7.22–7.38 (m, 4H, ArH), 7.82 (d, J = 9.2 Hz, 4H,
C20-ArH, C60-ArH), 8.14 (d, J = 9.2 Hz, 4H, C30-ArH,
C50-ArH), 8.71 (s, 2H, NH), 10.23 (s, 2H, NH); 13C NMR
(75 MHz, DMSO-d6): d (ppm) 47.1, 120.6, 124.4, 126.3,
Dataset
A dataset of fourteen (thio)urea derivatives (11–24,
Scheme 1) and their cytotoxic activities (Table 1) were
used for QSAR studies. The IC50 (lM) values of cytotox-
icity against cancer cell lines (HuCCA-1, A549, HepG2,
and MOLT-3 cell lines) were converted to pIC50 by taking
the negative logarithm to the base of 10 of the IC50 values.
Although the size of the dataset comprising 14 compounds
may be rather small, several recent QSAR investigations
demonstrated the feasibility of such studies on small
datasets with sample size of n = 12, 13, 19, and 23
(Podunavac-Kuzmanovic and Cvetkovic, 2011, Vahdani
and Bayat, 2011, Saghaie et al., 2012; Sawant et al., 2012).
126.5, 128.4, 138.4, 141.9, 146.2, 180.4. IR (UATR) cm-1
:
3341, 1595, 1533, 1507, 1327, 1305, 1246, 1110, 974, 854.
TOF–MS m/z: 497.1054 (Calcd for C22H21N6O4S2:
497.1060).
1,10-(1,3-phenylenebis(methylene))bis(3-(4-
chlorophenyl)thiourea) (24)
Quantum chemical and molecular descriptors calculation
A mixture of xylylenediamine 10 (136.19 mg) and
4-chloroisothiocyanate (339.26 mg) in dichloromethane
was stirred at room temperature for 12 h to provide 24
(385.12 mg, 81 %). mp: 149–150 °C. Rf 0.50 (40 % ace-
Molecular structures of compounds were constructed by
means of GaussView (Dennington et al., 2003). An initial
geometrical optimization of the structures was performed
at the semi-empirical AM1 (Austin Model 1) level fol-
lowed by a subsequent optimization at the density func-
tional level employing the Becke’s three-parameter hybrid
method with the Lee–Yang–Parr correlation functional
(B3LYP) together with the 6-31g(d) basis set of the
Gaussian 03W package (Frisch et al., 2004). Thirteen
quantum chemical descriptors (Parr et al., 1978, 1999; Parr
and Pearson, 1983; Thanikaivelan et al., 2000) as obtained
from low-energy conformers were comprised of the total
energy of the molecule (Etotal), energy of the highest
occupied molecular orbital (EHOMO), energy of the lowest
unoccupied molecular orbital (ELUMO), and total dipole
moment (l) of the molecule. Additional quantum chemical
descriptors were obtained by performing mathematical
operations on the EHOMO and ELUMO as described by the
following set of equations.
1
tone-hexane). H NMR (300 MHz, DMSO-d6): d (ppm)
4.72 (s, 4H, CH2), 7.19–7.37 (m, 4H, ArH), 7.34 (d,
J = 8.8 Hz, 4H, C20-ArH, C60-ArH), 7.46 (d, J = 8.8 Hz,
4H, C30-ArH, C50-ArH), 8.25 (s, 2H, NH), 9.70 (s, 2H,
NH); 13C NMR (75 MHz, DMSO-d6): d (ppm) 47.1, 124.8,
126.0, 126.3, 128.0, 128.2, 128.4, 138.2, 138.9, 180.8. IR
(UATR) cm-1: 3262, 3200, 1593, 1533, 1489, 1310,
1090, 972, 828. TOF–MS m/z: 475.0570 (Calcd for
C22H21Cl2N4S2: 475.0579).
Cytotoxic assay
The cytotoxic assay was performed as previously described
by Tengchaisri and co-workers (Tengchaisri et al., 1998).
