Hydrophobic substituents on isatin derivatives enhance their inhibition against bacterial peptidoglycan glycosyltransferase activity

Article abstract of DOI:10.1016/j.bioorg.2020.103710

Moenomycin A, the well-known natural product inhibitor of peptidoglycan glycosyltransferase (PGT), is a large amphiphilic molecule of molecular mass of 1583 g/mol and its bioavailablity as a drug is relatively poor. In searching for small-molecule ligands

Full text of DOI:10.1016/j.bioorg.2020.103710

Journal Pre-proofs
Hydrophobic Substituents on Isatin Derivatives Enhance Their Inhibition
against Bacterial Peptidoglycan Glycosyltransferase Activity
Yong Wang, Wing-Lam Cheong, Zhi-Guang Liang, Lok-Yan So, Kin-Fai
Chan, Pui-Kin So, Yu-Wai Chen, Wing-Leung Wong, Kwok-Yin Wong
PII:
S0045-2068(20)30105-X
DOI:
https://doi.org/10.1016/j.bioorg.2020.103710
YBIOO 103710
Reference:
To appear in:
Bioorganic Chemistry
Received Date:
Revised Date:
Accepted Date:
14 January 2020
23 February 2020
26 February 2020
Please cite this article as: Y. Wang, W-L. Cheong, Z-G. Liang, L-Y. So, K-F. Chan, P-K. So, Y-W. Chen, W-L.
Wong, K-Y. Wong, Hydrophobic Substituents on Isatin Derivatives Enhance Their Inhibition against Bacterial
Peptidoglycan Glycosyltransferase Activity, Bioorganic Chemistry (2020), doi: https://doi.org/10.1016/j.bioorg.
2020.103710
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. Please note that, during the production process, errors
may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier Inc.

Hydrophobic Substituents on Isatin Derivatives Enhance Their Inhibition against
Bacterial Peptidoglycan Glycosyltransferase Activity
Yong Wang‡, Wing-Lam Cheong‡, Zhi-Guang Liang, Lok-Yan So, Kin-Fai Chan, Pui-Kin So,
Yu-Wai Chen, Wing-Leung Wong and Kwok-Yin Wong*
State Key Laboratory of Chemical Biology and Drug Discovery, Department of Applied Biology
and Chemical Technology, The Hong Kong Polytechnic University, Hunghom, Kowloon, Hong
Kong, P. R. China.
‡ The authors contributed equally to this work.
* Corresponding author
E-mail: kwok-yin.wong@polyu.edu.hk
Abstract
Moenomycin A, the well-known natural product inhibitor of peptidoglycan glycosyltransferase
(PGT), is a large amphiphilic molecule of molecular mass of 1583 g/mol and its bioavailablity as
a drug is relatively poor. In searching for small-molecule ligands with high inhibition ability
targeting the enzyme, we found that the addition of hydrophobic groups to an isatin-based
inhibitor of bacterial PGT significantly improves its inhibition against the enzyme, as well as its
antibacterial activity. The improvement in enzymatic inhibition can be attributed to a better
binding of the small molecule inhibitor to the hydrophobic region of the membrane-bound
bacterial cell wall synthesis enzyme and the plasma membrane. In the present study, a total of 20
new amphiphilic compounds were systematically designed and the relationship between
molecular hydrophobicity and the antibacterial activity by targeting at PGT was demonstrated.
The in vitro lipid II transglycosylation inhibitory effects (IC50) against E. coli PBP1b and MICs
of the compounds were investigated. Optimized results including MIC values of 6 g/mL for
MSSA, MRSA, B. subtilis and 12 g/mL for E. coli were obtained with an isatin derivative 5m
which has a molecular mass of 335 g/mol.
1. Introduction
The World Health Organization’s recent surveillance data on antibiotic resistance revealed
high levels of resistance in a number of serious bacterial infections across many countries [1]. The
development of antimicrobial resistance occurs so quickly that it was estimated by 2050 the
number of death caused by antibiotic resistance will exceed that caused by cancer [2]. Thus, there
1

is an urgent need to develop the next generation antibiotics based on new drug targets or
therapeutic mechanisms.
The peptidoglycan glycosyltransferase (PGT) that transfers the disaccharide-peptide of
lipid II to the growing glycan chain in bacterial cell wall synthesis is regarded as a potential new
drug target for new antibiotic development [3]. The PGT domain of penicillin-binding protein is
anchored on the outer surface of the bacterial cell membrane by a transmembrane helix (Fig. 1)
[4-6]. Lipid II, the substrate of PGT, is a large amphiphilic molecule containing a hydrophilic
disaccharide-peptide head and a hydrophobic undecaprenyl diphosphate tail for its anchoring to
the bacterial cell membrane during polymerization [7-9]. Moenomycin A (MoeA), the well-
known inhibitor for PGT activity, is also an amphiphilic molecule consisting of a hydrophilic
pentasaccharide head and a hydrophobic C25-lipid moiety (Fig. 2) [10, 11]. However, the poor
pharmacokinetic properties of MoeA prevents its direct utilization as antibiotic for human, and no
PGT inhibitors are available for clinical trial or used currently as antimicrobial agents [12].
Nonetheless, previous understanding on the chemistry and biological functions of MoeA [10, 12]
and its analogues provide helpful information for rational design of PGT inhibitors [13-15].
Recently, a small molecule-based inhibitor adopting tryptamine as the main molecular scaffold
has been reported [16].
Fig. 1. The bacterial peptidoglycan biosynthesis. E. coli lipid II and PBP1b (PDB ID: 3VMA)
were adopted in this figure.
2

The active site of PGT is hydrophilic in nature, which interacts with the disaccharide-
peptide unit of lipid II or the pentasaccharide moiety of MoeA, while the adjacent surface is
hydrophobic, allowing it to immerse in the membrane surface [5]. The structure of PGT also
revealed an exposed hydrophobic region close to the active site, which most likely engages in the
interaction with the lipid chains of the substrate while it enters the active site [5]. We therefore
reason that an effective inhibitor of PGT is likely to be amphiphilic in nature and its
hydrophobicity has to be designed in a well-balanced manner in order to maximize the interaction.
As shown in Fig. 2, a small molecular fragment of guanidine-tagged isatin is selected to design
an effective PGT inhibitor due to its known antibacterial and moderate PGT inhibitory activities
[17-19], and a non-polar group (R) with adjustable length can be linked to the nitrogen atom of γ-
lactam to tune its hydrophobicity. In this study, a new series of isatin-based amphiphilic PGT
inhibitors were synthesized and their Quantitative Structure-Activity Relationship with respect to
the substituent (R group) of the molecules was studied.
HO
NHAc
HO
HO
O
O
O
O
AcHN
O
P
OH
O
O
P
O
HN
HO
O
O
HO
O
O
HO
NH
O
8
2
H
N
E. coli lipid II
NH2
COOH
O
NH
O
H
N
OH
O
NH2
O
O
NHAc
OH
O
H2N
O
OHO
O
HO
OH
O
O
OH
NH
HO
O
O
NH2
O
OH
HN
AcHN
HO
O
O
OH
Hydrophilic unit
O
HN
P
N
HO
OH
O
O
OH
H
O
O
O
O
OH
N
R
Substituents of different
hydrophobicities
Amphiphilic small molecule
MoeA (M.W. = 1583 Da)
Fig. 2. Chemical structures of E. coli lipid II, MoeA and a guanidine-tagged isatin amphiphilic
derivative.
3

