Boosting the efficacy of anti-MRSA β-lactam antibiotics via an easily accessible, non-cytotoxic and orally bioavailable FtsZ inhibitor

Article abstract of DOI:10.1016/j.ejmech.2018.11.052

The rapid emergence of methicillin-resistant Staphylococcus aureus (MRSA) strains has undermined the therapeutic efficacy of existing β-lactam antibiotics (BLAs), prompting an urgent need to discover novel BLAs adjuvants that can potentiate their anti-MRSA activities. In this study, cytotoxicity and antibacterial screening of a focused compound library enabled us to identify a compound, namely 28, which exhibited low cytotoxicity against normal cells and robust in vitro bactericidal synergy with different classes of BLAs against a panel of multidrug-resistant clinical MRSA isolates. A series of biochemical assays and microscopic studies have revealed that compound 28 is likely to interact with the S. aureus FtsZ protein at the T7-loop binding pocket and inhibit polymerization of FtsZ protein without interfering with its GTPase activity, resulting in extensive delocalization of Z-ring and morphological changes characterized by significant enlargement of the bacterial cell. Animal studies demonstrated that compound 28 had a favorable pharmacokinetic profile and exhibited potent synergistic efficacy with cefuroxime antibiotic in a murine systemic infection model of MRSA. Overall, compound 28 represents a promising lead of FtsZ inhibitor for further development of efficacious BLAs adjuvants to treat the staphylococcal infection.

Full text of DOI:10.1016/j.ejmech.2018.11.052

Accepted Manuscript
Boosting the efficacy of anti-MRSA β-lactam antibiotics via an easily accessible, non-
cytotoxic and orally bioavailable FtsZ inhibitor
Hok Kiu Lui, Wei Gao, Kwan Choi Cheung, Wen Bin Jin, Ning Sun, Jason W.Y. Kan,
Iris L.K. Wong, Jiachi Chiou, Dachuan Lin, Edward W.C. Chan, Yun-Chung Leung,
Tak Hang Chan, Sheng Chen, Kin-Fai Chan, Kwok-Yin Wong
PII:
S0223-5234(18)31011-0
DOI:
https://doi.org/10.1016/j.ejmech.2018.11.052
EJMECH 10910
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 17 October 2018
Revised Date: 5 November 2018
Accepted Date: 21 November 2018
Please cite this article as: H.K. Lui, W. Gao, K.C. Cheung, W.B. Jin, N. Sun, J.W.Y. Kan, I.L.K. Wong,
J. Chiou, D. Lin, E.W.C. Chan, Y.-C. Leung, T.H. Chan, S. Chen, K.-F. Chan, K.-Y. Wong, Boosting
the efficacy of anti-MRSA β-lactam antibiotics via an easily accessible, non-cytotoxic and orally
bioavailable FtsZ inhibitor, European Journal of Medicinal Chemistry (2018), doi: https://doi.org/10.1016/
j.ejmech.2018.11.052.
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
1
2
3
Boosting the efficacy of anti-MRSA β-lactam antibiotics
via an easily accessible, non-cytotoxic and orally
bioavailable FtsZ inhibitor
‡
,1
‡,1
‡,1
1
1
1
4
5
6
Hok Kiu Lui, Wei Gao, Kwan Choi Cheung, Wen Bin Jin, Ning Sun, Jason W. Y. Kan,
1
1
3
1
1
Iris L. K. Wong, Jiachi Chiou, Dachuan Lin, Edward W. C. Chan, Yun-Chung Leung, Tak
1
,2
1,3
,1
,1
Hang Chan, Sheng Chen, Kin-Fai Chan* and Kwok-Yin Wong*
1
7
8
9
State Key Laboratory of Chemical Biology and Drug Discovery and Department of Applied
Biology and Chemical Technology, The Hong Kong Polytechnic University, Hung Hom,
Kowloon, Hong Kong SAR, China
2
10
Department of Chemistry, McGill University, Montreal, Quebec, H3A 2K6, Canada
3
1
1
2
Shenzhen Key Laboratory for Food Biological Safety Control, Food Safety and Technology
Research Centre, The Hong Kong PolyU Shenzhen Research Institute, Shenzhen, China
1
‡
13
14
15
16
17
18
These authors contributed equally.
* Corresponding authors: Kin-Fai Chan and Kwok-Yin Wong
For K. F. C., Tel: +852 34008684, Fax: +852 23649932, email: kf.chan@polyu.edu.hk,
ORCID ID: 0000-0003-1475-288X.
For K. Y. W., Tel: +852 34008686, Fax: +852 23649932, email: kwok-yin.wong@polyu.edu.hk,
ORCID ID: 0000-0003-4984-7109.
1

ACCEPTED MANUSCRIPT
19
Abstract
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
The rapid emergence of methicillin-resistant Staphylococcus aureus (MRSA) strains has
undermined the therapeutic efficacy of existing β-lactam antibiotics (BLAs), prompting an
urgent need to discover novel BLAs adjuvants that can potentiate their anti-MRSA activities. In
this study, cytotoxicity and antibacterial screening of a focused compound library enabled us to
identify a compound, namely 28, which exhibited low cytotoxicity against normal cells and
robust in vitro bactericidal synergy with different classes of BLAs against a panel of multidrug-
resistant clinical MRSA isolates. A series of biochemical assays and microscopic studies have
revealed that compound 28 is likely to interact with the S. aureus FtsZ protein at the T7-loop
binding pocket and inhibit polymerization of FtsZ protein without interfering with its GTPase
activity, resulting in extensive delocalization of Z-ring and morphological changes characterized
by significant enlargement of the bacterial cell. Animal studies demonstrated that compound 28
had a favorable pharmacokinetic profile and exhibited potent synergistic efficacy with
cefuroxime antibiotic in a murine systemic infection model of MRSA. Overall, compound 28
represents a promising lead of FtsZ inhibitor for further development of efficacious BLAs
adjuvants to treat the staphylococcal infection.
35
36
37
38
39
40
2

