Discovery of natural berberine-derived nitroimidazoles as potentially multi-targeting agents against drug-resistant Escherichia coli

Article abstract of DOI:10.1007/s11426-017-9169-4

A series of natural berberine-derived nitroimidazoles as novel antibacterial agents were designed, synthesized and characterized by nuclear magnetic resonance (NMR), infrared spectra (IR), and high resolution mass spectra (HRMS) spectra. The antimicrobial evaluation showed that some target molecules exhibited moderate to good inhibitory activities against the tested bacteria and fungi including clinical drug-resistant strains isolated from infected patients. Especially, 2-fluorobenzyl derivative 8f not only gave strong activity against drug-resistant E. coli with the minimal inhibitory concentration (MIC) value of 0.003 mM, 33-fold more active than norfloxacin, but also exhibited low toxicity toward RAW 264.7 cells and less propensity to trigger resistance. The aqueous solubility and ClogP values of target compounds were investigated to elucidate the structureactivity relationships. Molecular docking and quantum chemical studies for compound 8f rationally explained its antibacterial effect. The further exploration of antibacterial mechanism revealed that the highly active compound 8f could effectively permeabilize E. coli cell membrane and intercalate into DNA isolated from resistant E. coli to form 8f-DNA complex that might block DNA replication to exert the powerful bioactivities. Compound 8f could also selectively address resistant E. coli from a mixture of various strains.

Full text of DOI:10.1007/s11426-017-9169-4

SCIENCE CHINA
Chemistry
•ARTICLES• . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . May 2018 Vol.61 No. 5: 557–568
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .https://doi.org/10.1007/s11426-017-9169-4
Discovery of natural berberine-derived nitroimidazoles as
potentially multi-targeting agents against drug-resistant
Escherichia coli
Guo-Biao Zhang, Swetha Kameswari Maddili, Vijai Kumar Reddy Tangadanchu,
Lavanya Gopala, Wei-Wei Gao, Gui-Xin Cai* & Cheng-He Zhou*
Institute of Bioorganic & Medicinal Chemistry, Key Laboratory of Applied Chemistry of Chongqing Municipality, School of Chemistry and
Chemical Engineering, Southwest University, Chongqing 400715, China
Received September 26, 2017; accepted November 1, 2017; published online December 28, 2017
A series of natural berberine-derived nitroimidazoles as novel antibacterial agents were designed, synthesized and characterized
by nuclear magnetic resonance (NMR), infrared spectra (IR), and high resolution mass spectra (HRMS) spectra. The anti-
microbial evaluation showed that some target molecules exhibited moderate to good inhibitory activities against the tested
bacteria and fungi including clinical drug-resistant strains isolated from infected patients. Especially, 2-fluorobenzyl derivative
8f not only gave strong activity against drug-resistant E. coli with the minimal inhibitory concentration (MIC) value of
0.003 mM, 33-fold more active than norfloxacin, but also exhibited low toxicity toward RAW 264.7 cells and less propensity to
trigger resistance. The aqueous solubility and ClogP values of target compounds were investigated to elucidate the structure-
activity relationships. Molecular docking and quantum chemical studies for compound 8f rationally explained its antibacterial
effect. The further exploration of antibacterial mechanism revealed that the highly active compound 8f could effectively
permeabilize E. coli cell membrane and intercalate into DNA isolated from resistant E. coli to form 8f-DNA complex that might
block DNA replication to exert the powerful bioactivities. Compound 8f could also selectively address resistant E. coli from a
mixture of various strains.
berberine, nitroimidazole, Escherichia coli, DNA, membrane permeabilization
Citation:
Zhang GB, Maddili SK, Tangadanchu VKR, Gopala L, Gao WW, Cai GX, Zhou CH. Discovery of natural berberine-derived nitroimidazoles as
potentially multi-targeting agents against drug-resistant Escherichia coli. Sci China Chem, 2018, 61: 557–568, https://doi.org/10.1007/s11426-017-
9169-4
1 Introduction
bilizing resistance elements in microbial populations, which
directly led to the emergence and spread of resistant strains
like methicillin-resistant Staphylococcus aureus (MRSA),
vancomycin-resistant Enterococcus faecium (VRE), and
fluoroquinolone-resistant Escherichia coli (E. coli) [2].
Seriously, E. coli has been extensively found to have the
recently emerged and globally disseminated resistance to
various clinical antibacterial drugs such as fluoroquinolones,
aminoglycosides, trimethoprim-sulfamethoxazoles, and car-
bapenems due to the overexpression of multidrug efflux
pumps and the enhanced function of cell membrane. The
The emergence of antibiotics brought a vital revolution to-
ward previously fatal microbial infections, and the discovery
of antibacterial sulfonamides and penicillins further spurred
the development of additional classes of antimicrobial
agents, which successfully kept infectious diseases in control
[1]. However, the misuse and inappropriate use of various
antibiotics have created unprecedented conditions for mo-
*Corresponding authors (email: gxcai@swu.edu.cn; zhouch@swu.edu.cn)
© Science China Press and Springer-Verlag GmbH Germany 2017 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . chem.scichina.com link.springer.com

