Micelles via self-assembly of amphiphilic beta-cyclodextrin block copolymers as drug carrier for cancer therapy

Article abstract of DOI:10.1016/j.colsurfb.2019.110425

We developed intelligent, star-shaped amphiphilic β-cyclodextrin (β-CD) co-polymer nanocarriers to circumvent the poor drug loading and water-solubility of β-CD. The secondary hydroxyl groups of β-CD were methylated to improve solubility, and the primary hydroxyl groups were conjugated with mPEG-b-PCL-SH through disulfide linkage to amplify the hydrophobic cavity and enhance the stability of the nanocarrier. A series of amphiphilic β-CD block copolymers (CCPPs) differing in molecular weights were synthesized that could self-assemble into core-shell nanospheres measuring 50–70 nm in water. The different CCPP carriers were screened for their drug loading, encapsulation and release efficiencies, and CCPP-2 showed the highest drug loading capacity of 31.9% by weight. These nanocarriers accumulated at the tumor site through the EPR effect and released the drug in a controlled manner in the reductive tumor microenvironment, with negligible premature leakage and side effects. Therefore, CCPP-2 shows significant potential as a smart and efficient nanovehicle for anticancer drug delivery.

Full text of DOI:10.1016/j.colsurfb.2019.110425

ColloidsandSurfacesB:Biointerfaces183(2019)110425
Contents lists available at ScienceDirect
Colloids and Surfaces B: Biointerfaces
journal homepage: www.elsevier.com/locate/colsurfb
Micelles via self-assembly of amphiphilic beta-cyclodextrin block
copolymers as drug carrier for cancer therapy
Xiufang Li, Hui Liu, Jianbing Li, Zhiwei Deng, Lingjun Li, Junjun Liu, Jing Yuan, Peiru Gao,
Yanjing Yang⁎, Shian Zhong⁎
School of Chemistry and Chemical Engineering, Central South University, Changsha 410083, China
A R T I C L E I N F O
A B S T R A C T
Keywords:
We developed intelligent, star-shaped amphiphilic β-cyclodextrin (β-CD) co-polymer nanocarriers to circumvent
the poor drug loading and water-solubility of β-CD. The secondary hydroxyl groups of β-CD were methylated to
improve solubility, and the primary hydroxyl groups were conjugated with mPEG-b-PCL-SH through disulfide
linkage to amplify the hydrophobic cavity and enhance the stability of the nanocarrier. A series of amphiphilic β-
CD block copolymers (CCPPs) differing in molecular weights were synthesized that could self-assemble into core-
shell nanospheres measuring 50–70 nm in water. The different CCPP carriers were screened for their drug
loading, encapsulation and release efficiencies, and CCPP-2 showed the highest drug loading capacity of 31.9%
by weight. These nanocarriers accumulated at the tumor site through the EPR effect and released the drug in a
controlled manner in the reductive tumor microenvironment, with negligible premature leakage and side effects.
Therefore, CCPP-2 shows significant potential as a smart and efficient nanovehicle for anticancer drug delivery.
Amphiphilic β-CD
Block copolymer
Drug carrier
Reductive stimulus responsive
Drug loading capacity
1. Introduction
has been extensively utilized to construct drug carriers [17], and DOX
loaded into β-CD exhibits a stronger anti-proliferative effect compared
Cancer remains a major cause of morbidity and mortality worldwide
despite advances in therapeutic modalities [1]. Most malignancies still
rely on conventional chemotherapy, which however is limited by the
poor water solubility, non-selective in vivo distribution, severe systemic
side effects [2], and premature leakage [3,4] of the drugs. Therefore, it
is necessary to design intelligent drug delivery system (IDDS) to cir-
cumvent the above limitations and improve the treatment outcome in
cancer patients. Amphiphilic block polymeric micelles have gained
considerable attention as nano-scale drug-delivery systems for myriad
clinical applications due to their ability to enhance the uptake of poorly
water-soluble drugs, achieve sustained drug release, and specifically
target tumor tissues [5–9]. However, studies have reported low dox-
orubicin (DOX) loading capacities (< 10%) of simple amphiphilic block
copolymers like PEG-b-PCL, PEG-b-PLA and PEG-b-PCL-b-PEG [10–14].
The loading capacity was also not significantly improved in the new
generation star-shaped amphiphilic copolymers containing poly-
amidoamine dendrimer cores and PEG-b-PCL arms [15]. Beta-cyclo-
dextrins (β-CD) are native cyclic oligomers composed of seven gluco-
to free DOX due to higher cellular uptake [18,19]. However, β-CD has
limited clinical utility due to poor water solubility and sub-optimal
drug-loading capacity [20]. The former can be attributed to the for-
mation of strong intermolecular hydrogen bonds between β-CD units
that reduces hydrogen bonding with water molecules. Therefore, sev-
eral chemically modified β-CDs such as methylated β-CDs [21], hy-
droxyalkylated β-CDs [22] and ionic β-CDs [23], have been synthesized
that have improved water solubility on account of the disrupted in-
termolecular hydrogen bonds between the secondary hydroxyl groups
of parent β-CDs. However, enhanced water solubility of these copoly-
mers does not necessarily lead to higher drug-loading capacity, which
can only be substantially augmented by increasing the volume of the
hydrophobic cavity. Therefore, micelles self-assembled from amphi-
philic β-CD-based polymers have been designed, which offer several
advantages compared to the parent β-CDs like larger core-shell struc-
ture, tunable physicochemical properties, prolonged blood circulation
[24,25], high water solubility and the ability to solubilize hydrophobic
drugs [26], high drug loading capacity and high encapsulation effi-
ciency. Qiu et al. [27] designed star-like PLA-b-PEG copolymers-based
nanosystems using the β-CD core, which exhibited enhanced stability
and increased drug loading capacity. Xiao et al. [28] also used a β-CD-
pyranose units linked by α-1, 4-glycosidic bonds, and have
a
hydrophilic exterior and hydrophobic interior [16] giving them the
appearance of hollow, truncated cones. The hydrophobic cavity of β-CD
⁎ Corresponding authors.
E-mail addresses: yangyanjing@csu.edu.cn (Y. Yang), zhongshian@aliyun.com (S. Zhong).
https://doi.org/10.1016/j.colsurfb.2019.110425
Received 5 May 2019; Received in revised form 16 July 2019; Accepted 3 August 2019
Availableonline08August2019
0927-7765/©2019ElsevierB.V.Allrightsreserved.

X. Li, et al.
ColloidsandSurfacesB:Biointerfaces183(2019)110425
based initiator to construct cationic star polymers with 21ACSPs arms
that demonstrated excellent biocompatibility. Therefore, the drug
loading capacity of β-CD can be significantly enhanced by combining it
with a block copolymer. Stimuli-responsive anti-cancer drug delivery
systems that can specifically release the cargo at the tumor in response
to the low pH [29,30], redox imbalance [31,32] and temperature
[33,34] of the tumor microenvironment have gained considerable at-
tention in recent years. These unique stimuli can be utilized to trigger
the release of chemotherapeutic agents into the tumor cells while
avoiding drug leakage into healthy cells. Therefore, the β-CD-block
copolymer drug carriers also need to possess stimulus-responsive bonds
in order to achieve targeted enrichment and release of the che-
motherapeutic agents.
