Design, synthesis, biological evaluation and in silico studies of certain aryl sulfonyl hydrazones conjugated with 1,3-diaryl pyrazoles as potent metallo-β-lactamase inhibitors

Article abstract of DOI:10.1016/j.bioorg.2020.104386

Based on a structure-guided approach, aryl sulfonyl hydrazones conjugated with 1,3-diaryl pyrazoles were designed to target metallo-β-lactamases (MBLs), using Klebsiella pneumoniae NDM-1 as a model. The in vitro MBLs inhibition showed remarkable inhibition constant for most of the designed compounds at a low micromolar range (1.5–16.4 μM) against NDM-1, IMP-1 and AIM-1 MBLs. Furthermore, all compounds showed promising antibacterial activity against (K+, K1-K9) resistant clinical isolates of K. pneumoniae and were able to re-sensitize resistant K. pneumoniae (K5) strain towards meropenem and cefalexin. Besides, in vivo toxicity testing exhibited that the most active compound was non-toxic and well tolerated by the experimental animals orally up to 350 mg/kg and up to 125 mg/kg parenterally. The docking experiments on NDM-1 and IMP-1 rationalized the observed in vitro MBLs inhibition activity. Generally, this work presents a fruitful matrix to extend the chemical space for MBLs inhibition. This aids in tackling drug-resistance issues in antibacterial treatment.

Full text of DOI:10.1016/j.bioorg.2020.104386

Journal Pre-proofs
Design, synthesis, biological evaluation and in silico studies of certain aryl
sulfonyl hydrazones conjugated with 1,3-diaryl pyrazoles as potent metallo-β-
lactamase inhibitors
Marwa M. Shaaban, Hanan M. Ragab, Kenichi Akaji, Ross P. McGeary,
Alaa-Eldin A. Bekhit, Waleed M. Hussein, Julia L. Kurz, Bassma H. Elwakil,
Salma A. Bekhit, Tamer M. Ibrahim, Mona A. Mahran, Adnan A. Bekhit
PII:
S0045-2068(20)31684-9
DOI:
https://doi.org/10.1016/j.bioorg.2020.104386
YBIOO 104386
Reference:
To appear in:
Bioorganic Chemistry
Received Date:
Revised Date:
Accepted Date:
5 August 2020
6 September 2020
15 October 2020
Please cite this article as: M.M. Shaaban, H.M. Ragab, K. Akaji, R.P. McGeary, A.A. Bekhit, W.M. Hussein, J.L.
Kurz, B.H. Elwakil, S.A. Bekhit, T.M. Ibrahim, M.A. Mahran, A.A. Bekhit, Design, synthesis, biological
evaluation and in silico studies of certain aryl sulfonyl hydrazones conjugated with 1,3-diaryl pyrazoles as potent
metallo-β-lactamase inhibitors, Bioorganic Chemistry (2020), doi: https://doi.org/10.1016/j.bioorg.2020.104386
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.
© 2020 Published by Elsevier Inc.

Design, synthesis, biological evaluation and in silico
studies of certain aryl sulfonyl hydrazones conjugated
with 1,3-diaryl pyrazoles as potent metallo-β-
lactamase inhibitors
Marwa M. Shaaban1, Hanan M. Ragab1, Kenichi Akaji2, Ross P. McGeary3, Alaa-Eldin A. Bekhit4,
Waleed M. Hussein3,5, Julia L. Kurz3, Bassma H. Elwakil6, Salma A. Bekhit7, Tamer M. Ibrahim8*,
Mona A. Mahran1, Adnan A. Bekhit1,9,10*
1Department of Pharmaceutical Chemistry, Faculty of Pharmacy, Alexandria University, Alexandria
21521, Egypt
2Department of medicinal Chemistry, Kyoto Pharmaceutical University, Japan
3School of Chemistry and Molecular Biosciences, The University of Queensland, Brisbane, QLD
4072, Australia
4Food Sciences, University of Otago, Dunedin, New Zealand
5Department of Pharmaceutical Organic Chemistry, Faculty of Pharmacy, Helwan University, Ein
Helwan, Helwan, Egypt

6Department of Medical Laboratory Technology, Faculty of Allied Medical Science, Pharos
University in Alexandria, Egypt
7High Institute of Public Health, Alexandria University, Alexandria 21568, Egypt
8Department of Pharmaceutical Chemistry, Faculty of Pharmacy, Kafrelsheikh University,
Kafrelsheikh 33516, Egypt.
9Cancer Nanotechnology Research Laboratory (CNRL), Faculty of Pharmacy, Alexandria
University, Alexandria 21521, Egypt
10Pharmacy Program, Allied Health Department, College of Health and Sport Sciences, University
of Bahrain, P.O. Box 32038, Kingdom of Bahrain
Corresponding authors:
*Tamer M. Ibrahim,
E-mail: tamer.ibrahim2@gmail.com
tamer_mohamad@pharm.kfs.edu.eg
*Adnan A. Bekhit,
E-mail: adnbekhit@pharmacy.alexu.edu.eg
Highlights



New conjugates of aryl sulfonyl hydrazone and 1,3-diaryl pyrazole were designed and synthesized.
They showed promising inhibition against NDM-1, IMP-1 and AIM-1 MBLs.
They exhibited promising ex vivo antibacterial activity against (K+, K1-K9) resistant clinical isolates of
K. pneumoniae.


They were able to re-sensitize resistant K. pneumoniae (K5) strain towards meropenem and
cefalexin.
The in vivo toxicity testing of the most active compound was non-toxic and well tolerated.



The docking experiments on NDM-1 and IMP-1 rationalized the observed in vitro MBLs inhibition
activity.
The in silico predictions displayed promising pharmacokinetics and drug-likeness profiles of the
active compounds.
KEYWORDS
sulfonyl hydrazones; pyrazoles; MBLs; NDM-1; IMP-1; AIM-1; K. pneumonia; inhibition assay;
liposome; docking
ABSTRACT
Based on a structure-guided approach, aryl sulfonyl hydrazones conjugated with 1,3-diaryl pyrazoles
were designed to target metallo-β-lactamases (MBLs), using Klebsiella pneumoniae NDM-1 as a
model. The in vitro MBLs inhibition showed remarkable inhibition constant for most of the designed
compounds at a low micromolar range (1.5 – 16.4 µM) against NDM-1, IMP-1 and AIM-1 MBLs.
Furthermore, all compounds showed promising antibacterial activity against (K+, K1-K9) resistant
clinical isolates of K. pneumoniae and were able to re-sensitize resistant K. pneumoniae (K5) strain
towards meropenem and cefalexin. Besides, in vivo toxicity testing exhibited that the most active
compound was non-toxic and well tolerated by the experimental animals orally up to 350 mg/kg and
up to 125 mg/kg parenterally. The docking experiments on NDM-1 and IMP-1 rationalized the
observed in vitro MBLs inhibition activity. Generally, this work presents a fruitful matrix to extend
the chemical space for MBLs inhibition. This aids in tackling drug-resistance issues in antibacterial
treatment.

