New compounds for a good old class: Synthesis of two Β-lactam bearing cephalosporins and their evaluation with a multidisciplinary approach

Article abstract of DOI:10.1016/j.bmc.2019.115302

Antimicrobial resistance is spreading massively in the world and is becoming one of the main health threats of the 21st century. One of the possible strategies to overcome this problem is to modify the known classes of antibiotics in a rational way, with the aim of tuning their efficacy. In this paper, we present the synthesis and the evaluation of the biological activity of a series of two β-lactam bearing cephalosporin derivatives, in which an additional isolated azetidinone ring, bearing different substituents, is joined to the classical cephalosporanic nucleus by a chain of variable length. A computational approach has been also applied in order to predict the molecular interactions between some representative derivatives and selected penicillin-binding proteins, the natural targets of β-lactam antibiotics. All these derivatives are active against Gram-positive bacteria, with MIC₁₀₀ comparable or even better than that of the reference antibiotic ceftriaxone, and show no or very low cytotoxic activity on different cell lines. Overall, these molecules appear to be able to exert their activity in particular against microorganisms belonging to some of the species more involved in the development of multidrug resistance.

Full text of DOI:10.1016/j.bmc.2019.115302

Journal Pre-proofs
New compounds for a good old class: synthesis of two Β-lactam bearing
cephalosporins and their evaluation with a multidisciplinary approach
Giovanni Vigliotta, Deborah Giordano, Anna Verdino, Ivana Caputo, Stefania
Martucciello, Annunziata Soriente, Anna Marabotti, Margherita De Rosa
PII:
S0968-0896(19)31363-X
https://doi.org/10.1016/j.bmc.2019.115302
BMC 115302
DOI:
Reference:
To appear in:
Bioorganic & Medicinal Chemistry
Received Date:
Revised Date:
Accepted Date:
14 August 2019
13 December 2019
31 December 2019
Please cite this article as: G. Vigliotta, D. Giordano, A. Verdino, I. Caputo, S. Martucciello, A. Soriente, A.
Marabotti, M. De Rosa, New compounds for a good old class: synthesis of two Β-lactam bearing cephalosporins
and their evaluation with a multidisciplinary approach, Bioorganic & Medicinal Chemistry (2019), doi: https://
doi.org/10.1016/j.bmc.2019.115302
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. Please note that, during the production process, errors
may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
©
2019 Published by Elsevier Ltd.

New compounds for a good old class: synthesis of two Β-lactam bearing
cephalosporins and their evaluation with a multidisciplinary approach
Giovanni Vigliotta1, Deborah Giordano1,2, Anna Verdino1, Ivana Caputo1, Stefania
Martucciello1, Annunziata Soriente1, Anna Marabotti1* and Margherita De Rosa1*
1
Department of Chemistry and Biology “A. Zambelli”, University of Salerno, Via
Giovanni Paolo II, 132, 84084 Fisciano (SA), Italy
2
Present address: Laboratory of Bioinformatics and Computational Biology, Institute
of Food Science, National Research Council of Italy, Via Roma, 64, 83100 Avellino, Italy
*
Correspondence: amarabotti@unisa.it; Tel. +39 089 969583 (A.M.);
maderosa@unisa.it; Tel. +39 089 969553 (M.D.R.)
Abstract: Antimicrobial resistance is spreading massively in the world and is becoming one
of the main health threats of the 21st century. One of the possible strategies to overcome
this problem is to modify the known classes of antibiotics in a rational way, with the aim of
tuning their efficacy. In this paper, we present the synthesis and the evaluation of the
biological activity of a series of two β-lactam bearing cephalosporin derivatives, in which an
additional isolated azetidinone ring, bearing different substituents, is joined to the classical
cephalosporanic nucleus by a chain of variable length. A computational approach has been
also applied in order to predict the molecular interactions between some representative
derivatives and selected penicillin-binding proteins, the natural targets of β-lactam
antibiotics. All these derivatives are active against Gram-positive bacteria, with MIC100
comparable or even better than that of the reference antibiotic ceftriaxone, and show no or
very low cytotoxic activity on different cell lines. Overall, these molecules appear to be able
to exert their activity in particular against microorganisms belonging to some of the species
more involved in the development of multidrug resistance.
Keywords: Cephalosporins; Gram-positive; bifunctional compounds; penicillin-binding
proteins; docking.

1
. Introduction
In the 21st century, one of the most frightening threats for the human health is the
widespread diffusion of antimicrobial resistance, caused by the abuse and misuse of
1
antibiotics in the last decades. This threat is increased by the difficulty in finding new classes
of molecules active against microorganisms, since the last class of antibiotics was discovered
2
about 30 years ago. Therefore, while awaiting the discovery of a brand-new antibiotic
compound, it is essential to widen the panel of existing drugs, trying to develop innovative
molecules from the old available scaffolds.
Nearly 90 years after their discovery, β-lactam antibiotics are still widely used because they
3
couple high specificity and potency with relatively low toxicity and costs of productions. The
key of their mechanism of action is their resemblance with the natural substrate of penicillin-
binding proteins (PBPs), bacterial enzymes belonging to different classes depending on their
molecular mass, structure and different additional activities, involved in the synthesis and
4
maintenance of the peptidoglycan, an essential component of the prokaryotic cell wall. The
antibiotics of this class act as suicide inhibitors for PBPs, forming a covalent intermediate that
blocks the activity of these enzymes, causing ultimately the lysis of the bacteria.
Unfortunately, soon after the introduction of these drugs in clinics, bacteria resistant to their
activity were discovered, and to date, multi-drug resistance is common among many
microorganisms, especially the so-called ESKAPE pathogens (from the initials of the following
bacteria: Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae,
5
Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter species). Therefore,
many different classes of β-lactam antibiotics have been created, either to overcome the
6
increasing bacterial resistance or to improve their pharmacological features.
As a part of our long-standing research for new potential antibiotics7-12, during the last years
our efforts have been devoted to the development of an innovative class of bifunctional β-
lactam compounds. In particular, recently we have developed the prototype of a class of new
7
-aminocephalosporanic acid (7-ACA) derivatives, containing an additional β-lactam ring with
different substituents, joined to the amino-nitrogen of 7-ACA scaffold via an amide bond.
This new compound shows an antibacterial activity against Gram-positive bacteria and no
cytotoxicity towards human cells, even at a concentration higher than the one inhibiting the
1
3
growth of bacteria. Moreover, a recent computational study predicted an increased
resistance of this molecule towards β-lactamases, the enzymes responsible for the most part
1
4
of the bacterial resistance for these compounds.
In the present work, we have performed a rational modification of the root bifunctional
pharmacophore, on the basis of our previous studies, by introducing substituents on the two
phenyl rings bound to the isolated 2-azetidinone moiety and by modulating the length of the
chain connecting this ring to the 7-ACA moiety, and we have tested the effects of our
modifications on selected strains of Gram-positive and Gram-negative bacteria, as a proof of
concept. We have corroborated this approach with the evaluation of the cytotoxicity of these
new compounds and the prediction of the interactions of selected compounds, by using a
computational biology approach.
2
. Results