2. Results and discussions
2.1 Synthesis of inhibitors
Our previous study on isatin derivatives [19] as PGT inhibitors identified compound 3d as
a relatively potent inhibitor. We reasoned that the effectiveness of 3d could be caused by a good
balance of hydrophilicity and hydrophobicity betwwen the polar aminoguanidine group and the
non-polar naphthyl group. In order to explore the molecular hydrophilicity influence in the
structure-activity relationship of PGT inhibitors in-depth, twenty new amphiphilic isatin-based
inhibitors were systematically designed by regulating the hydrophobicity of the molecules via R
group modifications. These small-molecular based inhibitors were synthesized through two
different synthetic routes as shown in Scheme 1. For inhibitor 3a-3c, they were synthesized similar
to 3d as previous;y reported [19] by reacting the primary amines with chloroacetyl chloride in
acetonitrile to yield 1, followed by the reaction with isatin or 5-methylisatin in DMF with
potassium carbonate to form 2. Compounds 3a-3d were obtained by reacting 2 with
aminoguanidine HCl in acetic acid as the solvent under reflux conditions. The compounds were
purified by flash column chromatography and isolated with 52-61 % yields.
Inhibitors 5a-5p were synthesized by the Chan-Evans-Lam coupling reaction [20-22]. The
intermediates 4 with a desirable substituent (R2 group) were prepared through the reaction of
phenylboronic acids and isatin with Cu(OAc)2 as the catalyst. Compound 4 was reacted with
aminoguanidine in acetic acid at room temperature to give 5a-5p. The yields of 5a-5c were very
low in the reaction with aminoguanidine because esterification of acetic acid with the hydroxyl
groups of 4 also occurred under reflux condition. However, this side reaction can be reduced
significantly when the reaction is taken place under room temperature for 4 h. The isolated yields
for the compound were 29-36 %, which is comparable with their analogues 5d-5p (35-56 %
yields).
As shown in Fig. 2, the overall hydrophobicity of the isatin-based inhibitors can be tuned
by varying the substituent (R group). The hydrophobicity of this series of 20 compounds were
estimated by calculating the averaged LogP (cLogP) values with four different reported
calculation methods [23-26]. The averaged cLogP values of the compounds ranged from -1.9 to
4.4. The molecular structure and cLogP values are listed in Table 1 and S2. In general, compounds
with the substituent R containing a terminal hydroxyl group showed lower averaged cLogP values,
while those compounds possessing a terminal alkoxyl group showed medium averaged cLogP
values whereas higher averaged cLogP values of compounds were obtained with a terminal alkyl
group of different chain lengths.
4

O
H
N
R1
O
N
O
1
O
R'
H
O
N
R1
Cl
Cl
O
R1 NH2
N
Cl
i
ii
H
O
R'
2
O
R1
NH
3a, R' = -CH3, R1 =
3b, R' = -H, R1 =
N
H
NH2
N
H2N
N
O
NO2
H
iii
N
R'
HN
NH
H2N
3c, R' = -H, R1 =
3d, R' = -H, R1 =
3
CF3
R2
NH
N
O
NH2
R2
N
H
N
H2N
N
H
R2
B
O
O
N
HN
v
iv
HO
OH
O
O
NH
5
H2N
4
O
O
5j R2 =
5d R2 =
5e R2 =
5f R2 =
5g R2 =
OH
5a R2 =
5b R2 =
5c R2 =
OH
O
5k R2 =
5l R2 =
OH
O
O
5m R2 =
5n R2 =
O
5h R2 =
5i R2 =
5o R2 =
5p R2 =
O
o
Scheme 1. Synthetic routes of isatin derivatives. (i) MeCN, 3h; (ii) DMF, K2CO3, 80 C, 12 h;
(iii) acetic acid / reflux, 2 h; (iv) Cu(OAc)2, Et3N, CH2Cl2, 48 – 72 h; (v) acetic acid / room
temperature, 4h.
5

2.2 Antimicrobial susceptible testing and their structure-activity relationship
To study the structure-activity relationship between the R group (Fig. 2) of the isatin
derivatives and their antibacterial potencies, their antibacterial activities targeting PGT were
examined with Gram-positive S. aureus, B. subtilis and Gram-negative E. coli, including the
methicillin-resistant S. aureus strain (ATCC BAA-41). The MIC results and averaged cLogP
values are listed in Table 1. Starting with compounds with methanediylamidyl linkage conserved
(3a-3d, Scheme 1), 3a with a terminal m-nitrobenzyl group and its R’ group replaced from hydro
group to methyl group did not further improve the antibacterial activities compared to its parent
molecule (R’ = H), suggested that the R’ of 3a may not be a promising position for further
modifications. However, 3c showed slight improvement to the antibacterial activity on the 3
strains (24 g/mL on S. aureus; 12 g/mL on B. subtilis and 48 g/mL on E. coli) while 3d showed
even better potencies on S. aureus (12 g/mL), the rigid non-polar naphthyl group of 3d should
play an important role on the enhanced activities.
In compounds 5a-5p, the substituent groups (R2) can be further categorized as polar
hydroxyl (5a – 5c), polar alkoxyl (5d – 5i) and non-polar (5j – 5p) groups (see Scheme 1), to offer
a wide range of hydrophobicity for bioactivity comparison. MIC results (Table 1) showed that all
compounds with polar hydroxyl and alkoxyl groups (except 5i) had no significant antibacterial
activities against both Gram-positive and Gram-negative bacteria. These results may indicate that
both the hydroxyl and alkoxyl groups in the substituent cannot achieve a well-balanced
hydrophobicity for the amphiphilic inhibitors. Interestingly, the antibacterial potencies increased
significantly when we adopted non-polar side groups (5j – 5p). Considering 5j and 5k, both with
naphthyl substituents, but 5k (m, p-naphthyl) is 1-fold more potent than 5j (o, m-naphthyl) against
the two Gram-positive bacteria (S. aureus and B. subtilis). This may suggest that the branch at
para-position of the phenyl ring is more potent than that at ortho-position. For 5l – 5p, the
compounds are analogues with different length of alkyl chains extended from the ortho-position
of the phenyl ring of the substituent R2. 5l with a 4-ethylphenyl substituent showed only moderate
MICs, which are similar to that of 3a and 5j. However, when increasing the alkyl chain length
(5m: n = 4), the antibacterial activities against Gram-positive bacteria were found to improve
drastically, MICs against S. aureus and B. subtilis were both lowered to 6 g/mL. Interestingly,
5m also showed good antibacterial activity against E. coli (12 g/mL).
An attempt to further enhance the antibacterial activities by increasing the alkyl chain
length from butyl to hexyl and octyl group, however, did not result in further improvement of the
MIC against S. aureus. These compounds even showed weaker activities against E. coli (MIC =
6

48 g/mL for 5n and > 96 g/mL for 5o compared to 12 g/mL for 5m) and the trend is obviously
different from that of S. aureus. Similar observation has also been found for MoeA, which shows
excellent antibacterial activities against S. aureus and B. subtilis but no observable activity against
E. coli because the high molecular mass MoeA cannot pass through the periplasmic region of the
Gram negative bacteria [27]. We therefore postulated that the highly hydrophobic compounds 5n
and 5o may behave in a similar manner as MoeA in that they stay mainly within the outer
membrane rather than diffusing into the periplasmic space. On the other hand, the much smaller
periplasmic space between the inner membrane and the peptidoglycan layer in Gram-positive
bacteria (S. aureus and B. subtilis) represents a much lower hindrance to the entrance of
hydrophobic compounds. Based on the isatin-based compounds examined in this report, a suitable
cLogP range of 2.3 - 4.4 was found for the PGT inhibitors.
7