INTRODUCTION
Antibiotic resistance is an increasing problem worldwide affecting human health[1]. A growing
number of infections such as pneumonia, tuberculosis, gonorrhoea and salmonellosis are becoming
harder to treat as standard antibiotics used for their treatment are exhibiting low effectiveness [2]. β-
lactam antibiotics is a cheap and effective class of antimicrobial agents. Despite the diversity of their
structures, they share a common four-membered β-lactam ring, which functions as the active
pharmacophore for this class. The production of β-lactamase enzymes (β-lactamases) is the most
common resistance strategy that bacteria use against β-lactam antibiotics [3]. These enzymes can
hydrolyze the β-lactam ring, rendering this class of antibiotics inactive. Based on amino acid
sequences, β-lactamases have been grouped into four molecular classes (A, B, C and D)[4]. The
three major classes A, C and D are called serine-β-lactamases (SBLs) due to the presence of a serine
residue in their structure, which reacts with the carbonyl carbon of the β-lactam ring resulting in ring
opening. These A, C and D β-lactamases are active against many penicillins, cephalosporins and
monobactams[5]. The class B is known as metallo-β-lactamases (MBLs) and are characterized by
the presence of a divalent zinc ion. These enzymes use zinc bound hydroxyl group to inactivate most
of penicillins, cephalosporins and carbapenems [6]. Based on sequence identity and zinc ion
dependence, MBLs are classified into three subclasses B1, B2 and B3. Interestingly, the enzymes of
B1 subclass (including IMP-1, VIM-2 and NDM-1) have emerged as the most clinically significant
[7, 8]. For instance, New Delhi metallo-β-lactamase (NDM-1) represents a global threat to human
health due to its ability to hydrolyze almost all β-lactam antibiotics[7, 9, 10]. In addition, subclass
B3 MBLs, such as SMB-1 and AIM-1, were also reported to be widely distributed among
pathogenic bacteria, resulting in a serious clinical threat [11].

Accordingly, there are two main strategies to overcome the resistance mechanisms against β-lactam
antibiotics and preserve their utility. The first strategy is the discovery of novel β-lactam antibiotics
that can resist bacterial enzymatic inactivation. The second is a combination therapy that involve
combining the β-lactam with an agent that can inhibit β-lactamases so the partner β-lactam can
safely reach the target site [12]. Furthermore , a recent strategy employs liposomal drug
formulations to overcome non-enzymatic resistance through fusion interaction between the
liposome-containing antibiotic and bacterial membrane phospholipids [13]. Generally, combination
therapy has been the frequently used to overcome the enzymatic resistance. However, all clinically
available inhibitors, such as clavulanic acid, sulbactam, tazobactam and Avibactam, are SBL
inhibitors and no reports of clinically useful MBL inhibitors are available [14]. Therefore, there is an
urgent need for the development of MBL inhibitors [15].
Several compounds have been reported as potential MBL inhibitors, including compounds, which
can coordinate with the zinc ions in the enzyme active site, such as carboxylic acids and thiol
containing molecules [16, 17]. Furthermore, N-aryl sulfonyl hydrazones are reported to show
promising inhibitory activity against IMP-1 and NDM-1 enzymes [16, 18, 19]. It is found that the
enzyme inhibitory activity significantly increases when the sulfonyl hydrazone is substituted with
bulky aromatic substituents possessing little electron withdrawing capacity [20]. In addition, the
reported design of N-aryl sulfonyl hydrazone derivatives is motivated by the resemblance of their
backbone to cephalosporins, the natural substrates of MBLs [18, 19] (Figure 1). The modeled
structures of N-aryl sulfonyl hydrazine derivatives appeared to coordinate with both zinc ions via the
oxygen atoms of the sulfonyl group [18]. However, only few studies investigated the MBL
inhibitory activity of pyrazole derivatives [16, 21].

This encouraged us endeavoring to extend the chemical space for MBL inhibitory activity using
conjugates of sulfonyl hydrazones (II) [15] and N-aryl pyrazoles (III) [16, 21] to produce IV
(Figure 1). As per a structure-guided approach to target MBLs, our attention dedicated mainly on
the co-crystal structure of K. pneumonia NDM-1 as a model target. The idea is to selectively
coordinate with at least one zinc ion by an oxygen atom of the sulfonyl hydrazone moiety.
Moreover, the 3-arylpyrazolyl moiety would be packed against the hydrophobic wall of the flexible
B3 loop and surrounded by the catalytic hydrophobic residues. Globally, this approach would
subsequently block the catalytic function of the NDM-1. A summary of the design approach is
illustrated in Figure 1. We synthesized only focused series of these compounds as a starting
approach in this study to confirm our design. To experimentally validate such approach, we tested
the ability of our designed compounds to experimentally inhibit NDM-1, as well as IMP-1, and
AIM-1 MBLs.
Figure 1. Design approach of sulfonyl hydrazones conjugated with N-aryl pyrazoles (IV). The Ki
values are for the competitive mode. The postulated binding poses are an output of docking

