All-active antitumor micelles via triggered lipid peroxidation

Article abstract of DOI:10.1016/j.jconrel.2018.08.003

Traditional antitumor nanomedicines have been suffering from the poor tumor targeting (ca. 1%) by the enhanced permeability and retention (EPR) effect, and the low drug loading (<5%). It was postulated that engineering all-active nanoplatform could increa

Full text of DOI:10.1016/j.jconrel.2018.08.003

Accepted Manuscript
All-active antitumor micelles via triggered lipid peroxidation
Min Gao, Xuan Meng, Xuliang Guo, Jundong Zhu, Aiping Fan,
Zheng Wang, Yanjun Zhao
PII:
S0168-3659(18)30455-3
doi:10.1016/j.jconrel.2018.08.003
COREL 9412
DOI:
Reference:
To appear in:
Journal of Controlled Release
Received date:
Revised date:
Accepted date:
29 March 2018
14 July 2018
1 August 2018
Please cite this article as: Min Gao, Xuan Meng, Xuliang Guo, Jundong Zhu, Aiping Fan,
Zheng Wang, Yanjun Zhao , All-active antitumor micelles via triggered lipid peroxidation.
Corel (2018), doi:10.1016/j.jconrel.2018.08.003
This is a PDF file of an unedited manuscript that has been accepted for publication. As
a service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof before
it is published in its final form. Please note that during the production process errors may
be discovered which could affect the content, and all legal disclaimers that apply to the
journal pertain.

ACCEPTED MANUSCRIPT
All-Active Antitumor Micelles via Triggered Lipid
Peroxidation
1
,†
1,†
1
1
1
1,2
,1
Min Gao, Xuan Meng, Xuliang Guo, Jundong Zhu, Aiping Fan, Zheng Wang, Yanjun Zhao*
1
School of Pharmaceutical Science & Technology, Tianjin Key Laboratory for Modern Drug Delivery
High Efficiency, and Collaborative Innovation Center of Chemical Science and Engineering (Tianjin),
Tianjin University, Tianjin 300072, China
&
2
State Key Laboratory of Medicinal Chemical Biology (Nankai University), Tianjin 300071, China
†
M. Gao and X. Meng contributed equally

To whom correspondence should be addressed.
Dr. Yanjun Zhao
School of Pharmaceutical Science & Technology, Tianjin University
9
2 Weijin Road, Nankai District, Tianjin 300072, China
Tel: +86-22-2740 7882, Fax: +86-22-2740 4018
Email: zhaoyj@tju.edu.cn

ACCEPTED MANUSCRIPT
ABSTRACT
Traditional antitumor nanomedicines have been suffering from the poor tumor targeting (ca. 1%) by the
enhanced permeability and retention (EPR) effect, and the low drug loading (< 5%). It was postulated
that engineering all-active nanoplatform could increase the therapeutic efficacy to enable the nanocarrier
function as both vehicle and active ingredient. To achieve this, a photosensitizer, Ce6 was encapsulated
within polymeric micelles with unsaturated fatty acids as the building blocks. Upon light irradiation, the
singlet oxygen produced by Ce6 induced lipid peroxidation, resulting in the generation of both active
free radicals and aldehydes. These supplementary radicals could exert cytotoxic effect for direct killing
tumor cells. The aldehyde end-products induced significant cell cycle arrest at G phase in 4T1 cells.
2
The peroxidation process also facilitated the on-demand disassembly of micelles and rapid release of
Ce6 to maximize the therapeutic effect of singlet oxygen. These all-active micelles showed a
significantly enhanced cytotoxicity with the half maximal inhibitory concentration (IC ) of 0.6 ± 0.2
5
0
μg/mL in contrast to the control micelles at 3.4 ± 0.5 μg/mL. The improved antitumor efficacy of the all-
active micelles was also demonstrated in the 4T1 tumor-bearing mice in vivo. The current work provides
a facile approach to enhance the antitumor efficacy of PDT nanomedicine using the biocompatible fatty
acids, which can be applied to various antitumor drugs and unsaturated lipids.
KEYWORDS: Micelles, all-active, fatty acid, self-assembly, lipid peroxidation.

ACCEPTED MANUSCRIPT
1
. INTRODUCTION
Solid tumor is characterized with a complex and abnormal microenvironment to fuel tumor
progression and resistance to various therapies [1, 2]. Nanomedicine can improve the anti-tumor therapy
with several commercialized products being developed [3-5]. The nanoformulations can increase the
therapeutic efficacy via manifold approaches, including solubilization of hydrophobic drugs, extended
systemic circulation, passive and/or active tumor targeting, and combinational delivery of multiple
agents [6-8]. However, the drug loading in traditional nanomedicines is very poor; it is usually less than
5
% (w/w) [9-11]. In addition, the actual dose reaching the tumor tissue by passive targeting (i.e.
enhanced permeability and retention/EPR effect) has been extremely limited and the typical dose in the
tumor was ca. 1% (w/w) upon intravenous delivery [12-14]. Namely, the majority mass of the
nanoformulations is of no use in terms of therapeutic efficacy. Hence, frequent dose administration is
necessary to boost up the drug concentration in the tumor, resulting in poor patient compliance.
Moreover, the repeated dosing would produce substantial accumulation of the toxic agents in healthy
organs via biodistribution, hence leading to severe adverse effects [15]. Site-specific all-active
nanomedicine brings hope to address this challenging issue as the toxicity is restricted to the disease site
and the content of active species available in the tumor are raised.
Photodynamic therapy (PDT) is a kind of site-specific delivery system that involves molecular oxygen,
photosensitizer, and light irradiation [16]. Upon laser irradiation, the photosensitizer produces highly
active reactive oxygen species (ROS), e.g. singlet oxygen to exert the cytotoxic effect. Since the light
treatment is only applied to the tumor tissue, the side-effects to healthy organs can be minimized.
Nanoscale PDT combines the benefits of ROS therapy and nanocarriers, but it still faces the problem of
short half-life and limited diffusion distance of singlet oxygen when the photosensitizer is confined in
the nanocarrier [17-19]. Our recent work reported a size-expandable micellar nanocarrier that can

