Chromanyl-isoxazolidines as Antibacterial agents: Synthesis, Biological Evaluation, Quantitative Structure Activity Relationship, and Molecular Docking Studies

Article abstract of DOI:10.1111/cbdd.12653

Regio- and stereoselective 1,3-dipolar cycloadditions of C-(chrom-4-one-3-yl)-N-phenylnitrones (N) with different mono-substituted, disubstituted, and cyclic dipolarophiles were carried out to obtain substituted N-phenyl-3′-(chrom-4-one-3-yl)-isoxazolidines (1-40). All the synthesized compounds were assayed for their in vitro antibacterial activity and display significant inhibitory potential; in particular, compound 32 exhibited good inhibitory activity against Salmonella typhymurium-1 & Salmonella typhymurium-2 with minimum inhibitory concentration value of 1.56 μg/mL and also showed good potential against methicillin-resistant Staphylococcus aureus with minimum inhibitory concentration 3.12 μg/mL. Quantitative structure activity relationship investigations with stepwise multiple linear regression analysis and docking simulation studies have been performed for validation of the observed antibacterial potential of the investigated compounds for determination of the most important parameters regulating antibacterial activities.

Full text of DOI:10.1111/cbdd.12653

Chem Biol Drug Des 2016; 87: 213–223
Research Article
Chromanyl–isoxazolidines as Antibacterial agents:
Synthesis, Biological Evaluation, Quantitative Structure
Activity Relationship, and Molecular Docking Studies
Gagandeep Singh1, Anuradha Sharma2, Harpreet
USA and is considered to be a growing threat to commu-
nity health (2). The availability of several classes of antibi-
Kaur3 and Mohan Paul S. Ishar1,*
otics such as oxazolidinones (linezolid) and quinolones
(ciprofloxacin) could fill this need; however, most clinical
1Bio-Organic and Photochemistry Laboratory, Department
of Pharmaceutical Sciences, Guru Nanak Dev University,
Amritsar 143 005, Punjab, India
MRSA isolates are sufficiently resistant against these
classes, rendering them ineffective in successfully curing
many infections (3–14). Hence, the discovery and develop-
ment of novel antimicrobial agents with optimized pharma-
cological profile together with increased activity towards
resistant strains and with a mode of action different from
existing drugs is of considerable interest.
2University Institute of Pharmaceutical Sciences, Panjab
University, Chandigarh 160014, India
3Department of Microbiology, Guru Nanak Dev University,
Amritsar 143 005, Punjab, India
*Corresponding author: Mohan Paul S. Ishar,
mpsishar@yahoo.com
It is well known that introduction of fluorine atom into bio-
logically active substances may lead to improved meta-
bolic stability and consequent pharmacological properties
and an increase in therapeutic effect (3–13). Therefore,
fluorine atom at 6th position in the fluoroquinolone struc-
ture has invariably been associated with improved
potency of a range of fluoroquinolones for bacterial infec-
tions and enhances the binding affinity to biological tar-
gets (14–18). The bio-isosteric equivalence of chromone
nucleus and quinolone moiety has been established
(19,20). Several fluorine-substituted chromones have been
proven to be effective against a variety of biological tar-
gets (Figure 1) (3–13,16).
Regio- and stereoselective 1,3-dipolar cycloadditions
of C-(chrom-4-one-3-yl)-N-phenylnitrones (N) with dif-
ferent mono-substituted, disubstituted, and cyclic
dipolarophiles were carried out to obtain substituted
N-phenyl-30-(chrom-4-one-3-yl)-isoxazolidines
(1-40).
All the synthesized compounds were assayed for their
in vitro antibacterial activity and display significant
inhibitory potential; in particular, compound 32 exhib-
ited good inhibitory activity against Salmonella typhy-
murium-1 & Salmonella typhymurium-2 with minimum
inhibitory concentration value of 1.56 lg/mL and also
showed good potential against methicillin-resistant
Staphylococcus aureus with minimum inhibitory con-
centration 3.12 lg/mL. Quantitative structure activity
relationship investigations with stepwise multiple linear
regression analysis and docking simulation studies
have been performed for validation of the observed
antibacterial potential of the investigated compounds
for determination of the most important parameters
regulating antibacterial activities.
Chromone derivatives possessing heterocyclic substituents
at 2 and 3 positions have been reported to possess anti-
allergic activity, muscular relaxation effect, and (3–13,16)
antimicrobial activity. Recently, we had also reported chro-
manyl–isoxazolidines as apoptosis inducers through the
mitochondrial-dependent pathway in HL-60 cells (21).
Key words: 1,3-dipolar cycloaddition, 3-chromanyl–isoxazo-
lidines, ADME properties, antibacterial activity, quantitative
structure activity relationship
Based on the above reports on the therapeutic importance
of substituted chromones and antimicrobial potential of
isoxazolidines (22,23), herein we report the synthesis and
antibacterial evaluation of some new derivatives of chro-
manyl–isoxazolidines for their antibacterial potential. Quan-
titative structure activity relationship (QSAR) evaluations
have been performed to develop the predication models
for the identification of lead. The lead molecule was further
evaluated using molecular docking studies for better
understanding of the crucial interactions of the ligand with
the enzyme. This would greatly help to elucidate the prob-
able mode of action and further drug designing/develop-
ment.
