Synthesis, SAR and molecular docking study of novel non-β-lactam inhibitors of TEM type β-lactamase

Article abstract of DOI:10.1016/j.bmcl.2017.02.025

The novel classes of acylated phenoxyanilide and thiourea compounds were investigated for their ability to inhibit TEM type β-lactamase enzyme. Two compounds 4g and 5c reveal the inhibition potency in micromolar range and show their action by non-covalent binding in the vicinity of the TEM-171 active site. The structure activity relationship around carbon chain length and different substituents in ortho- and para-positions of acylated phenoxyanilide as well as molecular modelling study has been performed.

Full text of DOI:10.1016/j.bmcl.2017.02.025

Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis, SAR and molecular docking study of novel non-b-lactam
inhibitors of TEM type b-lactamase
Roman L. Antipin a, ,1, Daria A. Beshnova b,1, Rostislav A. Petrov a, Anna S. Shiryaeva a, Irina P. Andreeva a,
⇑
Vitaly G. Grigorenko a, Maya Yu. Rubtsova a, Alexander G. Majouga a,c, Victor S. Lamzin b, Alexey M. Egorov a
a Chemistry Department, M.V. Lomonosov Moscow State University, Lenin Hills, 1, bld. 3, Moscow 119991, Russia
b European Molecular Biology Laboratory (EMBL), c/o DESY, Notkestrasse 85, Hamburg 22607, Germany
c National University of Science and Technology MISiS, Leninsky ave. 4, Moscow 119049, Russia
a r t i c l e i n f o
a b s t r a c t
Article history:
The novel classes of acylated phenoxyanilide and thiourea compounds were investigated for their ability
to inhibit TEM type b-lactamase enzyme. Two compounds 4g and 5c reveal the inhibition potency in
micromolar range and show their action by non-covalent binding in the vicinity of the TEM-171 active
site. The structure activity relationship around carbon chain length and different substituents in ortho-
and para-positions of acylated phenoxyanilide as well as molecular modelling study has been performed.
Ó 2017 Published by Elsevier Ltd.
Received 15 November 2016
Revised 10 February 2017
Accepted 13 February 2017
Available online xxxx
Keywords:
TEM type b-lactamase
b-Lactamase inhibitors
Molecular docking
Acylated phenoxyaniline
Thiourea
Since the discovery of penicillin in 1928, b-lactam compounds
have become one of the largest classes of antibiotics used world-
wide for the treatment of a variety of infectious diseases.1 How-
ever, in recent years, their effectiveness has been diminished by
the proliferation of bacterial species unsusceptible to these antibi-
otics. Antimicrobial resistance is a serious and growing threat to
public health and has been observed for almost all known antibi-
otic agents today, with a large number of them becoming ineffec-
tual.2–4 The increase in international trade and human mobility has
facilitated the widespread dissemination of such species.5 b-Lac-
tams, such as penicillins, cephalosporins, carbapenems and
monobactams are the most widely prescribed class of antibiotics
in the world today.6 This widespread use – and indeed misuse,
arising from inappropriate prescribing and poor control of distribu-
tion – has contributed to the growth of b-lactam resistance.7 The
most prevalent resistance mechanism is the production of b-lacta-
mases.8,9 These enzymes hydrolyse the antibiotic agents at their
strained lactam bond, thereby inactivating them. Lactamase inhibi-
tors have been co-administered with b-lactam antibiotics for a
number of years, and their inhibitory action results in the reduc-
tion of antibiotic degradation and improved therapeutic effects.10
The majority of these inhibitors are b-lactams and their presence
in the biosphere for many years indicates that enzymes naturally
resistant to them may already have evolved. Even if not, the devel-
opment of resistant species may require only a few mutations and
may arise rapidly upon therapeutic use.11
There is a pressing need for new inhibitors of b-lactamases,
preferably not based on the established b-lactam paradigm. A
number of novel, non-b-lactam inhibitors of various lactamase iso-
forms have been discovered.11–13 Some of these bind to the
enzyme active site while others interact elsewhere and possibly
inhibit the enzyme function allosterically.14,15 The latter mecha-
nism may be a particularly interesting route to finding improved
inhibitors. However, it is clear that novel non-lactam inhibitors
will require significant efforts in laboratory-based screenings,
structural examinations and computational analyses.