ACCEPTED MANUSCRIPT
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
Introduction
β-Lactam antibiotics (BLAs), the life-saving drugs that have long been widely used to treat
lethal bacterial infection, are arguably one of the most important classes of therapeutic drugs in
the history of human medicine. Although there is currently a rich collection of BLAs available
for clinical use, the twin threats of global overuse of BLAs and rapid emergence of multidrug-
resistant pathogenic bacteria have led to dramatic erosion of the therapeutic efficacy of the entire
1
-3
classes of BLAs including penicillins, cephalosporins, and even carbapenems. Indeed, some
examples of these multidrug-resistant pathogenic bacteria include the community- and
healthcare-associated methicillin-resistant Staphylococcus aureus (MRSA), which cause an
4
alarming patient mortality of over 11,000 deaths in the United States annually. Consequently,
the scarcity of effective treatment options of BLAs has created an urgent need not only for the
development of next generation BLAs but also for the discovery of BLAs adjuvants that can
make recalcitrant multidrug-resistant MRSA more susceptible to existing BLAs. Augmenting
BLAs with a second agent has been proven clinically as one of the most effective strategies to
5
-6
restore the efficacy and extend the lifespan of this important class of antibiotics. The well-
known examples include the combination of FDA-approved β-lactamase inhibitors, such as
clavulanic acid, sulbactam, tazobactam, avibactam and vaborbactam, with BLAs, providing
highly effective treatment options in restoring the efficacy of BLAs against Gram-negative
7
bacteria that have acquired diverse β-lactamase enzymes. Clinical BLAs resistance in MRSA,
however, is primarily mediated by acquiring another penicillin-binding protein Pbp2a with
markedly reduced affinity for all classes of BLAs. Development of new BLAs combination
treatment paradigm to boost the clinical efficacy of these important drugs against MRSA would
undoubtedly strengthen current infectious disease management.
3

ACCEPTED MANUSCRIPT
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
The bacterial cell division machinery involves many essential proteins that are extremely
8
-9
sensitive to perturbation by small molecules.
Among those cell division proteins, the
filamenting temperature-sensitive mutant Z (FtsZ) protein has been extensively studied as a drug
1
0-11
target for the discovery of antibacterial agents.
Over the past decade, different classes of FtsZ
12
13
inhibitors have been discovered, such as natural product chrysophaentin A, berberine,
1
4
15
16
naphthol analogues, guanosine triphosphate (GTP) derivatives and benzimidazoles. During
the process of cell division, monomeric FtsZ proteins undergo self-activating GTP-dependent
polymerization to produce FtsZ filaments and contractile Z-ring at the mid-cell, followed by the
17
constriction and depolymerization to give rise to two identical daughter cells.
1
8
Pioneering studies by Tan have nicely demonstrated that a FtsZ-specific inhibitor PC190723
1
9-20
(1)
acts synergistically with imipenem both in vitro and in a murine model of MRSA
infection (Figure 1). The underlying mechanism of synergy was unclear. However, it was
proposed to be driven by the initial delocalization of FtsZ filaments after the treatment of 1,
resulting in subsequent delocalization of the penicillin-binding proteins, which are important
1
8, 21
bacterial enzymes that are involved in peptidoglycan biosynthesis of bacterial cell wall.
Combined with the recent findings that the treadmilling of FtsZ filaments controls both the
2
2-23
location and activity of the septal peptidoglycan synthesizing enzymes,
these findings thus
provide a rational basis for exploring much wider chemical space of FtsZ inhibitors that can be
developed as efficacious anti-MRSA BLAs adjuvants. Despite the highly hydrophobic nature
2
4
and suboptimal drug-like properties of 1, its structurally similar derivatives TXA707 (2) ,
2
5-26
27-28
29
TXA6101 (3)
, and N-Mannich type prodrugs, such as TXY541 (4),
TXY436 (5) and
4

ACCEPTED MANUSCRIPT
2
4, 30-31
32
87
88
89
90
91
92
93
94
TXA709 (6),
as well as succinate prodrug 7 with enhanced in vitro and in vivo activities,
have been further pursued as anti-staphylococcal agents for clinical evaluation (Figure 1).
Although such prodrug approach has partially improved the aqueous solubility and
pharmacokinetic (PK) properties of the parental drugs 1 - 3, the intrinsic chemical instabilities
and multistep chemical synthesis of prodrugs 4 - 7 may remain a major obstacle for fully
unleashing their potential to be used in clinical practice. Therefore, alternative strategies remain
to be devised to exploit FtsZ inhibitors as anti-MRSA BLAs adjuvants.
9
5
6
9
Figure 1. Chemical structures of FtsZ inhibitors. Parental drugs 1 - 3 and their prodrugs 4 - 7 are
97
indicated in red and black respectively.
5

ACCEPTED MANUSCRIPT
9
8
9
9
We have previously reported the identification of a series of novel FtsZ inhibitors through the
3
3-36
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
computer-aided structure-based virtual screening
and the cell-based screening of natural
3
7-39
product library.
Several quinuclidine-based FtsZ inhibitors were also found to exhibit strong
synergistic effect against MRSA strains when combined with BLAs, suggesting that FtsZ protein
maybe a desirable “potentiation drug target” of BLAs to boost their anti-staphylococcal
4
0
activity. In the present study, by use of compounds 1 and 8 as a starting template, we sought to
systematically design, synthesize and screen a focused compound library of 47 candidates and
identified a new class of FtsZ inhibitors which exhibited easy accessibility, low cytotoxicity and
safe, favorable PK profile, and most importantly, potent in vitro and in vivo synergistic
antimicrobial activity when used in combination with existing BLAs against MRSA.
Results and Discussion
1. Compound design and chemical synthesis
As shown in Figure 1, PC190723 (1) was constructed from a 2,6-difluorobenzamide and a 6-
chlorosubstituted thiazolopyridine moiety joined by an ether linkage at the C-3 position of the
phenyl ring. Compound 8 possesses the same 2,6-difluorobenzamide warhead but with a
different n-nonyloxy tail. For the sake of comparison, both compounds were also synthesized as
41-42
positive controls according to the previous reports.
Resulted from the inspiration of their
4
3-48
chemical structures as well as other related studies of FtsZ inhibitors,
our molecular design
strategies are: (1) to replace the C-3 ether linkage with other functional groups such as secondary
or tertiary amine, amide and triazole because these groups usually offer more favorable
physicochemical properties than ether; (2) to replace the thiazolopyridine moiety with other
6