Table 1. MIC values (g/mL) and averaged cLogP values of the small molecular inhibitors.
S. aureus
S. aureus
B. subtilis
E. coli
Averaged
cLogP
Compound
R group
(ATCC 29213)
(ATCC BAA-41)
168
(ATCC 25922)
O
3a
48
48
24
12
48
48
24
12
24
24
12
12
96
>96
48
-1.90
1.53
1.94
1.85
N
H
NO2
O
O
O
3b
3c
3d
N
H
N
H
CF3
> 96
N
H
OH
OH
5a
5b
5c
5d
5e
5f
> 96
> 96
> 96
> 96
96
> 96
> 96
> 96
> 96
> 96
> 96
96
> 96
> 96
> 96
24
> 96
> 96
> 96
> 96
96
1.08
1.10
1.76
1.64
1.94
2.00
OH
O
O
O
> 96
96
> 96
8

O
5g
5h
O
> 96
96
> 96
96
96
48
> 96
> 96
1.59
2.39
O
5i
24
48
12
> 96
2.49
O
5j
48
24
48
24
24
12
48
48
2.38
2.36
5k
5l
5m
5n
5o
48
6
48
6
12
6
48
12
2.20
2.97
3.69
4.40
2.90
6
6
6
48
6
6
6
> 96
48
5p
6
6
3
< 0.75
> 96
1.5
MoeA
Not available
Not available
< 0.75
< 0.75
> 96
6
Not determined
Not determined
Ampicillin
< 0.75
9

2.3 Lipid II-based in vitro transglycosylation assay
To study the inhibitory effects of the compounds against PGT, fluorescent lipid II-based
transglycosylation inhibition assays were conducted using purified E. coli PBP1b [28] as the
targeting enzyme. The reaction rate of transglycosylation was fast in Triton X-100 buffered
solution and about 35 % lipid II (substrate) was consumed within 5 min (Fig. S3). Five inhibitors
(5d, 5e, 5l, 5m and 5o) were selected for the inhibitory assays (Table 2). Interestingly, the IC50 of
the compounds are highly correlated to their MIC values.
5d has an extraordinary high IC50 of 2156.1 M, indicated that it is virtually not a PGT
inhibitor. This result is also consistent to its MIC values of > 96 g/mL obtained against 3 selected
bacteria species, meaning that it has no observable antibacterial activities. A similar analogue 5e,
however, showed a more than 10-fold decrease in IC50 (203.4 M). The difference of in vitro
activities is reflected in the MIC value of B. subtilis, which was lowered to 24 g/mL. 5e is
practically not effective against S. aureus and E. coli. 5l, 5m and 5o are analogues with non-polar
substituents of different alkyl chain length (n = 2, 4 and 8, respectively). The IC50 of 5l is 125.2
M which is slightly lower than that of 5e, at the same time its MIC values showed a 1-fold
decrease against the 3 selected bacteria species; the IC50 of 5m is reduced to 82.8 M and its MICs
were also significantly lowered to 6 mg/mL for S. aureus and B. subtilis and 12 g/mL for E. coli
respectively. These results support that the inhibition effect of the compounds on PGT is closely
related to their antibacterial activities.
Increase in the alkyl chain length of the compounds in general enhances the PGT inhibitory
activity. Although 5o has the lowest IC50 (12.4 M) among the compounds, its antibacterial
activities against Gram-positive bacteria did not further improve compared with that of 5m. On
the other hand, the MIC value against Gram-negative E. coli dropped significantly (> 96 g/mL).
This inconsistency might be explained by its too high hydrophobicity (cLogP = 4.4). Its diffusion
to the PGT target is hindered by the inefficiency to pass through the periplasmic region of the
Gram-negative bacteria. As a result, the antibacterial activity against E. coli is compromised
regardless of its stronger PGT inhibition ability. The cytotoxicity of two representative
compounds 3d and 5m were also evaluated against human fibroblast cells (Detroit 551). Their
IC50 values were found to be 2.2 and 1.9 g/mL, respectively.
10

Table 2. In-vitro lipid II transglycosylation inhibitory constants (IC50) against E. coli PBP1b of
some compounds and their MICs.
MIC (g/mL)
B. subtilis
Compound
IC50 (M)
S. aureus
(ATCC BAA-41)
> 96
E. coli
(ATCC 25922)
168
5d
2156.1
203.4
125.2
82.8
> 96
24
> 96
96
5e
96
48
6
5l
12
48
5m
5o
6
12
12.4
6
6
> 96
N.D.
Vancomycin*
1.01
N.D.
N.D.
*Positive control with a reported IC50 value of 0.38 M using a modified lipid II substrate [29].
2.4 Molecular modelling
From the results obtained, we reason that the small amphiphilic inhibitor 5m is able to
mimic certain properties of MoeA, in which the hydrophilic unit could bind to the active-site
pocket of PGT whereas the alkyl chain could interact with the hydrophobic transmembrane region
of the enzyme. The binding pose between E. coli PBP1b and 5m was thus investigated by
molecular docking using AutoDock Vina [30]. One of the interaction modes depicted in Fig. 3
showed that the hydrophilic guanidyl moiety (shown in blue-violet) is able to bind to the active
site pocket of PGT, the same region occupied by the phosphate group of MoeA (shown in cyan),
the well-known inhibitor of PGT. In addition, the hydrophobic unit of 5m (4-butylphenyl group)
points to the direction of the transmembrane domain of PBP1b. Detailed interaction between 5m
and individual residues of PGT are shown in Fig. S4.
11

Fig. 3. Binding pose between MoeA (skeleton shown in cyan), 5m (skeleton shown in yellow)
and E. coli PBP1b model. The hydrophobic regions and hydrophilic regions of the PBP1b model
were indicated in red and grey, respectively. One of the best interaction modes of 5m bound with
the hydrophilic active-site pocket (indicated in grey) of PGT was calculated (see Fig. S4 for
details). The guanidyl moiety of 5m overlapped with the phosphate moiety of MoeA, suggested
that they competed for the same active site pocket. The hydrophobic moiety of 5m was pointing
downward and in similar direction as transmembrane domain.
3. Conclusion
We demonstrated an amphiphilic approach to design potent inhibitors targeting at PGT.
Our results showed that using guanidine-tagged isatin as the core scaffold accompanied by a
hydrophobic group can lead to PGT inhibitors with broad-spectrum antibacterial activities against
both Gram-positive and Gram-negative bacteria. The antibacterial activities of the amphiphilic
inhibitors can be tuned by adjusting the overall hydrophobicity of the molecules. MICs of 6 g/mL
for MSSA, MRSA and B. subtilis and 12 g/mL for E. coli (12 g/mL) were obtained with an
inhibitor 5m of averaged cLogP value 2.97. Our findings suggest that an effective inhibitor of
PGT should be amphiphilic in nature consisting of a hydrophilic moiety to interact with the active
site of the enzyme and a hydrophobic group to interact with the transmembrane region of the
enzyme as well as the plasma membrane which will assist the inhibitor molecule to approach the
active-site pocket of PGT. Nevertheless, in order to maintain bioavailability, the hydrophobicity
of the inhibitor molecule should not be too high to prevent it from entering the bacterial cell. The
12