experiments. Polar and non-polar regions of the binding site were presented by red and green colored
molecular surface, respectively.
RESULTS AND DISCUSSION
Chemistry
The synthetic pathway of the designed compounds 4a-m is shown in Scheme 1.
R2
R2
O
S
HN
N
R1
CHO
O
R1
R1
N
NH2NH2
N
4a. R1 = H, R2 = H, R3 = H
O
O
N
N
4b. R1 = H, R2 = H, R3 = Br
4c. R1 = H, R2 = H, R3 = NO2
4d. R1 = H, R2 = CH3, R3 = H
4e. R1 = H, R2 = CH3, R3 = CH3
4f. R1 = H, R2 = Br, R3 = H
4g. R1 = H, R2 = Cl, R3 = H
4h. R1 = CH3, R2 = H, R3 = H
4i. R1 = CH3, R2 = H, R3 = Br
4j. R1 = CH3, R2 = H, R3 = NO2
4k. R1 = CH3, R2 = CH3, R3 = H
4l. R1 = CH3, R2 = Br, R3 = H
4m. R1 = CH3, R2 = Cl, R3 = H
S
NH2
S
N
Cl
O
O
H
1a, b
2a, b
R3
R3
1a: R1 = H
1b: R1 = CH3
1a: R1 = H
1b: R1 = CH3
4a-m
3a-g
3a. R2 = H, R3 = H
3b. R2 = H, R3 = Br
3c. R2 = H, R3 = NO2
3d. R2 = CH3, R3 = H
3e. R2 = CH3, R3 = CH3
3f. R2 = Br, R3 = H
3g. R2 = Cl, R3 = H
Scheme 1. scheme representing the synthesis of sulfonyl hydrazone derivatives 4a-m
The sulfonyl hydrazone derivatives were reported to be prepared by condensation of aldehyde with
arylsulfonohydrazide by various methods and under different conditions. The most common
methods are either heating under reflux in absolute ethanol only[18], or in the presence of catalytic
amount of triethyl amine [22], piperidine [23], acetic acid [24], or conc. HCl [25]. Other solvents
could be used such as dioxane [26], benzene [27], tetrahydrofuran [28], or methanol [29]. Also,
microwave-assisted irradiation was reported for the synthesis of sulfonyl hydrazine derivatives [30].
In the present study, our designed compounds 4a-m were obtained by condensation of aryl

sulfonohydrazides 2a,b with various aldehydes 3a-g in hot ethanol [18]. This employs green and
environment - friendly conditions.
IR spectra of compounds 4a-m showed characteristic absorption band at 3100-3300 cm-1 attributed
to the NH group and two sharp bands at 1320-1370 cm-1 characteristic for SO2 group. They also
lacked the characteristic absorption band of aldehydic carbonyl group, which confirmed the
synthesis of the target compounds 4a-m. 1H-NMR spectra of the target compounds 4a-m
demonstrated the most deshielded singlet peak at 10.71-11.41 ppm attributed to D2O-exchangeable
NH proton. They also showed two characteristic singlet peaks at 8.02-9.09 ppm attributed to CH=N
and pyrazole methine proton. Spectra of compounds 4h-m displayed singlet peak at 2.36 ppm
attributed to three methyl protons of toluenesulfonohydrazide. 13C-NMR spectrum of compound 4k
showed two characteristic peaks at 21.34 and 21.47 ppm attributed to two CH3, another characteristic
peak at 143.79 ppm attributed to CH=N. Other signals appeared at their expected chemical shifts.
EIMS of compound 4f showed molecular ion peak at m/z 482 (M+∙ +2) corresponding to
C22H1781BrN4O2S; m/z 480 (M+∙) corresponding to C22H1779BrN4O2S in the ratio (1:1) and the base
peak at m/z 370. Representative IR, 1H-NMR, 13C-NMR and EIMS charts can be found in the
Supplementary Material.
Biological evaluation
In vitro MBLs inhibition activity against IMP-1, NDM-1, and AIM-1
The inhibitory effects of the synthesized compounds 4a–m were assessed against three MBLs (IMP-
1, NDM-1, and AIM-1). Compounds showing promising inhibition effects (>80% against IMP-1)
were selected for further kinetic studies to determine their inhibition constants (Ki values) and

modes of inhibition, i.e. competitive or uncompetitive inhibition. The observed % inhibition and
inhibition constants are shown in Table 1.
Table 1. Inhibition percentage and Inhibition constants of the tested compounds against IMP-1,
NDM-1 and AIM-1 enzymes.
% Inhibition
AIM-1
NDM-1
IMP-1
Comp
No.
Kica
Kiucb
(µM)
Kic
Kic
AIM-1
NDM-1 IMP-1
Kiuc
(µM)
Kiuc
(µM)
(10µM) (5µM)
(5µM)
(µM)
(µM)
(µM)
4a
4b
4c
4d
4e
4f
0%
26%
81%
74%
59%
73%
73%
77%
75%
74%
82%
77%
81%
37%
87%
92%
62%
76%
81%
85%
83%
76%
74%
81%
89%
NDc
ND
ND
ND
ND
ND
57%
42%
26%
42%
52%
47%
46%
43%
48%
64%
53%
13.5±5.2 11.9±1.7 3.6±1.9
6.5±1.5
4.8±1.2
9.3±1.2
5.1±1.3
ND
9.3±1.2
ND
11.8±2.4 6.1±0.8
12.2±1.6 2.8±1.1
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
12.8±1.8 14.1±3.5 5.2±1.7
10.8±1.6 12.6±3.9 4.6±1.4
13.6±2.1 3.8±1.4
10.6±2.4 2.8±0.7
7.1±1.6
3.9±0.8
2.2±0.3
ND
4g
4h
4i
8.9±2.6
ND
16.2±3.4 2.6±1.2
6.4±0.8
ND
2.4±0.2
ND
ND
ND
4j
13.2±3.6 14.2±4.2 3.6±0.6
12.4±2.8 13.6±3.6 6.8±2.2
4.4±0.8
ND
ND
4k
4l
16.4±2.5 5.4±1.4
4.8±0.9
10.8±2.4
8.5±2.4
16.0±3.0 2.2±0.4
2.2±1.1
4.3±1.8