Received 1 May 2015, revised 27 July 2015 and accepted for
publication 30 July 2015
Despite widespread distribution of methicillin-resistant Sta-
phylococcus aureus (MRSA), it was initially not considered
a major threat due to the availability of vancomycin-related
antibiotics that were effective for nearly two decades
against serious MRSA infections (1). However, nowadays
MRSA has become a serious problem, as the annual
mortality attributed to MRSA exceeds that of AIDS in the
ª 2015 John Wiley & Sons A/S. doi: 10.1111/cbdd.12653
213

Singh et al.
raphy using 60–120 mesh silica and 1–8% EtOAc in
hexane as eluent. The reactions with disubstituted/cyclic
dipolarophile were carried out under the same reaction
conditions and products have been isolated by column
chromatography. The reported yields were based on iso-
lated pure products and products were characterized by
spectroscopic techniques (1H NMR, 13C NMR, IR, and
Mass).
50-acetoxy-20-phenyl-30-(6,8-dichloro-chrom-4-one-
3yl)-isoxazolidine (11)
Off white semi-solid (86%), 1H NMR (300 MHz, CDCl3): d
8.35 (s, 1H, C2H), 7.96 (bs, 1H, C5H), 7.37 (dd,,1H,
J = 8.5, 5.4 Hz, C7H), 7.28 (d, 1H, J = 8.5 Hz, C8H),
7.13-6.81 (m, 5H, Ar-H), 4.77 (dd, 1H, J = 10.2, 5.4 Hz,
Figure 1: Examples of fluorine-substituted antimicrobial agents.
0
Methods and Materials
C5ꞌH), 4.42 (dd, 1H, J = 7.8, 2.5 Hz, C3 H), 3.13 (dd, 1H,
J = 8.7, 6.3 Hz, OCH2), 2.91 (ddd, 1H, Jgem = 12.8 &
0
Chemicals
J = 10.2, 7.8 Hz, C4 Ha), 2.62 (ddd, 1H, Jgem = 12.8 &
J = 5.4, 2.5 Hz, C4 Hb); 13C NMR (75 MHz, CDCl3): d
0
Starting materials and reagents were purchased from
commercial suppliers and used after further purification
(crystallization/ distillation). H- and 13C NMR spectra were
175.1, 170.7, 167.8, 149.8, 147.5, 134.3, 129.4, 128.6,
124.2, 124.0, 122.5, 116.3, 114.6, 75.6, 62.9, 52.0, 39.9;
HRMS (ESI, m/z): calculated for C20H15Cl2NO5 [M + H]+,
420.0400, found 420.0397.
1
recorded at room temperature (30 °C) either on a JEOL
(300/75 MHz) NMR spectrometer or a Bruker Advance
500 MHz (500/125 MHz) spectrometer, and chemical
shifts (d) are reported in ppm from tetramethylsilane (TMS)
used as the internal standard and coupling constant (J)
values in Hertz. IR spectra were recorded on Shimadzu
8400S FT-IR spectrophotometer (KBr, cmꢀ1). Mass spec-
tra, EI, and ESI methods were recorded on Shimadzu
GCMS-QP-2000A, Bruker Daltonics Esquire 300 mass
spectrometer, and Bruker micrOTOF Q II Mass spectrome-
ter (LC MS/MS). All melting points are uncorrected and
measured in open glass-capillaries using Veego Precision
Digital Melting Point Apparatus.
The compounds (1–40) were synthesized following the
general procedure as described above with the yield of
75–90%.
Biological activity evaluation
In vitro antibacterial studies of all the synthesized com-
pounds (1–40) were carried out against Gram-negative
bacterial strains such as Salmonella typhymurium-1
(MTCC-1251) & Salmonella typhymurium-2 (MTCC 98)
and MRSA (Gram-positive bacterial strain) by disk-diffusion
assay. Standard bacterial strains were obtained from
Microbial Type Culture Collection (MTCC), Institute of
Microbial Technology (IMTECH), Chandigarh, and the clini-
cal isolate methicillin-resistant Staphylococcus aureus
(MRSA) was obtained from Postgraduate Institute of Medi-
cal Education and Research (PGIMER), Chandigarh, India.
The activity of compounds was determined in comparison
to standard antibiotic discs of gentamicin in triplicate. Pre-
warmed Mueller-Hinton agar plates were inoculated with
106 CFU/mL of test bacteria. Each compound was dis-
solved in DMSO (1 mg/ml) and then 30 lL of each was
pipetted onto sterile paper discs (6 mm diameter) placed
on the surface of inoculated agar plates. Plates were incu-
bated at 37 °C for 24 h. Activity was expressed as the
diameter of the inhibition zone (mm) produced by the
compounds. DMSO was used as negative control. MIC of
compounds exhibiting considerable activity was evaluated.
The initial optical density (OD) of the medium was mea-
sured by spectrophotometer at 600 nm. The test strains
were incubated in nutrient broth until the OD reached 0.4–
0.6. Then the different concentrations of compounds
(0.78, 1.56, 3.12, 6.25, 12.5, 25, 50 lg/mL) were tested
General method for synthesis of C-(chromon-3-yl)-
N-phenylnitrone (N)
Substituted 3-formylchromone (3.0 g, 2.8 mmol) was dis-
solved in dry benzene (30 mL), and to the clear solution,
N-phenyl-hydroxylamine (4.08 g, 2.8 mmol) was added
and the contents were slightly heated and allowed to
stand at room temperature. After 30 min, N-phenylnitrone
(N) separated out as a light yellow solid, which was fil-
tered. It was used immediately to prevent its rearrange-
ment.