Briefly, cell lines suspended in RPMI 1640 containing
123

Med Chem Res (2013) 22:4016–4029
4021
Table 1 Cytotoxic activity of compounds 11–24
Compound
Cytotoxic activity (IC50, lM)a
HuCCA-1
A549
HepG2
MOLT-3
11
19.08 0.35
84.25 0.71
[181.98b
80.30 1.70
75.65 1.91
[181.98b
25.45 0.58
44.70 1.73
37.60 0.58
42.42 2.89
60.79 4.04
52.78 0.58
152.14 1.41
[171.35b
12.54 0.22
25.14 0.55
[181.98b
12
13
14
[138.34b
65.02 2.12
129.11 6.65
[149.34b
21.72 0.91
38.85 0.95
50.69 1.93
78.72 5.89
119.53 2.15
66.91 0.53
54.14 1.25
118.77 1.59
186.08 2.59
3.36 0.09
1.62 0.01
0.041 0.004
ND
15
139.23 0.21
[149.34b
16
17
[165.37b
[165.37b
18
[171.35b
[171.35b
19
[181.34b
[181.34b
22.70 5.00
34.68 3.46
[180.00b
20
[173.42b
[173.42b
21
[180.00b
[180.00b
22
[191.05b
[191.05b
[191.05b
23
34.24 9.90
52.58 4.24
ND
88.61 8.48
[105.16b
ND
7.51 0.75
7.63 0.96
27.18 5.29
0.42 0.03
24
Etoposidec
Doxorubicind
0.64 0.21
0.70 0.03
Cancer cell lines are HuCCA-1: Human cholangiocarcinoma cancer cells, A549: Human lung carcinoma cell line, HepG2: Human hepatocellular
carcinoma cell line and MOLT-3: Human lymphoblastic leukemia cell line
ND not determined
a
The assays were performed in triplicate
b
IC50 value with [lM derived from IC50 value [50 lg/mL denotes inactive compound
c,d
Etoposide and/or doxorubicin were used as the reference drugs
In addition, the mean absolute atomic charge (Qm)
was calculated from the Mulliken population analysis
as described the following equation (Karelson et al.,
1996):
Electron affinity (EA):
EA ¼ ꢁELUMO
Ionization potential (IP):
ð1Þ
ð2Þ
ð3Þ
ð4Þ
ð5Þ
IP ¼ ꢁEHOMO
The energy difference of HOMO and LUMO:
N
X
Qm ¼
jQaj=N
ð9Þ
a¼1
HOMO ꢁ LUMOgap ¼ EHOMO ꢁ ELUMO
where |Qa| and N are the absolute value of the charges on
all atoms and the total number of atoms present in the
compound, respectively.
Mulliken electronegativity (v):
v ¼ ðEHOMO þ ELUMOÞ=2
Hardness (g):
Molecular descriptors from Dragon span 22 categories
of descriptors as follows: Constitutional descriptors,
Topological descriptors, Walk and path counts, Connec-
tivity indices, Information indices, 2D autocorrelation,
Edge adjacency indices, Burden eigenvalues, Topological
charge indices, Eigenvalue-based indices, Randic molec-
ular profiles, Geometrical descriptors, RDF descriptors,
3D-MoRSE descriptors, WHIM descriptors, GETAWAY
descriptors, Functional group counts, Atom-centered
fragments, Charge descriptors, Molecular properties, 2D
binary fingerprints, and 2D frequency fingerprints.
g ¼ ðELUMO ꢁ EHOMOÞ=2
Electrophilicity (x):
2
x ¼ ðEHOMO þ ELUMO=2Þ =2g
Softness (S):
ð6Þ
ð7Þ
ð8Þ
S ¼ 1=2g
Electrophilicity index (xi):
xi ¼ l2=2g
123

4022
Med Chem Res (2013) 22:4016–4029
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P
n
Multiple linear regression (MLR)
i¼1 ðpi ꢁ aiÞ
RMS ¼
ð12Þ
n
MLR models were calculated by the following equation:
X
where pi is the predicted output, ai is the actual output, and
n is the number of compounds in the data set.
Y ¼ B0 þ
BnXn
ð10Þ
where Y is the pIC50 values of the compounds, B0 is the
intercept, and Bn are the regression coefficients ofdescriptors
Xn. QSAR models were generated using the MLR method as
performed using Waikato Environment for Knowledge
Analysis (Weka), version 3.4.5 (Witten et al., 2011).
Results and discussion
Chemistry
1,3-Disubstituted urea derivatives (11–22) were readily syn-
thesized by treatment with equimolar amounts of amines
(6–9) and iso(thio)cyanates including 4-nitrophenylisothio-
cyanate, 4-chlorophenylisothiocyanate or 4-chlorophenyl-
isocyanate in appropriate solvents (dichloromethane or
acetone) (Scheme 1). Whereas, two equivalents of isothio-
cyanate were employed for xylylenediamine (10) to achieve
bis-thioureas (23 and 24). Structures of all the obtained
compounds were characterized by 1H NMR, 13C NMR, IR,
and HRMS. Infrared spectra of all the synthesized compounds
had strong N–H absorption (3202–3375 cm-1), and the urea
carbonyl showed absorption stretching band values in the
Data sampling
The dataset was separated into two sets: training set and
testing set. Leave-one-out cross-validation (LOO-CV) is a
popular data sampling method for QSAR studies, while
model validation was performed using correlation coeffi-
cient and root mean square error. The LOO-CV method
leaves out one compound as the testing set, whereas the
remaining N-1 compounds were used as the training set.
This process was repeated iteratively until all samples were
used as the testing set.
range of 1627–1665 cm-1 13C NMR spectral analysis of
.
ureas and thioureas showed typical signals for carbonyl and
thiocarbonyl carbons in the range of d 155–156 ppm and d
179–182 ppm, respectively.
Statistical analysis
Calculated descriptors from Dragon that are constant or
near-constant were firstly removed by the Dragon software.