4m
43%
81%
86%
12.6±4.3 15.5±2.1 1.5±0.5
3.7±1.0
6.3±1.7
15.1±2.7
aKic: inhibition constant for the competitive inhibition mode
bKiuc: inhibition constant for the uncompetitive inhibition mode
cND: Not determined.
Most of the tested compounds exhibited both competitive and uncompetitive inhibition modes. This
mixed inhibition mode indicated that these inhibitors were capable of both binding in the active site
of MBLs (competitive inhibition) and forming a ternary enzyme-substrate-inhibitor (ESI) complex
which inhibits hydrolysis of the substrate (uncompetitive inhibition) [31]. These examined
compounds showed promising inhibition constant at a low micromolar range against NDM-1 (2.6-
6.8 µM) for the competitive mode, as well as against IMP-1 for both the competitive and non-
competitive modes via 2.2-5.4 µM and 2.2-7.1 µM, respectively. However, inhibition constant range
is slightly weaker against AIM-1 and for the non-competitive mode of NDM-1. This certainly
indicates broad-spectrum inhibitory activity including both B1 and B3 MBL.
It is worth mentioning that the inhibition constant against the three MBLs is potentiated by the
introduction of bulky substituents at positions R2 and R3 as demonstrated by compounds 4b, 4f, 4i
and 4l (bearing a bromo substituent), compounds 4g and 4m (having a chloro substituent) as well as
compounds 4c and 4j (containing a nitro group). These compounds showed remarkable inhibition
constant which can be rationalized by the increased hydrophobicity and optimum favorable
interactions with the hydrophobic B3 loop as discussed in the Molecular Modeling section for both
NMD-1 and IMP-1 enzymes. Also, 4-tolyl sulfonyl hydrazone derivatives (compounds 4h, 4j-m)

and their congeneric phenyl sulfonyl hydrazone derivatives demonstrated roughly comparable
inhibition constants. This attributed to the ability of their pheny/tolyl groups to form arene-H
interactions with the His122 coordinating to one zinc ion (Molecular Modeling section and Figure
2), and hence stabilizing the binding pose to block the catalytic site for NDM-1 for instance.
Generally, owing to the novelty of such hybrid strategy as MBLs inhibitor chemotype, we would
follow further studies on them in the near future with broader collection of substituents, in order to
extract sharper insights of the structure-activity relationships.
Antibacterial activity against K. pneumoniae
The synthesized compounds 4a-m were screened for their antimicrobial activity against resistant K.
pneumoniae clinical isolates (K+, K1-K9) [32]. The aim of this study was to determine the intrinsic
antibacterial activity of these compounds on selected resistant clinical isolates that showed multi-
drug resistance to the clinically available antibiotics. Minimum inhibitory concentrations (MIC) of
the tested compounds are shown in Table 2.
Table 2. Antibacterial activity testing against K. pneumoniae clinical isolates.
MICa(µg/ml)
K+ K1 K2 K3 K4
MIC(µM)
K4 K5
159 159
66
143 143 143
Comp
4a
K5
64
64
64
64
64
64
32
64
32
32
K6 K7 K8 K9 K.+ K1
K2
80
K3
K6
159
133
143
77
K7
80
66
36
38
K8
80
K9
159
133
143
77
32
32
32
32
32
32
32
32
16
16
32
32
32
32
32
32
32
16
32
32
32
64
64
64
64
64
32
32
64
32
64
64
64
32
32
32
64
64
64
64
64
32
64
64
64
32
64
32
32
16
64
64
64
32
64
32
64
64
32
32
32
32
16
16
64
32
64
16
32
16
32
32
32
64
64
64
32
16
32
32
64
64
64
32
64
32
64
64
32
32
80
66
72
77
74
66
73
77
32
35
80
66
72
77
74
66
73
38
65
69
159
133
143
154
149
133
73
4b
4c
133 133
66
72
4d
4e
154
149
133
73
77
74
66
154
149
66
154
149 149 149 149
4f
66
66
133
73
38
65
69
66
146
154
65
4g
4h
4i
146 146
146 146
77
154
77
65
35
154
65
154
65
38
65
35
129 129
69 139
4j
69
69
69

4k
4l
32
32
16
-
32
32
32
-
32
64
64
-
64
64
64
-
32
64
32
-
64
32
32
64
32
-
64
64
32
-
32
32
32
-
32
64
64
-
74
65
35
-
74
65
71
-
74
149
74
149
65
74
149
74
65
71
-
74
129
142
-
129 129 129
129 129
4m
64
142 142
71
-
142
2670
2947
71
-
71
-
MEMa
LEXb
1024
1024
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
aMEM: meropenem
bLEX: cefalexin
Results from antibacterial screening showed that 4i, 4j and 4m were the most active antimicrobial
agents against clinical isolate K+ compared with the antibiotics meropenem and cefalexin. Likewise,
4h showed the best antimicrobial activity against clinical isolate K1, while 4j and 4g were most
active against K2. For the clinical isolate K3, 4f and 4e displayed the best antimicrobial activity. 4j
was the most active against the clinical isolate K4, while 4i and 4l exhibited the best antimicrobial
activity against the clinical isolate K5. For the clinical isolate K6, 4i and 4f displayed the best
antimicrobial activity. 4j and 4c were the most active against the clinical isolate K7, while 4h was
the most active against the clinical isolate K8. For the clinical isolate K9, 4i and 4f exhibited the best
antimicrobial activity. The rest of the compounds showed moderate to low antibacterial activity. It
was found that clinical isolate K3 and K5 exhibited the highest resistance level to the test
compounds.
MBLs inhibition and re-sensitization of resistant K. pneumoniae(K5)
The synthesized inhibitors 4a-m were tested at sub-inhibitory concentration in combination with the
clinically available β-lactam antibiotics (cefalexin and meropenem) against the most resistant
clinical isolate of K. pneumoniae (K5) to investigate whether these inhibitors would be capable of re-
sensitizing K5 to the effect of antibiotics. The use of sub-inhibitory concentration was important to
confirm that the inhibitor would not serve as partner to antibiotic in combating the bacteria[33].

Checking the clinically available antibacterial combinations of antibiotics and β-lactamase inhibitors
to determine the molar ratio of antibiotic to inhibitor revealed that there is no available standard ratio
(the molar ratio is 1.4:1 in UNASYN®[34], 2:1 in AUGMENTIN®[35] and 5:1 in ZOSYN®[36]).
Accordingly, the molar ratio was selected to be 2:1.
Minimum inhibitory concentration (MIC) values of the combination against K5 were determined.
Also, improvement in antibiotic activity (% re-sensitization of k. pneumonia K5 to antibiotics) was
calculated according to the following equation (Eq.1). The combination of MIC values and %
improvement in the antibiotic activity are shown in Table 3.
Initial MIC ― Final MIC
Improvement in antibiotic activity (%) =
* 100
(Eq.1)
Initial MIC
Interestingly from Table 3, all tested compounds effectively sensitized resistant k. pneumonia K5 to
the effect of cephalexin and meropenem, producing 95.9-99.8% improvement in MIC values of these
antibiotics. Compound 4m was the most active MBLs inhibitor producing approximately 99.8%
improvement in the activity of both cephalexin and meropenem, while compound 4e showed 99.5%
increase in the meropenem activity. These results confirm the previously mentioned results of in
vitro MBLs inhibition activity of these compounds.
Table 3. Re-sensitization of resistant K. pneumoniae (K5) clinical isolate to cephalexin and
meropenem by MBLs inhibitors.
Improvement
Final MIC (µg/ml)
Final MIC (µM)
in antibiotic
activity (%)
Comp
No.
Comp+MEM (A)a Comp+LEX (B) b Comp in (A)c MEM in (A)d Comp in (B)e LEX in (B)f MEM LEX
4a
4b
16
32
64
64
13.7
25.6
27.4
51.3
58.3
54.4
117
109
98.9
98.1
96.0
96.3

