Synthesis, in vitro antitubercular activity and 3D-QSAR of novel quinoxaline derivatives

Article abstract of DOI:10.1111/j.1747-0285.2011.01246.x

Twenty new quinoxalines bearing azetidinone and thiazolidinone groups were synthesized by cyclocondensation of Schiff bases of quinoxaline-2, 3-dione and were characterized with several analytical tools. They were tested against Mycobacterium tuberculosis

Full text of DOI:10.1111/j.1747-0285.2011.01246.x

ª 2011 John Wiley & Sons A/S
doi: 10.1111/j.1747-0285.2011.01246.x
Chem Biol Drug Des 2011; 78: 988–998
Research Article
Synthesis, In Vitro Antitubercular Activity and
3D-QSAR of Novel Quinoxaline Derivatives
Ayarivan Puratchikody1, Ramalakshmi
of the later alone (3). As resistant strains of M. tuberculosis have
slowly emerged, treatment failures also emerge. The current front-
line therapy for tuberculosis consists of administering three or more
different drugs (usually isoniazid, rifampin, pyrazinamide and etham-
Natarajan2, Mohanapriya Jayapal3 and
Mukesh Doble3,*
1Department of Pharmaceutical Technology, Anna University
butol) over extended periods of time (4). Problems because of multi-
drug-resistant TB arise, and it becomes necessary to develop new
therapeutic agents to treat the new forms of the disease.
Tiruchirappalli, Tiruchirappalli 620 024, Tamilnadu, India
2Department of Pharmaceutical Chemistry, C. L. Baid Metha College
of Pharmacy, Chennai 600097, Tamilnadu, India
3Department of Biotechnology, Indian Institute of Technology
Madras, Chennai 600036, Tamilnadu, India
Quinoxaline nucleus is an imperative scaffold that is not only syn-
thetically important but also possesses a wide range of promising
biological activities. This ring is reported to possess antitubercular
activity (5–7). A literature survey showed that over 500 quinoxaline
derivatives were tested by the Tuberculosis Antimicrobial Acquisi-
tion and Coordinating Facility program including simple substituted
quinoxalines and their corresponding 1,4-di-N-oxides (8,9). Many of
these compounds possess excellent antitubercular activity. Among
the various heterocyclic compounds, quinoxalines form an attractive
biologically active molecule as these are a part of various antibiot-
ics such as echinomycin, levomycin and actinoleutin (10,11). The an-
titubercular activity for Mannich bases of quinoxaline against
M. tuberculosis has been reported (12).
*Corresponding author: Mukesh Doble, mukeshd@iitm.ac.in
Twenty new quinoxalines bearing azetidinone and
thiazolidinone groups were synthesized by cyclo-
condensation of Schiff bases of quinoxaline-2, 3-di-
one and were characterized with several analytical
tools. They were tested against Mycobacterium
tuberculosis H37Rv at a concentration of 10 lg ⁄ mL
by Microplate Alamar Blue Assay method. Quinoxa-
line derivatives with 2-chloro, dimethylamino and
nitro substitutions exhibited in vitro activity, com-
parable to that of the drug, isoniazid. Three-dimen-
sional quantitative structure–activity relationship
studies indicated that electrostatic and steric field
descriptors could explain the observed activity.
The developed model fits the data well and has
good predictive capability (r2 = 0.81, q2 = 0.71,
F = 27.06, r2_pred = 0.84, r2m = 0.84, r2 BS = 0.80).
Electronegative groups play an important role in
the antitubercular activity.
Azetidinones and thiazolidinones are structural subunit of penicillin.
Azetidinone is a four-membered nitrogen containing heterocycle
possessing powerful antimicrobial (13–15), anti-inflammatory (16)
and antitubercular activities (17,18). Thiazolidinone is an interesting
heterocycle with various pharmacological activities. They are
reported to possess a wide range of biological activities including
antibacterial, antifungal, antiHIV, etc. (19).
Key words: azetidinone, antitubercular activity, thiazolidinones,
three-dimensional, electron withdrawing groups, quantitative structure–
activity relationship, quinoxaline
Based on this logic, an attempt to synthesize some novel com-
pounds containing quinoxaline ring fused to azetidinones and thia-
zolidinones and study their antitubercular activity. With the aim to
establish the influence of substitutions on the antitubercular activ-
ity, 3D-quantitative structure–activity relationship (3D-QSAR) was
developed. Quantitative structure–activity relationship helps to
understand the structural features that are necessary for the antitu-
bercular activity of compounds, and it serves as a guide in the
design of more potent molecules. Use of QSAR for understanding
antibacterial, antifungal and antiTB activities of small molecules
has been well documented (20–25).
Abbreviations: 3D-QSAR, three-dimensional quantitative struc-
ture–activity relationship; CFU, colony-forming units; MIC, minimum
inhibitory concentration; MLR, multiple linear regression.
Received 15 October 2009, revised 8 September 2011 and accepted for
publication 17 September 2011
Tuberculosis is one of the most common infectious diseases known
to human. One-third of the world population is infected with Myco-
bacterium tuberculosis, and the World Health Organization (WHO)
estimates that within the next 20 years, approximately thirty million
people will be infected with the bacillus (1). With the global spread
of HIV, similar increase in the incidence rate of tuberculosis (TB) and
mortality is to be feared (2). Coinfection with TB and HIV can
increase the risk of death twice when compared with death because
Materials and Methods
Instruments
All chemicals were purchased from Sigma Aldrich (Bangalore, India).
Uncorrected melting points were determined using
a Sigma
988

Antitubercular Activity of Quinoxalines
scientific apparatus (Coimbatore, India). 1H and 13C NMR spectra
were measured on a Bruker AV-500 instrument (300 MHz; Bruker,
Mumbai, India) using tetramethylsilane (TMS) as an internal stan-
dard and DMSO-d6 as solvent. Mass spectra were recorded on a
GCMS QP 5000 Shimadzu (Shimadzu Analytical (India) Pvt. Ltd,
Mumbai, India). Elemental analysis was performed on Perkin Elmer
240 CHN analyzer (Perkin Elmer Life Sciences Inc., Boston, MA,
USA). IR was recorded on ABB Bomem FTIR Spectrometer MB104
with KBr pellet (ABB Bomem FTIR, Faridabad, India).
Compound 1a (3.42 g, 2 mmol) was added to a mixture of aniline
(1.85 mL) and ethanol (10 mL) and stirred magnetically at room
temperature for 30 min. To this well-stirred mixture, few drops of
glacial acetic acid were added, and the stirring was continued for
further 10 min. Then, it was irradiated in microwave oven (400 W)
for 2 min. The reaction mixture was poured into crushed ice. A
solid was collected by filtration, washed with water and recrystal-
lized from ethanol to yield 5.1 g of 3-((phenylimino)-3,4-dihydroqui-
noxaline)-one (1b) (yield 85%).
To a stirred solution of compound 1b (2.37 g, 1 mmol) and o-
phenylenediamine (1.08 g) in 10 mL of ethanol, few drops of glacial
acetic acid were slowly added at room temperature. This mixture
was irradiated with microwave oven for 2 min. The crude product
was filtered off, washed with water, dried by applying vacuum and
Synthesis of compounds
The synthetic route is shown in Scheme 1. The yield of the
compounds was not optimized. Intermediate quinoxaline-2,3-dione
(1a) was synthesized based on reported procedure (26).
NH2
NH2
H
N
H
N
N
O
O
H2N
i
+
+
N
H
O
N
H
1a
1b
ii
H
N
N
N
RCHO
1(d-m)
+
N
H
NH2
iii
1c
H
N
N
N
N
H
N
1(n-w)
H
N
R
N
N
iv
R
Cl
N
H
v
N
O
2(n-w)
H
N
N
N
R
S
N
H
N
O
3(n-w)
Scheme 1: Reagent and conditions (i) and (ii) EtOH, glacial acetic acid, microwave 400 W, 2 min.; (iii) DMF, glacial acetic acid microwave
400 W, 45 seconds; (iv) chloroacetyl chloride, 1,2-dioxan, triethylamine microwave 400 W, 3 min; (v) thioglycolic acid, EtOH microwave 400 W,
4 mins.
