Hypoxia-activatable nano-prodrug for fluorescently tracking drug release in mice

Article abstract of DOI:10.1007/s11426-020-9880-7

Chemotherapy is one of the commonly used methods to treat various types of cancers in clinic by virtue of its high efficiency and universality. However, strong side effects and low concentration of conventional drugs at the tumor site have always been important factors that plague the chemotherapy effects of patients, further precluding their practical applications. Thereof, to solve the above dilemma, by integration of anticancer drug (nitrogen mustard, NM) into an NIR fluorophore (a dicyanoisophorone derivative), an intelligent prodrug NIR-NM was developed via molecular engineering strategy. Prodrug NIR-NM stimulated in hypoxia condition exhibits significantly higher toxicity to cancer cells than normal cells, essentially reducing the collateral damage to healthy cells and tissues of nitrogen mustard. More importantly, the nanoparticle prodrug FA-lip@NIR-NM showed the advantages of the high accumulation of drug at tumor site and long-circulation capacity in vivo, which endowed it the ability to track the release of the active chemotherapeutic drug and further treat solid tumors.[Figure not available: see fulltext.].

Full text of DOI:10.1007/s11426-020-9880-7

SCIENCE CHINA
Chemistry
•
ARTICLES•
March 2021 Vol.64 No.3: 499–508
https://doi.org/10.1007/s11426-020-9880-7
Hypoxia-activatable nano-prodrug for fluorescently tracking drug
release in mice
1
,2†
1,3†
4
2
1
1
Haidong Li , Qichao Yao , Zhongji Pu , Jeewon Chung , Haoying Ge , Chao Shi ,
1
1
1
1
1,3
4
Ning Xu , Feng Xu , Wen Sun , Jianjun Du , Jiangli Fan , Jingyun Wang ,
2
*
1,3*
Juyoung Yoon & Xiaojun Peng
1
State Key Laboratory of Fine Chemicals, Dalian University of Technology, Dalian 116024, China;
2
Department of Chemistry and Nanoscience, Ewha Womans University, Seoul 03760, Korea;
3
Ningbo Institute of Dalian University of Technology, Ningbo 315016, China;
4
School of Bioengineering, Dalian University of Technology, Dalian 116024, China
Received August 6, 2020; accepted September 21, 2020; published online January 18, 2021
Chemotherapy is one of the commonly used methods to treat various types of cancers in clinic by virtue of its high efficiency and
universality. However, strong side effects and low concentration of conventional drugs at the tumor site have always been
important factors that plague the chemotherapy effects of patients, further precluding their practical applications. Thereof, to
solve the above dilemma, by integration of anticancer drug (nitrogen mustard, NM) into an NIR fluorophore (a dicyanoiso-
phorone derivative), an intelligent prodrug NIR-NM was developed via molecular engineering strategy. Prodrug NIR-NM
stimulated in hypoxia condition exhibits significantly higher toxicity to cancer cells than normal cells, essentially reducing the
collateral damage to healthy cells and tissues of nitrogen mustard. More importantly, the nanoparticle prodrug FA-lip@NIR-NM
showed the advantages of the high accumulation of drug at tumor site and long-circulation capacity in vivo, which endowed it the
ability to track the release of the active chemotherapeutic drug and further treat solid tumors.
NIR fluorescent probe, activatable, hypoxia, prodrug, imaging
Citation:
Li H, Yao Q, Pu Z, Chung J, Ge H, Shi C, Xu N, Xu F, Sun W, Du J, Fan J, Wang J, Yoon J, Peng X. Hypoxia-activatable nano-prodrug for
fluorescently tracking drug release in mice. Sci China Chem, 2021, 64: 499–508, https://doi.org/10.1007/s11426-020-9880-7
1
Introduction
in tumor sites is also the bottleneck of treatment effect [6,7].