4c
4d
4e
4f
16
16
8
32
64
64
64
64
64
32
16
32
13.2
13.5
6.68
12.8
13.3
54.1
50.7
26
26.4
27
28
56
99.0
98.9
99.5
99.0
99.0
95.9
96.2
98.1
98.0
98.1
96.1
96.1
96.3
96.1
96.1
98.2
99.1
98.1
57.6
56.9
54.4
56.6
57.6
26.9
13.8
28.4
115
114
109
113
115
53.8
27.7
56.9
13.4
25.6
26.6
108
101
52.1
53.4
16
16
64
64
32
32
4g
4h
4i
4j
4k
26.7
4l
64
4
32
4
50.7
3.28
101
26.9
3.49
53.8
6.98
96.2
99.8
98.1
99.8
4m
6.57
MEM: meropenem
LEX: cefalexin
aComp+MEM (A): combination of test compound and meropenem with a molar ratio of 1:2, respectively.
bComp+LEX (B): combination of test compound and cefalexin with a molar ratio of 1:2, respectively.
cComp in (A): compound concentration in the Comp+MEM combination.
dMEM in (A): meropenem concentration in the Comp+MEM combination.
eComp in (B): compound concentration in the Comp+LEX combination.
fLEX in (B): cefalexin concentration in the Comp+LEX combination.
In vivo study of the synthesized inhibitors combined with antibiotics cefalexin and meropenem
using rats infected with K. pneumoniae (K5).
The most active inhibitor 4m combined with antibiotics (meropenem and cefalexin) as well as
liposomal combination of 4m and meropenem (discussed in the next section) were further subjected
to in vivo antibacterial testing using rats infected with K. pneumoniae (K5). The number of survived
animals is shown in Figure 2.

Figure 2. In vivo survival curve of rats infected with K. pneumoniae (K5). Different groups are
illustrated in the legend. LEX is cefalexin and MEM is meropenem antibiotics.
One animal from each study group of therapy presented with pneumonia was killed at the beginning,
with median counts of 7.0 log10 CFU/g of lung. It was found that all the positive and negative control
animal groups died within 4 to 5 days from the start of the experiment, while animals receiving
conventional and liposomal antibiotic-inhibitor combinations were alive during the treatment period.
Full bacterial clearance was observed after 6 days in case of groups receiving conventional
combinations and after 5 days in group receiving liposomal combination.
In vivo acute toxicity testing.
The most active inhibitor 4m was tested for its toxicity in mice at concentrations up to 350
mg/Kg/day. The treated mice did not show any toxicity signs after treatment. In addition, there was
no significant difference in the weight of the mice and no death cases was recorded during 3 days of
observation post administration of the test compound. Thus, it could be concluded that compound
4m was non-toxic and was well tolerated by the experimental animals orally up to 350 mg/kg and up
to 125 mg/kg parenterally.
Preparation of dual drug loaded liposomes
The antibiotic meropenem and the most active inhibitor 4m were encapsulated into liposomes, and
dual drug loaded liposomes were developed by ethanol injection method through injecting an
ethanolic lipid solution in water. Initially, liposomes were prepared using lipoid S100 and
cholesterol for efficient incorporation of meropenem and compound 4m. The mean particle size of
dual drug-loaded liposomes was 116.2 nm, which is a slightly higher than blank liposomes (88.44
nm) (Table 4). The slight increase in particle size of the drug-loaded liposomes might confirm the

entrapment of free drugs into liposomes. On the other hand, the zeta potential of loaded liposomes
was -14.3 mV indicating a good colloidal stability. Both meropenem and 4m were successfully
encapsulated into liposomes with encapsulation efficiency of 100 % w/w (Table 4).
Table 4. Characterization of dual drug loaded liposomes.
Formula
Particle size
(nm)
polydispersity
index
Zeta potential
(mV)
encapsulation
efficiency
Blank liposomes
88.44
116.2
0.185
0.324
-12.8
-14.3
-
Dual drug loaded
liposomes
100%
Superior to the conventional antibiotic-4m combination (Figure 2), the liposomal formulation
displayed 99.9% re-sensitization of k. pneumonia K5 to the effect of meropenem. Theoretically, the
encapsulation of the meropenem would enhance the antimicrobial activity due to fusional
interactions between the bacterial membrane phospholipids and liposome [37].
Molecular Modeling
Molecular Docking
The aim of this section in our study is to provide in silico insights to rationalize the observed in vitro
MBLs inhibition activity based on our design approach.
Selection of a protein model
Unlike AIM-1, there are several X-ray crystal structures for IMP-1 and NDM-1 complexed with
different ligands of various topological and size features. This reflects structural data and flexibility
features of the amino acids involved in the binding event, and subsequently, an appropriate model
complexed with comparable ligand to our compounds can be selected. For instance, B3 (hairpin)
loop is reported to be highly flexible for all MBL subclasses B1 to facilitate substrate binding and