ACCEPTED MANUSCRIPT
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
commercially available building blocks of low molecular weight for easy accessibility and rapid
chemical synthesis; (3) to vary the position or reduce the number of fluorine group on the phenyl
ring for investigating the influence of fluorine atom on antimicrobial potency; (4) to replace the
amide group at C-1 position of phenyl ring with other bioisosteric functional groups for
4
9
providing more potential hydrogen bonding interactions with the FtsZ protein.
All newly designed compounds were synthesized as depicted in Scheme 1 and 2. As illustrated
in Scheme 1, the chemical synthesis was initiated with the formation of amide and reduction of
nitro group from a commercially available 2,6-difluoro-3-nitrobenzonic acid (9) following the
5
0-51
reported procedures.
The key intermediate 2,6-difluoro-3-aminobenzamide (10) thus
obtained in large quantity and high yield was further treated respectively with a wide range of
commercially available aryl aldehydes, alkyl bromides, alkenyl bromides, substituted benzyl
bromides or alkyl acid chloride to afford the desired products in one step with moderate to good
yield, allowing a series of compounds to be prepared rapidly for biological study. Reductive
alkylation of 10 with various commercially available aryl aldehydes in the presence of p-
toluenesulfonic acid (p-TsOH) as a catalyst in methanol followed by treatment of sodium
cyanoborohydride afforded the 3-aminobenzamide derivatives 11 - 21 in one-pot with good
yield. Furthermore, alkylation of 10 under the basic condition with different fluoro- or chloro-
substituted benzyl bromides using acetonitrile (ACN) as solvent furnished the mono- and di-
benzyl substituted 3-aminobenzamide derivatives 22 - 25 in good yield. It is worthy to mention
that these mono- and di-benzyl substituted 3-aminobenzamides can be easily purified by using
flash column chromatography simply due to their large polarity difference. Mono-alkylation of
10 with different alkyl bromides or alkenyl bromides gave secondary alkyl or alkenyl substituted
7

ACCEPTED MANUSCRIPT
1
1
1
1
1
1
1
1
1
1
1
1
44
45
46
47
48
49
50
51
52
53
54
55
3-aminobenzamide derivatives 26 - 30 and 32 - 35 in good yield. For amide 31, a different
approach was used. It was successfully prepared in two steps with high yield via the conversion
of nonanoic acid to acid chloride by treating with oxalyl chloride followed by subsequent
reaction of the acid chloride with 3-aminobenzamide 10. Further methylation of 27 and 28 with
dimethyl sulphate under basic condition using ACN as solvent afforded the tertiary 3-
aminobenzamide derivatives 36 - 37 in good yield. 4-Bromosubstituted 3-aminobenzamide
derivative 38 was prepared by the treatment of 28 with excess molecular bromine for 12 h at
room temperature in good yield. Similarly, a small group of 1,4-disubstituted 1,2,3-triazole
derivatives 40 - 42 was accessed in two steps with good yield by the initial formation of azide 39
from 3-aminobenzamide 10 followed by regioselective Cu(I) catalyzed azide-alkyne
5
2
cycloaddition reaction in refluxing tetrahydrofuran (THF) with various terminal alkynes.
8

ACCEPTED MANUSCRIPT
1
56
9

ACCEPTED MANUSCRIPT
o
1
1
1
1
1
1
1
57
58
59
60
61
62
63
2 3 2
Scheme 1. (a) (i) SOCl , cat. DMF, reflux, 2 h; (ii) 30% NH solution, 0 C, 1 h; (iii) SnCl , conc.
o
HCl, 0 C to r.t., 12 h; (b) aryl aldehydes, cat. p-TsOH, MeOH, r.t., 2 h, then NaBH CN, r.t., 12
3
h; (c) For 22 - 25, various benzyl bromides, K CO , ACN, reflux, 4 h; For 26 - 30 and 32 - 35,
2
3
various alkyl or alkenyl bromides, K CO , cat. KI, ACN, reflux, 4 h; For 31, nonanoyl chloride,
2
3
o
Py/DCM, 0 C, 4 h; (d) 28, Br , DCM, r.t., 12 h; (e) Me SO , K CO , ACN, reflux, 12 h; (f) conc.
2
2
4
2
3
o
HCl, NaNO , 0 C, 0.5 h, then NaN , r.t., 4 h; (g) terminal alkynes, cat. Cu(PPh ) Br, THF,
2
3
3 3
reflux, 14 h.
1
64
1
65
10

