Biomaterials 41 (2015) 53e68

Contents lists available at ScienceDirect

Biomaterials
journal homepage: www.elsevier.com/locate/biomaterials

Dual tumor targeting with pH-sensitive and bioreducible

polymer-complexed oncolytic adenovirus
Chang Yoon Moon a, 1, Joung-Woo Choi a, 1, Dayananda Kasala a, 1, Soo-Jung Jung a,
Sung Wan Kim a, b, Chae-Ok Yun a, *
a
b

Department of Bioengineering, College of Engineering, Hanyang University, 222 Wangsimni-ro, Seongdong-gu, Seoul, Republic of Korea
Center for Controlled Chemical Delivery, Department of Pharmaceutics and Pharmaceutical Chemistry, University of Utah, Salt Lake City, UT 84112, USA

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 27 August 2014
Accepted 8 November 2014
Available online

Oncolytic adenoviruses (Ads) have shown great promise in cancer gene therapy but their efcacy has
been compromised by potent immunological, biochemical, and specic tumor-targeting limitations. To
take full advantage of the innate cancer-specic killing potency of oncolytic Ads but also exploit the
subtleties of the tumor microenvironment, we have generated a pH-sensitive and bio-reducible polymer
(PPCBA)-coated oncolytic Ad. Ad-PPCBA complexes showed higher cellular uptake at pH 6.0 than pH 7.4
in both high and low coxsackie and adenovirus receptor-(CAR)-expressing cells, thereby demonstrating
Ad-PPCBA's ability to target the low pH hypoxic tumor microenvironment and overcome CAR dependence for target cell uptake. Endocytic mechanism studies indicated that Ad-PPCBA internalization is
mediated by macropinocytosis instead of the CAR-dependent endocytic pathway that internalizes naked
Ad. VEGF-specic shRNA-expressing oncolytic Ad complexed with PPCBA (RdB/shVEGF-PPCBA) elicited
much more potent suppression of U87 human brain cancer cell VEGF gene expression in vitro, and
human breast cancer MCF7 cell/Matrigel plug vascularization in a mouse model, when cancer cells had
been previously infected at pH 6.0 versus pH 7.4. Moreover, intratumorally and intravenously injected
RdB/shVEGF-PPCBA nanocomplexes elicited signicantly higher therapeutic efcacy than naked virus in
U87-tumor mouse xenograft models, reducing IL-6, ALT, and AST serum levels. These data demonstrated
PPCBA's biocompatibility and capability to shield the Ad surface to prevent innate immune response
against Ad after both intratumoral and systemic administration. Taken together, these results demonstrate that smart, tumor-specic, oncolytic Ad-PPCBA complexes can be exploited to treat both primary
and metastatic tumors.
2014 Elsevier Ltd. All rights reserved.

Keywords:
Cancer gene therapy
Oncolytic adenovirus
pH-sensitive
Bioreducible polymer
Systemic administration
Tumor microenvironment

1. Introduction
In the last two decades, oncolytic adenoviral (Ad) approaches
have evolved as promising therapeutic strategies against cancer
because of their numerous biochemical advantages [1e4]. A primary advantage is that Ads are cancer-specic. Innate defense
mechanisms against viral replication inside normal cells include
endogenous tumor suppression proteins such as p53, pRb, and
p14ARF. These proteins are defunct in certain cancer cells,
establishing the selectivity of Ad against these tumors [5]. Not
only do Ads destroy their host cells directly at the end of their lytic
* Corresponding author. Department of Bioengineering, College of Engineering,
Hanyang University, 222 Wangsimni-ro, Seongdong-gu, Seoul, Republic of Korea.
Tel.: 82 2 2220 0491; fax: 82 2 2220 4850.
E-mail address: chaeok@hanyang.ac.kr (C.-O. Yun).
1
These authors contributed equally to this study.
http://dx.doi.org/10.1016/j.biomaterials.2014.11.021
0142-9612/ 2014 Elsevier Ltd. All rights reserved.

cycles, but they also facilitate the spread of their viral progeny
throughout the tumor through cell lysis, leading to further
infection and subsequent eradication of neighboring cancerous
cells [6,7]. In addition to their oncolytic properties, Ads are among
the most potent viral vectors in gene therapeutics. Ads have high
gene transduction efciency in both dividing and non-dividing
cells, an effective nuclear entry mechanism, and the capacity to
be concentrated at high titers [8]. As a result, several clinical trials
have reported success using local Ad administration for cancer
gene therapy [9e11].
Despite their favorable characteristics, Ad-based treatments face
numerous challenges in realizing maximum therapeutic potential
with systemic administration. When delivered systemically, Ad is
rapidly cleared from the bloodstream through three major immunological mechanisms: 1) innate immune response induced by
macrophages and dendritic cells, 2) adaptive immune response by

54

C.Y. Moon et al. / Biomaterials 41 (2015) 53e68

neutralizing anti-Ad antibody, and 3) recruitment of viral particles

by Kupffer cells in the liver, causing hepatocyte infection and acute
liver toxicity. Moreover, Ad interacts with platelets and erythrocytes, potentially causing further side effects and toxicity [12]. In
addition to immunological barriers, Ad's transduction efciency is
severely limited by the level of coxsackie and adenovirus receptor
(CAR) expression in tumor cells. The CAR is a 46 kDa transmembrane protein that specically interacts with Ad surface proteins and is the primary mechanism through which Ad enters
target cells. Secondary interactions in Ad infection involve virus
binding via Arg-Gly-Asp (RGD) motifs on their capsid penton bases
with target cell surface receptors, especially avb3 and avb5 integrins.
Viral binding lead to clathrin-coated pit formation on the target cell
surface, which facilitates virus internalization inside an endosome
[13]. Due to Ad dependency on the CAR for infecting cells, the
therapeutic efcacy of Ad intratumoral and systemic injections are
compromised in cancer types whose cells express low levels of CAR.
One strategy that has been devised to overcome those
biochemical and immunological barriers inherent to using the Ad
vector for anti-cancer gene therapy is to develop smart Ad
nanocomplexes by modifying the viral surface with non-viral
systems [14e17]. These non-viral additions advantageously
compensate for native Ad vector limitations by conferring the
virus with increased tumor-specicity, reduced immunogenicity,
effective reproducibility, and a simple quality control process [3].
Several Ad nanocomplexes have been developed to maximize the
targeting efciency while minimizing harmful side effects
[18,19]. Polyethylene glycol (PEG) has been one of the prominent
polymers to be conjugated with Ad to reduce liver uptake and
extend its half-life in the circulation. Further, PEGylation of Ad
has demonstrated not only to reduce innate immune response
against Ad but also to prevent host antibodies from neutralizing
Ad [20].
Ad encapsulation with a cationic polymer is another effective
strategy to overcome CAR dependence and immunological barriers. Because cell plasma membranes and Ad are both negatively
charged, the positively charged polymer is ionically attracted to
the virus and the net positive Ad surface charge facilitates viral
entry into host cells via electrostatic interaction. Poly(ethylenimine) [21] and poly(L-lysine) [22] have been the two
cationic polymers frequently used to improve Ad transduction in
gene therapy approaches. However, these polymer-coated Ads
remain inefcient in terms of tumor selectivity because their
positive charge also increases their nonspecic uptake into
normal cells. In addition, these cationic polymers are nondegradable in vivo and induce severe cytotoxicity due to the
strong adsorption of positively charged polymer with negatively
charged cell membranes [23].
Hypoxia is a pathological characteristic of solid tumor local
environments, and decreased oxygen availability negatively impacts cancer gene therapy and also chemotherapy. In hypoxic tissues such as solid tumors, anaerobic glycolysis is upregulated to
compensate for reduced ATP production by oxidative phosphorylation. Lactic acid accumulates in and acidies the local extracellular environment [24]. Thus, a smart Ad nanocomplex that
targets specic aspects of the tumor microenvironment such as the
local tissue pH can enhance vector tumor-selectivity and safety. In
this study, we designed and synthesized a pH-sensitive polymer
with bioreducible disulde bond (mPEG-piperazine-N, N0 -cystaminebisacrylamide; PPCBA), in which piperazine and N, N0 -cystaminebisacrylamide acts as a pH-sensitive and bioreducible
moiety, respectively. We then investigated the in vitro transduction
efcacy and cellular uptake mechanism of Ad complexed with
PPCBA and the in vivo therapeutic efcacy of oncolytic Ad encapsulated by PPCBA.

