LED 209 conjugated chitosan as a selective antimicrobial and potential anti-adhesion material

Article abstract of DOI:10.1016/j.carbpol.2018.11.044

The rapid emergence of antibiotic-resistant Gram-negative bacteria (GNB) is becoming a global healthcare concern, and it urgently needs novel strategies to match the clinical challenge. In this work, we conjugated chitosan (CS) with LED 209, a highly selective inhibitor of QseC of GNB, to create the novel selective antimicrobial agent CS/LED. The data of FT-IR, NMR and elemental analysis for CS/LED conjugates proved the successful conjugation of CS with LED 209. Interestingly, the fluorescence signal detected in MDR-E. coli of CS/LED-FITC was about 2 times than that of CS-FITC at 3 h. The results shown that compared with CS, CS/LED exhibited higher selective antimicrobial on MDR-E. coli. Moreover, CS/LED exhibited the lower selectivity and cytotoxicity to mammalian cell than CS. Additionally, an unexpected enhancement of anti-adhesion activity against MDR-E. coli was determined by cellulose membrane coating CS/LED. The results demonstrated that CS/LED could reduce the adhesion of bacteria to the cellulose membrane by about 67.8%, while CS only reduced by about 45.3%. The dressings coated with CS/LED possessed the stronger ability to prevent microbial adhesion compared to the CS-coated dressing. Our present work firstly demonstrated that CS/LED had a highly selective activity and anti-adhesion activity against MDR-E. coli, which offered a potent and selective antimicrobial for combating multidrug-resistant GNB infections.

Full text of DOI:10.1016/j.carbpol.2018.11.044

Accepted Manuscript
Title: LED 209 conjugated chitosan as a selective
antimicrobial and potential anti-adhesion material
Authors: Zongbao Zhou, Tao Chen, Naibin Mei, Bing Li,
Zunkai Xu, Lei Wang, Xiaoying Wang, Shunqing Tang
PII:
S0144-8617(18)31378-X
DOI:
Reference:
https://doi.org/10.1016/j.carbpol.2018.11.044
CARP 14295
To appear in:
Received date:
Revised date:
Accepted date:
31 July 2018
13 October 2018
15 November 2018
Please cite this article as: Zhou Z, Chen T, Mei N, Li B, Xu Z,
Wang L, Wang X, Tang S, LED 209 conjugated chitosan as a selective
antimicrobial and potential anti-adhesion material, Carbohydrate Polymers (2018),
https://doi.org/10.1016/j.carbpol.2018.11.044
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

LED 209 conjugated chitosan as a selective
antimicrobial and potential anti-adhesion
material
1
1
1
1
1
Zongbao Zhou , Tao Chen , Naibin Mei , Bing Li , Zunkai Xu , Lei
2
1
1, *
Wang , Xiaoying Wang and Shunqing Tang
1Biomedical Engineering Institute, Jinan University, Guangzhou 510632,
P. R. China.
2Department of Pharmacology, Utrecht Institute for Pharmaceutical
Sciences, Utrecht University, 3568 CG Utrecht, Netherlands.
*Corresponding Author:
Shunqing Tang, Biomedical Engineering Institute, Jinan University,
Guangzhou 510632, P. R. China.;
E-mail address: tshunqt@jnu.edu.cn (S. Tang).
Highlights:

LED 209 conjugated CS as the novel selective antimicrobial and
anti-adhesion agent.

The conjugate exhibited higher selective antimicrobial on MDR-
E. coli than CS.
1


The conjugate could reduce MDR-E. coli adhesion and prevent its
biofilm formation.

The conjugate offered a potent antimicrobial for combating GNB
infections.
Graphical Abstract:
Abstract:
The rapid emergence of antibiotic-resistant Gram-negative bacteria
(GNB) is becoming a global healthcare concern, and it urgently needs
novel strategies to match the clinical challenge. In this work, we
conjugated chitosan (CS) with LED 209, a highly selective inhibitor of
QseC of GNB, to create the novel selective antimicrobial agent CS/LED.
The data of FT-IR, NMR and elemental analysis for CS/LED conjugates
proved the successful conjugation of CS with LED 209. Interestingly, the
fluorescence signal detected in MDR-E. coli of CS/LED-FITC was about
2
times than that of CS-FITC at 3 h. The results shown that compared
2

with CS, CS/LED exhibited higher selective antimicrobial on MDR-E.
coli. Moreover, CS/LED exhibited the lower selectivity and cytotoxicity
to mammalian cell than CS. Additionally, an unexpected enhancement of
anti-adhesion activity against MDR-E. coli was determined by cellulose
membrane coating CS/LED. The results demonstrated that CS/LED could
reduce the adhesion of bacteria to the cellulose membrane by about
67.8%, while CS only reduced by about 45.3%. The dressings coated
with CS/LED possessed the stronger ability to prevent microbial adhesion
compared to the CS-coated dressing. Our present work firstly
demonstrated that CS/LED had a highly selective activity and anti-
adhesion activity against MDR-E. coli, which offered a potent and
selective antimicrobial for combating multidrug-resistant GNB infections.
Keyword: Chitosan; LED 209; multidrug-resistant Gram-negative
bacteria; selective activity; anti-adhesion activity
1
. Introduction
Microbial infection is a serious worldwide threat to human health,
which kills millions of people every year (Chassaing & Cascales, 2018;
Flemming et al., 2016; Schillaci et al., 2017). Over the past few decades,
the evolution of antibiotics with an excellent profile of safety has
contributed highly to the improvement of the quality of life. Nonetheless,
the excessive use and misuse of antibiotics have resulted in the
emergence of antibiotic-resistant bacterial strains, especially antibiotic-
resistant Gram-negative bacteria (GNB), that is becoming a major health
3

