Colloids and Surfaces B: Biointerfaces 117 (2014) 346353

Contents lists available at ScienceDirect

Colloids and Surfaces B: Biointerfaces

journal homepage: www.elsevier.com/locate/colsurfb

A simple and general method for preparing antibody-PEG-PLGA

sub-micron particles using electrospray technique: An in vitro study
of targeted delivery of cisplatin to ovarian cancer cells
Meng-Yi Bai a,b, , Sheng-Zhong Liu a
a
b

Graduate Institute of Biomedical Engineering, National Taiwan University of Science and Technology, Taipei 10607, Taiwan, ROC
Adjunct appointment to the Department of Biomedical Engineering, National Defense Medical Center, Taipei 11490, Taiwan, ROC

a r t i c l e

i n f o

Article history:
Received 24 September 2013
Received in revised form 28 January 2014
Accepted 23 February 2014
Available online 12 March 2014
Keywords:
Electrospray
Nanotechnology
Controlled release
Drug delivery
Biomaterials

a b s t r a c t
Cisplatin-encapsulating maleimide-polyethylene glycol- Poly(d,l-lactic-co-glycolide) (cis-encapsulating
mal-PEG-PLGA) particles were produced using the electrospray technique and bioconjugated with CD44
monoclonal antibody, targeting the counterpart receptor. The produced suspension of cis-encapsulating
CD44-PEG-PLGA particles contains an antibody loading of 12.6515.17 g/mL and efciently targets a
CD44-overexpressed ovarian cancer cell line, such as CP70 and SKOV-3, within 6 h of treatment, which
was determined by Bradford assay, immunouorescence analysis, and confocal laser scanning microscopic (CLSM) study. Most importantly, no tedious multi-step bioconjugation procedures are needed to
synthesize mal-PEG-PLGA vehicles for antibody and drug loading, avoiding the undesirable hydrolysis
of mal-PEG moiety and so successfully generating the cis-encapsulating mal-PEG-PLGA particles within
one step. After conjugation of the CD44 antibody, the produced cis-encapsulating CD44-PEG-PLGA particles exhibited a better anti-proliferative ability against ovarian cancer cells compared to free form of
cisplatin and PLGA particles without CD44 conjugation. Notably, the cis-encapsulating CD44-PEG-PLGA
particles have approximately 10-14% greater the anti-proliferative ability against CP70 and SKOV-3 cells
at a concentration of 1.25 M, which falls within the concentrations used in chemotherapy. The proposed antibody-functionalization strategy represents an excellent platform for preparing particles with
targeting ability in cancer therapy in vitro or in vivo.
2014 Elsevier B.V. All rights reserved.

1. Introduction
Owing to the poor selectivity of the chemotherapeutic agent
toward normal cells and cancerous cells, most patients who receive
chemotherapy experience side effects such as depressed immune
system, hair loss, and gastrointestinal distress. Increasing the selective uptake of chemotherapeutic agents by tumor tissue would
be a promising mean of overcoming these drawbacks. Following
the discovery of specic markers that are selectively expressed
on cancerous cells or expressed at a much lower level in normal
cells than in cancerous cells, many targeting approaches have been
developed, such as an active targeting method that is based on
the ligandreceptor interaction. The active targeting method covalently conjugates the targeting moiety (e.g., an antibody) that is
a ligand for a specic marker via an appropriate spacer to the

Corresponding author at: TR-917, AAEON Building, no. 43 Keelung Road, Section
4, Taipei 10607, Taiwan, ROC. Fax: +886 2 2730 3733.
E-mail address: mybai@mail.ntust.edu.tw (M.-Y. Bai).
http://dx.doi.org/10.1016/j.colsurfb.2014.02.051
0927-7765/ 2014 Elsevier B.V. All rights reserved.

drug carrier. For instance, Dhara et al. designed a prostate specic

membrane antibody (PSMA) aptamer-polyethylene glycol (PEG)poly(d,l-lactic-co-glycolide) (PLGA) nanoparticles, for the targeted
delivery of cisplatin to prostate cancer cells [1]. Rafat et al. reported
on PEG-PLA microparticle for the delivery of a transactivator of
transcription-enhanced green uorescent protein (Tat-EGFP) to
retinal cells [2]. Hu et al. demonstrated lactoferrin conjugated PEGPLGA nanoparticles for delivery to the brain, and demonstrated
its efcacy in the treatment of Parkinsons disease [3]. Recently,
many works have also developed HER-2 (Human Epidermal Growth
Factor Receptor Type 2) antibody-anchored PEG-PLGA nanoparticles for targeting Her-2 positive breast cancer in vitro or in vivo
[4,5]. Some cases simply used the PEG-PLGA particles as carriers of proteins, such in cases of cystic brosis, or for pulmonary
delivery [68]. Although all of these PEG-PLGA particles with or
without antibody conjugation indeed showed superior targeting of
the site of interest or killing of cancer cells while having less of
an effect on normal cells, they require the application of a tedious
coupling protocol to conjugate antibody or PEG onto the surface
of the PLGA particles to enable the ligandreceptor interaction