We developed a novel amphiphilic star-like reductive stimulus-re-
sponsive drug carrier using β-CD core and polyethylene glycol (PEG)/
polycaprolactone (PCL) arms. The primary hydroxyl groups of β-CD
were conjugated to the PEG-PCL blocks and hydrosulphonyls, and the
secondary hydroxyl groups were chemically modified with iodo-
methane The methyl groups and PCL acted as the hydrophobic core and
PEG served as the hydrophilic shell, and the cavity resulting from the
self-assembly of the amphiphilic β-CD block copolymers (CCPP) into
core-shell nanospheres provided sufficient space for DOX. CCPP-2
showed high drug loading capacity of 31.9%, in addition to rapid but
targeted drug release under reductive stimulus. The amphiphilic drug
carrier also showed enhanced anti-tumor effect and low systemic toxi-
city in a tumor-bearing mouse model. The overall procedure for the
synthesis of amphiphilic drug carrier combined with reduction-trig-
gered drug release was designed for drug delivery to tumor tissues
(Scheme 1).
stannous octoate, 3,3′-thiodipropionic acid (DTPA), N,N'-dicyclohex-
ylcarbodiimide (DCC), tertbutyldimethylsilyl chloride (TBDMS-Cl),
bromine (Br2) and 4-dimethylaminopyridine (DMAP) were obtained
from Energy Chemical Technology Co., Ltd. (Shanghai, China). Sodium
methoxide was purchased from Maya Reagent Company (Zhejiang,
China), methoxypolyethylene glycols (MPEG; Mw: 2000) was purchased
from Sinopharm Chemical Reagent Co. Ltd, 4′, 6-Diamidino-2-pheny-
lindole (DAPI) from Molecular Probes (USA), and 1, 1′-Dioctadecyl-3, 3,
3′, 3′-tetramethyl indotricarbocyanine iodide (Dir) from Biotium
Company (USA). DL-Dithiothreitol (DTT), sodium hydride (NaH), io-
domethane (CH3I), coumarin-6 (C6) and DOX were obtained from
Aladdin Chemical Company (Shanghai, China). Fetal bovine serum
(FBS), Dulbecco's modified eagle medium (DMEM) and RPMI 1640
medium were purchased from Biological Industries (Israel), and peni-
cillin/streptomycin and 0.25% Trypsin-EDTA from Invitrogen (USA).
Cell Counting Kit-8 (CCK-8) was procured from Shanghai Seven Seas
Biotechnology Co. Ltd. All reagents and solvents were of analytical
purity. The water used for all experiments was doubly distilled.
2.2. Synthesis and characterization of CCPP
The three-step synthesis of 2,3−CH3O-β-CD-(s-s-PCL-b-mPEG)7 or
CCPP is outlined in Fig. 1. A series of thiol-terminated mPEG-b-PCL
copolymers incorporating PCL blocks of different molecular weights
(MW) were first synthesized by ring-opening polymerization of ε-ca-
prolactone with mono-hydroxyl-terminated mPEG initiator and stan-
nous octoate catalyst, followed by sequential esterification and 1,4-di-
thio-threitol (DTT) reduction (Fig. S1). The β-cyclodextrin derivative
per-6-SH-2,3−OCH3-β-CD was then synthesized by methylating the
secondary hydroxyls and mercaptoylating the primary hydroxyls
through four-step affinity substitution reactions (Fig. 1). Finally, the
thiol-terminated mPEG-b-PCL were conjugated to the -SH groups of per-
6-SH-2,3−OCH3-β-CD by H2O2-mediated oxidation (Fig. 1) to obtain
CCPP. As a control, the non-methylated β-CD-(s-s-PCL-b-mPEG)7 or CPP
was also synthesized using the same methods. The respective co-
2. Materials and methods
2.1. Reagents
β-CD, triphenylphosphine (PPh3), thiourea, I2, ε-Caprolactone,
Scheme 1. Schematic illustration of (A) self-assembly of the amphipathic β-CD block copolymer micelle; (B) the mechanism of anticancer drug vehicles in tumor
microenvironment.
2

X. Li, et al.
ColloidsandSurfacesB:Biointerfaces183(2019)110425
micelles were calculated according to the following formula:
weight of drug in micelles
weight of drug loaded micelles
DLC(%) =
EE(%) =
× 100%
(1)
(2)
weight of drug in micelles
weight of drug in feed
× 100%
2.4. Measurement of critical micelle concentration
The critical micelle concentration (CMC) of the polymers were es-
timated using pyrene as the probe [35–37]. Pyrene solution in methanol
was transferred through a series of volumetric flasks to evaporate the
methanol and achieve a final concentration is 6 × 10−7 M. CCPP or
CPP micelle solutions of different concentrations were added to the
flasks, and the mixture was kept at 25 °C overnight. The fluorescence
spectra were measured using the FLS920 spectrometer at the excitation
wavelength of 335 nm, and emission wavelengths of 372 and 383 nm.
The slits were set at 10 (excitation) and 2.5 nm (emission). The CMC
was estimated as the cross-point when extrapolating the intensity ratio
I372/I383 at low and high concentration regions.
2.5. Size and zeta potential analysis of CCPP nanoparticles
The average size and the size distribution of the micelles were de-
termined by dynamic light scattering (DLS). Briefly, CCPP or CPP mi-
celles of different Mw were suspended in distilled water at 25 °C to
prepare respective solutions of 0.5 mg/mL, and sonicated for 30 s. The
micellar solutions were filtered through a 0.45 μm mesh, and measured
using Zetasizer Nano-ZS (Malvern Instruments) equipped with a 633 nm
He-Ne laser using back-scattering detection. The zeta potential of the
micelles were tested in the same instrument after diluting the particle
suspension in deionized water.
Fig. 1. Detailed synthetic routes of a series of CCPP and CPP copolymers.
polymer solutions were dialysed for 48 h, and freeze-dried into the
powder form. Further details are included in the Supporting Informa-
tion.
Infrared spectra of the samples were obtained using a Fourier
transform infrared spectrometer (Nicolet, USA), and 1H NMR spectra
were recorded with a 400 or 500 MHz spectrometer (Bruker, Germany).
The Mw and polydispersity of the copolymers were determined using
Perkin-Elmer Series-200 gel permeation chromatograph (GPC)
equipped with PLgel 5 μm Mixed 2D columns. THF was used as the
eluent at a flow rate of 1 mL/min and a series of narrow polystyrene
standards were used for calibratiing the columns. High-resolution
Transmission electron microscopy (TEM) samples were prepared by
placing a drop of the sample solution in H2O (concentration: 1 mg/mL)
onto ultrathin carbon film. TEM was performed by using a JEOL 2100 F
microscope (JEOL, Japan) with an accelerating voltage of 200 kV.
2.6. In vitro DOX release
DOX-loaded CCPP or CPP (10 mg) were suspended in 2 mL phos-
phate buffer saline (PBS, pH = 7.4) with or without 10 mM DTT, and
transferred to a dialysis bag (3500 MWCO). The latter was placed in
8 mL PBS (with/without 10 mM DTT), and dialysis was carried at
37.5 °C. One millilitre aliquots were drawn at certain time intervals,
and replaced with the same volume of fresh buffer. The content of DOX
in the eluents were measured in terms of absorbance (Abs) at 490 nm,
and calculated as: y (Abs) = 0.0143x + 0.019 (x: μg/mL). The drug
release rate from the copolymer was calculated with the following
formula:
2.3. Preparation of DOX-loaded micelles
mt
m
Release rate (%) =
×100%
The blank micelles were first obtained by dissolving 50 mg of the
different CCPPs or CPPs in 5 mL DMSO, and each was added dropwise
into 25 mL deionized water under constant magnetic stirring. After
stirring for 24 h at the ambient temperature, the solutions were dia-
lyzed for 24 h to remove DMSO. The DOX-loaded micelles were ob-
tained by dissolving 10 mg of the DOX·HCl in 1 mL DMSO, and 5.32 μL
trimethylamine was added to neutralize HCl under constant magnetic
stirring. After stirring for 24 h at the ambient temperature, the color of
the solution slowly turned deep purple from the initial red. 10 mg of the
different CCPPs or CPPs was added and the resulting mixtures were
allowed to stir for 12 h, and each was added dropwise into 5 mL deio-
nized water under constant magnetic stirring. After stirring for 24 h at
the ambient temperature, the solutions were dialyzed for 48 h to re-
move DMSO. The blank and DOX-loaded micelles were lyophilized, and
1 mg of each was dissolved in 5 mL methanol and sonicated for 30 min.