issue of great concern (Lam et al., 2016; Munguia & Nizet, 2017; van
Harten, Willems, Martin, & Hendrickx, 2017). Notably, GNB strains have
a strong capacity to adhere to and form biofilm on abiotic surfaces,
including catheters and prosthetic devices, and this inherent characteristic
may bring unforeseen medical accidents (Cloutier, Mantovani, & Rosei,
2015; Raphel, Holodniy, Goodman, & Heilshorn, 2016). Despite
enormous efforts devoted to exploring the solution of infections
contribute to bacteria, scant researches can commendably address GNB
(Domalaon, Idowu, Zhanel, & Schweizer, 2018). Consequently, the
requirement of novel agents with strong selective against GNB is urgently
needed.
Chitosan (CS) is obtained from chitin, the second most abundant
polysaccharide, and has become a sought-after biomolecule due to its
cationicity and the presence of multiple functionality in the molecule
(Anitha et al., 2014). In addition, CS has emerged as an extremely
important biomaterial owing to its versatile biological activity, excellent
biocompatibility and complete biodegradability, as well as low toxicity
(Dash, Chiellini, Ottenbrite, & Chiellini, 2011). Both in vitro and in vivo
studies have demonstrated that CS possesses several fascinating
biological activities that may avail to several diseases including immune
dysregulation, metabolic abnormality, microbial infection and so forth
(Zargar, Asghari, & Dashti, 2015).
Noteworthy plenty of reports relate to variety of its biological
properties are available, one of the most unique biological properties of
4

CS is its antibacterial activity. The spectrum of antimicrobial activity of
CS and its derivatives extends to include fungi, yeasts and bacteria.
Although antimicrobial activity of CS has been well documented, its
mode of action has hitherto remained just vaguely unclear (Verlee,
Mincke, & Stevens, 2017). The great amount of literatures on various CS-
treated microorganisms do suggest that CS probably affects on the
microbial cell membrane. CS can not only inhibit bacterial adhesion
itself, but also serve as a drug carrier, making it a very intriguing material
to consider for future use (Rabea, Badawy, Stevens, Smagghe, &
Steurbaut, 2003), however, quite a few studies have illustrated that CS is
less sensitive to GNB than gram-positive bacteria, since a possible
binding of CS to lipopolysaccharides (LPS) (Wrighton, 2018), merely
confined to the outer membrane of GNB, would not significantly affect
the bacterial envelope (Badawy, Rabea, Taktak, & El-Nouby, 2016; Yue
et al., 2017). Chemical modification is an efficient approach to improve
antimicrobial activity for CS (Sashiwa & Aiba, 2004; H. Wang, Qian, &
Ding, 2017). Therefore, a promising strategy is to modify CS with
substitute that plays a role in the selective anti-GNB or anti-adhesion
activity.
Recent researches revealed that targeting microbial virulence has
become an effective and convincing strategy for combating bacterial
infection (Monedero, Collado, & Rodriguez-Diaz, 2018). QseC is a
conserved membrane histidine sensor kinase that has been found in at
least 21 important human and plant GNB by now (Papenfort & Bassler,
5

2016). Furthermore, the QseC receptor exerts a crucial role in promoting
the virulence of pathogens. LED 209 has been confirmed to be a highly
selective inhibitor of QseC, which remarkably inhibits the virulence of
several pathogens in vitro and in vivo. In addition, LED 209 is not toxic
and does not inhibit pathogen growth, which, therefore, may not induce
drug resistance. Furthermore, LED 209 has been evidenced that it could
resist biofilm formation in several multidrug-resistant clinical isolates of
recurrent urinary tract infections (rUTIs) to prevent bacterial adhesion
(Curtis et al., 2014; Rasko et al., 2008). Recently, Luo’s team reported
that a novel conjugation, G3 PAMAM-LED 209, had antimicrobial
potential against most GNB and could decrease the GNB virulence by
inhibiting QseC receptor, as well as reduced the original cytotoxicity of
G3 PAMAM (Xue et al., 2015). Given the benefits of LED 209, CS-LED
209 conjugates appear to be exceedingly attractive, nevertheless, so far,
there is none information available regarding it.
LED 209 is highly selective to the membrane-bound histidine kinase
QseC and intensively reduces the expression of key virulence genes of
several Gram-negative pathogens (Rasko et al., 2008). Unfortunately,
LED 209 does not possess functional groups for its further application.
Therefore, it is extremely significant to design reasonable functional
groups on its structure. According to previous research (Xue et al., 2015),
carboxyl group was successfully introduced at positions of the benzene
ring of LED 209. In addition, the carboxylated LED 209 compounds still
kept great potential to inhibit the expression of virulence in GNB without
6

apparent cytotoxicity to mammalian cells. Noticeably, the meta-LED 209-
COOH exhibited higher activity against the QseC receptor para-LED
2
09-COOH. Thus, the meta-LED 209-COOH was synthesized to
conjugate with CS by a covalent amide bond in this research.
In this study, we firstly synthesized the conjugation of CS with LED
2
2
09 through substitution of some amino groups on CS with meta-LED
09-COOH. The chemical structure of CS/LED was characterized by FT-
1
IR, H NMR and elemental analysis. These results indicated that LED
2
09 had been successfully conjugated onto CS chains. In addition, the
degrees of substitution (DS) of the CS/LED conjugates were also
calculated (0.08～0.24). High performance liquid chromatography
(HPLC) results shown that there was no meta-LED 209-COOH released
from the conjugates under the corresponding conditions in the activity
tests. The antibacterial activity of CS and CS/LED against E. coli, MDR-
E. coli, P. aeruginosa and S. aureus were studied. Furthermore, the
conjugate also exhibited high selectivity on MDR-E. coli and low
cytotoxicity to mammalian cells. Importantly, this novel conjugate could
availably inhibit the formation of biofilm by reducing the adhesion of
MDR-E. coli. This investigation demonstrated the CS/LED conjugate
provided a viable strategy against GNB infections.
.
2
2
. Materials and methods
.1 Materials
CS, with an average molecular weight of 50 kDa and a degree of
7