We aim to investigate non-b-lactam inhibitors of b-lactamases
for their complementarity to the structure of TEM type b-lacta-
mase from Escherichia coli. A novel acylated phenoxyaniline com-
pound showing competitive and reversible inhibition of TEM-171
b-lactamase and hypothetically bound in the vicinity of enzyme
active site was recently described.16 Here, we report on the use
of acylated phenoxyanilide chemical scaffold for a design of new
potent TEM-171 b-lactamase inhibitors.
⇑
Corresponding author.
E-mail address: antipin@org.chem.msu.ru (R.L. Antipin).
1
These authors contributed equally.
http://dx.doi.org/10.1016/j.bmcl.2017.02.025
0960-894X/Ó 2017 Published by Elsevier Ltd.
Please cite this article in press as: Antipin R.L., et al. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.02.025

2
R.L. Antipin et al. / Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
The structure of acylated phenoxyaniline (4g) is shown in
R2
R2
OH
F
O
Table 1 and the numbering reflects the synthesis steps presented
in Scheme 1. The inhibitor contains two chlorine atoms in ortho-
and para-positions, and two methylene groups in the side carbon
chain. We synthesised a number of analogues of 4g, investigated
their inhibitory potential, and carried out a molecular modelling
study in order to find compounds with improved affinity and struc-
ture-activity relationships (SAR). We investigated the effect of dif-
ferent substituents at a benzene ring in acylated phenoxyanilide
and the variation of its carbon chain length on b-lactamase
inhibition.
i
ii
+
R1
O2N
R1
NO2
1a-d
R2
2a-d
2a: R1=H, R2=H; 2b: R1=Cl, R2=H;
2c: R1=Br, R2=H; 2d: R1=Cl, R2=Cl;
R2
O
O
iii
O
O
R1
R1
NH2
NH (CH2)n OH
3a-d
4a-h
The synthesis of target compounds is outlined in Scheme 1. The
final compounds, 4a–i, were obtained in three steps.
4a: R1=H, R2=H, n=2; 4b: R1=H, R2=H, n=3; 4c: R1=Cl, R2=H, n=2;
4d: R1=Cl, R2=H, n=3; 4e: R1=Br, R2=H, n=2; 4f: R1=Br, R2=H, n=3;
4g: R1=Cl, R2=Cl, n=2; 4h: R1=Cl, R2=Cl, n=3;
iv, v
In the first step, 4-fluoronitrobenzene was reacted with differ-
ent phenols to obtain phenoxy-nitrobenzene compounds 2a–d
(Scheme 1). Second, nitro-compounds were reduced to the corre-
sponding amines, phenoxyanilines. In the third step, an acylation
was carried out. For this reaction, amines 3a–d were treated with
appropriate carboxylic acid anhydrides to obtain target com-
pounds 4a–i. The results were confirmed by spectral data. The IR,
1H NMR, mass spectra, and elemental analysis results were in
agreement with the obtained structures.
Cl
O
O
OH
Cl
NH
O
4i
Scheme 1. The synthesis of potential non-b-lactam inhibitors. (i = NaH/DMF/50 °C/
12 h, ii = N2H4*H2O/Ra-Ni/EtOH/60 °C/8 h, iii = acid anhydride/toluene/reflux/8 h,
iv = ethylchloroxalate/Et3N/THF, v = EtOH/H2O/HCl).
To investigate the effect of different lengths of the side carbon
chain and different substituents in ortho- and para-positions of
the phenoxy part, compounds 4a–i were evaluated for their ability
to inhibit TEM-171 b-lactamase (Table 1).