General procedure for reactions of N-
phenylnitrone (5) with various dipolarophiles
Reactions of N-phenylnitrones with monosubstituted dipo-
larophiles were carried out by mixing N-phenylnitrone (N,
0.75 mmol) with dipolarophile (2.25 mmol) in dry CH2Cl2
(40 mL), and the reaction mixture was stirred under the
exclusion of moisture, until all the N-phenylnitrone was
consumed (TLC). The solvent was removed under vac-
uum, and the residue was purified by column chromatog-
214
Chem Biol Drug Des 2016; 87: 213–223

Chromanyl-isoxazolidines as Antibacterial Agents
for the inhibition of growth of these microbes, in separate
tubes. The 10-mL tubes each containing 5 mL of nutrient
broth and 1 mL of different concentrations of compounds
were incubated for 24 h with shaking at 180 rpm using a
rotary shaker. Each tube corresponding to different con-
centrations was observed and concentration showing
apparently no turbidity was considered to be the MIC of
the respective compound.
DNA gyrase. All the ligand molecules were designed and
the structures were analyzed using ChemDraw Ultra 10 3D.
The minimization of ligands and protein was performed by
universal force field and amber force field respectively.
Binding site was defined by the selecting the active amino
acids of the enzyme structure and the grid with its dimen-
sions (51 9 53 9 46) were defined.
Physicochemical and ADME properties
Quantitative structure–activity relationship
ADME properties were calculated using QIKPROP v3. 6 tool
of Schrodinger Software. It predicts both physicochemi-
cally significant descriptors and pharmacokinetically signifi-
cant properties. QIKPROP provides ranges for comparing a
exacting molecule’s properties with those of 95% of
known drugs. QIKPROP also flags 30 types of reactive func-
tional groups that may cause false positives in high
throughput screening (HTS) assays. It also evaluates the
suitability of analogs based on Lipinski’s rule of five, which
is essential to ensure drug like pharmacokinetic profile
while using rational drug design. All the analogs were neu-
tralized before being used by QIKPROP.
QSAR analysis applies statistical methods to describe the
relationship between chemical structure and biological
activities of a series of chromanyl isoxazolidines quantita-
tively. The groups of calculated thermodynamic descriptors
included bend energy, heat of formation, torsion energy,
boiling point, melting point, Gibbs free energy, Henry’s law
constant, ideal gas thermal capacity, exact mass, and
molecular weight. Steric descriptors derived were steric
energy, connolly accessible area, connolly molecular area,
connolly solvent excluded volume, molar refractivity, and
ovality; apart from this partition coefficient calculated as
LogP. Electronic descriptors include kinetic energy, poten-
tial energy, and total energy. Molecular topology descrip-
tors included balaban index, cluster count, molecular
topological index, num rotatable bonds, polar surface area,
radius, shape attribute, sum of degrees, sum of valence
degrees, topological diameter, total connectivity, total
valence connectivity, and Wiener index.
Results and Discussion
Chemistry
The C-(chrom-4-one-3-yl)-N-phenylnitrones (N) were pre-
pared in good yields by reacting various substituted 3-for-
mylchromones and N-phenylhydroxylamine (1 molar
equivalent) in dry benzene with constant stirring, followed
by removal of the solvent under vacuum (24). The synthe-
sized nitrones were reacted with various monosubstituted
dipolarophiles by stirring in dry dichloromethane at room
temperature leading to isoxazolidines (1-29, Scheme 1)
(21,24).
Molecular docking study
The comparative and automated docking studies were per-
formed using GRID-based ligand docking with energetics
(GLIDE) program (version 6.3, Schrodinger, LLC, New York,
NY, USA) to determine the best in silico conformation of
synthesized molecule inside the active site. The standard
drugs such as ciprofloxacin and linezolid were drawn and
docked back into the protein in order to determine the relia-
bility of the docking method. The crystal structure of the
topo II DNA gyrase with its co-crystal ciprofloxacine (PDB
ID: 2XCT) was selected as target, to understand the nature
of the interaction between the most potent molecule and
Further, the synthesized nitrones (N) were reacted with
disubstituted and cyclic dipolarophiles, under similar set of
conditions to obtain isoxazolidines (30-40) (21). All the
products were separated and purified by column chro-
matography and characterized by modern spectroscopic
1
techniques, that is, H & 13C NMR and Mass (Scheme 2).
Scheme 1: Reaction of various nitrones (N) with monosubstituted dipolarophiles leads to chromanyl-isoxazolidines (1–29).
Chem Biol Drug Des 2016; 87: 213–223
215

Singh et al.
Scheme 2: Reaction of various nitrones (N) with disubstituted and cyclic dipolarophile.
Antibacterial activity
tuted and bicyclic isoxazolidines displayed much higher
In vitro antibacterial studies of all the synthesized com-
pounds (1-40) were carried out against Gram-negative
bacterial strains such as Salmonella typhymurium-1
(MTCC-1251) & Salmonella typhymurium-2 (MTCC 98)
and Gram-positive strain such as MRSA bacterial strain by
disk-diffusion assay (25). The activity of compounds was
determined in comparison with standard antibiotic disks of
Gentamycin. Minimum inhibitory concentration (MIC) of
compounds exhibiting considerable activity (Table 1) was
determined using serial dilution method (26). The initial
optical density (OD) of the medium was measured by
spectrophotometer at 600 nm.
inhibitory potential against S. typhi 1 & S. typhi 2 than
positive control, that is, Gentamycin. In particular, com-
pound 32 exhibited good inhibitory activity against S. typhi
1 & S. typhi 2 with MIC value of 1.56 lg/mL and also
showed good potential against MRSA with MIC of
3.12 lg/mL.