molecular design of an effective PGT inhibitor should be a well balance of the hydrophilicity and
hydrophobicity of the amphiphilic molecule.
4. Acknowledgments
This work was supported by the General Research Fund of the Hong Kong Research
Grants Council (PolyU 153338/16P), the Innovation and Technology Commission, the Ministry
of Science and Technology of China, and The Hong Kong Polytechnic University. We gratefully
acknowledged the support of the University Research Facilities on Life Sciences and Chemical
and Environmental Analysis of The Hong Kong Polytechnic University. KYW acknowledged the
support from the Patrick S.C. Poon endowed professorship.
5. Experimental
5.1
Chemicals
Phenylboric acid derivatives, isatin, 5-methylisatin and aminoguanidine HCl were purchased from
Aldrich. Dansyl-chloride was purchased from Sigma-Aldrich. Cell culture media were purchased
from Oxoid; bacteria strains were purchased from American Type Culture Collection (ATCC).
Moenomycin A (MoeA) was purified from flavomycin purchased from International Laboratory
as described [31].
5.2
Synthetic procedures for the new isatin-based inhibitors
Compounds 3a – 3d were obtained by following the reported procedures [19].
General procedures for the synthesis of compound 5a – 5p: To a mixture of phenylboronic acid
(0.30 mmol), isatin (0.33 mmol), copper(II) acetate (0.30 mmol) and triethylamine (125 L) in
dichloromethane (10 mL), dried molecular sieve (4Ǻ) was added. The mixture was then allowed
to react with stirring for 60 h at room temperature. Excessive isatin was removed by silica gel
chromatography using a mobile phase of ethyl acetate: petroleum ether = 1:4. Some vacuum dried
product (0.15 mmol) was further reacted with aminoguanidine HCl salt (0.27 mmol) with acetic
acid (20 mL) as the solvent. The reaction mixture was allowed to stir for 2 h at room temperature.
The crude products were precipitated by the addition of diethyl ether (200 mL) into the reaction
mixture and were collected by filtration. The pure product was isolated as a yellowish solid after
the purification with silica gel chromatography using a mobile phase of methanol:
dichloromethane = 1: 4.
13

5.2.1 (E)-2-(3-(2-carbamimidoylhydrazineylidene)-5-methyl-2-oxoindolin-1-yl)-N-(3-
nitrophenyl)acetamide (3a): The reaction was taken place for 23 h. Isolated yield 52 %. 1H NMR
(400 MHz, CD3OD) : 2.40 (s, 3H, Ar-CH3), 4.73 (s, 2H), 6.97 (d, J = 8.1 Hz, 1H), 7.31 (d, J =
7.6 Hz, 1H), 7.59 (t, J = 8.2 Hz, 1H), 7.66 (s, 1H), 7.92 (d, J = 8.1 Hz, 1H), 8.02 (d, J = 8.1 Hz,
1H), 8.61 (s, 1H). 13C NMR (125 MHz, CD3OD) δ: 21.01, 30.74, 44.00, 110.76, 115.50, 119.88,
120.40, 123.28, 126.63, 131.10, 134.02, 134.88, 138.66, 140.63, 142.83, 149.88, 162.79, 167.60.
HRMS calculated for C18H18N7O4: m/z: 396.1415 [M+H]+, found 396.1405.
5.2.2 (E)-2-(3-(2-carbamimidoylhydrazineylidene)-2-oxoindolin-1-yl)-N-(3-
ethylphenyl)acetamide (3b): The reaction was taken place for 23 h. Isolated yield 56 %. 1H NMR
(400 MHz, CD3OD) : 1.23 (t, J = 7.6 Hz, 3H), 2.64 (q, J = 7.6 Hz, 2H), 4.71 (s, 2 H), 7.00 (d, J
= 8.1 Hz, 1H), 7.09 (d, J = 8.0 Hz, 1H), 7.23 (q, J = 7.5 Hz, 2H), 7.38 - 7.43 (m, 2H), 7.51 (t, J =
7.9, 1H), 7.82 (d, J = 7.5 Hz, 1H). 13C NMR (125 MHz, CD3OD) δ: 16.05, 29.86, 43.60, 111.01,
112.31, 118.62, 120.64, 122.97, 124.12, 124.82, 125.28, 129.90, 133.76, 139.21, 144.57, 146.49,
157.67, 162.80, 166.82. HRMS calculated for C19H21N6O2: m/z: 365.1721 [M+H]+; found
365.1732.
5.2.3 (E)-2-(3-(2-carbamimidoylhydrazineylidene)-2-oxoindolin-1-yl)-N-(3-
(trifluoromethyl)phenyl)acetamide (3c): The reaction was taken place for 23 h. Isolated yield 61
1
%. H NMR (400 MHz, CD3OD) : 4.75 (s, 2H), 7.10 (d, J = 8.0 Hz, 1H), 7.23 (t, J = 7.6 Hz,
13
1H), 7.43 – 7.45 (m, 1H), 7.49 - 7.57 (m, 2H), 7.78-7.84 (m, 2H), 8.01 (s, 1H). C NMR (125
MHz, CD3OD) δ: 43.65, 111.02, 117.49, 120.37, 121.90, 122.90, 124.24, 124.88, 130.93, 132.14,
132.40, 133.73, 138.64, 140.18, 145.19, 157.67, 162.80, 167.27. HRMS calculated for
C18H16F3N6O2: m/z: 405.1281 [M+H]+, found 405.1289.
5.2.4 (E)-2-(3-(2-carbamimidoylhydrazineylidene)-2-oxoindolin-1-yl)-N-(naphthalen-2-
1
yl)acetamide (3d): The reaction was taken place for 23 h. Isolated yield 62 %. H NMR (400
MHz, CD3OD) : 4.78 (s, 2H), 7.13 (d, J = 7.9 Hz, 1H), 7.24 (t, J = 7.6 Hz, 1H), 7.41 - 7.54 (m,
13
3H), 7.59 (d, J = 8.9, 1H), 7.77-7.87 (m, 4H), 8.20 (s, 1H). C NMR (125 MHz, CD3OD) δ:
43.68, 111.06, 118.01, 120.40, 121.00, 122.88, 124.84, 126.28, 127.63, 128.57, 128.63, 129.76,
132.29, 133.72, 135.18, 136.74, 139.86, 145.29, 147.29, 162.83, 167.07. HRMS calculated for
C21H19N6O2: m/z: 387.1564 [M+H]+, found 387.1573.
5.2.5 (E)-2-(1-(4-hydroxyphenyl)-2-oxoindolin-3-ylidene)hydrazine-1-carboximidamide (5a):
14

