ROS-responsive “smart” polymeric conjugate: Synthesis, characterization and proof-of-concept study

Article abstract of DOI:10.1016/j.ijpharm.2019.118655

New approaches integrating stimuli-responsive linkers into prodrugs are currently emerging. These “smart” prodrugs can enhance the effectivity of conventional prodrugs with promising clinical applicability. Oxidative stress is central to several diseases, including cancer. Therefore, the design of prodrugs that respond to ROS stimulus, allowing a selective drug release in this condition, is fairly encouraging. Aiming to investigate the ROS-responsiveness of prodrugs containing the ROS-cleavable moiety, Thioketal (TK), we performed proof-of-concept studies by synthesizing ROS-responsive conjugate, namely mPEG-TK-Cy5, through exploiting Cy5 fluorescent dye. We demonstrated that, differently to non-ROS-responsive control conjugate (mPEG-Cy5), mPEG-TK-Cy5 shows a selective release of Cy5 in response to ROS in both, ROS-simulated conditions and in vitro on glioblastoma cells. Our results confirm the applicability of TK-technology in the design of ROS-responsive prodrugs, which constitutes a promising approach in cancer treatment. The translatability of this technology for other diseases treatment makes this a highly relevant and promising approach.

Full text of DOI:10.1016/j.ijpharm.2019.118655

International Journal of Pharmaceutics 570 (2019) 118655
Contents lists available at ScienceDirect
International Journal of Pharmaceutics
journal homepage: www.elsevier.com/locate/ijpharm
ROS-responsive “smart” polymeric conjugate: Synthesis, characterization
and proof-of-concept study
Natalia Oddone , Francesca Pederzoli , Jason T. Duskey , Chiara A. De Benedictisb,c,
a
b
a
b,c,d
a
a
a
Andreas M. Grabrucker
, Flavio Forni , Maria Angela Vandelli , Barbara Ruozi ,
a,⁎
Giovanni Tosi
a
Nanotech Lab TeFarTI Group, University of Modena and Reggio Emilia, Department of Life Sciences, Modena, Italy
Department of Biological Sciences, University of Limerick, Limerick, Ireland
Bernal Institute, University of Limerick, Limerick, Ireland
b
c
d
Health Research Institute (HRI), University of Limerick, Limerick, Ireland
A R T I C L E I N F O
A B S T R A C T
Keywords:
ROS stimulus
ROS-responsive conjugate
TK-technology
Cancer treatment applicability
New approaches integrating stimuli-responsive linkers into prodrugs are currently emerging. These “smart”
prodrugs can enhance the effectivity of conventional prodrugs with promising clinical applicability. Oxidative
stress is central to several diseases, including cancer. Therefore, the design of prodrugs that respond to ROS
stimulus, allowing a selective drug release in this condition, is fairly encouraging. Aiming to investigate the ROS-
responsiveness of prodrugs containing the ROS-cleavable moiety, Thioketal (TK), we performed proof-of-concept
studies by synthesizing ROS-responsive conjugate, namely mPEG-TK-Cy5, through exploiting Cy5 fluorescent
dye. We demonstrated that, differently to non-ROS-responsive control conjugate (mPEG-Cy5), mPEG-TK-Cy5
shows a selective release of Cy5 in response to ROS in both, ROS-simulated conditions and in vitro on glio-
blastoma cells. Our results confirm the applicability of TK-technology in the design of ROS-responsive prodrugs,
which constitutes a promising approach in cancer treatment. The translatability of this technology for other
diseases treatment makes this a highly relevant and promising approach.
1. Introduction
through a “simple” drug administration, allowing a more efficacious
and less toxic treatment.
Drug delivery systems (DDS) aiming to improve the solubility and
In terms of pathologies, cancer constitutes one of the most lethal
diseases of high worldwide incidence and difficulty to be cured (Kim
et al., 2014; Xu et al., 2018). The efficacy of current treatments is still
limited by side effects, with chemotherapy as first-line therapy of
choice against this disease (Chakraborty et al., 2018; Dragojevic et al.,
2015), being well-known for the lack of selectivity as well as the hy-
drophobicity of the majority of chemotherapeutics (Chakraborty et al.,
2018).
Amongst the number of possible targets and features which could be
exploited with the aim of creating a more selective and “smart” drug
delivery, great attention was given to reactive oxygen species (ROS)
such as hydrogen peroxide, hydroxyl radicals, singlet oxygen, super-
oxide, nitric oxide, nitroxyl and nitrogen dioxide, since their con-
centrations are increased in several diseases(Gupta et al., 2012), e.g.,
inflammatory diseases, neurodegenerative diseases and cancer(Salim,
2017; Wang and Michaelis, 2010). In cancer, it is estimated that the
bioavailability of poorly water-soluble drugs have been broadly de-
veloped (Tiwari et al., 2012). Amongst them, polymer-drug conjugates
or prodrugs were demonstrated to be a straightforward strategy to in-
crease the solubility and selectivity of certain drugs, enhancing their
effectivity as well as reducing their adverse effects (Dragojevic et al.,
2
015). The potential of this approach is incremented by the possibility
of further selective engineering of the prodrug by incorporating a
cleavable linker between the polymer and drug, which can be en-
zymatically hydrolyzed inside cells or be responsive to a stimulus, thus
being more tailored to the pathological features of a given disease
(
Chang et al., 2016). Therefore, by carefully planning the design of the
prodrug featured by the presence of a stimulus- responsive moiety, able
to be activated within the stimulus event, the selective “smart” release
of drugs at pathological sites can be obtained.
The great potential of this approach lies in the real possibility of
creating novel approaches for curing diseases more selectively than
2 2
intracellular concentration of H O can reach 10–100 μM, while in
⁎
Corresponding author.
E-mail address: gtosi@unimore.it (G. Tosi).
https://doi.org/10.1016/j.ijpharm.2019.118655
Received 4 June 2019; Received in revised form 26 August 2019; Accepted 30 August 2019
Available online 31 August 2019
0378-5173/ © 2019 Elsevier B.V. All rights reserved.

N. Oddone, et al.
International Journal of Pharmaceutics 570 (2019) 118655
Scheme 1. Syntheses Scheme. A. Synthesis of TK-C.L. B. Synthesis of ROS-responsive polymer mPEG-TK-COOH.
normal cells, the estimated H
2
O
2
concentration is between 0.001 and
All the solvents used, 3-Mercaptopropionic acid (3-MPA), di-
chloromethane (DCM), dimethylformamide (DMF), acetone, hydrogen
peroxide (H O ) and hexane, were of analytical grade and used without
2 2
0
.7 μM (Hagen et al., 2012). For instance, gliomas are characterized by
enhanced oxidative stress, which is a feature that increases the re-
sistance to chemotherapy (Sharma et al., 2007).
further purification. A Milli-Q water system (Millipore, Bedford, MA,
USA), supplied with distilled water, provided high-purity water (18 V).
F12-K and trypsin (TrypLE Select Enzyme (1X)) were purchased from
Gibco. DMEM High Glucose, fetal bovine serum (FBS), and peni-
cillin–streptomycin were purchased from Sigma-Aldrich. CellROX™
Green Reagent was purchased from Thermo Fisher Scientific. The pri-
mary antibodies were purchased from Cell Signaling Technology (Rab5
(C8B1), Rabbit mAb), Thermo Fisher Scientific (EEA1, Rabbit mAb),
and Abcam (Clathrin heavy chain, Rabbit pAb). Secondary Alexa Fluor
conjugated antibodies were purchased from Life Technologies.