558 . . . . . . . . . . . . . . . . . . . . Zhang et al. Sci China Chem May (2018) Vol.61 No.5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 558
drug-resistant E. coli as an increasing menace could lead to
urinary tract infection, cause injury of ureter contractility and
form intracellular bacterial communities in the bladder [3].
Therefore, the World Health Organization initiated a global
action plan calling on all countries to adopt measures against
drug-resistant microbes, and the discovery of new anti-
microbial drugs with novel mechanisms has been an urgent
need to combat resistant strains.
previous work that the introduction of triazole [15], imida-
zole [16] and benzimidazole [17] to berberine backbone
could result in highly bioactive molecules, a series of hybrids
of berberine and nitroimidazole via an ethylenic bond bridge
were designed and synthesized for the first time. The design
of target compounds was mainly from following three as-
pects: (1) reduction of aromatic ring of berberine to convert
the planar structure into the bended skeleton was expected to
increase molecular flexibility and solubility; (2) the ethylenic
bond-bridged new π-conjugated skeleton was constructed to
compensate the reduced aromatic ring and maintain the DNA
binding affinity of target molecules; (3) various aliphatic
substituents and benzyl moieties as regulators to ameliorate
physicochemical properties were introduced to investigate
their effects on antimicrobial activities. It was expected that
these novel hybrids of two different antibacterial fragments
could not only increase solubility and bioavailability but also
strengthen DNA binding affinity and membrane permeabi-
lization, thus exhibiting good potency against bacteria,
especially the drug-resistant strains. The natural berberine
was used as starting material to construct berberine-derived
nitroimidaozles. The target molecules were confirmed by
spectral analysis, screened for the antibacterial and anti-
fungal activities in vitro, and investigated for the aqueous
solubility and ClogP values in order to elucidate the struc-
ture-activity relationships. The cell toxicity and antibacterial
resistance were further evaluated for the highly active
compound. Molecular docking and quantum chemical stu-
dies were employed to rationally explain the potent anti-
bacterial effect of the most active molecule. Additionally, the
cell membrane permeabilization and binding behavior to
DNA, isolated from sensitive bacteria, toward the highly
active compound were studied to further explore the possible
antibacterial mechanism.
The antimicrobial drug discovery is quite difficult prob-
ably due to the poor permeation of compounds across the
microbial cell membrane [4]. Some natural products, which
can break through the penetration barriers, are historically
significant as lead compounds for antimicrobials [5]. The
natural berberine, occurring as a mainly pharmacological
component in various plants including Rhizoma coptidis,
Phellodendron amurense and Barberry, has long been uti-
lized as Chinese traditional medicine for the treatment of
diarrhea since 2000 years ago [6]. The good antibacterial
efficiency of berberine might be on account of its large π-
conjugated backbone binding to both single- and double-
stranded DNA in microorganisms through π-π stacking and
electronic interactions [7]. The long-historical use of ber-
berine has not obviously triggered the development of drug-
resistance till today, which showed that the bacteria still
cannot identify and respond to its antibacterial mechanisms.
However, the poor solubility and bioavailability of berberine,
which was probably attributed to the rigidly planar structure,
greatly limited its clinical use [8]. Some researches sug-
gested that the break of planar structure should be favorable
for improving solubility and antibacterial activities [9].
Nitroimidazole is an important moiety widely presented in
clinical drugs, such as dimetridazole, metronidazole, ni-
morazole and tinidazole, with broad spectrum activity
against bacteria and parasites [10], and its unique aromatic
five-membered heterocycle is favorable to effectively inter-
act with biologically important species such as DNAs, en-
zymes and receptors [11]. It is generally considered that the
nitro group will be reduced to reactive radical species, which
can directly react with cellular components to inhibit the
growth of bacteria [12]. However, resistance has occurred
due to the decreased expression and activity of reductive
enzymes in target cell. A lot of work revealed that the
modification of nitroimidazole by large structural fragments
might sterically stabilize nitro group to strengthen the in-
teraction with reductive enzymes and enhance the metabo-
lism [13]. Furthermore, the introduction of nitroimidazole,
which might be beneficial for the improvement of solubility
and bioavailability, has been an important and prevalent
strategy to modify natural products to be the potential clin-
ical drugs.
2 Experimental
2.1 Materials and measurement
Melting points of the synthesized compounds were measured
by the use of X-6 melting point apparatus (Beijing Focus
Instrument CO., Ltd., China) and were uncorrected. Thin
layer chromatography (TLC) analysis was performed
through pre-coated silica gel plates. Bruker RFS100/S
spectrophotometer (Bio-Rad, USA) was employed to record
Fourier transform infrared spectra (FT-IR) using potassium
bromide pellets in the 4000–400 cm−1 range. Nuclear mag-
netic resonance (NMR) spectra were carried out using Bru-
ker AV 600 spectrometer (Germany) and tetramethylsilane
(TMS) as an internal standard. The chemical shifts were
recorded with the unit of parts per million (ppm), the signals
were described as singlet (s), doublet (d), triplet (t) as well as
multiplet (m) and coupling constants (J) were expressed in
In view of this situation and the important function of
heterocyclic azoles [14] by the use of non-covalent bonds in
exerting biological activity, as a further extension of our

3. . . . . . . . . . . . . . . . . . . . . . . . . . . Zhang et al. Sci China Chem May (2018) Vol.61 No.5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 559
hertz (Hz). The following abbreviations were used to des-
ignate fragments: Ber=berberine, Im=nitroimidaozle, Ph=
phenyl. The high resolution mass spectra (HRMS) were
achieved by IonSpec FTICR mass spectrometer (Waters
Micromass, USA) using ESI resource. All fluorescence
spectra were recorded on F-7000 spectrofluorimeter (Hi-
tachi, Japan) and UV spectra were obtained through TU-
2450 spectrophotometer (Puxi Analytic Instrument Ltd. of
Beijing, China), both of which were equipped with 1.0 cm
quartz cells. The microbial strains were obtained from Si-
chuan Provincial People’s Hospital (Chengdu, China) and
neutral red (NR) was obtained from Sigma-Aldrich (USA).
The RAW 264.7 cells were obtained from Kunming wildlife
cell bank of the Chinese Academy of Sciences, Kunming,
China. Tris(hydroxymethyl)methylamine, sodium chloride
and hydrochloric acid were analytical purity. The masses of
samples were weighed using microbalance with a resolution
of 0.1 mg. All other chemicals and solvents were commer-
cially available and directly used without further purifica-
tion.
rially diluted. RAW 264.7 cells were grown as monolayer in
medium (90% DMEM medium, 10% fetal bovine serum, 1%
penicillin/streptomycin) and maintained at 37 °C in a hu-
midified atmosphere containing 5% CO2. Cells were de-
tached from culture flasks with 0.25% trypsin and 0.03%
ethylene diamine tetraacetic acid (EDTA) and resuspended
in fresh culture medium at a density of 1.5×105 cells/mL. By
using a Falcon 96-well, flat-bottom plate, 100 μL of the cell
suspension was added to each of the wells and the cells were
incubated for 24 h. The tested cells were treated with com-
pound 8f and berberine in triplicate at concentrations of 8,
16, 32, 64, 128, 256, and 512 μg/mL, respectively. After
incubation with compounds for 72 h, 50 μL of a 2.5 mg/mL
solution of 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2-H-
tetrazolium bromide (MTT) in phosphate-buffered saline
(PBS) was added to each well and further incubated for
3–4 h. The supernatant was removed, and the cells were
dissolved in 150 μL of DMSO. Absorbance values were
measured at 490 nm by microplate reader.
2.5 Resistance study
2.2 Synthesis
The representative compound 8f was selected to investigate
the developing rate of bacterial resistance according to the
reported method. We exposed a standard strain of resistant E.
coli toward increasing concentrations of compound 8f from
sub-MIC (0.5×MIC, MIC is the minimal inhibitory con-
centration) for sustained passages and determined the new
MIC values of compound 8f for each passage of E. coli. The
initial MIC value of compound 8f and norfloxacin was de-
termined against E. coli as mentioned above in antibacterial
assay. For the next MIC experiment, the bacterial dilution
was prepared by using the bacteria from sub-MIC con-
centration of the compound (0.5×MIC). After a 12 h in-
cubation period, again bacterial dilution was prepared by
using the bacterial suspension from sub-MIC concentration
of the compound (0.5×MIC) and assayed for the next MIC
experiment. The process was repeated for 15 passages. The
MIC values for compound 8f against each passage of E. coli
were determined.
The detailed synthetic procedures and spectral data for
compounds 6–8 were available in the Supporting Informa-
tion online.
2.3 Antibacterial assay
The berberine-derived nitroimidazoles 6–8 were evaluated
for their antibacterial activities against six Gram-positive
bacteria (Methicillin-Resistant Staphylococcus aureus N315,
Enterococcus faecalis, Staphylococcus aureus, Staphylo-
coccus aureus ATCC25923, Bacillus subtilis ATCC6633,
Micrococcus luteus ATCC4698) and six Gram-negative
bacteria (Klebsiella pneumoniae, Escherichia coli, Escher-
ichia coli ATCC25922, Pseudomonas aeruginosa, Pseudo-
monas aeruginosa ATCC27853, Acinetobacter baumannii)
including drug-resistant strains (without ATCC number)
isolated from infected patients. The bacterial suspension was
adjusted with sterile saline to a concentration of 1×105 CFU
measured by nephelometer. Initially the compounds were
dissolved in dimethyl sulfoxide (DMSO) to prepare the stock
solutions, then the tested compounds and reference drugs
were prepared in Mueller-Hinton broth (Guangdong Huaikai
Microbial Sci. & Tech Co., Ltd, Guangzhou, China) to obtain
the required concentrations. These dilutions were inoculated
and incubated at 37 °C for 24 h.
3 Results and discussion
3.1 Chemistry
The berberine-derived nitroimidazoles 6–8 were synthesized
via multi-step reactions from natural berberine as outlined in
Scheme 1. Initially, berberrubine 1 was easily obtained
through selective demethylation of commercial berberine in
high yield of 89.8%, and the latter was further reduced by
excessive sodium borohydride to generate structurally ben-
ded derivative 2 with the yield of 57.0%. Compound 2 was
further formylated by hexamethylenetetramine (HMTA)
2.4 Cytotoxicity assay
The stock solutions of compound 8f and berberine
(1024 μg/mL, in DMSO) were prepared in medium and se-