ACCEPTED MANUSCRIPT
rapidly release the photosensitizer and singlet oxygen upon photo-triggering [20]. Nevertheless, the
activity of such nanosystem is only from the drug (e.g. photosensitizer). Further integration of the
concepts of site-specific activation of therapeutic agents with all-active nanomedicine (e.g. drug
nanocrystals) would considerably improve the anti-tumor outcome with decreased adverse effects [21].
Unfortunately, not every drug can be engineered to nanocrystals due to their diverse physicochemical
properties.
Lipid peroxidation is a phenomenon that often refers to the oxidative degradation of polyunsaturated
fatty acids [22]. During this process, both lipid radicals and lipid peroxyl radicals were generated, which
is also accompanied with the production of reactive aldehydes, e.g. malondialdehyde (MDA) and 4-
hydroxynonenal (HNE) [23, 24]. The presence of singlet oxygen in PDT can speed up this chain
reaction process and the obtained lipid oxidation products thus induce cytotoxicity and tissue injury [25,
2
6]. In addition, HNE involves multiple antitumor mechanisms, hence acting as a promising cytotoxic
agent [27]. It was envisioned that incorporating lipid peroxidation mechanism in nanoscale PDT
antitumor system could create a tumor-specific all-active nanomedicine for enhanced antitumor efficacy.
Since the fatty acids are hydrophobic, they can be engineered as the biocompatible building block of
nanocarrier via self-assembly to realize triggered activation of nanocarrier [28, 29].
It was hypothesized that lipid peroxidation could induce the triggered disassembly of fatty acids-
bearing polymeric micelles, rapid release of PDT payload, and generation of supplementary MDA and
HNE for efficacy amplification. The aim of this work was to tailor-make fatty acids-bearing micellar
nanocarrier as a proof-of-concept of tumor-specific all-active nanomedicines (Scheme 1). Three
different types of fatty acids were selected and covalently conjugated to the multivalent methoxy
poly(ethylene glycol)-co-poly(lysine) copolymer (mPEG-PLys), generating a amphiphilic conjugate that
could self-assemble into micelles. The model photosensitizer, chlroin e6 (Ce6) was physically loaded

ACCEPTED MANUSCRIPT
within the micelle core. Upon light irradiation, Ce6 produces singlet oxygen that not only acts as the
ROS to inhibit tumor growth, but also initiates the lipid peroxidation process and releases other types of
cytotoxic products from the vehicle for supplementary therapeutic effect. Therefore, every part of the
micellar nanomedicine (i.e. both the drug and the vehicle) would have an antitumor effect for a
maximized efficacy. In addition, the lipid peroxidation could induce micelle disassembly and hence
enhance Ce6 release from nanocarriers to maximize the effect of singlet oxygen [18]. As the application
of light trigger can be spatially and temporally controlled, the adverse effects of this all-active
nanosystem should be minimal.
Scheme 1. Illustration of all-active micellar nanocarrier for enhanced antitumor therapy.
Polyunsaturated fatty acids are conjugated with methoxyl poly(ethylene glycol)-co-poly(lysine) to
produce amphiphilic conjugates that can self-assemble into micelles and physically encapsulate the
photosensitizer, chlorin e6 (Ce6). Upon laser irradiation, the generated singlet oxygen by Ce6 can
induce lipid peroxidation and micelle disassembly. The peroxidation products can act as the second
source of active agents besides singlet oxygen, i.e. both the drug and vehicle is active upon photo-
triggering.

ACCEPTED MANUSCRIPT
2
. MATERIALS AND METHODS
2
.1 Materials
L-lysine β-benzyl ester was purchased from Bide Pharmatech Ltd. (Shanghai, China). Methoxy
poly(ethylene glycol) amine (mPEG-NH , 5,000 Da) was sourced from JenKem Technology Co., Ltd.
2
(Beijing, China). Triphosgene was obtained from InnoChem Science & Technology (Beijing, China).
Singlet Oxygen Sensor Green (SOSG) (Invitrogen, USA) was purchased from Thermofisher (Beijing,
China). Other chemicals and fluorescent probes came from Sigma-Aldrich (Beijing, China). All standard
solvents were purchased from Guangfu Fine Chemical Research Institute (Tianjin, China) unless
otherwise specified. All chromatographic grade solvents came from Concord Tech Co., Ltd. (Tianjin,
China). Dulbecco’s modified Eagle’s medium (DMEM, Gibco, NY, USA), fetal bovine serum (Hyclone,
USA), penicillin-streptomycin (Gibco, USA) and MDA assay kit were purchased from Baibei Biotech
(Tianjin, China). Female BALB/c mice were sourced from Vital River Laboratory Animal Technology
Co. Ltd. (Beijing, China).
2
.2 Polymer synthesis
Lys(Z)-NCA was synthesized by referring to the literature procedure [30]. Briefly, in a round-bottom
flask fitted with a stir bar, L-lysine β-benzyl ester (1 g, 3.57 mmol) was suspended in anhydrous THF
bubbled with argon flux. After adding 0.5 equiv. of triphosgene, the suspension was stirred at 55°C for 2
h. Then, the filtrate of reaction mixture was concentrated and crystallized in hexane to achieve the
1
purified Lys(Z)-NCA (yield: 78%). H NMR(400 M, DMSO-d , ppm): 9.09 (s, 1H), 7.32 (m, 5H), 5.00
6
(s, 2H), 4.41 (m, 1H), 2.99 (m, 2H), 1.69 (m, 2H), 1.39-1.29 (m, 4H).
The ring opening polymerization (ROP) of Lys(Z)-NCA was initiated by mPEG-NH in the mixed
2
solvent of dry DMF and CHCl . Briefly, Lys(Z)-NCA (0.85 g, 2.77 mmol) and mPEG-NH (1 g, 0.20
3
2
mmol) were dissolved in 10 mL anhydrous DMF and 10 mL anhydrous CHCl under argon atmosphere.
3