20 Å away from the catalytic Ser70 OH group. Both TEM-1 and
TEM-171 enzymes have very similar biochemical characteristics18
and thus we expect that the results of the docking on one enzyme
would also be applicable to the other one. The sulphate ion and all
water molecules were removed from the 1xpb model, hydrogen
atoms were added, and the structure was then minimised using
the OPLS-2005 force field of the Glide software.19 The site for the
docking was defined by the coordinates of the ligand 6-alpha-
hydroxymethylpenicillanate bound to the TEM-1 b-lactamase
(PDB ID 1TEM).
It was reported16 that compound 4g has moderate inhibition of
TEM-171 b-lactamase. Removing the chlorine atom in the ortho-
(compounds 4c, 4d) and also in the para-position (compounds
4a, 4b) almost completely eliminated its affinity to the enzyme.
However, removing the chlorine atom in the ortho-position and
substituting it in the para-position by bromine led to only a mod-
erate inhibition (compounds 4e, 4f). Compounds 4h, 4i with a dif-
ferent length of the side carbon chain also showed moderate
inhibition of the enzyme. A variation of the number of methylene
groups in this chain has almost no effect on the inhibitory poten-
tial. The presence of chlorine atoms in both ortho- and para-posi-
tions seems essential for good inhibition and none of the tested
substituents have shown improvement compared to the com-
pound 4g.
The 3D models of the compounds 4a–4i, 5b and 5c were pre-
pared using the LigPrep tool from the Schrödinger software19
,
which involved the generation of ionisation states at pH 7.0 2.0
as well as the generation of tautomers. The compounds were
docked into the TEM-1 structure using Glide SP scoring function.20
Water molecules bound to a protein are often displaced upon
ligand binding or mediate the formation of protein-ligand H-
bonds.21,22 The Glide extra-precision (XP) scoring function allows
to enumerate the energy contribution in protein-ligand docking
provided by a displacement of water molecules in the protein
active site and by applying desolvation penalties when polar or
charged groups cannot hydrogen bond effectively.23 Since Glide
To understand the inhibitor binding mechanism we employed
molecular docking. The crystal structure of TEM-171 b-lactamase
is not available and, therefore, the model of highly homologous
TEM-1 b-lactamase (PDB ID 1xpb) was used. The sequence of
TEM-1 differs from TEM-171 by only one single mutation,
Val84Ile18, which is located on the protein surface and is about
Table 1
The initial hydrolysis rates of the chromogenic substrate CENTA17 by TEM-171 b-lactamase in the presence of synthesised compounds 4a–i. Concentrations of the enzyme,
substrate and the inhibitors were 10 nM, 100
l
M and 100
lM, respectively.
Compound
R1
R2
n
V0
(l
M/s)
Relative activity
decrease (%)
Docking score
(kcal/mol)
R2
O
O
O
R1
NH (CH2)n OH
TEM-171
0.74 0.11
0.0
b-lactamase, control
4a
4b
4c
4d
4e
4f
4g
4h
4i
H
H
H
H
H
H
H
H
Cl
Cl
Cl
2
3
2
3
2
3
2
3
0
0.73 0.11
0.73 0.11
0.70 0.11
0.73 0.11
0.53 0.08
0.53 0.08
0.46 0.07
0.51 0.08
0.51 0.08
1.4 0.2
1.4 0.2
5.4 0.8
1.4 0.2
À3.9
À3.8
À5.1
À5.0
À4.8
À4.8
À5.3
À5.1
À5.0
Cl
Cl
Br
Br
Cl
Cl
Cl
28
28
38
31
31
4
4
6
5
5
Please cite this article in press as: Antipin R.L., et al. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.02.025

R.L. Antipin et al. / Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
3
XP mode has been designed for a refinement of good ligand
poses20, only top-scoring ligands were further docked using XP
mode. The docked models were visualised using Chimera.24 Fig. 1
shows the docking pose of compounds 4a–d with the lowest dock-
ing scores. The docking score (GlideScore) is an empirical scoring
function that approximates the ligand binding free energy. It is
based on the empirical ChemScore function25, but includes a
steric-clash term and buried polar terms devised by Schrödinger
to penalise electrostatic mismatches. A brief description of GlideS-
core is given in the Glide User Manual.20
All compounds were docked to the active site cavity, the dis-
tance between the O atom of the hydroxyl group of catalytic
Ser70 and the closest atom of the ligand C19 is 3.6 Å (Fig. 1A).