560 . . . . . . . . . . . . . . . . . . . . Zhang et al. Sci China Chem May (2018) Vol.61 No.5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 560
Scheme 1 The synthesis of berberine-derived nitroimidazoles 6–8. Reagents and conditions: (i) 190 °C/vaccumn, 45 min; (ii) methanol, NaBH4, 25 °C, 4 h;
(iii) (a) HMTA, TFA, 120 °C, 6 h; (b) 10% H2SO4, 25 °C, 2 h; (iv) K2CO3, chloroacetonitrile, CH3CN, 80 °C, 2 h; (v) cyanoethyl nitroimidazole 5, piperidine,
CH3CN, refluxed, 12 h; (vi) aliphatic halides, K2CO3, CH3CN, 80 °C, 12 h; (vii) substituted benzyl halides, K2CO3, CH3CN, 80 °C, 6 h.
1
using trifluoroacetic acid (TFA) as solvent to conveniently
and efficiently produce the corresponding aldehyde 3 with
moderate yield of 78.5%. The condensation of intermediate 3
and cyanoethyl nitroimidazole 5, which was readily prepared
from 2-methyl-5-nitroimidazole, could afford compound 6 in
the presence of piperidine, and the further O-alkylation of
hydroxyl compound 6 with various aliphatic and aromatic
halides was performed to achieve target compounds 7 and 8
in 12.8%–49.5%. All the new compounds were confirmed by
NMR, IR and HRMS spectra. Furthermore, the purity of
target molecules was evaluated through quantitative nuclear
magnetic resonance (QNMR) method using 1,3,5-trioxane as
internal standard, which demonstrated that all the target
compounds exerted a purity of at least 95%.
were observed at the expected regions. In H NMR spectra,
compound 6 gave singlet at 3.99 ppm assigned to O-CH3
protons on berberine backbone. The O-alkylation of mole-
cule 6 by alkyl chains to yield berberine derivatives 7a–7e
resulted in little upfield shifts of O-CH3 protons
(3.94–3.87 ppm), while the replacement of alkyl chain by
aralkyl moiety (8a–8f) would led to little downfield shifts
(3.98–3.91 ppm). The chemical shifts of CH3 protons at-
tached in nitroimidazole ring were lower (2.58–2.55 ppm)
than that of O-CH3, which might be attributed to the electron-
withdrawing oxygen atom. The peaks for C=CH proton in
the target molecules 6−8 appeared at 7.69–7.64 ppm, which
were higher than normal alkene protons because of the new
large π-conjugated skeleton. All the other protons were ob-
served at the appropriate chemical shifts with expected in-
tegral values. In 13C NMR spectra, berberine-derived
nitroimidaozles 6–8 gave expected signals at 13.8–13.7 ppm
which should be assigned to the methyl carbon linked to
nitroimidazole ring, while the chemical shifts of O-CH3
carbon in berberine skeleton were higher (56.5–56.0 ppm)
due to the presence of oxygen atom. No large difference was
found in the 13C NMR chemical shifts for 2- and 3- C in the
berberine backbone (146.4–146.2 ppm). The signals for 5- C
in nitroimidazole ring appeared at 147.6–147.2 ppm, which
3.2 Spectral analysis
In IR spectra, all the newly synthesized molecules 6–8 gave
two broad absorption in 3167–3126 cm−1 and 1662–
1619 cm−1 that suggested the presence of ethylenic group.
The characteristic C≡N bands in all berberine-derived ni-
troimidazoles appeared in the region of 2218–2210 cm−1.
The strong absorption between 1486–1483 cm−1 indicated
the presence of nitro group. All the other absorption bands

5. . . . . . . . . . . . . . . . . . . . . . . . . . . Zhang et al. Sci China Chem May (2018) Vol.61 No.5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561
indicated little upfield shifts in comparison with normal
aromatic carbons because of the presence of electron-with-
drawing nitro group. All other carbons ideally gave the
signals at the expected regions.
3.3 Analysis of X-ray diffraction
The deprotonation of nitroimidazole 4 by potassium carbo-
nate would produce a tautomeric equilibrium of A and B
isomers in theory (Scheme 2), and subsequently reacted with
chloroacetonitrile could generate cyanoethyl nitroimidazoles
5a and 5b. In order to understand the structures of the re-
sulting isomer products 5a or 5b, give rational explanation
for these two isomers and further figure out the absolute
configuration of target compounds, the single crystal of cy-
Figure 1 X-ray structure of berberine-derived nitroimidazole compound
7h (color online).
anoethyl-modified derivative 7h was cultivated and X-ray
diffraction analysis showed that the berberine backbone was
combined with nitroimidazole through the 3-N position in
imidazole nucleus instead of the 1-N position, which in-
dicated that it should be isomer 5b as the main product.
Moreover, it was confirmed that the Z configuration and s-
trans conformation were adopted on account of the large
steric hindrance of nitroimidazole moiety in berberine-de-
rived nitroimidazole compound 7h (Figure 1).
tested bacteria was sensitive to compound 7a with short ethyl
chain. When the length of alkyl chain increased, the decyl
derivative 7e could effectively suppress the growth of P.
aeruginosa ATCC27853 with MIC value of 0.012 mM,
which was much lower than that of reference drug nor-
floxacin (MIC=0.100 mM), while the short unsaturated hy-
drocarbon modified compounds 7f and 7g gave better
inhibitory activity against the tested bacteria than short alkyl
derivatives 7a and 7b. Obviously, the hydroxyl and carbonyl
incorporated compounds 7i and 7j also exerted profound
antibacterial effects with low MIC values toward clinical
drug-resistant bacteria. These results dramatically suggested
that the long length of aliphatic chain should be favorable for
the inhibition of bacteria, which might be attributed to the
improved lipophilicity, and the introduction of unsaturated
groups could also significantly strengthen antibacterial ac-
tivity. Furthermore, the hydroxyl and carbonyl groups
smoothly gave positive influence on the bioactivity, which
might be on account of the formation of hydrogen bonds
between compound and biologically important species.
The in vitro antifungal screening revealed that some pre-
pared berberine nitroimidazoles exhibited good efficacies on
the inhibition of the tested fungal strains. Compound 6 dis-
played moderate to good activities with MIC values ranging
from 0.016 to 1.003 mM. In the series of 7a–7j, compound
7e bearing decyl chain generally gave good anti-C. albicans
activity with MIC value of 0.025 mM. The replacement of
decyl group by ethoxyl fragment, which generated com-
pound 7i, resulted in better efficiency to suppress C. tropi-
calis with MIC value of 0.015 mM, which was 23-fold and
56-fold more potent than reference drugs berberine and
fluconzole, respectively. Especially, the 2-fluorobenzyl de-
rivative 8f showed profound effect against A. fumigatus with
MIC value of 0.026 mM.
3.4 Antimicrobial activity
Table 1 showed that some berberine-derived nitroimidazoles
displayed more potent efficacy in comparison with nor-
floxacin and berberine against the tested bacteria, which
suggested that the introduction of nitroimidaozle to berberine
backbone could greatly optimize the antibacterial activities.
The 9-hydroxyl derivative 6 exhibited much better inhibitory
effect against resistant bacterial strains E. coli and P. aeru-
ginosa with lower MIC value of 0.032 mM than berberine
(MIC=0.688 mM). Subsequently, the O-alkylation of active
precursor 6 by hexyl bromide led to the hexyl hybrid 7c with
strong efficiency against E. faecalis (MIC=0.027 mM),
which was 32-fold and 13-fold more potent than norfloxacin
and berberine, respectively. In contrast with 7c, none of the
Scheme 2 Possible mechanism for the syntheses of cyanoethyl ni-
troimidazole 5.