2
.1. Chemistry
The synthetic route to the target compounds 23a-d, 24a, 24c-d, 25a-b, 26a and 27a (Figure
1
3
1
) was designed according to our previous studies with some modifications. The strategy
utilized a convergent approach based on the coupling of a variously functionalized β-lactam
ring (13-17) with commercial (+)-7-ACA. Compounds 7-10 and 13a-16a (Figure 1) were
prepared following the previously described methodology and their spectral data matched
1
5-18
with those reported in literature.
R
R2
R
R2
R
1
2a-d
O
O
O
n
H2N
Cl
OMe
R1
i
R1
H
ii
+
N
N
O
N
O
a: n=1
b: n=2
c: n=5
d: n=7
R
R2
+
R1
O
n
OMe
O
OMe
R2
n
R1
(3R, 4S)
(3S, 4R)
4
5
6
: R =R = H
7: R=R =R =H
1 2
rac-13a-d:R=R1=R2=H
rac-14a,c-d: R=R1=H, R2=OMe
rac-15a-b:R=H, R1=R2=Me
rac-16a:R=NO2, R1=H, R2=OMe
rac-17a:R=Cl, R1=H, R2=OMe
1
2
3
: R= H
: R= NO2
: R= Cl
1
2
: R =H, R = OMe
8: R=R =H, R =OMe
1
2
1
2
: R =R = Me
9: R=H, R =R =Me
1 2
1
2
1
0: R=NO2, R =H, R =OMe
1
2
1
1:R=Cl, R =H, R =OMe
1
2
iii
R2
R
R2
R
R2
R1
R
R2
R
R1
R1
R1
iv, v
H
4
N
O
N
O
O
N
O
H
H
O
H
S
+
N
H
O
S
+
O
3
N
6
O
N
O
+
O
OH
7
O
n
n
n
H
H
N
N
O
n
OH
H
S
(3R, 4S)
(3S, 4R)
O
-
O
-
O Na+
O Na
H2N
(
3S, 4R, 6R, 7R)
O
(3R, 4S, 6R, 7R)
O
O
N
rac-18a-d:R=R1=R2=H
rac- 19a,c-d: R=R1=H, R2=OMe
rac- 20a-b:R=H, R1=R2=Me
2
2
2
2
3a-d:R=R =R =H
O
O
1
2
OH
4a,c-d: R=R =H, R =OMe
O
1
2
5a-b:R=H, R =R =Me
7-ACA
rac- 21a:R=NO2, R1=H, R2=OMe
rac-22a:R=Cl, R1=H, R2=OMe
1
2
6a:R=NO2, R =H, R =OMe
1 2
2
7a:R=Cl, R =H, R =OMe
1 2
Reagents and conditions : (i)EtOH, reflux; (ii) nBu3N, toluene, reflux; (iii) LiOH, THF/H2O, rt; (iv) DCC, Et3N, CH2Cl2; (v) CH3COCH3,
NaHCO3, H2O
Figure 1. Reagents and conditions for the synthesis of target compounds.
The synthesis of the variously functionalized β-lactam rings (13-17) included a common
origin: a Staudinger [2+2] cycloaddition reaction between a properly substituted imine (7-11)
and a ketene prepared in situ from the appropriate acyl chloride (12a-d). The imines (7-11)
were obtained by condensation of commercially available aldehydes (1-3) and anilines (4-6)
in ethanol at reflux. As regards derivatives 12a-d, the compounds 12a and 12b were prepared
starting from the commercially available succinic and glutaric anhydride respectively (see
Supporting Information), while 12c and 12d were commercially available. Under our selected
experimental conditions, racemic trans-β-lactams (rac-13-17) were formed exclusively in the
1
11
Staudinger reaction as confirmed by H-NMR spectra and according to our previous reports.
Next, the treatment of these compounds with LiOH solution provided the corresponding
acids (rac-18-22) in quantitative conversion, which were directly subjected to the coupling
with commercially available (+)-7-ACA using DCC as a coupling reagent and Et N. The
3
subsequent transformation into sodium salt provided target compounds (23a-d, 24a, 24c-d,
2
5a-b, 26a, 27a) as diastereomeric mixture with (3S, 4R, 6R, 7R) and (3R, 4S, 6R, 7R) absolute
configuration at C-3, C-4, C-6 and C-7, respectively, and ensured their better water solubility.
Since the separation of diastereomers was disappointing and the computational study

1
3
reported in our previous paper evidenced no significant differences among the predicted
binding affinities to all PBPs for both diastereoisomers, we decided to perform the biological
assays with the diastereomeric mixtures.
2
.2 Antimicrobial Activity
The antimicrobial activity of compounds 23a-d, 24a, 24c-d, 25a-b, 26a and 27a was evaluated
by comparative analysis of both the MIC and the MLD determined for non-pathogenic Gram
negative strains Escherichia coli JM107 and Salmonella enterica LT2 ATCC 700720, and Gram
positive strains Staphylococcus aureus (a pathogenic hospital isolate) and Bacillus sp. (a non-
pathogenic environmental strain). These strains are the same bacteria used in our previous
1
3
work , to have a direct comparison with our previous results. The analysis was also extended
to ceftriaxone, a third-generation cephalosporin with a broad-spectrum activity (Table 1).
All compounds were very effective against Gram-positive microorganisms, with MIC100
ranging from values ≤0.25 to 32 μg/ml, but not against Gram-negative (up to 256 μg/ml). The
most active non-elongated antibiotics were 24a and 23a, whose MIC and MLD for S. aureus
were significantly lower than those of ceftriaxone are (MIC100 about 3.5 and 2 times lower for
2
4a and 23a, respectively).
The elongation of the carbon chain did not change the action spectrum of compounds,
however, it significantly increased both the activity towards Gram-positive and the bacterial
sensitivity to the antibiotics compared to ceftriaxone. As indicated in Table 1 for S. aureus
the MIC100 and MLD of 23d were 2.5 and 3 times lower than those of 23a, respectively, and
5
.5 and 7 times lower than those of ceftriaxone. On the Bacillus sp. the bacteriostatic activity
was proportional to the chain length, reaching the lower values for 23c, whose MIC100 value
was 5 times lower than that of 23a and 1.5 lower than that of ceftriaxone. In contrast,
compounds elongation significantly decreased the bactericidal activity, also in comparison to
ceftriaxone (see MLD). Analysis of compounds 24 evidenced that 24d was the more efficient
against both S. aureus and Bacillus sp. with MIC100 values more than 6 and 3 times lower than
2
4a, respectively. Compound 24d was much more active than ceftriaxone, even compared
to compounds 23c-d. In fact, the MIC100 for S. aureus and Bacillus sp. were 22 times and 3
times lower than those of ceftriaxone are, respectively. This antibiotic also presented the
higher bactericidal activity against S. aureus, compared to all tested compound and
ceftriaxone, while for Bacillus sp. the more lethal was the 24c.
Table 1. Antimicrobial activity of elongated and non-elongated compounds and comparison with
ceftriaxone (MICs and MLDs are in μg/ml). The reported values were determined from 0 to 256 μg/ml
for each antibiotic, by at least three independent experiments, each in triplicate. In some cases, only
the ranges of values at which the toxic doses fall have been determined. Except for >256 μg/ml, the
symbols “<” or “>” without the indication of the interval of concentrations mean that MIC and/or MLD
are very close to the values to which they refer, and in at least one of the independent experiments it
1
13
was found to be the same. : data already published in our previous work.