M.-Y. Bai, S.-Z. Liu / Colloids and Surfaces B: Biointerfaces 117 (2014) 346353

or to prevent uptake by the reticuloendothelial system (RES),

respectively [9]. The typical coupling protocol conjugates functional groups between the antibody/or PEG and PLGA particles. This
kind of covalent bonding can provide a stable and persistent link
that immobilizes the antibody or PEG but usually requires a tedious
procedure to construct the antibody-PEG-PLGA particles. Our previous studies have shown that the electrospray system (ES) can
generate drug encapsulating PLGA particles with monodispersed
size [10]. The aim of this work is to translate the strategy to a simple
method for fabricating antibody-PEG-PLGA particles.
The CD44-PEG-PLGA sub-micron particles were formed by
ES technique. These PLGA-based particles were also prepared
using conventional bioconjugation technique and compared to
ES-generated ones. According to a series of characterizations
that involved Fourier transform infrared spectroscopy (FTIR) and
nuclear magnetic resonance spectroscopy (NMR), the maleimide
moiety (an end group of the starting material maleimide-PEG for
antibody conjugation) was determined to be easily hydrolyzed
using the conventional bioconjugation protocol. In contrast, the
maleimide and PEG moiety were simultaneously incorporated into
the ES-generated PLGA particles and remained intact throughout
the ES process. To the best of our knowledge, our study is the rst
work to use the ES technique to generate the antibody-PEG-PLGA
particles for targeted delivery.
2. Materials and methods
2.1. Materials
Cis-diammineplatinum
(II)
dichloride
(cis,
crysSigmaAldrich),
dicyclohexylcarbodiimide
(DCC,
talline,
99%,
SigmaAldrich),
N-hydroxysuccinimide
(NHS,
98%,
SigmaAldrich), and N,N-diisopropylethylamine (DIPEA, 99.5%,
SigmaAldrich) were used as purchased without purication.
The solvents used in the experiments were acetone (Ace, ACS
reagent, Mallinckrodt), acetonitrile (ACN, 99.9%, J. T. Baker),
dichloromethane (DCM, 99%, TECO Inc., Taiwan), and dimethylformamide (DMF, 99.0%, SigmaAldrich). The encapsulation
material was poly(D,L-lactic-co-glycolide) acid terminated (PLGA,
lactide:glycolide = 50:50, Mw 700017,000, SigmaAldrich);
the crosslinker was maleimide-polyethylene glycol-NH2 (malPEG-NH2 , Mw 3500, Jenkem Technology, China); the stabilizer
was poly(vinyl alcohol) (PVA, average Mw 13,00023,000, 98%
hydrolyzed, SigmaAldrich). The CD44 monoclonal antibody
(CD44, eBioscience, Australia) was used as purchased for antibody
conjugation. The CD44-FITC monoclonal antibody (CD44-FITC,
eBioscience, Australia) was used as purchased for immunouorescence staining. Imject maleimide conjugation buffer (pH 7.2,
Thermo Scientic, USA) and a sulfhydryl addition kit (kit no.
23460, Thermo Scientic, USA) were used to add thiol groups
to CD44 antibody. Coomassie Plus (Bradford) assay kit (kit no.
23236SPCL, Thermo Scientic, USA) was used for Bradford assay. A
1 PBS solution was prepared using phosphate buffer saline (PBS,
10X, UniRegion Bio-tech) by diluting 1 mL of 10X PBS in 9 mL of
deionized H2 O and stirring for 1015 min. A 1 wt.% of PVA was
added to the 1X PBS solution to stabilize the dispersion of particles
(designated as PBS(PVA)).
2.2. Preparation of mal-PEG-PLGA using conventional
bioconjugation technique
The following standard synthesis method was used. First, 0.03 g
of PLGA was dissolved in 4 mL of DCM. Next, 0.38 g of DCC and
0.21 g of NHS was added to the PLGA solution to activate the carboxylate groups on the terminal of PLGA chain. After thorough

347

mixing, the reaction mixture was sealed in a 50 mL round-bottom

ask and deoxygenated under a vacuum. The reaction mixture was
then stirred at 200 rpm at room temperature for 4 h. After the predetermined reaction time, the reaction mixture was poured into
ice-cold ether/methanol cosolvent (80/20, v/v) for precipitation.
After decanting the supernatant, the isolated PLGA precipitate was
further washed with 20 mL of ice-cold ether/methanol co-solvent
to remove excess DCC and NHS. Before the step of mal-PEG-NH2
conjugation, the activated PLGA precipitate was freeze-dried and
preserved in powder form. Subsequently, 0.09 g of activated PLGA
powder, 0.12 mL of DIPEA, and 0.013 g of mal-PEG-PLGA were
co-dissolved in 12 mL of DCM. Again, the reaction mixture was
immediately sealed and deoxygenated under a vacuum. The reaction mixture was then stirred at 200 rpm at room temperature for
17 h. Finally, 50 mL of the ice-cold ether/methanol cosolvent (80/20,
v/v) was poured into the reaction mixture to isolate the resultant
mal-PEG-PLGA. The educt powders were separated by centrifugation at 15000 rpm for 12 min. After removing the supernatant,
20 mL of ice-cold ether/methanol co-solvent was used for washing.
The nal precipitate was freeze-dried to obtain the crude product of
mal-PEG-PLGA. For further purication, this crude product was dissolved in 2 mL of DCM and then poured into the 40 mL of ice-cold
ether/methanol co-solvent for further precipitation. The puried
mal-PEG-PLGA was used in the subsequent characterizations.