562 . . . . . . . . . . . . . . . . . . . . Zhang et al. Sci China Chem May (2018) Vol.61 No.5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 562
Table 1
MIC (mM) for berberine-derived nitroimidaozles 6−8 against Gram-positive and Gram-negative bacteria
a), b)
c)
Gram-positive bacteria
Gram-negative bacteria
Compound
S. aureus
B. subtili-
M. luteus
K. pneu-
moniae
E. coli
P. aeru- P. aeruginosa A. bau-
MRSA E. faecalis S. aureus
E. coli
ATCC25923 sATCC6633 ATCC4698
ATCC25922 ginosa ATCC27853
mannii
0.064
0.483
0.230
0.054
0.417
0.399
0.237
0.236
0.474
0.029
0.115
0.777
0.805
0.818
0.204
0.818
0.007
0.688
0.025
6
0.255
0.967
0.459
0.437
0.209
0.100
0.119
0.473
0.118
0.117
0.029
0.049
0.201
0.818
0.409
0.409
0.013
0.128
0.242
0.918
0.027
0.052
0.025
0.237
0.118
0.474
0.059
0.918
0.777
0.805
0.051
0.204
0.409
0.210
0.344
0.802
0.128
0.483
0.459
0.437
0.834
0.100
0.237
0.236
0.118
0.117
0.230
0.388
0.402
0.409
0.818
0.818
0.420
0.011
0.200
0.032
0.967
0.459
0.437
0.209
0.100
0.119
0.473
0.118
0.117
0.029
0.024
0.201
0.818
0.409
0.409
0.013
0.172
0.050
0.255
0.483
0.459
0.437
0.209
0.798
0.237
0.473
0.237
0.469
0.459
0.388
0.805
0.409
0.818
0.204
0.210
0.344
0.050
1.021
0.242
0.459
0.437
0.209
0.399
0.237
0.473
0.474
0.469
0.459
0.388
0.402
0.818
0.818
0.409
0.420
0.344
0.013
0.064
0.483
0.230
0.109
0.209
0.399
0.474
0.473
0.118
0.938
0.115
0.388
0.805
0.102
0.818
0.409
0.052
0.344
1.603
0.032
0.121
0.115
0.437
0.104
0.100
0.119
0.236
0.118
0.059
0.029
0.777
0.050
0.409
0.818
0.204
0.003
0.688
0.100
0.128
0.242
0.459
0.109
0.417
0.399
0.474
0.236
0.118
0.469
0.115
0.388
0.402
0.204
0.409
0.102
0.105
0.688
0.003
0.032
0.121
0.918
0.874
0.834
0.100
0.119
0.473
0.947
0.235
0.115
0.777
0.805
0.818
0.818
0.409
0.420
0.688
0.013
0.032
0.242
0.918
0.874
0.104
0.012
0.237
0.473
0.947
0.235
0.115
0.194
0.805
0.818
0.102
0.818
0.026
0.344
0.100
7a
7b
7c
7d
7e
7f
7g
7h
7i
7j
8a
8b
8c
8d
8e
8f
Berberine 0.688
Norfloxacin 0.013
a) Minimal inhibitory concentrations were determined by micro broth dilution method for microdilution plates; b) MRSA, methicillin-resistant Staphylococcus aureus (N315); E.
faecalis, Enterococcus faecalis; S. aureus, Staphylococcus aureus; S. aureus ATCC25923, Staphylococcus aureus (ATCC25923); B. subtilis ATCC6633, Bacillus subtilis
ATCC6633; M. luteus ATCC4698, Micrococcus luteus (ATCC4698); c) K. pneumoniae, Klebsiella pneumoniae; E. coli, Escherichia coli; E. coli ATCC25922, Escherichia coli
ATCC25922; P. aeruginosa, Pseudomonas aeruginosa; P. aeruginosa ATCC27853, Pseudomonas aeruginosa ATCC27853; A. baumannii, Acinetobacter baumannii.
Table 2 ClogP values and evaluation of solubility of berberine-derived
nitroimidaozels (mean±SD, n=3)
a), b)
3.5 Analysis of ClogP values and aqueous solubility
Aqueous
solubility
Aqueous
solubility
Compound ClogP
Compound ClogP
Lipophilicity/hydrophilicity directly exerts important influ-
ence in various biological processes of bioactive molecules
including transportation, distribution, metabolism and secretion
[20]. Therefore, it is quite necessary to have a good knowl-
edge of lipophilicity/hydrophilicity to predict the bioactivity
of drugs. The theoretically calculated values of logP (ClogP)
for all the berberine-derived nitroimidaozles were obtained
by the use of software ChemDraw Ultra 14.0 (Table 2).
Obviously, most of the berberine-derived nitroimidazoles
were lipophilic (ClogP=2.34–8.07), which was related to the
length of aliphatic chain and the number of halogen atom on
benzene ring. The relationship of all the target compounds
between ClogP and antibacterial effect against drug-resistant
E. coli was successfully investigated as shown in Figure 2. In
general, the evaluated compounds with high values of ClogP
usually showed poor inhibitory activities. The relatively
lower lipophilicity/hydrophilicity partition coefficient
(ClogP<5.5) remarkably resulted in profound efficiency
against E. coli, which suggested that the introduction of li-
pophilic moiety was favorable for the antibacterial activities.
Numerous promising drug candidates commonly suffered
6
2.83
3.83
4.89
5.95
7.01
8.07
3.72
4.08
2.34
5.04±0.27
4.79±0.06
1.65±0.34
6.49±0.14
0.21±0.01
0.09±0.02
5.31±0.09
3.34±0.17
6.71±0.12
7i
2.43
2.66
6.50
4.82
5.79
5.79
5.79
5.22
8.13±0.31
4.14±0.17
0.11±0.04
7.09±0.11
1.65±0.19
2.01±0.20
0.13±0.15
4.54±0.08
7a
7b
7c
7d
7e
7f
7j
8a
8b
8c
8d
8e
7g
7h
8f
Berberine
−0.77 4.08±0.02
a) ClogP values were calculated by ChemDraw Ultra 14.0; b) in μg/mL, de-
termined in pH 7.4 phosphate buffer, at 30 °C.
from poor aqueous solubility, which would lead to in-
complete absorption and hinder their application in clinical
cases. Thus, aqueous solubility is an important parameter for
potential drug [21]. The aqueous solubility of all the target
compounds was evaluated at physiological pH to further
elucidate the affection of various substituents on solubility
and antibacterial activity, and the obtained results were

