Investigation of synergistic antimicrobial effects of the drug combinations of meropenem and 1,2-benzisoselenazol-3(2H)-one derivatives on carbapenem-resistant Enterobacteriaceae producing NDM-1

Article abstract of DOI:10.1016/j.ejmech.2018.06.007

The worldwide prevalence of NDM-1-producing bacteria has drastically undermined the clinical efficacy of the last line antibiotic of carbapenems, prompting a need to devise effective strategy to preserve their clinical value. Our previous studies have shown that ebselen can restore the efficacy of meropenem against a laboratory strain that produces NDM-1. Here we report the construction of a focused compound library of 1,2-benzisoselenazol-3(2H)-one derivatives which comprise a total of forty-six candidate compounds. The structure-activity relationship of these compounds and their potential to serve as an adjuvant to enhance the antimicrobial efficacy of meropenem against a collection of clinical NDM-1-producing carbapenem-resistant Enterobacteriaceae isolates was examined. Drug combination assays indicated that these derivatives exhibited synergistic antimicrobial activity when used along with meropenem, effectively restoring the activity of carbapenems against the resistant strains tested in a Galleria mellonella larvae in vivo infection model. The mode of inhibition of one compound, namely 11_a38, which was depicted when tested on the purified NDM-1 enzyme, indicated that it could covalently bind to the enzyme and displaced one zinc ion from the active site. Overall, this study provides a novel 1,2-benzisoselenazol-3(2H)-one scaffold that exhibits strong synergistic antimicrobial activity with carbapenems, and low cytotoxicity. The prospect of application of such compounds as carbapenem adjuvants warrants further evaluation.

Full text of DOI:10.1016/j.ejmech.2018.06.007

Accepted Manuscript
Investigation of synergistic antimicrobial effects of the drug combinations of
meropenem and 1,2-benzisoselenazol-3(2H)-one derivatives on carbapenem-
resistant Enterobacteriaceae producing NDM-1
Wen Bin Jin, Chen Xu, Qipeng Cheng, Xiao Lin Qi, Wei Gao, Zhiwei Zheng, Edward
W.C. Chan, Yun-Chung Leung, Tak Hang Chan, Kwok-Yin Wong, Sheng Chen, Kin-
Fai Chan
PII:
S0223-5234(18)30497-5
10.1016/j.ejmech.2018.06.007
EJMECH 10477
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 7 February 2018
Revised Date: 11 April 2018
Accepted Date: 1 June 2018
Please cite this article as: W.B. Jin, C. Xu, Q. Cheng, X.L. Qi, W. Gao, Z. Zheng, E.W.C. Chan, Y.-C.
Leung, T.H. Chan, K.-Y. Wong, S. Chen, K.-F. Chan, Investigation of synergistic antimicrobial effects of
the drug combinations of meropenem and 1,2-benzisoselenazol-3(2H)-one derivatives on carbapenem-
resistant Enterobacteriaceae producing NDM-1, European Journal of Medicinal Chemistry (2018), doi:
10.1016/j.ejmech.2018.06.007.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. 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.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
1
2
3
4
Investigation of Synergistic Antimicrobial Effects of the Drug
Combinations of Meropenem and 1,2-Benzisoselenazol-3(2H)-
one Derivatives on Carbapenem-Resistant Enterobacteriaceae
Producing NDM-1
‡
,1
‡
,1
1
1
1
3
5
Wen Bin Jin, Chen Xu, Qipeng Cheng, Xiao Lin Qi, Wei Gao, Zhiwei Zheng, Edward W.
1
1
1,2
1
,1,3
6
7
C. Chan, Yun-Chung Leung, Tak Hang Chan, Kwok-Yin Wong, Sheng Chen,* and Kin-
,1
Fai Chan*
1
8
9
0
State Key Laboratory of Chirosciences and Department of Applied Biology and Chemical
Technology, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong SAR,
China
1
1
2
1
Department of Chemistry, McGill University, Montreal, Quebec, H3A 2K6, Canada
3
1
2
3
Shenzhen Key Laboratory for Food Biological Safety Control, Food Safety and Technology
1
Research Centre, The Hong Kong PolyU Shenzhen Research Institute, Shenzhen, China
‡
1
4
5
These authors contributed equally.
1
*Corresponding authors: Sheng Chen and Kin-Fai Chan
1
6
7
For S. C., Tel: +852 34008795, Fax: +852 23649932, email: sheng.chen@polyu.edu.hk
For K. -F. C., Tel: +852 34008684, Fax: +852 23649932, email: kf.chan@polyu.edu.hk
1
1

ACCEPTED MANUSCRIPT
1
ABSTRACT
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
The worldwide prevalence of NDM-1-producing bacteria has drastically undermined the clinical
efficacy of the last line antibiotic of carbapenems, prompting a need to devise effective strategy
to preserve their clinical value. Our previous studies have shown that ebselen can restore the
efficacy of meropenem against a laboratory strain that produces NDM-1. Here we report the
construction of a focused compound library of 1,2-benzisoselenazol-3(2H)-one derivatives which
comprise a total of forty-six candidate compounds. The structure-activity relationship of these
compounds and their potential to serve as an adjuvant to enhance the antimicrobial efficacy of
meropenem against a collection of clinical NDM-1-producing carbapenem-resistant
Enterobacteriaceae isolates was examined. Drug combination assays indicated that these
derivatives exhibited synergistic antimicrobial activity when used along with meropenem,
effectively restoring the activity of carbapenems against the resistant strains tested in a Galleria
mellonella larvae in vivo infection model. The mode of inhibition of one compound, namely
11_a38, which was depicted when tested on the purified NDM-1 enzyme, indicated that it could
covalently bind to the enzyme and displaced one zinc ion from the active site. Overall, this study
provides a novel 1,2-benzisoselenazol-3(2H)-one scaffold that exhibits strong synergistic
antimicrobial activity with carbapenems, and low cytotoxicity. The prospect of application of
such compounds as carbapenem adjuvants warrants further evaluation.
1
1
1
1
1
1
1
1
1
19
20
21
Keywords
Carbapenem-resistant Enterobacteriaceae, NDM-1 inhibitors, Ebselen derivatives, Carbapenem
adjuvants, Synergistic activity, Drug combination
2