ACCEPTED MANUSCRIPT
o
The reaction was maintained at 35 C for 3 days and then the reaction mixture was precipitated in ice-
1
cold diethyl ether to yield mPEG-PLys(Z) (yield: 79%). H NMR (400 MHz, DMSO-d , ppm): 7.33 (m,
6
5
0H), 4.99 (s, 20H), 4.20 (d, 10H), 3.51 (m, 457H), 3.22 (s, 3H), 2.94 (s, 20H), 1.62-1.22 (m, 60H).
The benzyl group in mPEG-PLys(Z) was de-protected by 33% (w/v) HBr/acetic acid at ambient
temperature to obtain mPEG-PLys. Specifically, 7.6 mL TFA was added to a round bottle flask to
dissolve the polymer mPEG-PLys(Z) (0.69 g, 0.08 mmol) under argon protection. Then 0.76 mL 33%
(w/v) HBr/acetic acid was injected dropwise to the above solution and the reaction was maintained for 2
h. Then the polymer was purified by precipitation using glacial diethyl ether followed by filtration. The
1
obtained product was further washed twice with diethyl ether and dried under vacuum. (yield: 86%). H
NMR (400 MHz, DMSO-d , ppm): 4.22 (m, 10H), 3.51 (m, 457H), 3.23 (s, 3H), 2.76 (s, 20H), 1.71-
6
1
.37 (m, 60H).
To generate conjugated polymers, a series of fatty acid differing in chemical structure were selected
including linoleic acid, α-linolenic acid and arachidonic acid. The multivalent conjugation of fatty acid
to the polymer backbone (mPEG-PLys) was achieved through the classical amidation reaction. In short,
linoleic acid (0.59 mL, 1.9 mmol) was activated by EDC·HCl (0.37 g, 1.9 mmol) and HOBt (0.26 g, 1.9
mmol) in 20 mL DCM at ambient temperature under the protection of argon for 1 h. Then the mixture of
mPEG-PLys (0.12 g, 0.019 mmol) and DIPEA (0.33 mL, 5.7 mmol) in 10 mL DCM was transferred
o
dropwise to the activated fatty acid solution, followed by continuous stirring at 35 C under argon for 24
h. Afterwards, the condensed crude product was purified through precipitation in ice-cooled diethyl
ether thrice and dialyzing against water (molecular weight cut-off/MWCO: 1,000 Da) for 24 h before
1
freeze-drying. (yield: 79%). H NMR (400 MHz, DMSO-d , ppm): 5.33 (m, 16H), 4.20 (m, 10H), 3.50
6
(m, 457H), 3.24 (s, 3H), 2.74 (s, 20H), 2.26-2.01 (m, 32H), 1.75-1.12 (m, 116H), 0.83 (s, 12H). The α-
linolenic acid and arachidonic acid conjugated polymers were synthesized using the same protocol for

ACCEPTED MANUSCRIPT
1
linoleic acid conjugate. Conjugate polymer of α-Linolenic acid (yield: 78%): H NMR (400 MHz,
DMSO-d , ppm): 5.31 (m, 24H), 4.17 (m, 10H), 3.51 (m, 457H), 3.24 (s, 3H), 2.80 (s, 20H), 2.33-2.00
6
1
(
m, 40H), 1.73-1.07 (m, 100H), 0.84 (s, 12H). Arachidonic acid conjugated polymer (yield: 73%): H
NMR (400 MHz, DMSO-d , ppm): 5.32 (m, 32H), 4.19 (m, 10H), 3.51 (m, 457H), 3.24 (s, 3H), 2.78 (s,
6
2
0H), 2.31-1.86 (m, 48H), 1.71-1.04 (m, 92H), 0.83 (m, 12H).
2
.3 Micelle Generation and Characterization
The molecular weight and their distribution of the polymers were determined by a Waters 1525 gel
permeation chromatography (GPC) system. THF was used as the mobile phase with the elution rate at 1
o
mL/min. The temperature was maintained at 35 C and polystyrene was used as the standard to make the
calibration curve. The self-assembly of amphiphilic polymer-fatty acid conjugates utilized the typical
dialysis method [31]. In brief, mPEG-PLys-Fat was dissolved in DMF, followed by dialysis against
water with regular replacement under argon protection. The micelles were lyophilized ready for use. The
Ce6-loaded micelles were produced using the similar method aforementioned. Excess amount of Ce6
together with polymer conjugates (mPEG-PLys-Fat) were dissolved in DMF and dialyzed against water;
after centrifugation and filtration, the micelle solution were lyophilized ready for further use. The Ce6
content in the micelles was analyzed by the high performance liquid chromatography (HPLC) using a
previously published method [20]. The aggregation number (n ) of the micelles was determined via the
agg
Debye plot using our previously published method [32]. The hydrodynamic size of all types of micelles
2
with or without laser treatment (660 nm, 100 mW/cm ) was also determined by the Malvern Zetasizer
Nano ZS instrument. The measurements were performed at ambient temperature in triplicate. To
confirm the physical stability of this singlet oxygen responsive system, all the Ce6 loaded micelles (1
mg/mL) were dissolved in PBS buffer (pH 7.4) and monitored the corresponding particle size every
other day up to a week. The micelle core diameter and morphology at the absence or presence of laser

ACCEPTED MANUSCRIPT
2
irradiation (660 nm, 100 mW/cm ) were observed and photographed by a JEM-100CX II transmission
electron microscope (TEM). In detail, the sample solution (5 mg/mL, 20 µL) was placed onto copper
grids with carbon support and air-dried prior to image acquisition. The critical micelle concentration
(CMC) of conjugate micelles was determined using a published method wherein pyrene was the
fluorescence probe [33]. Micellar aqueous solutions (0.4-400 µg/mL) were supplemented with fixed
pyrene probe at a molar concentration of 0.5 µM. The excitation wavelength was at 333 nm and the
spectral slit width was 5 nm. The emission spectra were recorded from 350 nm to 450 nm. The I /I
3
84 373
ratio was plotted against the logarithm of micelle concentration and the CMC value was calculated from
the flexion point of the curve.
2
.4 UV-vis Spectroscopy and Raman Spectroscopy
The aqueous solutions of RM -Ce6 and CM-Ce6 with Ce6 concentration at 20 μM were prepared
AA
and investigated by an ultraviolet-visible (UV-vis) spectrophotometer (Agilent, Cary 60). The spectra
2
were recorded after irradiation (660 nm, 100 mW/cm ) in several intervals. Raman spectra were
acquired in air at room temperature with a confocal Raman micro spectrometer (RENISHAW, inVia
reflex). The test was under the output of 320 mW and excitation wavelength of 785 nm He-Ne laser
illumination, with 50 × optical objectives (Olympus) and 100 s integration time. The spectrometer was
−
1
calibrated with a characteristic line of Si at 520cm . Arachidonic acid samples were dissolved in
ethanol with the concentration of 2 mM. The RM -Ce6 samples were aqueous solution with the AA
AA
concentration at 2 mM.
2
.5 In Vitro Release
o
The release of Ce6 from both control and responsive micelles were performed at 37 C in a set of static
2
Franz diffusion cells with ca. 2 cm effective permeation area. Each sample was carried out in triplicate.
The receptor chamber was filled with pH 7.4 disodium hydrogen phosphate and citric acid buffer and