The remaining descriptors were standardized according to
the following equation:
Cytotoxic activity
In assessing the cytotoxicity (Tengchaisri et al., 1998),
synthesized (thio)urea analogs 11–24 were evaluated
against HuCCA-1, A549, HepG2, and MOLT-3 human
cancer cell lines as summarized in Table 1. All the tested
compounds displayed inhibition activities against MOLT-3
cell line, except for 1-(4-chlorophenyl)-3-phenethylurea
(13). However, analog 13 selectively showed potent
activity against the HepG2 cell with an IC50 value of
37.60 lM. Significantly, all urea and thiourea derivatives
of phenylethylamines (11–13), homoveratylamines
(14–16), pyridylalkylamines (17, 19 and 20), as well as
xylylenediamines (23 and 24) displayed inhibitory prop-
erties against HepG2. Particularly, the bis-thiourea analogs
(23 and 24), pyridylurea (19) and phenylurea (11), showed
higher cytotoxic potency than the control drug, etoposide.
The analogs 23 and 24 have comparable IC50 values of
7.51 and 7.63 lM, respectively, and are the most potent
compounds. Apparently, all the pyridine analogs (17–22)
were found to be inactive toward HuCCA-1 and A549
cells. Thioureas bearing NO2 substituent exhibited higher
cytotoxicities than analogs containing Cl group as seen for
11 versus 12, 17 versus 18, 20 versus 21, and 23 versus 24.
Such electron-withdrawing effect was observed for
xij ꢁ ꢀxj
stn
ij
x
¼
ð11Þ
P
N
2
i¼1 ðxij ꢁ ꢀxjÞ =N
where xsijtn is the standardized value, xij is the value of each
sample, ꢀxj is the mean of each descriptors, and N is the
sample size of the dataset.
Standardized descriptors were subjected to further var-
iable selection by means of the Unsupervised Forward
Selection (UFS) algorithm of UFS, version 1.8 for
removing multi-collinear and redundant descriptors. The
UFS algorithm was previously described by Whitley et al.
(2000) and previously used by Nantasenamat et al. (2005).
In addition, important descriptors as correlated with cyto-
toxic activity of (thio)urea derivatives were further selected
by a stepwise multiple linear regression which as per-
formed by means of SPSS Statistics 18.0 (SPSS Inc. USA).
The QSAR model was evaluated by means of the two
statistical parameters comprised of correlation coefficient
(r), which corresponded to the degree of correlation between
the predicted and experimental values as a relative measure
of predictive performance and root mean square error (RMS)
as measuring the predictive error of the model. The RMS was
calculated according to the following equation:
123

Med Chem Res (2013) 22:4016–4029
4023
cytotoxic sulfonamides of anthranilic acid (Doungsoong-
nuen et al., 2011). Cytotoxic activity as compared between
urea and thiourea moieties depending on the type of cancer
cell lines. The thio compounds showed superior potency
toward HuCCA-1, A549, and MOLT-3 cells (12 versus 13,
15 versus 16 as well as 21 versus 22), whereas urea analogs
displayed higher activity against HepG2 cell (12 versus 13,
15 versus 16, and 18 versus 19). It was noted that the
inhibitory activities against HuCCA-1, HepG2, and MOLT-
3 cells were reduced when the phenyl ring of thiourea (11
and 12) was replaced by the polar dimethoxyphenyl ring (14
and 15) and also by the pyridine ring (17 and 18). It is
plausible to conclude that the hydrophobic group is
responsible for the observed cytotoxic activities of analogs.
All investigated cells were inhibited by thiourea analogs 11,
12, 15, and 23. Interestingly, 1,10-(1,3-phenylenebis(meth-
ylene))bis(3-(4-chlorophenyl)thiourea) (24) was shown to
be the most potent analog against MOLT-3 cells.
obtained in three steps: (i) Descriptors were calculated by
means of Dragon to obtain 3,224 variables where constant or
near-constant descriptors were removed, (ii) the remaining
1,447 descriptors were standardized according to Eq. (11)
and were subjected to further removal of multi-collinear and
redundant descriptors the pair of variables of which dis-
played a squared multiple correlation coefficient (R2) greater
than 0.99 as implemented in the UFS software which resulted
in 13 molecular descriptors comprising of B02[C-O],
B06[N-S], B09[C-N], F05[C-N], G2p, GVWAI-50,
More27e, PJI2, RDF030e, RDF140m, RDF145m,
RDF130v, and T[O..O], and (iii) the 13 molecular descrip-
tors in combination with 11 quantum chemical descriptors
generated from Gaussian 03W software were subjected for
further selection of significant descriptors using stepwise
regression of SPSS (version 18.0). The selected important
descriptors and its definition as obtained from the stepwise
regression are shown in Table 2 and were subsequently used
for generating QSAR models for predicting the cytotoxic
activity against cancer cell lines (HuCCA-1, A549, HepG2,
and MOLT-3 cell lines) by Eq. (10). Models 1–4 for pre-
dicting cytotoxicity against the four cancer cell lines and
model evaluations are presented in Table 3.