7. . . . . . . . . . . . . . . . . . . . . . . . . . . Zhang et al. Sci China Chem May (2018) Vol.61 No.5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563
discovery process [22]. The highly active derivative 8f was
further evaluated for its toxicity toward RAW 264.7 cells by
colorimetric cell proliferation MTT assay, and berberine was
used as the positive control. The cytotoxic results shown in
Figure 4 revealed that the cell viability against 2-fluor-
obenzyl derivative 8f of RAW 264.7 cells was at least more
than 85% within concentration of 16 μg/mL (at this con-
centration, compound 8f could effectively inhibit the growth
of resistant E. coli), which indicated that this molecule ex-
erted good selectivity for E. coli over mammalian cells.
However, the cell viability was lower against berberine than
compound 8f in all the tested concentrations, which implied
that the highly bioactive compound 8f exhibited lower cell
toxicity than its precursor natural berberine.
Figure 2 The calculated partition coefficients vs. antibacterial effect
against drug-resistant E. coli. (color online).
3.7 Development of resistance
Bacterial resistance is usually defined as natural drug re-
sistance and induced drug resistance. In comparison with
natural drug resistance that often results from the lack of
target sites for antibiotics, the induced drug resistance can
lead to much more severe issue due to the structural and
bioactive change of functional biomacromolecules such as
DNA, enzymes, efflux pump and cell membrane [23].
Therefore, the evaluation of novel promising drug candidates
for inducing bacterial resistance is of great significance [24].
In the present work, most of the berberine-derived ni-
troimidazoles remarkably exhibited significant effect against
clinical drug-resistant bacteria, such as MRSA, E. faecalis
and E. coli. To further evaluate the potency of the highly
active compound 8f toward multidrug-resistant bacteria, the
susceptible pathogen E. coli was chosen to develop re-
sistance against compound 8f, and norfloxacin was used as
the positive control. The standard strain was exposed to the
increasing concentrations of compound 8f from MIC
shown in Table 2. Most of the berberine-derived ni-
troimidazoles gave relatively better water solubility than
berberine while some molecules exhibited poor solubility,
which might be attributed to the hydrophobic aliphatic chain
and benzyl group. For example, compound 7e with lipophilic
decyl chain logically exhibited the lowest aqueous solubility
(0.09 μg/mL). Furthermore, the contribution of aqueous so-
lubility to the inhibitory potency against resistant E. coli was
scientifically expressed in Figure 3, which remarkably
showed that the highly aqueous solubility (>4 μg/mL) could
be favorable for the antibacterial activity. In summary, it has
demonstrated that the low lipophilicity and high hydro-
philicity would be useful for enhancing antibacterial activ-
ities.
3.6 Cell toxicity
Assessment of toxicity is a critical component in the drug
Figure 4 Cytotoxic assay of the active compound 8f and berberine on
RAW 264.7 cells by MTT methodology. Each data bar represents an
average of three parallels, and error bar indicates one standard deviation
from the mean (color online).
Figure 3 Aqueous solubility vs. antibacterial effect against drug-resistant
E. coli. (color online).

564 . . . . . . . . . . . . . . . . . . . . Zhang et al. Sci China Chem May (2018) Vol.61 No.5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 564
(0.003 mM) for sustained passages and the new MIC values
were investigated against each passage of E. coli.
As shown in Figure 5, there was no obvious substantial
variation in MIC values for compound 8f during all the 16
passages of E. coli strain, whereas norfloxacin significantly
gave increasing MIC values after the 9th passage and even
around 16-fold increase in MIC was observed after 15th
passage. These obtained results showed that clinical drug-
resistant E. coli was more difficult to develop resistance
toward compound 8f than the clinical drug norfloxacin.
3.8 Electronic properties
The concept of highest occupied molecular orbital (HOMO)
Figure 5 Evaluation of resistance development against compound 8f in
drug-resistant E. coli. (color online).
and lowest unoccupied molecular orbital (LUMO), which is
fundamentally prominent as it gives basic understanding for
the chemical stability and reactivity of a given molecule, is
recently applied to the rational design of drug molecules
[25]. The HOMO can be regarded as the outermost electron
containing orbital that tends to give these electrons away as
an electron donor, while the LUMO can be considered as the
innermost orbital containing free sites to serve as electron
receptor [26]. Extending the concept into drug-receptor
binding system, the HOMOs of compounds 7c, 8b and 8f
were mainly localized on berberine backbone and the LU-
MOs were primarily focused on the nitroimidazole ring, as
shown in Table 3. These results manifested that the elec-
trophilic interaction might occur between positively charged
biological targets and berberine backbone, whereas the ni-
troimidazole ring could bind to the negatively charged re-
sidues of receptor. Besides, it was observed that the
substituents at 9-position of berberine skeleton did not di-
rectly contribute to the HOMO and LUMO, which suggested
that these groups might be primarily used to regulate the
physicochemical characterizations.
The quantum chemical parameters of compounds 7c, 8b
and 8f were calculated to predict the possible biological
properties (Table 3). It was known that the electronegative
molecule can be a good π-electron acceptor, which further
evidenced that the natural berberine-derived nitroimidazoles
7c, 8b and 8f with relatively high electronegativity might
interact with various biological species through π-π stacking.
Moreover, the suitable global hardness and softness theore-
tically explained the low toxicity.
The molecular electrostatic potential (MEP) surface gen-
erally gives an observation of the charged surface area [27],
which can convictively probe the hydrophilicity of mole-
cules, electrostatic interaction between compounds and
biological targets and the orientation of drug candidates for
a)
Table 3 Atomic orbital HOMO-LUMO composition and quantum chemical parameters of compounds 7c, 8b and 8f
Compound
7c
8b
8f
HOMO (eV)
E=−5.58
E=−5.70
E=−5.58
LUMO (eV)
MEPs
E=−2.38
E=−2.58
E=−2.39
Electronegativity
Global hardness
Global softness
3.98
1.60
0.62
4.14
1.56
0.64
3.99
1.59
0.63
a) The 3D structure of the molecules were built by Sybyl 2.0 and Gauss 09, and were optimized through the B3LYP hybrid functional and standard 6-31G(d) basis set (color
online).