Thus, polymer prodrugs which respond to a ROS stimulus may
confer selectivity in cancer therapy with the activated delivery of drugs
in such target sites (Lee and Kataoka, 2009). Currently, several ROS-
responsive polymer platforms based on poly (propylene sulfide), sele-
nium/tellurium, polyoxalate, poly(proline), boronic ester, and thioketal
(
TK) are being investigated (Song et al., 2013). Their ROS responsive
mechanisms rely on either a transition from hydrophobic to hydrophilic
states or a cleavage reaction of the polymer platforms (Saravanakumar
et al., 2017). TK is a subclass of thioether group, which is cleaved to
thiol-containing groups upon the exposure to ROS (McEnery et al.,
2
016). In vitro studies on TK ROS-responsiveness to several ROS types
2
2
.2. Synthesis
(
Shim and Xia, 2013) (i.e., H , hydroxyl radical and superoxide)
2 2
O
showed the ability of the TK moiety to be cleaved by the most relevant
ROS that are encountered in pathological conditions. Most importantly,
the byproducts originated from their cleavage are rapidly cleared from
the circulation (Martin et al., 2014; Zhang et al., 2017). Therefore, they
constitute good candidates as moieties for the design of “smart” ROS
responsive polymers.
Another limitation of current chemotherapeutics is based on the low
water solubility that strongly hampers the administration and bioa-
vailability of drugs. One of the most exploited approaches to overcome
this limitation consists of the use of poly(ethylene glycol) (PEG), a
synthetic water-soluble polymer, FDA approved for its use in pharma-
ceutical formulations (Turecek et al., 2016). The conjugation of PEG
with either biotherapeutic drugs or small molecules, known as PEGy-
lation, was demonstrated to improve drug solubility and stability and to
prolong the circulation time of drugs, reducing their renal clearance
.2.1. ROS-cleavable thioketal containing linker (TK-C.L.) synthesis
The synthesis of ROS-cleavable thioketal linker (Scheme 1A) was
performed according to previous works without further modifications
(
(
Li et al., 2016a,b; Yue et al., 2016). Briefly, a mixture of 3-MPA
49.1 mmol) and anhydrous acetone (98.2 mmol) was saturated in dry
hydrogen chloride atmosphere and stirred at room temperature for 4 h.
After that, the reaction was quenched by placing the mixture on an ice-
salt mixture until the crystallization was complete. The crystals were
filtered by vacuum, washed gently with hexane and cold water, and
finally dried in vacuum. The product was fully characterized by
1
H
NMR and ESI-MS.
2
.2.2. ROS-responsive mPEG-TK-COOH polymer synthesis
mPEG-TK-COOH synthesis was performed according to Li et al. with
some modifications (Li et al., 2016a) and is shown in Scheme 1B.
Briefly, mPEG-NH (0.1 mmol), TK-C.L. (1 mmol), EDC. HCl (1.2 mmol)
(
Hu and Tirelli, 2012; Li and Mahato, 2017). Accordingly, combining
2
TK linkers and PEG conjugation may produce administrable and se-
lective prodrugs produced for the delivery of antitumor drugs in cancer
treatment. In order to create stable proofs-of-concept regarding the
applicability of this technology, in this study, we decided to use a
and NHS (1.2 mmol) were well dissolved in DCM (5 mL) and stirred at
room temperature for 48 h. The reaction was stopped at 48 h; the sol-
vent was removed by rotary evaporation and the product precipitated
by co-solvent precipitation in DCM and diethyl ether. After drying in
vacuum, the product was dissolved in DMF and dialyzed against dis-
tilled water for 72 h by using an MWCO: 3500 Da dialysis membrane.
After freeze-drying, the expected mPEG-TK-COOH was collected as a
2
fluorescent dye (NIR fluorescent dye, Cy5-NH ) to be conjugated to PEG
and TK in order to give clear of the technology’s feasibility and success
in selective ROS-mediated cleavage. Using in vitro validation of Cy5
ability to be selectively released in response to cancer pathological
conditions (glioblastoma and neuroblastoma cells), we therefore in-
vestigated the potential of TK-technology to be used in cancer treat-
ment.
1
white powder. The derivatization yield was determined from H NMR
characterization data and indirectly by quantifying non-reacted mPEG-
NH
2
through Quantitative Ninhydrin Assay.
The determination of the derivatization yield through 1H NMR was
achieved by integrating peak data at chemical shifts: 3.2 and 3.48 ppm,
which correspond to protons of –CH groups next to amine group (of
un-reacted mPEG-NH ) and amide bond (formed by the covalent con-
jugation between mPEG-NH and TK-C.L.), respectively.
2
. Materials and methods
2
2
2.1. Materials
2
The determination of the derivatization yield by colorimetric assay
with ninhydrin was performed by placing 100 μL of mPEG-TK-COOH
solutions (2.5–4 mg/mL) into the wells of a 96-well plate and 75 μL of
ninhydrin colour reagent, respectively. Then, the plate was incubated at
80 °C for 30 min. After incubation, the plate was cooled to room tem-
perature, and 100 μL of stabilizing solution (50% v/v ethanol) was
2
Methoxy-polyethylene glycol amine (mPEG-NH , Mw 5000 Da) and
mPEG propionic acid (Mw 5000 Da) were purchased from JenKem
Technology and Sigma-Aldrich, respectively. N-hydroxy succinimide
NHS) and N-(3-dimethylaminopropyl)-N-ethylcarbodiimide hydro-
chloride (EDC. HCl) were provided by Sigma-Aldrich and used directly.
(
2

N. Oddone, et al.
International Journal of Pharmaceutics 570 (2019) 118655
Scheme 2. Synthesis scheme of ROS-responsive polymer fluorescent conjugate, mPEG-TK-Cy5.
added to each well. Finally, the absorbance was measured at 570 nm
through a Multiskan Microplate Reader (Spectrum Finstruments®).
removed by rotary evaporation, and after drying the precipitate in a
vacuum desiccator, the product was dissolved in DMF and dialyzed
(MWCO: 3500 Da) against the same solvent for 24 h and against MilliQ
water for 48 h, respectively. The conjugates were isolated by freeze-
drying and kept in a desiccator.
Different concentrations of mPEG-NH
for the generation of a calibration curve from which the concentration
of non-reacted mPEG-NH on the sample and therefore its percentage,
2
in aqueous solution were used
2
were calculated. Linearity was assumed in the range of 0.015–1.5 mg/
mL (r2 = 0.9989). mPEG-TK-COOH derivatization yield was achieved
2
.2.4. Determination of mPEG-TK-Cy5 derivatization yield and Cy5-
by subtracting 100% to % mPEG-NH
2
.
loading on mPEG-TK-Cy5 and mPEG-Cy5 fluorescent conjugates.
The derivatization yield of mPEG-TK-Cy5 was indirectly determined
by quantifying the non-reacted mPEG-TK-COOH on mPEG-TK-Cy5
sample through RP-HPLC. The RP-HPLC apparatus comprised a Model
PU2089 pump provided with an injection valve with a 50 µL sample
loop (7725i, Jasco), UV–Vis detector (UV1575, Jasco) and fluorescent
detector (FP-2020 Plus, Jasco). UV–Vis detector and fluorescence de-
tector were set to λ = 210 nm and λexc = 640 nm/ λem = 660 nm,
respectively, and column temperature to 37 °C. The analysis was per-
formed in gradient conditions with mobile phase A- Water 100%: TFA
2
.2.3. Synthesis of ROS-responsive and non-ROS-responsive fluorescent
conjugate: mPEG-TK-Cy5 and mPEG-Cy5
mPEG-TK-COOH (Scheme 2) or mPEG propionic acid (Scheme 3)
(
(
11.46 µmol) were activated with EDC.HCl (57.4 µmol) and NHS
57.3 µmol) in 1.5 mL of DCM for an hour. Then, Cy5-NH (12.8 µmol)
2
diluted in 0.5 mL of DCM was added to the mixture and stirred for 48 h
at room temperature. The product was precipitated by co-solvent pre-
cipitation in DCM/cold diethyl ether on ice-bath. Diethyl ether was
0.1% and B-ACN 100%: TFA 0.1%, at a flow rate of 1 mL/min. The
samples were injected on HC-C18 (250 mm × 4.60 mm, Agilent
Technologies) and eluted following this gradient: 0–5 min, 70% A;
5–20 min, 30% A; 20–21, 30% A; 21–25 min, 70% A and 25–30, 70% A.