S. aureus
Bacillus sp.
E. coli
S. enterica
Compound
MIC50
MLD
/
100
MIC50/
100
MLD
MIC50
MIC100
MLD
MIC50
MIC100
MLD
N
O
H
H
S
0
.75
2.5
>3
<0.25
2.5
>2.5
>256
>256
N
H
O
O
O
O
O
N
1
2
3a
O
-
+ O
O Na
O
N
N
N
O
2
H
>0.25,
<0.5
H
O
S
N
H
2
.5
≥3
<0.25
<0.25
<0.25
<0.25
<0.12
<0.12
<0.25
≤1
1
>32
≥32
>256
>256
>256
>256
>256
>256
>256
>256
>256
>256
>256
>256
>256
>256
>256
>256
>256
>256
>256
O
N
2
3b
-
+O
O Na
O
O
5
H
>1.5,
<2
>3.5,
<4
H
O
S
>
4
0.50
0.75
0.75
N
H
O
N
23c
-
+ O
O Na
O
O
7
H
H
O
S
<
<
0.25
0.25
1
≥1
>16
N
H
O
N
2
3d
-
+ O
O Na
O
O
N
O
H
H
S
1.5
>1.5
≥128
≥0.75
>2, <4
≥3.5
64
O
O
O
N
O
+ O
H
N
2
4a
O
-
O Na
O
O
N
O
>1.75,
<2
>0.25,
<0.35
H
H
S
0
.75
1.75
<0.25
>4, <8
>4, <6
32
N
O
5
H
N
2
4c
4d
O
-
+ O
O Na
O
S
O
N
O
H
H
O
>
0.12
0.25
>16
>16
>32
0.25
3.5
2
N
H
O
7
N
2
-
+ O
O Na
O
N
O
H
H
S
>
1
O
N
O
H
N
2
5a
O
-
+ O
O Na
O
N
O
H
H
S
1
O
2
N
H
O
N
2
5b
O
-
+ O
O Na
O
O
NO2
N
O
H
H
S
>
4, <8
8
32
>256
O
N
H
O
+ O
N
2
6a
O
-
O Na
O
O
Cl
N
O
H
H
O
S
>
>
4, <8
2, <4
32
>32
>7
<4
32
>256
>256
O
N
O
+ O
H
N
2
7a
-
O Na
O
Ceftriaxone
<0.25/0.25
<0.25/0.25
5.5
<0.25
0.75
≥1
>
1
≥2
2
.3. Cell cytotoxicity

1
3
We tested compounds on MRC5, one of the two human cell lines previously used , at
different concentrations ranging from 0.5 to 50 μg/ml. We observed that no compounds
were cytotoxic in the range 0.5-25 μg/ml except in the case of 25a and 24d that both
significantly reduced cell viability of about 20% at 25 μg/ml. A slight (about 20%) significant
reduction of cell viability was observed at the highest dose (50 μg/ml) for several compounds
(Figure 2A). We also tested compounds 23b-c, 24a and 25b on a further human cell line, Calu-
1
, and only observed a reduction of cell viability at the highest concentration of compounds
2
3b and 23c (Figure 2B).
Figure 2. Cytotoxicity of cephalosporin derivatives. (A) Effect on cell viability of MRC5 cells after 24 h
exposure to compounds 23b-c, 24a, 24c-d, 25a-b, 26a, 27a (dose-range 0.5–50 μg/ml). Data for 23a
1
3
were previously published. (B) Effect on cell viability of Calu-1 cells after 24 h exposure to compounds
3b-c, 24a, 25b (dose-range 0.5–50 μg/ml). Data are reported as mean ± SD from two or three
2
independent experiments, each in triplicate. Vehicle (DMSO), at the highest concentrations used (0.4%
v/v), induced a reduction of cell viability of about 15% (not shown). *p<0.05 vs. the respective vehicle.
2
.4. Computational data
The results of the computational analysis of the interactions of two representative
compounds, 23d and 24a, with some representative PBPs from both Gram-positive and
1
9-24
Gram-negative species (Supplementary Table S1)
are reported in Tables 2-5. In
1
3
agreement with our previous work , there are no significant differences neither among the
binding energies predicted for the enzymes belonging to the Gram-positive bacteria and
those predicted for Gram-negative bacteria, nor between the binding energies predicted for
the two diastereoisomers obtained from the synthesis of the compounds. Indeed, the error
made by AutoDock in the prediction of the binding energies is about 1-1.5 kcal/mol, i.e. the

same order of magnitude of these differences. Moreover, the β-lactam ring belonging to the
-ACA moiety is still predicted to be more reactive towards acylation with respect to the
7
additional, isolated β-lactam ring, as the binding energies predicted for this moiety are
consistently lower. In agreement with the microbial assays, the energies predicted for the
interaction of 23d with the enzymes are better than ceftriaxone and those of its precursor
1
3
2
3a (the compound 8 reported in our previous publication ). This is evident in particular for
the reaction of the 7-ACA moiety. On the contrary, the binding energies predicted for the
interaction of compound 24a are similar or even slightly worse than those of compound 23a,
1
3
and in line with those obtained for ceftriaxone .
The detailed interactions of 23d and 24a with all the enzymes tested are shown in
1
3
Supplementary Tables S2-S5. As it was already shown for the precursor 23a , H-bonds
prevail among other interactions for both compounds and involve mainly the polar groups
linked to the 7-ACA nucleus, the amide group linking 7-ACA to the chain with the isolated
azetidinone ring, and the carbonyl groups of the azetidinone rings, especially in their open
form, when they interact with the catalytic Ser residue. A comparison of these interactions
among all the active sites of the enzymes tested shows that these H-bonds are made with
polar residues, in particular Ser, Thr and Asn, and often also with positively charged residues
(Lys and Arg). Moreover, in the majority of the active sites, the two phenyl rings on the
isolated azetidinone moiety are involved in π interactions of different nature with aromatic
residues (Phe, Tyr) and sometimes with charged or polar residues (Arg, Lys, Glu)
(Supplementary Tables S6-S9). Interestingly, in the complexes between 24a and the enzymes
from Gram-positive bacteria, the methoxy group bound to one of the two phenyl rings is
rarely involved in interactions (Supplementary Tables S2 and S3) whereas, in the complexes
with the enzymes from Gram-negative bacteria, that methoxy group is more frequently
involved in H-bonds with polar residues such as Gln, Ser and Asn, or even charged residues
(Supplementary Tables S4 and S5). On the other hand, the elongated chain of compound 23d
shows an enhanced conformational mobility and often points outside the active site of the
enzymes, making contacts with other residues far from it, on the protein surface.