ACCEPTED MANUSCRIPT
stirred by a magnetic bar. 2 mL micelle solution was placed in the donor chamber with ca. 1 µmol of
Ce6 applied dose. A regenerated cellulose membrane (MWCO 3,500Da) was clamped between two
2
chambers. The release study was performed in laser off and laser on (660 nm, 100 mW/cm , and 20 min)
conditions. Subsequently, 0.5 mL receiver fluid was collected at pre-determined time points. The same
volume of fresh buffer was replenished after each collection to keep the whole volume constant. Finally,
the collected samples were subjected to HPLC analysis for determination of Ce6 and the released Ce6
was plotted against time.
2
.6 Extracellular ROS Detection
SOSG analysis: The solutions of Ce6, RM -Ce6 and CM-Ce6 at a Ce6 detection concentration of 0.2
AA
μM in deuterium water were blended with a singlet oxygen sensor green (SOSG) fluorescent probe at a
final concentration of 10 μM in equal volumes; the samples (200 μL) were then pipetted into a black 96-
2
well plate, respectively. The samples were then subjected to a laser beam (660 nm, 100 mW/cm ) and
the fluorescent intensities (E : 504 nm, E : 525 nm) were recorded at 0, 0.5, 1.5, 2.5, 3.5 and 5.5 min
x
m
post-irradiation by a microplate reader (TECAN, Infinite 200 PRO). Each sample was prepared and
determined in triplicate.
DCFH-DA analysis: DCFH-DA was hydrolyzed to DCFH by 0.01N NaOH aqueous solution at room
temperature and diluted to 10 μM with disodium hydrogen phosphate and citric acid buffer. Then,
RM -Ce6 and CM-Ce6 dispersed in the above buffer with a Ce6 concentration of 0.2 μM were mixed
AA
with DCFH solution in equal volumes, respectively. As DCFH can turn to highly fluorescent DCF upon
oxidation by reactive oxygen species, 200 μL mixed solution were pipetted into a blank 96-well plate
and the fluorescent intensities (E : 485 nm, E : 530 nm) of DCF were recorded using the microplate
x
m
2
reader upon irradiation (660 nm, 100 mW/cm ) at 0, 0.5, 1.5, 2.5, 3.5 and 5.5 min intervals (n = 3). The
samples dispersed in dimethyl sulfoxide were performed in the same method.

ACCEPTED MANUSCRIPT
ESR analysis: The electron spin resonance experiments were performed using DMPO and TEMP as
spin traps on an ESR instrument (Bruker, EMX plus). For the detection of DMPO-•OH signal, CM-Ce6
and RM -Ce6 were dissolved in a 100 mM DMPO aqueous solution with a Ce6 concentration at 125
AA
2
μM. The ESR spectra of samples with/without laser treatment (660 nm, 100 mW/cm , and 5 min) were
recorded at room temperature using an X-band spectrometer which was operated at 9.83 GHz. The
conditions for measuring the ESR spectra were as follows: microwave power, 10 mW; field modulation,
1
0
.05 mT at 100 kHz; scan time, 100 s. The detection of TEMP adducts with O generated by Ce6, CM-
2
Ce6 and RM -Ce6 was implemented as the condition of DMPO system.
AA
2
.7 Detection of Aldehyde End-products
The generated aldehydes via lipid oxidation were confirmed through fluorescence derivatization
techniques [34]. Briefly, 8 mM alpha-furil and 3.0 M ammonium acetate were prepared in methanol and
glacial acetic acid, respectively. Both RM -Ce6 and CM-Ce6 were dissolved in methanol with the
AA
concentration at 2 mM. Subsequently, 250 μL sample solution with/without irradiation (660 nm, 100
2
mW/cm , 20 min) was transferred to a round-bottom flask individually followed by adding 300 μL
o
ammonium acetate (3.0 M) and 800 μL alpha-furil (8.0 mM). The mixture refluxed at 90 C under argon
for 30 min and then cooled down to record the fluorescence spectra (E : 250 nm, E : 280-550 nm) by a
x
m
fluorescence spectrophotometer (Edinburgh Instruments, FLS980).
2
.8 Cell Culture and Animal Model
Murine breast cancer 4T1 cell line was kindly offered by the State Key Laboratory of Medicinal
Chemical Biology, Nankai University. The cells were cultured in DMEM supplemented with 10% fetal
o
bovine serum and 1% penicillin/streptomycin in an incubator with 5% CO at 37 C. Female BALB/c
2
mice were used under protocols approved by the Animal Ethics Committee of Tianjin University. To

ACCEPTED MANUSCRIPT
6
develop the tumor model, 2 × 10 4T1 cells suspended in PBS were subcutaneously injected into the
3
right flanks of each mouse. The mice were used when tumor volumes reached about 50-100 mm .
2
.9 Cellular Uptake, Cytotoxicity and Cell Cycle
T1 cells were seeded on 35-mm plates at 8×10 cells/well for confocal imaging. After 36 h’s
4
4
incubation, the cells were treated with control or fatty acid micelles (Ce6 dose = 5 μg/mL) for 4 h. After
2
washing with PBS three times, the cells were irradiated by the 660 nm laser for 10 min (100 mW/cm )
and then labeled with 4’, 6-diamidino-2-phenylindole (DAPI). Cells without laser treatment were used
as the control. The excitation wavelengths of confocal laser were 633 nm for Ce6 (red) and 405 nm for
DAPI (blue).
The cell viability was measured by the standard MTT assay, the 4T1 cells were seeded on 96-well
3
plates at a density of 4 × 10 cells/well and treated with free Ce6, CM-Ce6, and RM -Ce6, respectively
AA
(n = 6). The Ce6 dose ranged from 1 to 10 μg/mL. After incubation for 4 h, the cells of laser groups
were washed with PBS and added with fresh medium. Then the light irradiation was applied for 10 min
2
(
660 nm, 100 mW/cm ). Samples without laser treatment were used as control. After incubation for 24
h, the cell medium was replaced with MTT solution. After 30 min’s incubation, the absorption at 490
nm was obtained using the microplate reader (TECAN, Infinite 200 PRO). The MTT assay results for
AA and RMAA and CM were obtained with the same methods as that for the Ce6 system. The dose of
AA and CM was also consistent with Ce6 system. In terms of cell cycle analysis, 4T1 cells were
incubated with the above three samples at the Ce6 concentration of 3 μg/mL. The groups with laser
irradiation were treated using the same condition of MTT assay. After 24 h’s incubation, the cells were
washed with PBS and detached with trypsin. Then the cells were washed with PBS again and fixed in
o
ethanol at 4 C overnight before the analysis. For the cell cycle analysis, the fixed cells were carefully
washed with PBS and treated with PI-staining solution containing 100 μg/mL RNase and 50 μg /mL