The Chromatographic peak areas of the samples were recorded and
analyzed using a JASCO software (JascoBorwin 1.5). Un-reacted mPEG-
TK-COOH content was calculated against a calibration curve performed
with mPEG-TK-COOH polymer (linearity was assumed in the
0.3–9.5 mg/mL; r2 = 0.9956 range).
Cy5 content on mPEG-Cy5 and mPEG-TK-Cy5 conjugates was ex-
pressed as µg of Cy5 per mg of conjugate and was determined by
measuring the fluorescence of conjugate solutions (1–2.5 mg/mL) in
methanol. The measurements were performed in a fluorometer (FP-
8
200, Jasco) with λexc = 620 nm and λem = 654 nm. Cy5 content on
the conjugates was obtained against a calibration curve of Cy5-amine in
methanol (linearity was assumed in the 0.103–1.03 mg/mL;
r2 = 0.9903 range).
2
2
.3. Characterization
Scheme 3. Synthesis of non-ROS-responsive (control) polymer fluorescent
conjugate, mPEG-Cy5.
.3.1. Nuclear magnetic resonance (NMR)
TK-C.L. 1H NMR spectra were acquired on Avance400- Bruker
3

N. Oddone, et al.
International Journal of Pharmaceutics 570 (2019) 118655
spectrometer in CDCl
3
and d6-DMSO solvents with TMS as an internal
PBS and fixed with 4% paraformaldehyde solution (PFA) in 1X PBS. Cell
nuclei were counterstained with DAPI and coverslips subsequently
mounted using Vecta Mount (Vector Laboratories, USA). The cells were
observed using a confocal laser-scanning microscope (Zeiss LSM710).
CellROX fluorescence of confocal images of C6, DI TNC1 and SH SY5Y
cells treated with either of the conjugates was quantified using ImageJ
(National Institutes of Health), by measuring at least 20 cells per con-
dition and cell line.
standard. On the other hand, mPEG-TK-COOH, mPEG-Cy5 and mPEG-
1
TK-Cy5 H NMR spectra were obtained in Avance III 600 HD- Bruker in
3
CDCl (mPEG-TK-COOH) and d6-DMSO (mPEG-Cy5 and mPEG-TK-
Cy5) solvents. The identification of all proton signals in mPEG-TK-
COOH polymer was completed through the information collected by the
1
13
following analysis: H, C, Homonuclear correlation spectroscopy
(
(
(
COSY), Heteronuclear single-quantum correlation spectroscopy
HSQC) and Heteronuclear multiple-bond correlation spectroscopy
HMBC) NMR.
2.5.3. mPEG-TK-Cy5 in vitro Cy5 release studies on C6 and SH SY5Y cells
5
Cells were seeded (1 × 10 cells/mL) on poly-L-lysine (0.1 mg/mL;
2
.3.2. MALDI-TOF/TOF mass spectrometry
mPEG-TK-COOH, mPEG-TK-Cy5, and mPEG-Cy5 mass spectra were
Sigma-Aldrich) coated glass coverslips in a 24-well plate and incubated
at 37 °C for 24 h. The cells were then incubated for 72 h with either
mPEG-Cy5 or mPEG-TK-Cy5 at equivalent Cy5 concentrations of 0.9
and 4.5 µg/mL, respectively. After the incubation period, cells were
fixed with 4% PFA at RT for 15 min and processed for im-
munocytochemistry. After washing with three times with 1X PBS con-
taining 0.2% Triton X-100 at RT, blocking was performed with a
blocking solution (BS: 10% FBS in 1X PBS) for 1 h at RT. Subsequently,
samples were incubated with primary antibody in BS at RT for 2 h. After
washing three times with 1X PBS, incubation with Alexa488 secondary
antibody dissolved in BS for 1 h was followed. After washing three
times with 1X PBS, Cell nuclei were counterstained with DAPI and
coverslips subsequently mounted using Vecta Mount (Vector
Laboratories, USA). The cells were observed using a confocal laser
scanning microscope (Zeiss LSM710). Confocal images were subse-
quently analyzed by Image J and Image- Pro 10, through applying a
software tool that permits to plot the fluorescence intensity along a line
segment traced in a region of interest. In order to evaluate cytoplasmic
Cy5 intensity, a line segment was traced through the cytoplasm of se-
lected cells, obtaining a plot of fluorescence distribution along the line.
acquired with Flex control- Ultraflex TOF/TOF, MALDI-TOF/TOF mass
spectrometer). Briefly, the samples for MALDI-TOF analysis were pre-
pared by mixing 2,5-DHB matrix in water/EtOH (9:1, v/v) and polymer
or conjugate in water in a 1:1 v/v ratio. 1 µL of each sample was de-
posited onto the plate, and after drying, the plate was introduced into
the instrument. The instrument was operated in positive charge and
reflection mode, and each mass spectrum was acquired by averaging
8
0–800 consecutive laser shots.
2.4. ROS-responsive validation studies of mPEG-TK-COOH and mPEG-TK-
Cy5 in ROS-simulated conditions
2.4.1. ROS-mediated TK cleavage from mPEG-TK-COOH polymer in ROS-
simulated conditions
This procedure was exploited in order to assess the ability of TK to
be cleaved under ROS-simulated conditions and it was adapted and
modified from Yue et al. (Yue et al., 2016) mPEG-TK-COOH
(
1
10.82 μmol of thioketal groups) was incubated in water with either
00 µM, 1 mM, 100 mM or 400 mM H
2
O
2
2
and 3.2 μM CuCl at 37 °C for
4
8 h. Once the incubation was completed, the polymer was dialyzed
2.6. Statistics
(
dialysis membrane MWCO: 3500 Da) overnight against MilliQ water.
The polymer inside the dialysis bag was finally freeze- dried and ana-
lyzed through H NMR (400 MHz, CDCl ).
3
Statistical analysis was performed with Prisma version 5. All data
are shown as mean of at least three experiments ± SD if not indicated
otherwise. For comparisons, unpaired t-tests or one/two-way ANOVA
with Bonferroni post-hoc test for pairwise comparison were performed.
Statistically significant differences are indicated in the figures as fol-
1
2
.4.2. mPEG-TK-Cy5 ROS-responsiveness in ROS-simulated conditions
mPEG-TK-Cy5 or mPEG-Cy5 (2 mg) were incubated in a mixture
containing 400 mM H and 3.2 μM CuCl in DMF/H O (5.4:4.6, v/v)
at 37 °C. This protocol was adapted and modified from Yue et al.
2016), Li et al. (2016a,b). At selected times: 2, 6, 24, and 48 h, samples
*
*
***
2
O
2
2
2
lows: *p ≤ 0.05, p ≤ 0.01 and p ≤ 0.001.
(
3. Results
aliquots of 50 μL were injected on RP-HPLC. The RP-HPLC system used
in these studies is described in Section 2.2.4.
3.1. Synthesis and characterization of TK-C.L. and mPEG-TK-COOH
2
2
.5. In vitro studies of mPEG-TK-Cy5 on cells
TK-C.L. was obtained with a yield of 25% (Scheme 1A). The gen-
eration of by-products that were successfully eliminated during the
purification process accounts for this low yield. Characterization was
.5.1. Cell culture
1
1
C6 Rat Glioblastoma cells (ATCC) were cultured in F-12 k culture
obtained by H NMR, showing H signals with chemical shifts: 2.75 (t,
4H); 2.51 (t, 4H); 1.54 (s, 6H) in d6-DMSO and 2.932 (t, 4H), 2.702 (t,
4H) and 1.623 (s, 6H) in deuterated dichloromethane, respectively (Fig.