2. Materials and methods

2.1. Cell culture and Ad generation
All cell lines were cultured in Dulbecco's modied Eagle's medium (DMEM;
GIBCO-BRL, Grand Island, NY) supplemented with 10% fetal bovine serum (FBS)
(GIBCO-BRL) at 37  C in a humidied atmosphere containing 5% CO2. A human
embryonic kidney cell line (HEK 293) expressing the Ad E1 region, human brain
cancer cell lines (U343, U87), and human breast cancer cell line (MCF7) were purchased from the American Type Culture Collection (Manassas, VA). Green uorescence protein (GFP)-expressing replication-incompetent Ad (dE1/GFP) and VEGFtargeted replication-competent Ad mutated in E1A and deleted in E1B regions
(RdB/shVEGF) were generated in HEK 293 cells and puried by the CsCl gradient
centrifugation [25,26]. Viral particle (VP) numbers were calibrated to optical density
measured at 260 nm (OD260) and 1 absorbance unit equaled 1012 VP/mL. Puried
viruses were stored at 80  C until use.
2.2. Synthesis of pH-sensitive and bioreducible polymer (PPCBA)
To generate the pH-sensitive and bioreducible polymer, mPEG-acrylate
(Mn 2.0 kDa, 0.10 mmol) (SigmaeAldrich, St. Louis, MO) and N, N0 -cystaminebisacrylamide (1.0 mmol) (PolySciences Inc., Warrington, PA) were dissolved
in methanol (5.0 mL). Subsequently piperazine (1.0 mmol) (SigmaeAldrich) was
added, and the reaction mixture was stirred at 50  C for 48 h in the dark under a
100% nitrogen atmosphere, cooled to room temperature, and the polymer (PPCBA)
was precipitated in excess diethyl ether for two times and dried under vacuum at
room temperature for 2 days. Finally, PPCBA was characterized by 1H NMR in CDCl3
(Mercury Plus 300 MHz; Varian Inc., Palo Alto, CA). Further, gel permeation chromatography analysis (GPC) showed average molecular weight Mw 2.85 kDa,
PDI 1.87. GPC measurements were performed using a Waters-GPC, Water 515
pump, Water 410 RI, and thermostated (35  C), TSK gel G3000 G5000 columns,
0.3 M NaAc aqueous solution (pH 4.4) plus methanol (70/30 v/v) was used as eluent
at a ow rate of 1.0 mL/min. The average molecular weights were calculated against
low polydispersity PEG standards. In addition, the matrix assisted laser desorption/
ionization-time of ight mass spectrum (MALDI-TOF) was performed. It showed that
the molecular weight of PPCBA polymer was 4.04 kDa.
2.3. Characterization of naked Ad and PPCBA-complexed Ad
The average size and zeta potential of dE1/GFP-PPCBA complex were measured
with a Zetasizer 3000HS (Malvern Instruments Inc., Worcestershire, UK) with a
HeeNe laser beam (633 nm, xed scattering angle of 90 ) at 25  C. To characterize
the electrostatic interaction of dE1/GFP with PPCBA polymer, various concentrations
of PPCBA polymer (5e50 mg/mL) were gently pipette-mixed with dE1/GFP Ad
(2  1010 VP) and incubated at room temperature for 30 min. After the formation of
dE1/GFP-PPCBA complex, phosphate-buffered saline (PBS, pH 7.4) was added to nal
volume of 1.0 mL before analysis. The sizes and zeta potential values were presented
as average values of three measurements.
2.4. Gel-retardation assay
To assess the PPCBA-encapsulation of Ad, various concentrations (5e50 mg/mL)
of PPCBA-coated dE1/GFP were electrophoresed (80V, 30 min) in 0.8% (w/v) agarose
gel in 1 TAE buffer (10.0 mM Tris/HCl, 1% (v/v) acetic acid, and 1.0 mM EDTA)
containing ethidium bromide. Viral DNA was visualized using a ChemiDoc gel
documentation system (Syngene, Cambridge, UK).
2.5. Transmission electron microscopy (TEM) imaging
The morphology of naked dE1/GFP and dE1/GFP-PPCBA complexes were characterized by TEM (JEM-2000EXll, JEPL; Nikon, Tokyo, Japan) at 200 kV after incubating for 30 min at different pH conditions (7.4 and 6.0).
2.6. Transduction efciency assay
Transduction efciency of PPCBA-complexed Ad was assessed by quantifying
GFP expression by uorescence-activated cell sorting (FACS) analysis in CAR ()
(U343) and CAR () (MCF7) cells. Each cell line was seeded at a density of 1  105
cells/well in 12-well plates for 24h before Ad transduction. Cells were transduced
with naked dE1/GFP or dE1/GFP-PPCBA complexes at 20 (U343) or 200 (MCF7) MOI,
in both pH 7.4 and pH 6.0, for 30 min. After 48 h of incubation at 37  C, cells were
observed by uorescence microscopy (Olympus IX81; Olympus Optical, Tokyo,
Japan). Cells were also assessed with a BD FACScan analyzer (BectoneDickinson, San
Jose, CA) using CellQuest software (BectoneDickinson); data from 10,000 events
were collected for further analysis and represented the values of relative uorescence intensities. Data represent the means and standard deviations of three
measurements.
2.7. Competition assay with adenoviral ber protein
U343 cells were plated at 1  105 cells/well in 12-well plates. After 24 h, cells
were incubated in serum-free DMEM with Ad ber protein (0.2 and 2 mg/mL) or PBS
(vehicle) at 4  C for 1 h. Naked dE1/GFP or dE1/GFP-PPCBA complexes (50 mg/mL

C.Y. Moon et al. / Biomaterials 41 (2015) 53e68

PPCBA) were added to the cells at an MOI of 50 (dE1/GFP) or 20 (dE1/GFP-PPCBA)
and incubated at 37  C for 1 h cells were then washed with ice-cold PBS and replaced
with DMEM/5% FBS. After incubating for 48 h, cells were observed by uorescence
microscopy, and GFP expression was quantied by ow cytometry and analyzed
with CellQuest software.
2.8. Cellular uptake mechanism/endosomal escape of the Ad-PPCBA complexes

55

2.13. Measurement of inammatory cytokines and toxicity studies

The inammatory immune response and toxicity were examined after intravenous injection of PBS, PPCBA (50 mg/mL), dE1/GFP (2  1010 VP), or dE1/GFPPPCBA (50 mg/mL PPCBA; 2  1010 VP) into BALB/c mice (8 weeks, Charles River
Korea Inc., Seoul, Korea). At 6 h post-injection, serum was harvested by retroorbital
bleeding and interleukin 6 (IL-6) levels were quantied using an IL-6 ELISA kit (R&D
Systems, Minneapolis, MN). At 3 days post-injection with PBS, naked dE1/GFP
(2  1010 VP), or dE1/GFP-PPCBA complex (50 mg/mL PPCBA; 1  1010 VP), mice were
sacriced and serum was collected for aspartate aminotransferase (AST) and alanine
aminotransferase (ALT) levels, which were determined at Neodin Corporation
(Seoul, Korea) as a measure of hepatic toxicity.

Cells were pre-treated with chlorpromazine (CPZ; clathrin-mediated endocytosis inhibitor; 0.2 and 1 mM), genistein (caveolin-mediated endocytosis inhibitor;
1.25 and 5 mM), NH4Cl (macropinocytosis inhibitor; 0.5 and 1 mM), or balomycin A1
(Bf-A) endosomal escape inhibitor; 5 and 20 mM) at pH 6.0 and 7.4 for 30 min. Naked
dE1/GFP (50 MOI for U343 and 500 MOI for MCF7) or dE1/GFP-PPCBA complexes (20
MOI for U343 and 200 MOI for MCF7) (50 mg/mL PPCBA) were added in the presence
or absence of the inhibitors for an additional 2 h. Then, cells were washed with PBS
and incubated for 48 h in DMEM/5% FBS. Cells were then observed by uorescence
microscopy, and GFP expression was quantied by ow cytometry and analyzed
with CellQuest software.

Data are expressed as the mean standard deviation (SD). The ManneWhitney
test was performed for statistical comparisons (SPSS v.18.0 software; SPSS/IBM,
Chicago, IL). Differences were considered statistically signicant when P-values
were <0.05.

2.9. VEGF quantication

3. Results

Human VEGF-A was quantied using the human VEGF Quantikine Immunoassay
kit (R&D Systems, Minneapolis, MN), according to manufacturer recommendations.
Serial dilutions of a known concentration of puried recombinant human VEGF-A
were used to establish a standard curve. Briey, U87 cells were plated onto sixwell plates in DMEM/5% FBS. At 50% conuence, cells were infected with PBS,
RdB/shVEGF, or RdB/shVEGF-PPCBA (50 mg/mL PPCBA; pH 6.0 or 7.4) at 1 or 10 MOIs.
Conditioned media were harvested 48 h after infection, and secreted VEGF was
measured by ELISA [26].
2.10. MTT assay
To evaluate the cancer cell-killing effect of RdB/shVEGF or RdB/shVEGF-PPCBA at
pH 6.0 or 7.4, U343 and MCF7 cells grown to 50% conuence in 24-well plates were
infected with naked RdB/shVEGF or RdB/shVEGF-PPCBA (50 mg/mL PPCBA) at 20
MOI for U343 and 200 MOI for MCF7 cells, and incubated at 37  C. Two days post
infection, 250 mL of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide
(MTT; SigmaeAldrich, St. Louis, MO) at 2 mg/mL in PBS was added to each well and
incubated at 37  C for 4 h, then the supernatant was discarded and the precipitate
was dissolved in 1.0 mL dimethylsulfoxide (DMSO). Optical density of the dissolved
formazan product was then read on a microplate reader at 540 nm. Non-infected
cells served as negative controls.
2.11. Ex vivo Matrigel plug assay
MCF7 cells (2  105) were plated in 6-well plates and infected with naked RdB/
shVEGF or RdB/shVEGF-PPCBA (50 mg/mL PPCBA), or PBS (vehicle, negative control)
at either pH 6.0 or 7.4. After 6 h, cells were trypsin-harvested and washed three
times with 5 mL of Hank's buffered saline. Cells were then mixed with 600 mL of cold
Matrigel (BD-bioscience, Sanjose, and CA) and injected subcutaneously above the
ank region of the male athymic nude mice (Charles River Korea Inc., Seoul, Korea).
The injected Matrigel rapidly formed a single solid gel plug. After 14 days, animals
were sacriced and the skin of each mouse was pulled back to expose the intact
Matrigel plug. To quantify blood vessel formation, Matrigel plugs were xed with
zinc xation solution, parafn-embedded in parafn, and sections were treated with
puried monoclonal rat anti-mouse CD31 (platelet/endothelial cell adhesion
molecule 1; BD Biosciences Pharmingen, San Diego, CA), and then with goat anti-rat
IgG-HRP. All slides were counterstained with Meyer's hematoxylin. The expression
levels of CD31 were analyzed semi-quantitatively with MetaMorph image analysis
software (Universal Image Corp., Westchester, PA).
2.12. Assessment of antitumor efcacy
To compare the anti-tumor efcacy of naked RdB/shVEGF and RdB/shVEGFPPCBA, U87 tumor xenografts were established subcutaneously by injecting
1  107 cells under the abdominal skin of 6- to 8-week-old female athymic nude
mice (Charles River Korea Inc.). Once tumors size reached approximately
100 mm3 in volume, mice were randomized into three groups (PBS, RdB/shVEGF,
and RdB/shVEGF-PPCBA) and injected intratumorally with 30 mL PBS or an
equivalent volume of PBS containing 5  109 VP of RdB/shVEGF or RdB/shVEGFPPCBA (50 mg/mL PPCBA) three times every other day. To evaluate antitumor
efcacy via systemic administration, U87 tumor-bearing mice were injected with
200 mL PBS, RdB/shVEGF (2  1010 VP/200 mL), or RdB/shVEGF-PPCBA (2  1010
VP/200 mL) via tail vein on three occasions spaced every other day. The length (L)
and width (W) of the tumor were measured every other day with a caliper to
calculate tumor growth. Tumor volume was calculated according to the following
formula: tumor volume 0.523 LW2. All animal studies were conducted under
the auspices and approval of Hanyang University institutional animal care and
use committee.

