New potent antibacterials against Gram-positive multiresistant pathogens: Effects of side chain modification and chirality in linezolid-like 1,2,4-oxadiazoles

Article abstract of DOI:10.1016/j.bmc.2014.10.037

The effects of side chain modification and chirality in linezolid-like 1,2,4-oxadiazoles have been studied to design new potent antibacterials against Gram-positive multidrug-resistant pathogens. The adopted strategy involved a molecular modelling approach, the synthesis and biological evaluation of new designed compounds, enantiomers separation and absolute configuration assignment. Experimental determination of the antibacterial activity of the designed (S)-1-((3-(4-(3-methyl-1,2,4-oxadiazol-5-yl)phenyl)-oxazolidin-2-one-5-yl)methyl)-3-methylthiourea and (S)-1-((3-(3-fluoro-4-(3-methyl-1,2,4-oxadiazol-5-yl)phenyl)-oxazolidin-2-one-5-yl)methyl)-3-methylthiourea against multidrug resistant linezolid bacterial strains was higher than that of linezolid.

Full text of DOI:10.1016/j.bmc.2014.10.037

Bioorganic & Medicinal Chemistry 22 (2014) 6814–6825
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
New potent antibacterials against Gram-positive multiresistant
pathogens: Effects of side chain modification and chirality in
linezolid-like 1,2,4-oxadiazoles
b
a
c
b
Cosimo G. Fortuna a, , Roberto Berardozzi , Carmela Bonaccorso , Gianluigi Caltabiano , Lorenzo Di Bari ,
⇑
Laura Goracci d, Annalisa Guarcello e,f, Andrea Pace e,f, , Antonio Palumbo Piccionello e,f
Gennaro Pescitelli b, Paola Pierro e,f, Elena Lonati g, Alessandra Bulbarelli g, Clementina E. A. Cocuzza h,
,
⇑
Giuseppe Musumarra a, Rosario Musumeci h,
⇑
a Dipartimento di Scienze Chimiche, Università degli Studi di Catania, Viale Andrea Doria 6, I-95125 Catania, Italy
b Dipartimento di Chimica e Chimica Industriale, Università di Pisa, Via Risorgimento, 35, I-56126 Pisa, Italy
c Laboratori de Medicina Computacional, Unitat de Bioestadística, Facultat de Medicina, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain
d Laboratory for Chemometrics and Cheminformatics, Chemistry Department, University of Perugia, Via Elce di Sotto 10, I-06123 Perugia, Italy
e Dipartimento di Scienze e Tecnologie Biologiche Chimiche e Farmaceutiche (STEBICEF), Sez. Chimica, Università degli Studi di Palermo, Viale delle Scienze Ed. 17—Parco D’Orleans II,
I-90128 Palermo, Italy
f Istituto EuroMediterraneo di Scienza e Tecnologia (IEMEST), Via Emerico Amari 123, I-90139 Palermo, Italy
g Dipartimento di Scienze della Salute, Università di Milano-Bicocca, Via Cadore 48, I-20900 Monza (MB), Italy
h Dipartimento di Chirurgia e Medicina Traslazionale, Università di Milano-Bicocca, Via Cadore 48, I-20900 Monza (MB), Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
The effects of side chain modification and chirality in linezolid-like 1,2,4-oxadiazoles have been studied
to design new potent antibacterials against Gram-positive multidrug-resistant pathogens. The adopted
strategy involved a molecular modelling approach, the synthesis and biological evaluation of new
designed compounds, enantiomers separation and absolute configuration assignment. Experimental
determination of the antibacterial activity of the designed (S)-1-((3-(4-(3-methyl-1,2,4-oxadiazol-5-
yl)phenyl)-oxazolidin-2-one-5-yl)methyl)-3-methylthiourea and (S)-1-((3-(3-fluoro-4-(3-methyl-1,2,4-
oxadiazol-5-yl)phenyl)-oxazolidin-2-one-5-yl)methyl)-3-methylthiourea against multidrug resistant
linezolid bacterial strains was higher than that of linezolid.
Received 25 August 2014
Revised 20 October 2014
Accepted 24 October 2014
Available online 1 November 2014
Keywords:
Drug design
Linezolid
Antibiotics
Ó 2014 Elsevier Ltd. All rights reserved.
Multidrug-resistant bacteria
Enantiomers
1. Introduction
action and was not cross-resistant with existing resistant mecha-
nisms in bacteria. Moreover, linezolid is characterized by excellent
The urgent need for new antibiotics due to the increase in the
frequency of bacterial infections with Gram-positive multidrug-
resistant (MDR) strains—such as methicillin-resistant Staphylococ-
cus aureus (MRSA), vancomycin-resistant enterococci (VRE) and
penicillin-resistant Streptococcus pneumoniae (PNSSP)—in both
the hospital and the community settings, has not been met by
the development of effective and broad spectrum antibiotics.1
Among them linezolid, a synthetic oxazolidinone which entered
the market in 2000, appeared to have a unique mechanism of
pharmacokinetic properties with 100% oral bioavailability and
effective penetration at therapeutic concentrations to almost every
organ in the body is making a valuable option in the treatment of
both community and hospital acquired infections.2 However, since
2001 linezolid resistance began to appear3–5 and an intense search
for novel oxazolidinones with promising antibacterial activity
started and structure–activity relationships (SAR) have been exten-
sively reported.6–12 In order to rationalize the site of modifications,
the structure of linezolid is usually divided into four portions, as
elucidated in Scheme 1.
It was suggested that fluorination of phenyl B-ring flanking a
morpholine or piperazine C-ring improved the bacterial activity
and that a C-5 acylaminomethyl group in the side chain was essen-
tial for activity.13 Subsequently, substitution of the C-5 side-chain
⇑
Corresponding authors. Tel.: +39 0957385010; fax: +39 095580138 (C.G.F.); tel.:
+39 091 23897543; fax: +39 091 596825 (A.P.); tel.: +39 02 64488103; fax: +39 02
64488250 (R.M.).
E-mail addresses: cg.fortuna@unict.it (C.G. Fortuna), andrea.pace@unipa.it
(A. Pace), rosario.musumeci@unimib.it (R. Musumeci).
http://dx.doi.org/10.1016/j.bmc.2014.10.037
0968-0896/Ó 2014 Elsevier Ltd. All rights reserved.