9. . . . . . . . . . . . . . . . . . . . . . . . . . . Zhang et al. Sci China Chem May (2018) Vol.61 No.5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 565
their activity [28]. In the present work, the MEP maps of
target compounds 7c, 8b and 8f were investigated and the
results, shown in Table 3, suggested that an electronegative
area (in red) can be observed at the nitroimdazole ring, which
strongly demonstrated that the long pair from nitroimidazole
probably oriented toward the outer part of the molecule, thus
easily interacting with various enzymes and receptors in
biological systems.
DNA, which was isolated from drug-resistant E. coli through
a four-step process, were performed using neutral red (NR)
as a spectral probe by UV-Vis spectroscopy.
The detection of hyperchromism and hypochromism by
absorption spectroscopy, which are critical spectral features
to distinguish the change of DNA double helical structure, is
one of the most important and useful method in DNA-
binding studies [32]. The observed strong hyperchromism
might be attributed to the denaturation of DNA caused by the
strong interaction between the electronic states of inter-
calating chromophore and that of the DNA base. The UV-Vis
absorption spectra were recorded with gradually increasing
amount of compound 8f at a fixed concentration of DNA. As
shown in Figure 7. The UV-Vis spectra revealed that the
maximum absorption peak of DNA at 260 nm showed pro-
portional increase and slight blue shift with the increasing
concentration of derivative 8f. Remarkably, the spectral in-
formation suggested that the measured absorption value of
8f-DNA complex was a little greater than that of simply sum
of free DNA and free compound 8f. This demonstrated a
constructional change in DNA duplex, which might be
characteristic of non-covalent interactions between DNA and
complexes resulting in some disassembly of DNA helix and
the exposure of some DNA bases. In general, the observed
spectral properties could be explained by the potent inter-
action of compound 8f with E. coli DNA.
3.9 Molecular docking
Molecular docking study as an effective method is widely
employed in drug discovery for investigating binding modes
of small molecules to biological species [29]. To rationalize
the obtained antibacterial activity and elucidate the anti-E.
coli action mechanism, molecular docking study was suc-
cessfully performed between compound 8f and E. coli DNA
polymerase III (PDB ID: 2POL) that was of great importance
for the synthesis of E. coli DNA. As shown in Figure 6, the
nitro group of compound 8f was in close proximity to the
residue GLY-219 through hydrogen bonds with a distance of
1.6 Å and the binding energy was as low as −5.32 kJ/mol,
which might be accordant with MEP effect. This interaction
demonstrated that compound 8f might disturb the synthesis
of E. coli DNA through binding to DNA polymerase III, thus
exhibiting profound antibacterial potency.
Based on the changes in the absorption spectra of DNA
upon binding to 8f, the binding constant K could be calcu-
lated by the Eq. (1), where A0 and A represent the absorbance
of DNA in the absence and presence of compound 8f at
260 nm, ξC and ξD-C are the absorption coefficients of com-
pound 8f and 8f-DNA complex, respectively. The plot of A0/
(A−A0) versus 1/[compound 8f] is constructed by using the
3.10 Interaction with DNA
Deoxyribonucleic acid (DNA) is of great importance for
storing information to continue the life and can also be a
potent drug target due to the presence of multiple sites in-
teracting with drug molecules, which is usually exploited as
an attractive approach for the rational design and construc-
tion of novel and effective drugs [30]. It has been reported
that berberine and its derivatives could effectively interact
with DNA through various binding modes [31]. Therefore, to
explore the possible antimicrobial mechanism, the in vitro
binding studies of highly active compound 8f with E. coli
Figure 7 Interaction of resistant E. coli DNA with highly active molecule
8f at different concentrations was recorded by UV absorption spectra (pH
7.4, T=303 K). Inset: comparison of absorption at 260 nm between the 8f-
DNA complex and the sum values of free DNA and free compound 8f.
[DNA]=5.18×10−5 M, and [compound 8f]=0−1.21×10−5 M for curves (a–h)
respectively at increment of 0.174×10−5 M (color online).
Figure 6 Molecular modeling of compound 8f and E. coli DNA poly-
merase III (color online).

566 . . . . . . . . . . . . . . . . . . . . Zhang et al. Sci China Chem May (2018) Vol.61 No.5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 566
absorption titration data and linear fitting (Figure S2, Sup-
porting Information online), yielding the binding constant,
K=2.85×105 L/mol, R=0.998, SD=0.089 (R is the correlation
coefficient, SD is standard deviation).
3.11 Bacterial membrane permeabilization
The exploitation of bacterial membrane active agents is an
increasingly important approach to solve the problem of
bacterial resistance for that the membrane-active property
mostly blocks the development of resistance [34]. Therefore,
the clinical drug-resistant E. coli was rationally chosen to
investigate the ability of highly active compound 8f to per-
meabilize the bacterial cell membrane using propidium io-
dide (PI) as dye, which could successfully pass through the
membrane of compromised bacterial cells and form a com-
plex with DNA to fluoresce. As shown in Figure 9, the
fluorescence intensity of experimental group (in the presence
of compound 8f) presented stability after first experienced an
upward trend with the increase in the time, while the control
group almost showed no change. The observed results de-
monstrated that compound 8f could effectively permeabilize
the membrane of resistant E. coli.
A0
1
C
C
=
+
×
(1)
K[Q]
A
A0
C
C
D
C
D
C
To further elucidate the interaction between compound 8f
and E. coli DNA, neutral red (NR, a planar phenazine dye
with a confirmed intercalative binding mode with DNA) was
employed as a spectral probe due to its lower toxicity and
higher stability. Therefore, NR was chosen as the spectral
probe to investigate the binding mode of compound 8f with
DNA in present work, and the absorption spectra of the NR
dye upon the addition of DNAwere shown in Figure S3. The
absorption peak of NR at around 460 nm gradually decreased
with the increasing concentration of DNA, and a new band at
around 530 nm appeared, which could be attributed to the
formation of the new DNA-NR complex [33]. The isosbestic
point at 500 nm provided evidence for the formation of
DNA-NR complex.
3.12 Selective inhibition of bacteria
Further investigation for the mode of binding between NR
and compound 8f was carried out via the competitive inter-
action of NR and 8f with DNA. As shown in Figure 8, an
apparent intensity increase was observed around 460 nm
along with the increasing concentration of compound 8f.
Besides, when compared with the absorption around 460 nm
of free NR in the presence of the increasing concentrations of
DNA (Figure S3), the absorbance at the same wavelength
exhibited the reverse process (inset of Figure 8). These re-
sults strongly demonstrated that compound 8f could effec-
tively intercalate into the double helix of DNA by
substituting NR in the DNA-NR complex.
Microorganisms are able to coexist in mixed populations and
can cause complex bacterial infections. Generally, in order to
treat these complex infections, various types of antimicrobial
drugs will be jointly used, and as a result, excess use of
antibiotics accelerates the development of drug-resistance
[35]. Specifically addressing a single bacterial strain from a
mixture of multiple strains is a fascinating scientific chal-
lenge and a necessity in healthcare and provides an important
tool in microbiology research [36]. Thus, the selectively
inhibitory activity of the highly active compound 8f with
MIC value of 0.003 mM against drug-resistant E. coli was
evaluated through incubating mixture strains (including
Gram-positive bacteria S. aureus, Gram-negative bacteria P.
aeruginosa and fungi C. albicans, respectively) on agar plate
containing 0.003 mM of berberine-derived nitroimidazole
8f. As the colony morphology was shown in Table 4, in the
Figure 8 The competitive reaction between highly active molecule 8f and
NR with E. coli DNA was determined by UV absorption spectra. [DNA]=
5.18×10−5 M, [NR]=2×10−5 M, and [compound 8f]=0−1.21×10−5 M for
curves (a–h) respectively at increment of 0.174×10−5. Inset: absorption
spectra of the system with the increasing concentration of 8f in the wa-
velength range of 350–550 nm absorption spectra of competitive reaction
between compound 8f and NR with E. coli DNA (color online).
Figure 9 Membrane permeabilization of resistant bacterial E. coli toward
compound 8f (12×MIC) (color online).