2.14. Statistical analysis

3.1. Synthesis of PPCBA polymer

The synthesis of pH-sensitive and bioreducible polymer (PPCBA)
is illustrated in Fig. 1A. The PPCBA polymer was synthesized
through Michael-type addition polymerization [27]. The reaction of
mPEG-acrylate and N, N0 -cystaminebisacrylamide with piperazine
formed mPEG-b-Pip-CBA (PPCBA). The chemical structure of PPCBA
was conrmed by 1H NMR (300 MHz, CDCl3) with its typical
resonance peaks at 2.3e2.8, and 3.6 ppm which corresponds to eN
(CH2eCH2) methylene protons of piperazine, and -S-CH2CH2eNHeCOe protons of CBA, respectively, along with characteristic peaks of mPEG (Fig. 1B). The absence of any proton signals
between 5 and 6.5 ppm indicates that no residual of acrylamide or
acrylate monomers are present in nally synthesized PPCBA polymer. Further, the average molecular weight was determined by gel
permeation chromatography (GPC). The observed average molecular weight was Mw 2.85 kDa with a PDI of 1.87. Furthermore, the
matrix-assisted laser desorption/ionization-time of ight mass
(MALDI-TOF) spectrum showed that the molecular weight of PPCBA
copolymer was 4.04 kDa (Fig. S1). This increased molecular weight
of PPCBA in comparison to mPEG (2 kDa) conrms the formation of
PPCBA copolymer. The ratios of eOCH3 of PEG and
eSeSeCH2eCH2eNHeCOe of alkyl chain were calculated by integrals of the two peaks, and value was identied as 1:6 (Fig. 1B)
[28]. These results suggest that the polymer is close to the expected
composition, in a good agreement with MALDI-TOF data.
3.2. Physico-chemical characterization of PPCBA and Ad-PPCBA
complexes
Potential cytotoxicity of the PPCBA polymer was evaluated by
MTT assay. Fig. 2A shows that PPCBA is not cytotoxic to mammalian
MCF7 cells, even at high concentrations (i.e., 100 mg/mL). GFPexpressing non-replicating Ad (dE1/GFP) was then complexed
with PPCBA polymer by ionic interactions between the carboxylic
acid groups of external amino acids in the hexon of Ad capsids and
tertiary amino groups of the PPCBA copolymer. The average size
and surface charge of the complexes with increasing concentrations of PPCBA were assessed by dynamic light scattering (DLS) and
zeta potential analysis (Fig. 2B). The diameter of the naked Ad
increased from 104 6.64 nm to 175 5.2 nm when Ad was
complexed with 50 mg/mL of PPCBA. The surface charge of the
complex
also
increased
from
the
negative
value
of 17.53 mV 0.40 mV of naked Ad to the positive value of
7.31 0.78 mV after coating Ad with PPCBA at 50 mg/mL. These
results indicate that coating of Ad viral particles with cationic
PPCBA polymer concentration-dependently increased the particle

56

C.Y. Moon et al. / Biomaterials 41 (2015) 53e68

Fig. 1. Schematic representation of PPCBA structure and characterization. (A) Schematic representation of synthesis of pH-sensitive and bioreducible polymer (PPCBA). (B) 1H NMR
spectrum of PPCBA polymer in CDCl3 (300 MHz).

diameter and zeta potential value, conrming that Ad adequately

complexes with PPCBA through electrostatic interaction.
In order to nd the optimal PPCBA concentration for Ad complexing, we performed a gel retardation assay (Fig. 2C). Retardation
of dE1/GFP-PPCBA complexes began at 25 mg/mL PPCBA and particles were completely immobilized at 50 mg/mL PPCBA. Based on
this result, the PPCBA concentration of 50 mg/mL was used to make
a complex with Ad for subsequent experiments, unless otherwise
stated.
The morphology of naked Ad and dE1/GFP-PPCBA complexes
was examined by TEM. Fig. 2D shows photomicrographs of naked
Ad at pH 7.4 and dE1/GFP-PPCBA complex at PPCBA concentration
of 50 mg/mL at both pH 7.4 and 6.0. The naked Ad shows the
characteristic hexon structure and icosahedral shape of the viral
particle. On the other hand, we observed a tendency for clusters of
Ad particles to be enveloped by a tensile chunk of PPCBA polymer,
leading to generation of Ad-PPCBA complexes. This observation
suggests that the cationic PPCBA polymer is encapsulating Ad and
linking the negative surface charge of multiple viral particles
[19,29,30].
Colloidal stability of dE1/GFP-PPCBA complexes was next evaluated in PBS buffer after incubating at room temperature for 3 h,
9 h, 18 h, and 24 h. As shown in Fig. S2, the average size of dE1/GFPPPCBA (175.1 nm) was not changed for 24 h, suggesting that PPCBA

polymer-coated Ad complex has a good colloidal stability under the

physiological saline condition. We also examined the bioreducibility of the dE1/GFP-PPCBA complexes by treating with
reducing agent dithiothreitol (DTT) at 37  C, for 2 h. As shown in
Fig. S3, the average size of dE1/GFP-PPCBA complexes after treatment with DTT was reduced to 116.1 nm, which is close to that of
naked Ad. In contrast, the treatment of naked Ad with DTT did not
change the size of the naked Ad. These results clearly demonstrate
that CBA block in PPCBA copolymer can be degraded upon exposure
to redox microenvironment, subsequently release Ad in cytosol.
3.3. Enhanced and pH-dependent transduction efciency of AdPPCBA
To evaluate the transduction efciency of dE1/GFP-PPCBA
complexes in comparison to naked Ad, we transduced CAR ()
U343 and CAR () MCF7 cancer cells at pH 7.4 and 6.0 with naked
Ad or dE1/GFP-PPCBA complexes at an MOI of 20 (U343) or 200
(MCF7). As shown in Fig. 3A, the transduction efciency of dE1/
GFP-PPCBA complex escalated by increasing the concentration of
PPCBA polymer used to make the Ad complex, at both pH conditions. Importantly, dE1/GFP-PPCBA complex showed higher GFP
expression at pH 6.0 than at pH 7.4 in both cell types, demonstrating enhanced transduction at lower pH. However, no pH-

C.Y. Moon et al. / Biomaterials 41 (2015) 53e68

57

120.0

Cell viability (%)

100.0
80.0
60.0
40.0
20.0
0.0
-

10

20

100 g/mL

50

Concentration of PPCBA polymer (g/mL)

10

200

Diameter (nm)

160

0
120
-5

80
-10

40

Zeta potential (mV)

-15

-20

dE1/GFP

10

25

50

Concentration of PPCBA polymer (g/mL)

Concentration of PPCBA polymer (g/mL)

dE1/GFP

200 nm

dE1/GFP

10

25

dE1/GFP-PPCBA (pH 7.4)

X100,000

200 nm

X100,000

50

dE1/GFP-PPCBA (pH 6.0)

200 nm

X 100,000

Fig. 2. Cytotoxicity of PPCBA polymer and physico-chemical properties of Ad-PPCBA complexes. (A) Cell cytotoxicity of the PPCBA polymer in MCF-7 cells was evaluated by MTT
assay. (B) The average size (nm) and zeta potential value (mV) of the PPCBA-complexed Ad were measured at various concentrations of PPCBA polymer (5e50 mg/mL) and 2  1010
VP in 1 mL PBS, at room temperature. The data are representatives of three independent experiments performed in triplicate. Data represents the mean SD for three replicates. (C)
Gel retardation assay of the Ad-PPCBA complex with various concentrations of PPCBA polymer. (D) TEM images of naked Ad and Ad-PPCBA complex. The Ad and Ad-PPCBA were
incubated with PBS at pH 7.4 and 6.0 for 30 min.