Additionally, in solid tumors, hypoxia is due to the pro-
liferation of cancer cells driven by uncontrolled oncogene,
altered metabolic characteristics, and abnormal tumor blood
vessels, which is an important sign of tumor progression [8–
14]. There are ample evidences to show that the poorly
oxygenated tumor is related to poor prognosis of cancer
patients because of the emergence phenomena of the in-
creased tumor angiogenesis, anti-apoptosis and multi-drug
resistance in the malignant hypoxic microenvironment [15–
Chemotherapy is the use of chemical drugs for the treatment
of tumor cells, which is mostly used in the management of
the middle and late stages of various malignant tumors in
clinic, owing to its inherent high efficiency and universality
[1,2]. Unfortunately, many conventional anticancer drugs
kill cancer cells and are highly toxic to normal cells, which
make patients suffer side effects from losing healthy cells [3–
5
]. The low concentration accumulation of anticancer agent
1
8]. Intriguing, prodrugs have low systemic toxicity and side
effects before reaching the tumor site; and they are activated
by tumor overexpressed substances (e.g., enzymes, redox
†
*
These authors contributed equally to this work.
Corresponding authors (email: jyoon@ewha.ac.kr; pengxj@dlut.edu.cn)
©
Science China Press and Springer-Verlag GmbH Germany, part of Springer Nature 2021
chem.scichina.com link.springer.com

500
Li et al. Sci China Chem March (2021) Vol.64 No.3
species, and antigen) to exert antitumor efficacy, which have
become one of the hot spots of clinical trials [19–25].
Thereof, it is of great significance to develop chemother-
apeutic prodrugs that can be used in tumor anaerobic mi-
croenvironment.
To maximize the anticancer efficacy of chemotherapy
prodrugs, understanding the underlying mechanism of hy-
poxia and related micro-environmental parameters in tumor
is vital [26–29]. Previous studies have demonstrated that
some intracellular reductases such as DT-diaphorase [30,31],
nitroreductase [32–34], and azo reductase [12,35,36], are
overexpressed in anaerobic cells and solid tumors. Hence,
some elegant prodrugs decorated with the quinone, ni-
trobenzene/nitroimidazole, or azo group as the hypoxia-
activatable moieties have been reported in recent years
Interestingly, under simulated anaerobic conditions (2% O2),
prodrug NIR-NM could be activated by the overexpressed
reductases in cell, generating the active therapeutic drug for
selectively killing tumor cells. Additionally, the amphiphilic
block copolymer DSPE-PEG2K-FA, soy lecithin, choles-
terol, and NIR-NM were chosen to construct the nanosized
drug delivery system FA-lip@NIR-NM, which possessed a
satisfactory size of 170.3 nm, high stability as well as en-
capsulation efficiency. More importantly, the nanoparticle
FA-lip@NIR-NM mainly accumulated at tumor sites of tu-
mor-bearing BABL/c mice owing to the enhanced perme-
ability and retention effect (EPR) and displayed the
prolonged blood circulation in vivo, which might be em-
ployed for in-situ real-time tracking drug release and antic-
ancer platform.
[12,32,35,37–41]. For instance, Kim et al. [20,35,38,42]
have designed some prodrugs with superior performance for
the treatment of tumors in vivo. Wu and coworkers [12]
described a DT-diaphorase-activatable prodrug for monitor-
ing camptothecin (CPT, an anticancer agent) and inhibiting
tumor growth. Although great efforts have been made in the
past decades to develop prodrugs, the available prodrugs for
tumor treatment are still in their infancy and there is still
much room for the development of new prodrugs for clinical
use.
As we know, nitrogen mustard (NM) is a class of antitumor
drugs used clinically, which has achieved outstanding effi-
cacy [43,44]. However, when it kills tumor cells in vivo, it is
inevitably highly toxic to healthy cells, causing the strong
side effect. To solve the above deficiencies, by integrating
nitrogen mustard into the NIR fluorophore [45], we herein
used the molecular engineering strategy to develop an azo
reductase-activatable prodrug NIR-NM, as shown in Scheme
2 Experimental
2
.1 Materials and instruments
All the reagents used were acquired from commercial sup-
pliers and were used without further purification unless
1
13
otherwise noted. H/ C-NMR spectra were measured on a
Bruker 400 MHz spectrometer (Germany). Dynamic light
scattering (DLS) was recorded with a Nano-ZS (Nanozs90,
UK). High pressure nanometer homogeneous instrument
(AH-NANO, ATS Engineering Inc.) was used to construct
nanoparticle. The fluorescence quantum yield (Φ ) for
f
compound was tested using an Absolute PL Quantum Yield
Spectrometer (HAMAMATSU C11347, Japan). pH mea-
surements were obtained with pH meter (Ohaus Starter 2100,
USA). Cell imaging was performed on FV 3000 confocal
microscopy (Olympus, Japan). In vivo imaging was gathered
with the small-animal imaging equipment (NightOWL II
LB983, Germany). Flash column chromatography was car-
ried out through silica gel. The interferential reagents were
1
. As a proof-of-concept, prodrug NIR-NM emitted the NIR
fluorescence of 650 nm upon sodium dithionite activation,
which was verified by both fluorescence and mass spectra.