2.3. Fabrication of cis-encapsulating mal-PEG-PLGA or

cis-encapsulating CD44-PEG-PLGA particles using ES system
The ES system was set up as described elsewhere [10]. To
form cis-encapsulating mal-PEG-PLGA particles, 5 mg of cis was codissolved with 0.0040.016 g of mal-PEG-NH2 and 0.1 g of PLGA in
2.5 mL of cosolvent (Ace/DMF = 1/4, v/v) to generate a stock solution
for ES processing. The stock solution was fed from a syringe pump
(New Era Pump System, Inc., Syringe Pump NE-300). The capillary
tube used in this experiment had an inner diameter (ID) of 0.012 in.
(300 m) and an outer diameter (OD) of 0.031 in. (785 m). A
spray electrical eld was established between the capillary tube
and particle collection substrate by applying a positive voltage to
the spray head assembly using a DC high-voltage power supply
(Bertan, Model 250B-20R) and electrically grounding the collection
substrate. Depending on the application, the collection substrate in
this setup may be a metal plate or a saturated cis-containing imject
maleimide conjugation buffer (4 mg/mL). The capillary tube was
approximately 8.5 cm from the collection substrate, and the applied
voltage ranged from 8 to 11 kV. Next, the electrospray modes of
the system were monitored by viewing the liquid meniscus at
the exit of the capillary tube. Additionally, the meniscus was illuminated with a diffusing light, and its shape was observed with
a microscopic system that consisted of a microscope lens (Retro
Zoom 65, Model: 30-43-10), a digital camera (Model: STC-620PWT,
SENTECH), and a high resolution monitor. The CD44-PEG-PLGA particles were produced by using the abovementioned mal-PEG-PLGA
particles as precursors because the maleimide moieties on the particle are excellent sites for protein coupling. Firstly, commercially
available CD44 protein was converted to sulfhydryl-modied protein using the sulfhydryl addition kit (Thermo Fisher Scientic Inc.,
USA). The sulfhydryl-modied protein was produced according
to the manufacturers instructions. The as-synthesized sulfhydrylmodied CD44 protein and the pre-formed mal-PEG-PLGA particles
were immediately mixed together in an imject maleimide conjugation buffer (Thermo Fisher Scientic Inc., USA) for coupling
thiol-containing antibody to mal-PEG-PLGA particles. After 2 h of
reaction, the entire reaction mixture was subject to centrifugation to obtain CD44-PEG-PLGA particles and the precipitate was
washed with imject maleimide conjugation buffer to eliminate the

348

M.-Y. Bai, S.-Z. Liu / Colloids and Surfaces B: Biointerfaces 117 (2014) 346353

unbonded sulfhydryl-modied CD44 protein. Finally, the obtained

CD44-PEG-PLGA particles were freeze-fried and stored at 20 C.
2.4. In vitro anti-proliferative abilities of a series of
cis-encapsulating PLGA particles on CP70 and SKOV-3 cell lines
The CP70 and SKOV-3 cells were maintained as monolayer cellular cultures in RPMI-1640 medium (Biological Industries Inc.),
supplemented with 10% fetal bovine serum (Gibco), 0.1% NaHCO3
(Sigma), 0.1% HEPES (Sigma), 1% NEAA (Biological Industries Inc.),
1% sodium pyruvate (Biological Industries Inc.), and 0.1% penicillinatreptomycin (Biological Industries Inc.) at 37 C in a 5% CO2
atmosphere. To produce the culture of CP70 cells, an extra addition of 1% insulin was given in the RPMI-1640 medium. In a typical
culture procedure, the aforementioned cells were seeded onto a 96well plate at a concentration of 5 104 cells/mL (100 L/well) in the
RPMI growth medium. The cells were incubated for 24 h at 37 C in a
5% CO2 atmosphere. After 24 h of culturing, 25 L of 100 M, 50 M,
25 M, 12.5 M, and 6.25 M of cis-encapsulating PLGA particle
suspensions were added to each well to treat the cells with the drug.
Since each well already contained 100 L of RPMI growth medium,
the nal drug concentrations in the wells were 20 M, 10 M, 5 M,
2.5 M, and 1.25 M, respectively, after thorough mixing of the
cis-encapsulating PLGA particle suspensions and growth medium.
Following 72 h of incubation, the supernatant was removed from
each well and the cells were washed with 100 L of PBS solution.
After gentle shaking for thorough washing, 100 L of PBS solution was withdrawn from each well. Subsequently, 120 L of MTS
solution was added to each well. The 96-well plate was then incubated for another 0.51 h. After gentle shaking of the well plate,
the optical density (OD) of the contents of each well was recorded
using an Epoch spectrophotometer at a wavelength of 490 nm. The
cell viability was calculated as the ratio of the recorded OD values
according to the following equation.
Cell viability =

OD of drug-treated group
100%
OD of medium only group

2.5. Receptor expression assays

The CD44 expression levels were qualitatively and quantitatively determined by confocal laser scanning microscopy (CLSM)
and immunouorescent staining. For CLSM imaging, 5 103 cells
were seeded in a glass-bottom dish (diameter, 35 mm). The cells
were incubated for 24 h at 37 C in a 5% CO2 atmosphere. Subsequently, the medium was removed from the dish and the attached
monolayer of the cellular cultures at the bottom of the dish was
xed using 4 wt% of paraformaldehyde for 15 min at 37 C [11].
Afterward, all medium was removed and the xed monolayer of
cellular cultures was washed with 100 L of PBS solution. Finally,
the xed monolayer of cellular cultures was co-incubated with
100 L of CD44-FITC RPMI solution (250 g/mL) at 4 C for 30 min.
After the predetermined time, the nuclei of cells were stained
with 10 g/mL 4 ,6-diamidino-2-phenylindole (DAPI; Invitrogen),
according to the manufacturers instructions, and were imaged
using the 10 objective lens of an LSM 510 confocal microscope
(Zeiss). For immunouorescent staining, cells were seeded in a 96well plate at a concentration of 2 105 cells/mL (100 L/well) in
the RPMI growth medium. The cells were incubated for 24 h at 37 C
in a 5% CO2 atmosphere. After 24 h of culturing, the medium was
removed from each well, and the attached monolayer of cellular
cultures at the bottom of the dish was washed with PBS solution.
Next, the attached monolayer of cellular cultures was co-incubated
with 100 L of CD44-FITC (250 g/mL) at 4 C for 30 min. Subsequently, the medium was removed, and the attached monolayer of
the cellular cultures was washed thrice with PBS solution. Finally,

0.1 mL of fresh serum-free RPMI medium was added to each well.