58

C.Y. Moon et al. / Biomaterials 41 (2015) 53e68

A
Concentration of PPCBA polymer (g/mL)

U343
dE1/GFP

10

25

50

pH 7.4

X 100
pH 6.0

200 m

MCF7
dE1/GFP

X 100

Concentration of PPCBA polymer (g/mL)

10
25

50

pH 7.4

X 100
pH 6.0

200 m

X 100

400

U343

300

200

100

0
dE1/GFP

10

25

50

Concentration of PPCBA polymer (g/mL)

dE1/GFP-PPCBA (pH7.4)

dE1/GFP (pH 6.0)

dE1/GFP-PPCBA (pH 6.0)

Normalized GFP expression level

Normalized GFP expression level

dE1/GFP (pH7.4)

MCF7

1.E+05
1.E+04
1.E+03
1.E+02
1.E+01
1.E+00
dE1/GFP

10

25

50

Concentration of PPCBA polymer (g/mL)

Fig. 3. Enhanced and pH-dependent transduction efciency of Ad-PPCBA. (A) Fluorescence images of U343 and MCF7 cells transduced with naked dE1/GFP or dE1/GFP-PPCBA
complex (5, 10, 25, and 50 mg/mL of PPCBA polymer) at pH 7.4 and 6.0. Original magnication: 100. (B) At 48 h post transduction, cells were analyzed for GFP expression by
ow cytometry analysis. The data are representatives of three independent experiments performed in triplicate. Bars represent mean SD.

dependent differences were observed for naked Ad. This observation was supported by quantitative FACS analysis of GFP, in
which PPCBA-complexing enhanced Ad transduction efciency by
3.6-fold at pH 6.0 and 2.7-fold at pH 7.4 relative to naked Ad in
U343 cells treated with dE1/GFP-PPCBA (50 mg/mL) (Fig. 3B). An
even more profound pH-dependent increase in transduction efciency was observed in dE1/GFP-PPCBA-treated MCF7 cells, which

showed a 475.2-fold increase at pH 6.0 and a 4.4-fold increase at

pH 7.4 relative to naked Ad, at 50 mg/mL concentration. Taken
together, these results imply that dE1/GFP-PPCBA complex can
enhance Ad transduction efciency regardless of CAR expression
by cancer cells, and that transduction efciency is further
enhanced under acidic conditions reective of the tumor
environment.

C.Y. Moon et al. / Biomaterials 41 (2015) 53e68

3.4. Cellular uptake mechanism studies of Ad-PPCBA complex

We then sought to elucidate the cellular uptake mechanism of
dE1/GFP-PPCBA complexes. First, a CAR competition assay was performed to investigate the extent of CAR dependency of the nanocomplex in its target cell uptake. Since the transduction efcient of
naked Ad was signicantly lower than dE1/GFP-PPCBA complex, we
used 50 MOI for naked Ad and 20 MOI for dE1/GFP-PPCBA complexes
to compensate the low transduction efciency of naked Ad. Pretreatment of U343 cells with CAR-binding Ad ber protein substantially reduced subsequent GFP transduction efciency of naked Ad in a
concentration-dependent manner (67% and 90% decreases with 0.2
and 2 mg/ml Ad ber protein pretreatment, respectively) (Fig. 4A). Cell
pretreatment with Ad ber protein also inhibited dE1/GFP-PPCBA
transduction efciency when performed at pH 7.4. In contrast, GFP
expression in cells transduced with dE1/GFP-PPCBA at pH 6.0 was not
blocked by Ad ber protein, indicating that cellular uptake of dE1/
GFP-PPCBA was not mediated by CAR/Ad ber interactions at low pH.
Next, we pretreated U343 cells with various cellular uptake inhibitors: 1) CPZ: an inhibitor of clathrin-mediated endocytosis [31]
(0.2 and 1 mM); 2) genistein, an inhibitor of caveolae-mediated endocytosis [32] (1.25 and 5 mM); and 3) NH4Cl, an inhibitor of macropinocytosis [33] (0.5 and 1 mM). As shown in Fig. 4B, dE1/GFP-PPCBA
complex at both pH 7.4 and pH 6.0 were not affected by CPZ, which
correlated with unchanged GFP expression levels in cells pretreated
with CPZ before transduction. In contrast, CPZ concentrationdependently decreased GFP expression in U343 cells transduced
with naked Ad, conrming Ad's dependence on clathrin-mediated
endocytosis for cell entry [34]. Genistein treatment had no inhibitory effect on the cellular uptake of either naked Ad at pH 7.4 or of dE1/
GFP-CBA complexes at either pH 6.0 or 7.4, indicating that none of
these vectors use caveolin-mediated endocytosis for cellular uptake.
Cell pretreatment with NH4Cl revealed that macropinocytosis is
a major endocytic pathway for both naked and polymer-complexed
dE1/GFP-PPCBA (Fig. 4C). At 1 mM, NH4Cl signicantly decreased
GFP expression in cells transduced with naked Ad at pH 7.4 and by
dE1/GFP-PPCBA at both pH 6.0 and 7.4.
3.5. Evaluation of endosome escaping capacity of Ad-PPCBA
complex
The vacuolar H ATPase inhibitor balomycin A1 (Bf-A) prevents,
acidication of Ad, which is a crucial factor for viral escape from
endosomes. To assess the endosome-escaping capability of AdPPCBA complexes, U343 and MCF-7 cell lines were pre-treated
with 5 and 20 mM of Bf-A for 30 min and then transduced with
naked dE1/GFP (50 MOI for U343 and 500 MOI for MCF7) at pH 7.4 or
dE1/GFP-PPCBA complexes (20 MOI for U343 and 200 MOI for MCF7)
at either pH 6.0 or 7.4. As shown in Fig. 5A, Bf-A signicantly
inhibited U343 cell transduction by both naked Ad and dE1/GFPPPCBA complexes, independent of extracellular pH, by z 85%
versus respective non-Bf-A-treated controls. In MCF7 cells, 20 mM BfA potently inhibited GFP expression induced by both naked Ad and
dE1/GFP-PPCBA treatment at pH 6.0 (Fig. 5B). This suggests that
endosome acidication is critical for efcient gene transfer efciency of both dE1/GFP-PPCBA complexes and naked Ad. We presume that rapid endosomal escape allows Ad-PPCBA complexes to
efciently avoid lysosomal degradation, resulting in enhanced gene
transfer efciency.
3.6. In vitro tumor cell killing and VEGF suppression efcacy of
oncolytic Ad complexed with PPCBA (RdB/shVEGF-PPCBA)
After elucidating its transduction efciency and uptake mechanism, we evaluated the effect of oncolytic Ad complexed with

59

PPCBA on cancer cell killing. RdB/shVEGF is an oncolytic Ad that

expresses VEGF-specic shRNA, and has previously been demonstrated to have efcient anti-angiogenic and anti-tumor activity by
suppressing VEGF expression in target cancer cells [26]. Infection of
U87 cells with naked RdB/shVEGF naked oncolytic Ad suppressed
VEGF expression by these cells in a viral concentration-dependent
manner (Fig. 6A). Similarly, RdB/shVEGF-PPCBA complex
decreased the VEGF level in a viral dose-dependent manner,
demonstrating that RdB/shVEGF complexed in PPCBA is functionally active in expressing VEGF-specic shRNA, which inhibits
angiogenic VEGF secretion. Importantly, RdB/shVEGF-PPCBA complex more potently suppressed VEGF gene expression at pH 6.0
compared to pH 7.4 (p < 0.05), conrming that enhanced infection
efciency of RdB/shVEGF-PPCBA under acidic conditions enhances
therapeutic siRNA functionality.
Because CAR expression is often lost as cancers progress, and
this loss has been one of major hurdles to use anti-cancer Ad vectors in the clinical setting, we compared the oncolytic effect of RdB/
shVEGF-PPCBA complexes in both CAR-positive U343 and CARnegative MCF7 cell lines. As shown in Fig. 6B, naked RdB/shVEGF
killed U343 cells very efciently, whereas MCF7 cells were resistant
to naked RdB/shVEGF. This demonstrated that naked oncolytic Ad
depends on tumor cell CAR expression for oncolytic activity. Similarly, at pH 7.4, RdB/shVEGF-PPCBA complexes effectively killed
CAR U343 cells but not MCF7 cells, implying that at this pH,
PPCBA-Ad complex uptake by tumor cells relies on target cell CAR
expression. In marked contrast, RdB/shVEGF-PPCBA complex at pH
6.0 elicited potent cell killing efcacy in both U343 and MCF7,
implying that RdB/shVEGF-PPCBA complex can bypass the
requirement for tumor cell CAR expression to infect these cells. This
suggests that, in the hypoxic and acidied tumor microenvironment, even CAR-decient tumor cells can be efciently infected by
PPCBA-Ad complexed vectors.