Scheme 1 Schematic illustration for the fabrication and activation process of FA-lip@NIR-NM in vivo (color online).

Li et al. Sci China Chem March (2021) Vol.64 No.3
501
prepared according to published literatures [46,47].
pressure. Finally, the residue was purified by silica gel col-
umn chromatography to obtain 41 mg prodrug NIR-NM
(
Yield 26%).
2
.2 Synthesis and characterization of prodrug NIR-
1
H NMR (400 MHz, CDCl ): δ 7.96 (d, J=8.7 Hz, 2H),
.91 (d, J=8.4 Hz, 2H), 7.63 (d, J=8.4 Hz, 2H), 7.08 (d, J=
3
NM
7
Synthetic procedure of prodrug NIR-NM is illustrated in
Figure 1 with detailed procedure as follows.
5.0 Hz, 2H), 6.89 (s, 1H), 6.80 (d, J=8.9 Hz, 2H), 3.87 (t, J=
6.9 Hz, 4H), 3.71 (t, J=6.8 Hz, 4H), 2.62 (s, 2H), 2.49 (s,
2
H), 1.10 (s, 6H).
1
3
2
.2.1 Synthesis and characterization of compound 2
C NMR (100 MHz, CDCl ): δ 169.06, 153.53, 149.14,
3
Compound 1 (3.62 g, 20 mM) was dissolved in 5 mL POCl3.
The mixture was stirred at 110 °C for 1 h. When the reaction
dropped to room temperature, 10 g ice was added into the
mixture. After that, the mixture was extracted with ethyl
acetate (3×15 mL) three times. The organic phase was dried
with anhydrous Na SO for overnight and vacuum filter. The
144.63, 137.00, 136.24, 129.95, 128.37, 125.87, 124.06,
123.12, 113.45, 112.67, 111.82, 79.03, 53.55, 43.01, 40.25,
39.25, 32.05, 29.70, 28.04.
+
ESI-HRMS: m/z calcd. for [C H Cl N ] 518.1873,
2
9
30
2
5
found, 518.1863.
2
4
crude product was purified through silica gel column chro-
2
.3 Construction of liposome and encapsulation effi-
matography to obtain 3.52 g compound 2 (Yield 81%).
ciency
1
H NMR (400 MHz, CDCl ): δ 7.33–7.22 (m, 2H), 6.79 (t,
3
J=7.3 Hz, 1H), 6.71 (d, J=8.1 Hz, 2H), 3.73 (dd, J=10.7,
The liposome was prepared based on our reported method
[49]. In brief, soy lecithin (80 mg), cholesterol (10 mg),
DSPE-PEG2K-FA (10 mg, Figure S1, Supporting Informa-
tion online), and prodrug (2 mg) were dissolved in chloro-
form (20 mL). The preparation process of liposome was
referred to Figure 2(a). And, the encapsulation efficiency
was measured by the following equation:
4
.0 Hz, 4H), 3.63 (dd, J=10.6, 4.0 Hz, 4H).
+
+
ESI-MS: m/z calcd. for C H Cl N [M+H] 218.04,
10
14
2
found, 218.07.
2
.2.2 Synthesis of compound 4 and 5
Compound 4 and 5 were synthesized according to our pre-
vious literature reports [48].
Encapsulation efficiency (%)=C /C ×100
(1)
m
0
where, C was the concentration of prodrug NIR-NM cal-
m
2
.2.3 Synthesis and characterization of prodrug NIR-NM
culated by the standard work curve in method; C represents
0
Compound 6 (87 mg, 0.3 mM) was dissolved in 1.5 mL HCl/
THF (v/v, 1/1). The mixture was stirred at 0°C for 0.5 h.
After that, NaNO (24 mg, 0.35 mM) dissolved in H O
the original concentration of prodrug NIR-NM.
2
2
2
.4 Determination of the quantum yield
(
0.25 mL) was added into the mixture through drop by drop.