S1A and B). Thus, TK-C.L. was properly characterized, confirming the
obtention of the correct chemical structure, in agreement to those re-
medium with 20% FBS and 1% penicillin/streptomycin. DI TNC1 Rat
Astrocyte cells (ATCC) were cultured in DMEM High Glucose medium
with 10% FBS and 1% penicillin/streptomycin. SH SY5Y
Neuroblastoma cells (Sigma Aldrich) were cultured in Ham’s
F12:EMEM (1:1) with 2 mM Glutamine, 1% Non Essential Amino Acids
6
ported previously in d -DMSO (Yue et al., 2016). In addition, we con-
(
NEAA) and 15% FBS. All cell lines were kept at 37 °C under a humi-
firmed the molecular weight to be 252.04 g/mol, which corresponds
dified atmosphere containing 5% CO
2
.
precisely to the theoretical molecular weight of TK-C.L. (252.35 g/mol)
(
Fig. S1C).
The synthesis of ROS-responsive mPEG-TK-COOH polymer
2.5.2. Comparative in vitro studies of mPEG-TK-Cy5 on C6, DI TNC1 and
SH SY5Y cells
(Scheme 1B), successfully led to the intended product in very high yield
(78.16%). Furthermore, the derivatization yield obtained, measured by
two different approaches (Ninhydrin assay and 1H NMR), was about
84%. The full characterization of the product was achieved by NMR
analysis, as 1H, C, Homonuclear correlation spectroscopy (COSY),
Heteronuclear single-quantum correlation spectroscopy (HSQC) and
Heteronuclear multiple-bond correlation spectroscopy (HMBC) NMR
(Fig. 1). The intense peaks at 3.66 ppm and 3.44 ppm correspond to
C6, DI TNC1 and SH SY5Y cells were seeded (100.000 cells/mL) on
poly-L-lysine (0.1 mg/mL; Sigma-Aldrich) coated glass coverslips in a
24 well plate and incubated at 37 °C until 80% confluence was reached.
13
Subsequently, the medium was replaced, and 0.9 µg/mL of mPEG-Cy5
or mPEG-TK-Cy5 were added to the corresponding wells and incubated
for 24 h. Before cell fixation, the cells were incubated with CellROX™
Green Reagent (5 µM) for 30 min. Then, the cells were rinsed with 1X
4

N. Oddone, et al.
International Journal of Pharmaceutics 570 (2019) 118655
Fig. 1. 1H NMR characterization of mPEG-TK-COOH in CDCl
. The chemical shifts of protons from the groups on the polymer are identified by letters.
3
protons of –CH
tively. The peak at 3.48 ppm corresponds to protons of –CH
located next to the amide bond formed between mPEG-NH and TK-C.L.
This signal was shifted respective to the signal of protons from the –CH
group on mPEG-NH , which was found to be 3.2 ppm (Fig. S2). This
confirms the formation of covalent bonds between mPEG-NH and TK-
2
O and –CH
3
O groups of the mPEG polymer, respec-
from TK (δ = 1.62 ppm).
2
group
2
3.3. mPEG-TK-Cy5 and mPEG-Cy5 synthesis and characterization
2
2
Both fluorescent conjugates (mPEG-TK-Cy5 and control mPEG- Cy5)
2
(
Scheme 2 and 3, respectively) were synthesized with product yields
C.L. The peak at 1.6 ppm corresponds to the chemical shift of –CH
3
over 90% (93 and 96%, respectively), with mPEG-TK-Cy5 derivatiza-
tion yield around 30%. The Cy5 content of mPEG-TK-Cy5 and mPEG-
Cy5 conjugates was 9.3 ± 3.1 µg/mg and 4.5 ± 1.4 µg/mg, respec-
tively.
group protons of TK-C.L, in accordance with 1H NMR data previously
reported by Li et al. (2016a,b), describing a similar chemical shift of
–
3
CH proton groups of TK in d6-DMSO at 1.52 ppm. The mPEG-TK-
COOH molecular weight calculated by MALDI-TOF was 5275.95 g/mol,
very close to the theoretical molecular weight expected for this polymer
The characterization of either of the conjugates was performed by
H NMR (Fig. S3) and MALDI-TOF. Chemical shifts in the range of 7.15
1
(
5234.3 g/mol).
to 7.7 ppm, which correspond to proton signals from the aromatic
groups of Cy5, have been identified in both spectra: mPEG-Cy5 and
mPEG-TK-Cy5 (Fig. S3A and B). In addition, the characteristic chemical
3.2. mPEG-TK-COOH responsiveness to ROS
3
shift of TK –CH protons can be observed in mPEG-TK-Cy5 (Fig. S3A).
The mean molecular weight calculated through MALDI-TOF for mPEG-
TK-Cy5 and mPEG-Cy5 was 5851.4 g/mol and 5519.2, respectively.
Both molecular weights are close to the theoretical molecular weights
expected for either of the conjugates: 5871.66 g/mol and 5634.7 g/mol.
To assess the ability of mPEG-TK-COOH to respond to ROS, the
polymer was incubated in water containing different concentrations of
(100 µM, 1 mM, 100 mM, and 400 mM) and CuCl (3.2 µM), for
8 h at 37 °C. These reagents produce hydroxyl radicals by reduction of
H
4
H
2
O
2
2
+
+
+++
2
O
2
catalyzed by the conversion of Cu
to Cu
to hydroxyl ra-
dical by a Fenton- like reaction (Goldstein et al., 1993). The purified
polymers obtained at the end of incubation were analyzed using 1
NMR, showing that the chemical shift at 1.58 ppm, corresponding to
protons of TK –CH groups, was decreased with increasing concentra-
tions of H (100 µM- 100 mM) and no longer detected at 400 mM of
(Fig. 2). This confirms that mPEG-TK-COOH responded to hy-
droxyl radicals by TK moiety cleavage under these experimental con-
ditions. Acetone, which is a byproduct originated from Thioketal ROS-
triggered cleavage, can also be detected as a broad peak at a chemical
3.4. Cy5 release from mPEG-TK-Cy5 in ROS-simulated conditions
H
Studies on the selective Cy5 release from mPEG-TK-Cy5 upon ROS
stimuli were performed by incubating mPEG-TK-Cy5 or mPEG-Cy5 in
3
2
O
2
ROS-simulated conditions (DMF/H
and 3.2 µM CuCl ) as well as in control conditions (DMF/H
v/v without H and CuCl ) at physiological temperature (37 °C). At
2
O (5.4:4.6) v/v with 400 mM H
2 2
O
H
2
O
2
2
2
O (5.4:4.6)
2
O
2
2
different times, aliquots of either of the conjugates’ samples were re-
trieved and analyzed by RP-HPLC.
shift of ~2.0 ppm on 100 mM and 400 mM H
result is also in line with the findings of Yue et al. (2016), who estab-
lished TK bond cleavage from TL-CPT (TK-C.L. conjugated to the an-
2
O
2
spectra (Fig. 2). This
Considering mPEG-TK-Cy5, the results after incubation in ROS-si-
mulated conditions (Fig. 3A) clearly evidence a time-dependent de-
crease of the mPEG-TK-Cy5 peak (r.t. at 19 min) with the concomitant
increase of a peak that corresponds to Cy5 released (r.t. at 18 min)
detected by both fluorescence and UV–vis. At 2 and 6 h of incubation,
the percentage of Cy5 release was 20.9% and 62.7%, respectively. At
24 h of incubation, a plateau of the Cy5 release was reached with 98.2%
of Cy5 released. When the same mPEG-TK-Cy5 conjugate was incubated
ticancer drug Camptothecin), due to the disappearance of the –CH
group proton signal (δ = 1.47–1.54 ppm) with detection of acetone
δ = 2.2 ppm). Similarly, Shim and Xia (2013) demonstrated the ROS
responsiveness of TK moiety on poly (amino thioketal) (PATK) poly-
mers by the time-related disappearance of the –CH group proton signal
3
(
3
5

N. Oddone, et al.
International Journal of Pharmaceutics 570 (2019) 118655
Fig. 2. mPEG-TK-COOH 1H NMR spectra after 48 h- incubation in ROS-simulated conditions at 0 µM, 100 µM, 1 mM, 100 mM and 400 mM H
scheme on the upper right displays the Fenton-like reaction which accounts for hydroxyl radical generation, as well as the cleavage site on mPEG-TK-COOH polymer.