3.7. In vivo therapeutic efcacy of RdB/shVEGF-PPCBA complexes

To conrm whether RdB/shVEGF-PPCBA complexes' in vitro
anti-tumor therapeutic potency translates to the in vivo scenario,
we rst evaluated the anti-angiogenic effect of RdB/shVEGF-PPCBA
by the Matrigel plug assay. MCF7 cells were infected with either
naked RdB/shVEGF or RdB/shVEGF-PPCBA either at pH 6.0 or at pH
7.4, and were mixed with cold Matrigel and injected into mice
subcutaneously above the ank region. After 14 days, Matrigel
plugs were excised and analyzed for vessel quantication. Matrigel plugs mixed with cells transduced with PBS, naked RdB/
shVEGF, or RdB/shVEGF-PPCBA (pH 7.4) showed similar amounts of
neovascularization. However, plugs containing MCF7 cells that had
been infected with RdB/shVEGF-PPCBA at pH 6.0 exhibited
dramatically reduced vascularization (p < 0.001) (Fig. 7A).
Next, we assessed the in vivo tumor-killing ability of RdB/
shVEGF-PPCBA. Mice with U87 xenograft tumors were established and were treated intratumorally with PBS, naked Ad, or RdB/
shVEGF-PPCBA. As shown in Fig. 7B, tumor growth was signicantly inhibited in the RdB/shVEGF-PPCBA group compared to the
naked RdB/shVEGF group (p < 0.01). The respective mean tumor
sizes at Day 23 after treatment with PBS, naked RdB/shVEGF Ad,
and RdB/shVEGF-PPCBA were 1984 439 mm3, 720 119 mm3,
and 177 73 mm3. At Day 23, the corresponding tumor growth
inhibition percentages for RdB/shVEGF and RdB/shVEGF-PPCBA
groups were 63.7% and 91.1% of PBS controls. Taken together,
these results demonstrated potent in vivo anti-angiogenic and
tumor-killing activities of RdB/shVEGF-PPCBA's, thereby highlighting their potential therapeutic efcacy as anti-cancer
treatments.

60

A
0

0.2

CP Z

Ad fiber knob protein

2

g/ml

dE1/GFP

0.2

Genistein
1

1.25

5 M

dE1/GFP

X 100

X 100

dE1/GFP-PPCBAp (pH 7.4)

dE1/GFP
-PPCBA
(pH 7.4)

X 100

X 100

dE1/GFP-PPCBA (pH 6.0)

200 m

X 100
200 m

X 100

800

dE1/GFP
dE1/GFP-PPCBA (pH 7.4)

700

dE1/GFP-PPCBA (pH 6.0)

600
500
400
300

***

***

200

***

***

100
0

0.2

0.2

0.2

g/ml

Relative normalized GFP expression (%)

GFP expression (mean)

dE1/GFP
dE1/GFP-PPCBA (pH 7.4)
120

dE1/GFP-PPCBA (pH 6.0)

100
80

**
60
40
20
0

0.2
CPZ

1.25

genestein

0.2
CPZ

1.25

genestein

0.2
CPZ

1.25

genestein

Fig. 4. Cellular uptake mechanism studies of Ad-PPCBA complex. (A) CAR competition assay. U343 cells were transduced naked dE1/GFP or dE1/GFP-PPCBA complexes (50 mg/mL PPCBA) in the presence and absence of ber protein (0.2
and 2 mg/mL) specic to CAR. At 48 h post infection, cells were analyzed for GFP expression by ow cytometry analysis. The data are representatives of three independent experiments performed in triplicate. Bars represent mean SD.
***P < 0.001 versus naked dE1/GFP-treated group. Original magnication: 100. (B, C) Endocytosis pathway analysis with chlorpromazine (CPZ), genistein, or ammonium chloride. U343 cells were pre-treated for 30 min with CPZ,
genistein, or ammonium chloride at indicated concentrations. Naked dE1/GFP or dE1/GFP-PPCBA complex (50 mg/mL PPCBA) was added in the absence or presence of the inhibitors for an additional 2 h. Cells were then observed for GFP
expression by uorescence microscopy and ow cytometry. Original magnication: 100. Bars represent mean SD. **P < 0.01 versus naked dE1/GFP-treated group.

C.Y. Moon et al. / Biomaterials 41 (2015) 53e68

dE1/GFP
-PPCBA
(pH 6.0)

C.Y. Moon et al. / Biomaterials 41 (2015) 53e68

61

Ammonium chloride
0

0.5

1 M

dE1/GFP

X 100
dE1/GFP-PPCBA (pH 7.4)

X 100
dE1/GFP-PPCBA (pH 6.0)

X 100

Relative normalized GFP expression (%)

200 m

120
100
80

**

**
**

60

40
20
0

0.5
dE1/GFP

0.5

dE1/GFP
-PPCBA (pH 7.4)

0.5

dE1/GFP
-PPCBA (pH 6.0)

Fig. 4. (continued).

To evaluate the therapeutic efcacy of systemicallyadministered oncolytic Ad-PPCBA complexes, nude mice bearing
U87 xenograft tumors were injected intravenously with PBS, naked
RdB/shVEGF, or RdB/shVEGF-PPCBA. As shown in Fig. 7C, tumor
growth was signicantly inhibited in the RdB/shVEGF-PPCBA group
versus the naked RdB/shVEGF group (p < 0.001). The respective
mean tumor sizes at Day 27 after intravenous injection with PBS,
RdB/shVEGF, or RdB/shVEGF-PPCBA were 2144 167 mm3,
1777 484 mm3, and 752 35 mm3. This corresponded to tumor
growth inhibitions for RdB/shVEGF and RdB/shVEGF-PPCBA groups
by 17.1% and 65.0%, respectively versus PBS controls. These results
not only conrm the immune evading effect of encapsulating Ad
with PPCBA, but also demonstrate the potent therapeutic antitumor potential of RdB/shVEGF-PPCBA after systemic adminis
tration.

3.8. Evasion of immune response and attenuated liver toxicity

in vivo
One of the challenges inherent to administering therapeutic
viral gene vectors in vivo is innate immune system activation after
recognizing Ad particles in the blood. Based on the gel retardation
assay and the TEM images of dE1/GFP-PPCBA complexes, we hypothesized that PPCBA encapsulation of Ad particles might enable
Ad-PPCBA complexes to evade this immune response by creating a
stealth Ad. To evaluate the evasive character of Ad-PPCBA complexes, we measured serum levels of the inammatory cytokine IL6 as a proxy for innate immune system activation against Ad. BALB/
c mice were injected intravenously with PBS only or an equivalent
volume containing 2  1010 VP of naked dE1/GFP or dE1/GFPPPCBA, and measured serum IL-6 by ELISA at 6 h after injection

62

A
U343

Bafilomycin A1

Bafilomycin A1
5

MCF7

20 M

20 M

dE1/GFP

dE1/GFP

X 100

X 100

dE1/GFP-PPCBA (pH 7.4)

dE1/GFP-PPCBA (pH 7.4)

dE1/GFP-PPCBA (pH 6.0)

200 m

X 100

200 m

X 100

dE1/GFP

dE1/GFP

dE1/GFP-PPCBA (pH 7.4)

1200
1000
800
600
400

***

200
0

20

20

20 M

400
350

100
80
60
40
20

***

20

20

20 M

300
250
200
150
100
50

***

20

20

Relative normalized GFP expression (%)

GFP expression (mean)

1400

dE1/GFP-PPCBA (pH 6.0)

120

GFP expression (mean)

1600

Relative normalized GFP expression (%)

dE1/GFP-PPCBA (pH 7.4)

C.Y. Moon et al. / Biomaterials 41 (2015) 53e68

X 100

X 100
dE1/GFP-PPCBA (pH 6.0)

20 M

dE1/GFP-PPCBA (pH 6.0)

120
100
80
60
40
20

***
0

20

20

20 M

Fig. 5. Assessment of endosome escaping capacity of Ad-PPCBA complex. U343 (A) and MCF (B) cells were pre-incubated at 4
for 30 min with 5 and 20 mM of balomycin A1 (Bf-A), and then transduced with naked dE1/GFP at pH 7.4
or dE1/GFP-PPCBA complexes (50 mg/mL PPCBA) at either pH 6.0 or 7.4. At 48 h post transduction, cells were analyzed for GFP expression by uorescence microscopy and ow cytometry. Original magnication: 100. The data are
representatives of three independent experiments performed in triplicate. Bars represent mean SD. ***P < 0.001 versus naked dE1/GFP-treated group.
C

C.Y. Moon et al. / Biomaterials 41 (2015) 53e68

VEGF expression level (g/ml)

63

RdB/shVEGF

3.5

RdB/shVEGF-PPCBA (pH7.4)

RdB/shVEGF-PPCBA (pH 6.0)

2.5
2
1.5

0.5
0
1

1: PBS
2: RdB/shVEGF 1 MOI
3: RdB/shVEGF 10 MOI
4: RdB/shVEGF-PPCBA pH(7.4) 1 MOI
5: RdB/shVEGF-PPCBA pH(7.4) 10 MOI
6: RdB/shVEGF-PPCBA pH(6.0) 1 MOI
7: RdB/shVEGF-PPCBA pH(6.0) 10 MOI

U343 (CAR(+))

MCF7 (CAR(-))

120

120

100

100

Cell viability (%)

Cell viability (%)

***
80
60

40
20

80
60
40
20

0
1

1: PBS
2: RdB/shVEGF
3: RdB/shVEGF-PPCBA pH(7.4)
4: RdB/shVEGF-PPCBA pH(6.0)
Fig. 6. VEGF suppression and cancer cell killing efcacy of RdB/shVEGF-PPCBA complexes. (A) U87 cells were infected with PBS, RdB/shVEGF, or RdB/shVEGF-PPCBA complex at 1 or
10 MOIs. VEGF concentration was measured in the culture supernatant at 48 h after infection by ELISA. The data are representatives of three independent experiments performed in
triplicate. Bars represent mean SD. *P < 0.05 versus naked RdB/shVEGF-treated group. (B) U343 and MCF7 cells were treated with PBS, RdB/shVEGF, or RdB/shVEGF-PPCBA
complex. At 48 h post infection, cell viability was then assessed by MTT assay. The PBS-treated group was set at 100%. The data are representatives of three independent experiments performed in triplicate. Bars represent mean SD. *P < 0.05, ***P < 0.001 versus naked RdB/shVEGF-treated group.