Cells were then analyzed using a luminescence scanner to read the
emission of uorescence (ex = 485 nm, ab = 538 nm).
2.6. Characterizations
The morphologies and structures of the mal-PEG-PLGA or
CD44-PEG-PLGA particles were analyzed using a scanning electron
microscope (SEM) that was operated at an accelerating voltage of
20 kV (JSM-6390L, JEOL). For charge dissipation purposes, the SEM
sample specimens were coated with an additional thin layer of
Their 1 H nuclear
Pt lm (with a thickness of approximately 80 A).
magnetic resonance (NMR) spectra were acquired using a Bruker
600 MHz high-resolution NMR spectrometer. All samples were dissolved in d-dimethyl sulfoxide. The attenuated total reectance
infrared (ATR-IR) spectra of the freeze-dried mal-PEG-PLGA particles were acquired using a Digilab FTS-1000 FTIR spectrometer that
was equipped with an ATR accessory. A constant clamping pressure was applied to mount all specimens to ensure optimal and
consistent surface contact with the diamond probe head. All ATRIR spectra were acquired from 36 cycles of accumulation to yield
a favorable S/N ratio. The calibration line of the CD44 antibody for
quantitative purpose was established according to the instructions
for use of the Bradford assay kit, and the optical density at 595 nm
of the CD44-PEG-PLGA particle suspension was recorded using an
Epoch Spectrophotometer (BioTek instrument, USA). The amount
of cis encapsulation in the PLGA particles was determined using
high performance liquid chromatography (HPLC, Jasco, JAPAN). The
HPLC was equipped with the following components: an analytical
column (DIKMA, C18 column, 150 mm 4.6 mm, beads: 5 m); an
autosampler (model AS-2055, Jasco, JAPAN); a pump (model PU2080, Jasco, JAPAN) set to deliver 1 mL/min; a UV detector (model
UV 2070; Jasco, JAPAN) set at 308 nm (which is the maximum
absorption wavelength of cis), and a data acquisition system (SISCQchrom). The mobile phase (ACN/DMF = 1/1, v/v) was pumped at
1 mL/min. The cis was detected by UV absorption at 308 nm at
a retention time of 1.67 min. Concentrations (g/L) of cis were
obtained from the integration of peak area of cis and the peak area
thus obtained was used in the calculation using a six-point calibration curve. All types of cis encapsulating PLGA particles were
pre-dissolved in 1 mL of mobile phase (ACN/DMF = 1/1, v/v). After
passing the resultant solutions through a 0.2 m polytetrauoroethene (PTFE) lter, a 20 L quantity of the solution was injected
into the HPLC for further analysis.
3. Results and discussion
3.1. Preparation of mal-PEG-PLGA using conventional
bioconjugation technique via a two-step coupling protocol
Fig. 1 shows a two-step coupling protocol for the synthesis
of mal-PEG-PLGA. Carboxylic acid-terminated PLGA was chosen
as the starting material. First, the carboxylic acid groups on the
PLGA chain were activated by DCC/NHS. Subsequently, the aminecontaining molecules, mal-PEG-NH2 , were added to react with
activated PLGA. The above protocol for coupling amine-containing
molecules to carboxylated particles is a viable starting point for a
method of immobilizing the desired protein and has been widely
used for bioconjugation. Fig. S1 shows the ATR-IR and NMR spectra
of the DCC/NHS-activated PLGA reactive intermediate. Fig. S1a
shows the ATR-IR spectrum of pure PLGA for comparison. Two
of the most prominently characteristic peaks of PLGA are one at
1777 cm1 , which is associated with the vibration mode of the
carbonyl group (C O group of the ester linkage) and one from
1000 cm1 to 1300 cm1 , which is a broad band associated with

M.-Y. Bai, S.-Z. Liu / Colloids and Surfaces B: Biointerfaces 117 (2014) 346353

Fig. 1. A two-step protocol using covalent coupling to carboxylate PLGA for the
synthesis of mal-PEG-PLGA.

the COC stretching vibration [12]. Fig. S1b shows the ATR-IR
spectrum of the DCC/NHS-activated PLGA reactive intermediate.
Fig. 1 shows that two reactive intermediates may have been
generated after the DCC/NHS activation step: a DCC-terminated
PLGA and an NHS-terminated PLGA. Therefore, a newly generated
secondary amine vibration mode at 1524 cm1 and the representative vibration mode of NHS at 1666 cm1 were observed in Fig. S1b.
Fig. S1c shows the DCC/NHS-activated PLGA, obtained by treating
PLGA with alkaline solution to generate more carboxylic acid
groups on PLGA for DCC/NHS activation. However, alkaline-treated
PLGA did not substantially increase the substitution of DCC/NHS.
Additionally, the 1 H proton NMR of the DCC/NHS-activated PLGA,
presented in Fig. S1e, revealed characteristic peaks of PLGA at 1.48
(3H, CH3 ), as determined by comparison with its mother PLGA
(Fig. S1d). Two extra prominently peaks at 2.76, and 2.81 ppm
were attributed to protons on the ve-member ring of NHS and the
proton on the tertiary carbon atom of the DCC ring, respectively.
Peaks that are marked by triangular symbols were associated with
the residue of DCC and ether [peaks at 2.59 (1H, CH, singlet),
2.92 (3H, CH3 , triplet), and 3.40 ppm (2H, CH2 , quartet)].
This 1 H proton NMR study of the DCC/NHS-activated PLGA was
consistent with the results of NMR prediction using ChemDraw
software (see Table S1S3). Fig. S1f shows the NMR spectrum of
DCC/NHS-activated PLGA, obtained by treating PLGA with alkaline
solution to generate more carboxylic acid groups on PLGA for
DCC/NHS activation but no further increase of DCC/NHS substitution was observed. Accordingly, the successful activation of the
carboxylate groups on the PLGA chain using the DCC/NHS protocol
was conrmed. Mal-PEG-NH2 was subsequently added to couple
with the DCC/NHS-activated PLGA intermediate. Fig. S2 shows the
ATR-IR and NMR spectra of the DCC/NHS-activated PLGA reactive
intermediate (Fig. S2a) after it has reacted with mal-PEG-NH2 (Fig.
S2c). Fig. S2b shows the ATR-IR spectrum of pure mal-PEG-NH2 for