In the case of monosubstituted isoxazolidines, compounds
bearing non-polar hydrophobic groups at C50 show good
to moderate antibacterial activity. For instance compounds
such as 11, 15, 18, and 22 display significant antibacterial
activities with MIC value of 6.25, 3.12, 3.12, and 6.25 lg/
mL, respectively, against S. typhi 1. Similarly, compounds
3, 6, and 25 bearing ethoxy, isobutoxy, and phenyl at C50
and fluoro substitution at C6 showed good inhibitory activ-
ity with MIC value of 12.5, 12.5, and 3.12 lg/mL, respec-
tively, against S. typhi 1. The inhibitory potential difference
in compound 3, 6, and 25 can be attributed to the differ-
ence in the hydrophobic character of ethoxy/isobutoxy/
All the synthesized chromanyl–isoxazolidine derivatives
exhibit good to moderate antibacterial activity with observ-
able variations due to different substitutions at C50 and
C6. The obtained results revealed that some of the com-
pounds possess excellent antibacterial activity against
selected strains. Among the evaluated molecules, disubsti-
216
Chem Biol Drug Des 2016; 87: 213–223

Chromanyl-isoxazolidines as Antibacterial Agents
Table 1: In vitro antibacterial activity of
synthesized compounds 1–40 against vari-
ous bacterial strains
Nitrone
MIC (lg/mL)
Salmonella
Salmonella
Comp. No. X1
X2 Dipolarophile
typhymurium-1 typhymurium-2 MRSA
1
2
3
4
5
6
7
8
H
CH3
F
Br
Cl
F
Br
CH3
H
Cl
Cl
H
Cl
H
Cl
H
Cl
F
CH3
H
Cl
F
CH3
H
F
Br
Cl
CH3
H
H
Cl
F
H
H
H
H
D; R = -OEt
R = -OEt
R = -OEt
R = -OEt
25
25
12.5
25
25
12.5
25
25
25
12.5
6.25
12.5
6.25
12.5
3.12
12.5
6.25
3.12
25
25
12.5
6.25
25
25
3.12
12.5
6.25
25
6.25
12.5
6.25
1.56
6.25
12.5
6.25
25
25
12.5
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
12.5
25
25
25
12.5
25
25
25
25
25
25
25
25
25
Cl R = -OEt
H
H
H
H
H
R = -OiBu
R = -OiBu
R = -OiBu
R = -CO2Me
R = -CO2Me
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Gentamycin
12.5
Cl R = -CO2Me
6.25
12.5
6.25
12.5
12.5
12.5
12.5
6.25
25
25
12.5
6.25
25
25
6.25
12.5
6.25
25
6.25
12.5
6.25
1.56
6.25
12.5
12.5
25
25
H
H
H
R = -CONH2
R = -CN
R = -CH2OH
Cl R = -CH2OH
H
H
H
H
H
H
H
H
H
H
H
R = -CO2Et
R = -CO2Et
R = -CO2Et
R = -CO2Et
R = -CO2Bu
R = -CO2Bu
R = -CO2Bu
R = -CO2Bu
R = -Ph
R = -Ph
R = -Ph
Cl R = -Ph
H
H
H
H
H
R = -Ph
R = -Pyridyl
a-Me-Styrene
a-Me-Styrene
a-Me-Styrene
3.12
Cl
CH3
F
Br
CH3
F
Cl a-Me-Styrene
25
25
25
25
25
25
25
25
H
H
H
H
H
H
H
a-Me-Styrene
Methyl methacrylate
Methyl methacrylate 25
Methyl methacrylate 25
N-Ph-maleimide
N-Ph-maleimide
N-Ph-maleimide
3.12
25
25
6.25
25
25
Br
CH3
1.56
1.56
0.39
MIC, minimum inhibitory concentration; MRSA, methicillin-resistant Staphylococcus
aureus.
phenyl at C50. Compound 25 also displayed significant
antibacterial activity against S. typhi 2 with MIC value of
6.25 lg/mL. Both strains S. typhi 1 & S. typhi 2 displayed
sensitivity against compound 29 bearing pyridyl at C50 with
MIC value 6.25 lg/mL (Table 1).
bioavailability (17,18). Thus, difference in inhibitory potential
of compounds 32 and 31 may be attributed to the pres-
ence of fluoro- and chloro- groups at C6. Compound 32,
a fluorinated analog also displays significant inhibitory
potential against MRSA with MIC value of 3.12 lg/mL.
Among bicyclic isoxazolidines, compound 38 showed
good inhibition against S. typhi 1 (MIC = 3.12 lg/mL).
Furthermore, in the case of disubstituted and bicyclic isox-
azolidines, compounds 31 and 32 showed good antibac-
terial activity against S. typhi 1 and S. typhi 2 with MIC
value of 6.25 and 1.56 lg/mL, respectively. It is widely
accepted that one or just a few atoms in an organic mole-
cule can dramatically alter its chemical and biological nat-
ure, including its potency, stability, lipophilicity, and
Among the various halogens at C6, fluoro substitution
shows distinct enhancement in the activity against all the
tested bacterial strains. Thus, all fluoro-substituted com-
pounds such as 3, 6, 18, 22, 25, 32, 35, and 38 exhibited
significant antibacterial activities against S. typhi 1, S. typhi
Chem Biol Drug Des 2016; 87: 213–223
217

Singh et al.
2, and MRSA strains in comparison with other analogs.