enhance catalytic function for the MBL [16, 38-40]. Upon ligand binding, the L3 loop is observed to
re-order and offer a hydrophobic wall for interaction [16]. Therefore, to select a model from the
Protein Data Bank (PDB) for NDM-1 and to account for its B3 loop flexibility, we extracted high
resolution X-ray coordinates of all PDB structures of NDM-1 from K. pneumonia. Then, a focus was
dedicated on the crystal structures complexed with ligands of molecular weight (Mwt) > 401, which
are comparable in topological features to our synthesized compounds (Mwt of our compounds
402.47 – 495.39). This ended up with 5 crystal structures roughly representing the available
conformational space of B3 loop involved in the ligand-protein binding event. Elucidating the
backbone and side chain conformations of B3 loop in these 5 structures (Figure S1, Supplementary
material), a major difference can be observed for the PDB structure complexed with the smallest
ligand (PDB code: 5N0H – ligand Mwt: 402.47) compared to other structures, while only minor
differences can be observed for the rest structures. Therefore, we selected two PDB structures as
models for our docking experiments representing two major conformations for the B3 loop of NDM-
1, (PDB code: 5O2E – ligand Mwt: 442) and (PDB code: 5N0H).
For IMP-1, we nominated the X-ray structure complexed with the heaviest ligand (Mwt: 429.52,
PDB code: 1DD6) due to limited availability of PDB structures compared to NDM-1, which would
represent a proper and best available conformation of the B3 loop.
Docking protocol and description of the docking poses
We employed two docking tools, AutoDock Vina[41] and MOE-Dock module[42], for extracting
consensus ranking of our designed compounds. For NDM-1, the results showed that the docking
against the PDB structure complexed with the heaviest ligand (PDB code: 5O2E) exhibited superior
scores compared to the structure complexed with the smallest ligand (PDB code: 5N0H; data no

shown). This is likely attributed to the fact that the topological and size features of our docked
compounds (Average Mwt 449.76 ± 31.17) is better accommodated in a bigger binding site
complexed with a ligand of comparable size (Mwt of X-ray ligand of PDB:5O2E is 442).
For both docking tools, similar conclusions can be outlined, and all best-ranked poses showed
interactions with the key catalytic zinc ions in the binding site. Also, narrow score range for both
docking tools can be observed for the docking poses of the in vitro active compounds 4b, 4c, 4f, 4g,
4h, 4k, 4l and 4m, as shown in Table 5. This is in coherence with the observed in vitro MBL
inhibition that displayed narrow low micromolar range of Kic values.
Elucidating the postulated binding poses and the interaction pattern, we focused on the most three
active compounds 4h, 4l and 4m for NDM-1, as shown in Figure 3. The postulated binding pose of
4h demonstrated noticeable interactions with both zinc ions via the sulfono group. 3-phenylpyrazolyl
moiety and the tolyl group attached to the sulfonohydrazide moiety exhibited favorable arene-H
contacts with Ile35 and His122, respectively. Also, the 3-phenylpyrazolyl moiety appeared to be
directed towards the hydrophobic side chains of Met69 and Phe70 forming the flexible B3 loop
signifying favorable hydrophobic interactions. Finally, the whole 1-phenylpyrazolyl moiety
appeared to be packed between the Ile35, His250, Lys211 and Ser251 indicating favorable contacts.
The postulated binding poses of 4m and 4l reproduced similar interactions to 4h pose. Furthermore,
their halogenated 3-arylpyrazolyl moiety would enhance the hydrophobic interactions with the
hydrophobic B3 loop of NDM-1. In general, these docking solutions recommended satisfactory in
silico binding affinity towards NDM-1 which is in-coherence with the competitive mode of
inhibition observed from the in vitro MBLs inhibition.

Table 5. The docking score of the active compounds in the binding site of NDM-1 and IMP-1.
NDM-1
IMP-1
Cpd.
MOE-Dock AutoDock Vina* MOE-Dock AutoDock Vina
4b
4c
4f
-8.17
-8.12
-8.00
-7.98
-8.03
-8.24
-8.26
-8.16
-8.2
-8.6
-8.6
-8.5
-8.4
-8.6
-9.2
-9.2
-7.12
-7.14
-6.89
-7.08
-7.66
-7.17
-7.53
-7.40
-9.5
-9.5
-8.7
-8.9
-9.4
-9.1
-8.4
-8.4
4g
4h
4k
4l
4m
*The docking score of AutoDock Vina is expressed as binding affinity (kcal/mol).

Figure 3. The docking poses of 4h, 4m and 4l in the binding site of NDM-1 (PDB code: 5O2E) for
A, B and C, respectively. Polar and non-polar regions of the binding site were presented by red and
green colored molecular surface, respectively. Zinc ions were represented as cyan spheres. Dashed
lines indicate favorable interactions. Non-polar hydrogen atoms were omitted for clarity.
Like NDM-1, we expounded the postulated binding poses of the most active compounds 4h, 4g and
4c in the binding site of IMP-1 (Figure 4). All binding poses appeared to coordinate with at least
one zinc ion via the sulfono group. Interestingly, the tolyl or phenyl group attached to the
sulfonohydrazide moiety demonstrated favorable arene-H or hydrophobic contact with the side chain
of Trp28 of the flexible B3 loop. Such docking solution suggests an acceptable in silico binding
affinity towards IMP-1 via interacting with its key residues. This is consistent with the competitive
mode of inhibition observed from the kinetic assays.

Figure 4. The docking poses of 4c, 4g and 4h in the binding site of IMP-1 (PDB code: 1DD6) for A,
B and C, respectively. Zinc ions were represented as cyan spheres. Dashed lines indicate favorable
interactions. Non-polar hydrogen atoms were omitted for clarity.
In silico physicochemical predictions:
The comprehensive discussion of this part can be found in the Supplementary material. In
conclusion, all molecules showed no alert to be a possible PAINS (pan-assay interference
compounds) [43], as shown in Table S1 in the Supplementary material. This signifies that the in
vitro determined biological results to be robust. Besides, drug-likeness properties (Table S1)
highlighted that all molecules do not display violations for the Lipinski and Veber rules [44]
signifying that these candidates would demonstrate good bioavailability profiles.