The absorption of the micelles at 490 nm (UV–vis spectra) was mea-
sured using a UV-2450 spectrophotometer (Shimadzu, Japan), and the
weight was calculated from the calibration curve. Finally, the drug
loading capacity (DLC) and encapsulation efficiency (EE) of the
where mt (g) was the amount of DOX released at a given time, and m (g)
was total amount of drug loaded.
2.7. Cell cytotoxicity assay
The cytotoxicity of CCPP-2, CPP-2, DOX@CCPP-2 and DOX@CPP-2
micelles was evaluated in the ovarian carcinoma (Skov 3) cells and
human embryonic kidney cells (HEK 293 T) (Xiangya Hospital of
Central South University, Changsha, China) with the CCK-8 assay. The
Skov 3 and HEK 293 T cells were respectively cultured in RPMI 1640
and DMEM, both supplemented with 10% FBS and 1% penicillin/
streptomycin, at 37 °C under 5% CO2. For the cytotoxicity assay, the
cells were seeded into 96 well plates at the density of 5 × 103 cells/well
and cultured for 24 h. After washing the cells thrice with PBS, 100 μL
fresh medium containing different concentrations of blank CCPP-2 or
CPP-2 (0, 50, 100, 200, 300, 500 and 1000 μg/mL) or the DOX (0, 1, 2,
3, 5, 7, 9, 12, 15, 18 μg/mL)-loaded micelles was added per well. After
3

X. Li, et al.
ColloidsandSurfacesB:Biointerfaces183(2019)110425
culturing for another 24 h, the used culture medium was replaced with
new medium. 10 μL CCK-8 was added to each well and the cells were
incubated further for 2 h. The optical density (OD) of the cell suspen-
sions at 450 nm was measured using a microplate reader (Bio-Rad,
USA). The cell viability was calculated as follows:
prepare cell suspension, and acquired by flow cytometry.
2.11. In vivo tumor treatment
All animal experiments were strictly implemented under the
guidelines approved by the Medical Laboratory Animal Management
Committee of the Xiangya School of Medicine. SPF female BALB/c nude
ODsample − ODblank
Cell viability (%) =
×100%
ODcontrol − ODblank
mice (18
3 g, 3–4 weeks old) were purchased from Hunan SJA
where sample, control and blank refers to the treated cells, the un-
Laboratory Animal Co. Ltd (animal quality certificate No.
43,004,700,041,110 and approval No. SCXK 2016-0002) and housed at
the Department of Laboratory Animals, Central South University. Each
mouse was injected subcutaneously with 5 × 106 cells/mL Skov 3 cells
in 200 μL into their flanks, and randomly divided into the untreated,
DOX, DOX@CCPP-2 and DOX@CPP-2 groups (n = 6). Once the tumor
grew to approximately 100 mm3, the mice were intravenously injected
with the suitable vehicle at the DOX or DOX-equivalent dose of 3 mg/kg
every two days. The untreated control mice received 0.5 mL/kg of
saline. The tumor dimensions and the body weight of mice were mea-
sured prior to every injection. The tumor volume (mm3) was calculated
as ab2/2, where a (mm) and b (mm) were respectively the longest and
shortest diameters.
treated cells and culture medium. Data were presented as means
(n = 6).
SD
2.8. Hemolysis assay
The biocompatibility of the different micelles was also tested in
terms of their hemolytic activity. Briefly, 5 mL sterile defibrinated
sheep blood (leaf Biology, Shanghai) was diluted in 50 mL PBS and
centrifuged at 500 g for 10 min. The RBCs were separated and washed
several times with PBS till the supernatant turned colorless. After the
last wash, a 10% (v/v) RBC suspension was prepared in PBS, and 100 μL
aliquots were mixed with 100 μL of varying concentrations (0.05, 0.1,
0.2, 0.3, 0.5, and 1 mg/mL in PBS) of the micelles. PBS and 1% Triton
X-100 were included as the negative and positive controls respectively.
The mixtures were gently whirled and incubated for 2 h at 37 °C under
5% CO2, and centrifuged at 3000 rpm for 5 min. The supernatants was
carefully removed, and 100 μL aliquots were transferred per well of a
96-well-plate. The hemoglobin released from the RBCs was measured at
540 nm using a microplate reader (Bio-Rad Instruments, USA), and the
hemolysis ratio (HR) of the RBCs (%) was calculated as follows:
2.12. In vivo distribution of CCPP-2 and CPP-2 micelles
Dir-loaded micelles were prepared via a similar protocol as de-
scribed in section 2.3, and used to track the systemic distribution of the
micelles. Two tumor-bearing mice were respectively injected with
100 μL Dir@CCPP-2 and Dir@CPP-2 (Dir dose of 10 μg/mL) via their
tail veins. The mice were imaged at 4, 12, 24 and 36 h after injection
with the IVIS Lumina II multispectral imaging system (Caliper, USA),
and sacrificed after the final imaging. Their vital organs (heart, liver,
spleen, lung, kidney) and tumors were extracted and imaged ex vivo to
evaluate in situ micellar accumulation.
To track the distribution of DOX in the tumor tissues and organs, the
differentially-treated mice (see section 2.11) were sacrificed four weeks
after treatment. The tumors and vital organs were weighed and
homogenized, and the amount of DOX in the homogenates was ex-
tracted by CHCl3. After vacuum evaporation of CHCl3, the DOX was re-
dissolved in chromatographic grade methanol and filtered through a
0.45 μm ultrafiltration membrane, the concentration of DOX in the
tissue fluid was determined by high performance liquid chromato-
graphy (HPLC; Agilent 1260, USA). The mobile phase consisted of
0.01 M KH2PO4, acetonitrile and acetic acid in the ratio of 45: 55: 0.27
(V: V: V), the flow rate was 1 mL/min and detection wavelength was
490 nm. The column used was Agilent-SB-C18 measuring
4.6 × 150 mm and the packing particle size was 5 μm.
ODsample − ODnegative control
Hemolysis (%) =
×100%
OD positive control − ODnegative control
The experiment was performed in triplicates.
2.9. Cellular uptake assay
Skov 3 cells and HEK 293 T cells were seeded into 96-well plates at
the density of 5000 cells/well, and treated with free or micelle-en-
capsulated DOX (1, 3, 5 and 9 μg/mL DOX) after 24 h of culture. After
another 24 h incubation, the cells were gently washed thrice with PBS,
fixed with methanol, and stained with 1 μg/mL DAPI. The intracellular
localization of DOX was tracked by a fluorescence inversion microscope
(Olympus IX 73, Japan).
2.10. Stimuli-responsive intracellular cargo release from the micelles
Cargo release from the amphiphilic micelles under reductive sti-
mulation was tracked using the fluorescent probe coumarin 6 (C6). The
C6-loaded micelles were prepared as the DOX-loaded micelles (see
section 2.3), and the loaded amount was measured by absorption at
460 nm according to calibration curve. Glutathione (GSH) and the
glutathione inhibitor L-buthionine-sulfoximine (BSO) were used as the
positive and negative controls respectively. The Skov 3 cells and HEK
293 T cells were seeded into 96-well plates at the density of 5000 cells/
well, and pre-treated with 250 μM BSO or 5 mM GSH 24 h later. After
incubating for 12 h or 4 h, 3 μg/mL C6@CCPP-2 or C6@CPP-2 was
added, and the cells were cultured for another 1 h or 3 h, respectively.