Chem Biol Drug Des 2011; 78: 988–998
989

Puratchikody et al.
recrystallized with ethanol to give 2.41 g of N-((E)-3-(phenylimino)-
506 (M+)1, 3%), 414 (25%), 311 (22%), 286 (31%), 221 (29%), 193
(30%), 103 (32%), 80 (100%), and 58 (47%). Anal. Calcd for
C29H22ClN5O2: C, 68.57%; H, 4.37%; N, 13.79%. Found: C, 68.54%;
H, 4.37%; N, 13.77%.
3,4-dihydroqunoxalin-2(1H)-ylidene)benzene-1,2-diamine(1c)
(yield
74%).
Compound 1c (3.27 g, 1 mmol) was dissolved in 10 mL of ethanol,
and then, the required aromatic aldehyde (1d–m) was added. Then
to this mixture, few drops of glacial acetic acid were added slowly
and stirred with a magnetic stirrer for 30 min. It was kept in micro-
wave oven for 3 min. The reaction was cooled to room tempera-
ture, washed with water, dried by MgSO4 and filtered. The filtrate
was evaporated to dryness to produce corresponding Schiff bases
(1n–w) (yield ranges between 58% and 77%), and the residue
was purified by silica gel column chromatography.
3-Chloro-4-(4-methoxyphenyl)-1-(2-((E)-((E)-3-
(phenylimino)-3,4-dihydroquinoxalin-2-(1H)-
ylidene)amino)phenyl)azetidin-2-one (2p)
Yield 80.01%; m.p. 268–272 ꢀC; IR (tmax, ⁄ cm, KBr): 3347(N-H),
1749(Azetidinone, C=O), 1655 (C=N), 1229(Ar-OCH3) and 764 (Azetid-
inone, C-Cl); 1H NMR(DMSO-d6, 300 MHz) d: 5.8 (1H, dd,
J = 5.7,10.7, CH-Cl of Azetidinone), 5.4 (1H, s, CH-N), 4.2 (1H, s, C-
NH) and 3.9 (1H, s, Ar-OCH3) p.p.m.; 13C (DMSO-d6) d: 62.32,
163.22, 157.142 and 63.45 p.p.m.; FAB-MS (m ⁄ z): 521 (M+,11%),
522 (M+ + 1, 7%), 520 (M+ ) 1, 5%), 415 (8%), 340 (7%), 206
(20%), 107 (38%), 80 (100%) and 77 (42%). Anal.Calcd for
C30H24ClN5O2: C 69.03%; H, 4.63%; N, 13.42%. Found: C, 69.03%;
H, 4.60%; N, 13.39%.
The Schiff base(s) 1n–w (0.01 mmol) was dissolved in 1,2-dioxan,
followed by the addition of chloroacetyl chloride (1.12 g,
0.01 mmol,) and triethylamine (1.01 g, 1 mmol) slowly in a round
bottom flask. The reaction mixture was stirred for 1 h, and then, it
was irradiated in a microwave oven for 3 min. The solid that
appeared after the addition of water was filtered off and dried to
get 2n–w (yield varying between 11.48% and 80.01%). The resi-
due was purified by silica gel column eluted with ethylace-
tate ⁄ petroleum ether (1:5) to yield respective azetidinones.
3-Choro-4-(4-chorophenyl)-1-(2-((E)-((E)-
3(phenylimino)-3,4-dihydroquinoxalin-2-(1H)-
ylidene)amino)phenyl)azetidin-2-one (2q)
Yield 70.20%; m.p. 180–182 ꢀC; IR (tmax, ⁄ cm, KBr): 3339 (N-H),
2889(C-H), 1743 (Azetidinone, C=O), 1655 (C=N), 765 (Azetidinone,
A mixture of Schiff base (s), 1n–w (0.01 mmol), thioglycolic acid
(1 mmol) and ethanol (10 mL) were irradiated in microwave oven for
4 min. The reaction mixture was poured into ice cooled water. The
solid thus obtained was removed by filtration, and the residue was
purified by silica gel column as mentioned above to afford 3n–w
(yield ranges between 64% and 88%). The chemical structures of
1
C-Cl); H NMR (DMSO-d6, 300 MHz) d: 5.4 (1H, dd, J = 5.8, 10.3,
CH-Cl of Azetidinone), 4.6 (1H, s, CH-N), and 4.0 (1H, s, C-NH)
p.p.m.; 13C (DMSO-d6) d: 163.65, 132.67, 117, 132.54 and 63.58
p.p.m.; FAB-MS (m ⁄ z): 525 (M+, 3%), 526 (M+ + 1, 5%), 524
(M+ ) 1, 5%), 434 (7%), 311 (26%), 243 (49%), 234 (35%), 191
(37%), 113 (38%), 84 (31%), 80 (100%) and 68 (9%). Anal. Calcd
for C29H21Cl2N5O: C, 66.17%; H, 4.02%; N, 13.28%. Found: C,
66.14%; H, 4.00%, N, 4.00%.
1
the synthesized compounds were determined on the basis of IR, H
NMR, 13C NMR, mass spectral data and elemental analysis.
3-Choro-4-(2-hydroxyphenyl)-1-(2-((E)-((E)-
3(phenylimino)-3,4-dihydroquinoxalin-2-(1H)-
ylidene)amino)phenyl)azetidin-2-one (2n)
3-Choro-4-(2-chlorophenyl)-1-(2-((E)-((E)-
Yield 52.66%, m.p. 270–272 ꢀC; IR (tmax, ⁄ cm, KBr): 3610 (Ar-OH),
3367 (N-H), 2960(C-H), 1741 (Azetidinone, C=O), 1665 (C=N) and 765
(Azetidinone, C-Cl); 1H NMR (DMSO-d6, 300 MHz) d: 7.7 (1H, d,
J = 5.8Hz, CH-Cl of Azetidinone), 5.2 (1H, s, CH-N), 7.6 (1H, s, Ar-
OH) and 6.44 (1H, dd, J = 1.8, 8.4 C-NH) p.p.m.; 13C NMR (DMSO-
d6) d: 173, 162, 131 and 121 p.p.m.; fast atom bombardment mass
spectroscopy (FAB-MS) (m ⁄ z): 507 (M+, 7%), 508 (M+ + 1, 6%), 506
(M+ ) 1, 5%), 416 (24%), 311 (22%) 235 (44%), 132 (27%), 106
(100%), and 77 (39%). Anal. Calcd for C29H22ClN5O2: C, 68.57%; H,
4.37%;N, 13.79%. Found: C, 68.54%; H, 4.37%; N, 13.77%.
3(phenylimino)-3,4-dihydroquinoxalin-2-(1H)-
ylidene)amino)phenyl)azetidin-2-one (2r)
Yield 27.97%; m.p. 220–222 ꢀC; IR (tmax, ⁄ cm, KBr): 3348 (N-H),
2995(C-H), 1757 (Azetidinone, C=O), 1658 (C=N), 1338 (C-N) and 747
(Azetidinone, C-Cl); 1H NMR (DMSO-d6, 300 MHz) d: 5.7 (1H, d,
J = 5.8, CH-Cl of Azetidinone), 4.9 (1H, dd, J = 1.76,8.34 CH-N) and
4.0 (1H, s, C-NH) p.p.m.; 13C (DMSO-d6) d: 163.21, 161.98, 135.33,
132.78 and 62.54 p.p.m.; FAB-MS (m ⁄ z): 525 (M+, 8%), 526
(M+ + 1, 6%), 524 (M+ ) 1, 5%), 448 (19%), 304 (18%), 290 (20%),
144 (42%), 91 (28%), 80 (100%), 90 (16%) and 45 (51%). Anal.
Calcd for C29H21Cl2N5O: C, 66.17%; H, 4.02%; N, 13.28%. Found: C,
66.14%; H, 4.60%; N, 13.30%.