deacetylation of 95%, was purchased from Zhejiang Maya Reagent Co.,
Ltd. (China). N-acetylsulfanilyl chloride, Aminobenzene, N-thiocarbonyl
diimidazole and m-aminobenzoic acid were analytical grade and were
purchased from Energy Chemical (China). Fetal bovine serum (FBS),
Dulbecco's modified Eagle medium (DMEM), phosphate-buffered saline
(PBS), and other culture media were obtained from Gibco BRL (USA).
Escherichia coli (E. coli, ATCC 25992), clinical isolates of multidrug-
resistant Escherichia coli (MDR-E. coli, J-20), Pseudomonas aeruginosa
(
P. aeruginosa, J-11) and Staphylococcus aureus (S. aureus, ATCC
5423) were kindly gifted by Prof. Jikai Wen (South China Agricultural
University, China). Multidrug resistant clinical isolates from patients
from the First Affiliated Hospital of Zhongshan University) with
2
(
recurrent urinary tract infections were collected and isolated from urine,
and the samples were characterized to determine their probable species.
Clinical samples were enriched in LB broth and isolated in MacConkey
agar to differentiate Gram-negative bacteria. Isolates were classified via
API 20e kits (bioMérieux, France). Probable species include
Pseudomonas, Klebsiella and Enterobacter species. All other reagents
used were of a highly pure grade and were bought from Shanghai Energy
Chemical Co., Ltd. (China).
2
2
.2 Preparation of CS/LED conjugates
.2.1 Synthesis of meta-LED 209-COOH
The meta-LED 209-COOH was synthesized in accordance with the
synthetic routes illustrated in the previous literatures (Xue et al., 2015),
8

and more detailed descriptions of the method could be found in the
supplementary material.
2.2.2 Synthesis of CS/LED conjugates
The preparation of CS/LED conjugates was implemented by using
EDC/NHS referring to the relevant literatures (Guo, Anderson, Bozell, &
Zivanovic, 2016; Xie, Hu, Wang, & Zeng, 2014), with minor
modifications. Initially, CS (0.50 g, 3.10 mmol) was dissolved in 10 mL
1% (v/v) acetic acid, and stirred overnight to obtain a clear CS solution
that was preserved for further use. Secondly, three proportions of meta-
LED 209-COOH (0.66 g, 1.55 mmol; 1.32 g, 3.10 mmol; 2.65 g, 6.20
mmol) were dissolved in 10 mL anhydrous ethanol, and added EDC﹒
HCl (1.19 g, 6.20 mmol)and NHS (0.71g, 6.20 mmol) successively to
completely dissolve through stirring 1 h , then slowly added the foregoing
CS solution. Furthermore, the reaction was kept at ambient temperature
for 24 h. After the grafting was completed, the excess acetone was added
to the reaction bottle to precipitate the product with different degree of
substitution (DS) (named CS/LED-1, CS/LED-2, CS/LED-3). Finally, the
product was centrifuged at 12,000 × g for 20 min, washed 3 times with
anhydrous ethanol and deionized water respectively, and freeze dried.
2
.2.3 Synthesis of fluorescein isothiocyanate (FITC)-labeled CS and
CS/LED
The synthesis of FITC-labeled CS or CS/LED was based on the
reaction between the isothiocyanate group of FITC and the amino group
of CS. The attachment of the FITC to the CS was implemented using a
9

published method with minor modification (Moussa, Tayel, & Al-Turki,
013). In brief, the FITC (20 mg in 10 mL anhydrous methanol) was
2
added to 20 mL 1% (w/v) CS or CS/LED in 1% (v/v) acetic acid solution,
then the reaction mixture was stirred for 24 h at room temperature in the
dark. To remove the unreacted FITC, the solution was dialyzed using
dialysis membrane (MWCO: 3 kDa) against distilled water for 1 week.
After every 12 h, the dialysis medium was replaced with a fresh one.
After dialysis, the reaction mixtures were kept at -80 °C for 24 h followed
by freeze-drying for 24 h.
2
2
.3 Characterization of CS/LED conjugates
.3.1 FT-IR analysis
Fourier-transform infrared (FT-IR) spectra of CS and CS/LED were
obtained using a Bruker VERTEX 70 FTIR spectrometer (Bruker,
-
1
Germany) at the wave numbers ranging from 4000 to 400 cm . All the
samples were mixed with anhydrous KBr to produce dried discs for
spectrum recording at 25 °C.
1
2
.3.2 H NMR analysis
1H NMR spectra were received on a 600 MHz NMR (Bruker,
Germany). All the solutions (5mg/mL) were prepared by dissolving CS
and CS/LED in 1% (v/v) solution of CD COOD in D O and D O,
3
2
2
respectively. In addition, LED 209 dissolved in DMSO-d
Tetramethylsilane (TMS) was used as an internal standard.
.3.3 Elemental analysis
Elemental analysis was performed by using a PerkinElmer EA 2400
6
.
2
1
0