ACCEPTED MANUSCRIPT
1
1
1
1
1
1
66
67
68
69
70
71
Scheme 2. (a) 43, 46 or 51, 1-bromononane, cat. KI, K
2
CO
3
, DMF, reflux, 12 h; (b) For 45a,
o
Me SO , K CO , ACN, reflux, 12 h; For 47b, methyl iodide, K CO , DMF, sealed tube, 60 C, 24
2
4
2
3
2
3
h; (c) hydroxylamine hydrochloride, NEt , MeOH/THF, reflux, 12 h; (d) For 49a, (i) acetic
3
o
anhydride, AcOH, 0 C, 12 h; (ii) H , Pd/C, MeOH, r.t., 12 h; (iii) conc. HCl, MeOH, reflux, 12
2
o
h; For 49b, (i) 2-chloroacetyl chloride, DCM, 0 C, 12 h; (ii) H , Pd/C, MeOH, r.t., 12 h; (e)
2
NaN , ZnCl , DMF/H O, reflux 12 h.
3
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
Preliminary screening of anti-staphylococcal activity of these compounds revealed that the 2,6-
difluoro-3-aminobenzamide derivatives 28 and 37 exhibited the most potent antimicrobial
activity, implying that the amine groups of secondary n-nonyl amine and tertiary n-nonyl
methylamine at C-3 position of phenyl ring are optimal substituents for the activity. Therefore, a
subseries of 3-aminobenzamides and structurally related derivatives bearing these two important
amino substituents was accessed next to investigate the effect of number and position of fluorine
as well as the bioisosteric replacement of amide group at C-1 position on their antimicrobial
activities. As shown in Scheme 2, mono-alkylation of aminobenzamides 43, 3-amino-2,6-
difluorobenzonitrile 46 and 2,4-difluoroaniline 51 with 1-bromononane under the basic condition
in dimethylformamide (DMF) at elevated temperature afforded the corresponding monoalkylated
aminobenzamides 44, 3-amino-2,6-difluorobenzonitrile 47a and 2,4-difluoroaniline 52 in good
yield respectively. Methylated aminobenzamide 45a and aminobenzonitrile 47b were further
prepared in good yield by treatment of 44a and 47a with dimethyl sulphate or methyl iodide
under basic medium. 2,6-Difluorobenzamidoximes 48, obtained from the reaction of
hydroxylamine hydrochloride with 3-amino-2,6-difluorobenzonitriles 47, were further converted
to the desired 2,6-difluorobenzamidines 49 in two steps with moderate yield. Similarly, treatment
11

ACCEPTED MANUSCRIPT
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
89
90
91
92
93
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
09
10
11
of 3-amino-2,6-difluorobenzonitrile 47a with sodium azide in the presence of zinc(II) chloride at
reflux temperature afforded the C-1 substituted tetrazole 2,6-difluoroaniline 50 in good yield.
Collectively, these types of compounds were easily obtained within 3 to 4 synthetic steps with a
reasonable overall yield by coupling various commercially available building blocks with 3-
aminobenzamides or 3-aminobenzonitrile, allowing rapid construction of compound library for
biological testing.
2. Evaluation of antibacterial and cytotoxic activities, SAR analysis and BLAs combination
studies
With this compound library in hand, we next determined their antibacterial and cytotoxic
activities simultaneously by measuring the minimal inhibitory concentrations (MICs) and the
half-maximal growth inhibition concentration (IC ) against two bacterial cells (E. coli 25922
5
0
and S. aureus 29213) and mouse fibroblasts L929 cell line respectively. The summarized results
are presented in Table 1, in which only compounds with MIC values against S. aureus less than
20 µg/mL are shown. Compounds 1 and 8 were used as a positive control. Both compounds
exhibited potent antibacterial activities against S. aureus with MIC of 0.5 to 1 µg/mL and low
levels of cytotoxicity against L929 cells (IC ≥ 90 µM), providing a relatively higher selectivity
5
0
index (SI) value (Entry 1 and 2 of Table 1). They were, however, completely inactive against the
Gram-negative E. coli even at a concentration of 64 µg/mL. These results were consistent with
1
9, 41
the previous reports.
Time-kill curve evaluation of compound 1 at 2 × and 4 × its MIC
against S. aureus ATCC BAA-41 confirmed its bactericidal mode of action, resulting in more
than 4-log reduction of cell viability within 7 h of drug treatment (Figure 2B). After 24 h drug
treatment, bacterial regrowth was not observed at all concentrations tested.
12

ACCEPTED MANUSCRIPT
2
2
12
13
a
Table 1. Antibacterial and cytotoxic activities of selected compounds.
2
14
b
Compound
No.
1
MIC µg/mL
IC50 µM
L929
> 100
90 ± 8
> 100
60 ± 10
> 100
92 ± 10
> 100
> 100
99 ± 12
> 100
> 100
Entry
NR2
SI
E. coli
> 64
> 64
> 64
> 64
> 64
> 64
> 64
> 64
> 64
> 64
> 64
> 64
> 64
> 64
> 64
> 64
S. aureus
0.5 (1.6)
1 (3.1)
1 (3.1)
1 (3.1)
4 (13)
4 (13)
4 (13)
5 (13)
7 (25)
14 (50)
15 (50)
16 (50)
16 (50)
17 (50)
17 (50)
19 (50)
1
2
3
4
5
6
7
8
9
N.A.
N.A.
> 63
29
> 32
19
> 8
7
> 8
> 8
4
8
28
37
30
nonylamino
methyl(nonyl)amino
non-2-ylamino
36
methyl(octyl)amino
methyl(nonyl)amino
bis(4-chlorobenzyl)amino
cis-non-2-en-1-ylamino
octylamino
48b
25b
34
1
1
1
1
1
1
0
1
2
3
4
5
6
27
> 2
> 2
> 2
> 2
2
49b
48a
50
11
12
methyl(nonyl)amino
nonylamino
> 100
> 100
99 ± 10
87 ± 15
> 100
nonylamino
3-(n-butyloxy)benzylamino
3-(n-pentyloxy)benzylamino
bis(4-fluorobenzyl)amino
2
> 2
1
22b
a
2
2
2
15
16
17
N.A., not applicable. SI, selectivity index, it was calculated using the formula IC50 (bµM)
L929/MIC value of S. aureus (µM). All experiments were performed in at least triplicates. µM
in the parentheses.
2
18
13

ACCEPTED MANUSCRIPT
2
2
19
20
Figure 2. Time-kill curves of (A) 28, (B) 1, combinations of methicillin (ME) and (C) 28 or (D)
1 against S. aureus ATCC BAA-41. The error bars indicate standard derivations from
measurements of triplicates. (E) Percentage of clinical MRSA isolates exhibiting synergistic
effect (FIC index ≤ 0.5) to combinations of different BLAs with 28. Twenty-eight MRSA strains
were tested in total. MR, meropenem; CX, cefuroxime; CL, cloxacillin; ME, methicillin; AM,
amoxicillin. (F) In vivo efficacy of intraperitoneal co-administering single agent of vehicle, CX,
2
2
2
2
2
21
22
23
24
25
14