C.G. Fortuna et al. / Bioorg. Med. Chem. 22 (2014) 6814–6825
6815
Table 1
O
Chemical-physical calculated values for compounds 1–5 a,b
O
a
O
N
N
H
N
Compound
pKa
LogPa
LogD 7.5a
Log (intrinsic solubility)b
O
1a
1b
2a
2b
3a
3b
4a
4b
5a
5b
—
—
—
—
0.2
0.4
3.2
3.3
1.7
1.8
1.7
1.9
1.2
1.4
0.2
0.4
3.2
3.3
1.7
1.8
1.7
1.9
1.2
1.4
ꢀ2.3
ꢀ2.3
ꢀ3.3
ꢀ3.3
ꢀ3.5
ꢀ3.5
ꢀ3.6
ꢀ3.6
ꢀ3.4
ꢀ3.3
F
C-ring
B-ring A-ring
Linezolid
C5 side-chain
10.37
10.37
10.93
10.93
—
Scheme 1. Structure and portions nomenclature of Linezolid.
of linezolid with thiocarbonyl groups (i.e. thioamide, dithiocarba-
mate, thiourea, and thiocarbamate) demonstrated that these com-
pounds are in vivo more active than linezolid.14 The introduction of
heteroaromatic moieties into the C5 side-chain was able to main-
tain or even enhance the antibacterial activity.15
—
a
Calculated using the MoKa 2.6 software.24
The logarithm of experimental intrinsic solubility (mol/Litre at 25 °C) Log
b
(intrinsic solubility) was predicted using the VolSurf + software.25
Considering the known properties of 1,2,4-oxadiazoles in bioac-
tive compounds16–18 and as biocompatible moiety in functional
biomaterials19–21, we have recently approached the synthesis of
linezolid-like compounds by introduction of a 1,2,4-oxadiazolyl
moiety. By following this approach, we have recently shown that
the replacement of the oxazolidinone (A-ring) with 1,2,4-oxadiaz-
ole resulted in the loss of activity,22 while replacement of the mor-
pholine or piperazine C-ring with 1,2,4-oxadiazole allowed to
maintain a satisfactory antibacterial activity against Gram-positive
multiresistant pathogens.23 Moreover, the presence of fluorine in
the benzene B-ring of these new oxadiazoles exhibited no signifi-
cant effect on activity and produced a slight decrease in cytotoxic-
ity, while C-5 acetamidomethyl or thioacetamidomethyl moieties
appeared essential structural requirements for activity.23 However,
in previous work22,23 the newly synthesized compounds were
tested as a racemic mixture. An homogeneous comparison with
the activity of Linezolid S enantiomer can only be achieved by
the measurement of the antibacterial activity of the S enantiomers
of new linezolid-like 1,2,4-oxadiazoles. In this context we here
report the design, synthesis, enantiomer separation, and in vitro
antibacterial evaluation for both S and R enantiomers to gain a dee-
per insight into the effects of C-5 side chain modification and
chirality.
The synthesis was achieved by initially following the amidoxime
route scheme for the construction of the 1,2,4-oxadiazole’s moi-
ety22,26 (Scheme 2).
For the subsequent step, the typical reactivity of 5-fluoroaryl-
1,2,4-oxadiazoles towards the nucleophilic aromatic substitu-
tion,27–30 was a key feature to access allylamine derivatives 9a,b
precursors for the construction of the oxazolidinone ring
(Scheme 3) through a iodo-cyclization reaction leading to oxazolid-
inones 10a,b.23
The synthetic pathway towards final derivatives 1–5 was
planned by taking advantage of the iodomethylene moiety of key
derivatives 10 as illustrated in Scheme 4.
In particular, methylidene derivatives 1 were obtained in good
yield and under mild conditions by a base promoted elimination
reaction. Isothiocyanates 2, azides 11, and triazoles 5, were instead
obtained through a nucleophilic substitution with either potassium
thiocyanate, sodium azide, or 1,2,3-triazole, respectively. Further
functionalization of derivatives 2 has been achieved by addition
of ammonia leading to the corresponding thioureas 3a,b. Reduction
of azides 11 into amines 12, followed by reaction with methylisoth-
iocyanate, yielded the corresponding methylthioureas 4.
2.2. Microbiological evaluation of compounds 1–5a,b
2. Results and discussion
The antibacterial tests against 3 MDR clinical isolates of MRSA,
compared with those against 2 standard ATCC S. aureus strains
reported in Table 2 show that 4a and 4b exhibit an excellent activ-
ity expressed in low Minimal Inhibitory Concentration (MIC) val-
ues against MRSA strains with potency comparable to that of
linezolid, while 5a and 5b are also active; other compounds are
inactive.
The presence of thioacetamidomethyl or 1,2,3-triazole moieties
in the side chain appear to be essential structural requirements
for activity, while the aromatic F substituent has no effect.
2.1. Synthesis and biological evaluation of new linezolid-like
1,2,4-oxadiazoles
In order to investigate the effects of side chain modification,
compounds 1–5a,b (see Chart 1) were synthesized and tested
against multidrug-resistant clinical isolates. In Table 1 we report
the calculated pKa, LogP, LogD75, and the logarithm of the pre-
dicted intrinsic solubility for compounds 1–5 a,b.
O
H3C
N
O
N
a: R = H
b: R = F
N
O
R
1a,b
O
Compound R2
H3C
N
O
N
N
2
3
4
5
NCS
NHC(=S)NH2
NHC(=S)NHCH3
1,2,3-triazol-1-yl
O
a: R1 = H
b: R1 = F
R1
2-5a,b
R2
Chart 1.

6816
C.G. Fortuna et al. / Bioorg. Med. Chem. 22 (2014) 6814–6825
O
R
CH3
N
H2N
CH3
NOH
i) Acetone / K2CO3
ii) Δ/solvent-free
Cl
N
O
F
F
R
6a,b
7
a: R = H
b: R = F
8a,b
Scheme 2.
CH3
i) (tBuOCO)2O
DMAP/ACN
CH3
CH3
N
N
N
NH2
K2CO3
O
I
N
N
N
O
O
O
N
ii) I2 / CH2Cl2
O
F
HN
R
R
R
9a,b
8a,b
10a,b
Scheme 3.
CH3
N
O
N
O
N
O
R
1a,b
ACN K2CO3
CH3
CH3
CH3
N
N
N
O
N
O
O
I
N
N
O
O
KSCN
NaN3
DMF
O
N
N
O
N
N
O
O
R
R
R
11a,b
2a,b
10a,b
N3
O
N C
S
PPh3
THF
NH3 CH3OH
1,2,3-triazole
CH3
N
O
N
CH3
CH3
O
N
R
N
O
N
N
O
O
N
O
O
N
12a,b
O
NH2
N
R
R
S
5a,b
CH3NCS
HN
N
NH2
N
3a,b
CH3
N
O
N
O
N
O
R
S
HN
NHCH3
4a,b
Scheme 4.

C.G. Fortuna et al. / Bioorg. Med. Chem. 22 (2014) 6814–6825
6817
Table 2
Antimicrobial activities, expressed in MIC values (
lg/ml) of compounds 1–5 a,b; linezolid was used as reference antibiotic agent
Compounds/bacterial strains
1a
1b
2a
2b
3a
3b
4b
4a
5a
5b
Linezolid
S. aureus ATCC 29213
S. aureus ATCC 700699 (MU 50)
S. aureus MRSA M923
S. aureus MRSA 433
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
50
>50
>50
>50
>50
>50
>50
>50
>50
>50
50
>50
>50
>50
>50
>50
6.25
1.6
6.25
3.125
6.25
6.25
1.6
6.25
3.125
6.25
25
3.125
25
12.5
12.5
25
1.6
25
6.25
12.5
<0.025
0.8
3.125
1.6
S. aureus MRSA F511
3.125
Interestingly, molecular modelling is capable to rationalize these
biological results at a molecular level (see below).
U2585. Moreover, according to the FLAP poses the fluorine atom
of the aromatic ring seems not to be important to generate specific
interactions. This is in agreement with the experimental findings
showing that, in the 1,2,4-oxadiazole linezolid-like compound ser-
ies, the presence of the fluorine atom does not play any role for
activity in 1,2,4-oxadiazole linezolid-like compounds.23
Comparison with linezolid must take into account that the
tested compounds were used as a racemic mixture, therefore their
antibacterial activity is underestimated as compared to the
potency of the pure active linezolid S enantiomer. Enantiomeric
separation and biological tests on pure enantiomers are therefore
required to achieve a proper comparison with linezolid activity.
Interestingly, only the four active compounds seem capable to
expose a hydrophobic moiety of the side chain towards the hydro-
phobic field of the cavity generated by the G2061 residue. Indeed,
the inactive compounds 1a–b, lacking the side-chain, cannot favor-
ably interact with the reported hydrophobic region; the other four
inactive compounds (2a–b and 3a–b) show a mismatch since they
point a polar group towards the hydrophobic region generated by
G2061. The importance of this hydrophobic interaction seems to be
confirmed by the different activity between compounds 4a–b and
compounds 3a–b. These compounds only differ by a methyl group,
which in compound 4 can point towards the G2061 hydrophobic
region. It is well known that linezolid resistance mechanism can
be ascribed to mutations of 23S rRNA. The universally conserved
residues in the linezolid binding pockets are G2061, A2451,
C2452, A2503, U2504, G2505, U2506, and U2585, whereas only
U2585 directly interacts with linezolid.42
2.3. Molecular modelling
Previous studies on 4-alkyliden-beta lactams with antibacterial
activity31 adopting a molecular modelling approach using Grid
Independent Descriptors (GRIND) based on the concept of a Virtual
Receptor Site (VRS), allowed identification attractive drug candi-
dates and contributed to the rationalization of functional group
effects in QSARs. The VRS approach was the only one available at
that time due to the lack of available pharmacological targets. In
2008 the X-ray structure of the oxazolidinone antibacterial agent
linezolid, bound to the 50S ribosomal subunit became available32
providing new opportunities to investigate of action of the above
antibacterials at a molecular level.
In the present study, the possible binding modes for compounds
1–5a,b into the target cavity derived from the 50S ribosomal sub-
unit from Deinococcus radiodurans co-crystallized with linezolid
(PDB code:3DLL)33 were investigated using FLAP (Fingerprints for
Ligands and Proteins). The FLAP algorithm was developed with
the aim to describe ligands and proteins based on a common refer-
ence framework.34 Thus, the natural application of this algorithm
was in the field of virtual screening, where FLAP is indeed very suc-
cessful.34–38 However, more recently the FLAP range of applicabil-
ity was extended including pharmacophore based approaches39,40
and similarity-based docking.23,37,41
It is interesting to note that the docking score in FLAP is not cal-
culated by evaluating the thermodynamics of the protein–ligand
interaction, but by considering the 3D complementarity between
the molecular interaction fields of the protein and those reported
for the ligands.
In order to study the mechanism of action of the most active
compounds, docking studies have been carried out on 4aS/R and
4bS/R (Chart 1), comparing their binding position with that of
the reference drug, Linezolid. A model of 23S subunit for Staphylo-
coccus aureus (see Section 4.1) has been made based over the 23S
subunit of the Deinococcus radiodurans 50S ribosomal subunit, in
complex with linezolid.33 Sequence alignment of D. radiodurans
and S. aureus shows that the bases directly involved in the binding
site of linezolid are 100% conserved between the two bacteria,
allowing the hypothesis that the designed oxadiazole compounds
dock at the same spot of Linezolid. Figure 2 shows the S-enantio-
mers of 4b and 4a (panel A and B, respectively), the most active
compounds, in complex with S. aureus. After minimization and a
0.5 ns molecular dynamics (see Section 4.1), both active com-
pounds mostly overlap linezolid but adjustment over S. aureus
binding site are observed. The new drugs share with the com-
pounds presented in a previous paper23 an aromatic C-ring moiety,
and this difference versus linezolid C-ring confers a more rigid sys-
tem with delocalized aromaticity between C-ring and B-ring. This
generates, see Figure 2A and B, a more organized packing of ribo-
somal A2451, C2452, U2504 and U2506 bases around C and B-
rings.
The B-ring of tested compounds is tightly sandwiched between
A2451, C2452 and U2506 thus facilitating the co-planar C-ring
moiety to interact, through hydrogen bonding involving the N(2)
of 1,2,4 oxadiazole moiety, with U2585 whose importance in the
linezolid binding has been documented.33 The stacking of the oxa-
zolidinone ring moiety (A-ring) with U2504 observed in the crystal
structure is also maintained in the present set of compounds. The
docking of 4bS and 4aS shows a rotation of the A-ring, with respect
to linezolid, of around 40° over the axis of the molecule (higher
than the one observed for compounds recently reported).23 This
rotation allows the new active compounds to gain four more
key interactions with the ribosome, with respect to linezolid,
likely accounting for the activity of these compounds versus
Using FLAP as a docking tool, the possible binding modes of
compounds 1–5a,b were evaluated by looking for the best similar-
ity between the target cavity and each one of the tested com-
pounds, as previously described.41 Here the aim was to find
possible interaction at the side-chain level that could be responsi-
ble for the difference in experimentally observed activities. The
structures of both R and S enantiomers for each tested racemic
mixture were modeled by FLAP. Only compounds possessing the
same relative configuration of C(5) of the oxazolidinone moiety
as Linezolid exhibited a satisfactory pose (in agreement with the
in silico design of the active enantiomeric forms discussed later
in the same section). These were the S enantiomers of compounds
3–4 and R enantiomers of compounds 2 and 5, as discussed below.
Figure 1 represents the best poses for the S enantiomer of active
(A) and inactive (B) compounds in the target cavity. The green field
represents the hydrophobic regions in the cavity, according to
FLAP.
As previously reported23, the 1,2,4-oxadiazole rings perfectly
mimic the morpholinic ring of linezolid in its interaction with