Table 2. Results of covalent docking between compound 23d and selected PBPs of Gram-positive
1
bacteria. :when the result with the best energy is not associated to the most populated cluster, the
2
energy and the number of poses of the most populated cluster is additionally reported. : the
3
additional β-lactam ring has been considered reactive towards the acylation. : the β-lactam ring of
4
the 7-ACA moiety has been considered reactive towards the acylation. : the results obtained for
1
3
ceftriaxone and previously published are reported here for sake of comparison.
3
VSL
1TVF
(PBP4 from S. aureus)
2J9P
(
PBP3 from S. aureus)
(PBP4a from B. subtilis)
Predicted ΔG
Predicted ΔG
(kcal/mol) and
number of
poses in the
Predicted ΔG
(kcal/mol) and
(
kcal/mol) and
number of
poses in the
selected
Total
number of
clusters
Total
number of
clusters
Total
number of
number of
poses in the
clusters
Compound 23d
selected
selected
1
1
1
cluster
cluster
cluster
-
-
-
-
12.75 (2)
11.61 (14)
15.41 (1)
15.36 (23)
-12.06 (12)
-10.90 (21)
-14.54 (2)
-14.33 (16)
(
3R,4S), ring A
3
4
4
22
-12.95 (32)
24
20
2
reactive
-12.81 (5)
-12.11 (21)
-12.25 (11)
(3R,4S), ring B
2
3
3
28
3
reactive
-12.28 (3)
(3S,4R), ring A
0
0
-12.55 (27)
22
21
22
29
30
28
2
reactive
-12.27 (15)
-15.07 (8)
-13.25 (12)
-11.81 (20)
-13.37 (13)
-11.70 (15)
-13.88 (8)
(3S,4R), ring B
3
reactive
-12.85 (22)
-
-
12.22 (17)
12.00 (20)
-11.75 (4)
Ceftriaxone4
-14.29 (36)
27
-10.65 (12)

Table 3. Results of covalent docking between compound 23d and selected PBPs of Gram-negative
1
bacteria. : when the result with the best energy is not associated to the most populated cluster, the
2
energy and the number of poses of the most populated cluster is additionally reported. : the
3
additional β-lactam ring has been considered reactive towards the acylation. : the β-lactam ring of
4
the 7-ACA moiety has been considered reactive towards the acylation. : the results obtained for
1
3
ceftriaxone and previously published are reported here for sake of comparison.
5
FGZ
2EX8
(PBP4 from E. coli)
3OCL
(
PBP1b from E. coli)
(PBP3 from P. aeruginosa)
Predicted ΔG
Predicted ΔG
Predicted ΔG
(kcal/mol) and
(
kcal/mol) and
number of
poses in the
selected
(kcal/mol) and
number of
poses in the
selected
Total
number of
clusters
Total
number of
clusters
Total
Compound
3d
number of
number of
poses in the
clusters
2
selected
1
1
1
cluster
cluster
cluster
-
12.61 (8)
-11.28 (4)
-10.51 (9)
-13.60 (18)
(
3R,4S), ring A
3
2
6
0
36
29
2
reactive
-12.39 (24)
-13.48 (19)
-
-
-
-
14.45 (4)
-13.35 (5)
-15.11 (1)
27
-13.85 (11)
(3R,4S), ring B
3
3
34
32
3
reactive
13.48 (19)
13.19 (12)
11.70 (12)
-13.30 (11)
-14.33 (10)
32
-12.86 (19)
(3S,4R), ring A
-12.47 (20)
2
reactive
-
13.83 (8)
-13.65 (20)
-13.56 (26)
-12.76 (11)
-11.77 (22)
-15.01 (4)
30
-14.51 (15)
(
3S,4R), ring B
3
5
27
3
reactive
-13.76 (25)
-
-
11.87 (4)
-13.99 (5)
25
Ceftriaxone4
30
14
11.39 (20)
-11.98 (18)

Table 4. Results of covalent docking between compound 24a and selected PBPs of Gram-positive
1
bacteria. : when the result with the best energy is not associated to the most populated cluster, the
2
energy and the number of poses of the most populated cluster is additionally reported. : the
3
additional β-lactam ring has been considered reactive towards the acylation. : the β-lactam ring of
4
the 7-ACA moiety has been considered reactive towards the acylation. : the results obtained for
1
3
ceftriaxone and previously published are reported here for sake of comparison.
3
VSL
1TVF
(PBP4 from S. aureus)
2J9P
(
PBP3 from S. aureus)
(PBP4a from B. subtilis)
Predicted ΔG
Predicted ΔG
(kcal/mol) and
number of
poses in the
Predicted ΔG
(kcal/mol) and
(
kcal/mol) and
number of
poses in the
selected
Total
number of
clusters
Total
number of
clusters
Total
Compound
4a
number of
number of
poses in the
clusters
2
selected
selected
1
1
1
cluster
cluster
cluster
-
10.91 (17)
10.01 (19)
-10.76 (20)
14
-9.52 (21)
(
3R,4S), ring A
1
8
-9.94 (47)
8
2
reactive
-
-
12.04 (1)
(
3R,4S), ring B
-
14.16 (24)
19
12
11
10
-11.75 (40)
18
10
10
3
reactive
-11.47 (32)
-
11.43 (10)
10.95 (24)
-10.47 (12)
-10.31 (44)
(3S,4R), ring A
1
9
-9.59 (28)
2
reactive
-
-
11.19 (18)
10.64 (25)
(3S,4R), ring B
-13.91 (21)
26
-11.70 (53)
3
reactive
-
-
-
12.22 (17)
12.00 (20)
-11.75 (4)
Ceftriaxone4
-14.29 (36)
27
22
28
-10.65 (12)