The reaction was taken place for 64 h. Isolated yield 32 %. 1H NMR (400 MHz, CD3OD) : 6.83
(d, J = 8.0 Hz, 1H), 7.00 (d, J = 8.8 Hz, 2H), 7.21 (t, J = 7.6 Hz, 1H), 7.28 (d, J = 8.8 Hz, 2H),
7.44 (t, J = 7.6 Hz, 1H), 7.84 (d, J = 7.2 Hz, 1H). 13C NMR (125 MHz, CD3OD) δ: 111.54, 117.40,
120.27, 122.87, 124.89, 125.45, 129.17, 133.57, 138.84, 146.60, 157.68, 159.30, 162.17. HRMS
calculated for C15H14N5O2: m/z: 296.1142 [M+H]+, found 296.1161.
5.2.5 (E)-2-(1-(4-(hydroxymethyl)phenyl)-2-oxoindolin-3-ylidene)hydrazine-1-
carboximidamide (5b): The reaction was taken place for 64 h. Isolated yield 29 %. 1H NMR (400
MHz, CD3OD) : 4.73 (s, 2H), 6.90 (d, J = 8.0 Hz, 1H,), 7.24 (t, J = 7.5 Hz, 1H), 7.43 - 7.50 (m,
3H), 7.58 - 7.62 (m, 2H), 7.86 (d, J = 7.4 Hz, 1H). 13C NMR (125 MHz, CD3OD) δ: 64.52, 111.49,
120.53, 122.92, 125.03, 127.63, 129.18, 133.22, 133.42, 138.52, 143.88, 145.92, 157.95, 161.85.
HRMS calculated for C16H16N5O2: m/z: 310.1299 [M+H]+, found 310.1301.
5.2.7 (E)-2-(1-(4-(3-hydroxypropyl)phenyl)-2-oxoindolin-3-ylidene)hydrazine-1-
carboximidamide (5c): The reaction was taken place for 64 h. Isolated yield 36 %. 1H NMR (400
MHz, CD3OD) : 1.92 (quin, J = 7.1 Hz, 2H), 2.82 (t, J = 7.6 Hz, 2H), 3.64 (t, J = 6.2 Hz, 2H),
6.90 (d, J = 8.0 Hz, 1H), 7.24 (t, J = 7.6 Hz, 1H), 7.40 - 7.49 (m, 5H), 7.86 (d, J = 7.2 Hz, 1H).
13C NMR (125 MHz, CD3OD) δ: 32.80, 35.35, 62.10, 111.53, 120.35, 123.02, 125.03, 127.60,
130.91, 131.95, 133.54, 138.65, 144.65, 146.07, 157.62, 161.92. HRMS calculated for
C18H20N5O2: m/z: 338.1612 [M+H]+, found 338.1614.
5.2.8 (E)-2-(1-(4-(methoxymethyl)phenyl)-2-oxoindolin-3-ylidene)hydrazine-1-
carboximidamide (5d): The reaction was taken place for 64 h. Isolated yield 41 %. 1H NMR (400
MHz, CD3OD) : 3.46 (s, 3H), 4.58 (s, 2H), 6.91 (d, J = 8.4 Hz, 1H), 7.24 (t, J = 7.6, 1H), 7.43 -
7.51 (m, 3H), 7.59 – 7.61 (m, 2H), 7.86 (d, J = 7.2, 1H). 13C NMR (125 MHz, CD3OD) δ: 58.59,
74.89, 111.51, 120.52, 122.96, 125.06, 127.62, 130.03, 133.44, 133.63, 138.50, 140.63, 145.84,
157.84, 161.82. HRMS calculated for C17H18N5O2: m/z: 324.1455 [M+H]+, found 324.1453.
5.2.9 (E)-2-(1-(4-ethoxyphenyl)-2-oxoindolin-3-ylidene)hydrazine-1-carboximidamide (5e):
The reaction was taken place for 64 h. Isolated yield 42 %. 1H NMR (400 MHz, CD3OD) : 1.45
(t, J = 7.0 Hz, 3H), 4.14 (q, J = 6.9 Hz, 2H), 6.84 (d, J = 8.0 Hz, 1H), 7.12 (d, J = 8.4, 2H), 7.22
(t, J = 7.6 Hz, 1H), 7.38 - 7.45 (m, 3H), 7.84 (d, J = 7.2, 1H). 13C NMR (125 MHz, CD3OD) δ:
15.08, 64.94, 111.51, 116.58, 120.34, 122.88, 124.93, 126.56, 129.11, 133.54, 138.77, 146.44,
15

157.73, 160.69, 162.11. HRMS calculated for C17H18N5O2: m/z: 324.1455 [M+H]+, found
324.1454.
5.2.10 (E)-2-(1-(4-(ethoxymethyl)phenyl)-2-oxoindolin-3-ylidene)hydrazine-1-carboximidamide
(5f): The reaction was taken place for 64 h. Isolated yield 41 %. 1H NMR (400 MHz, CD3OD) :
1.28 (t, J = 7.0 Hz, 3H), 3.64 (q, J = 7.0 Hz, 2H), 4.63 (s, 2H), 6.91 (d, J = 8.0 Hz, 1H), 7.24 (t, J
= 7.6, 1H), 7.43 - 7.51 (m, 3H), 7.59 - 7.61 (m, 2H), 7.86 (d, J = 7.8, 1H). 13C NMR (125 MHz,
CD3OD) δ: 15.48, 67.09, 72.93, 111.57, 120.42, 123.02, 125.08, 127.61, 130.00, 133.52, 133.58,
138.67, 141.04, 145.98, 157.64, 161.91. HRMS calculated for C18H20N5O2: m/z: 338.1612
[M+H]+, found 338.1614.
5.2.11 (E)-2-(1-(4-(methoxymethoxy)phenyl)-2-oxoindolin-3-ylidene)hydrazine-1-
carboximidamide (5g): The reaction was taken place for 64 h. Isolated yield 35 %. 1H NMR (400
MHz, CD3OD) : 3.52 (s, 3H), 5.29 (s, 2H), 6.86 (d, J = 8.0 Hz, 1H), 7.20 - 7.28 (m, 3H), 7.40 -
7.46 (m, 3H), 7.84 (d, J = 7.6, 1H). 13C NMR (125 MHz, CD3OD) δ: 56.37, 95.54, 114.48, 118.39,
120.37, 122.90, 124.97, 127.78, 129.08, 133.50, 138.61, 146.21, 157.79, 158.85, 162.02. HRMS
calculated for C17H18N5O3: m/z: 340.1404 [M+H]+, found 340.1407.
5.2.12 (E)-2-(1-(4-(tert-butoxymethyl)phenyl)-2-oxoindolin-3-ylidene)hydrazine-1-
carboximidamide (5h): The reaction was taken place for 64 h. Isolated yield 42 %. 1H NMR (400
MHz, CD3OD) : 1.35 (s, 9H), 4.60 (s, 2H), 6.90 (d, J = 8.0 Hz, 1H), 7.24 (t, J = 7.6 Hz, 1H),
7.43 - 7.48 (m, 3H), 7.59 – 7.61 (m, 2H), 7.86 (d, J = 7.6 Hz, 1H). 13C NMR (125 MHz, CD3OD)
δ: 26.61, 63.29, 73.82, 110.23, 119.04, 121.68, 123.74, 126.16, 128.52, 131.81, 132.25, 137.34,
140.94, 144.69, 156.28, 160.57. HRMS calculated for C20H24N5O2: m/z: 366.1925 [M+H]+, found
366.1926.
5.2.13 (E)-2-(1-(2-isobutoxyphenyl)-2-oxoindolin-3-ylidene)hydrazine-1-carboximidamide (5i):
The reaction was taken place for 64 h. Isolated yield 48 %. 1H NMR (400 MHz, CD3OD) : 0.79
[m, 6H], 1.85 [m, 1H], 3.79 [m, 2H], 6.59 (d, J = 8.0 Hz, 1H), 7.14 - 7.26 (m, 3H), 7.40 - 7.41 (m,
2H), 7.54 (t, J = 8.0 Hz, 1H), 7.84 (d, J = 7.2 Hz, 1H). 13C NMR (125 MHz, CD3OD) δ: 19.25,
19.27, 29.36, 75.81, 111.97, 114.57, 120.09, 122.13, 122.54, 122.75, 124.81, 130.66, 132.29,
133.60, 138.58, 146.58, 156.250, 157.69, 162.30. HRMS calculated for C19H22N5O2: m/z:
352.1768 [M+H]+, found 352.1768.
16