element analysis instrument (PE, USA). About 1 mg samples (CS and
CS/LED-1～3) were weighed by using an ultra-microbalance(Sartorius,
Germany), the samples were sealed and placed inside a nickel sleeve for
sample injection. Static combustion was conducted in pure oxygen,
meanwhile, Helium was used to carry the combustion products through
the analytical system. All the determinations were done in triplicate. The
particulate organic carbon and nitrogen mass percentages of CS and
CS/LED were obtained.
2.4 Zeta potential
The particle zeta potential of CS and CS/LED were measured at
5ꢀ°C using a Zetameter (Zetasizer nano Zs model, Malvern, UK), with a
2
refraction angle of 90°. All the solutions (1mg/mL) were prepared by
dissolving CS and CS/LED in deionized water and PBS, respectively, and
then detected the charges of CS and CS/LED by over 20 runs. Each
measurement was performed in triplicate.
2.5 Solubility test
25 mg CS, LED 209 and CS/LED-1~3 was dispersed in 5 mL
distilled water and different organic solvents including anhydrous
ethanol, acetone, diethyl ether, DMSO and DMF at room temperature,
respective. Afterwards, the suspension was ultrasonicated (300 W, 80
kHz) for 30 min to observe solubility of these compounds.
2.6 High performance liquid chromatography (HPLC)
The reverse-phase high-performance liquid chromatography (RP-
HPLC) was performed on an Agilent HPLC column (RP-C18, 250 × 4.6
11

mm) (Curcio et al., 2009; Xue et al., 2015).The mobile phase was 50:50
water/acetonitrile containing 0.5 % trifluoroacetic acid at a flow rate of
1.0 mL/min. Since it was used to analyze if there were any unreacted
small molecules in the macromolecular products, isocratic elution was
sufficient. The injection volume of samples was 20 μL. LED 209
identification was made by UV detection at 254 nm. Column temperature
and sample temperature was 25°C. The mobile phase was used as solvent
to dissolve and dilute the samples. First, the synthesized meta-LED 209-
COOH was used as sample 1 with a concentration of approximately 1
mg/mL. Second, 1mL from the last washing solution after grafting
dissolved into 10 mL mobile phase, and was used as sample 2. Last,
CS/LED dissolved in the aqueous solution of 1% (v/v) acetic acid as the
stock solution (10 mg/mL), and then the pre-preparation stock solution
was diluted two folds with sterile LB broth to add the selected wells (100
μL), and the microplate was incubated at 37 °C for 24 h. Subsequently, 50
μL solution was extracted from the selected well. After that, the extracted
solution dissolved into 1 mL mobile phase, and was used as sample 3. All
samples were filtered by microporous membrane (0.22 μm) before
testing. sample 1 was mainly used to determine the purity of the
compound meta-LED 209-COOH, sample 2 was mainly used to
determine whether the conjugation has diffused free meta-LED 209-
COOH, and sample 3 was mainly used to detect whether the polymer
(CS/LED) will release meta-LED 209-COOH in later activity tests.
2.7 Minimum inhibitory concentrations (MICs)
1
2

Two types of E. coli were employed to determine the MICs values of
CS and CS/LED using a micro-dilution assay in sterile 96-well
polypropylene microtiter plates according to the previous literatures
(Dang et al., 2018). In brief, CS and CS/LED-1~3 dissolved in the
aqueous solution of 1% (v/v) acetic acid as the stock solutions (10
mg/mL). The aqueous solution of 1% (v/v) acetic acid was used as a
control group. The inoculum of each bacterial suspension was prepared
by incubating the bacterial strain in LB broth at 37 °C for 12 h, followed
8
by a dilution until the OD610 value is 0.1 (～10 CFU/mL). After that, 100
mL of the bacterial suspension was added to each well in a 96-well
microplate. Subsequently, the pre-preparation stock solutions of CS and
CS/LED-1~3 were diluted two folds with new sterile LB broth to add an
identical volume, and the microplate was incubated at 37 °C for 24 h.
Finally, the absorbance at 610 nm of each well in the microplate was
recorded on a microplate reader (Bio-Rad, Hercules, CA, USA). For
confirmation, 50 μL from each well were spread on LB agar plates and
incubated at 37 °C for 24 h. After that, the number of Colony forming
units (CFUs) on LB agar plates was determined. The MICs tests were
repeated three times independently with duplicate samples each time.
2.8 Selectivity assay of CS/LED
Fluorescein isothiocyanate (FITC) labeled CS and CS/LED-3 were
employed to study the selectivity on the tested cells containing NIH-3T3
and MDR-E. coli. As previously reported (Xue et al., 2015), the detailed
procedure with slight modification was implemented according to the
1
3

following description.
4
In brief, NIH-3T3 cells (5 × 10 /mL) maintained in DMEM
containing 10% FBS were seed in sterile 24-well polypropylene
microplates with the cover glasses inside (14mm, Jingan Biological,
Shanghai, China), followed by incubation at 37 ℃for 12 h. After that, the
culture medium was removed and replaced with 0.5 mg/mL of CS-FITC
or CS/LED-FITC dissolving in DMEM. Subsequently, the microplates
were cultured in the darkness, and taken the cover glasses out of the
microplates at different time points. Then the cover glasses were washed
with PBS for three times, fixed with 4% paraformaldehyde for 30 min,
and examined by fluorescence microscopy. MDR-E. coli (～ 108
CFU/mL) cells were cultured in LB broth containing 0.5 mg/mL of CS-
FITC and CS/LED-FITC in sterile 96-well microplates at 37 ℃. The
samples were gathered by centrifugation at different time point, then
washed with PBS for three times, and added 50 μL 2.5% glutaraldehyde
to resuspend. At the last, the resuspended bacteria were dropped on the
glass slide, and air dried. Several images with the same exposure
parameters were taken from each glass under fluorescence microscopy,
and fluorescent quantification was achieved by using Image Pro Plus 6.0
software.
2.9 Cytotoxicity assay
The cell viability of CS and CS/LED to NIH-3T3 cells was
evaluated by a standard MTT assay (Abdulhadi Aljawish et al., 2014).
4
Briefly, NIH-3T3 cells (5 × 10 /mL) maintained in DMEM containing
1
4