This influence of flouro- on enhancement of antibacterial
activity is reminiscent of significance of fluoro substitution
at an identical position in antibacterial activity of fluoro-
quinolone class of compounds such as norfloxacin, ofloxa-
cin, ciprofloxacin etc., wherein fluoro substituent is shown
to improve the DNA gyrase inhibition up to 15- to 18-fold
over its 6-hydrogen analog (18).
r2 = 77.40, r2 (adj) = 71.50, r2 (pred) = 65.42, q2 = 74.39.
In model M-1, consensus of five type of descriptors was
found to be very significant for forecasting the biological
activity. The inhibitory activity is directly proportional to the
shape attribute, topological diameter, and total connectivity
as evidenced by the positive regression coefficient. This
model shows large negative correlation of the steric
parameter connolly molecular area and Henry’s law con-
stant. The physicochemical parameter such as molecular
refractivity is combination of steric bulk (volume) and polar-
izability of the substituents. A smaller MR indicates the
strong interactions within the enzyme and good possibility
in drug-receptor interactions. A negative correlation sug-
gests that a high polarizability is not desirable for the good
activity. The topological diameter has favorable contribu-
tion as evidenced by positive regression coefficient. It can
be clearly understood in the case of compounds 32, 33,
34, 35, 36, and 37 that more the topological diameter,
greater will be inhibitory concentration. Partition coefficient
(LogP), number of rotatable bonds, and Gibbs free energy
have an unfavorable contribution toward the binding inhibi-
tory activity.
Quantitative structure activity relationship
Quantitative structure activity relationship investigations
have been carried out to describe the relationship between
chemical structure and biological activities of a series of
chromanyl–isoxazolidines (1-40). Various thermodynamic,
electronic, steric, and molecular topology descriptors have
been calculated using CHEMDRAW 12 (Cambridge Soft Cor-
poration, Cambridge, MA, USA).
As no outliers were identified, therefore, it is important to
note that all these models were developed using the entire
set (n = 40). For the generation of models, the entire data
set was divided into training (n = 30) and test set (n = 10)
by K-means of clustering (Table S1). The whole data were
divided into four subgroups from each of which 25% of
compounds were selected as the members of the test set
(Table S2). Stepwise multiple linear regression analysis
method was employed to generate QSAR using MINITAB
software (27), and its validation was carried out by leave-
one-out (LOO) cross-validation method. The quality of the
model is indicated by the various parameters such as cor-
relation coefficient (r), Fisher’s statistics (F), standard error
of estimation (s), and adjusted coefficient of variation (r2adj).
The model proved to be robust and reliable as shown from
the values s = 0.608, r2 = 0.774, r2 (adj) = 0.715,
PRESS = 1.01, pred_r2 = 0.654, and q2 = 0.743 which
are much satisfactory than the stipulated value of 0.5. The
predictive potential of this model was determined by q2 of
the test set compounds, and it was found to be 74.3%
(Figure 2A,B).
The model M-2 is developed against S. typhi 2 for the
explanation of structure activity relationship.
A number of QSAR models were developed using all the
generated descriptors as independent variables for the
compounds against the S. typhi 1 and S. typhi 2. Out of
all, best two models (M-1 and M-2) were shown for the in-
depth understanding of the crucial factors required for the
activity.
pMIC50 ¼ 5:332 þ 0:00173 ðCritical VolumeÞ
ꢀ 0:158 ðShape AttributeÞ
ꢀ 0:0156 ðIdeal Gas Thermal CapacityÞ
ꢀ 0:099 ðNum Rotatable BondsÞ
ꢀ 0:071 ðMol RefractivityÞ
þ 0:048 ðHenry0s Law ConstantÞ
(M-2)
pMIC50 ¼ 2:278 þ 0:143 ðShape AttributeÞ
ꢀ 0:0087 ðConnolly Molecular AreaÞ
ꢀ 0:110 ðMol RefractivityÞ
r2 = 71.35, r2 (adj) = 63.87, r2 (pred) = 54.20, q2 = 73.29.
ꢀ 0:099 ðHenry0s Law ConstantÞ
þ 0:251 ðTopological DiameterÞ
In this model M-2, all the found descriptors have favorable
contribution toward enzyme inhibition except ideal gas
thermal capacity and molecular refractivity. Ideal gas ther-
þ 8873 ðTotal ConnectivityÞ
(M-1)
Training set (n = 30)
Test set (n = 10)
A
B
7
6
5
4
3
2
1
0
6
5
4
3
2
1
0
y = 0.773x + 1.134
r2 = 0.774
y = 0.739x + 1.383
r2 = 0.743
Figure 2: Correlation between observed
and calculated/predicted activity using
model 1 [M1] for (A) Training Set and (B)
Test Set.
4.5
5
5.5
6
4.5
5
5.5
6
Predicted pIC50
Predicted pIC50
218
Chem Biol Drug Des 2016; 87: 213–223

Chromanyl-isoxazolidines as Antibacterial Agents
mal capacity and critical pressure should be less to exhibit
higher activity as evidenced by negative regression coeffi-
cient. These findings suggest that the attenuation of these
parameters will lead to enhance inhibitory activity. Molecu-
lar refractivity is indirectly proportional to the inhibitory
activity. As fluorine has less molecular refractivity than
chlorine so that compound 32 is more potent than com-
pound 31. Compounds 31 and 32 showed good antibac-
in vitro activity and to determine the best in silico confor-
mation of most potent molecule. In our research group,
the bio-isosteric equivalence of chromone nucleus and
quinolone moiety has been established (19,20), and it is
pertinent to mention here that the chromone analogs
may follow same mode of action as of various fluoro-
quinolones such as ciprofloxacin and norfloxacin, by
inhibiting DNA gyrase; therefore, the crystal structure of
the topo II DNA gyrase with its cocrystal ciprofloxacin
(PDB ID: 2XCT) was selected as target, to understand
the nature of the interaction between the most potent
molecule and DNA gyrase (28). DNA gyrase, a type II
topoisomerase responsible for DNA replication and repair,
is essential in all bacteria and absent in eukaryotes (29).