The nuclei and GSH were respectively stained with Hoechst 33,342
(C1025, Beyotime) and thioltracker (T10095, Thermo Fisher Scientific)
for 30 min, and the fluorescence was tracked in real time by a High-
Content Imaging system (Perkin Elmer). To quantify C6 release in terms
of fluorescence intensity, the Skov 3 and HEK 293 T cells were seeded
into 6-well plates at the density of 2 × 105 cells/mL and treated as
described above. After 2 h, the cells containing C6-fluorescence were
digested with 0.25% (w/v) trypsin and washed thrice with PBS to
2.13. Histological examination and TUNEL assay
The mice were sacrificed on the 26th day post treatment, and the
tumors and major organs (heart, liver, spleen, lung and kidney) were
harvested, and fixed in 4% paraformaldehyde. The fixed specimens
were dehydrated, embedded in paraffin and cut into 5 μm-thick slices.
The sections were deparaffinized and stained with hematoxylin and
eosin (H&E) as per standard protocols, and observed under a light mi-
croscope. The apoptosis cells were detected using a TUNEL in situ Cell
Death Detection Kit according to the manufacturer’s instructions.
Briefly, the sections were deparaffinized, rehydrated and treated with
proteinase K for 15 min at room temperature, followed by incubation
with a TUNEL mixture at room temperature for 2 h. After counter-
staining with DAPI, the sections were observed with an inverted
fluorescence microscope.
4

X. Li, et al.
ColloidsandSurfacesB:Biointerfaces183(2019)110425
3. Results and discussion
3.2. Micelle formation and size distribution
3.1. Synthesis and characterization of CCPP and CPP
Since amphiphilic block copolymers are known to form core-shell
structure nanospheres with micellar properties [38], we hypothesized
similar characteristics of the CCPP and CPP copolymers. As shown in
Table S1, the CMCs of CCPP-1, 2and 3 were respectively 0.0296 mg/
mL, 0.0147 mg/mL and 0.0092 mg/mL, and that of CPP-1, 2 and 3 were
0.0244 mg/mL, 0.0143 mg/mL and 0.0094 mg/mL respectively.
Therefore, CMC values decreased with increasing Mw of PCL, indicating
that longer hydrophobic chains eased formation of amphiphilic mi-
celles. Furthermore, the nano-micelles formed with lower concentration
of the copolymers would have increased stability in circulation. We
tested micelle formation by the amphiphilic CCPP-2 and CPP-2 copo-
lymers and the hydrophilic mPEG shell by 1H-NMR. As shown in Fig.
S28 and S29, the PCL peaks disappeared in D2O solution while that of
mPEG remained visible, indicating that both copolymers assemble into
core-shell type nanospheres in water.
The size of the blank and DOX-loaded micelles ranged between 80
and 135 nm (Table S2). Not surprisingly, the DOX-loaded micelles were
larger than the corresponding blank micelles, indicating DOX entrap-
ment in the hydrophobic inner cavity. To further validate DOX en-
capsulation in the micelles, we analyzed the 1H-NMR spectrum of
DOX@CCPP-2 micelles in D2O (Fig. S30). Compared to the well-defined
peaks of free DOX in D2O (Fig. S30 (a)), DOX@CCPP-2 micelles lacked
signals for both DOX and PCL (Fig. S30 (c)), indicating their restricted
movement within the micellar core. Similar results were obtained for
DOX@CPP-2 (Fig. S31). Furthermore, the blank CCPP-2 and CPP-2
micelles had regular spherical morphology typical of hollow core-shell
structures, as shown in the TEM micrographs in Fig. 2A and B. The size
of the nanoparticles were ˜50–70 nm, making them suitable carriers for
hydrophobic drugs. In contrast, the DOX-loaded micelles had a more
“solid” appearance, indicating DOX encapsulation in the cavity (Fig. 2C
and D). The size of these loaded particles were ˜80-100 nm, which was
consistent with the DLS measurements (Fig. S32). After storage for 7
days, the DLS results (Fig. S33) showed that DOX@CCPP-2 and
DOX@CPP-2 in PBS had diameter of about 138 and 128 nm, respec-
tively. Compared to the previous DLS results of DOX@CCPP-2 and
DOX@CPP-2, they had no obvious change, which proved theirs good
stability to some extent. Finally, as shown in Fig. S34, the energy
spectra of CCPP-2 and CPP-2 also indicated micelle synthesis. Taken
together, both CCPP-2 and CPP-2 can self-assemble in water and en-
capsulate DOX in the hydrophobic pocket.
The synthetic routes of CCPP and CPP are illustrated in Fig. 1. The
compounds 1–8 were characterized by 1H NMR spectroscopy and FT-IR.
As shown in Fig. S2 and Fig. S3, two new peaks of the 1H NMR spectrum
appeared at 0.01 and 0.84 ppm, and the peak at 4.46 ppm disappeared,
indicating that the primary hydroxyl groups of β-CD were modified by
tert-butyldimethylsilyl. For compound 2, two new peaks appeared at
3.49 ppm and 3.65 ppm (Fig. S4), corresponding to the methyl group of
methyl iodide. Furthermore, the peaks at 5.75 ppm disappeared, cor-
responding to the modification of the second hydroxyl groups. The
disappearance of the peaks at 0.01 ppm and at 0.84 ppm indicated
synthesis of compound 3 (Fig. S5), and the appearance of a peak at
1.59–1.67 ppm corresponding to -SH confirmed replacement of the
bromide group with thiol to form compound 4 (Fig. S6). The formation
of compound 5 was indicated by an intense signal corresponding to PEG
at 3.64 ppm, four signals corresponding to PCL at 1.33–1.42 ppm,
1.61–1.68 ppm, 2.29–2.33 ppm and 4.04–4.08 ppm, and the dis-
appearance of the thiol signal at 1.59–1.67 ppm (Fig. S16-S18). Com-
pound
5
included the polymer series 2,3−OCH3-β-CD-(SS-PCLx-
mPEG2k)7, where x = 2k, 2.5k and 3k, of increasing molecular weights
(according to GPC traces shown in Fig. S25A), PCL chain length and
hydrophobic/hydrophilic ratios, and were accordingly designated as
CCPP-1, CCPP-2 and CCPP-3. The formation of compound 6 by the
iodination of the primary hydroxyl groups of β-CD was indicated by the
disappearance of the peak at 4.46 ppm (Fig. S20), while the appearance
of a new peak at 2.13 ppm confirmed the replacement of iodide with
thiol to form compound 7 (Fig. S21). Finally, as shown in Fig. S22-S24,
synthesis of compound 8 was validated by the appearance of signals
corresponding to PEG and PCL (see above), and disappearance of the
thiol signal at 2.13 ppm. Compound 8 included the amphiphilic poly-
mers β-CD-(SS-PCLx-mPEG2k)7, where x = 2k, 2.5k and 3k, with con-
trollable hydrophobic/hydrophilic ratios and were designated CPP-1,
CPP-2 and CPP-3 depending on the length of PCL chains. As with the
CCPP series, GPC traces (Fig. S25B) showed increasing molecular
weight from CPP-1 to CPP-3.
The FT-IR spectra of the β-CD and PEG/PCL compound series are
shown in Fig. S26. The appearance of a new peak at 2565 cm−1 (νS-H for
β-CD-(SH)7) (Fig. S26A (c)) indicated sulfhydration of the primary
hydroxyls of β-CD, while the peaks at 1467 cm-1 (νC-H for -C-(CH3)3)
and 1360 cm-1 (νC-H for Si-(CH3)2) indicated tert-butyldimethylsilyl
capping (Fig. S26B (b)). The enhanced peak at 2840 cm−1 (νC-H for
−OCH3) corresponded to methylation (Fig. S26B (c)), and the dis-
appearance of the peak at 1255 cm−1 after bromination (Fig. S26B (d))
3.3. Assessment of drug loading capacity and entrapment efficiency
The drug loading capacity (LC) and encapsulation efficiency (EE) of
the amphiphilic micelles were further assessed by UV–vis absorption.