Then, the reaction continued for 2 h. Following, compound 2
87 mg, 0.4 mM) dissolved in THF (0.25 mL) was added
The Φ of prodrug NIR-NM were measured by absolute PL
f
(
quantum yield spectrometer (HAMAMATSU, Japan). And,
they were calculated using the following equation:
slowly to the reaction solution. The mixture was stirred for
overnight, and the solvent was evaporated under reduced
Φ =PN /PN
abs
(2)
f
em
Figure 1 Synthetic procedure of prodrug NIR-NM (a) and activation process (b) (color online).

502
Li et al. Sci China Chem March (2021) Vol.64 No.3
Figure 2 Preparation and characterization of nanoparticle FA-lip@NIR-NM. (a) The schematic diagram of preparation process. DLS (b), TEM (c), and
stability (d) charaterization of FA-lip@NIR-NM (color online).
where, PNem represented the number of photons emitted of
molecules; PNabs represented the number of photons ab-
sorbed of molecules.
MTT method. Cells were cultured in medium with 0, 1, 2, 3,
4, 5, 8, 10, 15 and 20 μM of prodrug NIR-NM for 24 h under
a normal (21%) or hypoxia (2%) environment. Cells in
culture medium without prodrug NIR-NM were used as the
control, respectively. Six replicate wells were adopted for
each control and different experimental concentrations.
Next, MTT prepared in PBS was added to each well and the
plates were incubated at 37 °C for another 4 h in humidified
2
.5 Cell incubation and imaging
Hepatoma carcinoma cells (HepG-2 cells) and human cer-
vical carcinoma cell (HeLa cells) were incubated with Dul-
becco’s Modified Eagle’s medium (DMEM, Invitrogen).
Normal liver cells (LO2 cells) were grown in RPMI medium
incubator containing 95% air and 5% CO . The medium was
2
then carefully removed, and the purple crystals were lysed in
200 μL dimethyl sulfoxide (DMSO). Optical density (OD) of
solution was measured on a microplate reader (Thermo
Fisher Scientific) at 490 nm. Cell viability was expressed as
a percent of the control culture value, and it was calculated
using the following equation:
1
640. Both mediums were supplemented with 10% fetal
bovine serum (FBS, Invitrogen). HeLa cells were seeded in
confocal culture dishes and then incubated for 24 h at 37 °C
under a humidified environment containing 95% air and 5%
CO . Cells were treated with prodrug NIR-NM at 37 °C
2
under a normal (21%) or hypoxia (2%) environment, fol-
lowed by washing thrice using DMEM without FBS. Under
the confocal fluorescence microscope (Olympus FV3000,
Japan) with a 60× objective oil lens, prodrug NIR-NM was
excited at 488 nm and fluorescence emission at 650–720 nm
was gathered.
Cells viability (%)=(ODdye−ODblank)/
(ODcontrol−ODblank)×100
(3)
2
.7 In vivo tumor imaging
Female nude BABL/c mice (15–20 g) were purchased from
Dalian Medical University. All procedures were performed
in compliance with the guide for the care and use of la-
boratory animal resources and Dalian Medical University
animal care and use committee, and were approved by the
2
.6 MTT assays
Measurement of cell viability was performed by the standard

Li et al. Sci China Chem March (2021) Vol.64 No.3
503
institutional animal care and use committee of the Dalian
Medical University. To establish mice tumor model, the
HeLa and HepG-2 cells were chosen to transplant under the
armpit of approximately 15–20 g female nude BABL/c mice.
After about two week’s inoculation, the xenograft tumor
mice were given with prodrug NIR-NM within the period of
mice anesthesia. After that, in vivo imaging was performed
on a NightOWL II LB983 small animal imaging system,
using a 475 nm excitation and a 665 nm emission filter.
blood circulation time, was utilized to encapsulate prodrug
NIR-NM to form the nanoparticle FA-lip@NIR-NM. As
observed in Figure 2(a), we adopted multicomponent ele-
ments, including soy lecithin, cholesterol, and DSPE-
PEG2K-FA to build a robust delivery system, because it was
more stable in vivo than liposome alone. The encapsulation
efficiency of LDS loading NIR-NM was calculated to be
49.35%±2.01% by standard working curve (Figures S2 and
S3). And, the size of nanoparticle FA-lip@NIR-NM was
measured as 170.3 nm with the polydispersity index (PDI) of
0
.149 by DLS analysis (Figure 2(b)), assuring good per-
3
Results and discussion
meation and retention effect for increasing prodrug enrich-
ment at tumor sites in vivo. TEM imaging demonstrated the
morphology of nanoparticle FA-lip@NIR-NM dissolved in
aqueous solution (Figure 2(c)). Additionally, the size of nano
FA-lip@NIR-NM remained stable for 22 days (Figure 2(d)),
enabling nanoparticles to circulate in vivo for long periods of
time.