5.2.14 (E)-2-(1-(naphthalen-1-yl)-2-oxoindolin-3-ylidene)hydrazine-1-carboximidamide (5j):
The reaction was taken place for 62 h. Isolated yield 55 %. 1H NMR (400 MHz, CD3OD) : 6.42
(d, J = 7.9 Hz, 1H), 7.23 (t, J = 7.5 Hz, 1H), 7.34 (t, J = 7.4 Hz, 1H,), 7.50 - 7.71 (m, 5H), 7.92
(d, J = 7.4 Hz, 1H), 8.06 (d, J = 8.2, 1H), 8.11 (d, J = 8.2, 1H). 13C NMR (125 MHz, CD3OD) δ:
118.83, 120.46, 123.02, 123.58, 125.07, 126.97, 127.64, 128.05, 128.50, 129.86, 130.66, 131.01,
131.37, 133.67, 136.24, 138.63, 146.88, 157.74, 162.54. HRMS calculated for C19H16N5O: m/z:
330.1349 [M+H]+, found 330.1362.
5.2.15 (E)-2-(1-(naphthalen-2-yl)-2-oxoindolin-3-ylidene)hydrazine-1-carboximidamide (5k):
The reaction was taken place for 62 h. Isolated yield 50 %. 1H NMR (400 MHz, CD3OD) : 6.90
(d, J = 7.9 Hz, 1H), 7.22 (t, J = 7.5 Hz, 1H), 7.40 (t, J = 7.7 Hz, 1H), 7.51 - 7.58 (m, 3H), 7.85 -
8.04 (m, 5H). 13C NMR (125 MHz, CD3OD) δ: 111.68, 120.51, 123.03, 124.99, 125.15, 126.66,
128.11, 128.24, 128.99, 129.09, 130.89, 131.69, 133.60, 134.42, 135.02, 138.70, 146.06, 157.70,
162.06. HRMS calculated for C19H16N5O: m/z: 330.1349 [M+H]+, found 330.1364.
5.2.16 (E)-2-(1-(4-ethylphenyl)-2-oxoindolin-3-ylidene)hydrazine-1-carboximidamide (5l): The
reaction was taken place for 62 h. Isolated yield 55 %. 1H NMR (400 MHz, CD3OD) : 1.29 (t,
3H, J = 7.6 Hz), 2.74 (q, 2H, J = 7.6 Hz), 6.84 (d, J = 8.0 Hz, 1H), 7.20 (t, J = 7.5 Hz, 1H), 7.35
13
- 7.43 (m, 5H), 7.84 (d, J = 7.5, 1H). C NMR (125 MHz, CD3OD) δ: 16.08, 29.60, 111.58,
120.39, 122.95, 125.00, 127.64, 130.30, 131.85, 133.55, 138.73, 146.17, 146.69, 157.69, 161.96.
HRMS calculated for C17H18N5O: m/z: 308.1506 [M+H]+, found 308.1521.
5.2.17 (E)-2-(1-(4-butylphenyl)-2-oxoindolin-3-ylidene)hydrazine-1-carboximidamide
(5m):
The reaction was taken place for 62 h. Isolated yield 54 %. 1H NMR (400 MHz, CD3OD) : 0.97
(t, 3H, J = 7.4 Hz), 1.40 (m, 2H), 1.65 (m, 2H), 2.70 (t, J = 7.5 Hz, 2H), 6.83 (d, J = 8.0 Hz, 1H),
13
7.20 (t, J = 7.6 Hz, 1H), 7.33 - 7.41 (m, 5H), 7.83 (d, J = 7.5 Hz, 1H). C NMR (125 MHz,
CD3OD) δ: 14.10, 23.16, 34.59, 36.10, 111.31, 120.18, 122.79, 124.83, 127.22, 130.57, 131.57,
133.20, 138.14, 144.90, 145.65, 157.53, 161.55. HRMS calculated for C19H22N5O: m/z: 336.1819
[M+H]+, found 336.1830.
5.2.18 (E)-2-(1-(4-hexylphenyl)-2-oxoindolin-3-ylidene)hydrazine-1-carboximidamide
(5n):
The reaction was taken place for 62 h. Isolated yield 52 %. 1H NMR (400 MHz, CD3OD) : 0.94
(t, 3H, J = 6.9 Hz), 1.38 (m, 6H), 1.70 (m, 2H), 2.74 (t, J = 7.7 Hz, 2H), 6.89 (d, J = 8.0 Hz, 1H),
7.23 (t, J = 7.5 Hz, 1H), 7.38-7.46 (m, 5H), 7.85 (d, J = 7.5, 1H). 13C NMR (125 MHz, CD3OD)
17

δ: 14.42, 23.69, 30.04, 32.62, 32.88, 36.62, 111.36, 122.54, 124.02, 124.80, 127.59, 130.81,
132.10, 133.48, 138.12, 143.41, 144.45, 145.13, 161.63. HRMS calculated for C21H26N5O: m/z:
364.2132 [M+H]+, found 364.2145.
5.2.19 (E)-2-(1-(4-octylphenyl)-2-oxoindolin-3-ylidene)hydrazine-1-carboximidamide (5o): The
reaction was taken place for 62 h. Isolated yield 46 %. 1H NMR (400 MHz, CD3OD) : 0.90 (t,
3H, J = 6.7 Hz), 1.30 - 1.36 (m, 10H), 1.68 (m, 2H), 2.71 (t, J = 7.6, 2H), 6.85 (d, J = 8.0 Hz, 1H),
7.20 (t, J = 7.5 Hz, 1H), 7.38 - 7.42 (m, 5H), 7.83 (d, J = 7.4, 1H). 13C NMR (125 MHz, CD3OD)
δ: 14.44, 23.72, 30.35, 30.41, 30.58, 32.64, 33.04, 36.61, 111.59, 120.36, 122.97, 125.01, 127.52,
130.85, 131.83, 133.57, 138.74, 145.29, 146.16, 157.63, 161.96. HRMS calculated for
C23H30N5O: m/z: 392.2445 [M+H]+, found 392.2465.
5.2.20 (E)-2-(1-(4-(sec-butyl)phenyl)-2-oxoindolin-3-ylidene)hydrazine-1-carboximidamide
(5p): The reaction was taken place for 62 h. Isolated yield 56 %. 1H NMR (400 MHz, CD3OD) :
0.90 (t, J = 7.4 Hz, 3H), 1.32 (d, J = 7.2 Hz, 3H), 1.70 (m, 2H), 2.74 (m, 1H), 6.90 (d, J = 8.0 Hz,
1H), 7.23 (t, J = 7.6 Hz, 1H), 7.41 - 7.47 (m, 5H), 7.86 (d, J = 7.2 Hz, 1H). 13C NMR (125 MHz,
CD3OD) δ: 12.58, 22.32, 32.16, 42.84, 111.55, 120.44, 122.92, 124.98, 127.55, 129.49, 131.98,
133.46, 138.59, 146.04, 149.94, 157.80, 161.89. HRMS calculated for C19H22N5O: m/z: 366.1819
[M+H]+, found 336.1814.
5.3
Antimicrobial susceptible testing
To prepare fresh bacteria media, frozen stock bacterial strains were streaked on a TSA plate and
incubated at 37 oC overnight. A single colony was picked and added to corresponding media and
o
grew at 37 C for 5 h at 250 rpm. Working bacterial culture solutions were prepared by further
diluting 2 L of the grown media into 5 mL fresh media. The working bacterial culture solutions
were used immediately. Minimum inhibitory concentrations (MICs) were conducted using broth
micro-dilution method as described [19]. Three bacterial strains were used in this experiment:
MSSA (ATCC 29213), MRSA (ATCC BAA-41), B. subtilis 168 and E. coli (ATCC 25922). MHB
was the medium for B. subtilis and E. coli, CA-MHB was the medium for S. aureus.
5.4
Over-expression and purification of E. coli PBP1b
E. coli PBP1b (residues 58 - 604) was expressed and purified as described [5] without the second
purification step (size exclusion). The (His)6-tag was retained after purification. The final protein
concentration was measured by Bradford assay using bovine serum albumin as reference protein.
18