Table 5. Results of covalent docking between compound 24a and selected PBPs of Gram-negative
1
bacteria. : when the result with the best energy is not associated to the most populated cluster, the
2
energy and the number of poses of the most populated cluster is additionally reported. : the
3
additional β-lactam ring has been considered reactive towards the acylation. : the β-lactam ring of
4
the 7-ACA moiety has been considered reactive towards the acylation. : the results obtained for
1
3
ceftriaxone and previously published are reported here for sake of comparison.
5
FGZ
2EX8
(PBP4 from E. coli)
3OCL
(PBP1b from E. coli)
(PBP3 from P. aeruginosa)
Predicted ΔG
kcal/mol) and
number of
Predicted ΔG
(kcal/mol) and
Predicted ΔG
(kcal/mol) and
(
Total
number of
clusters
Total
number of
clusters
Total
number of
clusters
Compound
4a
number
of
number
of
2
poses in the
selected
poses in the
selected
poses in the
selected
1
1
1
cluster
cluster
cluster
-
10.99 (19)
-12.49 (11)
-12.05 (41)
(
3R,4S), ring A
-
11.32 (22)
17
10
13
12
16
2
reactive
-10.00 (34)
-13.36 (10)
-11.86 (15)
-11.66 (25)
(3R,4S), ring B
2
1
2
1
2
0
-12.65 (49)
-13.79 (30)
3
reactive
-11.31 (8)
-10.26 (25)
-11.99 (21)
-12.72 (9)
(3S,4R), ring A
12
19
2
reactive
-10.64 (31)
-12.26 (26)
-
13.00 (14)
12.57 (15)
(3S,4R), ring B
16
-13.79 (32)
13
3
reactive
-
-11.78 (33)
-12.76 (11)
-11.77 (22)
-
-
11.87 (4)
-13.99 (5)
Ceftriaxone4
30
14
25
11.39 (20)
-11.98 (18)
3
. Discussion
Starting from a literature survey of the basic structure-activity relationships in N-aryl-C4-aryl-
2
5-28
11,13
2
-azetidinones
and from our previous studies , we hypothesized that the nature of the
substituent groups on the aromatic rings of the additional β-lactam moiety in our compounds
could affect the interactions with the active sites and afford a stronger and more efficient
biological activity. Therefore, we designed and synthesized a series of novel bifunctional
cephalosporin analogues containing electron-withdrawing (NO , Cl) and/or electron-
2
donating groups (OMe) on the two phenyl rings at the 1- and 4-positions of the additional β-
lactam nucleus, in order to assess the substituent influence on the activity compared to
progenitor compound 23a. In the second instance, we hypothesized that the distance
between 7-ACA and the additional substituted β-lactam might also enhance the antibacterial
activity, possibly due to the increased conformational flexibility such as to fulfill the

geometric requirements of better interactions. Therefore, we introduced non-functionalized
polymethylene chains from one to seven carbons as flexible spacers.
The comparison of the microbiological results shows that all new synthesized compounds are
very effective against the Gram-positive selected bacteria and confirmed that carbon chain
elongation significantly increases their antimicrobial properties compared to ceftriaxone.
Particularly, the introduction of electron-donating OMe group on the phenyl ring at C-4
position of the additional β-lactam ring displays increased activity (bacteriostatic or both
bacteriostatic and bactericidal) compared to reference cephalosporin and 23a. However, the
simultaneous presence of OMe group at C-4 position and an electron-withdrawing group on
the phenyl ring at the 1-position completely cancels the positive effect of the first and results
in a significant reduction of the antimicrobial activity with respect to the basic compound
2
3a. The results display also the beneficial influence of lengthening the alkyl chain spacer in
the series 23a-d and 24a-d.
MIC and MLD also reveal that the compounds have a bactericidal action on S. aureus, similarly
to ceftriaxone (almost superimposable MIC and MLD values), while for Bacillus sp. is more
evident a bacteriostatic character (greater separation between MIC and MLD values). This
aspect underlines some differences in the antimicrobial mechanism between two genera,
which could be based on differences in the cellular PBPs and/or in the affinity with common
ones, both resulting in characteristic PBPs profiles of inhibition. The importance of the
combination of PBPs that are inhibited on the degree of cellular alteration and lethality has
been shown in both Gram-positive and Gram-negative bacteria (B. subtilis, E. coli,
4
, 29-31
Streptococcus pneumoniae).
We tested the cytotoxic effect of our compounds on MRC5 and Calu-1 cells. We found a
significant reduction of cell viability (not over 30%) only at concentrations at least 10 times
higher than MIC both in MRC5 and in Calu-1 cells, even if the two cell lines displayed a slight
5
0
different sensibility towards the compounds. However, the effect on cell viability of both cell
1
3
lines was comparable to that of 23a and of the commercial ceftriaxone on MRC5 cells.
Therefore, these compounds show no in vitro cytotoxicity on the tested cell lines at the doses
at which they exert a bacteriostatic and/or bactericidal activity.
The computational analysis is intended to provide an interpretation of the molecular
phenomena underlying the experimental data obtained in this work, although it is not
possible to compare them directly, given the different scale and the complexity of the
phenomena that can affect the microbiological assays. Moreover, given the extremely high
number of PBPs present in bacteria, this analysis does not claim to be exhaustive, but only to
provide suggestions for a better comprehension of the activity of this new class of
compounds.
From this analysis, these bifunctional molecules are predicted to bind with good affinity to
PBPs belonging both to Gram-positive and Gram-negative microorganisms, despite the
1
3-14
microbiological results. This is not unexpected since, as we postulated previously
,
probably the lack of activity of these compounds towards this last type of bacteria is due to
their external membrane that impairs the penetration of these big compounds in the
periplasmic space. The predicted binding energy of the elongated compound 23d for the two
PBPs belonging to S. aureus is consistently better than that predicted for the enzyme from B.