3
.1 Design of prodrug NIR-NM and construction of
FA-lip@NIR-NM
Molecular engineering strategy was carried out to construct
the smart prodrug with good biosafety for normal cells and
tissues. Nitrogen mustard, an antitumor drug used clinically
with outstanding efficacy, was chosen to install on NIR
fluorophore through the activatable “azo bond” to develop
the molecule NIR-NM, which was tactfully integrated with
diagnostic imaging and synergistic treatment. The synthetic
route of prodrug NIR-NM was illustrated in Figure 1(a), and
its characterization was fully verified by H NMR, C NMR,
and ESI-HRMS in the supporting information online (Figure
S7–S9). This engineered framework endowed the prodrug
NIR-NM non-fluorescent and non-toxicity, because the
presence of azo-bond made the energy dissipation in the
excited state easier and the formation of highly reactive
ethylene imine ions (high toxicity) more difficult. Upon re-
duction of azo bond by azo reductase overexpressed in the
solid tumor, compound 6 with the strong electron-donating
group was generated, achieving the recovery of NIR fluor-
escence, while the active chemotherapeutic drug (compound
3
.2 Optical properties
With prodrug NIR-NM in hand, its photo-physical properties
were investigated in various solvents. In the absorption
spectra, the free prodrug NIR-NM had relatively strong UV-
Vis absorption in the range of 460–490 nm (Figure 3(a)).
However, there was no obvious fluorescence peak even in
organic solvents (Figure 3(b)), which could be ascribed to the
conformational changes swiftly of the azo bond. In addition,
1
13
the absolute Φ values (below 0.7%, Table 1) in different
f
solvents were further verified that the fluorescence of pro-
drug NIR-NM was quenched thoroughly.
To validate prodrug NIR-NM activated under hypoxic
condition, sodium dithionite, a chemical mimic of azo re-
ductase, was selected for in vitro spectroscopy [15]. As de-
picted in Figure 4(b), after adding sodium dithionite in PBS
buffer (10 mM, pH 7.4), the emission peak of prodrug NIR-
NM increased steadily, which was due to the cleavage of azo
7
, Figure 1(b)) was released in situ.
To enhance tumor uptake and retention, liposome delivery
system (LDS) as a commonly used nanocarrier with long
Figure 3 Absorbance (a) and fluorescence spectra (b) of prodrug NIR-NM (10 μM) in various solvents (DCM, EA, ACN, MeOH, EtOH, DMF, DMSO, and
PBS) (color online).

504
Li et al. Sci China Chem March (2021) Vol.64 No.3
Table 1 Spectra data of prodrug NIR-NM in different solvents
including ions, reactive oxygen/sulfur species, amino acids
(Figure 4(d)), clearly demonstrating that drugs (nitrogen
mustard) did not leak easily under normal circumstances, as
expected. Prodrug NIR-NM displayed negligible pH effect
a)
b)
Solvent
DCM
EA
λabs (nm)
468
λem (nm)
649
ε
Φf (%)
0.52
0.46
0.42
0.47
0.44
0.41
0.42
0.35
0.4
465
617
0.6
(4.75–10.27, Figure S4), showing good ability to resist mi-
ACN
MeOH
EtOH
DMF
DMSO
PBS
464
620
0.7
croenvironment interference.