10% FBS were cultured in sterile 96-well microplates at 37 ℃under 5%
CO atmosphere for 24 h. Subsequently, the medium in the microplates
2
was supplemented with 100 μL different concentrations of CS and
CS/LED solution. Then, the plates were incubated at 37 °C for 24 h.
Ultimately, the absorbance at 540 nm of each well in the microplates was
recorded on a microplate reader.
2
2
.10 Anti-adhesion assay
.10.1 Colony forming units (CFUs) assay
One of the most widely used approaches to estimate microbial
viability is the determination of CFUs on agar media (Azeredo et al.,
017). Therefore, CFUs assay was employed to assess the total number of
2
adherent bacteria indirectly according to previous literatures (Amato et
4
al., 2018). Briefly, MDR-E. coli (5 × 10 /mL) maintained in LB broth
containing the cellulose dressings, coated by 500 µL of 1 mg/mL CS and
CS/LED, were cultured in 24-well microplates at 37 ℃for 24 h. In
addition, 500 µL of 1 mg/ mL meta-LED 209-COOH was added into
wells within uncoated cellulose membrane, and cultured under the same
conditions. After that, the coated dressings were carefully taken out of the
microplates, washed with PBS for three times, and put in glass tubes
containing 5 mL PBS, then detached adhering bacteria through sonication
for 5 min. After that, the ultrasonic solutions were centrifuged, diluted by
PBS and placed (50 µL) onto LB agar plates. After 20 h incubation at
37 °C, the number of CFUs on LB agar plates was determined.
2.10.2 Scanning Electron Microscopy (SEM) assay
1
5

The anti-adhesion ability of CS and CS/LED was evaluated by CS
and CS/LED coating commercial cellulose dressings. Therein, SEM has
been used to directly assess microbial adhesion extent in previous reports
(Amato et al., 2018; Richardson, Mordan, Figueiredo, Ng, & Gulabivala,
2009). In brief, the pre-prepared CS and CS/LED solutions (500 µL, 1
mg/mL) were dropped on the commercial cellulose dressings (φ14 mm),
the sterile water as the control. After drying, the cellulose dressings were
placed in 24-well microplates, and then added 1 mL MDR-E. coli (～ 108
CFU/mL) cells that were suspended in LB broth before. Following
incubation at 37 °C for 24 h, and taken carefully the dressings out of the
microplates, washed with PBS for three times, fixed with 2.5%
glutaraldehyde for 30 min, and gold sputtered. Several pictures were
taken from each dressing under SEM.
2.11 Statistical analyses
SPSS v.13.0.0 software (SPSS, Inc., Chicago, IL) was used for the
statistical analysis. Results are expressed as mean values (± SD).
Student's test and one-way analysis of variance (ANOVA) were used to
evaluate statistical significance as appropriate. A value of P< 0.05 was
considered statistically significant.
3
3
. Results and discussion
.1 Preparation of CS/LED conjugates
The meta-LED 209-COOH was synthesis via the route illustrated in
Scheme 1. Given the complex structure of the meta-LED 209-COOH, we
1
6

divided it into two parts to synthesize respectively. The thiourea in the
structure of the meta-LED 209-COOH could be obtained by the
condensation of isothiocyanate and amino group, therefore, we
synthesized compound IV containing amino group and compound V
containing isothiocyanate group to prepared our final compound VI. The
more detailed descriptions and detection data about the compounds from
Scheme 1 had been exhibited in the supplementary material.
3.2 Characterization of CS/LED conjugates
Due to its highly efficient and mild reaction conditions,
carbodiimide based chemical coupling reagents, such as 1-ethyl-3-(3-
dimethylaminopropyl) carbodiimide (EDC) and N- hydroxysuccinimide
(NHS) have been widely used for the chemical conjugation of CS with
phenolic acids (Liu, Pu, Liu, Kan, & Jin, 2017). The mechanism of the
grafting reaction by using EDC and NHS is illustrated in enormous
reports (Liu et al., 2016; Pasanphan & Chirachanchai, 2008). In general,
the grafting ratio of this reaction is significantly affected by reaction
conditions including initial molar ratio of CS/phenolic acid, reaction
temperature, reaction time and so on (Aytekin, Morimura, & Kida, 2011).
Among that, the most important factor is the initial molar ratio of
CS/phenolic acid. In addition, the excessive EDC and NHS contribute to
improve the grafting efficiency of the reaction (Woranuch & Yoksan,
2013). Accordingly, we selected three types of molar ratio of CS/LED
(2:1, 1:1, 1:2) to explore the grafting ratio of this reaction.
1
7

FT-IR spectroscopy is considered as a useful tool to confirm the
successful grafting of phenolic acid onto CS, and it can provide precise
structural information about the grafting pattern of CS/phenolic acid (Liu
et al., 2017). The FT-IR spectra of CS and CS/LED-1～3 were illustrated
in Fig. 1A. In the FT-IR spectrum of the pure CS, the characteristic bands
−
1
at 1654, 1591 and 1321 cm were observed, corresponding to C=O
stretching, N-H bending and C-N stretching, respectively(Lim & Hudson,
−
1
2
004). Moreover, the absorption bands at 1420 and 1380 cm were due
to CH bending and CH symmetrical deformation, respectively (Zhang et
al., 2009). Compared with the spectrum of CS, a new absorption band at
2
3
−
1
1
640 cm was observed in the spectrum of CS/LED-1～3, indicating the
presence of secondary amide (−CO−NH−) (Pasanphan & Chirachanchai,
008). This new band revealed the formation of amide bond between the
amino group of CS and the carboxyl group of meta-LED 209-COOH. In
2
−
1
addition, there was a new absorption band at 1724 cm in CS/LED
conjugates, indicating the ester bond formed between hydroxyl groups of
CS and carboxyl groups of meta-LED 209-COOH (Liu, Lu, Kan, Tang, &
Jin, 2013; Rui et al., 2017). These results indicated that LED had been
successfully conjugated onto CS chains.
1
NMR spectroscopy, especially H NMR spectroscopy, is believed to
be a reliable approach to prove the successful synthesis of CS/phenolic
1
acid (Xie et al., 2014). The H NMR spectrum of CS, CS/LED-3 and
1
LED 209 were illustrated in Fig. 1B. In the H NMR spectrum of the pure
CS, the characteristic single peak at 3.0 ppm (H-2), 4.4 ppm (H-1) were
1
8