QSAR modeling
A dataset of fourteen (thio)urea derivatives (Scheme 1) and
their cytotoxicity (Table 1) against cancer cell lines were
used in QSAR studies to elucidate the structure–activity
relationship between the molecular structure and cytotoxic
activity. Geometrical optimization of compounds were
initially performed by means of semi-empirical AM1 fol-
lowed by B3LYP/6-31g(d) to afford a set of quantum
chemical descriptors. In our previous QSAR studies, we
have successfully used descriptors generated from Dragon
and Gaussian 03W softwares for constructing QSAR
models (Nantasenamat et al., 2007a, b, 2008a, b; Prac-
hayasittikul et al., 2010; Suksrichavalit et al., 2009; Suv-
annang et al., 2011; Worachartcheewan et al., 2009, 2011,
2012). Furthermore, the procedures and applications of
QSAR model in life science from small molecules to
macromolecules have been reviewed by Nantasenamat
et al. (2009, 2010). Selected significant descriptors were
Results of QSAR model (model 1) for cytotoxic activity
against HuCCA-1 cell line showed that the r and RMS
values obtained from the training set were 0.9230 and
0.1154, respectively, and the r and RMS values of the
LOO-CV testing set were 0.8067 and 0.1844, respectively.
IP or the ionization potential is an important descriptor for
predicting the cytotoxicity of compounds. The definition
and values of the IP descriptor are shown in Tables 2 and 4,
respectively. We found that nitrothio compounds 11 and
23, which had high IP values (0.2216 and 0.2226, respec-
tively) also displayed higher cytotoxic activity as compared
to chlorothio compounds 12, 15, and 24 (having lower IP
values) as displayed in Table 4. The structure of com-
pounds 11 and 23 harbored NO2 as an electron-withdraw-
ing group at the para position of phenyl thiourea moiety.
This inductive effect facilitates N–H (urea) ionization
Table 2 Definition of descriptors used in QSAR model of cytotoxic activity of (thio)urea derivatives 11–24
Model Cytotoxic activity Descriptors Definition
Type
1
2
HuCCA-1 pIC50 IP
Ionization potential
Quantum chemistry
A549 pIC50
B02[C-O] Presence/absence of C-O at topological distance 02
2D binary fingerprints
RDF130v Radial distribution function-13.0/weight by atomic van der Waals volumes RDF description
g
Hardness
Quantum chemistry
3
4
HepG2 pIC50
MOLT-3 pIC50
B06[N-S] Presence/absence of N-S at topological distance 06
F05[C-N] Frequency of C-N at topological distance 05
2D binary fingerprints
2D frequency fingerprints
Quantum chemistry
EHOMO
Highest occupied molecular orbital energy
B06[N-S] Presence/absence of N-S at topological distance 06
F05[C-N] Frequency of C-N at topological distance 05
2D binary fingerprints
2D frequency fingerprints
123

4024
Med Chem Res (2013) 22:4016–4029
giving rise to compounds with high values of IP. Thus, the
reactive species formed plays a crucial role in cytotoxicity.
Plots of experimental versus predicted cytotoxic activities
of compounds against HuCCA-1 cell lines are shown in
Fig. 2a, while their predicted values are shown in Table 5.
A set of three descriptors comprising of B02[C–O],
RDF130v, and g were important variables that were used in
the construction of QSAR models (model 2) for predicting
the cytotoxicity against A549 cell lines as displayed in
Tables 2 and 3. It was found that the r and RMS values for
the training set were 0.9999 and 0.0013, respectively, while
the r and RMS values for the LOO-CV testing set were
0.5056 and 0.1342, respectively. We observed that the
active thio compounds 11, 12, 14, 15, and 23, except for 12,
displayed the presence of C–O at topological distance 02
and exhibited high value of RDF130v (radial distribution
function-13.0/weight by atomic van der Waals volumes) as
shown in Table 4. Particularly, dimethoxyphenyl derivative
of nitrothio (14) afforded the highest cytotoxicity. Plots of
the experimental versus predicted cytotoxic activities of
compounds against A549 cell lines are shown in Fig. 2b,
while their predicted values are presented in Table 5.
QSAR model 3 for predicting the cytotoxicity of com-
pounds against HepG2 cell lines was developed using
B06[N-S], F05[C-N], and EHOMO (Table 2). The resulting
QSAR equation and its corresponding predictive perfor-
mance is shown in Table 3, while the descriptor values of
compounds in the data set are shown in Table 4. The
performance of the QSAR model (Table 3) gave r and
RMS values for the training set of 0.9866 and 0.0590,
respectively, while the r and RMS values for the LOO-CV
testing set were 0.7132 and 0.2673, respectively. The
results showed that analogs with high cytotoxicity had low
E
HOMO, but high F05[C-N] (frequency of C-N topological
distance 05). The bis-thioureas 23 and 24 had the lowest
EHOMO values of -0.2226 and -0.2082, respectively, but
the highest F05[C-N] value of 12. It is notable that active
ureas against HepG2 cell line, all contain the electron-
withdrawing groups (NO2 and Cl). This might involve the
ease of N-H bond ionization and ultimately forming reac-
tive species, which may interact with cells and give rise to
cytotoxicity. Apparently, the bis-thio analogs 23 and 24
exerted the most potent cytotoxic activity. Plots of exper-
imental versus predicted activities of compounds as anti-
cancer agents against HepG2 cell lines are shown in
Fig. 2c, while their predicted values are given in Table 5.