6818
C.G. Fortuna et al. / Bioorg. Med. Chem. 22 (2014) 6814–6825
Figure 1. FLAP best poses for active (A) and inactive (B) compounds in the 3DLL binding site of linezolid. Active compounds: 4a–b and 5a–b; inactive compounds: 1a–b, 2a–b,
3a–b. The green field represents the hydrophobic regions in the cavity, according to the GRID molecular interaction field obtained using the DRY probe.
Figure 2. S. aureus model (light blue) in complex with: panel (A) 4b-S (yellow sticks) and panel (B) with 4a-S (cyan sticks). A small rotation over 4b/a-S axis is observed with
respect to (S)-linezolid (LZD-dark gray thinner sticks). As a result four new hydrogen bond form: (i) N-(4) of the 1,2,4-oxadiazole with ribose hydroxyl moiety of A2451; (ii)
the carbonyl oxygen from A-ring and the ribose hydroxyl group of G2505; and (iii) thiouric hydrogens from C5 side chain hydrogen bond both carbonyl oxygens from the
phosphate backbone of the ribosome of G2505.
Linezolid-resistant S. aureus strains. Indeed, with respect to linezo-
lid and to precedent active compounds,23 4bS and 4aS are charac-
terized by the presence of a –NHCH3 moiety in C-5. This generates
a new binding pose in which both thiouric hydrogens from C5 side
chain hydrogen bond both oxygens from the phosphate backbone
of the ribosome. Associated to this rotation are also (i) the hydro-
gen bond of the carbonyl oxygen from A-ring and the ribose hydro-
xyl group of G2505 (Escherichia coli numbering), and (ii) a gained
interaction of N(4) of the 1,2,4 oxadiazole with ribose hydroxyl
group of A2451, which is also involved in the stacking with B-ring.
Thus presence of –NHCH3 in C-5 side chain, force a rotation of the
new compounds allowing a new pose not observable for linezolid.
The methyl group of the C-5 ring is also playing a role. Indeed
compounds 3a and 3b, which feature hydrogen instead, are not
active. Figure 1 A and B, show the methyl is pointing to the aro-
matic ring of A2059, thus the presence of a free –NH2 group in
the thioureic moiety would favor a planar tautomeric @NH form,
in equilibrium with –NH2, which would break the key hydrogen
bond with the carbonyl oxygen from the phosphate backbone of
the ribosome and, moreover, would let appear a –SH in an
unfriendly hydrophobic environment, as is the one created by
G2061, A2059, A2058 and A2503.
It is worth noting that 4a differs from 4b by a fluorine atom at
B-ring. This fluorine is supposedly involved in coordinating with a
putative Mg2+ as described in Wilson at al.33 None of our docking
and/or experiments support differences due to fluorine atom.

C.G. Fortuna et al. / Bioorg. Med. Chem. 22 (2014) 6814–6825
6819
2.5
2
10
4a (S)
NVI579b
NVI579a
4a (R)
4a (S)
NVI579b
NVI579a
4a (R)
5
0
1.5
1
-5
0.5
0
-10
200
250
300
350
400
200
250
300
350
400
λ (nm)
λ (nm)
Figure 3. Experimental UV (left) and ECD (right) spectra of the two enantiomers of 4a (separated by HPLC) in acetonitrile (3 mM solutions); 0.02 cm cell. The absolute
configuration shown in the labels follows from the results discussed in the text.
Table 3
M06/6-31G(d) calculated geometries, relative free energies and Boltzmann populations at 300 K for the most stable conformations of (R)-4a
Mol
Structure
E (kcal/mol)
Pop % (300 K)
Fold1_cis
0.00
36.1
Fold1_trans
Fold2_trans
Fold2_cis
0.01
0.52
0.60
35.6
15.0
13.1
Ext_trans
Ext_cis
3.68
3.74
0
0
‘Fold’ and ‘Ext’ refer to the conformation of the C5-substituent containing the thiourea moiety; cis and trans refer to the orientation of the oxadiazole ring.