The protein complexes with compounds 4a and 4b are sta-
bilised by two H-bonds between the carboxyl group of the ligand
and the NH1/NH2 atoms of Arg275 and the main chain N atom
of Gly 242. The ligands 4c and 4d are docked to almost the same
location as 4a and 4b, but form an additional H-bond between
the O atom connecting two aromatic rings of the ligand and the
ND2 atom of Asn132. Some variation of the length of the side car-
bon chain in ligands 4a, 4b and 4c, 4d does not significantly affect
their docking.
Similarly to 4a–d, the compounds 4e–h also form two H-bonds
with the main chain NH groups of Arg275 and Gly242 (Fig. 2). In
addition, compounds 4f and 4h form an H-bond between the O
atom connecting two aromatic rings of the ligand and the NH
group of Asn132. We note that the Br atom of 4e, 4f participates
in the formation of a halogen bond with the OH group of Ser235.
Indeed, the C-BrÁ Á ÁO distance is 3.4 Å, the C-BrÁ Á ÁO angle is around
170°. When these bromine atoms are replaced by chlorine in com-
pounds 4h and 4i this halogen bond is no longer formed. Instead,
another halogen bond appears between the ligand chlorine atom
in the ortho- position and the carbonyl group of Asn170: the C-
ClÁ Á ÁO distance is 3.0 Å (4h) and 3.2 Å (4i), while the C-ClÁ Á ÁO angle
is 150° (4h) and 159° (4i). Additionally, the chlorine atom may par-
ticipate in the formation of an H-bond with hydroxyl group of
Ser70: the ClÁ Á ÁH distance is 2.7 Å and the C-ClÁ Á ÁH angle is 109°
Fig. 2. Docking of compounds to the active site of TEM-1 (A) compound 4e; (B)
compound 4f; (C) compound 4 h; (D) compound 4i. The hydrogen and halogen
bonds are shown by dashed black lines.
(4i). The C-XÁ Á ÁH (X = Cl, Br, I) contacts in biological systems are
characterised as weak H-bonding interactions. For a detailed inves-
tigation of the hydrogen bond acceptor capability of halogens the
reader is referred to.26
The observed dependence of the inhibition on the type of the
substituents in the ortho- and para-positions can be explained by
the formation of a halogen bond between the chlorine or bromine
atoms of the ligand and the oxygen atoms of the Ser235 and
Asn170 side chains. This is compatible with earlier studies where
halogen bonds were found to be important for modulation of inhi-
bition.27–33 We note that the presence of two chlorine atoms in the
inhibitor at the ortho- and para-positions provides the highest
inhibition of TEM-171 b-lactamase. At the same time, the investi-
gated variation of the length of the side carbon chain does not sig-
nificantly affect the inhibition.
Based on the results, we examined a further expansion of the
general scaffold of compound 4g towards thiourea compounds
5a–5c (Scheme 2), synthesised using a reaction of amines 3a–e
with isothiocyanates.
The pure compounds 5a–c were obtained as white solids. The
structures of compounds and their inhibitory activity data are
summarised in Table 2. The modelled binding of 5b, 5c is shown
in Fig. 3. The compounds form three similar H-bonds: between
the carboxyl group of the ligand and the main chain NH groups
of Arg275 and Gly242, as well as between the O atom connecting
two aromatic rings of the ligand and the carboxyl group of
Glu104. Compounds 5b and 5c form two weak H-bonds: C-
ClÁ Á ÁHH22-NH2 of Arg244 and C-ClÁ Á ÁHG-OG of Ser235.
R2
R2
O
vi
O
COOH
S
COOH
R1
NH2
R1
NH NH
SCN
3a, b, d
5a-c
5a: R1=H, R2=H; 5b: R1=Cl, R2=H; 5c: R1=Cl, R2=Cl;
Fig. 1. Docking of the compounds to the active site of TEM-1 b-lactamase. (A)
compound 4a; (B) compound 4b; (C) compound 4c; (D) compound 4d. The H-bonds
are shown by dashed black lines.