ACCEPTED MANUSCRIPT
propidium iodide (PI) for 30 min, followed by analysis using a flow cytometer (Becton & Dickinson,
FACSCalibur). The data were analyzed using the FlowJo 7.6.1 software.
2
.10 Intracellular ROS Detection
DCFH-DA probe: Intracellular ROS was detected using DCFH-DA which can passively diffuse into
the cells. Upon the cleavage of acetate groups by intracellular esterase, a fluorescent adduct (DCF) is
produced by oxidation. To analyze DCF signal, the 4T1 cells treated with free Ce6 and Ce6-loaded
micelles for 4 h were washed with PBS, followed by incubation with 20 µM DCFH-DA in DMEM for
3
2
0 min at 37˚C in darkness. After washing thrice with serum-free DMEM, cells were irradiated (660 nm,
2
0 mW/cm , 5 min) ready for imaging-taking.
SOSG probe: 4T1 cells were incubated with Ce6, CM-Ce6 and RM -Ce6 4 h, respectively. SOSG
AA
was dissolved in DMEM and added into the cells at the final concentration of 1 mM. After 3 h’s
2
incubation, the cells were photo-irradiated at 660 nm (20 mW/cm , 5 min), followed by CLSM analysis
of singlet oxygen.
HPF probe: Hydroxyphenyl fluorescein (HPF) was employed to detect ·OH. HPF is a cell membrane
permeable fluorescent probe, which can be converted to fluorescein by ·OH. After the formulation
treatment, cells were incubated with 50 µM HPF in DMEM for 30 min, followed by laser irradiation and
CLSM image-recording.
All analyses were performed using a LSM-710 confocal laser scanning microscope (Carl Zeiss,
Germany). Regarding DCF, SOSG and HPF signals, FITC filter was used with excitation wavelength of
4
88 nm.
2
.12 Intracellular lipid peroxidation detection

ACCEPTED MANUSCRIPT
The MDA assay kit was employed to determine the aldehydes produced by lipid peroxidation of
6
unsaturated fatty acid. 4T1 cells were seeded on the 60 mm dishes with the density of 2×10 cells/well,
followed by incubation for 36 h. After washing with PBS, the cells were incubated with CM-Ce6 and
RM -Ce6 for 4 h; the nanocarrier concentration was set at 10 μM. Then TBA solution was added into
AA
2
the lysis solution of detached cells after the laser treatment (660 nm, 100 mW/cm , 10 min). The
reacting adduct between MDA and thiobarbituric acid (TBA) was quantified colorimetrically (OD 532
nm) by the UV-vis spectrophotometer. The final MDA concentration was quantified according the
product protocol.
2
.13 Biodistribution
The samples (100 μL) of RM -Ce6, CM-Ce6, and free Ce6 were injected intravenously (i.v.) via the
AA
tail vein (n = 3) with the Ce6 concentration being fixed at 0.5 mg/mL. The kinetic fluorescent images of
mice were captured at 2 h, 4 h, 6 h, 8 h, and 24 h using a Cri Maestro in vivo imaging instrument
(Cambridge Research & Instrumentation, Inc., MA, USA). To see the ex vivo fluorescence distribution,
major organs and tumors were collected in the end of study, i.e. 24 h post dose administration, followed
by imaging and fluorescence recording.
2
.14 In vivo efficacy
To assess the biocompatibility of the Ce6, CM-Ce6 and RM -Ce6, hemolytic test was conducted. In
AA
brief, a mixture of 2% aqueous erythrocyte suspension and sample solution (1: 1, v/v) was incubated at
o
3
7 C for 3 h. The concentration of Ce6 for each sample was kept identical at 50 μM. Normal saline and
deionized water was used as the negative and positive control, respectively. Finally, the supernatant was
collected after centrifugation (15 min, 295 g), and assayed in the microplate reader at 540 nm (n = 3).
4
T1 tumor-bearing mice were i.v. injected with 100 μL of PBS, free Ce6, CM-Ce6 and RM -Ce6
AA
(concentration of Ce6 = 0.5 mg/mL), respectively. Laser irradiation was conducted at 8 h and 24 h post

ACCEPTED MANUSCRIPT
2
dosing (660 nm, 150 mW/cm , 20 min). The first time point was to match the peak of EPR effect. Each
group consisted of six mice. Tumor volume and body weight were measured every other day for 16
2
days. The tumor volume was calculated according to the following formula: width × length/2. Animals
were continually monitored and euthanized upon reaching the humane endpoints; then the survival rate
was calculated [35]. In the end of efficacy experiment, the tissue slices of tumors and main organs were
stained by H&E following the standard protocol and photographed using a Leica inversion microscopy
with a digital camera [36]. In terms of the degree of apoptosis, terminal deoxynucleotidyl transferase
dUTP nick end labeling (TUNEL) assay was carried out using an in situ cell death detection kit
(Promega Biotech Co., Ltd.) and then observed by the Carl Zeiss confocal microscope.
2
.15 Plasma Stability
The plasma stability of Ce6 loaded micelles were monitored by measuring the fluorescent intensity of
Ce6 [37]. In brief, CM-Ce6 and RM -Ce6 were dispersed in FBS containing 10% (v/v) PBS solution
AA
o
with a final Ce6 concentration of 25μg/mL and kept at 37 C. The fluorescent intensities (F ) at 650 nm
t
(x: 408 nm) were recorded at predetermined intervals and divided by the corresponding fluorescence at
0
h (F ). The fluorescence intensity ratio (F /F ) was plotted against time.
0
t
0
2
. 16 Statistical Analysis.
All data were presented as the mean ± standard deviation. Statistical analysis of sample differences
was carried out using Student’s t-test or analysis of variance (ANOVA) followed by Tukey’s post-hoc
analysis. Statistical significant was determined at p < 0.05.
3
. RESULTS AND DISCUSSION
3
.1 Preparation and Characterization of Micellar Nanocarriers.