(Fig. 8). Expectedly, systemic injection of naked Ad increased serum

IL-6 levels over two-fold compared to PBS controls. Intravenous
injection of PPCBA or dE1/GFP-PPCBA resulted in serum IL-6 levels
that were similar to levels in the PBS group (p > 0.05), demonstrating PPCBA's biocompatibility and capability to shield the Ad
surface from innate immune system recognition of, and response
against, Ad after systemic administration.
Another contributing factor in Ad's rapid clearance from the
blood is the recruitment and scavenger receptor-mediated phagocytosis of viral particles by hepatic Kupffer cells [35]. Cytokines
released by these specialized macrophages after Ad liver transduction (TNF and IL-6) cause early hepatotoxicity, which is

enhanced by hepatic chemokine-mediated neutrophil accumulation [36,37]. The extent of Ad-polymer complex-induced hepatotoxicity was assessed in vivo by serum levels of alanine
aminotransferase (ALT) and aspartate aminotransferase (AST),
which were measured in BALB/c mice 3 days after intravenous injection with PBS, naked dE1/GFP, or dE1/GFP-PPCBA (Fig. 8). Control
serum liver-enzyme levels were low but were markedly elevated by
naked Ad injection, whereby ALT and AST were respectively
increased by 49.9- and 7.6-fold versus the PBS group (p < 0.01 in
both cases). However, ALT and AST levels in dE1/GFP-PPCBA-treated
mice were not signicantly different from levels in PBS control
animals. In sum, the reduced circulating IL-6, ALT, and AST levels in

64

C.Y. Moon et al. / Biomaterials 41 (2015) 53e68

PBS

RdB/shVEGF

X 200

100 m

RdB/shVEGF-PPCBA (pH 7.4)

RdB/shVEGF-PPCBA (pH 6.0)

X 200

Normalized CD31
positive vessel (%)

120

***

100
80
60
40
20
0
1

1: PBS
2: RdB/shVEGF
3: RdB/shVEGF-PPCBA pH(7.4)
4: RdB/shVEGF-PPCBA pH(6.0)
Fig. 7. Potent antiangiogenic and antitumor efcacy of RdB/shVEGF-PPCBA. (A) Immunohistochemical staining of sections of matrigel plugs with anti-CD31 antibody. Representative photomicrographs are shown. Original magnication: 200. Blood vessels were counted in tumor tissues for each treatment group. Six images were analyzed per group. Bars
represent mean SD. ***P < 0.001 versus naked RdB/shVEGF or RdB/shVEGF-PPCBA at pH 7.4. (B) Antitumor therapeutic efcacy of naked RdB/shVEGF or RdB/shVEGF-PPCBA
complex in U87 glioblastoma xenograft established in nude mice through intratumoral (IT) injection. U87 tumor-bearing mice were injected with PBS, RdB/shVEGF (5  109
VP), or RdB/shVEGF-PPCBA (5  109 VP) complex three times spaced every other day. The arrow indicates administration of treatment. Results are expressed as mean SD (each
group, n 6). ***P < 0.001 versus PBS-treated group (RdB/shVEGF) or RdB/shVEGF-treated group (RdB/shVEGF-PPCBA) (C) Antitumor therapeutic efcacy of naked RdB/shVEGF or
RdB/shVEGF-PPCBA complex in U87 glioblastoma xenograft established in nude mice through intravenous (IV) injection. U87 tumor-bearing mice were injected with 200 mL PBS,
RdB/shVEGF (2  1010 VP), or RdB/shVEGF-PPCBA (2  1010 VP) via tail vein on three occasions spaced every other day. Tumor volume was measured every 2 days following
treatment. The arrow indicates administration of treatment. Results are expressed as mean SD (each group, n 6). ***P < 0.001 for versus PBS- or RdB/shVEGF-treated group.

Ad-PPCBA-injected mice versus the naked Ad-injected animals

attest to PPCBA's immune-escaping property.
4. Discussion
In considering the efcacy of viral or polymeric nanoparticle
gene therapy vectors to target metastatic tumors, three major
criteria require consideration: selectivity, transduction efciency,
and immunogenicity. Numerous vector modications have been

developed to maximize the vector's tumor targeting while

decreasing immunogenic side effects. Recent approaches include
non-specic Ad surface masking polymers such as poly(ethylenimine) (PEI) [21], PEG [22,38], N-(2-hydroxypropyl) methacrylamide (HPMA) [39,40], arginine-grafted bioreducible polymer
[18,30,41] and chitosan [42], alteration of endogenous Ad tropism
by ablation of CAR or integrin avb5 and avb5 binding properties
[43], and by attaching targeting moieties (e.g., tumor-homing
peptides and antibodies) [44e46]. With respect to Ad-polymer

C.Y. Moon et al. / Biomaterials 41 (2015) 53e68

65

IT injection

2500

PBS

Tumor volume (mm3)

RdB/shVEGF

2000

RdB/shVEGF-PPCBA

1500

***

1000
500

***

0
1

11

13

15

17

19

21

23 days

IV injection
2500

PBS

Tumor volume (mm3)

RdB/shVEGF

RdB/shVEGF-PPCBA

2000
1500

***

1000
500
0
1

11

13 15

17 19

21

23

25 27 (days)

Fig. 7. (continued).

complexes, however, most combinations fail to take full advantage

of their potential benets. For instance, while PEGylation of Ad
endows the virus with excellent shielding properties from immune response, PEG also decreases transduction efciency by
obstructing specic Ad ber-cell receptor interactions [20].
Furthermore, the polymers are reported not efciently biodegraded in vivo (i. e. 25 kDa PEI), thus inducing adverse effects. To
this end, we have introduced disulde bond in the CBA moiety of
PPCBA block copolymer to be degraded after cellular uptake and
release the viral vector in cytoplasm. In cytotoxicity assay, PPCBA
polymer was demonstrated not cytotoxic even at high concentration (100 mg/mL; Fig. 2A), whereas 25 kDa PEI shows cytotoxicity as low as 10 mg/mL [30], demonstrating the good safety
prole of PPCBA polymer. After release from PPCBA polymer by
reduction (dithols (eSeSe) to eSH) of CBA, Ad capsid trafcs
along microtubules towards the nucleus by interactions with the
motor protein dynein. Then, the partially disassembled virion
particle is transported to the nuclear membrane to release viral
DNA into the nucleus [17].

CAR dependence for Ad's cell entry remains an important issue

when considering systemic administration of Ad-polymer complexes. A benet of the CAR ablation strategy was reduced Ad liver
tropism and toxicity, suggesting that any vector that fully relies on
the CAR transduction pathway risks increased immunogenicity
caused by Kupffer cell recruitment, activation, and release of inammatory chemokines and cytokines. However, bypassing Ad
vector reliance on CAR-mediated target cell transduction without
equipping the vector with alternative endocytic pathway that is
cancer or cancer-environment specic would decrease Ad's excellent transduction capabilities, especially in CAR-positive tumors.
Furthermore, a degree of CAR reliance for Ad-polymer transduction
can serve as a tumor-selective feature and prevent Ad-polymer
infection of normal cells because most non-tumor cells do not express CARs. The ideal Ad-polymer complex would thus infect tumors via alternative cancer-specic transduction pathways that can
potentially function in conjunction with CAR, but also should
remain nonreactive to normal cells and should be encapsulated
against innate immune attack.

66

C.Y. Moon et al. / Biomaterials 41 (2015) 53e68

A
IL-6 expression level (mean)

200

180
160
140
120
100
80
60
40
20
0
PBS

dE1/GFP

PPCBA dE1/GFP-PPCBA
ALT

3000

**

AST

2500

Unit/L

2000

**

1500
1000
500
0
PBS

dE1/GFP

dE1/GFP-PPCBA

Fig. 8. Assessment of immune response and hepatotoxicity of Ad-PPCBA. (A) Induction

of IL-6 inammatory cytokine. At 6 h after intravenous injection of PBS, naked dE1/GFP
(2  1010 VP), PPCBA (50 mg/mL), or dE1/GFP-PPCBA complex (50 mg/mL PPCBA;
2  1010 VP) into mice, serum was collected and IL-6 levels were measured by ELISA.
*P < 0.05 for versus PBS-treated group. (B) Measurement of serum ALT and AST at 72 h
following intravenous administration of PBS, naked dE1/GFP (2  1010 VP), or dE1/GFPPPCBA complex (50 mg/mL PPCBA; 2  1010 VP). Data represent means SD and n 3
for each experimental condition. **P < 0.01 for versus naked dE1/GFP-treated group.