6820
C.G. Fortuna et al. / Bioorg. Med. Chem. 22 (2014) 6814–6825
The modeling binding mode studies encouraged to proceed
with racemate resolution, assignment of absolute configuration,
and microbiological tests.
40
30
Fold1_cis
Fold1_trans
Fold2_cis
2.4. Enantiomers separation
20
The enantiomers were separated by enantioselective HPLC
using chiral stationary phase on Daicel Chiralpak IA column with
n-hexane/isopropanol 70/30 as eluent and semipreparative separa-
tion was achieved by multiple injections (see Section 4). Each
enantiomer was obtained with ee values greater than 96%.
Under the same analytical HPLC conditions performed on race-
mic and S-linezolid the latter was revealed as the first eluting
enantiomer. This data allowed to reasonably correlate the relative
configurations of 4a and 4b first eluting peaks with that of S-lin-
ezolid. This hypothesis was confirmed by ECD experiments (see
below) performed on both enantiomers of 4a, whose first and sec-
ond eluting fractions were labelled as 4a(S) and 4a(R).
10
Fold2_trans
0
-10
-20
-30
-40
Rotational strengths for Fold1_cis
200
250
300
λ (nm)
350
400
2.5. Assignment of absolute configuration of the enantiomers of
4a
Figure 4. ECD spectra calculated with TDDFT method, CAM-B3LYP/TZVP//M06/6-
31G(d), on the first 4 lowest-energy minima of (R)-4a. Spectra were generated as
sum of Gaussians with 0.5 eV exponential half-width. Vertical bars represent
rotational strengths computed for the absolute minimum (in 10ꢀ39 cgs units).
The absolute configuration of the two enantiomers of com-
pound 4a was determined by comparing their electronic circular
dichroism (ECD) spectra with those calculated with time-depen-
dent density functional theory (TDDFT). The input geometries used
in TDDFT calculations were established on the basis of a conforma-
tional analysis run with molecular mechanics, followed by DFT
geometry optimizations. To ensure the reliability of the molecular
modelling results, these latter were checked against NMR data. The
above procedure may be regarded as a standard for assigning the
absolute configuration of moderately flexible organic molecules.43
The experimental UV–vis and ECD spectra of the two enantio-
mers of 4a in acetonitrile solution (Fig. 3) show the presence of
several bands mainly allied with the composite chromophore
which includes the phenyl ring, the oxadiazole ring, and the carba-
mate moiety. The ECD spectra for the two enantiomers of 4a sep-
arated by HPLC, are almost mirror images, as expected. The first
appearing ECD band at 270 nm (negative for (R)-4a the second-
eluted enantiomer) is very noisy because it is associated with a
strong absorption band and allied with a very small g factor
(around 5ꢁ10ꢀ5).
oxazolidine ring, while neither the conformation assumed by
the thiourea moiety nor the orientation of the oxadiazole do
play a decisive role.
Average absorption and ECD spectra were calculated by weight-
ing the four component spectra using Boltzmann populations at
300 K from DFT internal energies. The Boltzmann-averaged ECD
spectrum calculated for (R)-4a, (Fig. 5) is in very good agreement
with the experimental ECD spectrum of the second eluted enantio-
mer, (solid black spectrum in Figure 5) in terms of relative position
and intensity of almost all bands. Therefore the elution order for
the two enantiomers of 4a is established as: first eluted, (S)-4a;
second eluted, (R)-4a.
2.6. Enantiomers biological evaluation
Table 4 reports the in vitro activities for 4a and 4b racemic
mixtures and for the corresponding pure enantiomers versus 51
clinical bacterial strains (6 S. aureus ATCC and 45 MRSA) together
with those of linezolid. This table shows that the modeled (S)
1,2,4-oxadiazoles 4a and 4b exhibit the same activity of linezolid.
Therefore the modeling and the separation of the above com-
pounds provides additional tools to face the resistance of MRSA
clinical-isolates.
The relevance of these findings relies in the fact that the identi-
fication of new structures active against ATCC and MRSA clinical
isolates represents an urgent public health problem. However, lin-
ezolid resistant bacteria confined within hospital communities
have recently been isolated. The modelling aimed at the identifica-
tion of the most active 1,2,4-oxadiazole enantiomers is based on
the assumption that similarity of the structures of the new com-
pounds could predict for the same binding site and therefore the
activity at the cellular level is due to the same mechanism as lin-
ezolid. Consequently active enantiomers have also been tested
towards a few available linezolid resistant bacterial cultures.
Table 3 shows that the activity of the designed (S) 1,2,4-oxadiaz-
oles 4a and 4b against linezolid resistant strains is higher than that
of linezolid. Therefore the design and the synthesis of the new
compounds reported in this work expands the battery of tools
to tackle linezolid resistant strains, a possible epidemiological
problem in the not too distant future.
Enantiomer (R)-4a was arbitrarily chosen to run the molecular
modelling, whose results are summarized in Table 3, while the
details are given in the Computational Section. A conformational
search run with MMFF (Merck force field), followed by DFT geom-
etry optimizations at M06/6-31G(d) level, revealed a clear preva-
lence for the structures with the substituent at C5 oriented
axially and the thiourea moiety folded over the carbamate one.
Only four low-energy conformations seem to be populated at room
temperature, all of which show a ‘folded’ thiourea moiety and a
consistent conformation of the 5-membered oxooxazolidine ring,
and differ in the orientation of the oxadiazole ring and of the ter-
minal NH-methyl group.
The prevalence of folded conformations was also confirmed by
an NMR ROESY experiment (see Supporting information) where
diagnostic Overhauser effects were detected between: the thiourea
proton at 6.61 ppm and the aromatic protons H20 at 7.76 ppm (and
to a lesser extent H30 at 8.11 ppm); the methyl group at 2.89 ppm
and H30.
ECD spectra were calculated for the 4 DFT-optimized lowest-
energy minima of (R)-4a, with TDDFT method at CAM-B3LYP/
TZVP level. The four calculated spectra (shown in the Supporting
information) showed a consistent profile in the whole range.
This result demonstrates as expected, that the ECD spectrum
is substantially governed by the absolute conformation of the

C.G. Fortuna et al. / Bioorg. Med. Chem. 22 (2014) 6814–6825
6821
40
30
6
5
4
3
2
1
0
Calculated for (R)-NVI579
20
10
0
-10
-20
-30
-40
Calculated for (R)-NVI579
200
250
300
350
400
200
250
300
350
400
λ (nm)
λ (nm)
Figure 5. UV (left) and ECD (right) spectra calculated with TDDFT method, CAM-B3LYP/TZVP//M06/6-31G(d), as Boltzmann-weighted average on the first 4 lowest-energy
minima of (R)-4a. Spectra were generated as sum of Gaussians with 0.5 eV exponential half-width. The four component spectra are shown in the Supporting information.
Table 4
MIC values for racemic mixtures 4 and 5 and for the corresponding enantiomers
MIC-values
MIC-range
<0,06–>128 g/ml
Compounds
4b
4bS
4bR
4a
4aS
4aR
LZD
MIC-range
MIC50
MIC90
0,5–16
4
16
0,5–8
2
4
64–>128
>128
>128
1–16
8
16
0,5–8
2
4
128–>128
>128
>128
0,25–16
2
4
Tested strains: 6 ATCC (4 MSSA + 2 MRSA) e 45 MRSA all LZD-S
Tested strains: 12 MRSE all LZD-R
MIC-range
MIC50
MIC90
32–>128
64
128
8–16
8
8
>128
>128
>128
32–>128
64
>128
8–32
16
32
>128
>128
>128
32–64
32
64
2.7. Cell viability assay
100 lg/ml (Fig. 6). Furthermore, the cell viability is not affected
by 4aR enantiomer (Fig. 6). Whereas, both 4bS and 4bR induced
a slightly viability reduction (6del 12%) of HepG2 cells at the all
concentrations tested (Fig 6), which reached statistically signifi-
HepG2 cells are a suitable model of excretory system for drug
toxicity and targeting studies, because of their high degree of mor-
phological and functional differentiation in vitro. HepG2 were trea-
cance only after treatment with 4bR at 50 and 100
lg/ml. Treat-
ted with increasing concentrations (25–100
lg/ml) of both 4a and
ment with linezolid at 100 g/ml induced about 20% cell death
l
4b enantiomers. Linezolid was used as reference drug at 100
lg/ml
(Fig 6), while no negative effects on cell viability were observed
final concentration. Moreover, since 4aS/R and 4bS/R compounds
were dissolved in DMSO, we evaluated the potential cytotoxicity
of the solvent alone at the high percentage (0, 9%) employed. The
4aS enantiomer has a mild concentration-independent cytotoxic
effect on HepG2 cells. Indeed, although treatment with 4aS com-
pound lead to a statistically significant decrease in cell viability
after DMSO 0, 9% treatment (data not shown).
3. Conclusions
The effects of side chain modification and chirality in linezolid-
like 1,2,4-oxadiazoles have been studied to discover new
potent antibacterials against Gram-positive multidrug-resistant
at 25
lg/ml (p <0.01), it seems to not exert any effect at 50 and
120
100
**
**
**
**
80
60
40
20
0
0
25
50
100
g/mL
μ
4aS
4aR
4bS
4bR
Linezolid
Figure 6. Cell viability of 4aS/R, 4bS/R and Linezolid in HepG2 cells. ⁄ = p <0.05; ⁄⁄ = p <0.01.