3-Choro-4-(4-hydroxyphenyl)-1-(2-((E)-((E)-
3(phenylimino)-3,4-dihydroquinoxalin-2-(1H)-
ylidene)amino)phenyl)azetidin-2-one (2o)
Yield 62.80%; m.p. 200–202 ꢀC; IR (tmax, ⁄ cm, KBr): 3396 (Ar-OH),
3349 (N-H), 2960 (C-H), 1753 (Azetidinone, C=O), 1665 (C=N) and
3-Choro-4-(4-methylphenyl)-1-(2-((E)-((E)-
3(phenylimino)-3,4-dihydroquinoxalin-2-(1H)-
ylidene)amino)phenyl)azetidin-2-one (2s)
Yield 58.33%; m.p. 176–178 ꢀC; IR (tmax, ⁄ cm, KBr): 3341 (N-H),
2965 (C-H), 2850, 2917 (Ar-CH3); 1749 (Azetidinone, C=O), 1655
1
764 (C-Cl); H NMR (DMSO-d6, 300 MHz) d: 5.8 (1H, d, J = 5.8, CH-
Cl of Azetidinone), 5.4 (1H, s, CH-N), 4.2 (1H, s, C-NH) and 7.2 (1H,
s, Ar-OH) p.p.m.; 13C (DMSO-d6) d: 163.26, 162.55, 156.58, 149.34
and 62.43 p.p.m.; FAB-MS (m ⁄ z): 507 (M+, 5%), 508 (M+ + 1, 4%),
1
(C=N), 1439 (C-N) and 749 (Azetidinone, C-Cl); H NMR (DMSO-d6,
300 MHz) d: 5.5 (1H, d, J = 5.7, CH-Cl of Azetidinone), 4.6 (1H, s,
990
Chem Biol Drug Des 2011; 78: 988–998

Antitubercular Activity of Quinoxalines
CH-N), 3.9 (1H, s, C-NH) and 2.3 (3H, s, (Ar-CH3) p.p.m.; 13C (DMSO-
d6) d: 164.11, 165.52, 134.73 and 63.16 p.p.m.; FAB-MS (m ⁄ z): 505
(M+, 5%), 506 (M+ + 1, 4%), 504 (M+ ) 1, 7%), 428 (17%), 325
(27%), 235 (36%), 191 (37%), 132 (34%), 106 (100%) and 77(33%).
Anal. Calcd for C30H24ClN5O: C, 71.21%; H, 4.78%; N, 13.84%.
Found: C, 71.20%; H, 4.75%; N, 13.81%.
and 765 (Azetidinone, C-Cl); 1H NMR (DMSO-d6, 300 MHz) d: 5.8
(1H, s, CH-Cl of Azetidinone), 5.2 (1H, s, CH-N), 4.0 (1H, d, J = 1 .8,
C-NH), 4.9 (1H, s, Ar-OH) and 3.7 (3H, s, OCH3) p.p.m.; 13C (DMSO-
d6) d: 143.7, 150.2, 143.43, 63.79 and 56.21 p.p.m.; FAB-MS (m ⁄ z):
537 (M+, 4%), 538 (M+ +1, 4%), 536 (M+ ) 1, 3%), 431 (8%), 311
(6%), 221 (17%), 123 (30%), 80 (100%) and 90 (32%). Anal. Calcd
for C30H24ClN5O3: C, 66.97%; H, 4.53%; N, 13.00%. Found: C,
66.97%; H, 4.53%; N, 13.00%.
3-Choro-4-(4-(dimethylamino)phenyl)-1-(2-((E)-
((E)-3(phenylimino)-3,4-dihydroquinoxalin-2-(1H)-
ylidene)amino)phenyl)azetidin-2-one (2t)
2-(2-Hydroxyphenyl)-3-(2-((Z)-((E)-3-
Yield 11.48%; m.p. 190–192 ꢀC; IR (tmax, ⁄ cm, KBr): 3348 (N-
H),2967(C-H), 1437(C-N), 1753 (Azetidinone, C=O), 1655 (C=N), 1290
(N-(CH3)2) and 765 (Azetidinone, C-Cl); 1H NMR (DMSO-d6,
300 MHz) d: 5.3 (1H, s, CH-Cl of Azetidinone), 5.2 (1H, dd,
J = 1.7,8.34, CH-N), 4.0 (1H, s, C-NH) and 2.9 (6H, s, N-(CH3)2)
p.p.m.; 13C (DMSO-d6) d: 164.22, 162.34, 147.64, 135.56 and
40.3 p.p.m.; FAB-MS (m ⁄ z): 534 (M+, 9%), 535 (M + 1, 7%), 533
(M+ ) 1, 7%), 313 (16%), 222 (18%), 180 (28%), 132 (25%), 106
(100%), 103 (23%) and 77(42%). Anal. Calcd for C31H27ClN6O: C,
69.59%; H, 5.06%; N, 15.70%. Found: C, 69.55%; H, 5.06%; N,
15.70%.
(phenylimino)-3,4-dihydroquinoxalin-2-(1H)-
ylidene)amino)phenyl)thiazolidin-4-one (3n)
Yield 82%; m.p. 250–252 ꢀC; IR (tmax, ⁄ cm, KBr): 3367 (Ar-OH),
1
2993 (C-H), 1660 (C=O), 1451 (C=N) and 613 (C-S); H NMR (DMSO-
d6, 300 MHz) d: 4.5 (2H, s, NH), 3.5 (2H, s, CH2), 6.0 (1H, s, CH)
and 7.5(Ar-OH) p.p.m.; 13C (DMSO-d6) d: 162.02, 143.45,131.4, 119
and 110.6 p.p.m.; FAB-MS (m ⁄ z): 505 (M+ 6%), 506 (M+ + 1 3%),
504 (M+ ) 1 5%), 465(19%), 441 (18%), 367 (26%), 192 (25%), 132
(27%) and 91 (100%). Anal. Calcd for C29H23N5O2S: C, 68.89%; H,
4.59%; N, 13.85%. Found: C, 68.86%; H, 4.56%; N, 13.88%.
2-(4-Hydroxyphenyl)-3-(2-((Z)-((E)-3-
3-Choro-4-(3-nitrophenyl)-1-(2-((E)-((E)-
(phenylimino)-3,4-dihydroquinoxalin-2-(1H)-
ylidene)amino)phenyl)thiazolidin-4-one (3o)
3(phenylimino)-3,4-dihydroquinoxalin-2-(1H)-
ylidene)amino)phenyl)azetidin-2-one (2u)
Yield 45.03%; m.p. 212–214 ꢀC; IR (tmax, ⁄ cm, KBr):3347 (N-H),
1748 (Azetidinone, C=O), 1658 (C=N), 1557 (C-NO2) and 764 (Azetidi-
none, C-Cl); 1H NMR (DMSO-d6, 300 MHz) d: 5.7 (1H, d, J = 5.8,
CH-Cl of Azetidinone), 5.2 (1H, s, CH-N) and 3.8 (1H, s, C-NH)
p.p.m.; 13C (DMSO-d6) d: 163.01, 162.06, 162.06, 148.23, 135.31
and 62.74 p.p.m.; FAB-MS (m ⁄ z): 536 (M+, 7%), 537 (M+ + 1, 6%),
535 (M+ ) 1, 4%), 430 (8%), 310 (16%), 289 (12%), 225 (14%), 221
(18%), 144 (48%), 85(31%), 80 (100%) and 48(47%). Anal. Calcd for
C29H21ClN6O3: C, 64.87%; H, 3.94%; N, 15.65%. Found: C, 64.84%;
H, 3.91%; N, 14.51%.
Yield 69%; m.p. 238–240 ꢀC; IR (tmax
, ⁄ cm, KBr): 3349(N-H),
1
2885(C-H), 1614 (C=O), 1505(C=N) and 663 (C-S); H NMR (DMSO-
d6, 300 MHz) d: 4.3 (2H, s, NH), 3.7 (2H, s, CH2), 6.1 (1H, s, CH)
and 7.7 (Ar-OH) p.p.m.; 13C (DMSO-d6) d: 163.03, 156.4, 112.6,
133.4, 119. p.p.m.; FAB-MS (m ⁄ z): 505 (M+, 9%), 506 (M+ +1, 5%),
504 (M+ ) 1, 7%), 412(16%), 312 (15%), 221(18%), 192 (15%), 132
(13%) and 91 (100%). Anal. Calcd for C29H23N5O2S: C, 68.89%; H,
4.59%; N, 13.85%. Found: C, 68.87; H, 4.57; N, 13.88.