CONCLUSION
We synthesized new conjugates of aryl sulfonyl hydrazones with 1,3-diaryl pyrazoles in a
synthetically feasible and green method. Based on a structure-guided approach, we used Klebsiella
pneumoniae NDM-1 as a model to target MBLs. The in vitro MBLs inhibition confirmed our design
approach and showed outstanding inhibition constant for most compounds at a low micromolar
range against NDM-1, IMP-1 and AIM-1 MBLs. In addition, all compounds exhibited promising ex
vivo antibacterial activity against (K+, K1-K9) resistant clinical isolates of K. pneumoniae.
Furthermore, all compounds were able to re-sensitize resistant K. pneumoniae (K5) strain towards
meropenem and cefalexin with 95.9 – 99.8 % improved antibiotic activity. Further in vivo validation
showed that the infected rates with resistant K. pneumoniae (K5) were able to survive and exhibited
full bacterial clearance after 6 and 5 days when subjected to a conventional and liposomal
combination of meropenem and the MBL inhibitor 4m, respectively. Besides, in vivo toxicity testing
displayed that 4m was non-toxic and well tolerated by the experimental animals orally up to 350
mg/kg and up to 125 mg/kg parenterally. The docking experiments on NDM-1 and IMP-1
rationalized the observed in vitro MBLs inhibition. The in silico predictions displayed promising
pharmacokinetics and drug-likeness profiles of the active compounds. Generally, this work presents
a fruitful matrix to extend the chemical space for metallo-β-lactamases inhibition. This certainly
helps in tackling drug-resistance issues in antibacterial treatment.
EXPERIMENTAL
Chemistry
General procedure for the synthesis of sulfonyl hydrazide derivatives 2a,b.
The appropriate sulfonyl chloride (2.5mmol) was gradually added to a stirred solution of hydrazine
hydrate 99% (0.4 g, 0.4 ml, 12.5 mmol) in CH2Cl2 (5 ml). The reaction mixture was stirred for 15

min, then the pH was adjusted to approximately 11 using 10% aqueous Na2CO3. The aqueous phase
was extracted with CH2Cl2 (three times with 25 ml each). The combined organic layers were dried
over anhydrous Na2SO4, filtered and the solvent was removed under reduced pressure [18]. The
white solid obtained was dried and crystallized from ethanol.
Benzenesulfonohydrazide; 2a
White solid, yield (90%), m.p. 101-103 oC as reported [45].
4-toluenesulfonohydrazide; 2b
White solid, yield (92%), m.p. 109–110 oC as reported [46].
General procedure for the synthesis of sulfonyl hydrazone derivatives 4a-m.
A solution of the appropriate aldehyde 3a-j (0.5 mmol) in ethyl alcohol (5ml) was added to asolution
of the appropriate sulfonohydrazide 2a,b (0.5 mmol) in ethyl alcohol (5ml), and the reaction mixture
was warmed to 50-60oC and stirred for 30 min. The separated solid was then filtered, washed with
water, dried and crystallized from methylene chloride/ethanol. Some physicochemical and spectral
data for 4a-m are summarized in Table 6.
Table 6. Physicochemical and spectral data for the newly synthesized compounds 4a-m.
Molecular
Comp
4a
formula
(mw)a
M.P (°C) Yield (%) Elemental analysis %
C
H
N
S
C22H18N4O2S
402.47
110
75%
Calcd
Found
65.65 4.51 13.92 7.97
65.52 4.61 14.02 8.03

4b
4c
4d
C22H17BrN4O2S 188-190
481.37
93%
60%
88%
Calcd
Found
Calcd
Found
Calcd
Found
Calcd
Found
Calcd
Found
Calcd
Found
Calcd
Found
Calcd
Found
Calcd
Found
54.89 3.56 11.64 6.66
55
3.71 11.71 6.46
C22H17N5O4S
447.47
205-207
59.05 3.83 15.65 7.16
58.98 3.91 15.58 7.08
C23H20N4O2S
416.50
186-188
66.33 4.84 13.45
7.7
66.21 4.95 13.31 7.76
66.96 5.15 13.01 7.45
C24H22N4O2S
430.53
4e
4f
113-115
77%
72%
67.01 4.98
12.9
7.38
C22H17BrN4O2S 189-190
481.37
54.89 3.56 11.64 6.66
55.1
3.46 11.81 6.53
4g
4h
4i
C22H17ClN4O2S 179-180
436.91
84%
73%
80%
65%
60.48 3.92 12.82 7.34
60.31 3.82 12.97 7.19
C23H20N4O2S
416.50
173-175
66.33 4.84 13.45
7.7
66.39 4.68 13.54 7.64
55.76 3.87 11.31 6.47
55.68 3.97 11.23 6.61
59.86 4.15 15.18 6.95
C23H19BrN4O2S 199-200
495.4
4j
C23H19N5O4S
461.50
193-195
59.88 4.21 15.32
7

4k
4l
C24H22N4O2S
430.53
203-205
92%
82%
88%
Calcd
Found
Calcd
Found
Calcd
Found
66.96 5.15 13.01 7.45
67.22 5.19 12.88 7.28
55.76 3.87 11.31 6.47
C23H19BrN4O2S 224-225
495.40
55.9
3.74 11.48 6.31
4m
C23H19ClN4O2S 214-216
450.94
61.26 4.25 12.42 7.11
61.09 4.37 12.31 7.28
a Values are within ±0.4% of the calculated values.
N'-((1,3-Diphenyl-1H-pyrazol-4-yl)methylene)benzenesulfonohydrazide; 4a
IR (KBr, ν cm-1): 3188 (NH); 1598 (C=N); 1357, 1320 (SO2).
1H-NMR (400 MHz, DMSO-d6, δ ppm): 7.36 (t, J = 7.4 Hz, 1H, C-phenyl-C4-H); 7.45-7.65 (m,
10H, S-Phenyl-C3,4,5-H, C-phenyl-C2,3,5,6-H, N-phenyl-C3,4,5-H); 7.87 (d, J = 7.1 Hz, 2H, N-
Phenyl-C2,6-H); 7.96 (d, J = 7.9 Hz, 2H, S-phenyl-C2,6-H); 8.04 (s, 1H, Pyrazole-C5-H); 8.85(s, 1H,
CH=N ); 11.29 (s, 1H, NH, D2O-exchangeable).
N'-((1-(4-Bromophenyl)-3-phenyl-1H-pyrazol-4-yl)methylene)benzenesulfonohydrazide; 4b
IR (KBr, ν cm-1): 3148 (NH); 1591 (C=N); 1364, 1325 (SO2).
1H-NMR (400 MHz, DMSO-d6, δ ppm): 7.46-7.49 (m, 3H, C-phenyl-C3,4,5-H); 7.57-7.73 (m, 7H,
S-Phenyl-C3,4,5-H, C-phenyl-C2,6-H, bromophenyl-C2,6-H); 7.86 (d, J = 7.2 Hz, 2H, bromophenyl-
C3,5-H); 7.94 (d, J = 8.8 Hz, 2H, S-phenyl-C2,6-H); 8.02 (s, 1H, Pyrazole-C5-H); 8.89 (s, 1H, CH=N
); 11.31 (s, 1H, NH, D2O-exchangeable).