6822
C.G. Fortuna et al. / Bioorg. Med. Chem. 22 (2014) 6814–6825
pathogens. The adopted strategy involved molecular modelling,
the synthesis and biological evaluation of the resolved enantiomers
and absolute configuration assignment.
4.1.1.2. 3-(3-Fluoro-4-(3-methyl-1,2,4-oxadiazol-5-yl)-phenyl)-
5-(methyliden)-oxazolidin-2-one (1b). White solid (0.12 g,
84%); mp 205–208 °C; IR (Nujol)
1758, 1657 cmꢀ1 1H NMR
m
;
Docking studies show the presence of a –NHCH3 moiety in C-
5 in 4bS and 4aS likely to be responsible of a higher affinity of
these compounds for the ribosome with respect to linezolid. A
rotation of these compounds generates a new binding pose in
which both thiouric hydrogens from C5 side chain hydrogen
bond both oxygens from the phosphate backbone of the ribo-
some base U2504, accounting for the activity of these com-
pounds versus Linezolid-resistant S. aureus strains. Experiments
confirmed the computational prediction of enantiomer S being
the only active compounds probably due to the hypothesized
steric hindrance.
Experimental evaluation of the antibacterial activity of the
designed 4aS and 4bS against linezolid-susceptible S. aureus strains
showed that their potency is higher than that of linezolid, while
eukaryotic cell viability assay show that their cytotoxic power is
lesser than linezolid. That being so, future tests will be important
to evaluate the inhibitory effect on mitochondrial protein synthesis
(as known to be associated to oxazolidinones) of these promising
compounds in different cell lines representative of the mammalian
excretory system.
(300 MHz; CDCl3) d 2.49 (s, 3H), 4.53–4.62 (m, 1H), 4.67–4.69
(m, 2H), 4.93–4.95 (m, 1H), 7.41 (d, J1 = 10.1 Hz, J2 = 2.0 Hz, 1H),
7.69 (dd, J1 = 13.3 Hz, J2 = 2.0 Hz, 1H), 8.11 (t, J = 10.1 Hz, 1H). Anal.
Found (Calcd) for C13H10FN3O3 (%): C, 56.67 (56.73); H, 3.70 (3.66);
N, 15.31 (15.27)
4.1.2. General procedure for the preparation of compounds 2a,b
To a solution of 0.65 mmol of compound 10a or 10b in THF
(6 ml) was added KSCN (1.01 g, 10.42 mmol). The solution was stir-
red for 2 days at room temperature. The solvent was then removed
under vacuum. The residue was purified by chromatography yield-
ing the corresponding compounds 2a and 2b, respectively.
4.1.2.1. 3-(4-(3-Methyl-1,2,4-oxadiazol-5-yl)-phenyl)-5-(isothio-
cyanatomethyl)-oxazolidin-2-one (2a).
White solid (0.14 g,
67%); mp 166.4–173.9 °C; IR (Nujol)
m
1750, 1620 cmꢀ1
;
1H NMR
(300 MHz; CDCl3) d 2.48 (s, 3H), 3.30–3.45 (m, 2H), 4.02 (dd,
J1 = 9.6 Hz, J2 = 6.3 Hz, 1H), 4.37 (dd, J1 = 9.6 Hz, J2 = 9.0 Hz, 1H);
5.04–5.08 (m, 1H), 7.72 (d, J = 8.7 Hz, 2H), 8.15 (d, J = 8.7 Hz, 2H);
Anal. Found (Calcd) for C14H12N4O3S (%): C, 53.20 (53.16); H, 3.75
(3.82); N, 17.80 (17.71)
Interestingly cell viability assay performed in HepG2 cells
seems to indicate that these compounds are less cytotoxic than lin-
ezolid. Therefore the design and the synthesis of the new com-
pounds reported in this work provide new weapons in the battle
against linezolid resistant strains, which might represent an
important public health issue in a not too distant future.
4.1.2.2. 3-(3-Fluoro-4-(3-methyl-1,2,4-oxadiazol-5-yl)-phenyl)-
5-(isothiocyanatomethyl)-oxazolidin-2-one (2b).
White solid
(0.15 g, 68%); mp 199.0–201 °C; IR (Nujol)
m
1748, 1638 cmꢀ1
;
1H
NMR (300 MHz; CDCl3) d 2.51 (s, 3H), 3.30–3.46 (m, 2H), 4.01 (dd,
J1 = 9.3 Hz, J2 = 6.0 Hz, 1H), 4.36 (t, J = 9.3 Hz, 1H), 5.10 (m, 1H),
7.41 (dd, J1 = 8.7 Hz, J2 = 2.1 Hz, 1H), 7.69 (dd, J1 = 12.6 Hz,
J2 = 1.8 Hz, 1H), 8.11 (t, J = 8.7 Hz, 1H); Anal. Found (Calcd) for C14-
H11FN4O3S (%): C, 50.36 (50.30); H, 3.37 (3.32); N, 16.69 (16.76).
4. Experimental section
4.1. Materials and methods
4.1.3. General procedure for the preparation of compounds 3a,b
To a solution of 0.30 mmol of compound 2a or 2b in CH3OH
(5 ml) was added saturated ammonia methanolic solution (2 ml).
The solution was stirred for 2 h at 70 °C. The solvent was then
removed under vacuum. The residue was purified by chromatogra-
phy yielding the corresponding compounds 3a and 3b.
Melting points were determined on a Reichart-Thermovar hot-
stage apparatus and are uncorrected. FT-IR spectra were registered
in Nujol mull with a Shimadzu FTIR-8300 instrument. 1H NMR and
13C NMR spectra were recorded on a Bruker 300 Avance spectrom-
eter at indicated frequencies by using residual solvent peak as ref-
erence. Additional 1H NMR and ROESY (mixing time 0.6 s) spectra
were recorded at 600 MHz using an Agilent Inova 600 spectrome-
ter equipped with a triple resonance reverse probe with z-gradi-
4.1.3.1. 1-((3-(4-(3-Methyl-1,2,4-oxadiazol-5-yl)phenyl)-oxazoli-
din-2-on-5-yl)methyl)-3-thiourea (3a).
White solid (0.03 g, 30%);
ents. CD spectra were recorded with
a
Jasco J-715
mp 218.0–220.5 °C; IR (Nujol)
m
1759, 1617 cmꢀ1; 1H NMR (300 MHz;
spectropolarimeter, and UV spectra with a Jasco V-650 spectropho-
tometer, on ꢂ3 mM acetonitrile solutions using 0.01–0.02 cm cells.
Flash chromatography was performed by using silica gel (0.040–
0.063 mm) and mixtures of ethyl acetate and petroleum ether
(fraction boiling in the range of 40–60 °C) in various ratios. All
reagents were commercial. Compounds 1–5a,b were obtained as
previously reported.23
DMSO) d 2.39 (s, 3H), 3.24–3.37 (m, 2H), 3.94 (dd, J1 = 6.0 Hz,
J2 = 9.0 Hz, 1H), 4.31 (dd, J1 = 9.0 Hz, J2 = 9.0 Hz, 1H), 5.01 (dd,
J1 = 6.0 Hz, J2 = 9.0 Hz, 1H), 7.78 (d, J = 9.0 Hz, 2H), 8.08 (d, J = 9.0 Hz,
2H); Anal. Found (Calcd) for C14H15N5O3S (%): C, 50.51 (50.44); H,
4.60 (4.54); N, 21.12 (21.01).
4.1.3.2. 1-((3-(3-Fluoro-4-(3-methyl-1,2,4-oxadiazol-5-yl)phe-
nyl)-oxazolidin-2-one-5-yl)methyl)-3-thiourea (3b). White solid
(0.03 g, 30%); mp 220–227 °C; IR (Nujol)
NMR (300 MHz, CDCl3) d 2.53 (s, 3H), 3.19–3.30 (m, 2H), 3.91–4.00
(m, 1H), 4.30 (dd, J1 = J2 = 13.5 Hz, 1H), 5.01–5.11 (m, 1H), 7.45 (dd,
J1 = 13.2 Hz, J2 = 3.3 Hz 1H), 7.69 (dd, J1 = 19.5 Hz, J2 = 3.3 Hz, 1H),
8.15 (t, J = 12.0 Hz, 1H); Anal. Found (Calcd) for C14H14FN5O3S (%): C,
47.78 (47.86); H, 3.98 (4.02); N, 20.01 (19.93).
4.1.1. General procedure for the preparation of compounds 1a,b
To a solution of 0.52 mmol of compound 10a or 10b in CH3CN
(5 ml) was added K2CO3 (0.36 g; 2.60 mmol). The solution was stir-
red for 3 days at room temperature. The solvent was then removed
under vacuum. The residue was purified by chromatography yield-
ing the corresponding compounds 1a and 1b.
m ;
1759, 1617 cmꢀ1 1H
4.1.1.1. 3-(4-(3-Methyl-1,2,4-oxadiazol-5-yl)-phenyl)-5-(methyl-
4.1.4. General procedure for the preparation of compounds 4a,b
To a solution of 0.55 mmol of compound 12a or 12b in THF
(5 ml) was added CH3NCS (0.041 ml; 0.60 mmol) and triethyl-
amine (0.084 ml; 0.60 mmol). The solution was stirred for 3 hours
at room temperature. The solvent was then removed under
vacuum. The residue was purified by chromatography yielding
the corresponding compounds 4a and 4b.
iden)-oxazolidin-2-one (1a).
White solid (0.09 g, 68%); mp
176–178 °C; IR (Nujol)
m
1758, 1657 cmꢀ1
;
1H NMR (300 MHz;
CDCl3) d 2.43 (s, 3H), 4.47 (dd, J1 = 3 Hz, J2 = 9 Hz 1H), 4.67–4.69
(m, 2H), 4.88 (dd, J1 = 3 Hz, J2 = 9 Hz 1H), 7.68 (d, J1 = 9 Hz, 2H),
8.08 (dd, J1 = 9 Hz, 2H); Anal. Found (Calcd) for C13H11N3O3 (%): C,
60.80 (60.70); H, 4.35 (4.31); N, 16.36 (16.33).