QSAR model 4 for predicting the cytotoxicity against
MOLT-3 cell lines is shown in Table 3. It was found that
B06[N-S] and F05[C-N] were important descriptors for
generating the QSAR model (Tables 2, 3) and their values
Table 3 QSAR models of cytotoxic activities of (thio)urea derivatives 11–24
Model Equation
N
rTr
RMSTr rCV
RMSCV
1
2
3
4
HuCCA-1 pIC50 = 28.7806 (IP) - 7.8116
5
0.9230 0.1154 0.8067 0.1844
A549 pIC50 = -0.2975 (B02[C-O]) - 0.0457 (RDF130v) - 12.5278 (g) - 0.817
5 0.9999 0.0013 0.5056 0.1342
HepG2 pIC50 = -0.8233 (B06[N-S]) ? 0.0895 (F05[C-N]) - 11.6506 (EHOMO) - 4.4617 11 0.9866 0.0590 0.7132 0.2673
MOLT-3 pIC50 = -0.5365 (B06[N-S]) ? 0.1834 (F05[C-N]) - 2.5504 13 0.8898 0.2664 0.8370 0.3219
Table 4 Significant descriptors for predicting cytotoxic activities of (thio)urea derivatives 11–24
Compound
B02[C-O]
B06[N-S]
F05[C-N]
RDF130v
EHOMO
IP
g
11
12
13
14
15
16
17
18
19
20
21
22
23
24
1
0
0
1
1
1
1
0
0
1
0
0
1
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
5
5
0.096
0
-0.2216
-0.2089
-0.2151
-0.2013
-0.1967
-0.1972
-0.2202
-0.2075
-0.2141
-0.2146
-0.2017
-0.2077
-0.2226
-0.2082
0.2216
0.2089
0.2151
0.2013
0.1967
0.1972
0.2202
0.2075
0.2141
0.2146
0.2017
0.2077
0.2226
0.2082
0.0627
0.0848
0.1027
0.0530
0.0792
0.0939
0.0681
0.0848
0.0940
0.0677
0.0804
0.0845
0.0650
0.0845
5
0.514
0.718
0.069
0.639
0.492
0.032
0.622
0.253
0
5
5
5
6
6
6
4
4
4
1.103
0.46
12
12
0.299
123

Med Chem Res (2013) 22:4016–4029
4025
(a) -1.2
(b)-1.7
-1.8
-1.9
-2.0
-2.1
-2.2
-1.4
-1.6
-1.8
-2.0
-2.2
-2.2
-2.0
-1.8
-1.6
-1.4
-1.2
-2.15 -2.10 -2.05 -2.00 -1.95 -1.90 -1.85 -1.80 -1.75
Experimental activity (HuCCA-1 pIC
50
)
Experimental activity (A549 pIC )
50
(c)
(d) 0.0
-.6
-.8
-1.0
-1.2
-1.4
-1.6
-1.8
-2.0
-2.2
-2.4
-.5
-1.0
-1.5
-2.0
-2.5
-2.4
-2.2
-2.0
-1.8
-1.6
-1.4
-1.2
-1.0
-.8
-2.5
-2.0
-1.5
-1.0
-.5
0.0
Experimental activity (HepG2 pIC
50
)
Experimental activity (MOLT-3 pIC )
50
Fig. 2 Plot of the experimental and predicted anticancer activities
(pIC50) against HuCCA-1 (a), A549 (b), HepG2 (c), and MOLT-3
(d) cell lines for the training set (filled square; regression line is
represented as dotted line) and the leave-one-out cross-validated
testing set (open square; regression line is represented as solid line)
are summarized in Table 4. In the model, r and RMS
values obtained from the training set were 0.8898 and
0.2664, respectively, and r and RMS values of the LOO-
CV testing set were 0.8370 and 0.3219, respectively.
Similar results were observed as in the case of QSAR
model 3 for HepG2 cells. The most active bis-thio analogs
had the highest F05[C-N] value of 12. Most of the active
analogs showed the absence of N-S at topological distance
06. Plots of the experimental versus predicted activities of
compounds as anticancer agents against MOLT-3 cell lines
are shown in Fig. 2d while their predicted values are shown
in Table 5.
2-picolylamines, and xylylenediamines against HuCCA-1,
A549, HepG2, and MOLT-3 cell lines. It should be noted
that bis-thiourea analogs exhibited a notably higher activity
than monourea against HepG2 and MOLT-3 cell lines.