349

comparison. The amine-containing molecule was expected to couple with the activated carboxylate groups at the terminal of PLGA
chain to generate an amide linkage (CONH). Therefore, a small
peak corresponding to the NH stretching vibration was observed
at 3512 cm1 , and an absorption peak associated with the carbonyl
group was observed at 1735 cm1 (Fig. S2c). These peaks conrmed
the successful immobilization of the mal-PEG-NH2 chain onto
activated carboxylate groups at the terminal of the PLGA chain to
generate a mal-PEG-PLGA product. The C C of maleimide group
was designed for the subsequent antibody conjugation. However,
Fig. S2c shows that two characteristic peaks of maleimide were
absent: one near 1600 cm1 , which corresponds to the stretching vibration of C C, and one near 30003100 cm1 , which
corresponds to the unsaturated CH stretching vibration. Fig.
S2df shows the series of 1 H proton NMR spectrum of DCC/NHSactivated PLGA, mal-PEG-NH2 , and mal-PEG-PLGA acquired with
a 600 MHz high-resolution NMR spectrometer, which further
conrmed the unexpected disappearance of the maleimide end
group. As shown in Fig. S2f, the 1 H proton NMR spectrum of the
as-prepared mal-PEG-PLGA did not show the characteristic 1 H
peaks around 7.0 that are characteristics of the maleimide group,
unlike the 1 H proton NMR spectrum of the mal-PEG-NH2 (Fig.
S2e). However, in contrast to the 1 H proton NMR spectrum of
the DCC/NHS- activated PLGA (Fig. S2d), a strong resonance peak
at 3.5 ppm (hydrocarbon of methylene groups of PEG chain) was
observed after the reaction with mal-PEG-NH2 . The above studies
indicated that the conventional bioconjugation technique obtained
the mal-PEG-PLGA but apparently lost the maleimide group.
The spectral analysis results in Fig. S1S2 indicate that the
maleimide was likely lost from the PEG linkage using the commonly
used bioconjugation protocol. As Fig. S3a shows, we hypothesize
that the maleimide group underwent a hydrolysis reaction when
the pH value of its environment fell within the more alkaline
pH reaction condition (>pH 8.0), resulting in the conversion of
maleimide to maleic acid. Fig. S3bd shows the 1 H proton NMR
spectra of the supernatant of the mal-PEG-PLGA reaction mixture
after centrifugation, for maleimide, and for maleic acid, respectively. Although unsaturated protons on maleimide moiety were
characterized by a resonance peak at 6.88 ppm, this peak was not
observed in the spectrum of the as-prepared mal-PEG-PLGA (Fig.
S2f). Conversely, a small peak that was shifted upeld to 6.35 ppm
was observed in the 1 H proton NMR spectrum of the supernatant
(Fig. S3b). To assign this peak appropriately, the NMR spectra of
two model compounds (maleimide and maleic acid) were acquired
(Fig. S3cd). The chemical shifts of the unsaturated protons on
the maleimide ring and maleic acid were at 6.88 and 6.27 ppm,
respectively. These results conrm the speculation that the protein coupling site (maleimide group) readily leaves the PEG linkage
owing to the hydrolysis reaction. This observation arises because
of the long existing problem of the commonly used bioconjugation
technique that the mal-PEG is not quite stable in alkaline environment and can undergo ring opening to generate maleic acid
[13,14].
3.2. Preparation of cis-encapsulating CD44-PEG-PLGA particles
using ES technique and sulfhydryl CD44 antibody
Hydrolysis of the maleimide group is a serious problem that has
a large effect on the following antibody coupling; thus, a simple
and reliable method of preparing CD44-PEG-PLGA is needed. Fig. 2
displays a simple protocol for synthesizing CD44-PEG-PLGA polymer and in-situ producing its drug-encapsulating particles. Firstly,
all required materials, including PLGA, mal-PEG-NH2 , and cis were
combined into a stock solution for use in the ES process to generate a cis-encapsulating mal-PEG-PLGA particle suspension. To
prevent the hydrolysis of mal-PEG-NH2 , a commercially available

350

M.-Y. Bai, S.-Z. Liu / Colloids and Surfaces B: Biointerfaces 117 (2014) 346353

Fig. 2. A facile protocol for synthesizing CD44-PEG-PLGA polymer and in-situ producing its drug-encapsulating particles.

conjugation buffer with a pH value approximately 7.2 was used

as a particle collection solution. A preformed sulfhydryl CD44
antibody was then added to the cis-encapsulating mal-PEG-PLGA
particle suspension for antibody coupling. This new protocol does
not require tedious bioconjugation, and cis-encapsulating malPEG-PLGA particles are generated in a single step. Although the
mal-PEG-NH2 was not covalently bonded to the PLGA, its long
hydrocarbon chain became entangled with that of the PLGA, and
some of the mal-PEG was outcropped on the surface of the particle.
These moieties that remain on the surface of the particle provide
the functions of pegylation and antibody coupling.
3.3. Characterization of cis-encapsulating mal-PEG-PLGA
particles
Fig. 2 shows that the cis-encapsulating mal-PEG-PLGA particles
were rstly prepared for subsequent conjugation with the antibodies. Fig. 3 displays the surface morphology (SEM images) of the
cis-encapsulating mal-PEG-PLGA particles, which were spherical
and had an average diameter of 550 137 nm. The exact chemical composition of the produced cis-encapsulating mal-PEG-PLGA
particles was identied by FT-IR and NMR spectroscopic characterizations (see Fig. S4). Fig. S4ad compare the FT-IR spectra of
the cis-encapsulating mal-PEG-PLGA particles with those of PLGA,
mal-PEG-NH2 , and cisplatin. Fig. S4d shows the characteristic peaks
of all three components. (C O vibration mode at 1790 cm1 , C C
vibration mode at 1669 cm1 , CH vibration mode at 2996 cm1 ,
and NH stretching mode at 32003500 cm1 ). Notably, the characteristic peak of C C that is attributed to the maleimide moiety
is weak and markedly overlaps with the vibration mode of C O.
To ensure reliable detection of the maleimide moiety, an alternative NMR spectroscopic method was used (Fig. S4eh). Fig. S4h
shows the NMR spectrum of the cis-encapsulating mal-PEG-PLGA
particles. As a major difference from what has been reported in
contrast with Fig. S2f (i.e., production of mal-PEG-PLGA using

Fig. 3. The SEM images of the cis-encapsulating mal-PEG-PLGA particles (avg. diameter 550 137 nm) recorded under magnication ratios of (a) 10,000, and (b)
5000.