Scheme 2. The synthesis of thiourea compounds 5a–c.
Please cite this article in press as: Antipin R.L., et al. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.02.025

4
R.L. Antipin et al. / Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
Table 2
The initial hydrolysis rates of CENTA by TEM-171 b-lactamase in the presence of compounds 5a–c. The concentrations for the enzyme, substrate and the tested inhibitors were
10 nM, 100 mM and 100 mM, respectively.
Compound
R1
R2
V0
(lM/s)
Relative activity decrease (%)
Docking score (kcal/mol)
R2
O
COOH
S
R1
NH NH
TEM-171, control
0.74 0.11
0.59 0.09
0.53 0.11
0.37 0.07
0.0
5a
5b
5c
H
Cl
Cl
H
H
Cl
20.3 0.2
28.4 0.8
50.0 0.2
À5.8
À5.8
À5.5
Fig. 3. The docking of thiourea compounds to the active site of TEM-1 b-lactamase,
using Glide SP (yellow) and XP (cyan) scoring functions: (A) compound 5b; (B)
compound 5c. The H-bonds are shown by dashed black lines.
Fig. 4. Correlation between the docking score and the logarithm of inhibition rate
for all 12 compounds presented in Tables 1 and 2.
The distinguishing feature of compound 5c is that its second
chlorine atom forms three further weak H-bonds to the main chain
NH groups of Ala237, Ser70 and to the OH group of Ser70 (Fig. 3). A
high number of formed H-bonded contacts by the compound 5c
supports the observation of its highest inhibition potency (Ki = 48 -
development of this class of compounds. That may include incor-
porating additional H-bond acceptors to enable interactions with
N atom of Gly238 and/or Glu240.
l
M) among all compounds investigated in this work. In order to
take into account desolvation effects and possible water displace-
ment in the protein active site upon ligand binding, we further
refined molecular docking solution for top-scoring compounds
5b–5c (Table 2) using Glide XP mode. The binding poses predicted
by Glide SP and XP are very similar as shown in Fig. 3.
The docking results are consistent with our kinetic data
(Table 1). In addition, we observe dependence between the loga-
rithm of the decrease in the enzyme activity and the values of
the docking scores with a linear Pearson correlation coefficient of
69% (Fig. 4).
In conclusion, we used molecular docking and chemical synthe-
sis to rationalise the structure activity relationships of acylated
phenoxyanilide and thiourea classes of compounds as inhibitors
for TEM-class b-lactamase. The docking study allowed to propose
a binding mode for acylated phenoxyanilide and provided insights
for further inhibitor design. We found that the more hydrogen and/
or halogen bonds are formed between the ligand and the protein,
the higher the inhibition potency of the studied compounds. There-
fore, possible further modifications of the acylated phenoxyanilide
compounds may include incorporation of a -COOH group at the R1/
R2 positions in the ligand to enable the interactions with NH1 and
NH2 atoms of Arg244. We can also propose introducing additional
H-bond acceptors in the ligand’s side carbon chain to form interac-
tions with NH1/NH2 atoms of Arg275 and with main chain N
atoms of Gly242 and Arg241.
Acknowledgments
D.A.B. acknowledges the support of the EMBL Interdisciplinary
Postdocs (EIPOD) fellowship programme under Marie Skło-
dowska-Curie COFUND (grant number 291772) from the European
Commission. A.M.E. acknowledges the support of the Russian
Science Foundation (project no. 15-14-00014; work on production
of recombinant b-lactamase TEM-171) and the support of the Rus-
sian Foundation for Basic Research (project no. 15-54-74007; work
on inhibitory analysis). R.L.A. acknowledges the support of the Rus-
sian Foundation for Basic Research (project no. 14-04-01195; work
on organic synthesis).
A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2017.02.
025. These data include MOL files and InChiKeys of the most
important compounds described in this article.
References
1. Fisher JF, Meroueh SO, Mobashery S. Chem Rev. 2005;105:395.
2. Antimicrobial Resistance Global Report on Surveillance. World Health
Organization
<http://www.who.
int/drugresistance/documents/surveillancereport/en/> 2014.