Both bis-thioureas 23 and 24 displayed significant and
promising cytotoxic activity against HepG2 cells having
IC50 values much lower than the reference drug. Most urea
analogs displayed cytotoxicity against MOLT-3 cell
lines, particularly 1,10-(1,3-phenylenebis(methylene))bis(3-
(4-chlorophenyl)thiourea) (24) was the most potent com-
pound affording an IC50 value of 1.62 lM. To achieve
higher cytotoxicity of the 1,3-disubstituted ureas, it was
found that one of the substituent should be a phenylalkyl
group rather than a pyridylalkyl group while the other
substituent should be a phenyl ring with a 4-nitro electron-
withdrawing group. All these experimental results were
well correlated with the results from QSAR studies. The
results suggested that compounds with high IP and low
EHOMO values correspondingly led to high cytotoxicity as
Conclusion
This study investigated the cytotoxic properties of a series
of 1,3-disubstituted (thio)ureas (11–24) derived from phe-
nylethylamines, homoveratylamines, 2-pyridylethylamines,
123

4026
Med Chem Res (2013) 22:4016–4029
Table 5 The experimental and predicted activities (pIC50) of cytotoxic activity of (thio)urea derivatives 11–24
123

Med Chem Res (2013) 22:4016–4029
4027
Table 5 continued
HuCCA-1 human cholangiocarcinoma cancer cells, A549 human lung carcinoma cell line, HepG2 human hepatocellular carcinoma cell line and
MOLT-3 human lymphoblastic leukemia cell line, Exp. experimental activity, Pred. predicted activity
References
noted for HuCCA-1 and HepG-2 cell lines, respectively.
Analogs having high RDF value also displayed high cyto-
toxicity such as those of dimethoxyphenyl analogs, which
exhibited high cytotoxicity against A549 cells. This study
demonstrated an accessible approach for the synthesis of
promising bioactive analogs and the development of QSAR
models, which helps pave the way toward achieving novel
molecules with promising anticancer activity by providing an
understanding of their physicochemical properties.
Bourdais J, Omar A-MME (1980) Polycyclic azines. III. Synthesis of
3-aminoimidazo[1,5-a]pyridine derivatives by cyclodesulfuriza-
tion of N0-substituted-N-(2-pyridylmethyl)thioureas with dicy-
clohexylcarbodiimide. J Heterocyclic Chem 17:555–558
Dennington II R, Keith T, Millam J, Eppinnett K, Hovell WL,
Gilliland R (2003) GaussView, Version 3.09, Semichem, Inc.,
Shawnee Mission, KS, USA
Doungsoongnuen S, Worachartcheewan A, Pingaew R, Suksrichavalit
T, Prachayasittikul S, Ruchirawat S, Prachayasittikul V (2011)
Investigation on biological activities of anthranilic acid sulfon-
amide analogs. EXCLI J 10:155–161
Acknowledgments We gratefully acknowledge the research grant
supported by Srinakharinwirot University (B.E. 2556). This project
is supported by Office of the Higher Education Commission and
Mahidol University under the National Research Universities
Initiative.
Dwivedi C, Gupta TK, Parmar SS (1972) Substituted thiazolidones as
anticonvulsants. J Med Chem 15:553–554
Esteves-Souza A, Pissinate K, Nascimento MG, Grynberga NF, Eche-
varria A (2006) Synthesis, cytotoxicity, and DNA-topoisomerase
123

4028
Med Chem Res (2013) 22:4016–4029
inhibitory activity of new asymmetric ureas and thioureas. Bioorg
Med Chem 14:492–499
Nantasenamat C, Isarankura-Na-Ayudhya C, Naenna T, Prac-
hayasittikul V (2007a) Quantitative structure-imprinting factor
relationship of molecularly imprinted polymers. Biosens Bio-
electron 22:3309–3317
Nantasenamat C, Isarankura-Na-Ayudhya C, Tansila N, Naenna T,
Prachayasittikul V (2007b) Prediction of GFP spectral properties
using artificial neural network. J Comput Chem 28:1275–1289
Nantasenamat C, Isarankura-Na-Ayudhya C, Naenna T, Prac-
hayasittikul V (2008a) Prediction of bond dissociation enthalpy
of antioxidant phenols by support vector machine. J Mol Graph
Model 27:188–196
Nantasenamat C, Piacham T, Tantimongcolwat T, Naenna T,
Isarankura-Na-Ayudhya C, Prachayasittikul V (2008b) QSAR
model of the quorum-quenching N-acyl-homoserine lactone
lactonase activity. J Biol Syst 16:279–293
Nantasenamat C, Isarankura-Na-Ayudhya C, Naenna T, Prac-
hayasittikul V (2009) A practical overview of quantitative
structure–activity relationship. EXCLI J 8:74–88
Nantasenamat C, Isarankura-Na-Ayudhya C, Prachayasittikul V
(2010) Advances in computational methods to predict the
biological activity of compounds. Exp Opin Drug Discov
5:633–654
Nishizawa S, Yokobori T, Kato R, Yoshimoto K, Kamaishi T,
Teramae N (2003) Hydrogen-bond forming ionophore for highly
efficient transport of phosphate anions across the nitrobenzene–
water interface. Analyst 128:663–669
Parr RG, Pearson RG (1983) Absolute hardness: companion parameter
to absolute electronegativity. J Am Chem Soc 105:7512–7516
Parr RG, Donnelly RA, Levy M, Palke WE (1978) Electronegativity:
the density functional viewpoint. J Chem Phys 68:3801–3807
Parr RG, Szentpaly Lv, Liu S (1999) Electrophilicity index. J Am
Chem Soc 121:1922–1924
Podunavac-Kuzmanovic SO, Cvetkovic DD (2011) QSAR modeling
of antibacterial activity of some benzimidazole derivatives.