The enzyme catalyzes the interconversion of various
topological forms of DNA, an essential process in DNA
replication. The crystal structure of topoisomerase II DNA
gyrase with its cocrystal surrounding by active amino
acids is shown in Figure 4.
terial activity against S. typhi
1 & S. typhi 2 with
MIC = 6.25 and 1.56 lg/mL, respectively. The difference
in the inhibitory activity of the compounds 32 and 31
attributed to the presence of fluoro- and chloro- group,
respectively, at C6 position is due to the higher polar sur-
face area of the former. The quality of fit was justified by
the correlation graphs as shown in the Figure 3. The
model proved to be reliable and statistically significant as
the value of r2 and r2 (pred) is 71.35 and 54.20, respec-
tively.
Based on these models, it is observed that the predicted
antibacterial activities by our QSAR models were very
close to those experimentally observed, indicating that
these models can be safely applied for forecasting of more
effective hits having the same skeletal framework.
The comparative and automated docking studies were
performed using GLIDE program to determine the best in sil-
ico conformation of synthesized molecule inside the active
site (30). The standard drugs such as ciprofloxacin and
linezolid were drawn and docked back into the protein to
determine the reliability of the docking method. Figure 5
clearly shows the important interactions of the both stan-
dard drugs in the enzyme pocket.
Molecular docking studies
Considering the potent activity of synthesized com-
pounds (1–40) as new antibacterial agents observed in
in vitro study, it was worthwhile to perform molecular
docking studies (in silico virtual screening), to support the
The predicted binding interactions were analyzed, and the
hypothetical binding mode of all the ligands in the defined
Training set (n = 30)
Test set (n = 10)
A
B
7
6
5
4
3
2
1
0
6
5
4
3
2
1
0
y = 0.711x + 1.420
r² = 0.713
y = 1.815x – 3.726
r² = 0.732
Figure 3: Correlation between observed
and calculated/predicted activity using
model M4 for (A) Training Set and (B) Test
Set.
4.5
5
5.5
6
4.5
5
5.5
6
Predicted pIC50
Predicted pIC50
A
B
Figure 4: (A) The crystal structure of
topoisomerase II DNA gyrase with its
cocrystal found by site-map. (B) Active
amino acids surrounding the cocrystal in the
binding pocket.
Chem Biol Drug Des 2016; 87: 213–223
219

Singh et al.
A
B
Figure 5: (A) Hypothetical binding
orientation showing the important
interactions of the linezolid and (B)
ciprofloxacin in the enzyme pocket (2D-
VIEW).
grid of the crystal structure is shown in the Figure 6. Theo-
retically, all the synthesized compounds showed G score
ranging from ꢀ7.68 to ꢀ3.31 (Table S3).
Another moderately active compound 33 also shows hydro-
gen bonding with GLH-1088; however, there is no pie–pie
interaction with any amino acids (Figure 8). Therefore, the
activity may be attributed to inhibition of enzyme DNA gyrase,
which is crucial for the essential processes of DNA transcrip-
tion and cell replication.
Among the series, docking of DNA gyrase enzyme with 32
revealed that its dock score (ꢀ7.68) was quite comparable
with standard drug ciprofloxacin (dock score ꢀ6.37). The
docking binding model of compound 32 indicated that it
was well filled in the active site. The docked compound 32
formed hydrogen bonding with GLH 1088 and p–p inter-
action with DG 8 (Figure 7).
Physicochemical and ADME properties
On account of unfavorable absorption, distribution,
metabolism, and elimination (ADME) parameters, most of
Figure 6: Binding orientation of all ligands
in the binding pocket of the enzyme
structure.
A
B
Figure 7: (A) Binding orientation of the active compound 32 in the enzyme pocket showing hydrophilic (blue colored) and hydrophobic
map (brown colored). (B) Another pose of the same compound showing pie–pie interactions with DG 8 and hydrogen bonding interactions
with GLH 1088 (2D view).
220
Chem Biol Drug Des 2016; 87: 213–223

Chromanyl-isoxazolidines as Antibacterial Agents
A
B
Figure 8: (A) Hypothetical mode of compound 33 forming the hydrogen bond with GLH 1088 in the binding pocket of the enzyme. (B) In
the case of moderate active compound 33, there is no formation of pie–pie interactions (2D view).
the potential therapeutic agents fail to reach the clinic.
Therefore, computational study of synthesized
logs showed significant antibacterial activities in compar-
ison with other analogs. In particular, compound 32,
having fluoro substituent at C6 was found to have the
most potent inhibitory potential against S. typhi-1, S. ty-
phi-2, and methicillin-resistant Staphylococcus aureus,
which also points toward possibly an identical mode of
action as fluoroquinolones; supported by docking analy-
sis. Correlation of the results obtained from docking and
QSAR studies provides information of structural require-
ments for the enhanced activity. It can be used in
rationalization of the activity trends in the molecules
under study. Therefore, our findings support that disub-
stituted (30–34) and bicyclic (38-40) isoxazolidines have
promising potential as lead compounds for the design of
new antibacterial agent.
a
compounds 1–40 was performed using QIKPROP v3.6 tool
of Schrodinger software for the assessment of ADME
properties, and values obtained are depicted in Table 2b.