observed. Furthermore, the multiple peaks at 3.3-3.7 ppm (H-3, H-4, H-5,
H-6) verified the protons of glucosamine units (Eom, Senevirathne, &
Kim, 2012). Apparently, compared with CS, the spectrum of CS/LED-3
possessed a series of new peaks in the range of 6.0–8.0 ppm, which could
be confirmed to the protons of benzenoid moieties in LED 209 (Liu et al.,
2015). Therefore, coupled with FT-IR spectrum above, we could
guarantee the successful grafting of LED 209 with CS.
Although the successful grafting of LED 209 onto CS could be
proved by the above-mentioned methods, the detailed information about
the degrees of substitution (DS) of CS/LED was still unclear. For the
derivatives of CS, its biological activity is significantly related to its DS.
In addition, there are many methods for calculating its DS. Elemental
analysis is generally applied to determine DS of CS/phenolic acid (Dos
Santos, Caroni, Pereira, Da Silva, & Fonseca, 2009). It was assumed that
the degree of acetylation (DA) of CS remains unchanged during the
reaction. Thus, the molecular formula of pure CS can be expressed as
[(C
6
H
11NO
4
) (C
2
H O) DA] n. In addition, meta-LED 209-COOH could
2
react with free amino groups in CS, resulting in the change of its DS, so
the molecular formula of CS/LED samples (CS/LED-1~3) can be
expressed as [(C
6
H
11NO
4
) (C
2
H
2
O) DA (C20
H
15
N
3
O
3
2
S ) DS] n. Because of
the trace water in the samples interfere with the determination of oxygen
mass percentages of the samples, it is an appropriate way to calculate the
DS of CS/LED by the mass ratio between carbon and nitrogen in
1
9

elemental analysis. Therefore, according to the above the molecular
formula of CS and CS/LED, the DS of the samples could be calculated by
formula (1~3): (J. Wang & Wang, 2011; Yue et al., 2017):
total mass of carbon in CS (6+2×DA)×MC
=
total mass of nitrogen in CS MN
RC⁄N
=
(
(
(
1)
2)
3)
total mass of carbon in CS⁄LED
(6+2×DA+20×DS)×Mꢀ
=
total mass of nitrogen in CS⁄LED
(1+3×DS)×Mꢁ
R′C⁄N
=
(
ꢄꢅꢀ⁄ꢁ−ꢄꢀ⁄ꢁ)×Mꢁ
ꢂꢃ = 2
0×M −3×M ×ꢄ
ꢅ
ꢀ
ꢁ
ꢀ⁄ꢁ
In formulas (1～3), DS is degree of substitution. DA is degree of
acetylation of CS. M
nitrogen. RC⁄N is mass ratio between carbon and nitrogen of CS and
Rꢆ
is mass ratio between carbon and nitrogen of CS/LED-1～3.
C
and M are relative atomic mass of carbon and
N
C⁄N
The DS of the samples was estimated by the percentage of
particulate organic carbon and particulate organic nitrogen as
demonstrated in Table 1. The results showed that when the mole ratio of
LED 209 in the reaction was increased, the DS was increased. In general,
the grafting ratio of phenolic acid-g-chitosan is greatly affected by
reaction conditions. Many factors will affect the grafting ratio of phenolic
acid-g-chitosan(Aytekin et al., 2011; Schreiber, Bozell, Hayes, &
Zivanovic, 2013). Obviously, in this study, the initial molar ratio of
2
0

CS/meta-LED 209-COOH greatly affected the grafting ratio of the
conjugates. When the molar ratio was 3:1, the DS increased to 0.24,
which was the largest percentage of CS/LED-1～3. In addition, the
solubility of CS/LED-1～3 was investigated in six common solvents
including distilled water, anhydrous ethanol, acetone, diethyl ether,
DMSO and DMF. The results showed that CS did not dissolve in the
solvents, however, when conjugated with LED 209 facilitated increasing
solubility with increase in DS. The phenomena could be explained as that
the introduction of meta-LED 209 thwarted the formation of the hydrogen
bonding among the hydroxyl groups of CS, thereby the conjugates
become easy to dissolve in solvents (Guo et al., 2016). Taking into
account its good solubility and high grafting rate, CS/LED-3 was chosen
as the experimental object, and which was expressed as CS/LED in the
following content.
In addition, the zeta potential measurement indicated the charge
values of CS and CS/LED were 1.36 ± 0.98 and 0.69 ± 0.25 mV in
deionized water, respectively, and 2.69 ± 0.71 and -1.05 ± 0.42 mV in
PBS, respectively. And the result demonstrated the nearly neutral nature
of CS and CS/LED. The free amino group in CS is the key to its
bioactivity. Under physiological conditions, these free amino groups will
be protonated and transformed into positive charges. Correspondingly the
introduction of LED 209 may cause the decrease of the number of
protonated amino groups and shielding of the positive charges. Our
results proved this possibility. The zeta potential positive charge of
2
1