ACCEPTED MANUSCRIPT
The synthesis of mPEG-PLys copolymer employed the typical ring-opening polymerization of N-
carboxyanhydrides with mPEG-NH (5000 Da) as the initiator [38]. Three fatty acids with increasing
2
unsaturation degree, i.e. linoleic acid (LA), α-linolenic acid (ALA) and arachidonic acid (AA), were
conjugated to the backbone of poly(lysine) via amidation using a post-polymerization modification
approach (Scheme 2) [39]. Poly(lysine) functioned as a multivalent template to enable the attachment of
multiple fatty acids per polymer conjugate, which was advantageous over the approach of linking single
fatty acid with the PEG in terms of facile manipulation of micelle property [40]. The successful
1
synthesis of three types of conjugates was verified by H NMR analysis (Figure S1-S3). The mean
polymerization degree of poly(lysine) was 10 and the corresponding conjugation degree of all three fatty
acids were ca. 40%. The molecular weights of three conjugates were 7328 Da (LA), 7320 Da (ALA),
and 7424 Da (AA), respectively. The molecular weight of polymers was also analyzed by GPC (Table
S1). The low polydispersity of these polymers (below 1.2) indicated that the molecular weight was
narrowly distributed. The micelles preparation and drug loading employed a traditional dialysis method
[
31]. These ROS-responsive micelles (RM) were abbreviated as RM , RM , and RM , individually
LA ALA AA
to differentiate their side chains with increasing double bonds from LA to AA. The non-responsive
control micelles (CM) were made of mPEG-PLys(Z) with no fatty acids. The aggregation number (nagg)
of all three types of responsive fatty acid micelles was lower than that of CM micelles (Table S2), which
was presumed due to the lower conjugation degree of fatty acids. The Ce6 loading was quantified by
high performance liquid chromatography (HPLC) with no interference from conjugates (Figure S4).
The Ce6 loading values were 3.0 ± 0.4% (w/w) (CM), 2.9 ± 0.5% (w/w) (RM ), 3.0 ± 0.3% (w/w)
LA
(
RMALA), and 3.2 ± 0.5% (w/w) (RM ), responsively.
AA

ACCEPTED MANUSCRIPT
Scheme 2. Synthetic route of amphiphilic polymer-fatty acid conjugates.
As an indicator of the micelle stability, the critical micelle concentration (CMC) of three RM
were larger than that of CM (p < 0.05), which was because of the relatively low conjugation efficiency
of fatty acids (ca. 40%) in contrast to that of CM at 100% [41]. However, all three types of RMs
displayed the same stability; there was no significant difference among the CMC of these RMs (p > 0.05)
(Figure 1A). The three Ce6 loaded RMs exhibited particle size increase upon laser treatment (Figure
1
B), which was presumed due to the lipid peroxidation-induced micelle disassembly that was initiated
by the singlet oxygen produced by Ce6 [29, 42]. The extent of size increase was well correlated with the
unsaturation degree of fatty acids. The increase of double bond number should lead to a higher extent of
fatty acid oxidation and hence more evident micelle disassembly as a result of the impaired
amphiphilicity [20, 43]. The constant monitoring hydrodynamic size of micelles showed that the tailored
micelles are stable at the absence of irradiation (Figure S5). The slight size increase of RMs in contrast
to CM was thought as the oxidation effect induced by the oxygen in the air; such phenomenon should be
avoided when the samples were stored in an oxygen-free container. The serum stability of micelles plays
a key role in determining the in vivo performance of micelles [40]. Based on a fluorescence approach

ACCEPTED MANUSCRIPT
[
37], both the CM and RMs displayed good stability in plasma (Figure S6), which would be beneficial
for the subsequent in vivo efficacy assessment. The drug release profile from RMs concurred well with
the particle size analysis. At the presence of laser treatment, the rate and extent of Ce6 release from RMs
ranked as follows: RMAA > RMALA > RM , while they were very similar under dark (Figure 1C, 1D).
LA
RMAA contains more double bonds, which makes it highly sensitive to singlet oxygen oxidation,
resulting in the most distinct disassembly [44, 45]. The fatty acid type-dependent collapse of RMs upon
singlet oxygen challenge was further verified by transmission electron microscopy (Figure 2, Figure S7,
and Table S3). In contrast, the CM-Ce6 did not show such behavior because of the absence of fatty
acids (Figure S8). The polydispersity of poly(lysine) as well as the fatty acid architecture can affects the
amphiphilicity and self-assembly of conjugate polymers, resulting in a polydisperse micelle distribution
even at the absence of irradiation (Figure 2). Upon laser treatment, the fatty acid side chain was
oxidized to different degrees, leading to the diminishment of assembly driving force and hence micelle
aggregation. Therefore, the size of irradiated micelles was more polydisperse. In addition, the lipid
peroxidation process is accompanied with the generation of MDA that could function as a micelle cross-
linker to covalently react with the amine in the polymer backbone. All these factors work together,
contributing to the formation of large clusters in Figure 2. Although the laser irradiation-initiated RMs
disassembly and accompanied Ce6 release was a complex process that was associated with multiple
cascade events, the payload release mechanism was presumed as peroxidation-induced micelle
disassembly, which was demonstrated by the drug release profile and micelle stability with/without
irradiation treatment (Figure 1B-1D and Figure S5-S6).

ACCEPTED MANUSCRIPT
Figure 1. Interplay of micelle stability and drug release with or without light irradiation. (A) Critical
micelle concentration (CMC) assessment; (B) Hydrodynamic size of singlet oxygen-responsive micelles
2
with increasing duration of laser treatment (660 nm, 100 mW/cm ); (C, D) Photo-triggered Ce6 release
o
2
from micelles at 37 C with or without light irradiation (660 nm, 100 mW/cm , 20 min). Data are
presented as means ± standard deviation (n = 3). CM designates control micelles. RM , RMALA, and
LA
RMAA indicates responsive micelles (RM) containing linoleic acid (LA), α-linolenic acid (ALA) and
arachidonic acid (AA), respectively.
Figure 2. Transmission electron microscope (TEM) images of Ce6-loaded responsive micelles with and
2
without irradiation (660 nm, 100 mW/cm , 20 min). Scale bar: 100 nm.
3
.2 Triggered Lipid Peroxidation.

ACCEPTED MANUSCRIPT
Based on the particle sizing results, RMAA was selected as the model system for further assessment of
lipid peroxidation due to the highest unsaturation degree of AA among all three fatty acids. Upon
2
irradiation (660 nm, 100 mW/cm ), a new absorption peak at 306 nm emerged; its intensity increased
with continuous irradiation, which indicated the progressive AA peroxidation (Figure 3A). In contrast,
there was no such change for the Ce6-loaded CM upon irradiation since no fatty acids were present
−1
(
Figure 3B). The Raman spectroscopy of both RMAA and free AA exhibit a band at 3015 cm as a
consequence of the stretching mode of =C-H bond associated with the C=C double bonds [46]. In
addition, the intensity of this band is roughly proportional to the number of double bonds in the fatty
2
acids. Upon irradiation (660 nm, 100 mW/cm , 20 min), the diminishment of the band intensity at 3015
−
1
cm demonstrated the disappearance of C=C bonds and the progression of AA peroxidation for both
free AA and AA-bearing micelles (Figure 3C, 3D). During the oxidation process, the formation of
aldehydes products was also detected using a fluorescence-based approach (Figure S9) [34].
Figure 3. Spectroscopic analysis of photo-triggered peroxidation of arachidonic acid (AA) and
responsive micelles (RM ) in contrast to control micelles (CM). (A and B) UV-vis spectra of Ce6-
AA
2
loaded RMAA and CM upon laser irradiation for up to 5 min (660 nm, 100 mW/cm ); (C and D) Raman
spectra of Ce6-loaded RMAA and free AA with and without irradiation (660 nm, 100 mW/cm , 20 min).
2