ACCEPTED MANUSCRIPT
2
2
26
27
28 or combination of CX and 28 twice a day in a murine systemic infection model of MRSA
ATCC 43300.
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
In general, among all newly synthesized compounds, low levels of cytotoxicity against normal
cell L929 were observed with IC values ranged from 60 µM to > 100 µM, implying that these
5
0
compounds are potentially non-toxic and safe. Below 100 µM concentration, this class of
compounds is unlikely to interact with other protein targets and induce cellular toxicity. Their
IC values are at least twice the observed MIC values, in particular, compound 28 displayed the
5
0
highest SI of > 32 (Entry 3 of Table 1). Moreover, all newly synthesized compounds are also
completely inactive against Gram-negative E. coli (MIC > 64 µg/mL), perhaps it is due to the
intrinsic low permeability of compound itself to pass through the cell membrane of E. coli or the
membrane efflux pumps presented in the E. coli, causing them far from reaching the drug target.
More experiments on Gram-negative bacteria are required to test these hypotheses.
Among all the tested compounds, two compounds, namely 28 and 37, displayed comparable
anti-staphylococcal activity (MIC = 1 µg/mL) and selectivity index (≥ 19) with the positive
controls. Interestingly, both compounds possess the common structural features of a warhead of
2,6-difluorobenzamide and a hydrophobic tail of n-nonylamino group. Detailed structure-activity
relationships (SAR) analysis on the benzamide head and n-nonylamino tail revealed several
structural features that are crucial to maintain the anti-staphylococcal activity. For the benzamide
warhead, firstly, bioisosteric replacements of carboxamide group at C-1 position of compound 28
with other functional groups, such as N-hydroxycarboximidamide (compound 48a),
carboximidamide (compound 49a) and tetrazole (compound 50), weakened the antibacterial
15

ACCEPTED MANUSCRIPT
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
activity. Similarly, replacement of carboxamide group of compound 28 with carbonitrile
(compound 47a) or hydrogen (compound 52) even resulted in no antibacterial activity. Secondly,
both the position and the number of fluorine atom on the phenyl ring play a very important role
in the antibacterial potency. 2,6-Difluoro-substituted functional group of compound 28 exhibited
the most potent antibacterial activity while reducing the number of fluorine atom to one
(compound 44b) or zero (compound 44a) or varying the position of fluorine atoms to C-4 and C-
6 positions (44c) lost their antibacterial activity. Thirdly, the secondary (compound 28) or
tertiary (compound 37) amino groups at the C-3 position of the phenyl ring offered the most
potent antibacterial activity. Installation of less freely rotatable substituents at this position, such
as amide (compound 31) and 1,4-disubstituted triazole moieties (compound 41) dramatically
reduced the antibacterial activity. On the other hand, for the n-nonylamino tail, several structural
features, including the length, rigidity, bulkiness and lipophilicity, interfere the potency of
antibacterial activity. Replacing the optimal n-nonylamino group with a longer n-decylamino
group (compound 29) or shorter n-heptyl (compound 26) and n-octyl amino group (compound
27) of straight alkyl chains or branched 2-nonylamino group (30) diminished sharply in the
antibacterial activity. Moreover, increasing the chain rigidity by the introduction of alkene
(compounds 33 and 34) or benzyloxy ring (compounds 11 and 12) in the amino tail also
weakened their antibacterial activity significantly. Both decreasing the chain lipophilicity by
introducing an oxygen atom (compound 32) in the middle of the chain and increasing the chain
bulkiness by installing a phenyl ring at the terminal position (compound 35) of the alkyl chain
lead to no antibacterial activity. Taken together, compound 28 demonstrated the most promising
SI value among all tested compounds, it was selected for detailed biological characterization.
16

ACCEPTED MANUSCRIPT
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
Surprisingly, time-kill curve evaluation of compound 28 clearly indicated that its mode of
action is bacteriostatic (cells show arrested growth), but not bactericidal, because it required 16 ×
its MIC to kill the bacteria within 24 h (Figure 2A). After 24 h of drug treatment, bacterial
regrowth was observed at concentrations below 16 × its MIC. We next assessed the synergistic
effect of this compound in combination with a wide range of clinically used BLAs, including
penicillin-type antibiotics methicillin (ME), cloxacillin (CL) and amoxicillin (AM),
cephalosporin-type antibiotic cefuroxime (CX) and carbapenem-type antibiotic meropenem
(MR), against a panel of twenty-eight clinical MRSA strains. As shown in Table 2, some of
these strains exhibited a high level of drug resistance to multiple BLAs with MIC values ranging
from 2 µg/mL to 1024 µg/mL. Encouragingly, combination studies revealed that compound 28
demonstrated strong synergistic effect with all tested BLAs against these three clinical MRSA ,
with calculated fractional inhibitory concentration (FIC) index as low as 0.1 (Table 2).
Moreover, as shown in Figure 2E, 82%, 75%, 68%, 61% and 11% of clinical MRSA isolates
exhibited synergistic effect (FIC index ≤ 0.5) to the combinations of ME, CX, CL, AM and MR
antibiotics with compound 28 respectively (Table S1). These results suggested that compound
28 has a board spectrum for BLAs combination and is, therefore, an excellent BLAs adjuvant. In
addition, time-kill curve evaluation of the combination of compound 28 and ME revealed that
the mode of action is bactericidal (Figure 2C), which is similar to the combination of compound
1 and ME (Figure 2D). After 21 h of drug combination treatment, bacterial regrowth was not
observed at all concentrations tested for both combinations.
2
2
92
93
Table 2. Combination studies of compound 28 with various BLAs against selected clinically
a
2
94
isolated MRSA strains and calculated FIC index.
17