11. . . . . . . . . . . . . . . . . . . . . . . . . . Zhang et al. Sci China Chem May (2018) Vol.61 No.5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567
a)
Table 4
Selectively inhibitory effect against resistant E. coli of highly active derivative 8f
E. coli vs. S. aureus
E. coli vs. P. aeruginosa
E. coli vs. C. albicans
Presence of compound 8f
Absence of compound 8f
a) Representative photographs of colonies from mixed cultures incubated for 12 h in the presence and absence of 0.003 mM compound 8f.
presence of compound 8f, the three agar plates did not show
any growth at the E. coli sector, while normal growth was
observed at the other two sectors. Comparably, all the strains
were normally growing in the absence of compound 8f.
These results demonstrated that compound 8f exhibited the
potency to selectively inhibit the growth of resistant E. coli
from a mixture of microorganisms.
would exhibit distinct influence on the bioactivities and most
of the aliphatic substituents exerted relatively potent anti-
microbial property (Figure 10). However, 2-fluorobenzyl
compound 8f with an improved aqueous solubility (4.54±
0.08 μg/mL) not only exerted strong activity against drug-
resistant E. coli (MIC=0.003 mM), which was more active
than berberine and norfloxacin, but also exhibited low toxi-
city toward RAW 264.7 cells and less propensity to trigger
development of resistance. The molecular docking and
quantum chemical experiments suggested that compound 8f
might target E. coli DNA polymerase III through the for-
mation of hydrogen bonds. Berberine skeleton and ni-
troimidazole ring theoretically acted as the electron donor
and receptor, respectively, to bind with various biological
molecules. The further exploration of antibacterial mechan-
ism revealed that the highly active molecule 8f could ef-
fectively permeabilize resistant E. coli cell membrane and
intercalate into DNA isolated from resistant E. coli to form
8f-DNA complex (K=2.85×105 L/mol), thus blocking DNA
replication. Moreover, the active molecule 8f could selec-
4 Conclusions
In conclusion, a series of berberine-derived nitroimidazoles
were successfully developed by a convenient procedure from
commercially natural berberine. Their structures were con-
firmed by NMR, IR and HRMS spectra. The antimicrobial
evaluation in vitro indicated that some target molecules
showed good inhibitory effects against drug-resistant bac-
teria (E. coli, A. baumannii and E. faecalis) and drug-sus-
ceptible strains (S. aureus ATCC25923 and P. aeruginosa
27853). The structure-activity relationships revealed that
various substituents on 9-position of berberine skeleton
Figure 10 Preliminary summary of structure-activity relationships (color online).