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
Introduction
Carbapenem-resistant Enterobacteriaceae (CRE) comprise a family of Gram-negative bacteria
that exhibit antimicrobial resistance to multiple antibiotics, notably carbapenems which are the
drugs of last resort for treating lethal infections. CRE are the major causes of community- and
healthcare-associated bacterial infections, such as those in the urinary tract, bloodstream, surgical
site, and intra-abdominal region. Because of their phenotypic resistance to various antibiotics,
infections due to such “nightmare bacteria” are often difficult to treat, resulting in a significantly
increased mortality. According to a recent report published by the Centers for Disease Control
and Prevention (CDC), more than 9,000 hospitalized patients are infected by CRE annually and
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
half of these patients who developed CRE-associated bloodstream infections died eventually.
Moreover, about 600 death cases are due to infection by the two most common types of CRE,
carbapenem-resistant Klebsiella spp. and E. coli, annually. More importantly, recent surveillance
reports on antimicrobial resistance also provided evidence which suggests that there is increasing
2
-3
prevalence of CRE strains in the European countries and China. CRE infections have now
been regarded as a serious threat to public health. This situation has been be further deteriorated
with the recent emergence of MCR-1 positive, colistin-resistant CRE, which do not respond to
4
-6
all known antibiotics. A list of antibiotic-resistant "priority pathogens" released by the World
Health Organization (WHO) early this year indicates that CRE are among the most critical group
7
of bacteria which cause infections that require new antibiotics for treatment .
One of the principal resistance mechanisms of carbapenems in CRE involves the production of
carbapenemases, which are carbapenem-hydrolyzing β-lactamases that possess versatile
8
hydrolytic capacities to break down nearly all β-lactam antibiotics and render them ineffective.
3

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
These broad-spectrum carbapenemases belong to members of the class A, B, and D β-
lactamases, among which the most efficient carbapenemases are class B metallo-β-lactamases
9
(MBLs). MBLs employ a central zinc ion as essential cofactor to catalyze hydrolysis of the β-
lactam ring. The hydrolytic process of β-lactam antibiotics does not involve any covalent
enzyme-β-lactam intermediate but rather a nucleophilic attack on the water/hydroxyl ion which
is stabilized by the divalent zinc ion in the active site. MBLs can be further divided into three
subclasses (B1, B2 and B3) according to their amino acid sequence. As illustrated in Figure 1,
for example, New Delhi metallo-β-lactamase-1 (NDM-1) of subclass B1 MBLs contains a
dinuclear zinc centre and a water molecule in the active site, in which both zinc ions are
coordinated with the 3H site (His116, His118, His196) and the DCH site (Asp120, Cys221,
1
1
1
1
1
1
1
1
1
1
2
2
2
2
1
0
His263), and the water molecule is located in between both zinc ions. Clinically, subclass B1
MBLs are the most significant and prevalent carbapenemases, in particular, NDM-1. The rapid
worldwide dissemination of NDM-1-producing “superbugs” further emphasizes the significant
role of this type of carbapenemases in conferring antimicrobial resistance, and a pressing need
1
1
for development of effective NDM-1 inhibitors for clinical applications.
1
2
Despite the discovery of a NDM-1-producing K. pneumoniae isolate in year 2009 , huge
success has been made in combating organisms that produce β-lactamases through clinical usage
13
of β-lactam antibiotic/β-lactamase inhibitor combinations. In particular, development of small
1
4
molecules targeting NDM-1 have been actively pursued in the past decade. Numerous reports
1
5
16
on construction of NDM-1 inhibitors such as cyclic boronate 1, dicarboxylate heterocycle 2,
1
0, 17-18
19-20
21
thiol-containing carboxylic acids 3, 5-6,
metal chelator 4,
natural product 7 and
2
2
23
covalent inhibitors like cefaclor and chromone-3-carboxaldehyde, have appeared in the
4

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
literature (Figure 1). Despite the ongoing effort, clinically useful inhibitors of MBLs, in
particular NDM-1, are still not available. Development of clinically useful NDM-1 inhibitors has
been deemed technically challenging, with compound toxicity being a great concern.
We have previously identified a promising small molecule, covalent NDM-1 inhibitor, namely
ebselen (Eb) (Figure 1), which possesses an active scaffold of 1,2-benzisoselenazol-3(2H)-one,
through a cell-based screening assay against a laboratory strain of E. coli BL21 (NDM-1) that
harbors an isopropyl-β-D-thiogalactoside (IPTG) inducible full-length plasmid pET28b-blaNDM-
2
4
1
. Eb itself does not exhibit antibacterial activity towards the test strain. However, it exhibits
1
1
1
1
1
1
1
1
1
1
2
2
2
2
potent synergistic activity when used in combination with meropenem (Mem) and re-sensitizes
the resistant strain to a level equivalent that of the parental, non-NDM-1-producing E. coli BL21
strain, suggesting that it has great potential to be further developed into a new class of
carbapenem adjuvant. Despite its potent synergistic activity against the laboratory strain of E.
coli BL21 (NDM-1), further investigation of Eb revealed that its potency of inhibition towards
clinically isolated NDM-1 producing CRE strains was not as potent as expected. As shown in
Table 2, the minimum inhibition concentrations (MIC) screening of Mem in combination with
Eb against a panel of twenty-three clinical NDM-1 producing CRE isolates evidenced that the
MIC values of Mem only ranged from 16 µg/mL to 128 µg/mL, suggesting that the inhibitory
potency was only mild to moderate. The synergistic antimicrobial activity of Eb when used in
combination with meropenem was exhibited only in six clinical CRE isolates, with fractional
inhibitory concentration (FIC) indices being less than 0.5 (Table 2, Entry 4-6, 9-10 and 13).
Moreover, attempts to dissolve Eb in various aqueous formulations for in vivo efficacy assay was
found to be impractical. We reasoned that this might be due to the inferior ability of E. coli BL21
5

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
(NDM-1) in mimicking clinical CRE strain physiologically, and the intrinsically sub-optimal
drug-like properties of Eb.
In the present study, using Eb as a molecular template, we have constructed a focused compound
library with forty-four candidates of 1,2-benzisoselenazol-3(2H)-one derivatives and two
structurally related compounds, with the aims of exploring the structure-activity relationship
(SAR) of these potential carbapenem adjuvants and improving their physicochemical properties.
Another E. coli strain, Tg1 (NDM-1), is also generated to replace E. coli BL21 (NDM-1) for
preliminary cell-based screening assay. By evaluating the FIC indices of these derivatives when
tested in combination with Mem, the identified promising hits were selected for MIC screening
against clinical NDM-1 producing CRE isolates and for standard enzymatic assays to confirm
their inhibitory effect on the purified NDM-1 enzyme. The most effective derivative was further
tested to assess their in vivo efficacy and cytotoxicity on eukaryotic cells. By using electrospray
ionization mass spectrometry (ESI-MS) analysis, its mechanism of NDM-1 inhibition was also
investigated in details.
1
1
1
1
1
1
6