The daunting task of engineering the ideal Ad nanocomplex

therefore demands nesse and versatility in vector design, so that it
is ideally only active in the tumor microenvironment. Idiosyncratic
features of the tumor environment include hypoxia and acidication which can be used to target and activate suitably adapted
vectors. For instance, the acidic tumor environment has been
exploited in generating various pH-sensitive polymers for efcient
drug and gene delivery [24,47]. In addition, various proteins unique
to the hypoxic tumor such as hypoxia-inducible factor 1-a have
been targeted for enhanced Ad tumor recruitment and killing [10].
Despite the innovative development of various pH-sensitive polymers, they still suffer from lack of specicity and impeded interactions between the ligand and their target cell receptor possibly
due to functional group modications or steric hindrance from
conjugating targeting moieties with polymeric carriers. As a
compensatory strategy for the shortcomings of existent pHresponsive drug carriers, we engineered Ads ionically crosslinked to a pH-sensitive, bio-reducible polymer, PPCBA. Our data
demonstrate enhanced transduction efcacy of Ad-PPCBA complexes in vitro and elucidate Ad-polymer cellular trafcking
mechanisms, but also highlight the vector's in vivo therapeutic antitumor killing efciency when administered either intratumorally
or and intravenously.

Physical and chemical changes of Ad when conjugated with

PPCBA include: 1) a 1.7-fold increased particle size; 2) cationic
charge reversal of the capsid surface; and 3) encapsulation of Ad
particles by the polymer. Encapsulation of Ad by a cationic polymer,
as demonstrated with ABP, chitosan, and PNLG [18,19,42], is a useful
strategy to improve the Ad transduction efciency and DNA delivery because the positive charge of the polymer not only enables
the ionic interaction between the polymer and the virus but also
facilitates viral entry into host cells via electrostatic interaction. By
using this strategy, Ad-PPCBA complexes benet from enhanced
transduction efciency in both CAR-positive and -negative cells.
Transduction efciency studies revealed that the Ad-PPCBA vector
transduction efciency was modulated by extracellular pH, with
enhanced infectivity at pH 6.0 versus 7.4. Interestingly, the transduction efciency studies also demonstrated the vector's ability to
overcome CAR dependence, whereby transduction of CAR-negative
cancer cells in acidic conditions was signicantly elevated over
infectivity at physiological pH.
Cellular uptake mechanism studies revealed that dE1/GFPPPCBA enters cells via a macropinocytosis-mediated endocytosis,
which is a different route than naked Ad. In agreement with this
nding, macropinocytosis has been previously implicated as a
pivotal pathway in inducing CAR-independent endocytosis and
consequentially higher efciency of infection of CAR-negative cell
lines by Ad-polymer complexes [48]. Cell treatment with Bf-A,
which prevents viral escape from endosomes, signicantly
decreased gene transfer efciency of Ad-PPCBA complex (Fig. 5),
suggesting that Ad-PPCBA complexes require endosomal acidication for successful gene transduction, similar to what is required
by naked Ad.
Cancer cell killing efcacy was signicantly increased when
oncolytic Ad and PPCBA were complexed (RdB/shVEGF-PPCBA)
compared to killing by naked oncolytic Ad alone. This was most
likely due to enhanced cellular uptake and was not limited by target
cell CAR expression (Fig. 6). The RdB/shVEGF-PPCBA complex
decreased tumor cell VEGF expression in viral dose- and pHdependent manners; thus RdB/shVEGF complexed with PPCBA is
functionally active and reduces VEGF expression with higher efciency under acidic conditions characteristic of the tumor environment. Consistent with these ndings, RdB/shVEGF-PPCBA
complexes elicited a much greater antitumor efcacy versus naked
RdB/shVEGF-Ad in U87 xenografts, demonstrating the potent
therapeutic value of oncolytic Ad-PPCBA complexes. A key feature
of oncolytic Ad is its successive secondary replication in tumor
tissues. Therefore, the excellent antitumor activity of RdB/shVEGFPPCBA complexes may be attributable to the active replication of
oncolytic Ad in tumor tissues following efcient cancer cell infection in the hypoxic tumor microenvironment. Concurrent expression of siRNA specic against VEGF during viral replication
suppresses tumor-driven angiogenesis. Importantly, amplifying
therapeutic siRNA in a cancer cell-specic manner would decrease
potential side effects of siRNA while maximizing its therapeutic
value. Dual tumor targeting strategy by oncolytic Ad-mediated
cancer cell specic killing efcacy and tumor-microenvironment
(hypoxia)-specic Ad infection enhancement endowed by PPCBA
complex can further increase the therapeutic efcacy of Admediated cancer gene therapeutics with decrease of unwanted
side-effect.
Systemic Ad administration, in contrast to intratumoral injection, presents entirely different challenges: short blood retention
time, acute liver tropism and toxicity, and innate immune system
activation. In fact, the efcacy and safety of any Ad-conjugate vector
engineered for systemic administration have been primarily
dependent on the vector's immunogenicity. A leading strategy to
reduce immune responses to the Ad vector has been viral

C.Y. Moon et al. / Biomaterials 41 (2015) 53e68

encapsulation [3]. Conjugation of various polymers on the Ad surface such as PEGylation has fashioned stealth Ad particles by
masking Ad capsid molecules and reducing absorption of host
antibody to the virus [49]. Encapsulation of Ad with PPCBA results
in pH-dependent particle morphologies. By TEM, the polymer is
tightly wound around the Ad at pH 7.4, but appears loosely
expanded in acidic conditions (i.e., pH 6.0). Based on the TEM images, we investigated whether cloaking of Ad with PPCBA at neutral
conditions produces the similar stealth effects as does PEGylation. As an indicator of innate immune activation after Ad systemic
administration, serum IL-6 levels were signicantly reduced in AdPPCBA-injected mice versus animals that received naked Ad. This
decreased IL-6 following Ad-PPCBA complex administration may be
due to reduced vector uptake by Kupffer cells and polymershielding of the Ad surface to the immune system.
Reduced immunogenicity of Ad-PPCBA complexes was supported by in vivo studies in which intravenously-injected dB/
shVEGF-PPCBA's showed therapeutic efcacy against established
tumors in a mouse xenograft model. Conversely, the therapeutic
efcacy of naked oncolytic Ad was signicantly attenuated when it
was administered systemically, even though same oncolytic Ad
elicited marked antitumor efcacy when it was administered
intratumorally. This observation conrmed that systemically
delivered naked Ad is inactivated and cleared by immune system
and non-specic liver uptake, results in low anti-tumor efcacy.
Importantly, Ad-related liver toxicity as measured by serum ALT
and AST levels was nonexistent in animals that received systemic
administration of RdB/shVEGF-PPCBA. Taken together, this pHsensitive, PPCBA shielded, oncolytic Ad shows negligible immunogenicity and liver toxicity, and potent antitumor activity, which
confers great potential for in vivo clinical applications to selectively
target and kill both primary and metastatic tumors.
5. Conclusions
We have synthesized a pH-sensitive and bio-reducible polymer,
mPEG-Pip-CBA (PPCBA), and engineered a dual tumor targeting
polymer-complexed Ad vector system (Ad-PPCBA) that targets tumors via electrostatic interaction between the virus and target cell.
By taking advantage of dual tumor targeting ability from oncolytic
Ad's innate oncolytic prowess and the exibility of pH-sensitive
polymer, the oncolytic Ad-PPCBA complex demonstrated enhanced
therapeutic efciency in both CAR-positive and -negative cells, which
was signicantly increased in the low pH environment reective of
the tumor milieu. The potent therapeutic potential and minimal
toxicity of Ad-PPCBA should encourage the further development of
other smart, tumor-selective oncolytic Ad complexes that exploit
targetable features of the tumor microenvironment.
Acknowledgments
This work was supported by grants from the Ministry of Trade,
Industry & Energy (10030051 Dr. C-O. Yun), the National Research
Foundation of Korea (2010-0029220, 2013M3A9D3045879,
2013K1A1A2A02050188, Dr. C-O. Yun), The Korea Food and Drug
Administration (KFDA-13172-306, Dr. C-O. Yun), Basic Research
Programs
by
National
Research
Foundation
of
Korea
(2013R1A1A2012483, Dr. D. Kasala) and the National Institutes of
Health, USA (CA 107070, Dr. S-W. Kim).
Appendix A. Supplementary data
Supplementary data related to this article can be found online at
http://dx.doi.org/10.1016/j.biomaterials.2014.11.021.