CS/LED was less than that of CS. This result provided a guarantee for
later bioactivity experiments that positive charge was no longer the
primary factor affecting CS/LEDs.
3.3 High performance liquid chromatography (HPLC)
To confirm that whether the conjugates CS/LEDs release free
meta-LED 209-COOH in in the further experiments, HPLC was
chosen to prove this issue. HPLC systems equipped with reversed-
phase C18 column and UV detector have been used to distinguish
phenolic acid-g-chitosan from free phenolic acid in many
researches(Curcio et al., 2009; Liu et al., 2017). In addition, HPLC
analysis was often performed after the dialysis of reaction solution.
As shown in Fig. 2. The retention time of meta-LED 209-COOH is
1
2
2
.549 min, which is similar to the previous research result(Xue et al.,
015). In addition, The HPLC results showed the purity of meta-LED
09-COOH is 95.3%. Meanwhile, free meta-LED 209-COOH was
not detected in sample 2 and sample 3. These results demonstrated
that the conjugation had not diffused free meta-LED 209-COOH.
Combined with the above results, we believed that CS/LED had not
released meta-LED 209-COOH in further antibacterial experiments.
3.4 Antibacterial activities of CS/LED
To exam whether LED 209 conjugation with CS still remain the
antibiotic activity, MICs was determined to evaluate the ability of CS and
CS/LED against E. coli, MDR-E. coli, P. aeruginosa and S. aureus.
2
2

Among them, P. aeruginosa is Gram-negative bacteria without the QseC
receptor, and S. aureus is Gram-positive bacteria. As illustrated in Fig. 3,
the MICs values of CS/LED against P. aeruginosa and S. aureus were
obvious weaker than that of CS, mainly due to the decrease of the number
of protonated amino groups and shielding of the positive charges by LED
2
09 chains. Moreover, CS and CS/LED were similar sensitive to MDR-E.
coli and E. coli. According to precious literatures (A. Aljawish et al.,
014; Božič, Gorgieva, & Kokol, 2012; Chatterjee et al., 2015), the
antibacterial activity of CS was mainly because of the interaction between
positive charge on -NH groups and negatively charged bacteria
2
2
membranes. Compared with the pure CS, the introduction of LED 209,
which did not inhibit pathogen growth, decreased the number of free
amino groups (-NH ) in CS, which endowed CS/LED with much lower
2
antibacterial activity against P. aeruginosa and S. aureus. Meanwhile, the
introduction of LED 209, which possessed highly selective to the QseC
receptor existing in GNB, might enhance the targeting ability of CS to E.
coli and MDR-E. coli result in the antibacterial properties of CS/LED
could still remain.
3.5 Selectivity of CS and CS/LED on MDR-E. coli and NIH-3T3 cells
To study the selectivity of CS and CS/LED on the tested cells,
Fluorescein isothiocyanate (FITC) is chosen owing to its outstanding
ability to react with the amino group and easy to observe. The synthesis
of FITC-labeled CS or CS/LED was based on the reaction between FITC
2
3

isothiocyanate group and CS primary amino group. Due to the chemical
link between CS and FITC, the novel conjugate could be used as a
fluorescent agent for cell labeling through the green fluorescence
observed (Mi et al., 2007).
As demonstrated in Fig. 4A, compared with CS-FITC, the
fluorescence signal of CS/LED-FITC was more detected in MDR-E. coli,
and with the extension of time (1 h to 3 h), the differentia was more
remarkable, moreover, from 2 h, the concentration of CS/LED-FITC
adhering to bacteria membranes was higher than that of CS-FITC. The
results were in agreement with our hypothesis that introduction of LED
209 would promote CS to target the MDR-E. coli. It is well established
that LED 209 is a highly selective inhibitor of QseC, which is mainly
existed in most GNB but not in mammalian cells (Rasko et al., 2008).
Hence, the conjugation of LED 209 to CS enhanced the selectivity effects
of CS on GNB. As shown in Fig. 4B, Compared with CS-FITC, the
fluorescence signal of CS/LED-FITC was less detected in NIH-3T3 cells,
especially, the difference in fluorescence signal between CS-FITC and
CS/LED-FITC was utterly obvious at 8 h. The introduction of LED 209
to CS was correlated with the reduced positive charges on CS, which
might lead to a reduction in the sorption of CS/LED on mammalian cell.
To sum up, these results were consistent with our hypothesis that
introduction of LED 209 would promote CS to target GNB and reduce
the interaction with mammalian cells.
2
4

3.6 Cytotoxicity of CS and CS/LED
The cell viability of CS and CS/LED were illustrated in Fig. 5, the
cell viability of CS/LED to NIH 3T3 cells was similar to that of
unmodified CS. Notably, the cytotoxicity of CS/LED decreased slightly
compared with that of unmodified CS at high concentrations (5-1.25
mg/mL). The previous studies suggested that LED 209 in vitro and in
vivo did not present any evidence of toxicity, besides, it also had desirable
pharmacokinetics when administered orally to mice (Curtis et al., 2014).
In addition, CS was also a kind of safe and low toxic biomaterial.
Together, these data suggest that CS/LED have better selectivity on GNB
and minor cytotoxicity to mammalian cells.
3.7 Anti-adhesion assay
The anti-adhesion activity of the commercial cellulose dressings
coated with CS or CS/LED was assessed by culturing in MDR-E. coli for
4 h. The number of CFUs adherent for CS and CS/LED after 24 h of
2
incubation was demonstrated in Fig. 6, it is obvious that meta-LED 209-
COOH (B) had an anti-adhesion effect at the test concentration (1
mg/mL), and its effect is similar to that of pure CS. Moreover, the results
depicted that the anti-adhesion activity of the dressing coated with CS
was significantly lower compared to the uncoated dressing by reason of
the effective antibacterial activity of CS. Notably, dressings coated with
CS/LED possessed the stronger ability to prevent microbial adhesion
2
5