5.5
Lipid II-based in vitro transglycosylation assay
Lipid II was prepared as described [32]. Dansyl fluorophore was attached to lipid II for
spectroscopic detection (see structure and mass spectrum of dansyl-lipid II in Fig. S5) [33]. E.
o
coli PBP1b and compounds were pre-incubated at 30 C for 15 min in reaction buffer (50 mM
HEPES pH 7.5, 200 mM NaCl, 10 mM CaCl2) with additional 0.04 % Triton X-100, 25 % DMSO
and 1 U/L muramidase. Dansyl-lipid II (working concentration 2 M) was then added into
reaction mixture and mixed well to initiate the reaction. The reaction was allowed to stand for 5
min at 30 oC, and MoeA (working concentration 15 M) was added to quench the reaction. The
o
mixtures were stored at -80 C before HPLC analysis as described [33, 34]. 20 L of samples
were injected into an Agilent 1200 series equipped with a Supelco SAX1 anion-exchange column
and eluted using a gradient of NH4OAc/MeOH (20 mM to 500 mM). Half maximal inhibitory
concentrations (IC50) of selected compounds were calculated by GraphPad Prism 7 with the model:
Y=1/(1+10^((LogIC50-X)*h)) (Y, normalized response; X, log[inhibitor]; h, hillslope). The Dose-
response curve for in vitro transglycosylation inhibition assays were shown in Fig. S2. The isatin-
based compounds were measured in triplicate while vancomycin (positive control) was measured
in duplicate, and half maximal inhibitory concentrations (IC50) were calculated as described [16].
HPLC chromatograms of in vitro transglycosylation assay of 5m were shown in Fig. S1.
5.6
Molecular modelling
The docking study was using a E. coli PBP1b model built based on a protein crystal structure of
E. coli PBP1b (PBP ID: 3VMA [5]) with the missing residues 241-246 and 252-266 added by
Modeller [35] and Coot with reference to other PGT crystal structures (PBP ID: 3FWL [5], 3NB6
[36], 2OQO [37] and 6FTB [38]). The missing tail of MoeA in 3VMA was replenished based on
the MoeA ligand from 6FTB, and was adjusted to point into the membrane (cyan, Fig. 3 of the
main text).
Molecular model of the PGT domain of E. coli PBP1b in the presence of 5m was performed using
AutoDock Vina [30]. The structure of 5m was drawn and 3D coordinates were generated by
ChemDraw Ultra 8.0. Energy of 5m was minimized by Avogadro with Merck Molecular Force
Field 94 (MMFF94) [39]. PBP1b and 5m were treated by reported procedures [40], the interaction
between the hydrocarbon moiety and membrane did not count. The best mode with 5m bound to
active-site of PGT domain was selected for reporting. Results were shown in Fig. 3 and Fig. S4.
5.7
Cytotoxicity
19

Human fibroblast cells (Detroit 551) (American Type Culture Collection) were grown in Eagle’s
Minimum Essential Medium with 10% fetal bovine serum at 37°C in a humidified atmosphere of
5% CO2 in the air. Cells were seeded at ~5000 cells per well in 96-well plates and allowed to
settle overnight. The compounds were diluted into the growth medium with stated concentrations
and the plates were incubated for 3 days. Quantitative cell proliferation assays were performed
using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) reagent
according to the instructions provided by the manufacturer. All experiments were performed in
triplicate and the results are the average of three independent experiments.
References
[1] WHO report on surveillance of antibiotic consumption: 2016-2018 early implementation,
(2018).
[2] J. O’Neill, Tackling drug-resistant infections globally: final report and recommendations,
The Review on Antimicrobial Resistance, (2016).
[3] B. Ostash, S. Walker, Bacterial transglycosylase inhibitors, Curr Opin Chem Biol, 9 (2005)
459-466.
[4] C.Y. Huang, H.W. Shih, L.Y. Lin, Y.W. Tien, T.J. Cheng, W.C. Cheng, C.H. Wong, C. Ma,
Crystal structure of Staphylococcus aureus transglycosylase in complex with a lipid II analog
and elucidation of peptidoglycan synthesis mechanism, Proc Natl Acad Sci U S A, 109 (2012)
6496-6501.
[5] M.T. Sung, Y.T. Lai, C.Y. Huang, L.Y. Chou, H.W. Shih, W.C. Cheng, C.H. Wong, C. Ma,
Crystal structure of the membrane-bound bifunctional transglycosylase PBP1b from Escherichia
coli, Proc Natl Acad Sci U S A, 106 (2009) 8824-8829.
[6] A.L. Lovering, L.H. de Castro, D. Lim, N.C.J. Strynadka, Structural insight into the
transglycosylation step of bacterial cell-wall biosynthesis, Science, 315 (2007) 1402-1405.
[7] X.Y. Ye, M.C. Lo, L. Brunner, D. Walker, D. Kahne, S. Walker, Better substrates for
bacterial transglycosylases, J Am Chem Soc, 123 (2001) 3155-3156.
[8] L.Y. Huang, S.H. Huang, Y.C. Chang, W.C. Cheng, T.J. Cheng, C.H. Wong, Enzymatic
synthesis of lipid II and analogues, Angew Chem Int Ed Engl, 53 (2014) 8060-8065.
[9] Y. Qiao, V. Srisuknimit, F. Rubino, K. Schaefer, N. Ruiz, S. Walker, D. Kahne, Lipid II
overproduction allows direct assay of transpeptidase inhibition by beta-lactams, Nat Chem Biol,
13 (2017) 793-798.
[10] J.G. Taylor, X. Li, M. Oberthur, W. Zhu, D.E. Kahne, The total synthesis of moenomycin
A, J Am Chem Soc, 128 (2006) 15084-15085.
[11] Y. Rebets, T. Lupoli, Y. Qiao, K. Schirner, R. Villet, D. Hooper, D. Kahne, S. Walker,
Moenomycin resistance mutations in Staphylococcus aureus reduce peptidoglycan chain length
and cause aberrant cell division, ACS Chem Biol, 9 (2014) 459-467.
[12] B. Ostash, S. Walker, Moenomycin family antibiotics: chemical synthesis, biosynthesis,
and biological activity, Nat Prod Rep, 27 (2010) 1594-1617.
[13] T.J.R. Cheng, M.T. Sung, H.Y. Liao, Y.F. Chang, C.W. Chen, C.Y. Huang, L.Y. Chou,
Y.D. Wu, Y. Chen, Y.S.E. Cheng, C.H. Wong, C. Ma, W.C. Cheng, Domain requirement of
moenomycin binding to bifunctional transglycosylases and development of high-throughput
discovery of antibiotics, P Natl Acad Sci USA, 105 (2008) 431-436.
[14] C.M. Gampe, H. Tsukamoto, E.H. Doud, S. Walker, D. Kahne, Tuning the moenomycin
pharmacophore to enable discovery of bacterial cell wall synthesis inhibitors, J Am Chem Soc,
135 (2013) 3776-3779.
20