2-(2-Methoxyphenyl)-3-(2-((Z)-((E)-3-
(phenylimino)-3,4-dihydroquinoxalin-2-(1H)-
ylidene)amino)phenyl)thiazolidin-4-one (3p)
3-Choro-4-(furan-3-yl)-1-(2-((E)-((E)-
Yield 74%; m.p. 244–246 ꢀC; IR (tmax, ⁄ cm, KBr): 3365(N-H),
3(phenylimino)-3,4-dihydroquinoxalin-2-(1H)-
ylidene)amino)phenyl)azetidin-2-one (2v)
Yield 61.75%; m.p. 258–262 ꢀC; IR (tmax, ⁄ cm, KBr): 3349 (N-H),
2991(C-H), 1758 (Azetidinone, C=O), 1665 (C=N), and 765 (Azetidinone,
2992(C-H), 1613 (C=O), 1518(C=N), 1278(Ar-OCH3) and 720(C-S); 1H
NMR (DMSO-d6, 300 MHz) d: 4.2 (2H, s, NH), 3.4 (2H, s, CH2), 5.4
(1H, s, CH) and 4.8 (1H, s, OCH3) p.p.m.; 13C (DMSO-d6) d: 163.87,
150.6, 133.4, 130.3 and 119 p.p.m.; FAB-MS (m ⁄ z): 519(M+, 3%),
520 (M++ 1, 5%), 518 (M+ ) 1, 7%), 412 (31%) 288 (22%),130
(100%) 91 (36%) and 106 (29%). Anal. Calcd for C30H25N5O2S: C,
69.34%; H, 4.85%; N, 13.48%. Found: C, 69.31%; H, 4.77%; N,
13.35%.
1
C-Cl); H NMR (DMSO-d6, 300 MHz) d: 5.7 (1H, dd, J = 1.53, 8.23,
CH-Cl of Azetidinone), 5.1 (1H, d, J = 3.8, CH-N) and 4.0 (1H, s, C-NH)
p.p.m.; 13C (DMSO-d6) d: 162.04, 142.01, 129.26 and 62.87 p.p.m.;
FAB-MS (m ⁄ z): 481 (M+, 5%), 482 (M+ + 1, 5%), 480 (M+ ) 1, 3%),
414 (12%), 311 (7%), 221 (17%), 185 (23%), 170 (10%), 106 (100%)
and 91(39%). Anal. Calcd for C27H20ClN5O2: C, 67.29%; H, 4.18%; N,
14.53%. Found: C, 67.30%; H, 4.15%; N, 14.51%.
2-(4-Chlorophenyl)-3-(2-((Z)-((E)-3-(phenylimino)-
3,4-dihydroquinoxalin-2-(1H)-
ylidene)amino)phenyl)thiazolidin-4-one (3q)
Yield 71%; m.p. 251–253 ꢀC; IR (tmax, ⁄ cm, KBr):3369 (N-H), 2993
(C-H), 1613 (C=O), 1459 (C=N) and 612 (C-S); 1H NMR (DMSO-d6,
300 MHz) d: 4.3 (2H, s, NH), 3.4 (2H, s, CH2) and 5.8 (1H, s, CH)
p.p.m.; 13C (DMSO-d6) d: 164.32, 133.4, 131.56, 130.3 and
119 p.p.m.; FAB-MS (m ⁄ z): 523 (M+, 8%), 524 (M+ + 1, 5%), 522
3-Choro-4-(4-hydroxy-3-methoxyphenyl)-1-(2-
((E)-((E)-3(phenylimino)-3,4-dihydroquinoxalin-2-
(1H)-ylidene)amino)phenyl)azetidin-2-one (2w)
Yield 47.68%; m.p. 182–186 ꢀC; IR (tmax, ⁄ cm, KBr): 3389 (Ar-OH),
3348 (N-H), 1743 (Azetidinone, C=O), 1665 (C=N), 1229 (Ar-OCH3)
Chem Biol Drug Des 2011; 78: 988–998
991

Puratchikody et al.
(M+ ) 1, 7%), 412 (29%), 311 (50%), 211 (31%),106 (100%) and
91(49%). Anal. Calcd for C29H22ClN5OS: C, 66.47%; H, 4.23%; N,
13.36%. Found: C, 66.49%; H, 4.25%; N, 13.39%.
3-(2-((Z)-((E)-3-(Phenylimino)-3,4-
dihydroquinoxalin-2-(1H)-ylidene)amino)phenyl)-
2-(tetrahydrofuran-3-yl)thiazolidin-4-one (3v)
Yield 72%; m.p. 245–247ꢀC; IR (tmax, ⁄ cm, KBr): 3338 (N-H), 2287
1
(C-H), 1613 (C=O), 1508 (C=N), 1164 (Ar-CH) and 715 (C-S); H NMR
2-(2-Chlorophenyl)-3-(2-((Z)-((E)-3-(phenylimino)-
3,4-dihydroquinoxalin-2-(1H)-
(DMSO-d6, 300 MHz) d: 4.6 (2H, s, NH), 3.2 (2H, s, CH2) and 5.8
(1H, s, CH) p.p.m.; 13C (DMSO-d6) d: 162.88, 142.54, 132.8, 129.7
and 118.7 p.p.m.; FAB-MS (m ⁄ z): 479 (M+, 7%), 480 (M+ + 1, 8%),
478 (M+ ) 1, 6%), 311 (38%), 168 (24%), 106 (29%) and 91
(100%). Anal. Calcd for C27H21N5O2S: C, 67.62; H, 4.41; N, 14.50.
Found: C, 67.63; H, 4.40; N, 14.53.
ylidene)amino)phenyl)thiazolidin-4-one (3r)
Yield 88%; m.p. 234–236 ꢀC; IR (tmax, ⁄ cm, KBr): 3355 (N-H), 2995
(C-H), 1614 (C=O), 1507 (C=N) and 662 (C-S); 1H NMR (DMSO-d6,
300 MHz) d: 4.3 (2H, s, NH), 3.8 (2H, s, CH2), and 5.6 (1H, s, CH)
p.p.m.; 13C (DMSO-d6) d: 136.45, 136.4, 123.3 and 117 p.p.m.; FAB-
MS (m ⁄ z): 523 (M+, 7%), 524 (M+ +1, 13%), 522 (M+ ) 1, 11%),
488 (17%), 312 (19%) 106 (16%) and 91(100%) Anal. Calcd for
C29H22ClN5OS: C, 66.47%; H, 4.26%; N, 13.36%. Found: C, 66.51;
H, 4.26; N, 13.39.
2-(3-Methoxy-4-hydroxyphenyl)-3-(2-((Z)-((E)-3-
(phenylimino)-3,4-dihydroquinoxalin-2-(1H)-
ylidene)amino)phenyl)thiazolidin-4-one (3w)
Yield 64%; m.p. 278–280ꢀC; IR (tmax, ⁄ cm, KBr): 3375 (OH), 3345(N-
H), 1610 (C=O), 1515 (C=N), 1387 (Ar-CH2), 1245 (OCH3) and 685(C-
S); 1H NMR (DMSO-d6, 300 MHz) d: 4.0 (2H, s, NH), 2.3 (2H, s,
CH2), 4.6 (1H, s, CH),3.2 (3H, s, Ar-OCH3) and 5.3(1H, s, Ar-OH)
p.p.m.; 13C (DMSO-d6) d: 162.56, 150.6, 130.3 and 119 p.p.m.; FAB-
MS (m ⁄ z): 535 (M+, 6%), 536 (M+ + 1 2%), 534 (M+ ) 1, 6%), 488
(25%), 312 (31%), 123 (24%), 106 (100%), and 91(22%). Anal. Calcd
for C30H25N5O3S: C, 67.27%; H, 4.70%; N, 13.08%. Found: C,
67.24%; H, 4.68%; N, 13.04%.