67

References
[1] Wold WS, Toth K. Adenovirus vectors for gene therapy, vaccination and cancer
gene therapy. Curr Gene Ther 2013;13:421e33.
[2] Choi IK, Yun CO. Recent developments in oncolytic adenovirus-based immunotherapeutic agents for use against metastatic cancers. Cancer Gene Ther
2013;20:70e6.
[3] Choi JW, Lee JS, Kim SW, Yun CO. Evolution of oncolytic adenovirus for cancer
treatment. Adv Drug Deliv Rev 2012;64:720e9.
[4] Bachtarzi H, Stevenson M, Fisher K. Cancer gene therapy with targeted adenoviruses. Expert Opin Drug Deliv 2008;5:1231e40.
[5] Yun CO. Overcoming the extracellular matrix barrier to improve intratumoral
spread and therapeutic potential of oncolytic virotherapy. Curr Opin Mol Ther
2008;10:356e61.
[6] Guo ZS, Thorne SH, Bartlett DL. Oncolytic virotherapy: molecular targets in
tumor-selective replication and carrier cell-mediated delivery of oncolytic
viruses. Biochim Biophys Acta 2008;1785:217e31.
[7] Russell SJ, Peng KW, Bell JC. Oncolytic virotherapy. Nat Biotechnol 2012;30:
658e70.
[8] Harrop R, John J, Carroll MW. Recombinant viral vectors: cancer vaccines. Adv
Drug Deliv Rev 2006;58:931e47.
[9] Reid T, Galanis E, Abbruzzese J, Sze D, Andrews J, Romel L, et al. Intra-arterial
administration of a replication-selective adenovirus (dl1520) in patients with
colorectal carcinoma metastatic to the liver: a phase I trial. Gene Ther 2001;8:
1618e26.
[10] Kwon OJ, Kim PH, Huyn S, Wu L, Kim M, Yun CO. A hypoxia- and {alpha}fetoprotein-dependent oncolytic adenovirus exhibits specic killing of hepatocellular carcinomas. Clin Cancer Res 2010;16:6071e82.
[11] Yu W, Fang H. Clinical trials with oncolytic adenovirus in China. Curr Cancer
Drug Target 2007;7:141e8.
[12] Cichon G, Schmidt HH, Benhidjeb T, Loser P, Ziemer S, Haas R, et al.
Intravenous administration of recombinant adenoviruses causes thrombocytopenia, anemia and erythroblastosis in rabbits. J Gene Med 1999;1:
360e71.
[13] Meier O, Greber UF. Adenovirus endocytosis. J Gene Med 2003;5:451e62.
[14] Kim J, Kim PH, Kim SW, Yun CO. Enhancing the therapeutic efcacy of
adenovirus in combination with biomaterials. Biomaterials 2012;33:1838e50.
[15] Wang C, Pham PT. Polymers for viral gene delivery. Expert Opin Drug Deliv
2008;5:385e401.
[16] Kang E, Yun CO. Current advances in adenovirus nanocomplexes: more
specicity and less immunogenicity. BMB Rep 2010;43:781e8.
[17] Kasala D, Choi JW, Kim SW, Yun CO. Utilizing adenovirus vectors for gene
delivery in cancer. Expert Opin Drug Deliv 2014;11:379e92.
[18] Lee WJ, Kim YO, Choi IK, Rah DK, Yun CO. Adenovirus-relaxin gene therapy for
keloids: implication for reversing pathological brosis. Br J Dermatol
2011;165:673e7.
[19] Kim J, Li Y, Kim SW, Lee DS, Yun CO. Therapeutic efcacy of a systemically
delivered oncolytic adenovirus - biodegradable polymer complex. Biomaterials 2013;34:4622e31.
[20] Eto Y, Yoshioka Y, Mukai Y, Okada N, Nakagawa S. Development of PEGylated
adenovirus vector with targeting ligand. Int J Pharm 2008;354:3e8.
[21] Han J, Zhao D, Zhong Z, Zhang Z, Gong T, Sun X. Combination of adenovirus
and cross-linked low molecular weight PEI improves efciency of gene
transduction. Nanotechnology 2010;21:105106.
[22] Park JW, Mok H, Park TG. Physical adsorption of PEG grafted and blocked
poly-L-lysine copolymers on adenovirus surface for enhanced gene transduction. J Control Release 2010;142:238e44.
[23] Fischer D, Li Y, Ahlemeyer B, Krieglstein J, Kissel T. In vitro cytotoxicity testing
of polycations: inuence of polymer structure on cell viability and hemolysis.
Biomaterials 2003;24:1121e31.
[24] Lee ES, Gao Z, Bae YH. Recent progress in tumor pH targeting nanotechnology.
J Control Release 2008;132:164e70.
[25] Kim PH, Sohn JH, Choi JW, Jung Y, Kim SW, Haam S, et al. Active targeting and
safety prole of PEG-modied adenovirus conjugated with herceptin. Biomaterials 2011;32:2314e26.
[26] Yoo JY, Kim JH, Kwon YG, Kim EC, Kim NK, Choi HJ, et al. VEGF-specic short
hairpin RNA-expressing oncolytic adenovirus elicits potent inhibition of
angiogenesis and tumor growth. Mol Ther 2007;15:295e302.
[27] Ko J, Park K, Kim YS, Kim MS, Han JK, Kim K, et al. Tumoral acidic extracellular
pH targeting of pH-responsive MPEG-poly(beta-amino ester) block copolymer
micelles for cancer therapy. J Control Release 2007;123:109e15.
[28] Zhang CY, Yang YQ, Huang TX, Zhao B, Guo XD, Wang JF, et al. Self-assembled
pH-responsive MPEG-b-(PLA-co-PAE) block copolymer micelles for anticancer
drug delivery. Biomaterials 2012;33:6273e83.
[29] Park Y, Kang E, Kwon OJ, Hwang T, Park H, Lee JM, et al. Ionically crosslinked
Ad/chitosan nanocomplexes processed by electrospinning for targeted cancer
gene therapy. J Control Release 2010;148:75e82.
[30] Kim PH, Kim TI, Yockman JW, Kim SW, Yun CO. The effect of surface modication of adenovirus with an arginine-grafted bioreducible polymer on
transduction efciency and immunogenicity in cancer gene therapy. Biomaterials 2010;31:1865e74.
[31] Wang LH, Rothberg KG, Anderson RG. Mis-assembly of clathrin lattices on
endosomes reveals a regulatory switch for coated pit formation. J Cell Biol
1993;123:1107e17.

68

C.Y. Moon et al. / Biomaterials 41 (2015) 53e68

[32] Aoki T, Nomura R, Fujimoto T. Tyrosine phosphorylation of caveolin-1 in the

endothelium. Exp Cell Res 1999;253:629e36.
[33] Perumal OP, Inapagolla R, Kannan S, Kannan RM. The effect of surface functionality on cellular trafcking of dendrimers. Biomaterials 2008;29:3469e76.
[34] Meier O, Greber UF. Adenovirus endocytosis. J Gene Med 2004;6(Suppl. 1):
S152e63.
[35] Xu Z, Tian J, Smith JS, Byrnes AP. Clearance of adenovirus by Kupffer cells is
mediated by scavenger receptors, natural antibodies, and complement. J Virol
2008;82:11705e13.
[36] Muruve DA, Barnes MJ, Stillman IE, Libermann TA. Adenoviral gene therapy
leads to rapid induction of multiple chemokines and acute neutrophildependent hepatic injury in vivo. Hum Gene Ther 1999;10:965e76.
[37] Lieber A, He CY, Meuse L, Schowalter D, Kirillova I, Winther B, et al. The role of
Kupffer cell activation and viral gene expression in early liver toxicity after
infusion of recombinant adenovirus vectors. J Virol 1997;71:8798e807.
[38] O'Riordan CR, Lachapelle A, Delgado C, Parkes V, Wadsworth SC, Smith AE,
et al. PEGylation of adenovirus with retention of infectivity and protection
from neutralizing antibody in vitro and in vivo. Hum Gene Ther 1999;10:
1349e58.
[39] Fisher KD, Green NK, Hale A, Subr V, Ulbrich K, Seymour LW. Passive tumour
targeting of polymer-coated adenovirus for cancer gene therapy. J Drug Target
2007;15:546e51.
[40] Fisher KD, Stallwood Y, Green NK, Ulbrich K, Mautner V, Seymour LW. Polymer-coated adenovirus permits efcient retargeting and evades neutralising
antibodies. Gene Ther 2001;8:341e8.
[41] Kwon OJ, Kang E, Kim S, Yun CO. Viral genome DNA/lipoplexes elicit in situ
oncolytic viral replication and potent antitumor efcacy via systemic delivery.
J Control Release 2011;155:317e25.

[42] Kwon OJ, Kang E, Choi JW, Kim SW, Yun CO. Therapeutic targeting of chitosanPEG-folate-complexed oncolytic adenovirus for active and systemic cancer
gene therapy. J Control Release 2013;169:257e65.
[43] Yao XL, Yoshioka Y, Ruan GX, Chen YZ, Mizuguchi H, Mukai Y, et al. Optimization and internalization mechanisms of PEGylated adenovirus vector with
targeting peptide for cancer gene therapy. Biomacromolecules 2012;13:
2402e9.
[44] Yao X, Yoshioka Y, Morishige T, Eto Y, Narimatsu S, Kawai Y, et al. Tumor
vascular targeted delivery of polymer-conjugated adenovirus vector for cancer gene therapy. Mol Ther 2011;19:1619e25.
[45] Kim J, Nam HY, Kim TI, Kim PH, Ryu J, Yun CO, et al. Active targeting of RGDconjugated bioreducible polymer for delivery of oncolytic adenovirus
expressing shRNA against IL-8 mRNA. Biomaterials 2011;32:5158e66.
[46] Morrison J, Briggs SS, Green NK, Thoma C, Fisher KD, Kehoe S, et al. Cetuximab
retargeting of adenovirus via the epidermal growth factor receptor for
treatment of intraperitoneal ovarian cancer. Hum Gene Ther 2009;20:
239e51.
[47] Zhang X, Lin Y, Gillies RJ. Tumor pH and its measurement. J Nucl Med
2010;51:1167e70.
[48] Lee CH, Kasala D, Na Y, Lee MS, Kim SW, Jeong JH, et al. Enhanced therapeutic
efcacy of an adenovirus-PEI-bile-acid complex in tumors with low coxsackie
and adenovirus receptor expression. Biomaterials 2014;35:5505e16.
[49] Croyle MA, Chirmule N, Zhang Y, Wilson JM. Stealth adenoviruses blunt cellmediated and humoral immune responses against the virus and allow for
signicant gene expression upon readministration in the lung. J Virol
2001;75:4792e801.