ACCEPTED MANUSCRIPT
1
2
Figure 1. Crystal structure of NDM-1 (PDB ID: 3Q6X) showing the active site of dinuclear zinc
3
4
5
centre and water (W), along with the numbered functionally important amino acids residues; the
I
chemical structures of various NDM-1 inhibitors and their reported values of IC50 or K were
highlighted in red; Eb was labelled with the A, B and C rings.
6
7

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
Results and Discussion
1. Chemical synthesis
The synthetic routes of the target compound library of 1,2-benzisoselenazol-3(2H)-one
2
5
derivatives 11 are straightforward and concise, and depicted in Scheme 1. Starting from
commercially available anthranilic acid (8), diazotization with sodium nitrite in hydrochloric
acid followed by treatment with a basified solution of disodium diselenide (Na Se ) under inert
2
2
atmosphere afforded the 2,2'-diselanediyldibenzoic acid (9a) in good yield. It is worthy to note
that the water used to prepare Na Se solution should first be degassed to remove oxygen.
2
2
Further conversion of acid 9a to 2-(chlorocarbonyl)phenyl hypochloroselenoite (10) via
treatment with thionyl chloride at refluxing temperature in the presence of a catalytic amount of
dimethylformamide (DMF) was achieved in high yield. It is worthy to mention that the acid
chloride 10 is unstable and should be kept under the nitrogen atmosphere in a freezer for
prolonged storage. The synthesis of the final 1,2-benzisoselenazol-3(2H)-one derivatives 11 were
successfully accomplished via treatment of acid chloride 10 with various aryl or alkyl substituted
primary amines a1-a14, a17-27, a29-a36, a38-a39, amide a15, and sulfonamide a16 (Chart 1)
respectively in the presence of triethylamine with dichloromethane (DCM) as solvent, producing
moderate to good yield. Similarly, treatment of acid chloride 10 with D-Ala-D-Ala (a28) or
glycine (a37) in acetonitrile at room temperature furnished 11 with a carboxyl acid moiety.
Moreover, three derivatives of 11 were selected for further functionalization, such as alkylation
of 11_a36 with 1-butyl iodide in DMF to furnish 11 with an imidazolium moiety, and acidic
deprotection of Boc group of 11_a38 and 11_a39 to yield 11 with an amine group. These amine
derivatives were in turn converted to the desired carbamates derivatives 11_a43-11_a45 via
standard carbamoylation procedure with 3-methyl-1-(((1,1,1-trifluoro-2-methylpropan-2-
1
1
1
1
1
1
1
1
1
1
2
2
2
2
8

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
9
0
1
2
yl)oxy)carbonyl)-1H-imidazol-3-ium iodide or benzyl chloroformate. Furthermore, two
structurally related compounds benzo[f][1,2,4]-selenadiazepine-3,5(2H,4H)-dione (12) and 2-
phenyl-1,2-selenazolidin-3-one (13_a1) were also synthesized for testing. Treatment of acid
chloride 10 with urea in DCM afforded the compound 12 with a characteristic seven-members
2
6
1,2,4-selenadiazepane ring in moderate yield. In the case of compound 13-a1, starting from
commercially available 3,3'-diselanediyldipropanoic acid (9b), N-ethyl-N’-(3-
dimethylaminopropyl)carbodiimide (EDCI) catalyzed condensation of acid 9b with aniline (a1),
followed by tert-butyl hydrogen peroxide assisted cyclization, smoothly generated the desired
1,2-selenazolin-3-one 13_a1 in low yield.
1
1
1
Scheme 1. Chemical synthesis of 1,2-benzisoselenazol-3(2H)-one derivatives and structurally
a
related compounds .
9

ACCEPTED MANUSCRIPT
1
a
2
3
4
5
6
7
8
9
(a) (1) NaNO
SOCl , DMF (cat.), 0 °C to reflux, 3 h; (c) For compound 11_a2-11_a27, 11_a29-11_a36,
11_a38-11_a39, amine a2-a27, a29-a36, a38-a39, NEt , DCM, rt, 14 h; For compound 11_a28
2 2 2 2
, conc. HCl (aq), 0 °C, 30 min; (2) Na Se , H O, 0 °C to 60 °C., 14 h; (b)
2
3
and 11_a37, D-Ala-D-Ala (a28) and glycine (a37), ACN, rt, 48 h; (d) urea, DCM, acetonitrile, 0
°C to rt, 24 h; (e) 1-butyl iodide, DMF, rt, 14 h; (f) conc. HCl, MeOH, DCM, 0 °C to rt, 14 h; (g)
3-methyl-1-(((1,1,1-trifluoro-2-methylpropan-2-yl)oxy)carbonyl)-1H-imidazol-3-ium
NEt , CHCl , 14 h; (h) benzyl chloroformate, NEt , DCM, 0 °C to rt, 14 h; (i) (1) amine a1,
iodide,
3
3
3
EDCI, DCM, 0 °C to rt, 14 h; (2) t-BuOOH, MeOH, rt, 24 h.
10
11
12
Chart 1. Chemical structures of aryl amines, benzyl amines, amide, sulfonamide, amino acids
used in this study.
10

ACCEPTED MANUSCRIPT
1
2
3
4
5
2. Antimicrobial activities against E. coli Tg1 (NDM-1)
To generate a Mem resistant strain with a clean background for preliminary screening of
compounds, a parental E. coli Tg1 strain was transformed into a resistant strain of E. coli Tg1
11

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(NDM-1) by introducing a full-length blaNDM-1 β-lactamase gene isolated from a clinical K.
pneumoniae isolate. E. coli Tg1 (NDM-1) is believed to better mimic clinical CRE strain than E.
coli BL21 (NDM-1). In line with the Clinical and Laboratory Standards Institute (CLSI)
2
7
approved guidelines, antimicrobial tests were performed accordingly and the reproducibility
was within one 2-fold dilution. The MIC of Mem against this screening strain was found to be 64
µg/mL (Table 1, entry 1), which was 128-fold higher than the parental E. coli Tg1 (MIC of Mem
= 0.5 µg/mL), demonstrating that this strain was capable of producing the NDM-1 enzyme. The
antimicrobial activities of all compounds were evaluated by measuring the MICs against this
screening strain, and the results are presented in Table 1. In general, most of these compounds
exhibited relatively weak or even no antibacterial activity, with MICs ≥64 µg/mL (Entry 3-42),
suggesting that these compounds were well-tolerated, relatively less toxic and exhibiting little
off-target side effects. However, six compounds (Entry 43-48) were found to exhibit mild
antibacterial activity, with MIC values ranging from 4 µg/mL to 32 µg/mL, indicating that these
compounds are likely to undergo non-specific interaction with protein targets other than NDM-1,
and are not worthy to pursue further. Structurally, these compounds possess polar functional
groups, such as alcohol, amine and carboxylic acid; an exception is compound 11_a9, which has
a benzyl group.
1
1
1
1
1
1
1
1
1
1
2
Table 1. MIC screening of compounds 11-13, Eb and their combination with Mem against E.
a
coli Tg1 (NDM-1), calculated cLogP, FIC indices and synergistic efficiency of compounds
21
Entry
Cpd
R
cLogP
MIC (µg/mL)
FIC
Synergistic
12