The DOX-loaded micelles showed significantly higher absorption at
490 nm compared to the respective blank micelles (Fig. S35A and Fig.
S35C). The capsulation of DOX in the hydrophobic cavity of CCPP-2
and CPP-2 micelles was also exhibited by the slight red shift in the their
respective strongest absorption peaks compared to free DOX in water.
The drug loading behavior of the different micelles is summarized in
Table S1. CCPP-2 exhibited significantly higher EE and LC (83.9% and
31.9%) compared to CCPP-1 (62.4% and 25.4%) and CCPP-3 (71.9%
and 29.7%), indicating that the hydrophilic/hydrophobic ratio in the
amphiphilic β-CD block copolymer is a determining factor for high LC
and EE. Similarly, CPP-2 also showed higher EE and LC (76.6% and
28.6%) relative to CPP-1 (58.6% and 22.5%) and CPP-3 (68.7% and
26.8%). Therefore, the subsequent experiments were performed using
CCPP-2 and CPP-2.
validated (CH3O)14-β-CD-Br7 synthesis. Finally,
a new peak at
2720 cm−1 (νS-H for (CH3O)14-β-CD-(SH)7) (Fig. S26B (e)) was in-
dicative of (CH3O)14-β-CD-(SH)7 synthesis.. New peaks at 2945 (νC-H for
PCL), 2866 (νC-H for mPEG) and 1720 (νC=O for PCL) respectively
corresponded to PCL and mPEG (Fig. S26C (a) and C (b)), absence of a
peak at 3437 cm−1 (νO-H for mPEG-b-PCL) (Fig. S26C (c)) indicated
removal of the hydroxyl of mPEG-b-PCL by DTPA (Fig. S26C (d)), and
appearance of a peak at 2170 cm−1 (νS-H for mPEG2k-b-PCL2.5k-SH)
(Fig. S26C (e)) indicated reduction of mPEG-b-PCL-DTPA by DTT. The
FT-IR spectra of β-CD-(SH)7, (CH3O)14-β-CD-(SH)7 and mPEG2k-b-
PCL2.5k-SH are shown in Fig. S26D (a), (b) and (c). The peak at
2565 cm-1 (νS-H for β-CD-(SH)7) and 2170 cm−1 (νS-H for mPEG2k-b-
PCL2.5k-SH) disappeared and new ones appeared at 2942 (νC-H for PCL),
2885 (νC-H for mPEG) and 1730 (νC=O for PCL) upon oxidation (Fig.
S26D (d) and (e)). These results suggested that mPEG2k-b-PCL2.5k was
grafted onto β-CD through a disulfide bond. The FT-IR spectra of the
remaining two amphiphilic block copolymers are shown in the Fig. S27.
3.4. DOX release behavior in vitro
To demonstrate DOX release from the micelles under a reductive
stimulus, we compared the amount of free DOX in PBS (pH 7.4) and in
5

X. Li, et al.
ColloidsandSurfacesB:Biointerfaces183(2019)110425
Fig. 2. TEM images of (A) CCPP-2, (B) CPP-2, (C) DOX@CCPP-2 and (D) DOX@CPP-2.
under physiological conditions, but rapidly break under reducing con-
ditions which then accelerates CCPP-2 dissociation. While DOX@CPP-2
showed similar cumulative release rate in DTT, that in PBS was higher
compared to DOX@CCPP-2 under the same conditions. This could be
due to drug leakage from the DOX@CPP-2 micelles in physiological
conditions. Taken together, DOX@CCPP-2 displayed faster drug release
in the reductive environment, and no leakage under simulated phy-
siological conditions.
3.5. In vitro biocompatibility and hemolytic effect of the micelles
HEK 293 T cells and Skov 3 cells were incubated with different
concentrations of CPP-2 or CCPP-2 for 24 h, and the proportion of vi-
able cells were analyzed. As shown in Fig. 4A, more than 88% of the
cells survived even in the presence of 1 mg/mL of the micelles, in-
dicating excellent biocompatibility of the nanoparticles. In addition,
hemolysis assay was also performed to determine any potential detri-
mental effects of the copolymers on RBCs. As shown in Fig. 4B, the
percentage of hemoglobin release did not exceed 2% in the presence of
either the blank or DOX-loaded micelles. CCPP-2 and DOX@CCPP-2
showed < 1% hemolysis rate, even at the maximum concentration of
1 mg/mL, far below the international standard (5%), further validating
their good biocompatibility.
Fig. 3. DOX release curves under normal and reductive conditions from dif-
ferent micelles. Black square- DOX@CCPP-2 in 10 mM DTT; red circle-
DOX@CCPP-2 in PBS; blue triangle-DOX@CPP-2 in 10 mM DTT; teal inverted
triangle-DOX@CPP-2 in PBS.
DTT solution which simulated the reductive tumor microenvironment.
As shown in Fig. S36, DOX@CCPP-2 exhibited excellent release per-
formance under reductive stimulus comparing to the other two drug
carriers. Drug burst and lag release were not observed in the
DOX@CCPP-2. In contrast, the other two drug systems more or less
exposed weaknesses. While DOX@CPP-2 showed similar to the result of
DOX@CCPP-2 in comparison with DOX@CPP-1 and DOX@CPP-3 sys-
tems. In addition, DOX was released significantly faster from CCPP-2 in
the presence of DTT than in its absence, and the cumulative release rate
of DOX@CCPP-2 in DTT solution was > 50% within 8 h. In contrast,
the cumulative release rate in PBS was < 20% even after 100 h (Fig. 3).
This could be due to the fact that disulfide bonds are relatively stable
3.6. Cell uptake of the copolymer micelles
The cellular uptake of CCPP-2 and CPP-2 nanoparticles was tracked
by loading the fluorescent probe-C6 into the micellar cavity. In addi-
tion, the cells were also labeled with a thiol-tracker to detect disulfide
bond breakage in GSH and/or the mercapto groups on β-CD. As shown
in Fig. 5A, Skov 3 cells treated with C6@CCPP-2 or C6@CPP-2 showed
dual fluorescence, which enhanced significantly in the presence of GSH
in a concentration-dependent manner, indicating effective C6 release.
Addition of the GSH inhibitor BSO significantly weakened the fluores-
cence intensity of both probes, indicating that cargo release from these
6

X. Li, et al.
ColloidsandSurfacesB:Biointerfaces183(2019)110425
Fig. 4. (A) Viability of the HEK 293 T and Skov
3 cells incubated with varying concentrations of
CCPP-2 and CPP-2 for 24 h. (B) Hemolysis rate
of CPP-2, DOX@CPP-2, CCPP-2, DOX@CCPP-2
and DOX. (C–D) Viability of (C) HEK 293 T and
(D) Skov 3 cells incubated with varying con-
centrations of DOX, DOX@CPP-2 and
DOX@CCPP-2 for 24 h. *p < 0.05,**p <
0.01.
Fig. 5. (A–B) Representative fluorescence images of (A) Skov 3 and (B) HEK 293 T cells after incubation with C6@CCPP-2 and C6@CPP-2 for 1 h and 3 h. (C)
Representative flow cytometric histograms showing the cellular uptake of C6@CCPP-2 and C6@CPP-2 in Skov 3 and HEK 293 T cells. (D) Mean fluorescence intensity
of Skov 3 and HEK 293 T cells treated with C6@CCPP-2 and C6@CPP-2.