subtilis, especially for the 7-ACA moiety. This effect is evident also for compound 24a towards
PBP3 belonging to S. aureus, whereas in the case of the PBP4 the predicted affinities are
similar for S. aureus and B. subtilis. Overall, these data suggest that these compounds could
bind stronger to their target enzymes in S. aureus, and this might explain their more
pronounced bactericidal activity towards this Gram-positive organism.
Comparing the interactions of compound 24a with the residues of the active sites of the PBPs
with those of progenitor compound 23a, it is possible to note that the addition of the
methoxy group on the phenyl rings bound to the isolated azetidinone moiety seems to
increase the number of interactions of this moiety, especially with those enzymes belonging
to Gram-negative bacteria. Furthermore, the predicted binding affinity of the elongated
compound 23d for the PBPs is generally improved with respect to the precursor 23a,
especially when the β-lactam ring belonging to 7-ACA is the target for the acylation. We did
not perform a direct assay on isolated PBPs to prove it, nevertheless these data are in
agreement with the microbiological assays and suggest that the long chain of this compound,
by pushing the two moieties far away each other's, allows reducing the steric hindrance of
the portion not directly involved in the acylation with the enzyme, thereby favoring the
interaction of the other portion with the active site. From these data, it is not surprising that
compound 24d, including both modifications, is the one with the best antimicrobial activity.
Additionally, the docking results show that the number of clusters obtained for this
compound is very large, in line with the enhanced conformational flexibility of the connecting
chain. The position of this chain is quite different in the several results obtained for each
simulation (data not shown). In some complexes obtained with docking simulations, this
chain is bended and the free moiety can interact with the protein's surface, probably with
aspecific interactions, whereas in other cases, the chain is more elongated and protrudes
outside the active site (that usually in PBPs is formed by a large cavity at the protein's surface,
very accessible to the solvent). This suggests that, once one of the two moieties is covalently
bound to the target PBPs, the other moiety might move freely in the medium, being available
for the interaction with another target enzyme. The two parts are then potentially able to
act independently with fewer constraints than in the precursor 23a. However, it is important
to point out that these compounds are very large and that the long chain is characterized by
a high number of torsional degrees, reaching almost the maximum value that can be
managed by AutoDock. Further studies by using different approaches, e.g. molecular
dynamics simulations in solvent, could be applied to study this effect in more details. For the
same reasons, since these complexes violate some of the Lipinski's rules, more studies will
be necessary to develop drug-like compounds from this new class of bifunctional
cephalosporins.
4
. Materials and Methods
4
.1 Chemistry
Detailed experimental procedures utilized for the preparation of all new compounds,
including the intermediates, and spectral characterization are reported in the Supplementary
File S6. All new compounds were characterized by NMR spectroscopy (Bruker Avance-400
and BRUKER Avance-600 spectrometers), ESI-(+)-MS measurements (Waters 4 micro
quadrupole mass spectrometer equipped with electrospray ion source), elemental analysis

(FlashEA 1112 Series with Thermal Conductivity Detector) and melting points (DSC 2920 TA
INSTRUMENTS) as reported in the Supplementary File S6.
4
.2 Microbiological assays and bacterial strains
1
3
The compounds were tested on the same bacteria strains used previously , both Gram-
negative as E. coli JM109 and Salmonella enterica LT2 ATCC 700720, and Gram-positive as S.
aureus and Bacillus sp. The non-pathogenic strains E. coli JM109 and Salmonella enterica
subsp. Enterica serovar Typhimurium, strain LT2 ATCC 700720 were purchased from Promega
(http://www.promega.com/products) and Leibniz Institute DSMZ-German Collection of
Microorganisms and Cell Culture (http://www.dsmz.de), respectively. S. aureus and Bacillus
sp. were derived from the collection deposited in the microbiology laboratory directed by dr.
G. Vigliotta. The first is a pathogenic hospital isolate, while the second one is a non
pathogenic environmental strain.
The minimum concentration inhibiting growth (MIC ), the minimum concentration that
1
00
reduced growth by 50% compared to the control samples grown in the absence of antibiotic
MIC ) and the minimum lethal dose (MLD) of each antibiotic derivative were estimated as
(
5
0
1
3
described previously , according to the recommendations of the Clinical and Laboratory
Standards Institute (CLSI).
1
1,32
The toxic doses were determined from 0 to 256 μg/ml for each
antibiotic, by at least three independent experiments, each in triplicate.
Results were compared to that of ceftriaxone (Sigma-Aldrich, Milan, Italy), a reference
cephalosporin agent active towards a broad range of Gram-positive and Gram-negative
bacteria, in clinical use for the treatment of a variety of infections such as meningitis,
3
3-34
gonorrhea and community-acquired pneumonia.
4
.3 Evaluation of cytotoxicity
Cytotoxicity of compounds was monitored by the 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium (MTT) (Sigma-Aldrich) colorimetric test performed on MRC5, a human
hembryonic lung cell line, and on Calu-1 cells, a human lung cancer cell line (Interlab Cell Line
Collection, Istituto Nazionale per la Ricerca sul Cancro, Genoa, Italy), as previously
1
3,35
described.
triplicate.
For each compound the MTT assay was performed at least two times in
4
.4 Computational section
1
9-24
The structures of the PBPs used in this work
were downloaded from the Protein Data
3
6
Bank. These representative proteins were selected in order to include enzymes from both
Gram-positive and Gram-negative bacteria tested in this work, belonging to different classes
according to the Ambler classification. In addition, quality criteria of their structures were
4
1
3-14
taken into account, as it has been done in our previous works.
The structures of the representative compounds 23d and 24a were designed and saved in 3D
.pdb format using ChemDraw and Chem3D Pro 12.0 (Perkin Elmer). Since their chemical
synthesis produces a diastereoisomeric mixture (see Results), we designed and analysed both
diastereoisomers (3R, 4S) and (3S, 4R). Moreover, for each diastereoisomer, the reaction
with both β-lactam rings was simulated. Therefore, for each protein and each compound,
four different docking simulations were performed.

3
7
To perform covalent docking, the flexible side chain approach implemented in AutoDock v.
3
8
39
4
.2 was used. Chimera was used to modify the structures and, if necessary, to optimize
them, according to the protocol for the covalent docking. The parameters used to setup
1
3-14
simulations on compound 24a were the same used in our previous studies , whereas for
3d the maximum number of energy evaluations was increased to 25,000,000 to take into
2
account the high number of torsions in this molecule, and the dimensions of the box was
increased to 75x75x80 to widen the space available for the molecule to change its
conformation. The values of the interactions between compound 23a (previously identified
1
3
as compound 8 ) and the PBPs, or between ceftriaxone and PBPs were recalculated using
these new parameters, but they resulted not significantly different with respect to those
already published (data not shown). The conformations corresponding to the best energetic
and to the most populated cluster of poses obtained from covalent docking, were analysed
with Discovery Studio (DassaultSystèmes, 2015) to identify the most important interactions
between the enzymes and the two compounds.
5
. Conclusions
We have presented the synthesis and a multidisciplinary analysis of the biological effects of
different series of new cephalosporin derivatives bearing two β-lactam rings, with proven
antimicrobial activity on Gram-positive bacteria such as S. aureus and Bacillus sp.. These
compounds show to have bactericidal/bacteriostatic effects at concentrations at which they
are well tolerated by human cells. A computational analysis has allowed to dissect the
interactions that these compounds can have with their target enzymes, PBPs. All together,
these data suggest that these molecules are promising starting points for the development
of new antibiotics.
Funding: This research was funded by UNIVERSITY OF SALERNO, grant number FARB-
ORSA170308 and 183080 (to A.V., A.M.), FARB-ORSA177918 (to M.D.R.), and by ITALIAN
MINISTRY OF UNIVERSITY AND RESEARCH, grant FFABR2017 (to A.M., M.D.R.).
References
1
.
World Health Organization (WHO). Antimicrobial resistance: Global report on
surveillance; WHO Press: Geneva, Switzerland, 2014.
2
.
Silver, L.L. Challenges of antibacterial discovery. Clin Microbiol Rev 2011, 24, 71-109.
https://doi.org/10.1128/CMR.00030-10.
3
.
King, D.T.; Sobhanifar, S.; Strynadka, N.C.J. One ring to rule them all: Current trends
in combating bacterial resistance to β-lactams. Protein Sci 2016, 25, 787-803.
https://doi.org/10.1002/pro.2889.
4
.
Sauvage, E.; Kerff, F.; Terrak, M.; Ayala, J.A.; Charlier, P. The penicillin-binding
proteins: Structure and role in peptidoglycan biosynthesis. FEMS Microbiol Rev 2008, 32,
34-258. https://doi.org/10.1111/j.1574-6976.2008.00105.x.
2