2-(4-Methylphenyl)-3-(2-((Z)-((E)-3-(phenylimino)-
3,4-dihydroquinoxalin-2-(1H)-
ylidene)amino)phenyl)thiazolidin-4-one (3s)
Yield 73%; m.p. 252–254 ꢀC; IR (tmax
, ⁄ cm, KBr): 3339 (N-H),
2887(C-H), 1613 (C=O), 1515 (C=N), 1387 (Ar-CH3), 615 (C-S); 1H
NMR (DMSO-d6, 300 MHz) d: 4.5 (2H, s, NH), 3.7 (2H, s, CH2)
and 2.8 (1H, s, CH3) p.p.m.; 13C (DMSO-d6) d: 162.55,132.8,129.7
and 118.7 p.p.m.; FAB-MS (m ⁄ z): 503 (M+, 2%), 504 (M+ +1
16%), 502 (M+ ) 1, 4%), 412 (16%), 397 (15%), 311 (32%), 192
(20%), 158 (17%) and 91(100%). Anal. Calcd for C30H25N5OS: C,
71.55%; H, 5.00%; N, 13.91%. Found: C, 71.57%; H, 5.03%; N,
13.94%.
Anti-mycobacterial activity
Antitubercular activity was evaluated against M. tuberculosis H37
Rv ATCC27294 using Microplate Alamar Blue Assay (27). The test
was performed in black, clear-bottomed, 96-well microplates (black
view plates; Packard Instrument Company, Meriden, CT, USA) to
minimize the background fluorescence. Initial drug dilutions were
prepared in dimethylsulfoxide, and subsequent twofold dilutions
were performed in 0.1 mL of 7H9GC media in the microplates.
0.1 mL of 2.5 · 106 CFU ⁄ mL of M. tuberculosis H37 Rv in 7H9GC
was added to each well of the 96-well microtitre plate containing
the test compounds. Three control wells containing drug and med-
ium, bacteria and medium and medium alone were prepared. All
microtitre plates were incubated at 37 ꢀC. On the seventh day of
incubation, Alamar Blue dye solution (20 mL Alamar Blue solution
and 12.5 mL of 20% Tween-80) was added to all the wells, and
plates were re-incubated at 37 ꢀC for 24 h. Fluorescence was mea-
sured in a Victor II multilabel fluorometer (Perkin Elmer Life Sci-
ences Inc). Minimum inhibitory concentration (MIC) was determined
from the colour change.
2-(4-Dimethylaminophenyl)-3-(2-((Z)-((E)-3-
(phenylimino)-3,4-dihydroquinoxalin-2-(1H)-
ylidene)amino)phenyl)thiazolidin-4-one (3t)
Yield 68%; m.p. 228–230 ꢀC; IR (tmax, ⁄ cm, KBr):3342 (N-H), 2994
(C-H), 1611 (C=O), 1519 (C=N), 1295 (N-(CH3)2) and 617(C-S); 1H
NMR (DMSO-d6, 300 MHz) d: 4.1 (2H, s, NH), 3.7 (2H, s, CH2), 5.7
(1H, s, CH) and 2.9 (6H, s, N-(CH3)2) p.p.m.; 13C (DMSO-d6) d:
162.44, 133.1, 128.9and 119.6 p.p.m.; FAB-MS (m ⁄ z): 532 (M+, 4%),
533 (M+ + 1, 7%), 531 (M+ ) 1, 4%), 488 (23%), 312 (36%), 221
(32%), 120 (27%) and 106 (100%). Anal. Calcd for C31H26N6OS: C,
69.9%; H, 5.3%; N, 15.78%. Found, C, 69.96%; H, 5.33%; N,
15.75%.
2-(3-Nitrophenyl)-3-(2-((Z)-((E)-3-(phenylimino)-
3,4-dihydroquinoxalin-2-(1H)-
ylidene)amino)phenyl)thiazolidin-4-one (3u)
Yield 76%; m.p. 270–272 ꢀC; IR (tmax, ⁄ cm, KBr): 3361(N-H), 1600
(C=O), 1525 (C=N), 1349 (Ar-NO2), 688 (C-S); 3361(N-H), 1600 (C=O),
1525 (C=N), 1349 (Ar-NO2)and 688 (C-S); 1H NMR (DMSO-d6,
300 MHz) d: 4.7 (2H, s, NH), 3.2 (2H, s, CH2) and 5.6 (1H, s, CH)
p.p.m.; 13C (DMSO-d6) d: 143.68,133.1, 128.9 and119.6 p.p.m.; FAB-
MS (m ⁄ z): 534 (M+, 6%), 535 (M+ + 1, 5%), 533 (M+ ) 1, 4%), 312
(34%), 299 (16%), 158 (24%) and 91(100%). Anal. Calcd for
C29H22N6O3S: C, 65.15%; H, 4.15%; N, 15.72%. Found: C, 65.12%;
H, 4.12%; N, 15.70%.
Quantitative structure–activity relationship
The structure of the 20 quinoxalines (Table 1) was drawn, and their
minimum energy conformations were determined with MM+ force
field using HYPERCHEM version 8 software (Molecular Modelling Tool,
Hypercube Inc, Gainesville, FL, USA) (20). The 3D-QSAR studies
were performed using VLIFE MOLECULAR DESIGN SUITE 3.5 (VLife Sci-
ences Technologies Pvt. Ltd., Pune, India).
The molecules were aligned on the common fragment, namely qui-
noxiline. The MIC values in lM were converted to )log (MIC) and
992
Chem Biol Drug Des 2011; 78: 988–998

Antitubercular Activity of Quinoxalines
Table 1: The in-vitro antitubercular activity of compounds 2n–w and 3n–w against Mycobacterium tuberculosis H37Rv strain
H
N
H
N
N
N
N
N
R
R
S
Cl
N
H
N
H
N
N
O
O
(2n–w)
Compounds
(3n–w)
Minimum inhibitory concentration (lg ⁄ mL)
R
2n
50.00
OH
2o
2p
3.00
1.53
OH
CH
3
O
2q
2r
1.40
0.89
Cl
Cl
2s
2t
20.6
0.76
CH
3
CH
CH
–
3
N
3
O
2u
0.97
+
N
O
2v
10.1
11.8
O
CH
O
3
2w
OH
Chem Biol Drug Des 2011; 78: 988–998
993

Puratchikody et al.
Table 1: (Continued).
Compounds
R
Minimum inhibitory concentration (lg ⁄ mL)
3n
13.8
OH
3o
3p
3.70
1.38
OH
CH
3
O
3q
3r
1.38
0.76
Cl
Cl
3s
3t
19.00
0.67
CH
3
CH
CH
3
N
3
3u
1.77
O
+
N
–
O
3v
36.00
35
O
3w
CH
O
3
OH
NH
Isoniazid
0.46
O
NH
2
N
994
Chem Biol Drug Des 2011; 78: 988–998

Antitubercular Activity of Quinoxalines
were used as dependant variable in the regression equation. Six-
teen molecules were chosen at random in the training set and four
molecules in the test set (28). The former was used to generate
the model and the later for determining its predictive capability.
products), with the appropriate aldehyde(s) (1d–m) and intermedi-
ate
N-((E)-3-(phenylimino)-3,4-dihydroqunoxalin-2(1H)-ylidene)ben-
zene-1,2-diamine (1c), under microwave atmosphere. The cyclization
of these intermediates, 1n–w, with chloroacetyl chloride and thio-
glycollic acid leads to the formation of products, 2n–w and 3n–
w, respectively. During the course of the reaction, the thiazolidi-
none series, 3n-w, gave better yields than azetidinone series, 2n–
w. The yields of 2n–w were in the range of 11.48–80.01%,
whereas the yields of thiazolidinones 3n–w were 64.00–88.00%.