2 2
O concentration. The
The red rectangle depicted in the spectra indicates the chemical shift of methyl protons of TK moiety at 1.58 ppm, which decreased upon H
dependent ROS- triggered cleavage.
2 2
O – concentration
in control conditions, the conjugate’s peak remained unmodified (Fig.
S4), clearly confirming that this conjugate undergoes TK cleavage only
in the presence of ROS.
Considering the non-ROS-responsive mPEG-Cy5 conjugate exposed
to simulated ROS conditions, only the peak corresponding to mPEG-Cy5
TNC1 astrocyte cell line as a control that exhibits several characteristic
features of astrocytes, retaining a type 1 astrocyte phenotype (Pellerin
et al., 1997; Philippe et al., 2002). Thus, this astrocyte cell line con-
stitutes an appropriate control of normal astrocytes for comparative
studies. In order to confirm that the ROS levels on C6 cells are higher
(
r.t.18 min) was observed (Fig. 3C). Nevertheless, although Cy5 was not
than on DI TNC1 cells, a dye whose fluorescence is enhanced in ROS
TM
released from this conjugate due to the absence of the ROS-cleavable
linker, conjugate fluorescence intensity decreased with time upon ROS
exposure. To closer investigate this, curves of the percentage of mPEG-
Cy5 intensity related to the initial (t = 0 h) of samples incubated in ROS
simulated and control conditions (solvent mix without Fenton- like
reagents), respectively, were plotted (Fig. 3D). A significant difference
between these samples was found at 24 and 48 h of incubation
producing cells, CellROX
green reagent, was used in these studies.
Fig. 4 depicts the confocal images obtained after the incubation of the
conjugates in either of the cell lines. The CellROX signals on Fig. 4A-D
were quantified using ImageJ, confirming that CellROX fluorescence of
cells incubated with either mPEG-Cy5 or mPEG-TK-Cy5, was sig-
nificantly higher in C6 cells than in DI TNC1 cells (Fig. S5).
Regarding the intracellular distribution of mPEG-Cy5, a cluster
distribution of the Cy5 signal can be observed in both C6 and DI TNC1
cells (Fig. 4A and C), which indicates that mPEG-Cy5 is possibly inside
endo/lysosomal vesicles. In the case of mPEG-TK-Cy5 incubated C6
cells, we could observe Cy5 signal in the cytoplasm, which is no longer
vesicular distributed (Fig. 4B), suggesting that Cy5 was released from
mPEG-TK-Cy5 inside the endo/lysosomes and diffused into the cytosol
of these cells. However, in the case of astrocytes incubated with mPEG-
TK-Cy5, the Cy5 signal is also distributed in clusters, similar to mPEG-
Cy5, meaning that Cy5 was not released from mPEG-TK-Cy5 in these
cells. This demonstrates the selectivity of Cy5 release in an environment
with high levels of ROS, such as found in glioblastoma cell cultures, in
comparison to an environment with a low level of ROS, such as found in
astrocytes cells.
(
Fig. 3D). This result was then confirmed by measuring fluorescence
intensity in a fluorometer at 6 and 48 h of incubation (Fig. 3E). Re-
garding the total fluorescence intensity of the mPEG-TK-Cy5 samples,
which comprises mPEG-TK-Cy5 and released Cy5 fluorescence
(
Fig. 3E), the fluorescence did not vary throughout the experiment. In
addition, the total mPEG-TK-Cy5 fluorescence intensity at 48 h of in-
cubation was significantly higher than the total mPEG-Cy5 fluorescence
intensity. Taken together, these results suggest that, while Cy5 from
mPEG-Cy5 was quenched by ROS, TK linkages on mPEG-TK-Cy5 can be
acting as ROS scavengers, preventing Cy5 from being quenched.
3
.5. Comparative in vitro studies of mPEG-TK-Cy5 in C6 and DI TNC1 cells
To validate the obtained result that Cy5 release is primarily de-
pendent on the presence of ROS, and thus occurs in other tumor cell
lines with similar levels of ROS, we repeated the experiments using the
neuroblastoma SH SY5Y cell line. Cultivated SH SY5Y cells show
comparable levels of ROS to C6 cells (Fig. S6A). As before, cells were
incubated with mPEG-TK-Cy5 or mPEG-Cy5 for 24 h. Intriguingly, ROS
To evaluate the selective release of Cy5 in cancer cells versus
healthy cells, we incubated C6 glioblastoma (tumor) cells and DI TNC1
astrocyte (control) cells for 24 h with mPEG-TK-Cy5 or mPEG-Cy5.
Previous work reported that ROS is essential for rat C6 glioma cells to
grow (Hwang et al., 2018; Kim et al., 2017). As astrocytes can be the
cells of origin for glioblastomas(Zong et al., 2012), we used a rat DI
6

N. Oddone, et al.
International Journal of Pharmaceutics 570 (2019) 118655
Fig. 3. RP-HPLC studies of mPEG-TK-Cy5 and mPEG-Cy5 responsiveness to ROS. All data are shown as mean ± SD. A. Overlapped chromatograms of mPEG-TK-Cy5
injected samples at incubation times: 0 h (blue line), 2 h (pink line) and 6 h (orange line) in ROS simulated medium. B. Cy5 release curve from mPEG-TK-Cy5 with
time with and without ROS. C. Overlapped chromatograms of mPEG-Cy5 injected samples at incubation times: 0 h (blue line), 2 h (pink line) and 6 h (orange line) in
ROS simulated medium. Incubation times with significant differences with respect to the control group are indicated by asterisks (**p ≤ 0.01 and ***p ≤ 0.001).
These statistical results were achieved by applying unpaired t-test. D. mPEG-Cy5 intensity percentage related to initial (t = 0 h) up to 48 h of incubation in ROS
simulated conditions and in medium without ROS (control). Incubation times with significant differences with respect to the control group are indicated by asterisks
(*p ≤ 0.05). These statistical results were achieved by applying unpaired t-test. E. Total fluorescence of mPEG-TK-Cy5 and mPEG-Cy5 incubated in ROS simulated
conditions, measured by fluorometer. The significant difference in conjugates fluorescence intensity is indicated by an asterisk (*p ≤ 0.05) in the graph, and it was
determined by two-way ANOVA with Bonferroni post hoc test.
levels measured by CellROX signal intensity significantly decreased in
mPEG-TK-Cy5 treated cells (p = 0.045, t-test), confirming that the TK
linkage on mPEG-TK-Cy5 can act as ROS scavenger (Fig. S6B). More
importantly, we confirmed a significantly different distribution of Cy5
signals between cells treated with mPEG-TK-Cy5 and mPEG-Cy5 (Fig.
S6C). Similar to C6 cells, several vesicle-like signals were found in
mPEG-Cy5 treated cells. However, the Cy5 signal was more diffuse and
significantly (p = 0.0003, t-test) less clustered in cells treated with
mPEG-TK-Cy5 (Fig. S6C).
3.6. In vitro Cy5 release from mPEG-TK-Cy5 on C6 cells
In order to confirm the results obtained by incubating C6 and SH
SY5Y cells with mPEG-TK-Cy5 for 24 h, where we suggested that in
these cells, Cy5 was released from this conjugate in ROS level depen-
dent fashion, we performed a complementary experiment using C6 cells
7

N. Oddone, et al.
International Journal of Pharmaceutics 570 (2019) 118655
Fig. 4. Confocal images of C6 and DI TNC1 cells
incubated with either mPEG-Cy5 or mPEG-TK-
Cy5 at equivalent Cy5 concentrations of 0.9 µg/
mL for 24 h. A. mPEG-Cy5 incubated C6 cells. B.
mPEG-TK-Cy5 incubated C6 cells. C. mPEG-Cy5
incubated DI TNC1 cells. D. mPEG-TK-Cy5 in-
cubated DI TNC1 cells. The vertical panels show
the different channels (dyes) of every confocal
image. With channels as following: 488 nm
(CellROX), 405 nm (DAPI) and 633 nm (Cy5).