5
.
Santajit, S.; Indrawattana, N. Mechanisms of antimicrobial resistance in ESKAPE
pathogens. Biomed Res Int 2016, 2016, 2475067. https://doi.org/10.1155/2016/2475067.
6
.
Demain, A.L.; Solomon, N.A. (Eds). Antibiotics containing the β-lactam structure Part
II: Springer-Verlag: Berlin-Heidelberg, Germany, 2012.
7
.
Technikova-Dobrova, Z.; Damiano, F.; Tredici, S.M.; Vigliotta, G.; di Summa, R.;
Palese, L.; Abbrescia, A.; Labonia, N.; Gnoni, G.V.; Alifano, P. Design of mineral medium for
growth of Actinomadura sp. ATCC 39727, producer of the glycopeptide A40926: effects of
calcium ions and nitrogen sources. Appl Microbiol Biotechnol 2004, 65, 671-67.
https://doi.org/10.1007/s00253-004-1626-2
8
.
Vigliotta, G.; Tredici, S.M.; Damiano, F.; Montinaro, M.R.; Pulimeno, R.; di Summa, R.;
Massardo, D.R.; Gnoni, G.V.; Alifano, P. Natural merodiploidy involving duplicated rpoB
alleles affects secondary metabolism in a producer actinomycete. Mol Microbiol 2005, 55,
3
96-412. https://doi.org/10.1111/j.1365-2958.2004.04406.x
9
.
Grimaldi, M.; De Rosa, M.; Di Marino, S.; Scrima, M.; Posteraro, B.; Sanguinetti, M.;
Fadda, G.; Soriente, A.; D’Ursi, A.M. Synthesis of new antifungal peptides selective against
Cryptococcus neoformans. Bioorg Med Chem 2010, 18, 7985-7990.
https://doi.org/10.1016/j.bmc.2010.09.033.
1
0.
De Rosa, M.; Zanfardino, A.; Notomista, E.; Wichelhaus, T.A.; Saturnino, C.;
Varcamonti, M.; Soriente A. Novel promising linezolid analogues: Rational design, synthesis
and biological evaluation. Eur Med Chem 2013, 69, 779-785.
https://doi.org/10.1016/j.ejmech.2013.09.035.
J
1
1.
De Rosa, M.; Vigliotta, G.; Palma, G.; Saturnino, C.; Soriente, A. Novel penicillin-type
analogues bearing a variable substituted 2-azetidinone ring at position 6: Synthesis and
biological evaluation. Molecules 2015, 20, 22044-22057.
https://doi.org/10.3390/molecules201219828.
1
2. Tammaro, L.; Saturnino, C.; D'Aniello, S.; Vigliotta, G.; Vittoria, V. Polymorphic
solidification of Linezolid confined in electrospun PCL fibers for controlled release in topical
applications. Int J Pharm 2015, 25, 32-38. https://doi.org/10.1016/j.ijpharm.2015.04.070.
1
3.
Verdino, A.; Vigliotta, G.; Giordano, D.; Caputo, I.; Soriente, A.; De Rosa, M.;
Marabotti, A. Synthesis and biological evaluation of the progenitor of a new class of
cephalosporin analogues, with a particular focus on structure-based computational analysis.
PLoS One 2017, 12, e0181563. https://doi.org/10.1371/journal.pone.0181563.
1
4.
Verdino, A.; Zollo, F.; De Rosa, M.; Soriente, A.; Hernández-Martínez, M.Á.;
Marabotti, A. Computational analysis of the interactions of a novel cephalosporin derivative
with β-lactamases. BMC Struct Biol 2018, 18, 13. https://doi.org/10.1186/s12900-018-0092-
5
.
1
5.
Hwu, J.R.; Tseng, W.N.; Patel, H.V.; Wong, F.F.; Horng, D.N.; Liaw, B.R.; Lin, L.C. Mono-
deoxygenation of nitroalkanes, nitrones, and heterocyclic N-oxides by hexamethyldisilane
through 1,2-elimination: Concept of "Counterattack Reagent". J Org Chem 1999, 64, 2211–
2
218. https://doi.org/10.1021/jo981054i.

1
6.
Ibrahim, M.N.; Hamad, K.J.; Al-Joroshi, S.H. Synthesis and characterization of some
Schiff bases. Asian J Chem 2006, 18, 2404-2406.
1
7. De Rosa, M.; Soriente, A. Water opportunities: Catalyst and solvent in Mukaiyama
aldol addition of Rawal's diene to carbonyl derivatives. Tetrahedron 2011, 67, 5949-5955.
https://doi.org/10.1016/j.tet.2011.06.035.
1
8.
Zarei, M.; Mohamadzadeh, M. 3-Thiolated 2-azetidinones: synthesis and in vitro
antibacterial and antifungal activities. Tetrahedron 2011, 67, 5832–5840.
https://doi.org/10.1016/j.tet.2011.05.043.
1
9.
Kishida, H.; Unzai, S.; Roper, D.I.; Lloyd, A.; Park, S.Y.; Tame, J.R. Crystal structure of
penicillin binding protein 4 (dacb) from Escherichia coli, both in the native form and
covalently linked to various antibiotics. Biochemistry 2006, 45, 783-792.
https://doi.org/10.1021/bi051533t.
2
0.
Sauvage, E.; Duez, C.; Herman, R.; Kerff, F.; Petrella, S.; Anderson, J.W.; Adediran,
S.A.; Pratt, R.F.; Frère, J.M.; Charlier, P. Crystal structure of the Bacillus subtilis penicillin-
binding protein 4a, and its complex with a peptidoglycan mimetic peptide. J Mol Biol 2007,
3
71, 528-539. https://doi.org/10.1016/j.jmb.2007.05.071.
2
1. Sainsbury, S.; Bird, L.; Rao, V.; Shepherd, S.M.; Stuart, D.I.; Hunter, W.N.; Owens, R.J.;
Ren, J. Crystal structures of penicillin-binding protein 3 from Pseudomonas aeruginosa:
Comparison of native and antibiotic-bound forms. J Mol Biol 2011, 405, 173-184.
https://doi.org/10.1016/j.jmb.2010.10.024.
2
2.
Navratna, V.; Nadig, S.; Sood, V.; Prasad, K.; Arakere, G.; Gopal, B. Molecular basis for
the role of Staphylococcus aureus penicillin binding protein 4 in antimicrobial resistance. J
Bacteriol 2010, 192, 134-144. https://doi.org/10.1128/JB.00822-09.
2
3.
Yoshida, H.; Kawai, F.; Obayashi, E.; Akashi, S.; Roper, D.I.; Tame, J.R.; Park, S.Y.
Crystal structures of penicillin-binding protein 3 (PBP3) from methicillin-resistant
Staphylococcus aureus in the apo and cefotaxime-bound forms. J Mol Biol 2012, 423, 351-
3
64. https://doi.org/10.1016/j.jmb.2012.07.012.
2
4. King, A.M.; King, D.T.; French, S.; Brouillette, E.; Asli, A.; Alexander, J.A.; Vuckovic, M.;
Maiti, S.N.; Parr, T.R. Jr.; Brown, E.D.; Malouin, F.; Strynadka, N.C.; Wright, G.D. Structural
and kinetic characterization of diazabicyclooctanes as dual inhibitors of both serine-β-
lactamases and penicillin-binding proteins. ACS Chem Bio 2016, 11, 864-868.
https://doi.org/10.1021/acschembio.5b00944.
2
5.
lactams. In: Heterocyclic Scaffolds I. Topics in Heterocyclic Chemistry; Banik, B.; editor;
Springer: Berlin-Heidelberg, Germany, 2010, pp. 101-209.
https://doi.org/10.1007/7081_2009_12.
Troisi, L.; Granito, C.; Pindinelli, E. Novel and recent synthesis and applications of β-
2
6.
Mehta, P.D.; Sengar, N.P.S.; Pathak, A.K. 2-Azetidinone - A new profile of various
pharmacological activities. Eur Med Chem 2010, 45, 5541-5560.
J
https://doi.org/10.1016/j.ejmech.2010.09.035.