compared to the CS-coated dressing. The data displayed CS/LED could
reduce the adhesion of bacteria to the cellulose membrane by about
67.8%, while CS only reduced by about 45.3%. These findings were
determined by SEM in Fig. 7. Apparently, A large number of colonies
were observed on the surface of the uncoated dressing (Fig. 7B),
moreover, the mature biofilm could be seen distinctly (Fig. 7C). The
bacterial aggregates could also be observed on the surface of the CS-
coated dressing, but the number of bacteria colonies was less than that of
the uncoated dressing. However, although MDR-E. coli could be
adsorbed on the surface of the dressing coated with CS/LED, the further
aggregates did not form. According to the above findings, CS could
prevent bacterial adhesion by killing bacteria, but it could not hamper the
formation of the biofilm ulteriorly. Furthermore, CS/LED could
effectively prevent the formation of the biofilm to reduce bacterial
adhesion.
The main purpose of the present study is to explore novel strategies
to modify CS, which can significantly strength the antibacterial activity
of CS. Due to the most Gram-negative pathogens depends on the
virulence factors which is mainly controlled by QseC receptor (such as E.
coil), CS was conjugated with LED 209, a specific inhibitor of QseC
receptor, to generate a potent and selective antimicrobial biomaterial
which can target QseC receptor with lower cytotoxicity. Encouragingly,
CS/LED effectively inhibited the growth of most of the tested bacterial
strains. The MIC values of CS/LED were similar to that of CS, mainly
2
6

due to the decrease of the number of protonated amino groups and
shielding of the positive charges by LED 209 chains. In addition,
CS/LED showed higher selectivity to MDR-E. coli, indicating that the
three-dimensional structure of CS did not greatly affect the activity of
LED 209-COOH. Based on our results, we concluded that the
introduction of LED 209 enhanced the selectivity of CS to MDR-E. coli,
thereby increasing the concentration of the biomaterial aggregated on the
bacterial surface with the extension of time. This indicated the potential
of this biomaterial in targeting therapy for Gram-negative bacterial
infection. In addition, previous studies suggested that according to the
perceptible changes in host neurotransmitters, the pathogens could
modulate their virulence gene expression to ensure optimal host
colonization. Therefore, hindering the virulence gene expression of
bacteria could reduce bacterial adhesion and prevent the formation of
biofilm (Defoirdt, 2018; Johnson & Abramovitch, 2017; Kimmey &
Stallings, 2016). LED 209, as a novel anti-virulence prodrug, possessed
inhibitory effect on bacterial adhesion. The anti-adhesion effect of LED
209 also made CS/LED more advantageous in resisting bacterial
adhesion. Accordingly, our findings indicated that the CS/LED conjugates
might be helpful for solving the problem of GNB biofilm, which had
been believed to be an alarming threat to public health. Furthermore, the
means for the invading pathogen to gauge the physiological and
immunological state of the host are promising. Since these methods do
generally not inhibit microbial growth, interference with this signaling
2
7

will only stop virulence, not kill the microorganism. Consequently, this
anti-virulence approach may engender a lower evolutionary pressure
toward the development of drug resistance (Cloutier et al., 2015; Koo,
Allan, Howlin, Stoodley, & Hall-Stoodley, 2017). Relevant investigations
are in progress.
4
. Conclusions
In this study, CS/LED with different DS was synthesized via
1
carbodiimide based chemical coupling method. The FT-IR, H NMR and
elemental analysis confirmed the structure of CS/LED. Besides, CS/LED
exhibited a better solubility compare to CS. Our studies illustrated that,
compared with CS, CS-LED 209 showed higher selectivity to MDR-E.
coli and much lower selectivity to mammalian cell, as well as remained
low cytotoxicity to mammalian cell. Moreover, further research had
indicated that the conjugation could significantly hinder the formation of
MDR-E. coli biofilm to prevent bacterial adhesion. These finding
indicated that the three-dimensional structure of CS did not greatly affect
the activity of LED 209, and this conjugation could still maintain the
properties of CS and LED 209. Hence, this study demonstrated that the
conjugation of CS with LED 209 is a feasible strategy to synthesize a
multifunctional antimicrobial agent, which displays selective anti-GNB
and anti-adhesion activity.
2
8

Acknowledgments
This research was financially supported by the NNSFC31570964,
GDNSF2016A0303138, 201508020126, 2013 JCYJ20130401174221011,
Marine Economy Innovation and Development Project (GD2012-D01-
002), Marine Fishery Science and Technology Promotion Project of
Guangdong (A1201301C06).
References
Aljawish, A., Chevalot, I., Jasniewski, J., Revol-Junelles, A. M., Scher, J.,
&
Muniglia, L. (2014). Laccase-catalysed functionalisation of
chitosan by ferulic acid and ethyl ferulate: evaluation of
physicochemical and biofunctional properties. Food Chemistry,
161, 279-287.
Aljawish, A., Muniglia, L., Jasniewski, J., Klouj, A., Scher, J., &
Chevalot, I. (2014). Adhesion and growth of HUVEC endothelial
cells on films of enzymatically functionalized chitosan with
phenolic compounds. Process Biochemistry, 49(5), 863-871.
Amato, A., Migneco, L. M., Martinelli, A., Pietrelli, L., Piozzi, A., &
Francolini, I. (2018). Antimicrobial activity of catechol
functionalized-chitosan versus Staphylococcus epidermidis.
Carbohydrate Polymers, 179, 273-281.
Anitha, A., Sowmya, S., Kumar, P. T. S., Deepthi, S., Chennazhi, K. P.,
Ehrlich, H., Jayakumar, R. (2014). Chitin and chitosan in selected
2
9