ACCEPTED MANUSCRIPT
MIC (µg/mL)
FIC Index of combination
MRSA
Strain
No.
ME
+28
2
4
4
CL
+28
2
CX
+28
2
AM
+28
8
8
8
MR ME CL CX AM MR
+28 +28 +28 +28 +28 +28
2
8
ME
CL
CX
AM
MR
4
17
32 1024
64
16
2
1024
512
512
64
512
64
16
32
4
4
4
0.1 0.1 0.1 0.3 0.2
0.2 0.2 0.1 0.4 0.4
0.3 0.5 0.1 0.1 0.1
2
516
32
512
64
16
2
2
2
7
74
a
1
1024
2
2
2
2
2
3
95
96
97
98
99
00
ME, Methicillin; CL, cloxacillin; CX, cefuroxime; AM, amoxicillin; MR, meropenem. FIC
index is calculated by using the formulate FIC index = FIC (compound) + FIC (drug), where FIC
(compound) is the (MIC of compound in combination with drug)/(MIC of compound alone)
while FIC (drug) is the (MIC of compound in combination with drug)/(MIC of drug alone). The
combination is considered synergistic if the FIC Index ≤ 0.5. All experiments were performed in
at least triplicates.
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
3. In vivo efficacy of combination of CX and 28 against MRSA ATCC 43300
On the basis of in vitro data which show that compound 28 is broadly synergistic in
combination with various BLAs against diverse clinically relevant MRSA strains and relatively
non-cytotoxic to mouse peritoneal fibroblast L929 (IC50 > 100 µM), we next pursued the
synergistic efficacy of compound 28 in combination with CX when co-administered
intraperitoneally (IP) to a murine systemic infection model of MRSA. The preclinical model of
infection using MRSA ATCC 43300 has been frequently employed to predict the clinical
2
9
antibiotic efficacy. Among those BLAs that have been tested in vitro, CX was selected because
it is an oral antibiotic, which would enjoy a higher patient acceptance. MIC studies demonstrated
that combination of CX and 28 also exhibited strong synergistic effect against MRSA ATCC
43300 with a FIC index of 0.1, prompting us to carry out in vivo efficacy studies. Preliminary
dose regime studies indicated that CX and compound 28 co-administered IP both at 50 mg/kg
once a day provided a survival rate of 33%, but all the mice died at day 5 upon treatment by CX
or compound 28 as a single agent (Figure S52A). These preliminary results suggested that such
combination therapy is efficacious against MRSA ATCC 43300 although the survival rate was
only moderate. We reasoned that such low survival rate is likely attributed to the hydrophobic
18

ACCEPTED MANUSCRIPT
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
nature of compound 28 (cLogP = 5.0) that may cause high plasma protein binding and reduced
potency. Nonetheless, an adjusted dose regime of compound 28 (50 mg/kg) and CX (25 mg/kg)
at twice a day was tested next for improving the survival rate. As shown in Figure 2F, CX (25
mg/kg) and compound 28 (50 mg/kg) administered IP as a single agent only provided 70% and
40% survival rate respectively in treating mice with MRSA infection compared with the vehicle
treatment (50% survival rate). Encouragingly, IP co-administering both compound 28 and CX at
these dosages provided a significant increase in survival rate to 100% after 4 days of
combination therapy. In addition, no compound 28-CX-resistant mutants were identified among
the CFU recovered from the in vivo study and no obvious trauma around the injection site of
compound 28 was observed (Figure S52B). Collectively, these data provide strong evidence
supporting the hypothesis that compound 28 may provide an alternative strategy to develop as a
bactericidal BLAs combination agent that is efficacious against the clinical MRSA infection.
4. Validation of FtsZ protein as the drug target of compound 28
PC190723 (1) has been shown to inhibit the bacterial cell division process through targeting
the binding site at T7-loop of S. aureus FtsZ protein by using the protein-ligand crystal co-
1
8, 53
complex.
Structurally, compound 28 also possesses the same 2,6-difluorobenzamide
warhead, but with a more freely rotatable n-nonylamino substituent at the C-3 position of the
phenyl ring. Due to their overall structural unlikeness, the next question we need to answer is
whether this compound still bind to the same binding site at T7-loop of FtsZ protein and interfere
the cell division process in a similar way. To address this question, we sought to conduct the
following series of biochemical and microscopic studies to prove that the anti-staphylococcal
19

ACCEPTED MANUSCRIPT
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
activity of compound 28 reflects its ability to target T7-loop of S. aureus FtsZ protein and
interfere with the downstream cell division process.
4.1 Isolation of compound 28 resistant mutants for genetic studies and computational docking
studies
The frequency of resistance (FOR) assays indicated that bacterial cells of S. aureus ATCC
1717 were able to grow even in the presence of 4-fold or 16-fold MIC of compound 28 as a
single agent (Figure 3A), suggesting that potential genetic mutations in the target protein may
have been induced, resulting in drug resistance. However, no colony was observed for the plates
treated with the combination of 28 and CX after 48 h incubation, implying a reduced rate of drug
resistance development. Therefore, the most definitive approach for in vivo target identification
of compound 28 is through the drug resistance mapping analysis of compound 28-resistant
isolates, demonstrating that mutations in the target protein result in drug resistance. In this
connection, we have employed a large-inoculum approach in an effort to raise spontaneous
resistant mutants of S. aureus ATCC 29213 strains that are highly resistant to compound 28. This
approach successfully yielded three compound 28-resistant strains with MIC values of 32
µg/mL, 64 µg/mL and 128 µg/mL respectively (Figure 3B, upper part). The genetic materials in
each resistant strain as well as the wild-type strain were isolated and subjected to whole genome
sequencing followed by sequence alignment to identify any nucleotide changes. Surprisingly,
compared with the wild-type strain, the sequencing results indicated that all three compound 28-
resistant strains carried the same single nucleotide change of G786A, which is corresponding to
the amino acid substitution of M262I that mapped to the S. aureus FtsZ protein (Figure 3B,
upper part). Previous mutational analysis of PC190723 (1)-resistant mutants also identified
20