ACCEPTED MANUSCRIPT
Cpd Cpd + Mem
Index Efficiency
1
2
Mem
Eb
N.A.
N.A.
3.71
64
64
N.A.
16
N.A.
N.A.
0.500
0.043
3
11_a38
3.04
≥128
2
0.047
0.153
4
5
11_a2
4.42
3.65
≥128
≥128
4
8
0.094
0.188
0.139
0.088
11_a10
6
7
11_a15
11_a21
3.31
2.95
≥128
≥128
8
8
0.188
0.188
0.084
0.084
8
9
11_a26
11_a29
3.45
3.71
≥128
≥128
8
8
0.188
0.188
0.084
0.084
1
0
1
2
11_a31
11_a32
11_a34
2.66
2.74
0.77
≥128
≥128
≥128
8
8
8
0.188
0.188
0.188
0.084
0.084
0.084
1
1
1
1
3
4
11_a39
11_a25
3.33
3.32
≥128
≥128
8
8
0.188
0.188
0.080
0.076
N
N
n-Bu
1
1
5
6
11_a40
11_a43
-0.24 ≥128
8
8
0.188
0.188
0.076
0.073
4.68
≥128
1
7
11_a44
4.97
≥128
8
0.188
0.070
1
1
8
9
11_a23
11_a3
2.11
3.62
≥128
≥128
16
16
0.375
0.375
0.065
0.054
2
2
2
0
1
2
11_a6
2.67
3.16
2.01
≥128
≥128
≥128
16
16
16
0.375
0.375
0.375
0.054
0.054
0.054
11_a22
11_a36
2
2
3
4
11_a4
5.13
3.28
≥128
≥128
16
16
0.375
0.375
0.052
0.049
11_a16
13

ACCEPTED MANUSCRIPT
2
5
11_a27
2.89
≥128
16
0.375
0.049
2
2
6
7
11_a28
11_a20
1.71
2.82
≥128
≥128
16
16
0.375
0.375
0.049
0.047
2
2
8
9
11_a11
11_a12
4.57
5.62
≥128
≥128
16
16
0.375
0.375
0.043
0.043
Ph
3
3
3
0
1
2
11_a45
13_a1
3.52
2.52
2.21
≥128
≥128
≥128
16
32
32
0.375
0.750
0.750
0.043
0.024
0.018
N.A.
11_a13
3
3
3
3
4
5
11_a8
11_a30
11_a7
4.25
2.97
3.45
≥128
≥128
64
32
32
4
0.750
0.750
0.125
0.015
0.013
0.109
3
3
3
6
7
8
11_a33
12
2.06
0.84
3.99
64
64
64
8
8
8
0.250
0.250
0.250
0.107
0.107
0.087
N.A.
11_a14
3
4
9
0
11_a24
11_a17
3.17
2.60
64
64
16
16
0.500
0.500
0.041
0.039
4
4
1
2
11_a19
11_a18
3.12
3.12
64
64
16
16
0.500
0.500
0.039
0.039
4
4
3
4
11_a9
11_a5
3.73
3.45
32
32
8
8
0.375
0.375
0.058
0.052
4
4
4
4
5
6
7
8
11_a35
11_a37
11_a42
11_a41
1.16
1.39
1.55
1.24
32
8
16
2
0.750
0.281
0.281
1.063
0.022
0.091
0.091
-0.005
8
2
4
4
a
1
2
3
4
5
Compound cLogP were calculated using ChemDraw Ultra (Version 12.0). The synergistic
effect is determined by the FIC index, which is calculated as FIC (cpd) + FIC (Mem), where FIC
(cpd) is the (MIC of cpd in combination with Mem)/(MIC of cpd alone) while FIC (Mem) is
(MIC of cpd in combination with Mem)/(MIC of Mem alone). The drug combination is
considered synergistic if the FIC Index is ≤ 0.5. Synergistic efficiency is calculated by measuring
14

ACCEPTED MANUSCRIPT
1
2
3
–ln(FIC index)/non-hydrogen atom. N.A.: Not Applicable. All experiments were performed in at
least triplicates and inhibition of bacterial growth was assessed by naked eye upon incubation
overnight.
4
5
6
7
8
9
0
1
2
3
4
3. Combination studies with Mem against E. coli Tg1 (NDM-1) and clinical NDM-1 producing
CRE isolates
Next, to assess the synergistic activities of the test compounds when used in combination with
Mem, MICs of Mem were evaluated by mixing Mem with the compounds at a mass ratio of 1:1.
The FIC indices, which quantified the degree of interaction between antibiotic and adjuvant,
2
8
1
1
1
1
1
were calculated as previously described, with FIC indices ≤0.5 depicting synergistic
interaction. To better differentiate the compounds which have exactly the same FIC index values,
here we proposed, for the first time, a simple idea of synergistic efficiency, which is a similar
2
9
concept to the ligand efficiency as previously proposed. The synergistic efficiency is proposed
to be calculated by using the equation as shown below,
−
ln ꢀ( ꢁꢂꢃꢀꢄꢅꢆꢇꢈꢀꢉꢊꢀꢋꢀꢌꢉꢍꢇꢎꢏꢍꢇ)
Synergisticꢀefficiency =
ꢅꢏꢌꢐꢇꢑꢀꢉꢊꢀℎꢇꢋꢒꢓꢀꢋꢔꢉꢌꢀꢄꢅꢀꢔℎꢇꢀꢌꢉꢍꢇꢎꢏꢍꢇ
15
16
17
18
19
20
21
22
23
, where the heavy atom means non-hydrogen atom in a molecule. In fact, synergistic efficiency is
a normalized FIC index to enable comparisons of synergistic activities between different
molecular scaffolds but with the same FIC index. The larger the value of the synergistic
efficiency, the stronger the synergistic interaction will be. It provides a useful parameter for
medicinal chemists to choose a lead compound for further optimization. The results of Mem
MICs when tested in combination with the compounds, cLogP, calculated FIC indices and
synergistic efficiencies of all compounds are summarized in Table 1, in which compounds were
prioritized according to the FIC indices and synergistic efficiency respectively. The parental
compound Eb was used as a positive control for comparison purpose. The MIC of Mem in
15