C.G. Fortuna et al. / Bioorg. Med. Chem. 22 (2014) 6814–6825
6823
4.1.4.1. 1-((3-(4-(3-Methyl-1,2,4-oxadiazol-5-yl)phenyl)-oxazoli-
din-2-one-5-yl)methyl)-3-methylthiourea (4a). White solid
3364, 1732 cmꢀ1
(0.5 ns long) of complexes have been performed with the Sander
module of AMBER10.48,49,50 Restrictions over atoms farther than
20Å from binding site have been applied.
(0.15 g, 80%); mp 189.4–191.8 °C; IR (Nujol)
m
;
1H NMR (300 MHz; CDCl3) d 2.39 (s, 3H), 2.82 (bs, 3H), 3.82–4.00
(m, 3H), 4.20 (dd, J1 = 8.7 Hz, J2 = 6.0 Hz, 1H), 4.91 (bs, 1H), 7.77–
7.80 (m, 3H), 8.09 (d, J = 6.9 Hz, 2H); Anal. Found (Calcd) for C15H17-
N5O3S (%): C, 51.91 (51.86); H, 5.00 (4.93); N, 20.20 (20.16).
4.3. Enantiomers separation
4.3.1. Chiral chromatography
Analytical liquid chromatography was performed on a chro-
matograph equipped with a 20 lL loop injector, and UV detector.
The enantiomers were well resolved by enantioselective HPLC
using chiral stationary phase on Daicel Chiralpak IA column
(250 mm, 4.6 mm, particle size 5 lm) under normal phase condi-
tions with n-hexane/isopropanol 70/30 as eluent (flow rate 1 ml/
min and T 25 °C) and UV detection at 270 nm. Semipreparative
separation was performed by multiple injections (1 mg of sample
4.1.4.2.1-((3-(3-Fluoro-4-(3-methyl-1,2,4-oxadiazol-5-yl)phenyl)-
oxazolidin-2-one-5-yl)methyl)-3-methylthiourea (4b). White solid
(0.18 g, 88%); mp 170.7–172.4 °C; IR (Nujol)
m ;
3370, 1739 cmꢀ1 1H
NMR (300 MHz, DMSO) d 2.48 (s, 3H), 2.89 (bs, 3H), 3.89–4.07 (m,
3H), 4.24–4.30 (m, 1H), 4.89 (bs, 1H), 7.74 (s, 1H), 7.79 (dd,
J1 = 13.5 Hz, J2 = 2.1 Hz, 2H), 8.20 (t, J = 9.0 Hz, 2H); Anal. Found
(Calcd) for C15H16FN5O3S (%): C, 49.21 (49.31); H, 4.35 (4.41); N,
19.10 (19.17).
each run) with a 500 lL loop injector, each separated by thorough
washing cycles of the column. The sample was dissolved in the
mobile phase. Each enantiomer was obtained with ee values
greater than 96%.
Under the same analytical HPLC conditions performed on race-
mic and S-Linezolid the latter was revealed as the first eluting
enantiomer. This data allowed to correlate the relative configura-
tions of 4a and 4b first eluting peaks with that of S-linezolid.
4.1.5. General procedure for the preparation of compounds 5a,b
In a glass tube, to 0.45 mmol of compound 10a or 10b was
added 1,2,3-triazole (0.124 g; 1.8 mmol). The mixture was heated
until complete consumption of the starting material monitored
by TLC. The residue was purified by chromatography yielding the
corresponding compounds 5a and 5b.
4.3.2. Compound 4a
4.1.5.1. ((3-(4-(3-Methyl-1,2,4-oxadiazol-5-yl)phenyl)-oxazoli-
Retention time of first eluting peak: tr = 11.40 min; ee >98%.
Retention time of second eluting peak: tr = 20.70 min; ee = 96%.
din-2-on-5-yl)methyl)-4,5-dihydro-1H-1,2,3-triazole (5a). White
solid (0.11 g, 73%); mp 208–210 °C; IR (Nujol)
NMR (300 MHz; CDCl3) d 2.46 (s, 3H), 4.03 (dd, J1 = 6.3 Hz,
J2 = 9.3 Hz, 1H), 4.25 (dd, J1 = 9.3 Hz, J2 = 9.0 Hz, 1H), 4.82–4.83
(m, 2H), 5.08–5.14 (m, 1H), 7.59 (d, J = 9.0 Hz, 1H), 7.75 (s, 1H),
7.80 (s, 1H), 8.08 (d, J = 9.0 Hz, 1H); Anal. Found (Calcd) for
m ;
1751 cmꢀ1 1H
a
= 1.81.
4.3.3. Compound 4b
Retention time of first eluting peak: tr = 12.99 min; ee >98%.
Retention time of second eluting peak: tr = 22.69 min; ee = 93%.
C15H14N6O3 (%): C, 55.30 (55.21); H, 4.39 (4.32); N, 25.69 (25.75).
a
= 1.74.
4.1.5.2. ((3-(3-Fluoro-4-(3-methyl-1,2,4-oxadiazol-5-yl)phenyl)-
oxazolidin-2-on-5-yl)methyl)-4,5-dihydro-1H-1,2,3-triazole
4.3.4. Racemic LINEZOLID
(UV detection at 255 nm).
Retention time of first eluting peak (S): tr = 14.19 min.
Retention time of second eluting peak (R): tr = 16.34 min.
(5b).
m
White solid (0.10 g, 64%); mp 176.2–177.8 °C; IR (Nujol)
;
1751 cmꢀ1
1H NMR (300 MHz; CDCl3) d 2.48 (s, 3H), 4.03 (dd,
J1 = 9.3 Hz, J2 = 6.0 Hz, 1H), 4.25 (dd, J1 = 9.6 Hz, J2 = 9.0 Hz, 1H),
4.82–4.83 (m, 2H), 5.15–5.30 (m,1H), 7.27 (dd, J1 = 8.3 Hz,
J2 = 1.8 Hz, 1H), 7.56 (dd, J1 = 12.6 Hz, J2 = 1.8 Hz, 1H), 7.75 (s,
1H), 7.79 (s, 1H), 8.02 (t, J = 8.3 Hz, 1H); Anal. Found (Calcd) for
a
= 1.15.
4.3.5. S-LINEZOLID
(UV detection at 255 nm).
tr = 14.19 min.
C15H13FN6O3 (%): C, 52.37 (52.33); H, 3.85 (3.81); N, 24.47 (24.41).
4.2. Computational methods
4.4. Assignment of enantiomers absolute configuration
The computational tools employed in this work are mainly part
of FLAP package.34 In this paper, FLAP was used in the structure-
based mode and the template is the protein cavity obtained remov-
ing the linezolid molecule from the 3DLL X-ray structure. FLAP
compares the Molecular Interaction Fields (MIFs) of the cavity to
those of the test compounds ranking the poses according to a sim-
ilarity score. The probes used in this analysis to generate the MIFs
were the default probes DRY, O, N1 and H.44
4.4.1. UV and CD Experiments
The absorption UV and ECD spectra of the two enantiomers of
4a were measured on 3 mM acetonitrile solutions using 0.01–
0.02 cm cells.
4.4.2. Conformational searches and geometry optimizations
S
The model of wild type S. aureus 23S ribosomal rRNA (genebank
id: X68425) has been made with modeRNA45 using as a template
the 23S subunits of the 50S rRNA subunit of D. radiodurans (pdb
ID:3dll). Compounds 4a-S and 4b-S, as the most representative
compounds of this study, have been drawn with MOE.46 Minimiza-
tion and a conformational analysis of the compounds have been
performed with the same program. The complexes of S. aureus
23S ribosomal rRNA with drawn drugs 4a,b have been made taking
in consideration the similarity between designed compounds and
Linezolid, thus docking have been made manually, superposing
the best conformations over Linezolid, using PyMOL.47 The general
Amber force field46 and HF/6-31G⁄-derived RESP atomic charges
were used for the ligands. Minimizations and Molecular Dynamics
CH3
N
H
N
H
N
O
O
An initial geometry of 4a was built using arbitrary (R) configu-
ration. Conformational searches were run with the Monte Carlo
algorithm implemented in Spartan’10 using molecular mechanics
(Merck Molecular Force Field, MMFF). To speed up the process,
the molecule was first cut at the phenyl-oxadiazole bond (C40-
C100) to obtain the simplified model shown aside. In the Monte Car-
lo procedure, all rotatable bonds were allowed to vary systemati-