ACCEPTED MANUSCRIPT
3
3
3
3
3
3
3
63
64
65
66
67
68
69
several major amino acid substitutions that mapped to FtsZ protein, including G193D, G196A
1
8-19
and N263K (Figure 3B, lower part).
The amino acid substitution of M262I was found to be
located exactly at the same binding pocket of PC190723 (1), suggesting that compound 28 is
very likely to bind directly to the S. aureus FtsZ protein in the same manner as PC190723 (1).
Our mutational analysis is, therefore, consistent with the FtsZ protein being the antibacterial drug
target of compound 28.
3
3
70
71
Figure 3. (A) FOR studies of compound 28 alone and combination of compound 28 and CX
showing the number of colony and (B, upper part) Summary of MIC, DNA nucleotide changes
and amino acid substitutions of compound 28-resistant mutants and (B, lower part) Model of
compound 28 (blue sticks) docked into the T7-loop cleft of FtsZ using the crystal structure of S.
aureus FtsZ protein (PDB ID: 4DXD) with labelled helix 7 (H7), T7-loop and amino acid
residues G193, G196, M262, N263 and T309. The grey dotted line indicates the potential
3
3
3
3
3
72
73
74
75
76
21

ACCEPTED MANUSCRIPT
3
3
77
78
hydrogen bonding interaction between the C-3 amino group of compound 28 and the hydroxyl
group of T309.
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
To gain more insights into the potential binding site and binding pose of compound 28 in the S.
aureus FtsZ protein, computational docking studies of compound 28 using previously reported
18
crystal structure of S. aureus FtsZ protein (PDB ID: 4DXD) was conducted next. The results of
docking studies revealed that the highest docking score positioned compound 28 into a cleft
between the helix 7 (H7) and the C-terminal domain of FtsZ, which is in good agreement with
PC190723 (1) (Figure 3B, lower part). The 2,6-difluorobenzamide warhead of 28 was well-
situated in the hydrophobic pocket interacting with the T7-loop of FtsZ protein. A conventional
hydrogen bonding interaction was predicted to be established between the C-3 amino group of
compound 28 and the hydroxyl group of T309. The amino acid residues of M262, G193, G196
and N263 shown in Figure 3B were closely adjacent to the residues comprising the binding
pocket of 28 proposed by the docking study. These results suggested that potential amino acid
substitutions at this binding pocket are likely to be induced easily by small molecules that bind to
this pocket. The resultant changes appear to alter slightly the overall shape of this binding pocket
without interfering the normal function of FtsZ protein, resulting in compound 28 or PC190723
(1) no longer binding to FtsZ protein and causing drug resistance.
4.2 Effect on FtsZ protein polymerization and GTPase activity upon compound 28 treatment
Previous reports have shown that the antibacterial activities of PC190723 (1) are resulted from
the overstimulation of FtsZ protein polymerization through stabilizing the nonfunctional FtsZ
5
4
polymeric structures. In order to confirm whether compound 28 would exert similar effect on
22

ACCEPTED MANUSCRIPT
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
FtsZ protein polymerization, we next expressed and purified the S. aureus FtsZ protein for
assessment of its polymerization dynamics in the absence or presence of compound 28 using an
in vitro light scattering assay. In this assay, the monomeric FtsZ protein polymerization was
continuously monitored in the presence of GTP by a time-dependent increase in light scattering
as reflected by an increase in solution absorbance at 600 nm. The results of FtsZ protein
polymerization in the presence of compound 28 at concentrations ranging from 12.5 µM to 100
µM were shown in Figure 4A. Surprisingly, compound 28 potently inhibited the FtsZ protein
polymerization in a concentration-dependent manner, a behavior that is opposite to that of
PC190723 (1), which stimulates FtsZ protein polymerization at concentrations of 12.5, 25 and 50
µM in a dose-dependent manner (Figure S53). Surprisingly, compound 1 at 100 µM completely
inhibited FtsZ protein polymerization. On the other hand, compared with the vehicle control (1%
DMSO) at 500 seconds, compound 28 at 25 µM, 50 µM and 100 µM exhibited about 34%, 39%
and 47% inhibition of FtsZ protein polymerization respectively. These results suggested that
compound 28 is able to perturb the FtsZ protein polymerization in vitro.
23

ACCEPTED MANUSCRIPT
4
4
15
16
Figure 4. (A) Effect of compound 28 at different concentrations on the kinetics of S. aureus FtsZ
4
17
polymerization. The experiments were performed in triplicate with the symbols indicating the
24

ACCEPTED MANUSCRIPT
4
4
4
18
19
20
mean value. Electron micrographs of FtsZ polymer after the treatment of compound 28 at (B) 0
µM, (C) 100 µM and (D) 50 µM. The scale bar is 500 nm. (E) Effect of compound 28 at various
concentrations on the GTPase activity of S. aureus FtsZ protein.
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
To further demonstrate the effect of compound 28 on inhibition of FtsZ protein
polymerization, transmission electron microscopy (TEM) imaging of the compound 28 treated
and untreated S. aureus FtsZ protein was carried out to investigate the morphological change of
FtsZ filaments. S. aureus FtsZ protein treated with compound 28 at concentrations of 0, 50 and
100 µM in the presence of GTP were visualized in Figure 4B, 4D and 4C respectively. As
anticipated, there was considerable reduction in the extent of FtsZ filament formation upon
treatment of compound 28 compared to the untreated FtsZ protein. The magnitude of these
suppressing effects increases with the increasing concentration of compound 28. At 100 µM of
compound 28, the density of S. aureus FtsZ filaments was substantially reduced, producing
short, thin and single strand FtsZ filaments (Figure 4C), implying that compound 28 may block
the FtsZ protein polymerization in a longitudinal and lateral manner. In a sharp contrast, the
untreated S. aureus FtsZ protein showed a heavily dense network of FtsZ filaments (Figure 4B).
These results clearly indicated the highly efficient inhibition of S. aureus FtsZ assembly to form
filaments by compound 28 at a dose-dependent manner, which is consistent with the results of
light scattering assay.
The GTPase activity of FtsZ protein also plays an important role of assembling monomeric
FtsZ proteins by hydrolyzing GTP molecules as an important energy source for driving
polymerization. Compound 1 has been reported to inhibit directly the GTPase activity of FtsZ in
25