A 3D cubic lattice grid in the x, y and z directions was created
(dimension of the box 19.9 · 20.9 · 19.5 ꢀ) to encompass the
aligned molecules. The descriptors were calculated using a sp3 car-
bon probe atom with a van der Waals radius of 1.52 ꢀ, a charge
of +1.0 and a dielectric constant of 1.0 to generate steric, electro-
static and hydrophobic fields at each lattice point (29,30). Three
thousand five hundred and thirty descriptors that included all the
electrostatic, steric and hydrophobic properties inside this cubic lat-
tice were evaluated for all the twenty quinoxaline molecules. Sev-
eral literature reports give a description of descriptors that are
used in QSAR (31–33). The best set of descriptors from this large
pool was short listed using the stepwise forward variable selection
method for generating the multiple linear regression equation. The
cross-correlation limit of 0.5, q2 as the selection criteria and auto-
scaling method were adapted during the selection procedure. Sev-
eral statistical parameters such as, squared correlation coefficient
for the training set (r2), r2 PRESS, r2 BS and rm2 ðtestÞwere evalu-
ated. rm2 ðtestÞ is a modified r2 and is evaluated for better predictive
1
All the compounds were characterized by IR, H and 13C NMR, MS
and elemental analyses.
The IR spectra of the compounds, 2n–w, showed strong absorption
band in the range of 1750–1650 ⁄ cm characteristic of the b-lactam
carbonyl group, which are lower than the reported values
(1800 ⁄ cm) for ketones, aldehydes and other amides. This may be
due to the conjugation of unpaired electrons of the ring nitrogen
with carbonyl group in the b-lactam of azetidinone resulting in
increased single-bond character and lowering of the carbonyl single
absorption frequency (35). The absorption peaks because of C-Cl
appeared in the range of 749–765 ⁄ cm in the compounds 2n–w.
The carbonyl stretching vibrations for the thiazolidinones derivatives
(3n–w) appeared between 1660 and 1600 ⁄ cm. The absorption
bands ranging from 612 to 715 ⁄ cm indicate the presence of C-S
bond in the compounds 3n–w.
potential of the model. It is calculated using the formula,
p
rm2 ðtestÞ ¼ rt2 ꢀ ð1 ꢁ rt2 ) rt20), where rt20 and rt2 are the squared
correlation coefficients between the observed and the predicted val-
ues of the test set compounds with intercept set to zero and not
set to zero, respectively (34). Other statistical parameters that
describe the quality of the QSAR which were determined were cor-
relation coefficient of the regression line (between observed versus
predicted) with activity passing through the origin (R02), correlation
coefficient of regression line (between predicted versus observed)
with activity passing through the origin (R002 ), slope of the regres-
sion line (between observed versus predicted) with activity passing
through the origin (k) and slope of the regression line (between pre-
dicted versus observed) with activity passing through the origin (k')
(34). k and k' are expected to be between 0.95 and 1.05. r2 BS is
the bootstrap r2 that is the average r2 measured with different sets
of test and training sets. r2 PRESS, also known as r2_pre-
d = 1 ) (PRESS ⁄ SD), where SD is the sum of the squared devia-
tions between the activities of the test set and the mean activity
of the training set molecules and PRESS, is the squared deviations
between predicted and actual activity values for every molecule in
the test set. q2 is the cross-validated r2 based on leave one out
method (also known as internal validation).
In the proton NMR spectral data, all protons were seen according
to the expected chemical shift and integral values. The disappear-
ance of singlet at d 8.5 p.p.m. indicated the conversion of N=CH
proton of the Schiff's base to azetidinone and thiazolidinones. Fur-
ther, the CH-Cl peak is present in the compounds 2n–w, in the
range of d 5.3–7.7 p.p.m., demonstrating the formation of 2-azetidi-
none. CH2S signal is observed between d 2.8 and 3.8 p.p.m. for the
compounds 3n–w, which confirms the presence of 4-thiazolidinone.
The NH proton of title compounds appeared in the range of d 3.8–
4.7 p.p.m. In addition, the aromatic protons appeared as multiplet
peaks within the range d 6.3–8.9 p.p.m. in all the titled compounds,
2n–w and 3n–w. The peaks appearing at d 7.6, 7.2, 7.5 and
7.7 p.p.m. confirm the presence of OH in compounds 2n, 2o, 3n
and 3o, respectively. Likewise, the appearance of singlets at d 3.9
and 4.8 p.p.m. in compounds 2p and 3p is attributed to methoxy
group. The singlets at d 2.3 and 2.8 p.p.m. in compounds 2s and
3s indicated the presence of methyl group. The dimethylamino in
compounds 2t and 3t was confirmed by singlet produced at d
2.9 p.p.m. In 13C NMR, the carbon of CH-Cl of azetidinones (2n–w)
showed a singlet varying between d 40.3 and 62.3 p.p.m. The aro-
matic carbons were observed as multiplets in the range of d
135.31–173.10 p.p.m. Similarly, in thiazolidinone series, the carbon
of CH-S showed a singlet at d 41.6–45.8 p.p.m., and the aromatic
multiplets were found at d 164.32–117.10 p.p.m. The mass spectra
showed peaks corresponding to their molecular weights.
Results and Discussion
Synthesis
Scheme 1 outlines the synthetic strategy for compounds 2n–w
and 3n–w. The preparation of quinoxaline-2,3-dione (1a) was car-
ried out as previously reported. Reaction of 1a with aniline resulted
in the formation of 3-((phenylimino)-3,4-dihydroquinoxaline)-one
(1b). The intermediate N-((E)-3-(phenylimino)-3,4-dihydro quinoxalin-
2(1H)-ylidene)benzene-1,2-diamine (1c) was achieved through reac-
tion between 1b and o-phenylenediamine. The synthesis of Schiff
bases (1n–w) produced good yields (58.00–77.00% of isolated
Antitubercular activity and structure–activity
relationship
The MIC was determined for compounds 2n–w and 3n–w against
the M. tuberculosis strain H37Rv using the micro plate Alamar Blue
assay.
Chem Biol Drug Des 2011; 78: 988–998
995

Puratchikody et al.
The results of the antitubercular activity are presented in Table 1.
All the synthesized compounds exhibited an interesting activity pro-
file against the strain. The results revealed that the activity is con-
siderably affected by various substituents on the aromatic ring of
either 2-azetidinone or 4-thiazolidinone nucleus. The introduction of
a hydroxyl, methoxy, chloro, dimethylamino and nitro group on aro-
matic ring (2o, 3o, 2p, 3p, 2q, 3q, 2t, 3t, 2r, 3r, 2u and 3u)
resulted in compounds with an enhanced antitubercular activity
(MIC values ranging from 0.67 to 3.70 lg ⁄ mL). Amongst them, com-
pounds 2t, 3t, 2r, 3r and 2u (MIC values ranging from 0.67 to
0.97 lg ⁄ mL) exhibited significant activity when compared with the
first-line drug, isonictinyl acid hydrazide (INH) (MIC = 0.47 lg ⁄ mL).
This antitubercular activity may be attributable to the introduction
of electron withdrawing group on the aromatic ring. However, sub-
stitution by electron releasing methoxy group also resulted in com-
pounds with moderate antitubercular activity (2p and 3p). It is
interesting to note that the introduction of an electron-releasing
hydroxyl group in the p-position of phenyl ring resulted in moderate
activity (2o and 3o), whereas substitution of hydroxyl group in the
o-position is found to have complete loss of activity (2n and 3n).
It has been observed that the replacement of hydroxyl moiety by
electron releasing methyl group in the compounds 2s and 3s
showed mild activity. The compounds with electron withdrawing
substituents (chloro, dimethylamino and nitro) were found to be
more active than electron releasing methoxy, hydroxyl and methyl
moiety. Literature survey reveals that electron withdrawing or
donating groups amend the lipophilicity of the test compounds,
which in turn alters permeability across the bacterial cell membrane
(36). Further, compounds 2v and 3v substituted with a five mem-
bered furan structure did not show any considerable activity. Simi-
larly, introduction of methoxy and hydroxyl groups in compounds
2w and 3w leads to complete activity loss. The obtained results
revealed that the nature of the substituent and substitution pattern
on the benzene ring may have a considerable impact on the antitu-
bercular activity of the synthesized compounds. The influence of the
2-azetidinone nucleus in compounds 2n–o and 4-thiazolidinone
nucleus in compounds 3n–o has been studied on the biological
activity. The replacement in the core nucleus did not alter the antit-
ubercular activity to a greater extent.