6824
C.G. Fortuna et al. / Bioorg. Med. Chem. 22 (2014) 6814–6825
cally, and the five-membered ring was allowed to pucker. All
MMFF minima thus obtained were optimized with DFT method
using M06 functional and 6-31G(d) basis set. A set of 7 low-energy
structures was obtained within 5 kcal/mol from the lowest energy
structure. Of these, the two most stable structures have the substi-
tuent at C5 oriented axially with the thiourea moiety folded over
the carbamate one. The third minimum, having the same substitu-
ent oriented axially but in an extended conformation (i.e., pointing
away from the carbamate), is over 3.5 kcal/mol less stable than the
lowest-energy minimum. These 3 minima were employed to
reconstitute the original molecule by attaching the oxadiazole ring
in the two possible orientations, and re-optimizing the entire
structure. Thus, a set of 6 geometries was obtained reported in
Table 2.
Cell viability assay: Cells were seeded at 40,000cell/cm2 onto col-
lagen coated 24-well plates, and after two days treated with both
4a and 4b enantiomers with increasing concentrations (25–100g/
ml). After 48 h treatment, cell viability was evaluated with Presto-
BlueÒ Cell Viability Reagent, a cell permeable resazurin-based solu-
tion that modified by the reducing power of living cells turns red in
color. Briefly, PrestoBlueÒ was administrated to cells according to
the manufacturer’s protocol and incubated for 1 h at 37 °C. The
absorbance was measured with FLUOstar Omega (BMG Labtech)
multidetection microplate reader at wavelength of 570 nm and at
reference wavelength of 600 nm. All data are expressed as mean
SEM of 3 separate experiments performed in triplicate. The dif-
ferences were calculated by means Student’s t test. A p value
<0.05 was considered to be statistically significant.
4.4.3. NMR experiments
Acknowledgements
1H NMR and ROESY (mixing time 0.6 s) spectra were recorded at
600 MHz on a CD3CN solution of a racemic sample of 4a.
Compound 4a 1H NMR (600 MHz, CD3CN, 25 °C): d 2.40 (s, 3H,
ArCH3), 2.89 (bs, 3H, NH CH3), 3.97 (m, 3H, H6a, H6b, H4a), 4.14 (t,
1H, J = 9.3 Hz, H4b), 4.92 (m, 1H, H5), 6.49 (bs, 1H, CH2NH), 6.61 (bs,
1H, C@SNH), 7.76 (d, 2H, J = 8.6 Hz, H20), 8.11 (d, 2H, J = 8.6 Hz, H30).
Financial supports from the Italian MIUR within the ‘FIRB-Futu-
ro in Ricerca 2008’ Program—Project n. RBFR08A9V1 (to A.P., C.G.F.
and R.M.) and from the University of Catania for a post-doctoral
Fellowship (to L.G.) are gratefully acknowledged.
Supplementary data
4.4.4. CD calculations
Excited-state (ECD) calculations were run on the 6 geometries
described above, with (R) configuration, using the TDDFT method
with the CAM-B3LYP functional and TZVP basis sets. In one case,
a basis set augmented with diffuse functions (ma-TZVP) was also
employed to check the consistency with the non-augmented TZVP.
The first 30 excited states were computed in all cases. The calcu-
lated rotational strengths (dipole-length gauge) were converted
in ECD spectra by applying a Gaussian bandshape with 0.5 eV
exponential half-width. The component spectra are shown in
Figure 4. The spectra of the most stable structures were averaged
using the Boltzmann populations estimated at 300 K from M06/
6-31G(d) internal energies. The Boltzmann-weighted spectra is
shown in Figure 5.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmc.2014.10.037.
References and notes
1. Freire-Moran, L.; Aronsson, B.; Manz, C.; Gyssens, I. C.; So, A. D.; Monnet, D. L.;
Cars, O. Drug Resist. Updat. 2011, 14, 118.
2. Dryden, M. S. J. Antimicrob. Chemother. 2011, 66, 7.
3. Tsiodras, S.; Gold, H. S.; Sakoulas, G.; Eliopoulos, G. M.; Wennersten, C.;
Venkataraman, L.; Moellering, R. C.; Ferraro, M. J. Lancet 2001, 358, 207.
4. Auckland, C.; Teare, L.; Cooke, F.; Kaufmann, M. E.; Warner, M.; Jones, G.;
Bamford, K.; Ayles, H.; Johnson, A. P. J. Antimicrob. Chemother. 2002, 50, 743.
5. Seedat, J.; Zick, G.; Klare, I.; Konstabel, C.; Weiler, N.; Sahly, H. Antimicrob.
Agents Chemother. 2006, 50, 4217.
6. Vara Prasad, J. V. N. Curr. Opin. Microbiol. 2007, 10, 454.
7. Hutchinson, D. K. Curr. Top. Med. Chem. 2003, 3, 1021.
8. Gravestock, M. B. Opin. Drug Discovery Dev. 2005, 8, 469.
9. Renslo, A. R.; Luehr, G. W.; Gordeev, M. F. Med. Chem. 2006, 14, 4227.
10. Poce, G.; Zappia, G.; Porretta, G. C.; Botta, B.; Biava, M. Exp. Opin. Ther. Patents
2008, 18, 97.
11. Im, W. B.; Choi, S. H.; Park, Ju-Y.; Choi, S. H.; Finn, J.; Yoon, Sung-H Eur. J. Med.
Chem. 2011, 46, 1027.
12. Jo, Y. W.; Im, W. B.; Rhee, J. K.; Shim, M. J.; Kim, W. B.; Choi, E. C. Bioorg. Med.
Chem. 2004, 12, 5909.
13. Barbachyn, M. R.; Ford, C. W. Angew. Chem., Int. Ed. 2003, 42, 2010.
14. Jang, S.-Y. et al Bioorg. Med. Chem. Lett. 2004, 14, 3881.
15. Locke, J. B.; Finn, J.; Hilgers, M.; Morales, G.; Rahawi, S.; Kedar, G. C.; Picazo, J. J.;
Im, W.; Shaw, K. J.; Stein, J. L. Antimicrob. Agents Chemother. 2010, 54, 5337;
Phillips, O. A. Bioorg. Med. Chem. 2003, 35, 1156. Gravestock, M. B.; 2003, 13,
4179.
16. Bostroem, J.; Hogner, A.; Llinas, A.; Wellner, E.; Plowright, A. T. J. Med. Chem.
2012, 55, 1817.
17. Terenzi, A.; Barone, G.; Palumbo Piccionello, A.; Giorgi, G.; Guarcello, A.;
Portanova, P.; Calvaruso, G.; Buscemi, S.; Vivona, N.; Pace, A. Dalton Trans. 2010,
39, 9140.
18. Terenzi, A.; Barone, G.; Palumbo Piccionello, A.; Giorgi, G.; Guarcello, A.; Pace,
A. Inorg. Chim. Acta 2011, 373, 62.
19. Mandracchia, D.; Palumbo Piccionello, A.; Pitarresi, G.; Pace, A.; Buscemi, S.;
Giammona, G. Macromol. Biosci. 2007, 7, 836.
4.5. Microbiological assays
4.5.1. Bacterial strains
A total of 51 (4 MSSA and 47 MRSA) well characterized for their
antibiotic-susceptibility phenotype Staphylococcus aureus isolates
and 10 linezolid-resistant methicillin-resistant Staphylococcus
epidermidis (MRSE) were used for the determination of the
in vitro antibacterial activity.
4.5.2. Determination of minimum inhibitory concentrations
(MICs)
The in vitro antibacterial activity was studied by determining
their minimum inhibitory concentrations (MICs) by means of the
broth microdilution method according to the Clinical and Labora-
tory Standards Institute (CLSI) guidelines51 according to a previ-
ously described procedure.22,23 The in vitro antibacterial
activities were tested and compared to that of linezolid (Sigma–
Aldrich, Italy) used as a reference.
20. Palumbo Piccionello, A.; Guarcello, A.; Calabrese, A.; Pibiri, I.; Pace, A.; Buscemi,
S. Org. Biomol. Chem. 2012, 10, 3044.
21. Palumbo Piccionello, A.; Pitarresi, G.; Pace, A.; Triolo, D.; Picone, P.; Buscemi, S.;
Giammona, G. J. Drug Target. 2012, 20, 433.
22. Palumbo Piccionello, A.; Musumeci, R.; Cocuzza, C.; Fortuna, C. G.; Guarcello,
A.; Pierro, P.; Pace, A. Eur. J. Med. Chem. 2012, 50, 441.
23. Fortuna, C. G.; Bonaccorso, C.; Bulbarelli, A.; Caltabiano, G.; Rizzi, L.; Goracci, L.;
Musumarra, G.; Pace, A.; Palumbo-Piccionello, A.; Guarcello, A.; Pierro, P.;
Cocuzza, C. E.; Musumeci, R. Eur. J. Med. Chem. 2013, 65, 533.
24. Cruciani, G.; Milletti, F.; Storchi, L.; Sforna, G.; Goracci, L. Chem. Biodivers. 2009,
6, 1812.
4.5.3. Determination of cell viability
Cell cultures conditions: The effects of 4aS/R, 4bS/R and linezolid
on cells viability were studied in vitro on HepG2 cells, (hepatocel-
lular carcinoma). HepG2 were grown in Dulbecco’s modified eagles
medium (DMEM) supplemented with 10% heat inactivated foetal
bovine serum (FBS), 2 mM L-glutamine, 100 units/ml penicillin
and 100 lg/ml streptomycin. Cells were maintained on collagen
25. Cruciani, G.; Crivori, P.; Carrupt, P. A.; Testa, B. J. Mol. Struct. (Theochem) 2000,
503, 17.
26. Pace, A.; Pierro, P. Org. Biomol. Chem. 2009, 7, 4337.
coated 75-cm2 flask at 37 °C in a 5% CO2 atmosphere. All reagents
for cell culture were bought from Euroclone, Pero, Italy.