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
combination with Eb was 16 µg/mL with a calculated FIC index of 0.50 and synergistic
efficiency of 0.043 (Table 1, entry 2), exhibiting only moderate synergistic activity. Generally,
most of the compounds were found to exhibit stronger synergistic activity than Eb, with FIC
indices ranging from 0.047 to 0.375. These compounds were divided into two series to
investigate their SARs: (1) 1,2-benzisoselenazol-3(2H)-one derivatives with a 2-aryl or benzyl
side chain and (2) 1,2-benzisoselenazol-3(2H)-one derivatives with a flexible 2-alkyl side chain.
For series 1, installation of various functional groups on the phenyl ring C of Eb, such as 4-Cl, 4-
i
MeO, 4-CH
2
OH, 3-CO
2
H, 4-CO
2
H, and 4- Pr improved the synergistic activities dramatically
1
1
1
1
1
1
1
1
1
1
2
2
2
with FIC indices ranging from 0.094 to 0.375 (Entry 4, 19, 20, 23, 35, 44). Furthermore,
replacing the phenyl ring C of Eb with different benzyl groups also helped improve the
synergistic activities (Entry 5, 28, 29, 43). It is worthy to mention that replacing the freely
rotatable benzylic carbon with a carbonyl or sulfonyl group of high rigidity maintained the
synergistic activities (Entry 6, 24). Compound 12 with an unusual seven-members 1,2,4-
selenadiazepane ring also exhibited synergy (Entry 37). On the other hand, removal of the phenyl
ring B or replacing the phenyl ring C of Eb with pyridine ring or 3-Cl-4-MeO di-substituted
phenyl ring resulted in no synergistic activities (Entry 31-33). Among these derivatives, it should
be mentioned that compound 11_a2 with a 4-Cl phenyl group demonstrated the most promising
synergy, with a FIC index of 0.094 (Entry 4). However, its relatively high cLogP value of 4.42
revealed that it has lower hydrophilicity and poorer aqueous solubility than Eb, which may
impede in vivo efficacy study.
16

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
For series 2, replacing the 2-phenyl ring of Eb with various flexible alkyl side chains generated
many compounds of promising synergistic activities, with FIC indices ranging from 0.047 to
0.500 (Entry 3, 7-18, 21-22, 25-27, 30, 36, 38-42). Among these compounds, compound 11_a38
exhibited the most promising synergistic activity, with a FIC index of 0.047 (Entry 3), which is
about 10-fold that of Eb. At a concentration of 2 µg/mL, it could re-sensitize Mem resistant
screening strain back to the susceptible level. The bulky Boc group of 11_a38 was essential in
conferring synergistic activity as its replacement with other functional groups, such as 3,3-
dimethylbutanone (Entry 7), benzyl group (Entry 8), 2-thiophenylmethone (Entry 10),
dimethylphosphonate (Entry 12), trifluoro-Boc (Entry 16), benzyl carbamate (Entry 30) and
substituted phenylmethanones (Entry 14, 27, 34), resulted in weaker synergistic activities.
Installation of heterocyclic rings at the terminal position, like 4-tetrahydropyran (Entry 11), n-
butyl imidazolium (Entry 15), 1-piperidine (Entry 21), 1-imidazole (Entry 12) and 4-morpholine
(Entry 25, 40), also caused weaker synergy. The CH CH group of 11_a38 seems to be crucial
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
for strong synergistic activity as increasing its length to CH CH CH resulted in weaker synergy
2
2
2
(Entry 13).
On the basis of the favorable FIC index of compound 11_a38 and MIC of Mem, we tested
whether the observed synergy in the screening strain could also be reproduced in our in-house
collection of clinically isolated NDM-1-producing strains, including four E. coli, two K. oxytoca,
four C. freundii, nine E. cloacae, two K. pneumoniae and two M. morganii strains. These CRE
strains were all NDM-1 positive and highly Mem resistant, with half of the strains exhibiting
Mem MICs ≥128 µg/mL (Table 2). Apart from producing NDM-1, several of these strains also
produced other β-lactamases such as VIM-1, IMP-4, KPC-2, CTX-M-3, CTX-M-14, CTX-M-15,
17

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
TEM-1 and SHV-12. As illustrated in Table 2, compound 11_a38 demonstrated no antibacterial
activity itself (MICs ≥128 µg/mL) but strong synergy with Mem across this panel of clinical
isolates, except for KP04 and EL18, with FIC indices ranging from 0.09 to 0.25 (Table 2, entry
1, 17). Compared with the parental compound Eb, compound 11_a38 exhibited much stronger
synergy as well as the optimal cLogP value of 3.04 for the in vivo efficacy assay.
Table 2. MIC screening of compound 11_a38 and Eb in combination with Mem against
a
clinically isolated, NDM-1-producing CRE strains
9
Additional
β-lactamase
determinants
b5, b1, b7, b8
MIC (µg/mL)
11_a38
MIC (µg/mL)
CRE
Strains
FIC
index
FIC
index
Entry
Eb +
Mem
Mem 11_a38
Eb
+ Mem
1
2
3
4
5
6
7
8
9
KP04
KP14
EC06
EC33
EC34
EC45
KO03
KO16
CF05
CF17
CF35
CF43
EL07
EL08
EL09
EL10
EL18
EL19
≥128
≥128
≥128
64
≥128
≥128
≥128
≥128
≥128
≥128
≥128
≥128
≥128
≥128
≥128
≥128
≥128
≥128
≥128
≥128
≥128
≥128
32
0.50 ≥128
64
1.00
2.00
1.00
0.38
0.38
0.25
0.50
0.50
0.25
0.38
0.50
0.50
0.38
1.25
2.00
1.50
2.50
2.00
b5, b3, b8
b4, b5, b8
b6
8
8
0.13 ≥128 ≥128
0.13 ≥128
0.19 ≥128
0.09 ≥128
0.06 ≥128
0.13 ≥128
0.13 ≥128
0.13 ≥128
0.09 ≥128
0.13 ≥128
0.25 ≥128
0.19 ≥128
0.16 ≥128
64
16
16
16
32
32
16
16
32
32
16
32
8
b6, b7, b8
none
64
4
≥128
≥128
≥128
≥128
64
4
b8
8
b8
8
b8
8
1
1
1
1
1
1
1
1
1
0
1
2
3
4
5
6
7
8
b8
4
b4, b3, b8
none
≥128
≥128
64
8
16
8
b5, b7, b8
b5, b7, b8
b4, b5, b7, b8
b5, b7, b8
b2, b7
32
4
≥128
64
8
0.13 ≥128 ≥128
8
0.19 ≥128
0.63 ≥128
64
64
32
16
16
b5, b3
≥128
0.25 ≥128 ≥128
18