[15] J. Zuegg, C. Muldoon, G. Adamson, D. McKeveney, G. Le Thanh, R. Premraj, B. Becker,
M. Cheng, A.G. Elliott, J.X. Huang, M.S. Butler, M. Bajaj, J. Seifert, L. Singh, N.F. Galley, D.I.
Roper, A.J. Lloyd, C.G. Dowson, T.J. Cheng, W.C. Cheng, D. Demon, E. Meyer, W.
Meutermans, M.A. Cooper, Carbohydrate scaffolds as glycosyltransferase inhibitors with in
vivo antibacterial activity, Nat Commun, 6 (2015) 7719.
[16] I. Sosic, M. Anderluh, M. Sova, M. Gobec, I.M. Rascan, A. Derouaux, A. Amoroso, M.
Terrak, E. Breukink, S. Gobec, Structure-Activity Relationships of Novel Tryptamine-Based
Inhibitors of Bacterial Transglycosylase, J Med Chem, 58 (2015) 9712-9721.
[17] A. Medvedev, O. Buneeva, V. Glover, Biological targets for isatin and its analogues:
Implications for therapy, Biologics, 1 (2007) 151-162.
[18] S.K. Sridhar, M. Saravanan, A. Ramesh, Synthesis and antibacterial screening of
hydrazones, Schiff and Mannich bases of isatin derivatives, Eur J Med Chem, 36 (2001) 615-
625.
[19] Y. Wang, F.Y. Chan, N. Sun, H.K. Lui, P.K. So, S.C. Yan, K.F. Chan, J. Chiou, S. Chen,
R. Abagyan, Y.C. Leung, K.Y. Wong, Structure-based design, synthesis, and biological
evaluation of isatin derivatives as potential glycosyltransferase inhibitors, Chem Biol Drug Des,
84 (2014) 685-696.
[20] D.M.T. Chan, K.L. Monaco, R.P. Wang, M.P. Winters, New N- and O-arylations with
phenylboronic acids and cupric acetate, Tetrahedron Lett, 39 (1998) 2933-2936.
[21] D.A. Evans, J.L. Katz, T.R. West, Synthesis of diaryl ethers through the copper-promoted
arylation of phenols with arylboronic acids. An expedient synthesis of thyroxine, Tetrahedron
Lett, 39 (1998) 2937-2940.
[22] P.Y.S. Lam, C.G. Clark, S. Saubern, J. Adams, M.P. Winters, D.M.T. Chan, A. Combs,
New aryl/heteroaryl C-N bond cross-coupling reactions via arylboronic acid cupric acetate
arylation, Tetrahedron Lett, 39 (1998) 2941-2944.
[23] A. Daina, O. Michielin, V. Zoete, iLOGP: a simple, robust, and efficient description of n-
octanol/water partition coefficient for drug design using the GB/SA approach, J Chem Inf
Model, 54 (2014) 3284-3301.
[24] T. Cheng, Y. Zhao, X. Li, F. Lin, Y. Xu, X. Zhang, Y. Li, R. Wang, L. Lai, Computation of
octanol-water partition coefficients by guiding an additive model with knowledge, J Chem Inf
Model, 47 (2007) 2140-2148.
[25] S.A. Wildman, G.M. Crippen, Prediction of physicochemical parameters by atomic
contributions, J Chem Inf Comp Sci, 39 (1999) 868-873.
[26] I. Moriguchi, S. Hirono, Q. Liu, I. Nakagome, Y. Matsushita, Simple Method of
Calculating Octanol Water Partition-Coefficient, Chem Pharm Bull, 40 (1992) 127-130.
[27] E.R. Baizman, A.A. Branstrom, C.B. Longley, N. Allanson, M.J. Sofia, D. Gange, R.C.
Goldman, Antibacterial activity of synthetic analogues based on the disaccharide structure of
moenomycin, an inhibitor of bacterial transglycosylase, Microbiol-Uk, 146 (2000) 3129-3140.
[28] W.L. Cheong, M.S. Tsang, P.K. So, W.H. Chung, Y.C. Leung, P.H. Chan, Fluorescent
TEM-1 beta-lactamase with wild-type activity as a rapid drug sensor for in vitro drug screening,
Bioscience Rep, 34 (2014) 523-533.
[29] L. Chen, D. Walker, B. Sun, Y. Hu, S. Walker, D. Kahne, Vancomycin analogues active
against vanA-resistant strains inhibit bacterial transglycosylase without binding substrate, Proc
Natl Acad Sci U S A, 100 (2003) 5658-5663.
[30] O. Trott, A.J. Olson, AutoDock Vina: improving the speed and accuracy of docking with a
new scoring function, efficient optimization, and multithreading, J Comput Chem, 31 (2010)
455-461.
[31] M. Adachi, Y. Zhang, C. Leimkuhler, B. Sun, J.V. LaTour, D.E. Kahne, Degradation and
reconstruction of moenomycin A and derivatives: dissecting the function of the isoprenoid
chain, J Am Chem Soc, 128 (2006) 14012-14013.
21

[32] E. Breukink, H.E. van Heusden, P.J. Vollmerhaus, E. Swiezewska, L. Brunner, S. Walker,
A.J.R. Heck, B. de Kruijff, Lipid II is an intrinsic component of the pore induced by nisin in
bacterial membranes, J Biol Chem, 278 (2003) 19898-19903.
[33] B. Schwartz, J.A. Markwalder, S.P. Seitz, Y. Wang, R.L. Stein, A kinetic characterization
of the glycosyltransferase activity of Eschericia coli PBP1b and development of a continuous
fluorescence assay, Biochemistry-Us, 41 (2002) 12552-12561.
[34] C.Y. Liu, C.W. Guo, Y.F. Chang, J.T. Wang, H.W. Shih, Y.F. Hsu, C.W. Chen, S.K. Chen,
Y.C. Wang, T.J. Cheng, C. Ma, C.H. Wong, J.M. Fang, W.C. Cheng, Synthesis and evaluation
of a new fluorescent transglycosylase substrate: lipid II-based molecule possessing a dansyl-C20
polyprenyl moiety, Org Lett, 12 (2010) 1608-1611.
[35] B. Webb, A. Sali, Comparative Protein Structure Modeling Using MODELLER, Curr
Protoc Bioinformatics, 54 (2016) 5 6 1-5 6 37.
[36] S. Fuse, H. Tsukamoto, Y. Yuan, T.S. Wang, Y. Zhang, M. Bolla, S. Walker, P. Sliz, D.
Kahne, Functional and structural analysis of a key region of the cell wall inhibitor moenomycin,
ACS Chem Biol, 5 (2010) 701-711.
[37] Y. Yuan, D. Barrett, Y. Zhang, D. Kahne, P. Sliz, S. Walker, Crystal structure of a
peptidoglycan glycosyltransferase suggests a model for processive glycan chain synthesis, Proc
Natl Acad Sci U S A, 104 (2007) 5348-5353.
[38] A.S. Punekar, F. Samsudin, A.J. Lloyd, C.G. Dowson, D.J. Scott, S. Khalid, D.I. Roper,
The role of the jaw subdomain of peptidoglycan glycosyltransferases for lipid II polymerization,
Cell Surf, 2 (2018) 54-66.
[39] M.D. Hanwell, D.E. Curtis, D.C. Lonie, T. Vandermeersch, E. Zurek, G.R. Hutchison,
Avogadro: an advanced semantic chemical editor, visualization, and analysis platform, J
Cheminform, 4 (2012) 17.
[40] G.M. Morris, R. Huey, W. Lindstrom, M.F. Sanner, R.K. Belew, D.S. Goodsell, A.J. Olson,
AutoDock4 and AutoDockTools4: Automated docking with selective receptor flexibility, J
Comput Chem, 30 (2009) 2785-2791.
22

Highlights
. New amphiphilic isatin-based inhibitors targeting bacterial peptidoglycan
glycosyltransferase were synthesized and investigated.
. The hydrophobicity of the inhibitors showed significant influence on the inhibition potency
to the enzyme and antibacterial activity.
. The structure-activity relationship of the new isatin-based inhibitors was evaluated.
. The isatin-based inhibitors showed high potential to be developed as a broad-spectrum
antibiotic.
23

24

Declaration of interests
The authors declare that they have no known competing financial interests or personal relationships
that could have appeared to influence the work reported in this paper.
☐ The authors declare the following financial interests/personal relationships which may be considered
as potential competing interests:
25