M.-Y. Bai, S.-Z. Liu / Colloids and Surfaces B: Biointerfaces 117 (2014) 346353

351

conjugation technique), Fig. S4h presents the simultaneous

presence of cis encapsulation (4.11 ppm, 4.39 ppm, 4.67 ppm,
H6 Cl2 N2 Pt) and a maleimide moiety (6.90 ppm, H2 C CH2 ), both
of which are required for targeted drug delivery. The benets of
the proposed ES technique are its avoidance of a tedious synthetic
procedure for incorporating the mal-PEG chain into the PLGA material and its simultaneous encapsulation of the drug. Specically, it
avoids hydrolysis of the maleimide moiety by conventional bioconjugation, and the cis-encapsulating mal-PEG-PLGA particles are
obtained within a single step.
3.4. Characterization of cis-encapsulating CD44-PEG-PLGA
particles
The cis-encapsulating CD44-PEG-PLGA particles were synthesized using the preformed cis-encapsulating mal-PEG-PLGA
particles and sulfhydryl CD44 antibody. The antibody used herein
has a thiol functional group (-SH), which readily reacted with the
maleimide group via the formation of a thioether bond, such that
the antibody was nally conjugated to the maleimide-modied
substrate. Fig. S5d displays the dynamic light scattering result of
the cis-encapsulating CD44-PEG-PLGA particles, which shows an
average diameter of approximately 634 37 nm. This conjugation
is based on a Michael addition reaction, and the corresponding
mechanism and protocol is well studied and established [15]. Fig. S5
shows the samples of our produced cis-encapsulating CD44-PEGPLGA particles that were suspended in a conjugation buffer (see
optical image), and reveals that the antibody conjugation was in
the range 12.6515.17 g/mL, based on the Bradford assay. Firstly,
a calibration line of the CD44 antibody was established according
to the instructions for the Bradford assay kit (Fig. S5c), yielding
the equation Y = 0.01 + 0.00672X, R2 = 0.998, where Y is the OD of
the cis-encapsulating CD44-PEG-PLGA particles suspension and X
is the concentration of the CD44 antibody (g/mL). Fig. S5a shows
the the absorption spectrum of the cis-encapsulating CD44-PEGPLGA particle suspension after addition of Coomassie Brilliant Blue
G-250. The spectrum reveals a historic maximum approximating
595 nm with an optical density of 0.112 (15.17 g/mL, as calculated using the aforementioned equation). The amount of antibody
immobilized on the surface of the particles was further conrmed
by centrifuging the above cis-encapsulating CD44-PEG-PLGA particle suspension to obtain the supernatant that could contain the
free form of the antibodies (unreacted or detached). Fig. S5b shows
the absorption spectrum of the supernatant, which shows a weak
absorption peak at approximately 595 nm and an optical density
of 0.027 (approximately 2.52 g/mL, as calculated using the above
equation). Based on this quantitative assay of protein, the antibodies that were conjugated to the cis-encapsulating CD44-PEG-PLGA
particles fall within the concentration range 12.6515.17 g/mL.
3.5. In vitro inhibition of CP70 and SKOV-3 tumor cell growth
using cis-encapsulating CD44-PEG-PLGA particles
The encapsulation of cis in particles was used to determine
whether CD44-PEG-PLGA particles could serve as a translational
platform for cancer therapy. Several formulations were compared,
including (1) a free form of cis, (2) cis-encapsulating PLGA particles,
(3) cis-encapsulating PEG-PLGA particles, and (4) cis-encapsulating
CD44-PEG-PLGA particles dispersed in PBS(PVA) at different concentrations for use against the ovarian cancer cell lines CP70 and
SKOV-3. Fig. 4 shows the results of quantitative analyses of the
cytotoxicity of the aforementioned formulations toward CP70 and
SKOV-3. These analyses were performed under the same degree
of CD44 conjugation. After treating the CP70 and SKOV-3 cell
lines with varying concentrations of cis formulations for 72 h
(1.2520 M), the MTS assay was performed. Firstly, cells that

Fig. 4. Cis-encapsulating CD44-PEG-PLGA particle suspension effectively suppresses ovarian cancer cell proliferation at concentrations of 1.2520 M in (a) CP70
cell line and in (b) SKOV-3 cell line. After treating cells with a series of cis formulations for 72 h, MTS assay was performed to determine the cell viability. All cell
viabilities shown in this gure are normalized to control cells treated with equivalent amounts of RPMI-1640 medium. The PBS(PVA) represents the mixture of PBS,
PVA and RPMI-1640 medium. The cell viability of each group was expressed as
mean SD (n = 3/group). The p values * p < 0.05, ** p < 0.01, and *** p < 0.001 was considered statistically signicant with respect to the group of medium. The software
Sigma-Plus was employed for statistical analysis.

had been treated with equal amounts of PBS(PVA) medium, void

PLGA particles, or void PEG-PLGA particles were tested as controls
and found to have no signicant cytotoxic effect on cell viability of CP70 and SKOV-3. Although all four foumulations exhibited
dose- dependent anti-proliferative abilities within the concentration of micromolar range (1.2520 M), all cis-encapsulating
PLGA particles, with or without pegylation, exhibited poorer antiproliferative ability compared to the free cis formulation at all
tested concentrations. The only exceptions were cis-encapsulating
CD44-PEG-PLGA particles in CP70 and SKOV-3.This observation
reasonably corresponded to the drug release kinetics and the
cellular uptake efciency. Firstly, in the cis-encapsulating PLGA
particles, with or without pegylation, drug release by diffusion
was much slower compared to the free form of cis. Secondly,
the cell uptake efciency in the cis-encapsulating CD44-PEG-PLGA
particles-treated group was possibly much greater than that of the
non-CD44 conjugated PLGA particles-treated group, indicating that