Chem Ind Chem Eng Q 17:33–38
Fortin JS, Lacroix J, Desjardins M, Patenaude A, Petitclerc E,
C-Gaudreault R (2007) Alkylation potency and protein specific-
ity of aromatic urea derivatives and bioisosteres as potential
irreversible antagonists of the colchicine-binding site. Bioorg
Med Chem 15:4456–4469
ˆ ´
Fortin S, Moreau E, Lacroix J, Cote MF, Petitclerc E, C-Gaudreault R
(2010) Synthesis, antiproliferative activity evaluation and struc-
ture-activity relationships of novel aromatic urea and amide
analogues of N-phenyl-N0-(2-chloroethyl)ureas. Eur
Chem 45:2928–2937
J Med
Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA,
Cheeseman JR, Montgomery Jr JA, Vreven T, Kudin KN, Burant
JC, Millam JM, Iyengar SS, Tomasi J, Barone V, Mennucci B,
Cossi M, Scalmani G, Rega N, Petersson GA, Nakatsuji H, Hada
M, Ehara M, Toyota K, Fukuda R, Hasegawa J, Ishida M,
Nakajima T, Honda Y, Kitao O, Nakai H, Klene M, Li X, Knox
JE, Hratchian HP, Cross JB, Bakken V, Adamo C, Jaramillo
J,Gomperts R, Stratmann RE, Yazyev O, Austin AJ, Cammi R,
Pomelli C, Ochterski JW, Ayala PY, Morokuma K, Voth GA,
Salvador P, Dannenberg JJ, Zakrzewski VG, Dapprich S,
Daniels AD, Strain MC, Farkas O, Malick DK, Rabuck AD,
Raghavachari K, Foresman JB, Ortiz JV, Cui Q, Baboul AG,
Clifford S, Cioslowski J, Stefanov BB, Liu G, Liashenko A,
Piskorz P, Komaromi I, Martin RL, Fox DJ, Keith T, Al-Laham
MA, Peng CY, Nanayakkara A, Challacombe M, Gill PMW,
Johnson B, Chen W, Wong MW, Gonzalez C, Pople JA (2004)
Gaussian 03, Revision C.02, Gaussian, Inc., Wallingford CT
´
´
´
Gil MJ, Man˜u MA, Arteaga C, Migliaccio M, Encıo I, Gonzalez A,
Martınez-Merino V (1999) Synthesis and cytotoxic activity of
´
N-(2-pyridylsulfenyl)urea derivatives. A new class of potential
antineoplastic agents. Bioorg Med Chem Lett 9:2321–2324
Gnewuch CT, Sosnovsky G (1997) A critical appraisal of the
evolution of N-nitrosoureas as anticancer drugs. Chem Rev
97:829–1013
Hisaki I, Sasaki S-I, Hirose K, Tobe Y (2007) Synthesis and anion-
selective complexation of homobenzylic tripodal thiourea deriv-
atives. Eur J Org Chem 2007:607–615
Jiang J-D, Davis AS, Middleton K, Ling Y-H, Perez-Soler R, Holland
JF, Bekesi JG (1998a) 3-(Iodoacetamido)-benzoylurea: a novel
cancericidal tubulin ligand that inhibits microtubule polymeri-
zation, phosphorylates bcl-2, and induces apoptosis in tumor
cells. Cancer Res 58:5389–5395
Jiang JD, Roboz J, Weisz I, Deng L, Ma L, Holland JF, Bekesi JG
(1998b) Synthesis, cancericidal and antimicrotubule activities of
3-(haloacetamido)-benzoylureas. Anti-Cancer Drug Des 13:
735–747
Prachayasittikul S, Wongsawatkul O, Worachartcheewan A, Nanta-
senamat C, Ruchirawat S, Prachayasittikul V (2010) Elucidating
the structure–activity relationships of the vasorelaxation and
antioxidation properties of thionicotinic acid derivatives. Mol-
ecules 15:198–214
Saeed S, Rashid N, Jones PG, Ali M, Hussain R (2010) Synthesis,
characterization and biological evaluation of some thiourea
derivatives bearing benzothiazole moiety as potential antimicro-
bial and anticancer agents. Eur J Med Chem 45:1323–1331
Saghaie L, Sakhi H, Hassan Sabzyan H, Shahlaei M, Shamshirian D
(2012) Stepwise MLR and PCR QSAR study of the pharma-
ceutical activities of antimalarial 3-hydroxypyridinone agents
using B3LYP/6-311??G** descriptors. Med Chem Res. doi:
10.1007/s00044-012-0152-5
Jiang JD, Denner L, Ling YH, Li JN, Davis A, Wang Y, Li Y, Roboz
JL, Wang LG, Roman PS, Marcelli M, Bekesi G, Holland JF
(2002) Double blockade of cell cycle at G1-S transition and M
phase by 3-Iodoacetamido benzoyl ethyl ester, a new type of
tubulin ligand. Cancer Res 62:6080–6088
Sawant RL, Bansode CA, Wadekar JB (2012) In vitro anti-inflam-
matory potential and QSAR analysis of oxazolo/thiazolo pyrim-
idine derivatives. Med Chem Res. doi:10.1007/s00044-012-01
89-5
Karelson M, Lobanov VS, Katritzky AR (1996) Quantum-chemical