Merge results from the superposition of DAPI
and Cy5 channels. Scale bar = 50 µm.
Fig. 5. Confocal images of representative C6 cells incubated with either mPEG-Cy5 or mPEG-TK-Cy5. A. C6 cells incubated with either mPEG-Cy5 or mPEG-TK-Cy5 at
equivalent Cy5 concentrations of 0.9 and 4.5 µg/mL, respectively for 72 h. a) mPEG-Cy5 incubated at an equivalent Cy5 concentration of 0.9 µg/mL. b) mPEG-Cy5
incubated at an equivalent Cy5 concentration of 4.5 µg/mL. c) mPEG-TK-Cy5 incubated at an equivalent Cy5 concentration of 0.9 µg/mL. d) mPEG-TK-Cy5 at an
equivalent Cy5 concentration of 4.5 µg/mL. The vertical panels show the different channels (dyes) of every confocal image. With channels as following: 405 nm
(DAPI), 488 nm (Rab5), and 633 nm (Cy5). Merge results from the superposition of all the channels. Scale bar: 50 µm. B. C6 cells incubated with mPEG-Cy5 for 24 h.
The vertical panels show the different channels with 405 nm (DAPI), 488 nm (Clathrin, EEA1, or Rab5), and 633 nm (Cy5) as well as all channels merged. Arrows
indicate co-localizing signals of Clathrin, EEA1, or Rab5 with Cy5. Scale bar: 10 µm (Clathrin), 15 μm (EEA1 and Rab5).
8

N. Oddone, et al.
International Journal of Pharmaceutics 570 (2019) 118655
and testing two different equivalent concentrations of Cy5 with an
extended incubation time (72 h). Again, non-ROS-responsive mPEG-
Cy5 was also included in these studies, as control. Glioblastoma cells
were incubated for 72 h with either of the conjugates at equivalent Cy5
concentrations of 0.9 and 4.5 µg/mL, respectively. Confocal microscopy
studies showed that both conjugates were internalized by cells. In
mPEG-Cy5 incubated cells, a cluster-distribution of the red signal,
corresponding to Cy5, can be observed at any of the equivalent Cy5
concentrations assayed (Fig. 5A, a and b). These clusters correspond to
mPEG-Cy5 that was endocytosed by cells.
regions are associated with cytoplasm without Cy5, being in accordance
with the cluster endo/lysosome distribution of red fluorescence ob-
served inside these cells.
4. Discussion
In this study, we were able to synthesize and completely char-
acterize ROS-responsive mPEG-TK-COOH. To validate the usefulness of
this polymer as a polymer platform to obtain ROS-responsive polymer
conjugates for cancer treatment, we conjugated this polymer to the NIR
On the other hand, in the case of mPEG-TK-Cy5 incubated cells
2
functionalized fluorophore, Cy5-NH . The obtained fluorescent con-
(
Fig. 5A, c and d), Cy5 was released from the conjugates as almost all
jugate, mPEG-TK-Cy5, was then employed to conduct ROS-responsive
studies in ROS- simulated conditions as well as in vitro, on either as-
trocyte or glioblastoma cell lines, and neuroblastoma cells. We con-
firmed that the fluorescent dye was released from mPEG-TK-Cy5
polymer conjugates in a time-dependent manner, while in the case of
non-ROS-responsive control mPEG-Cy5, no release and therefore no
cleavage of Cy5 was achieved. Moreover, we were also able to show
another role of TK above the “smart” cleavage in the presence of ROS,
which is a protection against ROS, clearly evidenced by the ability of TK
in decreasing fluorescence quenching of Cy5 induced by ROS. Other
ROS-responsive moieties have been shown to act as ROS scavengers in
the literature, being used as antioxidants alone or in combination with
antioxidant drugs (Lee et al., 2015; Milcovich et al., 2017). Poole et al.,
the fluorescence signal is no longer cluster-distributed. Thus, this con-
firms that Cy5 from mPEG-TK-Cy5 conjugates was first released and
subsequently diffused into the cytoplasm. The differences on Cy5
fluorescence localization inside cells were clearly visible at the higher
Cy5 equivalent concentration (4.5 µg/mL) (Fig. 5A, b and d). While Cy5
from mPEG-Cy5 was clustered inside endocytic vesicles, the Cy5 signal
from mPEG-TK-Cy5 was mainly dispersed through the cytoplasm. These
results further confirm that Cy5, triggered by ROS in glioblastoma cells,
was detached from mPEG-TK-Cy5 with subsequent diffusion into the
cytoplasm. Endocytosed mPEG-Cy5 shows very limited co-localization
with clathrin, the early endosome marker EEA1, and the endosome
associated protein Rab5 (Fig. 5B) hinting at clathrin-dependent en-
docytosis as one way of cellular uptake. However, other mechanisms of
uptake such as pinocytosis seem to have a major contribution to the
internalization of mPEG-Cy5 and mPEG-TK-Cy5.
In mPEG-Cy5 incubated cells, Cy5 was not detached through ROS.
The line profile analysis that was performed for the cytoplasm of se-
lected cells on each confocal image shows that Cy5 fluorescence levels
of mPEG-TK-Cy5 incubated cells do not vary along the cytoplasm at any
of the Cy5 equivalent concentrations (Fig. 6C and D). On the contrary,
the line profile analysis of mPEG-Cy5 incubated cells reveals red
fluorescence peaks alternated by valleys with background red fluores-
cence at both Cy5 equivalent concentrations (Fig. 6A and B). The peaks
can be well correlated to mPEG-Cy5 inside vesicles while the valley
for instance, incubated H
Cur) loaded Poly (propylene sulfide) (PPS) microparticles. They ob-
2 2
served that blank PPS and Cur-PPS were both protective against H O –
2 2
O exposed 3 T3 fibroblasts with Curcumin
(
induced cytotoxicity to a similar extent. This demonstrated that blank
microparticles exhibited scavenger capacity and therefore suggested
that PPS and curcumin might be acting synergistically in preventing
ROS-induced cell death (Poole et al., 2015). Nevertheless, to our
knowledge, this is the first work that shows a protective role against
quenching using a fluorescent dye which is attached directly to a TK
moiety. We hypothesize that in the case of a drug that is similarly at-
tached to a polymer utilizing TK, the drug would be protected from
ROS, preserving its stability.
Fig. 6. Line profile of selected cell cytoplasm on C6 glioblastoma cells from confocal images. Upper panel: confocal images with regions depicting the selected cell of
each of the evaluated conditions. Down panel: Line profile fluorescence plots originated from the selected cells. A-B: mPEG-Cy5 at equivalent Cy5 concentrations of
0.9 and 4.5 µg/mL. C-D: mPEG-TK-Cy5 at equivalent Cy5 concentrations of 0.9 and 4.5 µg/mL. Scale bar: 10 µm.
9

N. Oddone, et al.
International Journal of Pharmaceutics 570 (2019) 118655
In vitro results obtained from glioblastoma, neuroblastoma and as-
Declaration of Competing Interest
trocytes cells were also in line with the results obtained in simulated
ROS conditions. mPEG-TK-Cy5 was selectively released in C6 and SH
SY5Y cells, which produce high levels of ROS, in contrast to the results
obtained by testing mPEG-TK-Cy5 on a “healthy” astrocyte cell line (DI
TNC1 cells) featured by lower ROS levels. Liu et al., who developed an
ROS-responsive amphiphilic dendrimer for the specific delivery of
siRNA in cancer cells, also comparatively measured the ROS levels
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgments
(
through CellROX® orange reagent) in tumor and non-tumor cell lines,
The authors thank Dr. Diego Pinetti and Dr. Cinzia Restani of CIGS,
University of Modena for having performed ESI-MS and NMR analysis,
respectively. Sarah Hudson and Dr. Edel Durack of Biopoint, Bernal
Institute, University of Limerick for the MALDI-TOF/TOF mass spec-
trometry measurements. Natalia Oddone was supported by PhD School
in Clinical and Experimental Medicine, University of Modena and
Reggio Emilia. The project is also granted by FAR UNIMORE and
MAECI grant 2019 (Nanomedicine for BBB crossing in CNS Oncologic
Pathologies).
confirming higher ROS levels in tumor cells (MCF-7 and PC-3) in
comparison to non-tumor cells (HEK and CHO). In addition, the gene
silencing of ROS-responsive siRNA amphiphilic dendrimer nanocom-
plex was considerably observed only in the High-ROS cells, MCF-7 and
PC-3 (Liu et al., 2016).