ACCEPTED MANUSCRIPT
1
9
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
a concentration-dependent manner with a half-maximal inhibitory concentration of 55 ng/mL.
5
5
On the contrary, Our laboratory and others did not observe such inhibitory effect. Compound 1
at 30, 50 and 100 µM concentrations even increased the GTPase activity by 47%, 29% and 15%
respectively (Figure S53E). On the other hand, as shown in Figure 4E, there was no significant
change of the GTPase activity for compound 28 even at the concentration of 100 µM, suggesting
that compound 28 is likely to perturb FtsZ protein polymerization through binding to the T7-loop
of FtsZ protein without interfering its GTPase activity.
4.3 Microscopic studies of bacterial morphology and localization of the Z-ring of B. subtilis and
S. aureus cells
Formation of Z-ring at the appropriate site of cytokinesis is one of the most important
5
6
prerequisites for bacteria to carry out cell division properly. Microscopic studies of previous
reports have demonstrated that small molecules, which block the Z-ring formation through
inhibition of FtsZ protein polymerization, at a sublethal concentration induced both iconic
elongated phenotype in rod-shaped B. subtilis cells and enlarged phenotype in spherical S.
aureus cells respectively. Moreover, an obvious septal delocalization of green fluorescent protein
(GFP)-tagged FtsZ polymers was also observed in both cells. As shown in Figure 5 and S54,
such morphological changes and septal delocalization of GFP-tagged FtsZ polymer after
treatment of compound 28 were confirmed. Fluorescent microscopic studies indicated that
fluorescent foci at the mid cell were observed in the presence of 1% DMSO, implying the proper
formation and localization of Z-ring at the appropriate division septum (Figure 5A and S54C).
Upon treatment of compound 28 or 1, multiple discrete foci throughout the whole elongated B.
subtilis 168 cells (Figure S54E), and enlarged S. aureus RN 4220 cells (Figure 5B and 5C)
26

ACCEPTED MANUSCRIPT
4
4
4
4
4
4
4
4
4
4
4
64
65
66
67
68
69
70
71
72
73
74
were observed respectively, demonstrating the markedly altered localization of Z-ring without
being specifically restricted to the division septum. Moreover, for the bacterial morphology,
elongated B. subtilis 168 cells (Figure S54B) and enlarged S. aureus ATCC BAA-41 cells
(Figure 5E and 5F) were observed respectively upon treatment of compound 28 or 1. These
results are consistent with other reported FtsZ inhibitors.
Combining all the studies related to compound 28, we reasonably propose that S. aureus FtsZ
protein is probably the drug target of compound 28 and it is likely to inhibit the S. aureus FtsZ
protein polymerization process through binding to the T7-loop of FtsZ protein, causing
subsequent delocalization of Z-ring and disruption of the cell division process.
4
4
75
76
Figure 5. Fluorescent microscopic study (upper panel) of FtsZ-GFP fusion stain of S. aureus RN
4220 cells in the presence of (A) 1% DMSO, (B) 4 × MIC of compound 28 and (C) 4 × MIC of
4
77
27

ACCEPTED MANUSCRIPT
4
4
4
4
78
79
80
81
compound 1. The scale bar is 7.5 µm and the red arrows indicated the fluorescent foci.
Histograms (lower panel) showing the normalized frequency distribution of cell volume of S.
aureus ATCC BAA-41 in the presence of (D) 1% DMSO, (E) 2 × MIC of compound 28 and (F)
2 × MIC of compound 1 with the indicated cell volumes at mode and mean respectively.
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
5
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
00
5. PK profile of compound 28
Oral bioavailability is one of the key considerations for developing bioactive molecules as
therapeutic agents. Lead compounds with poor oral bioavailability may result in low efficacy and
unpredictable drug response. Previous study indicated that the small intestine of rat and human
exhibit similar drug absorption profiles and transporter expression patterns, providing a more
5
7
easier prediction of oral drug absorption potential in human. In this connection, we sought to
reveal the rat plasma concentration-time profile of compound 28 upon intravenous injection (IV)
at a dose of 1 mg/kg and oral administration (PO) at a dose of 50 mg/kg (Figure 6, left). The rat
PK parameters of compound 28 are listed in Figure 6 (right). The results of PO indicated that
compound 28 exhibits a fast absorption (Tmax = 2 h) with a peak plasma concentration (Cmax) of
1.9 µg/mL. The terminal half-life (t ) representing the time required for systemic level of
1
/2
compound 28 to reduce by half following PO and IV administrations were 3 and 5.5 h
respectively. The area under the curve (AUC ) representing the total systemic drug exposure
0
-∞
for PO and IV were 10.3 and 1.6 mg/L·h respectively. Thus, the oral bioavailability (F) of
compound 28, which is the fraction of a compound that reaches systemic circulation following
oral administration, was moderate at 13%. Taken together, these PK parameters indicated that
compound 28 has moderate oral drug absorption in rat and compound 28 can serve as a lead for
further structural optimization.
28

ACCEPTED MANUSCRIPT
5
01
5
5
02
03
Figure 6. (A) The plasma concentration-time profile of compound 28 upon intravenous injection
5
5
04
05
(pink square) and oral administration (dark triangle) in rats and (B) PK parameters of compound
28.
5
5
5
5
5
5
5
5
5
5
5
5
5
06
07
08
09
10
11
12
13
14
15
16
17
18
Conclusion
In summary, a focused compound library of 3-aminobenzamides and structurally related
derivatives has been designed and synthesized for evaluation of their antibacterial and cytotoxic
activities against bacterial cells and normal cells. These compounds were easily obtained in 3 to
4 synthetic steps by coupling of various commercially available building blocks with 3-
aminobenzamides or 3-aminobenzonitrile, allowing rapid construction of the compound library
for SAR analysis. Our efforts have yielded a compound, 28, which exhibits low cytotoxicity
against normal cells and robust in vitro bactericidal synergy with different classes of BLAs
against a panel of multidrug-resistant clinical MRSA isolates with FIC index as low as 0.1.
Further target identification and mechanistic studies employing a series of genetic study,
computational docking, biochemical assays and microscopic studies have revealed that
compound 28 is likely to interact with the S. aureus FtsZ protein at the T7-loop binding pocket
29