7

X. Li, et al.
ColloidsandSurfacesB:Biointerfaces183(2019)110425
micelles requires a reductive environment which breaks the disulfide
linkages and disintegrates the micelles. Furthermore, the fluorescence
intensity of cells incubated with C6@CCPP-2 was less compared to the
C6@CPP-2-treated cells in the absence of GSH, which indicated
minimal leakage of C6 from the CCPP-2 micelles under normal condi-
tions. In contrast to the Skov 3 cells, the non-transformed HEK 293 T
cells treated with C6@CCPP-2 or C6@CPP-2 fluoresced weakly even in
the presence of GSH (Fig. 5B), signifying the cancer cell-specificity of
these micellar carriers. Similar results were obtained with flow cyto-
metry analysis, both in terms of the percentage of fluorescently labeled
cells and the mean fluorescence intensity (MFI) (Fig. 5C and 5D). Taken
together, the amphiphilic copolymer micelles are highly effective drug
carriers that can selectively release the cargo in cancer cells, and CCPP-
2 in particular can achieve controlled drug release in the reductive
tumor microenvironment, thereby minimizing drug leakage in circula-
tion.
3.7. Anticancer effects of the DOX carriers in vitro
To further assess the specific anticancer effects of DOX@CCPP-2 and
DOX@CPP-2, we evaluated their toxicity in both HEK 293 T cells and
Skov 3 cells. As shown in Fig. 4C and Fig. 4D, the half maximal in-
hibitory concentration (IC50) of free DOX in HEK 293 T cells and Skov 3
cells were ˜0.9 μg/mL and 1 μg/mL, respectively, which points to in-
discriminate toxicity in both malignant and normal cells. In contrast,
the encapsulated DOX exhibited minimal toxicity in the HEK 293 T
cells, while maintaining a strong killing effect on the Skov 3 cells. The
IC50 values of DOX@CCPP-2 and DOX@CPP-2 loaded with equivalent
amount of DOX (5 μg/mL) were respectively 2 μg/mL and 1 μg/mL in
Skov 3 cells, and over 18 μg/mL and ˜14 μg/mL in HEK 293 T cells.
Since the GSH content in the Skov 3 cells is 1000 times higher com-
pared to that in 293 T cells, the disulfide bonds of the drug-loaded
nanoparticles are sensitive to breakage in the reductive environment of
these cells. Consistent with the IC50 values, the survival rate of Skov 3
cells and HEK 293 T cells treated with DOX@CCPP-2 was 29%and 83%
respectively, and 36% and 67% after treatment with DOX@CPP-2.
Therefore, DOX@CCPP-2 was selectively toxic towards Skov 3 cells
without significantly effecting HEK 293 T cells. In contrast, premature
DOX leakage from DOX@CPP-2 lowered its efficacy against the cancer
cell line, but resulted in increased toxicity to the non-transformed cells..
Furthermore, Skov 3 cells treated with DOX@CCPP-2, DOX@CPP-2 and
DOX for 24 h showed varying degrees of nuclear damage (Fig. S37A, C,
E), while only free DOX and DOX@CPP-2 resulted in nuclear de-
formation in the HEK 293 T cells (Fig. S37D, F), DOX@CCPP-2 ex-
hibited low cytotoxicity towards HEK 293 T cells (Fig. S37B). Taken
together, DOX@CCPP-2 combines superior anticancer efficacy and
negligible cytotoxicity, on account of high drug load capacity, stimulus-
responsive drug release, low leakage and appropriate particle size.
Fig. 6. Representative images showing time-dependent accumulation of
Dir@CCPP-2 and Dir@CPP-2 in Skov 3 tumor-bearing mice (upper panels). Bar
graphs showing fluorescence intensity of Dir in major organs and tumors after
36 h (lower panel).
kidneys. In contrast, Dir@CPP-2 showed slightly higher non-specific
accumulation compared to Dir@CCPP-2, due to its poor enrichment
effect as indicated in previous experiments. Taken together, the ap-
propriate size and excellent stimulus responsiveness of Dir@CCPP-2
enabled its targeted accumulation and release in the cancer tissues,
with minimal leakage to the normal tissues.
3.9. Antitumor efficacy of DOX@CCPP-2 and DOX@CPP-2 in vivo
The tumor-bearing mice were treated with saline, free DOX,
DOX@CPP-2 or DOX@CCPP-2, and the tumor sizes and body weight
were monitored regularly to assess the antitumor effects of the drug
carriers in vivo. As shown in Fig. 7, DOX@CCPP-2 showed significantly
better antitumor effect than the other groups. Although compared with
saline group, the other three groups could significantly inhibit the in-
crease of tumor volume. DOX@CCPP-2 had better efficacy because of
the EPR effect when compared with free DOX group. Fig. 7A, B showed
representative photograph images of tumor-bearing mice and excised
tumors after injection of saline, DOX, DOX@CPP-2 and DOX@CCPP-2,
respectively. The represent mice photos reflected the change of tumor
size and the tumor morphology of each group. DOX@CCPP-2 displayed
the strongest antitumor effect among all groups through visual ob-
servation. The mice treated with saline (control group) exhibited rapid
tumor growth tendency. However, DOX, DOX@CPP-2 and DOX@CCPP-
2 treatments displayed different influences on tumors growth inhibi-
tions when comparing with those of control groups (Fig. 7D). Inter-
estingly, unlike the results of cell viability evaluation in vitro, pure DOX
only displayed moderate inhibition effect on tumor growth in vivo. It
was related to the fact that free DOX quickly permeated into all tissues
of mice after intravenous injection, resulting in potential toxic side ef-
fect on normal tissues, thus only small amount of DOX could reach
tumor site in vivo. On the other hand, free DOX had only short life time
in vivo, and would be cleared off from the host via blood circulation and
3.8. In vivo distribution of the drug-loaded micelles
To track the in vivo distribution of the micelles, mice were injected
intravenously with Dir@CCPP-2 or Dir@CPP-2, and imaged 4, 12, 24
and 36 h later using the IVIS Lumina II multispectral imaging system. In
addition, the tumors and major tissues were dissected and photo-
graphed after 36 h using this imaging system. As shown in Fig. 6, both
Dir@CCPP-2 and Dir@CPP-2 accumulated at the tumor sites in a time-
dependent manner, likely due to the EPR effect, and showed controlled
release in the tumor tissues. However, Dir@CPP-2 was also noticeable
in the normal organs due to its leakage in the circulation. In addition,
the ex vivo images of the tumors and other tissues of the Dir@CCPP-2-
treated mice showed relatively strong fluorescence signals in the liver
and spleen, which could be due to the Kupffer cells in the liver and
spleen macrophages. Nevertheless, the tumors tissues had 2.8-fold and
6.9-fold higher fluorescence intensity compared to the liver and spleen
respectively, and very low signals were recorded in the heart, lungs and
8

X. Li, et al.
ColloidsandSurfacesB:Biointerfaces183(2019)110425
Fig. 7. (A) Representative photographs of mice after treatment with saline, DOX, DOX@CPP-2 and DOX@CCPP-2 for 26 days. (B) Body weights of the differentially-
treated mice during the treatment regimen normalized to the initial values (*P < 0.05 and **P < 0.01). (C) Representative images of the excised solid tumors after
26 days of treatment. (D) Relative tumor volume in the different groups during the treatment (*P < 0.05 and **P < 0.01). (E) Concentration of DOX in tumor
tissues and major organs after treatment (*P < 0.05 and **P < 0.01).