ACCEPTED MANUSCRIPT
The ESR analysis showed that singlet oxygen was produced by all three samples (RM -Ce6, CM-
AA
Ce6, and free Ce6) upon irradiation in contrast to that under dark (Figure 4A, 4B). This was clearly
because of the presence of Ce6. However, RM -Ce6 also generated highly reactive hydroxyl radical
AA
post irradiation, but CM-Ce6 did not do so irrespective of laser treatment (Figure 4C, 4D) [47]. The
hydroxyl radical can be produced by the decomposition of hydroperoxide or lipid peroxide [48].
Regardless of the irradiation time, RM -Ce6 yielded more ROS than CM-Ce6 did by virtue of the
AA
involvement of lipid peroxidation and the production of various types of ROS besides singlet oxygen in
the former nanosystem (Figure 4E). Although both RM -Ce6 and CM-Ce6 could produce singlet
AA
oxygen upon laser treatment, the duration of irradiation made a difference. Both micellar systems
generated the same amount of singlet oxygen before 5.5 min, as evidenced by the fluorescence intensity
of the specific SOSG probe. After that time point, RM -Ce6 produced more “apparent” singlet oxygen
AA
because of the disassembly of micelles that enabled the rapid singlet oxygen diffusion out of the
micelles (p < 0.05) (Figure 4F).

ACCEPTED MANUSCRIPT
Figure 4. Triggered lipid peroxidation in vitro. (A) Electron spin resonance (ESR) spectra of the spin
2
adduct formed post laser treatment (660 nm, 100 mW/cm ) of RM -Ce6, CM-Ce6 and free Ce6 for 5
AA
min at ambient temperature (TEMP probe: 100 mM); (B) ESR spectra of RM -Ce6, CM-Ce6 and free
AA
Ce6 without irradiation (TEMP: 100 mM); (C) DMPO spin trapping ESR spectra of RM -Ce6 in water
AA
2
for •OH detection with or without irradiation (660 nm, 100 mW/cm , 5 min); (D) DMPO spin trapping
ESR spectra of CM-Ce6 with or without irradiation (660 nm, 100 mW/cm , 5 min); (E) Comparison of
2
the RM -Ce6 and CM-Ce6 micelles regarding ROS production in water upon irradiation with DCF as
AA
2
the probe (660 nm, 100 mW/cm ); (F) Singlet oxygen detection of RM -Ce6 and CM-Ce6 using SOSG
AA
probe under the same irradiation conditions.* p < 0.05.
3
.3 Cellular Uptake and Cytotoxicity.
Since Ce6 displays inherent fluorescence, no external probe is required to label the micelle. Both
types of nanocarriers could transport Ce6 into the 4T1 cells. Regarding RMAA formulation, the intensity
of red fluorescence of Ce6 was much higher upon irradiation in contrast to that under dark. In contrast,

ACCEPTED MANUSCRIPT
Figure 5. (A) CLSM images of 4T1 cells incubated with RM -Ce6 and CM-Ce6 by recording Ce6
AA
2
fluorescence. “Laser on” represents a condition of 660 nm, 100 mW/cm for 10 min (Scale bar: 50 μm);
(
B and C) Viability of 4T1 cells in response to free Ce6, CM-Ce6 and RM -Ce6 with or without
AA
irradiation (n = 5); (E) The IC50 values for Ce6, CM-Ce6 and RM -Ce6 at the presence of laser
AA
treatment, ** p < 0.01.
there was no such difference in terms of CM system (Figure 5A). This was presumed due to the AA
oxidation-induced transition of the hydrophobic block of polymer to relatively hydrophilic one, which
would eventually cause micelle disassembly and facilitated cargo release (Figure 2). The employment
of lipid oxidation to initiate nanocarrier damage has been well investigated [29, 49]. The singlet oxygen-
initiated polymer amphiphilicity change and micelle collapse was also reported recently [20]. The rapid

ACCEPTED MANUSCRIPT
Ce6 release from RMAA other than CM would overcome the barrier of short half-life and limited
diffusion radius of singlet oxygen [17-19]. This could ultimately enhance the cytotoxicity of RMAA
(Figure 5B-5D). RMAA showed certain degree of toxicity even without the light irradiation (Figure 5C),
which was presumed as the autonomous AA peroxidation initiated by the intracellular ROS (Figure
S10). Obviously, there was no such cytotoxicity in terms of placebo CM system (Figure S11).
Figure 6. Confocal microscopy images of 4T1 cells after incubation with Ce6, CM-Ce6 and RM -Ce6
AA
2
with and without light irradiation (660 nm, 20 mW/cm , and 5 min) for comparison of intracellular ROS
with the aid of DCFH-DA probe. Scale bar: 20 μm.
However, the cytotoxic effect of RMAA was not only from Ce6, but also from the lipid peroxidation
products. The intracellular confocal images clearly showed that RMAA produced more ROS compared to
its counterpart CM upon laser treatment (Figure 6). These ROS generated by RMAA included not only
singlet oxygen, but also hydroxyl radicals as an evidence of AA oxidation (Figure S12, S13). The
oxidative degradation of AA also generated MDA, a typical lipid peroxidation end-product (Figure S14).
These intracellular data agreed well with the results in aqueous buffers (Figure 3, 4 and Figure S9).
Therefore, the AA peroxidation process produced both ROS and aldehydes that could enhance the
cytotoxicity of micellar nanosystem, i.e. the vehicle was also active. The ROS obtained from AA