352

M.-Y. Bai, S.-Z. Liu / Colloids and Surfaces B: Biointerfaces 117 (2014) 346353

Fig. 5. Confocal images of (ac) CP70 and (df) SKOV-3cells incubated with CD44-FITC antibody for 30 min. The images from left to right direction correspond to the image
of nuclei stained with DAPI (blue), membrane stained with CD44-FITC (green) and merging of both. (For interpretation of the references to color in this gure legend, the
reader is referred to the web version of the article.)

more cis were engulfed by the cells. This nding suggests that the
cytoxicity of the cis-encapsulating CD44-PEG-PLGA particles was
possibly dominated by the receptor-mediated endocytosis, which
was responsible for intracellular drug delivery.
3.6. Receptor expression assays
To elucidate that the intracellular uptake of the drug vehicle
via receptor-mediated endocytosis was primarily responsible for
the superior cytotoxicity of the cis-encapsulating CD44-PEG-PLGA
particles, a series of cell studies were conducted to visualize the
penetration of the particles into the cells and the targeting effect of
the particles that were conjugated with the CD44 antibody. Fig. 5
shows a quantitative evaluation of the expression of the CD44
receptor on the cell membrane of the CP70 and SKOV-3 cells. First,
both cell lines were stained with CD44-FITC antibody via the ligandreceptor interaction. After thorough washing, the stained cells were
characterized using confocal laser scanning microscopy (CLSM).
Fig. 5af shows the CLSM results for both cell lines. The image
acquired at the FITC excitation wavelength shows green uorescence. As shown in Fig. 5, immunouorescence analysis revealed
that CD44 were expressed in both CP70 and SKOV-3 cell line, but the
immunouorescence response of SKOV-3 was apparently stronger
than that of CP70. To quantitatively determine the CD44 receptors that were expressed on the CP70 and SKOV-3 cell membranes,
CD44-FITC antibody was used to stain the cells that were seeded in
a well plate. The cells were then washed with PBS, and a luminescence scanner was used to measure the emission of uorescence
(ex = 485 nm, ab = 538 nm). This in vitro analysis revealed that the
extent of CD44 expression was approximately 1.47 times higher in
SKOV-3 cells than that in CP70 cells (Fig. S6, *** p < 0.001), which is
consistent with our previous immunohistochemical staining study
(see Fig. 5c and f).

3.7. In vitro cellular uptake: confocal microscopic study

The CP70 and SKOV-3 were treated with either cisencapsulating CD44-PEG-PLGA particles or with cis-encapsulating
mal-PEG-PLGA particles to evaluate their cellular uptake behavior
and the efciency of cellular uptake. Based on the drug release
prole study (Fig. S7), cells were co-cultured with the aforementioned particles for 6 h to ensure the release of the drug from
the carriers. Fig. S8 shows a series of confocal laser scanning
microscopic (CLSM) images of the CP70 and SKOV-3 cells after 6 h
co-incubation with either cis-encapsulating mal-PEG-PLGA particles or cis-encapsulating CD44-PEG-PLGA particles after washing
thrice with PBS. The cells were then xed, and their membranes
were stained (cellmaskTM , red stain, Invitrogen) for CLSM imaging.
Apparently, PLGA particles without CD44 antibody conjugation did
not bind to either CP70 or SKOV-3 cells, and most of the particles
were removed during the washing step. Fig. S8 (columns a,b,
designated as z-stack image) shows the deep z-scanning confocal
uorescence images obtained to visualize the penetration of PLGA
particles without CD44 antibody conjugation. The images were
taken at the midpoint of the cells treated with uorescent PLGA
particles without CD44 antibody conjugation. Notably, the z-stack
images of CP70 and SKOV-3 cells reveal no particle uptake by the
cells. Conversely, Both CP70 and SKOV-3 cells treated with uorescent cis-encapsulating CD44-PEG-PLGA particles revealed green
uorescence emission around each cell even after washing step
(see Fig. 6ab, as indicated by the arrowhead), which indicated that
adsorption of the uorescent cis-encapsulating CD44-PEG-PLGA
particles may have resulted from a ligand-receptor interaction.
Additional deep z-scanning confocal uorescence images acquired
at the midpoint of the both cells from the same specimens (Fig.
S8cd, designated as Z-stack image) under the same exciting
laser intensity and the same confocal microscope revealed green

M.-Y. Bai, S.-Z. Liu / Colloids and Surfaces B: Biointerfaces 117 (2014) 346353

353

Fig. 6. Confocal laser scanning microscope (CLSM) images of the (a) CP70 and (b) SKOV-3 cells after 6 h co-incubation with cis-encapsulating CD44-PEG-PLGA particles.

uorescence dots in the cytoplasm of both CP70 and SKOV-3 cells.