C.G. Fortuna et al. / Bioorg. Med. Chem. 22 (2014) 6814–6825
6825
27. Buscemi, S.; Pace, A.; Calabrese, R.; Vivona, N.; Metrangolo, P. Tetrahedron
2001, 57, 5865.
40. Cross, S.; Ortuso, F.; Baroni, M.; Costa, G.; Distinto, S.; Moraca, F.; Alcaro, S.;
Cruciani, G. J. Chem. Inf. Model. 2012, 52, 2599.
28. Buscemi, S.; Pace, A. Heterocycles 2004, 63, 1619.
29. Palumbo Piccionello, A.; Pace, A.; Pibiri, I.; Buscemi, S.; Vivona, N. Tetrahedron
2006, 62, 8792.
30. Palumbo Piccionello, A.; Pace, A.; Pierro, P.; Pibiri, I.; Buscemi, S.; Vivona, N.
Tetrahedron 2009, 65, 119.
31. Broccolo, F.; Cainelli, G.; Caltabiano, G.; Cocuzza, C. E. A.; Fortuna, C. G.; Galletti,
P.; Giacomini, D.; Musumarra, G.; Musumeci, R.; Quintavalla, A. J. Med. Chem.
2006, 49, 2804.
32. Ippolito, J. A.; Kanyo, Z. F.; Wang, D.; Franceschi, F. J.; Moore, P. B.; Steitz, T. A.;
Duffy, E. M. J. Med. Chem. 2008, 51, 3353.
33. Wilson, D. N.; Schluenzen, F.; Harms, J. M.; Starosta, A. L.; Connell, S. R.; Fucini,
P. PNAS 2008, 105, 13339.
34. Baroni, M.; Cruciani, G.; Sciabola, S.; Perruccio, F.; Mason, J. S. J. Chem. Inf.
Model. 2007, 47, 279.
35. Cross, S.; Baroni, M.; Carosati, E.; Benedetti, P.; Clementi, S. J. Chem. Inf. Model.
2010, 50, 1442.
41. Muratore, G.; Mercorelli, B.; Goracci, L.; Cruciani, G.; Digard, P.; Palù, G.;
Loregian, A. Antimicrob. Agents Chemother. 2012, 56, 6009.
42. Long, K. S.; Vester, B. Antimicrob. Agents Chemother. 2012, 56, 603.
43. http://dx.doi.org/10.1039/C1CS15036G http://dx.doi.org/10.1039/B515476F.
44. Kerns, Edward H.; Li Di Drug-like Properties. Concepts, Structure Design and
Methods: from ADME to Toxicity Optimization; Elsevier Academic Press:
Amsterdam, 2007.
45. Rother, M.; Rother, K.; Puton, T.; Bujnicki, J. M. Nucleic Acids Res. 2011, 39, 4007.
46. Chemical Computing Group, MOE 20011.10. Inc 2011.
47. Schrodinger, LLC The PyMOL Molecular Graphics System, Version 1.5r3 2010.
48. Case, D. A.; Darden, T. A.; Cheatham, T. E.; Simmerling, C. L.; Wang, J.; Duke, R.
E.; Luo, R.; Crowley, M.; Walker, R. C.; Zhang, W.; Merz, K. M.; Wang, B.; Hayik,
S.; Roitberg, A.; Seabra, G.; Kolossvary, I.; Wong, K. F.; Paesani, F.; Vanicek, J.;
Wu, X.; Brozell, S. R.; Steinbrecher, T.; Gohlke, H.; Yang, L.; Tan, C.; Mongan, J.;
Hornak, V.; Cui, G.; Mathews, D. H.; Seetin, M. G.; Sagui, C.; Babin, V.; Kollman,
P. A. Amber 10; University of California: San Francisco, 2008.
36. Simone Sciabola; Stanton, Robert V.; Mills, James E.; Flocco, Maria M.; Massimo
Baroni; Gabriele Cruciani J. Chem. Inf. Model. 2010, 50, 155.
37. Emanuele Carosati; Gabriele Cruciani; Alberto Chiarini; Roberta Budriesi;
Pierfranco Ioan; Raffaella Spisani; Domenico Spinelli; Barbara Cosimelli; Fabio
Fusi; Maria Frosini; Rosanna Matucci; Francesco Gasparrini; Alessia Ciogli;
Stephens, Philip J.; Devlin, Frank J. J. Med. Chem. 2006, 49, 5206.
38. Muratore, G.; Goracci, L.; Mercorelli, B.; Foeglein, A.; Digard, P.; Cruciani, G.;
Palù, G.; Loregian, A. PNAS 2012, 109, 6247.
49. Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D. A. J. Comput. Chem.
2004, 25, 1157.
50. Case, D. A.; Cheatham, T. E.; Darden, T.; Gohlke, H.; Luo, R.; Merz, K. M.;
Onufriev, A.; Simmerling, C.; Wang, B.; Woods, R. J. J. Comput. Chem. 2005, 26,
1668.
51. Clinical and Laboratory Standards Institute (CLSI). Methods for Dilution
Antimicrobial Susceptibility Tests for Bacteria That Grow Aerobically;
Approved Standard, 9th ed., M07-A9, 32, Wayne, Pennsylvenia, 2011.
39. Cross, S.; Baroni, M.; Goracci, L.; Cruciani, G. J. Chem. Inf. Model. 2012, 52, 2587.