2
7.
activities.
https://doi.org/10.2174/092986711797200480.
Galletti, P.; Giacomini, D. Monocyclic β-lactams: New structures for new biological
Curr Med Chem 2011, 18, 4265-4283.
2
8. Arya, N.; Jagdale, A.Y.; Patil, T.A.; Yeramwar, S.S.; Holikatti, S.S.; Dwivedi, J.; Shishoo,
C.J.; Jain, K.S. The chemistry and biological potential of azetidin-2-ones. Eur J Med Chem
2
014, 74, 619–656. https://doi.org/10.1016/j.ejmech.2014.01.002.
2
9. Denome, S.A.; Elf, P.K.; Henderson, T.A.; Nelson, D.E.; Young K.D. Escherichia coli
mutants lacking all possible combinations of eight penicillin binding proteins: viability,
characteristics, and implications for peptidoglycan synthesis. J Bacteriol 1999, 181, 3981–
3
993.
3
0.
Paik, J.; Kern, I.; Lurz, R.; Hakenbeck, R. Mutational analysis of the Streptococcus
pneumoniae bimodular class A penicillin-binding proteins. J Bacteriol 1999, 181, 3852–3856.
3
1. Scheffers, D.J.; Errington, J. PBP1 is a component of the Bacillus subtilis cell division
machinery. J Bacteriol 2004, 186, 5153–5156. https://doi.org/10.1128/JB.186.15.5153-
5
156.2004
3
2. National Committee for Clinical Laboratory Standards. Methods for dilution
antimicrobial susceptibility tests for bacteria that grow aerobically; approved standard, 5th
ed. Document M07-A5. CLSI: Wayne, PA, USA, 2001.
3
3.
Stevens, D.L.; Bisno, A.L.; Chambers, H.F.; Dellinger, E.P.; Goldstein, E.J.; Gorbach,
S.L.; Hirschmann J.V.; Kaplan, S.L.; Montoya, J.G.; Wade, J.C.; Infectious Diseases Society of
America. Practice guidelines for the management of skin and soft tissue infections: 2014
update by the Infectious Diseases Society of America. Clin Infect Dis 2014, 59, e10-e52.
https://doi.org/10.1093/cid/ciu444.
3
2
4.
Workowski, K.A.; Bolan, G.A. Sexually transmitted diseases treatment guidelines,
015. MMWR Recomm Rep 2015, 64, 1-137.
3
5. Martucciello, S.; Paolella, G.; Muzashvili, T.; Skhirtladze, A.; Pizza, C.; Caputo, I.;
Piacente, S. Steroids from Helleborus caucasicus reduce cancer cell viability inducing
apoptosis and GRP78 down-regulation. Chem Biol Interact 2018, 279, 43-50.
https://doi.org/10.1016/j.cbi.2017.11.002
3
6.
Berman, H.M.; Westbrook, J.; Feng, Z.; Gilliland, G.; Bhat, T.N.; Weissig, H.;
Shindyalov, I.N.; Bourne P.E. The Protein Data Bank. Nucleic Acids Res 2000, 28, 235-242.
https://doi.org/10.1093/nar/28.1.235.
3
7.
Bianco, G.; Forli, S.; Goodsell, D.S.; Olson, A.J. Covalent docking using Autodock: Two-
point attractor and flexible side chain methods. Protein Sci 2016,, 25, 295-301.
https://doi.org/10.1002/pro.2733.
3
8.
Morris, G.M.; Huey, R.; Lindstrom, W.; Sanner, M.F.; Belew, R.K.; Goodsell, D.S.;
Olson, A.J. AutoDock4 and AutoDockTools4: Automated docking with selective receptor
flexibility. J Comput Chem 2009, 30, 2785-2791. https://doi.org/10.1002/jcc.21256.

3
9.
Pettersen, E.F.; Goddard, T.D.; Huang, C.C.; Couch, G.S.; Greenblatt, D.M.; Meng, E.C.;
Ferrin, T.E. UCSF Chimera--A visualization system for exploratory research and analysis. J
Comput Chem 2004, 25, 1605-1612. https://doi.org/10.1002/jcc.20084.

Legend to figures
Figure 1. Reagents and conditions for the synthesis of target compounds.
Figure 2. Cytotoxicity of cephalosporin derivatives. (A) Effect on cell viability of MRC5 cells
after 24 h exposure to compounds 23b-c, 24a, 24c-d, 25a-b, 26a, 27a (dose-range 0.5–50
1
3
μg/ml). Data for 23a were previously published. . (B) Effect on cell viability of Calu-1 cells
after 24 h exposure to compounds 23b-c, 24a, 25b (dose-range 0.5–50 μg/ml). Data are
reported as mean ± SD from two or three independent experiments, each in triplicate.
Vehicle (DMSO), at the highest concentrations used (0.4% v/v), induced a reduction of cell
viability of about 15% (not shown). *p<0.05 vs. the respective vehicle.

Graphical abstract