These cells substantially differ from those treated with PLGA particles without CD44 antibody conjugation (Fig. S8ab, designated
as Z-stack image). Since expression of the CD44 receptor was 1.47
times higher in the SKOV-3 cells than in the CP70 cells, the SKOV-3
cells were greatly shrunk upon 6 h of treatment with the drug, and
they became spherical, but the CP70 cells retained a spindle-like
morphology of living CP70 cells. We believed that the receptormediated endocytosis facilitates and enhances the uptake of
particles into cells when the CD44-conjugated PLGA particles come
into contact with the overexpressed CD44 receptors on the surfaces of the cells. Accordingly, the susceptibility of SKOV-3 cells to
cisplatin was greatly improved, and the difference between the cell
viability of the cis-encapsulating CD44-PEG-PLGA particle-treated
group and the group treated with cis-encapsulating mal-PEG-PLGA
particles was greater than that in CP70 group. In summary, the
anti-proliferative ability of the cis-encapsulating CD44-PEG-PLGA
particles against CP70 and SKOV-3 cells was improved by 1014%
of upon treatment with 1. 25 M of cis, which concentration
falls within the range of concentrations that is clinically used in
chemotherapeutic injections (Kemoplat Injection, 0.81.6 M).
4. Conclusions
This work used ES technique to develop a simple and general
method of generating cis-encapsulating CD44-PEG-PLGA particles
with controlled sizes and excellent stability without precipitation in aqueous solution. Most importantly, the method requires
no tedious multi-step bioconjugation process and eliminates the
common problem of hydrolysis of the maleimide-PEG linker.
Furthermore, this work demonstrated that the cis-encapsulating
CD44-PEG-PLGA particles exhibited an enhanced cytotoxic effect,
such that the effective dosage of such particles against ovarian cancer cell lines was lower than the effective dosages of free form of
cis, cis-encapsulating PLGA particles, and cis-encapsulating PEGPLGA particles. This cis-encapsulating CD44-PEG-PLGA particles
can be used as drug carriers for the effective targeted delivery of
a chemotherapeutic agent in the case of CD44-positive ovarian
cancer therapy.
Acknowledgments
This work was partly funded by the National Taiwan University
of Science and Technology. Some of the research work was performed at the Research Facility of the National Defense Medical
Center. M.Y. Bai would like to thank the National Science Council

of Taiwan for nancial support under contract no. (NSC-101-2221E-011-053- and NSC-102-2221-E-011-070-). M.Y. Bai authored the
manuscript and prepared all gures. M.Y. Bai, and S. Z. Liu designed
and performed all experimental procedures, including setup of the
electrospray system, materials preparation, and protein assays.
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/j.colsurfb.
2014.02.051.
References
[1] S. Dhar, N. Kolishetti, S.J. Lippard, S.C. Farokhzad, Targeted delivery of a cisplatin
prodrug for safer and more effective prostate cancer therapy in vivo, Proc. Natl.
Acad. Sci. U. S. A. 108 (2011) 1850.
[2] M. Rafat, C.A. Clroux, W.G. Fong, A.N. Baker, B.C. Leonard, M.D. OConnor, C.
Tsildis, PEG-PLA microparticles for encapsulation and delivery of Tat-EGFP to
retinal cells, Biomaterials 31 (2010) 3414.
[3] K. Hu, Y. Shi, W. Jiang, J. Han, S. Huang, X. Jiang, Lactoferrin conjugated
PEG-PLGA nanoparticles for brain delivery: Preparation, characterization and
efcacy in Parkinsons disease, Int. J. Pharm. 415 (2011) 273.
[4] Y. Liu, K. Li, B. Liu, S.S. Feng, A strategy for precision engineering of nanoparticles
of biodegradable copolymers for quantitative control of targeted drug delivery,
Biomaterials 31 (2010) 9415.
[5] R. Dhankar, P. Rathee, A.K. Jain, S. Arora, M.S. Kumar, G. Rath, A.K. Saxena,
P.R. Sharma, G. Chashoo, A.K. Goyal, HER-2 targeted immunonanoparticles for
breast cancer chemotherapy, J. Appl. Pharm. Sci. 01 (2011) 132.
[6] Y.P. Li, Y.Y. Pei, X.Y. Zhang, Z.H. Gu, Z.H. Zhou, W.F. Yuan, J.J. Zhou, J.H. Zhu, X.J.
Gao, PEGylated PLGA nanoparticles as protein carriers: synthesis, preparation
and biodistribution in rats, J. Control. Rel. 71 (2001) 203.
[7] N. Vij, T. Min, R. Marasigan, C.N. Belcher, S. Mazur, H. Ding, K.T. Yong, I. Roy,
Development of pegylated PLGA nanoparticle for controlled and sustained drug
delivery in cystic brosis, J. Nanobiotechnol. 8 (2010) 22.
[8] B. Patel, V. Gupta, F. Ahsan, PEG-PLGA based large porous particles for pulmonary delivery of a highly soluble drug, low molecular weight heparin, J.
Control. Rel. 162 (2012) 310.
[9] L.B. Peppas, J.O. Blanchette, Nanoparticle and targeted systems for cancer therapy, Adv. Drug Deliv. Rev. 56 (2004) 1649.
[10] Y.H. Lee, F. Mei, M.Y. Bai, S. Zhao, D.R. Chen, Release prole characteristics
of biodegradable-polymer-coated drug particles fabricated by dual-capillary
electrospray, J. Control. Rel. 145 (2010) 58.
[11] J. Su, X.H. Xu, Q. Huang, M.Q. Lu, D.J. Li, F. Xue, F. Yi, J.H. Ren, Y.P. Wu, Identication of cancer stem-like CD44+ cells in human nasopharyngeal carcinoma
cell line, Arch. Med. Res. 42 (2011) 15.
[12] Z.X. Meng, Y.S. Wang, C. Ma, W. Zheng, L. Li, Y.F. Zheng, Electrospinning of
PLGA/gelatin randomly-oriented and aligned nanobers as potential scaffold
in tissue engineering, Mater. Sci. Eng. C 30 (2010) 1204.
[13] R.G. Barradas, S. Fletcher, J.D. Porter, The hydrolysis of maleimide in alkaline
solution, Can. J. Chem. 54 (1976) 1400.
[14] G. Shen, M.F.G. Anand, R. Levicky, X-ray photoelectron spectroscopy and
infrared spectroscopy study of maleimide-activated supports for immobilization of oligodeoxyribonucleotides, Nucl. Acids Res. 32 (2004) 5973.
[15] G.T. Hermanson, Bioconjugate Techniques, second ed., Academic Press, Amsterdam, 2008 (Chapter 18).