Figure 1: Stereo view of the molecular rectangular field grid
around the super-imposed quinoxiline with the location of the elec-
trostatic and steric descriptors used in the three-dimensional quan-
titative structure–activity relationship (dimensions of the box
19.9*20.9*19.5ꢀ).
1.0
0.5
0.0
–0.5
–1.0
–1.5
–2.0
–2.5
–2.0
–1.5
–1.0
–0.5
0.0
0.5
Observed -log(MIC)
Figure 2: Comparison of observed and predicted activities of the
compounds in the training and the test sets (with 95% prediction
interval) from template based alignment (closed and open cir-
cles = Training and test set compounds respectively).
The best 3D-QSAR model developed with the training set (n = 16)
and other statistics are given below.
)log (MIC) = 0.0201 ) 0.1857*E_412 ) 2.0065*S_366
r2 = 0.81, q2 = 0.71, r2 PRESS (r2_pred = 0.84, F = 27.06, rm2 = 0.84,
r2 BS = 0.80, k = 0.906, k'=0.995, R02 = 0.901, R020 = 0.901. All the
statistical parameters are good indicating that the model has good
predictive capability.
Figure 1 shows the stereo view of the rectangular field grid around
all the molecules, super-imposed (stacked) on the common template,
quinoxiline. The location of the electrostatic and steric descriptors
selected in the model is also shown in the figure.
E_412 and S_366 are the electrostatic and steric descriptors,
respectively. 59.8% of the contribution is from the former and the
rest from the latter descriptor. The electrostatic descriptor is nega-
tively correlated with activity indicating that electronegative groups
would increase activity. The steric descriptor is negatively correlated
with activity indicating less bulky groups would enhance activity.
Figure 2 shows the comparison of the observed and predicted (from
the 3D-QSAR) activities of the compounds in the training and the
test set (with the 95% prediction interval) from template-based
alignment. All the predictions lie within this prediction band. Similar
descriptors have been identified while studying the antiTB activity
of thiazolines and thiazolidinones (20).
996
Chem Biol Drug Des 2011; 78: 988–998

Antitubercular Activity of Quinoxalines
Table 2: Values of the descriptors selected in the model and
comparison of experimental and predicted activities
trostatic field regions were found to be important for M. tuberculo-
sis inhibitory activity. From the detailed analysis of the results of
above studies, it is concluded that antitubercular activity of the syn-
thesized compounds significantly depends on the presence of elec-
tronegative group. The results obtained from the 3D-QSAR can be
used to predict antitubercular activity of new compounds and also
to design novel quinoxaline derivatives with enhanced performance.
The current findings can help chemists and pharmacists for further
investigations in this field in search of potent antitubercular agents.
Experimental Predicted
activity
activity
Compound E_412
Training set
S_366
)log lM
)log lM
Residuals
2n
2p
2q
2s
2u
2v
2w
3n
3o
3p
3q
3r
10.0000 )0.1566 )1.6989
5.5800 )0.4293 )0.1847
4.0096 )0.4974 )0.1461
)1.5231
)0.1549
0.2734
)0.1758
)0.0298
)0.4195
)0.1802
0.4603
)0.1341
)0.3689
0.2987
)0.3146
0.1515
0.1344
0.1478
0.1041
5.1682 0.0966
4.2567 )0.1612 0.0132
3.9367 0.0793 )1.0043
)1.3139
)1.1337
)0.4471
)0.8702
)0.7029
)1.4386
)0.2536
)0.2914
)0.2743
)0.0286
)1.3829
)0.3690
)0.2031
)1.3839
Acknowledgments
6.0193 )0.1968 )1.0718
10.0000 )0.1986 )1.1399
4.2837 )0.2601 )0.5682
4.3829 )0.2504 )0.1399
4.5296 )0.2725 )0.1399
3.5086 )0.3005 0.1192
9.4975 )0.1799 )1.2788
4.8044 )0.2508 0.1739
3.2662 )0.1911 )0.2480
9.0376 )0.1368 )1.5560
The authors acknowledge SAIF, IIT, and Chennai for providing 1H
NMR and Mass spectra and BIF, Department of Biotechnology, IIT-
Madras for computational resources.
3s
3t
References
0.5429
)0.0449
)0.1721
3u
3v
1. WHO (2003) WHO Annual Report on Global TB Control Sum-
mary. Wkly Epidemiol Rec;78:121–128.
Test set
2o
2. De Cock K.M., Chaisson R.E. (1999) Will DOTS do it? A reap-
praisal of tuberculosis control in countries with high rates of
HIV infection. Int J Tuberc Lung Dis;3:457–465.
3. Gupta R.A., Gupta A.K., Soni L.K., Kaskhedikar S.G. (2007)
Rationalization of physicochemical characters of oxazolyl
thiosemicarbazone analogs towards multi-drug resistant
tuberculosis: a QSAR approach. Eur J Med Chem;42:1109–1116.
4. Janin Y.L. (2007) Antituberculosis drugs: ten years of research.
Bioorg Med Chem;15:2479–2513.
5. Esther V., Silvia P.S., Lidia M.L., Saioa A., Asunciꢁn B., Beatriz
S., Raquel V., Ignacio A., Antonio M. (2009) Selective activity
against Mycobacterium tuberculosis of new quinoxaline 1,4-di-
N-oxides. Bioorg Med Chem;17:385–389.
6. Esther V., Pablo R.D., Eduardo A.C., Antonio M. (2009) QSAR
analysis for quinoxaline-2-carboxylate 1,4-di-N-oxides as anti-
mycobacterial agents. J Mol Graph Model;28:28–36.
5.8931 )0.3923 )0.4771
3.1354 )0.5337 0.0506
5.1974 )0.4236 0.1191
10.0000 )0.1898 )1.5440
)0.2874
0.5087
)0.0953
)1.4563
)0.1897
)0.4581
0.2144
2r
2t
3w
)0.0877
Table 2 gives the values of the two descriptors, experimental and
predicted activities (in )log lM) and the residual error for the
twenty compounds.
The antitubercular activity is positively correlated (r = 0.47,
p < 0.05) with AlogP. ALogP is a measure of the hydrophobic –
hydrophilic balance of the compounds (37). This indicates that
hydrophobic quinoxiline derivatives would exhibit higher antitubercu-
lar activity. But this descriptor does not appear in the QSAR
because the other two properties (electrostatic and steric)
contribute much more to equation. Vincente et al. (38) also did not
observe lipophilicity in their QSAR while exploring 3-aryl quinoxa-
lines as antiTB compounds. The importance of hydrophobicity on
the antiTB activity of chalcones has been reported (22).
7. Elsa M., Saioa A., Silvia P., Enrique T., Ignacio A., Antonio M.
(2010) Synthesis and antimycobacterial activity of new quinoxa-
line-2-carboxamide 1,4-di-N-oxide derivatives. Eur
Chem;45:4418–4426.
J Med
8. Carta A., Corona P., Loriga M. (2005) Quinoxaline 1, 4-dioxide: a
versatile scaffold endowed with manifold activities. Curr Med
Chem;12:2259–2272.
9. Goldman R.C., Laughon B.E., Reynolds R.C., Secrist J.A., Maddry
J.A., Guie M.A., Poffenberger A.C., Kwong C.A., Ananthan S.
(2007) Programs to facilitate tuberculosis drug discovery: the
tuberculosis antimicrobial acquisition and coordinating facility.
Infect Disord Drug Targets;7:92–104.
10. Lee J.S., Waring M. J. (1978) Interaction between synthetic
analogues of quinoxaline antibiotics and nucleic acids. Changes
in mechanism and specificity related to structural alterations.