ACCEPTED MANUSCRIPT
oxidation could directly exert the toxicity since they are extremely active. The aldehyde end-products of
lipid peroxidation have been proved able to induce cell cycle arrest [50, 51]. This was the case in the
current study; upon laser irradiation, RMAA induced significant cell cycle arrest at G phase in contrast
2
to CM and free Ce6 (Figure S15).
Figure 7. Biodistribution of three formulations (Ce6, CM-Ce6, and RM -Ce6) in 4T1 tumor-bearing
AA
mice post intravenous administration. (A) The in vivo kinetics of Ce6 accumulation in tumors. (B) Semi-
quantitative ex vivo fluorescent imaging of tumor as well as healthy organs 24 h post dosing; (C) Kinetic
fluorescence intensity of tumors; (D) The fluorescence intensity of tumors and major organs 24 h post
administration. Heart (H), Liver (Li), Spleen (Sp), Lung (Lu), Kidney (K), Tumor (T).
3
.4 Biodistribution.
Taking advantage of the intrinsic fluorescence of Ce6, the biodistribution study was performed
without micelle labelling by probes. Both micellar formulations (RMAA and CM) displayed a
characteristic profile of nanomedicines accumulation in tumor. The Ce6 amount deposited in the tumor
tissue (i.e. the micelle mass) reached the peak 6 h post intravenous (i.v.) administration (Figure 7A, 7C).
This was due to the EPR effect; namely, nanoparticles sized between 100 nm and 200 nm usually could

ACCEPTED MANUSCRIPT
extravasate through vascular fenestrations of tumors [15, 52]. The transient openings and closings at the
leaky blood vessels were shown to contribute to the enhanced nanoparticle extravasation into tumors
[
53]. However, the extent of tumor drug deposition was much lower for the free Ce6 as the absence of
EPR effect-based passive targeting. Since RMAA and CM displayed similar size and surface charge, they
almost delivered the same mass of Ce6 to the tumor 24 h post i.v. injection (p > 0.05). Besides tumor, all
three formulations also accumulated in other healthy organs, e.g. heart, liver, spleen, lung, and kidney
(Figure 7B, 7D). Particularly, the drug accumulation in liver and spleen was substantial since both
organs were the major sites of nanoparticle elimination [54]. The micelle accumulation in different
organs is a kinetic process, which is also accompanied with payload release. Nevertheless, we only
collected the data 24 h post dosing to minimize the use of animals. At this time point, the high value of
micelle accumulation in tumor was presumed still as a consequence of EPR effect. As a semi-
quantitative approach, the fluorescence-based technique cannot give the information of accurate dose
percentage in reference to the injected dose, which is one limit of such method for in vivo application.
3
.5 In Vivo Efficacy.
Both types of micelles showed good biocompatibility with the degree of hemolysis at 1.6 ± 0.5%
(
RM ) and 0.9 ± 0.1% (CM), respectively in contrast to free Ce6 (7.6 ± 0.4%), indicating no hemolytic
AA
potential of micellar formulations for in vivo application (Figure S16) [55]. The four formulations
ranked as follows in terms of tumor growth inhibition: RMAA > CM > Ce6 > PBS (Figure 8A). As the
negative control, PBS showed no antitumor effect because of no active ingredient therein. Both RMAA
and CM were superior to the free Ce6 as a consequence of the passive tumor targeting effect. As
expected, RMAA defeated CM and exhibited the best antitumor outcome since its activity came from
both the drug (Ce6) and the vehicle (lipid peroxidation products). Moreover, the RMAA formulation
could release Ce6 more rapidly upon triggering due to the on-demand micelle disassembly, which could

ACCEPTED MANUSCRIPT
maximize the effect of Ce6 as the issue of short half-life and limited diffusion distance of singlet oxygen
was addressed. Such effect was also observed in other types of stimuli-responsive PDT micelles [20].
The in vivo results agreed well with the in vitro cytotoxicity data (Figure 5). As shown in Figure 8B,
untreated 4T1 tumor bearing mice (n = 6) displayed a median survival time of 25 days whereas RMAA-
Ce6 treatment improved the median survival time to 33 days (p < 0.01). The body weight of mice was
recorded accompanied with the tumor mass, showing no sign of toxicity for formulations of free Ce6,
CM and RMAA (Figure 8C). The weight of mice in PBS group started to decrease at the beginning of
second week post dose administration, which was due to reduced food uptake as a result of uncontrolled
tumor growth. At the end of efficacy study, the tumor tissues were collected and their weights ranked as:
RMAA < CM < Ce6 < PBS, demonstrating the supreme anti-tumor performance of all-active RMAA
(
Figure 8D). Both histological analysis and apoptosis investigation of the harvested tumor tissues at the
end of efficacy study verified the optimal antitumor efficacy of the all-active RMAA formulation (Figure
E). Additionally, no obvious histopathological change was observed in liver, lung, spleen, heart and
kidney as shown in Figure S17.
8

ACCEPTED MANUSCRIPT
Figure 8. In vivo antitumor efficacy of micelles in 4T1 tumor model. (A) Tumor growth curves of
different groups treated by four formulations (PBS, free Ce6, RM -Ce6, and CM-Ce6). (B) Survival
AA
curve (Kaplan–Meier) of different treatment groups. (C) Variation of body weight during the treatment
course. (D) Average tumor mass in the end of efficacy experiment. (E) Histological analysis of tumor
tissue by H&E staining (top) and apoptosis analysis of tumor tissue by TUNEL staining (bottom) in the
end of efficacy study. Data are presented mean ± standard deviation (n = 6), * p < 0.05, ** p < 0.01.

ACCEPTED MANUSCRIPT
4
. CONCLUSION
In summary, we demonstrated the all-active antitumor PDT nanomedicine using unsaturated fatty
acids as the key building block of polymeric micelles. Such micellar nanoformulations are advantageous
in terms of on-demand rapid cargo release and triggered lipid peroxidation. The former may liberate the
singlet oxygen out of the carrier to fully exert its action. The latter not only produces active ROS, but
also generates aldehydes that can regulate multiple signaling pathways to induce cell arrest and
apoptosis. As the traditional EPR effect can only deposit very limited nanomedicines (ca. 1%) in the
tumor tissue post i.v. administration. Among these 1% mass of nanomedicine, only the active agent (ca.
5
% mass) can exert the antitumor effect and the majority mass of nanomedicine shows no potency. The
current approach could fully utilize the 1% nanomedicine accumulated in the tumor by activate the
vehicle through triggered lipid peroxidation. Such proof-of-concept can be extended to a broad variety
of unsaturated lipids, photosensitizers, as well as different potential agents that can trigger the
peroxidation of vehicles.
ACKNOWLEDGMENT
We acknowledge the funding support from the National Basic Research Program of China
(2015CB856500), National Natural Science Foundation of China (21650110447), and the Tianjin
Research Program of Application Foundation and Advanced Technology (18JCZDJC35700).
Appendix a. Supplementary Data
Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.jconrel.2016.xx.xxx