On the other hand, apart from the selective release of Cy5, we found
that the CellROX signal was reduced on mPEG-TK-Cy5 treated SH SY5Y
cells, confirming the ROS-scavenging property of TK moiety as in the
case of other ROS-responsive moieties (Lee et al., 2015; Milcovich et al.,
2017; Poole et al., 2015).
Appendix A. Supplementary data
Further studies performed with mPEG-TK-Cy5 on C6 cells showed
that Cy5 was released from the conjugate, while on the contrary, mPEG-
Cy5 incubated cells did not show to be able of releasing Cy5. As men-
tioned before, Cy5 fluorescence in the case of mPEG-Cy5 incubated cells
came from endocytic vesicles while in the case of mPEG-TK-Cy5 in-
cubated cells, this fluorescence came mainly from the cytoplasm. In
addition, Cy5 fluorescence intensity of mPEG-Cy5 incubated on C6 cells
was found to be lower in comparison to mPEG-TK-Cy5 incubated cells
at any of the equivalent Cy5 concentrations assayed. As the structure of
both polymer conjugates is almost the same, except for the presence or
absence of the TK moiety, these differences in Cy5 fluorescence cannot
be attributed to a difference in cell uptake. However, the difference
might be attributed to self-quenching of Cy5 molecules from mPEG-Cy5
inside endocytosed vesicles. This phenomenon of self-quenching has
been recently described to occur in fluorophores which are attached to
molecules when they are densely packed into subcellular compartments
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.ijpharm.2019.118655.
References
Chakraborty, C., Sharma, A.R., Sharma, G., Sarkar, B.K., Lee, S.-S., 2018. The novel
strategies for next-generation cancer treatment: miRNA combined with chemother-
apeutic agents for the treatment of cancer. Oncotarget 9, 10164–10174. https://doi.
org/10.18632/oncotarget.24309.
Chang, M., Zhang, F., Wei, T., Zuo, T., Guan, Y., Lin, G., Shao, W., 2016. Smart linkers in
polymer–drug conjugates for tumor-targeted delivery. J. Drug Target. 24, 475–491.
https://doi.org/10.3109/1061186X.2015.1108324.
Dragojevic, S., Ryu, J.S., Raucher, D., 2015. Polymer-based prodrugs: improving tumor
targeting and the solubility of small molecule drugs in cancer therapy. Molecules 20,
21750–21769. https://doi.org/10.3390/molecules201219804.
Goldstein, S., Meyerstein, D., Czapski, G., 1993. The Fenton reagents. Free Radical Biol.
Med. 15, 435–445. https://doi.org/10.1016/0891-5849(93)90043-T.
Gupta, M.K., Meyer, T.A., Nelson, C.E., Duvall, C.L., 2012. Poly(PS-b-DMA) micelles for
reactive oxygen species triggered drug release. J. Control. Release 162, 591–598.
https://doi.org/10.1016/j.jconrel.2012.07.042.
(
Swiecicki et al., 2016). Wang et al., who develop ditelluride-containing
poly(ether-urethane) NPs for the encapsulation of DOX, reported a si-
milar result. They observed that DOX was quenched inside the en-
docytosed NPs and most importantly, they correlated the DOX detect-
able fluorescence in cells to the quantity of DOX released from the NPs
Hagen, H., Marzenell, P., Jentzsch, E., Wenz, F., Veldwijk, M.R., Mokhir, A., 2012.
Aminoferrocene-based prodrugs activated by reactive oxygen species. J. Med. Chem.
55, 924–934. https://doi.org/10.1021/jm2014937.
Hu, P., Tirelli, N., 2012. Scavenging ROS: superoxide dismutase/catalase mimetics by the
use of an oxidation-sensitive nanocarrier/enzyme conjugate. Bioconjugate Chem. 23,
(
Wang et al., 2016).
4
38–449. https://doi.org/10.1021/bc200449k.
Hwang, E., Sim, S., Park, S.H., Song, K.D., Lee, H.-K., Heo, T.-H., Jun, H.S., Kim, S.-J.,
018. Anti-proliferative effect of Zea mays L. cob extract on rat C6 glioma cells
2
5. Conclusions
through regulation of glycolysis, mitochondrial ROS, and apoptosis. Biomed.
Pharmacother. 98, 726–732. https://doi.org/10.1016/j.biopha.2017.12.115.
Kim, E.-J., Bhuniya, S., Lee, H., Kim, H.M., Cheong, C., Maiti, S., Hong, K.S., Kim, J.S.,
To confirm the ability of drugs to be released upon a ROS stimulus,
polymeric conjugates were successfully synthesized and characterized
based on the use of TK-technology. Employing the fluorescent dye Cy5-
2
NH , we validated the ROS responsive release of Cy5 from an mPEG-
2
014. An activatable prodrug for the treatment of metastatic tumors. J. Am. Chem.
Soc. 136, 13888–13894. https://doi.org/10.1021/ja5077684.
Kim, S.-J., Hwang, E., Yi, S.S., Song, K.D., Lee, H.-K., Heo, T.-H., Park, S.-K., Jung, Y.J.,
Jun, H.S., 2017. Sea buckthorn leaf extract inhibits glioma cell growth by reducing
reactive oxygen species and promoting apoptosis. Appl. Biochem. Biotechnol. 182,
TK-Cy5 conjugate in ROS-simulated conditions, also describing a pro-
tective role of TK, which acts as a “suicide” molecule in the case of ROS
undesired actions (i.e., quenching of fluorescence). Motivated by the
possible use of this ROS responsive polymer for application in cancer
therapy, we assayed mPEG-TK-Cy5 using glioblastoma and neuro-
blastoma cells and demonstrated that Cy5 can be released upon ROS
physiological concentrations in cancer cells. In addition, mPEG-TK-Cy5
in astrocyte cells, as control, did not show to release Cy5. This confirms
the selectivity of the Cy5 release in a “pathological” condition (tumor
cells) in contrast to a “healthy” condition (astrocyte cells). We conclude
that, due to their ability to selectively deliver drugs upon ROS stimulus
while maintaining drug stability, the application of TK-technology in
the design of conjugates constitutes a promising approach for future
therapeutics against cancer. Furthermore, taking into consideration
that high ROS levels are present in other pathologies than cancer as
well, e.g., immunological, cardiovascular and neurodegenerative dis-
eases, the TK-technology can also be applied in the design of conjugates
against these diseases.
1
663–1674. https://doi.org/10.1007/s12010-017-2425-4.
Lee, D., Park, S., Bae, S., Jeong, D., Park, M., Kang, C., Yoo, W., Samad, M.A., Ke, Q.,
Khang, G., Kang, P.M., 2015. Hydrogen peroxide-activatable antioxidant prodrug as a
targeted therapeutic agent for ischemia–reperfusion injury. Sci. Rep. 5, 16592.
https://doi.org/10.1038/srep16592.
Lee, Y., Kataoka, K., 2009. Biosignal-sensitive polyion complex micelles for the delivery of
biopharmaceuticals. Soft Matter 5, 3810–3817. https://doi.org/10.1039/B909934D.
Li, F., Mahato, R.I., 2017. Bioconjugate therapeutics: current progress and future per-
spective. Mol. Pharmaceut. 14, 1321–1324. https://doi.org/10.1021/acs.
molpharmaceut.7b00263.