metabolism. DOX@CPP-2 group showed a moderate inhibition effect on
tumor growth, since CPP-2 could partially reduce DOX diffusion on its
way heading for tumor site. DOX@CCPP-2 group displayed the most
notable tumor growth inhibition efficiency, with significantly smaller
tumor size than those of other groups. To investigate the potential
tissue toxicity of DOX@CCPP-2 in vivo, the body weight of nude mice
was also periodically monitored (Fig. 7C). For free DOX group, the mice
weights started to decrease on the 4th day, indicating that DOX had
severe side toxic effect on nude mice. The weight of mice in the saline
group first increased and then decreased with the increase of tumor,
this was due to large tumors likely interfered with normal physiological
functions, resulting in weight loss. The body weights of nude mice in-
creased with different degrees after administrations of DOX@CPP-2 and
DOX@CCPP-2, respectively. The result suggested that CPP-2 and CCPP-
2 have good biocompatibility in vivo and reduce the toxicity of DOX as
nanocarriers. Taken together, DOX@CCPP-2 displayed excellent tumor
clearance with low toxicity to normal tissues, on account of the EPR
effect and excellent stimulus responsiveness.
group was 2.9 times greater than in the tumors of the free DOX group,
and 4 times higher than that in the normal tissues. In contrast, the DOX
concentration was higher in the normal organs of the free DOX-treated
mice compared to the tumor tissues. Therefore, in the absence of tar-
geted and controlled release, DOX can cause severe side effects when it
enters the circulation. The potential toxic effects of DOX, DOX@CPP-2
and DOX@CCPP-2 on both normal and tumor tissues were also eval-
uated by HE staining (Fig. 8A). Compared to pure DOX, the en-
capsulated forms resulted in negligible toxicity in the heart, liver,
spleen, lung and kidney, and caused significant damage in the tumor
tissues. Furthermore, the TUNEL assay showed significantly higher
proportion of apoptotic cells in the tumor tissues of the mice treated
with DOX@CCPP-2 or DOX@CPP-2 compared to the free DOX group
(Fig. 8B). In addition, DOX@CCPP-2 resulted in higher apoptosis rate
compared to DOX@CPP-2. Therefore, consistent with the results so far,
DOX@CCPP-2 exhibited superior anticancer performance and low toxic
side effects due to its targeted accumulation and stimuli-responsive
release.
3.10. Drug distribution and toxicity in vivo
4. Conclusions
The DOX content in the different tissues was measured by HPLC
after extracting and homogenizing the tissue samples. As shown in
Fig. 7E, the DOX content in the tumor tissues of the DOX@CCPP-2
We constructed a novel amphiphilic micellar drug delivery platform
with a β-CD core and amphiphilic block copolymer arms, which showed
high drug loading capacity, targeted uptake in cancer cells, controlled
9

X. Li, et al.
ColloidsandSurfacesB:Biointerfaces183(2019)110425
Fig. 8. (A) Representative images of H&E stained tissues of mice treated with saline, DOX, DOX@CPP-2 and DOX@CCPP-2 for 26 days. Nuclei-blue and cytoplasm-
red. (B) Representative fluorescence images showing TUNEL-positive cells (green) in the tumor tissues of different treatment groups. DAPI stained nuclei-blue, scale
bar =100 μm.
drug release under reductive conditions and minimal toxic effects in
normal cells. In the tumor-bearing mouse model, the DOX-loaded mi-
celles were enriched at the tumor site via EPR, and resulted in sig-
nificant tumor clearance with negligible effects on the healthy tissues.
This is a highly promising nanovehicle for targeted anti-cancer therapy
and should be further tested clinically.
Central South University for technical support.
Appendix A. Supplementary data
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.colsurfb.2019.110425.
References
Declaration of Competing Interest
[1] H.L. Mao, Y.D. Xie, H.C. Ju, H.S. Mao, L. Zhao, Z. Wang, L. Hua, C.M. Zhao, Y.L. Li,
R.T. Yu, H.M. Liu, Design of tumor microenvironment responsive drug-drug micelle
for cancer radiochemotherapy, ACS Appl. Mater. Interfaces 10 (2018)
33923–33935.
[2] T.M. Sun, Y.S. Zhang, B. Pang, D.C. Hyun, M.X. Yang, Y.N. Xia, Engineered nano-
particles for drug delivery in cancer therapy, Angew. Chem. Int. Ed. 53 (2014)
12320–12364.
[3] X.X. Shi, X.Q. Ma, M.L. Hou, Y.E. Gao, S. Bai, B. Xiao, P. Xue, Y.J. Kang, Z.G. Xu,
C.M. Li, pH-Responsive unimolecular micelles based on amphiphilic star-like co-
polymer with high drug loading for effective drug delivery and cellular imaging, J.
Mater. Chem. B Mater. Biol. Med. 5 (2017) 6847–6859.
[4] C.X. Yan, Z.Q. Guo, Y.Y. Shen, Y. Chen, H. Tian, W.H. Zhu, Molecularly precise self-
assembly of theranostic nanoprobes within a single-molecular framework for in vivo
The authors declare no competing interests.
Acknowledgement
This project was financially supported by the National Natural
Science Foundation of China (No.21576295), the Fundamental
Research Funds for the Central Universities of Central South University
(2017zzts175) and the Open-End Fund for the Valuable, Precision
Instruments of Central South University (CSUZC201830). We thank the
MRI facility of School of Chemistry and Chemical Engineering of
10

X. Li, et al.
ColloidsandSurfacesB:Biointerfaces183(2019)110425
[22] J.A. Yáñez, M.L. Forrest, Y. Ohgami, G.S. Kwon, N.M. Davies, Pharmacometrics and
delivery of novel nanoformulated PEG-b-poly(ε-caprolactone) micelles of rapa-
mycin, Cancer Chemother. Pharmacol. 61 (2008) 133–144.
[23] F. Liu, A. John, Y. Li, H. Majeed, R. Liang, Y. Ma, J.G. Ma, F. Zhong, Chitosan/
sulfobutylether-β-cyclodextrin nanoparticles as a potential approach for tea poly-
phenol encapsulation, Food Hydrocoll. 57 (2016) 291–300.
[24] L. Liu, C.X. Li, X.C. Li, Z. Yuan, Y.L. An, B.L. He, Biodegradable polylactide/poly
(ethylene glycol)/polylactide triblock copolymer micelles as anticancer drug car-
riers, J. Appl. Polym. Sci. 80 (2010) 1976–1982.
[25] E. Pierri, K. Avgoustakis, Poly(Lactide)-Poly(Ethylene glycol) micelles as a carrier
for griseofulvin, J. Biomed. Mater. Res. Part B Appl. Biomater. 75 (2005) 639–647.
[26] J. Szeman, H. Ueda, J. Szejtli, E. Fenyvesi, Y. Machida, T. Nagai, Complexation of
several drugs with water-soluble cyclodextrin polymer, Chem. Pharm. Bull. 35
(1987) 282–288.
[27] L. Zhang, J.F. Lu, Y.M. Jin, L.Y. Qiu, Folate-conjugated beta-cyclodextrin-based
polymeric micelles with enhanced doxorubicin antitumor efficacy, Colloids Surf. B
Biointerfaces 122 (2014) 260–269.
tracking of tumor-specific chemotherapy, Chem. Sci. 9 (2018) 4959–4969.
[5] H.Y. Xu, P.M. Yang, H.F. Ma, W.D. Yin, X.X. Wu, H. Wang, D.M. Xu, Z. Xia,
Amphiphilic block copolymers-based mixed micelles for noninvasive drug delivery,
Drug Deliv. 23 (2016) 3063–3071.
[6] X.L. Wei, Y.G. Wang, W.F. Zeng, F. Huang, L. Qin, C.L. Zhang, W. Liang, Stability
influences the biodistribution, toxicity, and anti-tumor activity of doxorubicin en-
capsulated in PEG-PE micelles in mice, Pharm. Res. 29 (2012) 1977–1989.