descriptors in QSAR/QSPR studies. Chem Rev 96:1027–1044
Li J-N, Song D-Q, Lin Y-H, Hu Q-Y, Yin L, Bekesi G, Holland JF,
Jiang J-D (2003) Inhibition of microtubule polymerization by
3-bromopropionylamino benzoylurea (JIMB01), a new cancer-
icidal tubulin ligand. Biochem Pharmacol 65:1691–1699
Li H-Q, Lv P-C, Yan T, Zhu H-L (2009) Urea derivatives as
anticancer agents. Anticancer Agents Med Chem 9:471–480
Nantasenamat C, Naenna T, Isarankura-Na-Ayudhya C, Prac-
hayasittikul V (2005) Quantitative prediction of imprinting
factor of molecularly imprinted polymers by artificial neural
network. J Comput Aid Mol Des 19:509–524
Sondhi SM, Singh J, Rani R, Gupta PP, Agrawal SK, Saxena AK
(2010) Synthesis, anti-inflammatory and anticancer activity
evaluation of some novel acridine derivatives. Eur J Med Chem
45:555–563
Song D-Q, Wang Y, Wu L-Z, Yang P, Wang Y-M, Gao L-M, Li Y,
Qu J-R, Wang Y-H, Li Y-H, Du N-N, Han Y-X, Zhang Z-P,
Jiang J-D (2008) Benzoylurea derivatives as a novel class of
antimitotic agents: synthesis, anticancer activity, and structure–
activity relationships. J Med Chem 51:3094–3103
Suksai C, Pakawatchai C, Tuntulani T (2009) Synthesis and crystal
structure analysis of thiourea-pendant pyridines. J Chem Crys-
tallogr 39:348–352
123

Med Chem Res (2013) 22:4016–4029
4029
Suksrichavalit T, Prachayasittikul S, Nantasenamat C, Isarankura-
Na-Ayudhya C, Prachayasittikul V (2009) Copper complexes of
pyridine derivatives with superoxide scavenging and antimicro-
bial activities. Eur J Med Chem 44:3259–3265
Suvannang N, Nantasenamat C, Isarankura-Na-Ayudhya C, Prac-
hayasittikul V (2011) Molecular docking of aromatase inhibitors.
Molecules 16:3597–3617
Tengchaisri T, Chawengkirttikul R, Rachaphaew N, Reutrakul V,
Sangsuwan R, Sirisinha S (1998) Antitumor activity of triptolide
against cholangiocarcinoma growth in vitro and in hamsters.
Cancer Lett 133:169–175
Thanikaivelan P, Subramanian V, Raghava Rao J, Unni Nair B (2000)
Application of quantum chemical descriptor in quantitative
structure activity and structure property relationship. Chem Phys
Lett 323:59–70
Whitley DC, Ford MG, Livingstone DJ (2000) Unsupervised forward
selection: a method for eliminating redundant variables. J Chem
Inf Comput Sci 40:1160–1168
Witten IH, Frank E, Hall MA (2011) Data mining: practical machine
learning tools and techniques, 3rd edn. Morgan Kaufmann, San
Francisco
Worachartcheewan A, Nantasenamat C, Naenna T, Isarankura-
Na-Ayudhya C, Prachayasittikul V (2009) Modeling the activity
of furin inhibitors using artificial neural network. Eur J Med
Chem 44:1664–1673
Worachartcheewan A, Nantasenamat C, Isarankura-Na-Ayudhya C,
Prachayasittikul S, Prachayasittikul V (2011) Predicting the free
radical scavenging activity of curcumin derivatives. Chemometr
Intell Lab Syst 109:207–216
Worachartcheewan A, Prachayasittikul S, Pingaew R, Nantasenamat
C, Tantimongcolwat T, Ruchirawat S, Prachayasittikul V (2012)
Antioxidant, cytotoxicity, and QSAR study of 1-adamantylthio
derivatives of 3-picoline and phenylpyridines. Med Chem Res
21:3514–3522
Toth JE, Grindey GB, Ehlhardt WJ, Ray JE, Boder GB, Bewley JR,
Klingerman KK, Gates SB, Rinzel SM, Schultz RM, Weir LC,
Worzalla JF (1997) Sulfonimidamide analogs of oncolytic
sulfonylureas. J Med Chem 40:1018–1025
Vahdani S, Bayat Z (2011) A quantitative structure–activity relation-
ship (QSAR) study of anti-cancer drugs. Der Chemica Sinica
2:235–243
Valdes-Martinez J, Hernandez-Ortega S, Ackerman LJ, Li DT,
Swearingen JK, West DX (2000) Spectral and structural studies
of N-(2)-pyridylethyl-N0-arylthioureas. J Mol Struct 524:51–59
Zhang X, Lee YK, Kelley JA, Burke TR Jr (2000) Preparation of aryl
isothiocyanates via protected phenylthiocarbamates and appli-
cation to the synthesis of caffeic acid (4-isothiocyanato)phenyl
ester. J Org Chem 65:6237–6240
123