Li, Q., Wen, Y., Wen, J., Zhang, Y.-P., Xu, X.-D., Victorious, A., Zavitz, R., Xu, X., 2016a. A
new biosafe reactive oxygen species (ROS)-responsive nanoplatform for drug de-
livery. RSC Adv. 6, 38984–38989. https://doi.org/10.1039/C5RA25913D.
Li, Q., Wen, Y., You, X., Zhang, F., Shah, V., Chen, X., Tong, D., Wei, X., Yin, L., Wu, J.,
Xu, X., 2016b. Development of a reactive oxygen species (ROS)-responsive nano-
platform for targeted oral cancer therapy. J. Mater. Chem. B 4, 4675–4682. https://
doi.org/10.1039/C6TB01016D.
Liu, X., Wang, Y., Chen, C., Tintaru, A., Cao, Y., Liu, J., Ziarelli, F., Tang, J., Guo, H.,
Rosas, R., Giorgio, S., Charles, L., Rocchi, P., Peng, L., 2016. A fluorinated bola-
amphiphilic dendrimer for on-demand delivery of siRNA, via specific response to
reactive oxygen species. Adv. Funct. Mater. 26, 8594–8603. https://doi.org/10.
1
002/adfm.201604192.
10

N. Oddone, et al.
International Journal of Pharmaceutics 570 (2019) 118655
Martin, J.R., Gupta, M.K., Page, J.M., Yu, F., Davidson, J.M., Guelcher, S.A., Duvall, C.L.,
containing arylboronic ester and self-immolative motif: synthesis and degradation
study. Macromolecules 46, 8416–8425. https://doi.org/10.1021/ma401656t.
Swiecicki, J.-M., Thiebaut, F., Di Pisa, M., Gourdin -Bertin, S., Tailhades, J., Mansuy, C.,
Burlina, F., Chwetzoff, S., Trugnan, G., Chassaing, G., Lavielle, S., 2016. How to
unveil self-quenched fluorophores and subsequently map the subcellular distribution
of exogenous peptides. Sci. Rep. 6, 20237. https://doi.org/10.1038/srep20237.
Tiwari, G., Tiwari, R., Sriwastawa, B., Bhati, L., Pandey, S., Pandey, P., Bannerjee, S.K.,
2012. Drug delivery systems: an updated review. Int J Pharm Investig 2, 2–11.
https://doi.org/10.4103/2230-973X.96920.
Turecek, P.L., Bossard, M.J., Schoetens, F., Ivens, I.A., 2016. PEGylation of biopharma-
ceuticals: a review of chemistry and nonclinical safety information of approved drugs.
JPharmSci. 105, 460–475. https://doi.org/10.1016/j.xphs.2015.11.015.
Wang, X., Michaelis, E.K., 2010. Selective neuronal vulnerability to oxidative stress in the
brain. Front Aging Neurosci 2. https://doi.org/10.3389/fnagi.2010.00012.
Wang, Yangyun, Zhu, L., Wang, Yong, Li, L., Lu, Y., Shen, L., Zhang, L.W., 2016.
Ultrasensitive GSH-responsive ditelluride-containing poly(ether-urethane) nano-
particles for controlled drug release. ACS Appl. Mater. Interfaces 8, 35106–35113.
https://doi.org/10.1021/acsami.6b14639.
2
014. A porous tissue engineering scaffold selectively degraded by cell-generated
reactive oxygen species. Biomaterials 35, 3766–3776. https://doi.org/10.1016/j.
biomaterials.2014.01.026.
McEnery, M.A.P., Lu, S., Gupta, M.K., Zienkiewicz, K.J., Wenke, J.C., Kalpakci, K.N.,
Shimko, D.A., Duvall, C.L., Guelcher, S.A., 2016. Oxidatively degradable poly(thio-
ketal urethane)/ceramic composite bone cements with bone-like strength. RSC Adv.
6
, 109414–109424. https://doi.org/10.1039/C6RA24642G.
Milcovich, G., Contessotto, P., Marsico, G., Ismail, S., Pandit, A., 2017. Synthetic/ECM-
inspired hybrid platform for hollow microcarriers with ROS-triggered nanoporation
hallmarks. Sci. Rep. 7, 13138. https://doi.org/10.1038/s41598-017-13744-y.
Pellerin, L., Stolz, M., Sorg, O., Martin, J.-L., Deschepper, C.F., Magistretti, P.J., 1997.
Regulation of energy metabolism by neurotransmitters in astrocytes in primary cul-
ture and in an immortalized cell line. Glia 21, 74–83. https://doi.org/10.1002/(SICI)
1
098-1136(199709)21:1<74::AID-GLIA8>3.0.CO;2-1.
Philippe, J.-M., Dubois, J.-M., Rouzaire-Dubois, B., Cartron, P.-F., Vallette, F., Morel, N.,
2
3
002. Functional expression of V-ATPases in the plasma membrane of glial cells. Glia
7, 365–373. https://doi.org/10.1002/glia.10041.
Poole, K.M., Nelson, C.E., Joshi, R.V., Martin, J.R., Gupta, M.K., Haws, S.C., Kavanaugh,
T.E., Skala, M.C., Duvall, C.L., 2015. ROS-responsive microspheres for on demand
antioxidant therapy in a model of diabetic peripheral arterial disease. Biomaterials
Xu, L., Yang, Y., Zhao, M., Gao, W., Zhang, H., Li, S., He, B., Pu, Y., 2018. A reactive
oxygen species–responsive prodrug micelle with efficient cellular uptake and ex-
cellent bioavailability. J. Mater. Chem. B 6, 1076–1084. https://doi.org/10.1039/
C7TB02479G.
Yue, C., Zhang, C., Alfranca, G., Yang, Yao, Jiang, X., Yang, Yuming, Pan, F., de la Fuente,
J.M., Cui, D., 2016. Near-infrared light triggered ROS-activated theranostic platform
based on Ce6-CPT-UCNPs for simultaneous fluorescence imaging and chemo-photo-
dynamic combined therapy. Theranostics 6, 456–469. https://doi.org/10.7150/thno.
14101.
4
1, 166–175. https://doi.org/10.1016/j.biomaterials.2014.11.016.
Salim, S., 2017. Oxidative stress and the central nervous system. J. Pharmacol. Exp. Ther.
60, 201–205. https://doi.org/10.1124/jpet.116.237503.
3
Saravanakumar, G., Kim, J., Kim, W.J., 2017. Reactive-oxygen-species-responsive drug
delivery systems: promises and challenges. Adv. Sci. (Weinh) 4, 1600124. https://
doi.org/10.1002/advs.201600124.
Sharma, V., Joseph, C., Ghosh, S., Agarwal, A., Mishra, M.K., Sen, E., 2007. Kaempferol
Zhang, Yu., Guo, Q., An, S., Lu, Y., Li, J., He, X., Liu, L., Zhang, Yujie, Sun, T., Jiang, C.,
2017. ROS-switchable polymeric nanoplatform with stimuli-responsive release for
active targeted drug delivery to breast cancer. ACS Appl. Mater. Interfaces 9,
12227–12240. https://doi.org/10.1021/acsami.6b16815.
induces apoptosis in glioblastoma cells through oxidative stress. Mol. Cancer Ther. 6,
2
544–2553. https://doi.org/10.1158/1535-7163.MCT-06-0788.
Shim, M.S., Xia, Y., 2013. A reactive oxygen species (ROS)-responsive polymer for safe,
efficient, and targeted gene delivery in cancer cells. Angew. Chem. Int. Ed. 52,
Zong, H., Verhaak, R.G., Canoll, P., 2012. The cellular origin for malignant glioma and
prospects for clinical advancements. Expert Rev Mol Diagn 12, 383–394. https://doi.
org/10.1586/erm.12.30.
6
926–6929. https://doi.org/10.1002/anie.201209633.
Song, C.-C., Ji, R., Du, F.-S., Li, Z.-C., 2013. Oxidation-responsive poly(amino ester)s
11