[7] N. Nishiyama, K. Kataoka, Nanostructured devices based on block copolymer as-
semblies for drug delivery: designing structures for enhanced drug function, Adv.
Polym. Sci. 193 (2006) 67–101.
[8] S. Zeng, M.P. Xiong, Trilayer micelles for combination delivery of rapamycin and
siRNA targeting Y-box binding protein-1 (siYB-1), Biomaterials 34 (2013)
6882–6892.
[9] J.Q. Yan, T. Su, F.R. Cheng, J. Cao, H. Zhang, B. He, Multifunctional nanoparticles
self-assembled from polyethylenimine-based graft polymers as efficient anticancer
drug delivery, Colloids Surf. B Biointerfaces 155 (2017) 118–127.
[10] D. Sutton, S.H. Wang, N. Nasongkla, J.M. Gao, E.E. Dormidontova, Doxorubicin and
beta-lapachone release and interaction with micellar core materials: experiment
and modeling, Exp. Biol. Med. 232 (2007) 1090–1099.
[28] J.S. Li, Z.Q. Guo, J.Y. Xin, G.L. Zhao, H.N. Xiao, 21-Arm star polymers with different
cationic groups based on cyclodextrin core for DNA delivery, Carbohydr. Polym. 79
(2010) 277–283.
[11] T. Noh, Y.H. Kook, C. Park, H. Youn, H. Kim, E.T. Oh, E.K. Choi, H.J. Park, C. Kim,
Block copolymer micelles conjugated with anti-EGFR antibody for targeted delivery
of anticancer drug, J. Polym. Sci. Part A: Polym. Chem. 46 (2008) 7321–7331.
[12] G.B. Hong, R.X. Yuan, B.L. Liang, J. Shen, X.Q. Yang, X.T. Shuai, Folate-functio-
nalized polymeric micelle as hepatic carcinoma-targeted, MRI-ultrasensitive de-
livery system of antitumor drugs, Biomed. Microdevices 10 (2008) 693–700.
[13] X.T. Shuai, H. Ai, N. Nasongkla, S. Kim, J.M. Gao, Micellar carriers based on block
copolymers of poly(ε-caprolactone) and poly(ethylene glycol) for doxorubicin de-
livery, J. Control. Release 98 (2004) 415–426.
[14] N.V. Cuong, M.F. Hsieh, Y.T. Chen, I. Liau, Synthesis and Characterization of PEG-
PCL-PEG Triblock Copolymers as Carriers of Doxorubicin for the Treatment of
Breast Cancer, J. Appl. Polym. Sci. 117 (2010) 3694–3703.
[15] Y.Y. Zhou, L. Li, W. Chen, D. Li, N.C. Zhou, J.L. He, P.H. Ni, Z.B. Zhang, X.L. Zhu, A
twin-tailed tadpole-shaped amphiphilic copolymer of poly(ethylene glycol) and
cyclic poly(ε-caprolactone): synthesis, Self-assembly and Biomedical Application,
Polym. Chem. 9 (2018) 4343–4353.
[29] S. Mura, J. Nicolas, P. Couvreur, Stimuli-responsive nanocarriers for drug delivery,
Nat. Mater. 12 (2013) 991–1003.
[30] Y. Yang, L.G. Xu, W.J. Zhu, L.Z. Feng, J.J. Liu, Q. Chen, Z.L. Dong, J.X. Zhao,
M.Y. Chen, One-pot synthesis of pH-responsive charge-switchable PEGylated na-
noscale coordination polymers for improved cancer therapy, Biomaterials 156
(2018) 121–133.
[31] C. Hegedűs, K. Kovács, Z. Polgár, Z. Regdon, É. Szabó, A. Robaszkiewicz,
H. JayForman, A. Martner, L. Virág, Redox control of cancer cell destruction, Redox
Biol. 16 (2018) 59–74.
[32] C.L. Sun, X.Y. Li, X.D. Du, T. Wang, Redox-responsive micelles for triggered drug
delivery and effective laryngopharyngeal cancer therapy, Int. J. Biol. Macromol.
112 (2018) 65–73.
[33] M. De Smet, S. Langereis, S.V. Den Bosch, H. Grüll, Temperature-sensitive lipo-
somes for doxorubicin delivery under MRI guidance, J. Control. Release 143 (2010)
120–127.
[16] A.A. Christy, Water adsorption properties of free and dehydrated β-Cyclodextrin
studied by near infrared spectroscopy and gravimetry, Key Eng. Mater. 689 (2016)
143–147.
[34] X.S. Meng, B. Gui, D.Q. Yuan, M. Zeller, C. Wang, Mechanized azobenzene-func-
tionalized zirconium metal-organic framework for on-command cargo release, Sci.
Adv. 2 (2016) e1600480.
[17] L.Y. Hu, J.F. Han, W.Q. Qiao, X.K. Zhou, C.L. Wang, D.G. Ma, Y.N. Li, Z.Y. Wang,
Side-chain engineering in naphthalenediimide-based n-type polymers for high-
performance all-polymer photodetectors, Polym. Chem. 9 (2018) 327–334.
[18] V.R. Yadav, S. Prasad, R. Kannappan, J. Ravindran, M.M. Chaturvedi, L. Vaahtera,
J. Parkkinen, B. Aggarwal, Cyclodextrin-complexed curcumin exhibits anti-in-
flammatory and antiproliferative activities superior to those of curcumin through
higher cellular uptake, Biochem. Pharmacol. 80 (2010) 1021–1032.
[19] M. Sivasubramanian, T. Thambi, V.G. Deepagan, G. Saravanakumar, H. Ko,
Y.M. Kang, J.H. Park, Carboxymethyl dextran-cyclodextrin conjugate as the carrier
of doxorubicin, J. Nanosci. Nanotechnol. 13 (2013) 7271–7278.
[35] C.Y. Chen, T.H. Kim, W.C. Wu, C.M. Huang, H. Wei, C.W. Mount, Y.Q. Tian,
S.H. Jang, S.H. Pun, A.K.Y. Jen, pH-dependent, thermosensitive polymeric nano-
carriers for drug delivery to solid tumors, Biomaterials 34 (2013) 4501–4509.
[36] Q. Zhao, T.T. Pan, G. Xiang, Z.P. Mei, J.P. Jiang, G. Li, X.S. Zou, M.W. Chen,
D.Z. Sun, S.M. Jiang, Y.Q. Tian, Highly efficient ratiometric extracellular oxygen
sensors through physical incorporation of a conjugated polymer and PtTFPP in graft
copolymers, Sens. Actuators B Chem. 273 (2018) 242–252.
[37] W.D. Ke, W. Yin, Z.S. Zha, J.F. Mukerabigwi, W.J. Chen, Y.H. Wang, C.X. He,
Z.S. Ge, A robust strategy for preparation of sequential stimuli-responsive block
copolymer prodrugs via thiolactone chemistry to overcome multiple anticancer
drug delivery barriers, Biomaterials 154 (2018) 261–274.
[20] L. Szente, J. Szejtli, Highly soluble cyclodextrin derivatives: chemistry, properties,
and trends in development, Adv. Drug Deliver Rev. 36 (1999) 17–28.
[21] S. Aquino, N. Quach, C. Richardson, C. Mcdonald, The potential use of cyclodextrins
in parenteral formulations, J. Parenter. Sci. Technol. 44 (1990) 48–49.
[38] G.S. Kwon, M. Naito, M. Yokoyama, T. Okano, Y. Sakurai, K. Kataoka, Physical
entrapment of adriamycin in AB block copolymer micelles, Pharm. Res. 12 (1995)
192–195.
11