ACCEPTED MANUSCRIPT
1
2
2
2
2
9
0
1
2
3
EL22
EL27
b5, b7, b8
b4, b5, b7, b8
b7, b8
≥128
32
≥128
≥128
≥128
≥128
≥128
16
4
0.25 ≥128 ≥128
2.00
2.50
2.50
1.50
1.50
0.16 ≥128
0.16 ≥128
0.19 ≥128
0.05 ≥128
64
64
64
64
EL28
32
4
MM23
MM26
none
64
8
none
64
2
a
1
2
3
4
5
6
7
8
9
EC, Escherichia coli; KO, Klebsiella oxytoca; CF, Citrobacter freundii; EL, Enterobacter
cloacae; KP, Klebsiella pneumoniae; MM, Morganella morganii. Additional β-lactamase
determinants: b1: VIM-1, b2: IMP-4, b3: KPC-2, b4: CTX-M-3, b5: CTX-M-14, b6: CTX-M-15,
b7: TEM-1, and b8: SHV-12. The synergistic effect is depicted by the FIC index, which is
calculated as FIC (cpd) + FIC (Mem), where FIC (cpd) is the (MIC of cpd in combination with
Mem)/(MIC of cpd alone) while FIC (Mem) is (MIC of cpd in combination with Mem)/(MIC of
Mem alone). The drug combination is considered synergistic if the FIC Index ≤0.5. All
experiments were performed in at least triplicates; the degree of inhibition of bacterial growth
was determined with the naked eye after incubation.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
4. In vivo efficacy study of compound 11_a38 in combination with Mem
To shed light on the potential clinical benefits of compound 11_a38, evaluation of in vivo
toxicity and antimicrobial efficacy was performed by employing a Galleria mellonella infection
model. First, the tolerable dose of compound 11_a38 was investigated by injecting a solution of
compound 11_a38 formulated in 50% PEG solution at different dosages (0, 10 or 30 mg/kg) into
the hemocoel of larvae and the percent of survival were monitored at 12 h interval for 48 h. As
shown in Figure S51, 100% survival rates were observed after 48 h for all treatment groups,
indicating the relative safety of both the vehicle and the compound at the dosages tested. Next,
the therapeutic abilities of compound 11_a38 alone, Mem alone or their combinations to protect
the larvae against a lethal dose infection of clinical NDM-1-producing CRE isolate EL10 were
determined at single doses of 0 or 10 mg/kg in drug monotherapy or 2 + 2 mg/kg, 10 + 10 mg/kg
in drug combinations (Figure 2). 100% mortality of larvae were observed for the treatment
groups of the vehicle, 10 mg/kg monotherapy of compound 11_a38, and combination therapy at
2 + 2 mg/kg doses after 12 h. Encouragingly, 10 mg/kg monotherapy of Mem and combination
19

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
therapy at 10 + 10 mg/kg doses after 48 h resulted in 20% and 60% survival rate respectively,
demonstrating the excellent ability of compound 11_a38 to potentiate Mem antibacterial activity
against clinical NDM-1-producing CRE strains. Compared to the treatment groups of
monotherapy of compound 11_a38 and Mem at 10 mg/kg, it was found to be highly significant
(p < 0.05). These results suggest that the therapeutic efficacy of the combination of Mem and
compound 11_a38 should be further evaluated in a mouse infection model in future.
7
8
Figure 2. Kaplan-Meier survival analysis of monotherapy and combination therapy in protecting
Galleria mellonella larvae infected with a lethal dose of the clinical NDM-1-producing CRE
9
0
5
1
isolate EL10 (2.5 x 10 CFU/larva). Data are the means of three independent experiments.
11
12
13
14
15
16
5. Biochemical studies of compound 11_a38 with purified NDM-1 protein
To confirm whether the observed synergy is due to the inhibition of the NDM-1 activity of
compound 11_38, standard enzymatic assays using purified NDM-1 protein and colorimetric β-
2
4
lactamase substrate nitrocefin were performed as previously described. Compound 11_a38
demonstrated a potent in vitro dose-dependent inhibition of the NDM-1 enzyme, with a
calculated IC50 value of about 13 µM (Figure S49). Previous studies of Eb indicated that it
20

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
inhibited the activity of NDM-1 in a time-dependent and concentration-dependent manner. Based
on the similarity of the core structure of Eb and compound 11_38, we reasoned that both
compounds might exhibit similar NDM-1 inhibition mechanism. To test this hypothesis, the
rapid high dilution method was performed to test the reversibility of inhibition of compound
11_a38. The results revealed that approximately 40% residual NDM-1 activity was observed
after the rapid high dilution of a 20-minutes pre-incubation mixture when compared to positive
control with no addition of compound 11_a38 (Figure S50), implying that the inhibitory effect
of compound 11_a38 most likely involves slowly reversible binding to the NDM-1 protein.
Compared with Eb, which has 11% residual activity after the rapid high dilution, compound
11_a38 exhibited weaker binding to NDM-1. As illustrated in Figure 3A, inhibition of NDM-1
by compound 11_a38 was also time-dependent and concentration-dependent. Moreover, plots of
natural logarithm of percentage residual activity versus incubation time at different
concentrations of compound 11_a38 indicated that inhibition of DNM-1 follows pseudo-first
1
1
1
1
1
1
-1
order kinetics with calculated kinetic parameters K of 17 ± 2 µM, k
of 0.068 ± 0.007 min
I
inact
-1 -1
I
and kinact/K of 0.067 mM s respectively (Figure 3B).
16
21