Biochem J;173:129–144.
11. Bailey C., Michael J.W. (1998) DNA recognition by quinoxaline
antibiotics: use of base-modified DNA molecules to investigate
determinants of sequence-specific binding of triostin and Tan-
dam. Biochem J;330:8187.
Conclusion
Synthesis, spectral characterization, antitubercular activity and 3D-
QSAR analysis of a new series of 2-azetidinone and 4-thiazolidinone
derivatives of quinoxalines are reported here. The various 2-azetidi-
none derivatives were synthesized by cycloaddition with chloroace-
tyl chloride in the presence of triethylamine while the 4-
thiazolidinone derivatives of quinoxalines were synthesized by cyclo-
condensation of substituted arylidene derivatives of quinoxaline-2,3-
dione with thioglycollic acid. The in vitro antitubercular screening
results of the title compounds evidenced that compounds 2t, 3t,
2r, 3r and 2u may be considered promising for the development
of new antitubercular agents. 3D-QSAR results revealed that elec-
Chem Biol Drug Des 2011; 78: 988–998
997

Puratchikody et al.
12. Ramalingam P., Ganapaty S., Babu R.C. (2010) In vitro
antitubercular and antimicrobial activities of 1-substituted
quinoxaline-2,3(1H,4H)-diones. Bioorg Med Chem Lett;20:406–
408.
2 ⁄ 4-quinolinecarbaldehyde
Chem;14:7302–7310.
derivatives.
Bioorg
Med
26. Gris J., Glisoni R., Fabian L., Fernandez B, Moglioni A.G. (2008)
Synthesis of potential chemotherapic quinoxalinone derivatives
by biocatalysis or microwave-assisted Hinserg reaction. Tetrahe-
dron Lett;49:1053–1056.
27. Alistair K., Sudharsan S., Laurent K., Sandra L., Lynn G., James
C., Gurdyal S. (2005) Probing the mechanism of the Mycobacte-
rium tuberculosis b-Ketoacyl-Acyl carrier protein synthase III
mtfabh: factors influencing catalysis and substrate specificity. J
Biol Chem;280:32539–32547.
28. Raparti V., Chitre T., Bothara K., Kumar V., Dangrea S., Khacha-
nea C., Gorea S., Deshmanea B. (2009) Novel 4-(morpholin-4-yl)-
N'-(arylidene)benzohydrazides: synthesis, antimycobacterial activ-
ity and QSAR investigations. Eur J Med Chem;44:3954–3960.
29. Fegade J.D., Rane R.Y., Adhari V., Patil R. (2009) 3D-QSAR study
of benzylidene derivatives as selective cyclooxygenase-2-inhibi-
tors. DJNB;4:145–154.
30. Manvar A., Malde A., Verma I., Virsodiac V., Mishra A., Upadhy-
ay K., Acharyaa H., Coutinhob E., Shahc A. (2008) Synthesis,
anti-tubercular activity and 3D-QSAR study of coumarin-4-acetic
acid benzylidene hydrazides. Eur J Med Chem;43:2395–2403.
31. Sivakumar P.M., Prabakar P.K., Doble M. (2010) Synthesis, anti-
oxidant evaluation and QSAR studies of chalcones. Med Chem
Res;20:482–492.
13. Sharma M.C., Sahu N.K., Kohli D.V., Chaturvedi S.C., Smita S.
(2009) Synthesis, characterization and biological activities of
some 1-(nicotinylamino)-2 substituted azetidine-4-ones as poten-
tial antibacterial agents. Dig J Nanomat Bios;4:361–367.
14. Girija S.S., Boycie J.M. (2005) Synthesis of 2-azetidinones from
2-diazo-1, 2-diarylethanones and N-(2-thienylidene)imines as
possible antimicrobial agents. Farmaco;60:727–730.
15. Rajasekaran A., Periasamy M., Venkatesan S. (2010) Synthesis,
characterization and biological activity of some novel azetidinon-
es. J Dev Biol Tissue Eng;2:5–13.
16. Aliasghar J., Maaroof Z. (2006) Synthesis of novel N-sulfonyl
monocyclic b-lactams as potential antibacterial agents. Mole-
cules;11:49–58.
17. Vijaykumar M.M., Nagaraja T.S., Shameer H., Jayachandran E.,
Sreenivasa G.M. (2009) N-substituted-3-chloro-2-azetidinones:
synthesis and characterization of new novel anti-inflammatory
agents. J Pharm Sci Res;1:83–92.
18. Sandip J., Mahesh P., Veeresh M., Pradeepkumar R, Shivaling-
arao M., Satyanarayana D., Mangala G. (2009) Synthesis
and antimycobacterial activity of a novel series of isonicotinyl
hydrazide derivatives. Arch Pharm Chem Life Sci;342:723–
731.
32. Puratchi A., Doble M. (2009) QSAR studies on antiepileptic and
locomotor in vivo activities of 4,5-diphenyl-1H-imidazoles. Chem
Biol Drug Des;74:173–182.
19. Amit V., Shailendra K.S. (2008) 4-Thiazolidinone – a biologically
active scaffold. Eur J Med Chem;43:897–905.
20. Sivakumar P.M., Geetha Babu S.K., Doble M. (2008) Impact of
topological and electronic descriptors in the QSAR of pyrazine
containing thiazolines and thiazolidinones as antitubercular and
antibacterial agents. Chem Biol Drug Des;71:447–463.
21. Sivakumar P.M., Geetha Babu S.K., Doble M. (2007) QSAR stud-
ies on chalcones and flavonoids as antituberculosis agents using
genetic function approximation (GFA) method. Chem Pharm
Bull;55:44–49.
22. Sivakumar P.M., Doble M., Prabu S.S., Vanaja K. (2007)
Synthesis, antimycobacterial activity evaluation and QSAR stud-
ies of chalcone derivatives. Bioorg Med Chem Lett;17:1695–
1700.
23. Sivakumar P.M., Muthukumar T., Doble M. (2009) Antifungal
activity, mechanism and QSAR studies on chalcones. Chem Biol
Drug Des;74:68–79.
24. Sivakumar P.M., Sobana P., Doble M. (2009) Synthesis, biological
evaluation, mechanism of action and QSAR studies on chalcones
and chalcones as antibacterial agents. Chem Biol Drug
Des;73:403–415.
33. Puratchi A., Doble M. (2007) Antinociceptive and anti-inflamma-
tory and QSAR studies on 2-substituted 4,5-diphenyl-1H-imidaz-
oles. Bioorg Med Chem;15:1083–1090.
34. Aparna V., Jeevan J., Ravi M., Desiraju G.R., Gopalakrishnan B.
(2006) 3D-QSAR studies on antitubercular thymidine monophos-
phate kinase inhibitors based on different alignment methods.
Bioorg Med Chem Lett;16:1014–1020.
35. Silverstein R.M., Bassler G.C., Morrill T.C. (1981) Spectrometric
Identification of Organic Compounds, 4th edn. New York, USA:
John Wiley and Sons;123 p.
36. Bhusari K.P., Amnerkar N.D, Khedekar P.B, Kale M.K, Bhole R.P.
(2008) Synthesis and in vitro antimicrobial activity of some new
4-amino-N-(1,3-Benzothiazol-2-yl) benzenesulphonamide deriva-
tives. Asian J Research Chem;1:53–58.
37. Thompson S.J., Hattotuwagama C.K., Holliday J.D., Flower D.R.
(2006) On the hydrophobicity of peptides: comparing empirical
predictions of peptide log P values. Bioinformation;1:237–241.
38. Vincente E., Duchowicz P.R., Ortiz E.V., Monge A., Castro E.A.
(2010) Exploring 3-aryl quinoxaline-2-carbonitrile-1,4-di-N-oxides
activities against neglected disease with QSAR. Chem Biol Drug
Des;76:59–69.
25. Nayyar A., Malde A., Coutinho E., Jain R. (2006) Synthesis,
anti-tuberculosis activity, and 3D-QSAR study of ring-substituted-
998
Chem Biol Drug Des 2011; 78: 988–998