From all these parameters, it can be observed that
compound 32 exhibits significant ADME properties in the
acceptable range.
A molecule likely to be developed as an orally active drug
candidate should show no more than one violation of the
following four criteria: number of hydrogen bond acceptor
≤10, number of hydrogen bond donor ≤5, logP (octanol–
water partition coefficient) ≤5, and molecular weight ≤500.
All the synthesized molecules showed significant values for
the ADME properties analyzed (Table S4) and exhibited
drug-like characteristics based on Lipinski’s rule of 5;(31)
therefore, these compounds can be further developed as
oral drug candidates.
Acknowledgments
Financial support from CSIR, New Delhi (02(0082)/12/
EMR-II) to GS is gratefully acknowledged. UGC, New
Delhi, is acknowledged for grant to G.N.D.U. Amritsar
under UPE program.
Conclusion
Based on our previous work, we have synthesized series
of chromanyl–isoxazolidines (1-40), evaluated for their
antibacterial activity against S. typhi 1, S. typhi 2, and
MRSA bacterial strains. All the fluorine-substituted ana-
Conflict of interest
The authors declare they have no conflict of interests.
Table 2: In silico predicted ADME properties for good oral bioavailability of potent compound 32
Predicted physicochemical pharmacokinetic properties
ꢀ2
Entry
n-ON acceptors
4
n-OHNH donors
0
nROTB
3
PSA (A )
QP logP
5.683
QP logS
QP logK
QP logBB
0.405
MW
Comp.32
43.10
ꢀ6.651
ꢀ0.332
401.43
n-ON acceptors: number of H-bond acceptors; n-OHNH donors: number of H-bond donors; nROTB: number of rotatable bonds; PSA:
polar surface area; QP logP: predicted octanol/water partition coefficient; QP logS: predicted aqueous solubility; QP logK: prediction of
binding to human serum albumin; QP logBB: prediction of Blood/Brain partition coefficient; MW: molecular weight; ADME, absorption, dis-
tribution, metabolism, and elimination.
Chem Biol Drug Des 2016; 87: 213–223
221

Singh et al.
tion of 3-(5-phenyl-3H-[1,2,4]dithiazol-3-yl)chromen-4-
ones and 4-oxo-4H-chromene-3-carbothioic acid N-
phenylamides and their antimicrobial evaluation. Eur J
Med Chem;44:3209–3216.
References
1. Veale C.G.L., Zoraghi R., Young R.M., Morrison J.P.,
Pretheeban M., Lobb K.A., Reiner N.E., Andersen R.J.,
Davies-Coleman M.T. (2015) Synthetic analogues of
the marine bisindole deoxytopsentin: potent selective
17. Alaa A.M., Abdel A., Yousif A.A., Mohamed H.M.A.
(2011) Design, synthesis and antibacterial activity of
fluoroquinolones containing bulky arenesulfonyl frag-
inhibitors of MRSA pyruvate kinase.
J
Nat
ment: 2D-QSAR and docking study. Eur
J Med
Prod;78:355–362.
Chem;46:5487–5497.
2. Michalska K., Karpiuk I., Krol M., Tyski S. (2013) Recent
development of potent analogues of oxazolidinone
antibacterial agents. Bioorg Med Chem;21:577–591.
3. Diwakar S.D., Bhagwat S.S., Shingare M.S., Gill C.H.
(2008) Substituted 3-((Z)-2-(4-nitrophenyl)-2-(1H-tetra-
zol-5-yl) vinyl)-4H-chromen-4-ones as novel anti-MRSA
agents: synthesis, SAR and in-vitro assessment. Bioorg
Med Chem Lett;18:4678–4681.
18. Domagala J.M., Hanna L.D., Heifetz C.L., Hutt M.P.,
Mich T.F., Sanchez J.P., Solomon M. (1986) New
structure activity relationships of the quinolone antibac-
terials using the target enzyme. The development and
application of
a
DNA gyrase assay.
J
Med
Chem;29:394–404.
19. Ishar M.P.S., Singh G., Singh S., Sreenivasan K.K.,
Singh G. (2006) Design, synthesis, and evaluation of
novel 6-chloro-/fluorochromone derivatives as potential
topoisomerase inhibitor anticancer agents. Bioorg Med
Chem Lett;16:1366–1370.
4. Filler R., Kobayashi Y. (1982) Biomedicinal Aspects of
Fluorine Chemistry. Tokyo and Amsterdam: Kodansya
and Elsevier Biomedical.
5. Welch J.T., Eswarakrishman S. (1991) Fluorine in
Bioorganic Chemistry. New York: JohnWiley & Sons.
6. Filler R., Kobayashi Y., Yagupolskii L.M. (1993)
Organofluorine Compounds in Medicinal Chemistry and
Biomedical Applications. Amsterdam: Elsevier.
7. Kukhar V.P., Soloshonok V.A. (1995) Fluorine Contain-
ing Amino Acids: Synthesis and Properties. Chichester:
John Wiley & Sons Ltd.
8. Ojima I., McCarthy J.R., Welch J.T. (1996) Biomedical
Frontiers of Fluorine Chemistry. Washington, D.C.:
American Chemical Society.
9. Hiyama T. (2000) Organofluorine Compounds, Chem-
istry and Applications. Berlin: Springer.
10. Kirsch P. (2004) Modern Fluoroorganic Chemistry.
Weinheim: Wiley-VCH Verlag GmbH.