ACCEPTED MANUSCRIPT
1
2
3
Figure 3. (A) Time-dependent and concentration-dependent inhibition of NDM-1 enzyme by
compound 11_a38; (B) Hyperbolic plot of Kobs of compound 11_a38 versus concentration of
compound 11_a38.
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
To better illustrate the nature of potential interaction between NDM-1 protein and compound
11_a38, we examined the wild-type and denatured NDM-1 protein treated with compound
11_a38 respectively, using nano-ESI-MS. First, we obtained the ESI mass spectra of wild-type
and denatured NDM-1 protein for comparison (Figure 4). The deconvoluted ESI mass spectra of
wild-type and denatured NDM-1 protein revealed an intact mass of 25,995 Da and 25,866 Da
respectively, with a difference of 129 Da, suggesting that there were two zinc ions (2 × 65 Da) in
the wild-type NDM-1 protein. In contrast, the deconvoluted ESI-MS spectra of compound
11_a38-treated wild-type and denatured NDM-1 protein revealed a complex where the majority
of protein had a mass of 26,272 Da and 26,208 Da respectively, with a difference of 64 Da,
indicating that there was only one zinc ion in the compound 11_a38-treated wild-type NDM-1
protein. Since the molecular mass of compound 11_a38 is about 341 and there is only one
cysteine present in the NDM-1 protein, the protein mass difference between the denatured NDM-
1 protein and compound 11_a38-treated denatured NDM-1 protein is 342 Da, indicating that
only one covalent interaction event occurred between compound 11_a38 and a cysteine
molecule. The major peak (26,272 Da) in the deconvoluted ESI-MS spectra of compound
11_a38-treated wild-type NDM-1 protein, therefore, represented the sum of the molecular mass
of a denatured NDM-1 protein (25,866 Da), compound 11_a38 (341 Da) and one zinc ion (65
Da). These results are consistent with those of the parental compound Eb.
1
1
1
1
1
1
1
1
1
1
2
2
22

ACCEPTED MANUSCRIPT
1
2
Figure 4. Nano-ESI-MS analysis and their cartoon representations of wild-type (A) and
denatured (C) NDM-1 enzyme as well as compound 11_a38-treated wild-type (B) and denatured
(D) NDM-1 enzyme.
3
4
5
6
7
8
9
0
1
2
3
4
Based on the above data, the mechanism of inhibition of compound 11_a38 against NDM-1
protein was proposed. Both cationic divalent zinc ions in the active site of NDM-1 may act as
Lewis acids to interact with the carbonyl oxygen of compound 11_a38, causing the 1,2-
benzisoselenazol-3(2H)-one being susceptible to nucleophilic attack by the nearby thiol group of
cysteine. The thiol group of cysteine reacts with compound 11_a38 to form a new S-Se covalent
bond, leading to the loss of one zinc ion and therefore inhibiting the NDM-1 activity.
1
1
1
1
1
6. Cytotoxicity studies of compound 11_a38 against eukaryotic cells
As mentioned before, overcoming compound toxicity has been a great concern for the
development of clinically useful NDM-1 inhibitors. Although a recent report of phase 2 clinical
23

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
9
trial of Eb in an unrelated study indicated that treatment of Eb at a dose of 400 mg twice daily in
3
7
human was safe, a potential problem usually associated with cysteine modifying agents is their
high toxicity towards eukaryotic cells. To verify the safety of compound 11_a38, its toxicity was
tested against human cervical cancer HeLa cell line and mouse peritoneal fibroblast L929 cell
line respectively. As shown in Figure 5, compound 11_a38 exhibited relatively low cytotoxicity
against Hela and L929 cell lines, with a cell viability of 70% or higher at a concentration of 2
mg/mL, which is the effective synergistic concentration in the combination study. Microscopic
investigation of both cell lines also indicated that there were no morphological changes after
incubation with compound 11_a38 at these concentrations, displaying negligible cytotoxicity.
1
0
1
1
Figure 5. Cytotoxicity profile of compound 11_a38 against HeLa and L929 cell lines.
12
13
14
15
16
Conclusion
In this study, we demonstrated that structural modification of Eb through simple chemistry
allows us to generate a focused compound library of 1,2-benzisoselenazol-3(2H)-one derivatives
with forty-six candidates. Many compounds displayed stronger synergistic activity with Mem
24

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
(FIC index ≤0.5) than Eb as well as better physiochemical properties (cLogP < that of Eb) but
negligible antibacterial activity (MIC ≥128 µg/mL) when tested alone. One of these compounds
was also demonstrated to exhibit potent synergistic activities with Mem against a panel of
clinical NDM-1-producing CRE isolates. These promising results led to the efficacy testing of
the lead compound 11_a38 in the Galleria mellonella infection model. Investigation of the mode
of action suggests that compound 11_a38 exerted synergistic antimicrobial effect with
meropenem by covalently binding to the NDM-1 protein. In summary, because of its structural
simplicity, potent synergistic activity in combination with Mem, and low cytotoxicity towards
eukaryotic cells, a new class of 1,2-benzisoselenazol-3(2H)-one derivatives may represent
excellent leads for the development of next-generation carbapenem adjuvant co-therapy.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
Experimental section
Chemical Synthesis
1
All NMR spectra were recorded on a Bruker Advance-III spectrometer at 400 MHz for H, 101
1
3
19
MHz for C, and 376 MHz for F. All NMR measurements were carried out at room
temperature. The chemical shifts are reported as parts per million (ppm) in unit relative to the
resonance of CDCl , DMSO-d , D O, Acetone-d , or MeOH-d . Mass spectra of low-resolution
3
6
2
6
4
and high-resolution mass spectrometry were obtained on a Micromass Q-TOF-2 by electron
spray ionization (ESI) mode. All chemical reagents and organic solvents were reagent grade and
were used without further purification unless otherwise stated. The plates used for analytical
thin-layer chromatography (TLC) were E. Merck Silica Gel 60F254 aluminum-backed plates
(0.25 mm thickness) and were visualized under short and long UV light (254 and 365 nm) or
stained with acidified potassium permanganate solution followed by gentle heating. Column
25