N'-((1-(4-Nitrophenyl)-3-phenyl-1H-pyrazol-4-yl)methylene)benzenesulfonohydrazide; 4c
IR (KBr, ν cm-1): 3183 (NH); 1596 (C=N); 1335, 1321 (SO2).
1H-NMR (400 MHz, DMSO-d6, δ ppm): 7.49-7.52 (m, 3H, C-phenyl-C3,4,5-H); 7.58-7.68 (m, 5H,
S-Phenyl-C3,4,5-H, C-phenyl-C2,6-H); 7.87 (d, J = 7.1 Hz, 2H, S-phenyl-C2,6-H); 8.04 (s, 1H,
Pyrazole-C5-H); 8.26 (d, J = 9.2 Hz, 2H, nitrophenyl-C2,6-H); 8.38 (d, J = 9.2 Hz, 2H, nitrophenyl-
C3,5-H); 9.09 (s, 1H, CH=N ); 11.41 (s, 1H, NH, D2O-exchangeable).
N'-((1-Phenyl-3-(4-tolyl)-1H-pyrazol-4-yl)methylene)benzenesulfonohydrazide; 4d
IR (KBr, ν cm-1): 3168 (NH); 1597 (C=N); 1363, 1317 (SO2).
1H-NMR (400 MHz, DMSO-d6, δ ppm): 2.38 (s, 3H, CH3); 7.28 (d, J = 7.9 Hz, 2H, tolyl-C3,5-H);
7.36 (t, J = 7.4 Hz, 1H, N-phenyl-C4-H); 7.46-7.68 (m, 7H, S-Phenyl-C3,4,5-H, tolyl-C2,6-H, N-
phenyl-C3,5-H); 7.87 (d, J = 7.1Hz, 2H, N-Phenyl-C2,6-H); 7.95 (d, J = 7.6 Hz, 2H, S-Phenyl-C2,6-
H); 8.02 (s, 1H, Pyrazole-C5-H ); 8.83(s, 1H, CH=N ); 11.27 (s, 1H, NH, D2O-exchangeable).
N'-((1,3-Di-4-tolyl-1H-pyrazol-4-yl)methylene)benzenesulfonohydrazide; 4e
IR (KBr, ν cm-1): 3357 (NH); 1610 (C=N); 1367, 1330 (SO2).
1H-NMR (400 MHz, DMSO-d6, δ ppm): 2.35, 2.37 (2s, 3H each, two CH3); 7.28 (d, J = 8 Hz,
2H, C-tolyl-C3,5-H); 7.32 (d, J = 8 Hz, 2H, N-tolyl-C3,5-H); 7.51 (d, J = 8 Hz, 2H, C-tolyl-C2,6-H);
7.58-7.68 (m, 3H, phenyl-C3,4,5-H); 7.83 (d, J = 8.4 Hz, 2H, N-tolyl-C2,6-H); 7.88 (d, J = 7.2 Hz,
2H, phenyl-C2,6-H); 8.02 (s, 1H, Pyrazole-C5-H); 8.77(s, 1H, CH=N); 11.26 (s, 1H, NH, D2O-
exchangeable).

N'-((3-(4-Bromophenyl)-1-phenyl-1H-pyrazol-4-yl)methylene)benzenesulfonohydrazide; 4f
IR (KBr, ν cm-1): 3183 (NH); 1598 (C=N); 1343, 1305 (SO2).
1H-NMR (400 MHz, DMSO-d6, δ ppm): 7.38 (t, J = 7.4 Hz, 1H, N-phenyl-C4-H); 7.51-7.67 (m,
9H, S-Phenyl-C3,4,5-H, bromophenyl-C2,6-H, N-phenyl-C2,3,5,6-H); 7.85 (d, J = 7.1 Hz, 2H,
bromophenyl-C3,5-H); 7.94 (d, J = 7.7 Hz, 2H, S-phenyl-C2,6-H); 8.02 (s, 1H, Pyrazole-C5-H);
8.87(s, 1H, CH=N ); 11.34 (s, 1H, NH, D2O-exchangeable).
EIMS m/z (% relative abundance): 482 (12) (M+∙+2), 480 (18) (M+∙), 370 (100)
N'-((3-(4-Chlorophenyl)-1-phenyl-1H-pyrazol-4-yl)methylene)benzenesulfonohydrazide; 4g
IR (KBr, ν cm-1): 3167 (NH); 1538 (C=N); 1363, 1316 (SO2).
1H-NMR (400 MHz, DMSO-d6, δ ppm): 7.37 (t, J = 7.4 Hz, 1H, N-phenyl-C4-H); 7.50-7.69 (m,
9H, S-Phenyl-C3,4,5-H, chlorophenyl-C2,6-H, N-phenyl-C2,3,5,6-H); 7.85 (d, J = 7.2 Hz, 2H,
chlorophenyl-C3,5-H); 7.94 (d, J = 7.7 Hz, 2H, S-Phenyl-C2,6-H); 8.02 (s, 1H, Pyrazole-C5-H); 8.87
(s, 1H, CH=N); 11.33 (s, 1H, NH, D2O-exchangeable).
N'-((1,3-Diphenyl-1H-pyrazol-4-yl)methylene)-4-toluenesulfonohydrazide; 4h
IR (KBr, ν cm-1): 3199 (NH); 1598 (C=N); 1375, 1349 (SO2).
1H-NMR (400 MHz, DMSO-d6, δ ppm): 2.36 (s, 3H, CH3); 7.35-7.40 (m, 3H, tolyl-C3,5-H, N-
phenyl-C4-H); 7.44-7.55 (m, 5H, C-phenyl-C3,5-H, N-phenyl-C3,4,5-H); 7.64 (dd, J = 1.3, 1.9, 7.4 Hz,
2H, C-Phenyl-C2,6-H); 7.75 (d, J = 8.2 Hz, 2H, N-Phenyl-C2,6-H); 7.97 (d, J = 7.92 Hz, 2H, tolyl-
C2,6-H); 8.03 (s, 1H, Pyrazole-C5-H); 8.85 (s, 1H, CH=N ); 11.20 (s, 1H, NH, D2O-exchangeable).