11. Uneyama K. (2006) Organofluorine Chemistry. Oxford:
Blackwell.
12. Begue J-P., Bonnet-Delpon D. (2006) Recent
advances (1995–2005) in fluorinated pharmaceuticals
based on natural products. J Fluor Chem;127:992–
1012.
13. Kirk K.L. (2006) Fluorine in medicinal chemistry: recent
therapeutic applications of fluorinated small molecules.
J Fluor Chem;127:1013–1029.
14. Kim H.Y., Wiles J.A., Wang Q., Pais G.C.G., Lucien
E., Hashimoto A., Nelson D.M., Thanassi J.A., Podos
S.D., Deshpande M., Pucci M.J., Bradbury B.J. (2011)
Exploration of the activity of 7-Pyrrolidino-8-methoxy-
isothiazoloquinolones against methicillin-resistant Sta-
phylococcus aureus (MRSA). J Med Chem;54:3268–
3282.
15. Barbachyn M.R., Cleek G.J., Dolak L.A., Garmon
S.A., Morris J., Seest E.P., Thomas R.C., et al.
(2003) Identification of phenylisoxazolines as novel
and viable antibacterial agents active against gram-
positive pathogens. J Med Chem;46:284–302.
16. Raj T., Bhatia R.K., Sharma R.K., Gupta V., Sharma
D., Ishar M.P.S. (2009) Mechanism of unusual forma-
20. Mayer C., Janin Y.L. (2014) Non-quinolone inhibitors of
bacterial type-IIA topoisomerases: a feat of bioisoster-
ism. Chem Rev;114:2313–2342.
21. Singh G., Kaur A., Sharma V., Suri N., Sharma P.R.,
Saxena A.K., Ishar M.P.S. (2013) Synthesis and cyto-
toxicity evaluation of regioisomeric substituted N-phe-
nyl-30-(chrom-4-one-3-yl)-isoxazolidines: induction of
apoptosis through a mitochondrial-dependent path-
way. Medchemcommun;4:972–978.
22. Keri R.S., Budagumpi S., Pai R.K., Balakrishna R.G.
(2014) Chromones as a privileged scaffold in drug dis-
covery: a review. Eur J Med Chem;78:340–374.
23. Chakraborty B., Samanta A., Sharma P.K., Chhetri
M.S., Kafley S., Banerjee A., Sinha C. (2010) One pot
high yield synthesis of some novel isoxazolidine deriva-
tives using N-methyl-a-chloro nitrone in water and their
antibacterial activity. J Chem Pharm Res;2:727–739.
24. Ishar M.P.S., Singh G., Kumar K., Singh R. (2000) In-
vestigations on peri-, regio- and stereoselectivities in
thermal cycloadditions involving C-(4-Oxo-4H[1]ben-
zopyran-3-yl)-N-phenylnitrones: role of steric factors
and secondary interactions in 1,3-dipolar cycloaddi-
tions. Tetrahedron;56:7817–7828.
25. Mitic-Culafic D., Vukovic-Gacic B., Knezevic-Vukcevic
J., Stankovic S., Simic D. (2005) Comparative study
on the antibacterial activity of volatiles from sage (Sal-
via officinalis L.). Arch Biol Sci;57:173–178.
26. Hindler J., Mahon C.C., Manuselis G. (1995) Textbook
of Diagnostic Microbiology Special Antimicrobial Sus-
ceptibility Tests. China: Saunders Elsevier; 89.
27. Darlington R.B. (1990) Regression and Linear Models.
New York: McGraw-Hill.
28. Bax B.D., Chan P.F., Eggleston D.S., Fosberry A.,
Gentry D.R., Gorrec F., Giordano I., et al. (2010) Type
IIA topoisomerase inhibition by a new class of antibac-
terial agents. Nature;466:935–940.
29. Hameed S., Raichurkar A., Madhavapeddi P., Menasi-
nakai S., Sharma S., Kaur P., Nandishaiah R., Pan-
222
Chem Biol Drug Des 2016; 87: 213–223

Chromanyl-isoxazolidines as Antibacterial Agents
duga V., Reddy J., Sambandamurthy V.K., Sriram D.
(2014) Benzimidazoles: novel mycobacterial gyrase
inhibitors from scaffold morphing. ACS Med Chem
Lett;5:820–825.
Supporting Information
Additional Supporting Information may be found in the
online version of this article:
30. Halgren T.A., Murphy R.B., Friesner R.A., Beard H.S.,
Frye L.L., Pollard W.T., Banks J.L. (2004) Glide: a new
approach for rapid, accurate docking and scoring. 2.
Enrichment factors in database screening. J Med
Chem;47:1750–1759.
Table S1. Training set (n = 30) and test set (n = 10)
ligands obtained through K-Means clustering.
Table S2. K-Means clustering of ligands (n = 40) using
31. Lipinski C.A., Lombardo F., Dominy B.W., Feeney P.J.
(2012) Experimental and computational approaches to
estimate solubility and permeability in drug discovery
and development settings. Adv Drug Deliv Rev;64:4–17.
calculated descriptors.
Table S3. Dock Score for all the synthesized compounds
(1–40).
Notes
Table S4. In silico predicted ADME properties for good
oral bioavailability of compounds 1–40.
aSmall-Molecule Drug Discovery Suite 2014-2: Glide, ver-
€
sion 6.3, Schrodinger, LLC, New York, NY, 2014.
bSchrodinger, LLC. New York, USA: Schrodinger Inc.;
2008
Chem Biol Drug Des 2016; 87: 213–223
223