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
chromatographic purifications were carried out on MN silica gel 60 (230−400 mesh) with
gradient elution. Compound purity was determined by an Agilent 1100 series HPLC installed
with a normal phase Prep-Sil Scalar column (4.6 mm × 250 mm, 5 µm) at UV detection of 254
nm (reference at 450 nm). All tested compounds were shown to have >95% purity according to
the HPLC. 3-Methyl-1-(((1,1,1-trifluoro-2-methylpropan-2-yl)oxy)carbonyl)-1H-imidazol-3-ium
3
0
iodide was synthesized as described previously. Amines a20, a21, a25, a30, a31 and a34 were
prepared in two steps by mixing the corresponding acid chlorides, such as benzoyl chloride, 3,3-
dimethylbutanoyl chloride,
4-methylbenzoyl chloride, 2,4-difluorobenzoyl chloride and
thiophene-2-carbonyl chloride, or dimethyl chlorophosphate with amine a38 respectively,
3
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
followed by acidic treatment of TFA as previously reported. Amines a1-a19, a22-a24, a26-
a29, a33, a35-a39, anthranilic acid (8) and 3,3'-diselanediyldipropanoic acid (9b) are
3
8
39
42
42
39
40
commercially available. Compound 11_a2, 11_a3, 11_a5, 11_a7, 11_a9, 11_a10,
4
0
41
41
42
26
11_a14, 11_a23, 11_a35, 11_a37, and 12 have been reported in the literature previously.
Other compounds reported here are new compounds.
2,2'-Diselanediyldibenzoic acid (9a)
Anthranilic acid (8) (18 g, 131 mmol) was dissolved in a solution of 37% hydrochloric acid (30
mL) and H O (30 mL). After the solution was cooled to 0 °C, a solution of NaNO (10 g, 145
2
2
mmol in 10 mL H O) was added dropwise and the reaction mixture was stirred at 0 °C for 30
2
mins until all solid has been dissolved. The obtained diazonium salt was used immediately
without further purification. To a well-stirred mixture of selenium powder (5.1 g, 65 mmol) in
degassed H
2
O (50 mL) under a nitrogen atmosphere, a NaBH
4
aqueous solution (4.45 g, 130
mmol in 10 mL H
2
O) was added dropwise. After the Se powder has been completely dissolved,
26

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
another batch of 5.1 g Se powder was added to the solution in portions. The reaction mixture was
heated for 30 mins to ensure all the Se powder has been dissolved completely. The obtained
brownish-red solution was basified with NaOH (10 g) in an ice bath. After cooling the basified
solution to 5 °C, the diazonium salt prepared above was added dropwise at a rate at which the
reaction temperature was maintained below 10 °C. Afterwards, the mixture was stirred under the
nitrogen atmosphere at 60 °C for 11 h and another 3 h at room temperature. The mixture was
acidified with hydrochloric acid to pH = 3 ~ 4 and then filtered. The obtained brownish-red
precipitates were re-dissolved in NaOH solution (pH = 9 ~ 10) and any undissolved black solid
was removed by filtration. The brownish-red filtrate was re-acidified with hydrochloric acid to
pH = 1 and the red precipitates were filtered and washed twice with water. The desired product
1
1
1
1
1
1
1
1
1
1
2
2
2
1
was obtained by recrystallization in acetic acid as a light yellow solid (15.6 g) in 60% yield. H
NMR (400MHz, DMSO-d ) δ 13.56 (br. s., 1 H), 8.04 (d, J = 6.4 Hz, 1 H), 7.68 (d, J = 7.6 Hz, 1
6
1
3
H), 7.49 (t, J = 7.6 Hz, 1 H), 7.36 (t, J = 7.0 Hz, 1 H); C NMR (101MHz, DMSO-d ) δ 169.0,
6
134.1, 134.0, 132.0, 130.0, 129.2, 127.0.
2-(Chlorocarbonyl)phenyl hypochloroselenoite (10)
To a solution of SOCl (10 mL) in a three-necked round bottom flask under a nitrogen
2
atmosphere at 0 °C, compound 9a (2.0 g, 8 mmol) in portions and serval drops of DMF were
added. The mixture was heated to reflux at 80 °C for 3 h. After heating, excess SOCl was
2
removed by blowing with nitrogen and 20 mL hexane were added to the flask to extract the
product for three times. After evaporating the hexane under reduced pressure, a bright yellow
1
product of compound 10 (2.0 g) was obtained in 80% yield. H NMR (400MHz, CDCl
3
) δ 8.38
27

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
(d, J = 7.0 Hz, 1 H), 8.13 (d, J = 8.3 Hz, 1 H), 7.82 (t, J = 7.6 Hz, 1 H), 7.52 (t, J = 7.6 Hz, 1 H);
1
3
C NMR (101MHz, CDCl ) δ 172.6, 146.3, 136.2, 134.6, 129.0, 127.1, 126.7.
3
2-(4-Chlorophenyl)benzo[d][1,2]selenazol-3(2H)-one (11_a2)
To a well-stirred solution of 4-chloroaniline (a2) (0.57 g, 4.5 mmol) and NEt (5 mL) in dry
3
DCM (25 mL) was added dropwise a solution of compound 10 (0.51 g, 2.0 mmol) at room
temperature. The reaction mixture was stirred at room temperature for 14 h. The crude mixture
was obtained by removal of the organic solvents under reduced pressure and was re-dissolved in
2
DCM, followed by washing with H O for 3 times. The organic layers were combined, dried over
1
1
1
1
1
1
1
1
1
1
2
2
2
anhydrous MgSO , filtered and concentrated in vacuum. Purification of the product was
4
performed by flash column chromatography on silica gel to afford the desired compound 11_a2
1
(0.28 g) in 45% yield. H NMR (400MHz, CDCl ) δ 8.04 (d, J = 7.8 Hz, 1 H), 7.59 (d, J = 3.9
3
1
3
Hz, 2 H), 7.51 (d, J = 1.0 Hz, 2 H), 7.41 (t, J = 3.9, 7.8 Hz, 1 H), 7.32 (d, J = 1.0 Hz, 2 H); C
NMR (101MHz, CDCl ) δ 185.4, 134.7, 134.0, 129.4, 128.1, 126.7, 126.6, 125.2, 123.8, 121.0;
3
+
HRMS m/z calcd for (C13
H
8
ClNOSe + H) 309.9532, found 309.9531.
2-(4-Methoxyphenyl)benzo[d][1,2]selenazol-3(2H)-one (11_a3)
Following the experimental procedure for the preparation of compound 11_a2 described above,
but with 4-methoxyaniline (a3) (0.55 g, 4.5 mmol) as a starting material, compound 11_a3 (0.32
1
g) was obtained in 52% yield. H NMR (400MHz, DMSO-d ) δ 8.08 (d, J = 7.8 Hz, 1 H), 7.89
6
(d, J = 7.8 Hz, 1 H), 7.67 (t, J = 1.0 Hz, 1 H), 7.53 - 7.45 (m, 3 H), 7.01 (d, J = 8.8 Hz, 2 H), 3.79
1
3
(s, 3 H); C NMR (101MHz, DMSO-d ) δ 165.4, 157.8, 139.5, 132.7, 132.5, 128.8, 128.3,
6
28