568 . . . . . . . . . . . . . . . . . . . . Zhang et al. Sci China Chem May (2018) Vol.61 No.5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 568
14 (a) Zhang L, Peng XM, Damu GLV, Geng RX, Zhou CH. Med Res
tively inhibit the growth of E. coli from a mixture of various
strains. These results strongly demonstrated that the natural
Rev, 2014, 34: 340–437; (b) Rouf A, Tanyeli C. Eur J Med Chem,
2015, 97: 911–927; (c) Zhou CH, Wang Y. Curr Med Chem, 2012, 19:
berberine-derived nitroimidaozle 8f was a promising candi-
date as multi-targeting anti-E. coli drug.
239–280; (d) Dai L, Cui S, Damu GLV, Zhou C. Chin J Org Chem,
2013, 33: 224–244
15 Zhang SL, Chang JJ, Damu GLV, Fang B, Zhou XD, Geng RX, Zhou
CH. Bioorg Med Chem Lett, 2013, 23: 1008–1012
16 Wen SQ, Jeyakkumar P, Avula SR, Zhang L, Zhou CH. Bioorg Med
Chem Lett, 2016, 26: 2768–2773
17 (a) Jeyakkumar P, Liu HB, Gopala L, Cheng Y, Peng XM, Geng RX,
Zhou CH. Bioorg Med Chem Lett, 2017, 27: 1737–1743; (b) Lai TT,
Xie D, Zhou CH, Cai GX. J Org Chem, 2016, 81: 8806–8815
18 (a) Zhang FF, Gan LL, Zhou CH. Bioorg Med Chem Lett, 2010, 20:
1881–1884; (b) Peng XM, Kumar KV, Damu GLV, Zhou CH. Sci
China Chem, 2016, 59: 878–894
Acknowledgements This work was partially supported by the National
Natural Science Foundation of China (21672173, 21372186), Research
Fund for International Young Scientists from International (Regional) Co-
operation and Exchange Program of NSFC (81650110529), Chongqing
Special Foundation for Postdoctoral Research Proposal (Xm2016039,
Xm2017185), Program for Overseas Young Talents from State Adminis-
tration of Foreign Experts Affairs, China (WQ2017XNDX047), the Doc-
toral Fund of Southwest University (SWU111075), and the Research Funds
for the Central Universities (XDJK2017B015).
19 (a) Su T, Xie S, Wei H, Yan J, Huang L, Li X. Bioorg Med Chem,
2013, 21: 5830–5840; (b) Zhang L, Chang JJ, Zhang SL, Damu GLV,
Geng RX, Zhou CH. Bioorg Med Chem, 2013, 21: 4158–4169
20 Salas PF, Herrmann C, Cawthray JF, Nimphius C, Kenkel A, Chen J,
de Kock C, Smith PJ, Patrick BO, Adam MJ, Orvig C. J Med Chem,
2013, 56: 1596–1613
21 (a) Zelikin AN, Ehrhardt C, Healy AM. Nat Chem, 2016, 8: 997–
1007; (b) Dai LL, Zhang HZ, Nagarajan S, Rasheed S, Zhou CH. Med
Chem Commun, 2015, 6: 147–154
Conflict of interest The authors declare that they have no conflict of
interest.
Supporting information The supporting information is available online at
http://chem.scichina.com and http://link.springer.com/journal/11426. The
supporting materials are published as submitted, without typesetting or
editing. The responsibility for scientific accuracy and content remains en-
tirely with the authors.
22 (a) von Karstedt S, Montinaro A, Walczak H. Nat Rev Cancer, 2017,
17: 352–366; (b) Cui SF, Addla D, Zhou CH. J Med Chem, 2016, 59:
4488–4510
23 (a) Wolny-Koładka K, Lenart-Boroń A. Water Air Soil Pollut, 2016,
227: 146; (b) Hu J, Liu F, Ju H. Angew Chem Int Ed, 2016, 55: 6667–
6670
24 (a) Konai MM, Ghosh C, Yarlagadda V, Samaddar S, Haldar J. J Med
Chem, 2014, 57: 9409–9423; (b) Gao WW, Zhou CH. Future Med
Chem, 2017, 9: 1461–1464
25 (a) Abraham CS, Prasana JC, Muthu S. Spectrochim Acta A, 2017,
181: 153–163; (b) Dey T, Praveena KSS, Pal S, Mukherjee AK. J Mol
Struct, 2017, 1137: 615–625
26 Cheng Y, Avula SR, Gao WW, Addla D, Tangadanchu VKR, Zhang
L, Lin JM, Zhou CH. Eur J Med Chem, 2016, 124: 935–945
27 Zhang HZ, He SC, Peng YJ, Zhang HJ, Gopala L, Tangadanchu VKR,
Gan LL, Zhou CH. Eur J Med Chem, 2017, 136: 165–183
28 Gao WW, Rasheed S, Tangadanchu VKR, Sun Y, Peng XM, Cheng Y,
Zhang FX, Lin JM, Zhou CH. Sci China Chem, 2017, 60: 769–785
29 (a) Zhang SL, Chang JJ, Damu GLV, Geng RX, Zhou CH. Med Chem
Commun, 2013, 4: 839–846; (b) Yin BT, Yan CY, Peng XM, Zhang
SL, Rasheed S, Geng RX, Zhou CH. Eur J Med Chem, 2014, 71: 148–
159; (c) Fang XF, Li D, Tangadanchu VKR, Gopala L, Gao WW,
Zhou CH. Bioorg Med Chem Lett, 2017, 27: 4964–4969
30 (a) Shimazaki Y, Tanaka J, Kohara Y, Kamahori M, Sakamoto T.
Anal Chem, 2017, 89: 6305–6308; (b) Addla D, Wen SQ, Gao WW,
Maddili SK, Zhang L, Zhou CH. Med Chem Commun, 2016, 7: 1988–
1994
1
2
3
Garland M, Loscher S, Bogyo M. Chem Rev, 2017, 117: 4422–4461
Brown ED, Wright GD. Nature, 2016, 529: 336–343
(a) Petty NK, Ben Zakour NL, Stanton-Cook M, Skippington E,
Totsika M, Forde BM, Phan MD, Gomes Moriel D, Peters KM, Davies
M, Rogers BA, Dougan G, Rodriguez-Baño J, Pascual A, Pitout JDD,
Upton M, Paterson DL, Walsh TR, Schembri MA, Beatson SA. Proc
Natl Acad Sci USA, 2014, 111: 5694–5699; (b) Guo Y, Liu X, Li B,
Yao J, Wood TK, Wang X. J Bacteriol, 2017, 199: e00057
(a) Ling LL, Schneider T, Peoples AJ, Spoering AL, Engels I, Conlon
BP, Mueller A, Schäberle TF, Hughes DE, Epstein S, Jones M, La-
zarides L, Steadman VA, Cohen DR, Felix CR, Fetterman KA, Millett
WP, Nitti AG, Zullo AM, Chen C, Lewis K. Nature, 2015, 517: 455–
459; (b) O’Connell KMG, Hodgkinson JT, Sore HF, Welch M, Sal-
mond GPC, Spring DR. Angew Chem Int Ed, 2013, 52: 10706–10733
(a) Peng XM, Damu GLV, Zhou CH. Curr Pharm Des, 2013, 19:
3884–3930; (b) Silva LN, Zimmer KR, Macedo AJ, Trentin DS.
Chem Rev, 2016, 116: 9162–9236
Tillhon M, Guamán Ortiz LM, Lombardi P, Scovassi AI. Biochem
Pharmacol, 2012, 84: 1260–1267
(a) Jeyakkumar P, Zhang L, Avula SR, Zhou CH. Eur J Med Chem,
2016, 122: 205–215; (b) Mistry B, Keum YS, Noorzai R, Gansukh E,
Kim DH. J Iran Chem Soc, 2016, 13: 531–539
(a) Endeshaw M, Zhu X, He S, Pandharkar T, Cason E, Mahasenan
KV, Agarwal H, Li C, Munde M, Wilson WD, Bahar M, Doskotch
RW, Kinghorn AD, Kaiser M, Brun R, Drew ME, Werbovetz KA. J
Nat Prod, 2013, 76: 311–315; (b) Kumar A, Ekavali A, Chopra K,
Mukherjee M, Pottabathini R, Dhull DK. Eur J Pharmacol, 2015, 761:
288–297
4
5
6
7
8
31 Jiang H, Wang X, Huang L, Luo Z, Su T, Ding K, Li X. Bioorg Med
Chem, 2011, 19: 7228–7235
9
Duan JR, Liu HB, Jeyakkumar P, Gopala L, Li S, Geng RX, Zhou CH.
Med Chem Commun, 2017, 8: 907–916
32 (a) Peng LP, Nagarajan S, Rasheed S, Zhou CH. Med Chem Commun,
2015, 6: 222–229; (b) Zhang L, Kumar KV, Rasheed S, Zhang SL,
Geng RX, Zhou CH. Med Chem Commun, 2015, 6: 1303–1310; (c) Li
ZZ, Gopala L, Tangadanchu VKR, Gao WW, Zhou CH. Bioorg Med
Chem, 2017, 25: 6511–6522
33 Zhang Y, Damu GLV, Cui SF, Mi JL, Tangadanchu VKR, Zhou CH.
Med Chem Commun, 2017, 8: 1631–1639
34 Zhang L, Addla D, Ponmani J, Wang A, Xie D, Wang YN, Zhang SL,
Geng RX, Cai GX, Li S, Zhou CH. Eur J Med Chem, 2016, 111: 160–
182
10 (a) Peng XM, Cai GX, Zhou CH. Curr Top Med Chem, 2013, 13:
1963–2010; (b) Zhang HZ, Gan LL, Wang H, Zhou CH. Mini-Rev
Med Chem, 2017, 17: 122–166; (c) Liu HB, Gao WW, Tangadanchu
VKR, Zhou CH, Geng RX. Eur J Med Chem, 2018, 143: 66–84
11 Khalaj A, Nakhjiri M, Negahbani AS, Samadizadeh M, Firoozpour L,
Rajabalian S, Samadi N, Faramarzi MA, Adibpour N, Shafiee A,
Foroumadi A. Eur J Med Chem, 2011, 46: 65–70
12 Ang CW, Jarrad AM, Cooper MA, Blaskovich MAT. J Med Chem,
2017, 60: 7636–7657
13 Cui SF, Peng LP, Zhang HZ, Rasheed S, Vijaya Kumar K, Zhou CH.
Eur J Med Chem, 2014, 86: 318–334
35 Velema WA, van der Berg JP, Szymanski W, Driessen AJM, Feringa
BL. ACS Chem Biol, 2014, 9: 1969–1974
36 Zheng T, Nolan EM. J Am Chem Soc, 2014, 136: 9677–9691