Among the tested inhibitors, the 5c compound (4-([4-(2,4-
dichloro)phenoxyphenyl]-4-carboxyphenyl)thiourea) is the stron-
gest inhibitor against TEM-171 b-lactamase. The SAR and molecu-
lar modelling described here would be useful for further
3. Frieri M, Kumar K, Boutin A. J Infect Public Health. 2016. http://dx.doi.org/
10.1016/j.jiph.2016.08.007.
4. Baquero F, Tedim AP, Coque TM. Front Microbiol. 2013;4(15):1.
5. Van der Bij AK, Pitout JDD. J Antimicrob Chemother. 2012;67:2090.
6. Bush K, Bradford PA. Cold Spring Harb Perspect Med. 2016;6(a025247):1.
Please cite this article in press as: Antipin R.L., et al. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.02.025

R.L. Antipin et al. / Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
5
7. Roca I, Akova M, Baquero F, et al. New Microbes New Infect. 2015;6:22.
8. Tang S, Apisarnthanarak A, Hsu LY. Adv Drug Deliv Rev. 2014;78:3.
9. Bush K. Ann N.Y. Acad Sci. 2013;1277:84.
10. Pimenta AC, Fernandes R, Moreira IS. Mini Rev Med Chem. 2014;14:111.
11. Watkins RR, Papp-Wallace KM, Drawz SM, Bonomo RA. Front Microbiol. 2013;4
(392):1.
12. Drawz SM, Bonomo RA. Clin Microbiol Rev. 2010;23:160.
13. Ehmann DE, Jahic H, Ross PL, et al. Proc Natl Acad Sci. 2012;109:11663.
14. Horn JR, Shoichet BK. J Mol Biol. 2004;336:1283.
21. Sgrignani J, Novati B, Colombo G, Grazioso G. J Comput Aided Mol Des. 2015;29
(5):441.
22. London N, Miller RM, Krishnan S, et al. Nat Chem Biol. 2014;10(12):1066.
23. Friesner RA, Murphy RB, Repasky MP, et al. J Med Chem. 2006;49:6177.
24. Petterson EF, Goddard TD, Huang CC, et al. J Comp Chem. 2004;13:1605.
25. Eldridge MD, Murray CW, Auton TR, Paolini GV, Mee RP. J Comput-Aided Mol
Des. 1997;11:425.
26. Lu Y, Wang Y, Xu Zh, et al. J Phys Chem B. 2009;113:12615.
27. Wilcken R, Zimmermann MO, Lange A, Joerger AC, Boeckler FM. J Med Chem.
2013;56:1363.
15. Yu C, Shoichet BK. Nat Chem Biol. 2009;5(5):358.
16. Grigorenko VG, Andreeva IP, Rubtsova MYu, et al. Biochemie. 2017;132:45.
17. Bebrone C, Moali C, Mahy F, et al. Antimicrob Agents Chemother. 2001;45
(6):1868.
18. Grigorenko VG, Andreeva IP, Rubtsova MYu, Burmakin VV, Uporov IV, Egorov
AM. Mosc Univ Chem Bull. 2015;70(6):238.
28. Kantsadi AL, Hayes JM, Manta S, et al. Chem Med Chem. 2012;7:722.
29. Wilcken R, Liu X, Zimmermann MO, et al. J Am Chem Soc. 2012;134:6810.
30. Fedorov O, Huber K, Eisenreich A, et al. Chem Biol. 2011;18:67.
31. Hardegger LA, Kuhn B, Spinnler B, et al. ChemMedChem. 2011;54:2048.
32. Baumli S, Endicott JA, Johnson LN. Chem Biol. 2010;17:931.
33. Xu Z, Liu Z, Chen T, et al. J Med Chem. 2011;54:5607.
19. LigPrep, version 2.3, Schrödinger, LLC, New York (NY); 2009.
20. Glide, version 5.5, Schrödinger, LLC, New York (NY); 2009.
Please cite this article in press as: Antipin R.L., et al. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.02.025