469
619
0.4
457
622
0.3
3
.3 Mechanism verification
483
595
0.3
493
620
0.5
To verify the mentioned mechanism, Time Dependent Den-
sity Functional Theory (TDDFT) calculations based on
B3LYP/6-31G* level using Gaussian 16 programs were
performed. In prodrug NIR-NM at the optimized excitation
state, the apparent difference of highest occupied molecular
orbital (HOMO) and lowest unoccupied molecular orbital
(LUMO) electron distribution was found (Figure 5(a)). No-
tably, compared to the LUMO level, the π electrons at
HOMO level were mainly concentrated in the azo bond of
the prodrug NIR-NM. And, the oscillator strength value of 0
signified that the transition between the LUMO and HOMO
level was forbidden. For fluorophore 6, π electrons were laid
out across the entire molecular skeleton at both the LUMO
and HOMO level (Figure 5(b)). The large oscillator strength
460
662
0.3
5
−1
−1
a) Molar extinction coefficient (×10 mol cm L) of prodrug NIR-NM;
b) fluorescence quantum yields of prodrug NIR-NM.
bond between fluorophore and anticancer drug, eventually
resulting in the generation of NIR fluorescence signal and
drug release. In contrast, the fluorescence spectra of the so-
lution with only the prodrug remained essentially unchanged
within 150 min (Figure 4(a)). The variation trends of fluor-
escence intensity at 650 nm of prodrug NIR-NM with and
without sodium dithionite over time were shown in
Figure 4(c). Besides, prodrug NIR-NM had negligible
fluorescence responses towards common biological analytes,
Figure 4 Spectra response of prodrug NIR-NM (10 μM) in the absence (a) and presence (b) of sodium dithionite in phosphate buffer saline (PBS, 10 mM,
pH 7.4). (c) Fluorescence intensity changes of prodrug NIR-NM at 650 nm in the absence (red dot) and presence (black dot) of sodium dithionite as a function
of time. (d) The changes of fluorescence intensity at 650 nm of prodrug NIR-NM in the presence of various biological analytes. Insert (except for special
+
+
2+
2+
2+
+
−
−
−
−
−
−
instructions, the concentration is 300 µM) 1, blank; 2, Na ; 3, K ; 4, Ca ; 5, Ni ; 6, Mg ; 7, NH ; 8, F ; 9, Cl ; 10, Br ; 11, I ; 12, CH COO ; 13, HCO ;
1
4
3
3
2−
2−
−
−
2−
−
−
4, CO3 ; 15, S ; 16, H PO ; 17, NO ; 18, SO ; 19, SCN ; 20, NO ; 21, GSH; 22, Cys; 23, Hcy; 24, Ascorbic acid; 25, NO; 26: NaClO (100 µM); 27,
2 4 3 4 2
H O (100 µM); 28, Na S O . Error bar=±SD (n=3) (color online).
2
2
2
2
4

Li et al. Sci China Chem March (2021) Vol.64 No.3
505
Figure 5 TDDFT analysis of prodrug NIR-NM (a) and fluorophore 6 (b). (c, d) High resolution mass spectra (HRMS) analysis of prodrug NIR-NM towards
sodium dithionite (color online).
and small ∆E induced the generation of obvious NIR
emission. Besides, the HRMS analysis of the reaction be-
tween prodrug NIR-NM and sodium dithionite in PBS buffer
was also carried out. The apparent peaks (m/z) of 290.1657
and 289.1579 were observed in Figure 5(c), which was
q)), which was ascribed to the overexpression of azo re-
ductase in anaerobic cells, activating more prodrug mole-
cules. As a result, we could visualize drug release of NIR-
NM in-situ in cancer cells. Free nitrogen mustard quickly
transformed into highly toxic imine ions in cells, which then
covalently combined with guanine nitrogen to generate the
cross connection within the double strand of DNA or the
cross connection of different bases within the same strand of
DNA and led to cell apoptosis [50], as illustrated in
Figure 6(s).
+
+
consistent with the [M+H] (m/z=290.1652) and [M]
m/z=289.1579) of compound 6. In addition, the obvious
(
+
peaks (m/z) of 232.0534 represented the compound 7 ([M] ,
m/z=232.0534, Figure 5(d)). These data clearly verified the
supposed recognition mechanism of prodrug NIR-NM.
Subsequently, the treatment efficacy of prodrug NIR-NM
was evaluated at the cellular level by MTT assays. Under
normal oxygen state, the antitumor cells effect (HepG-2
cells, survival rate 77.7%) was more than normal cells (LO2
cells, survival rate 95.1%) with increasing dose (0–20 μM)
of prodrug NIR-NM (Figure 6(t)). Additionally, the survival
rate of cancer cells was significantly reduced to be 32.8%
incubated with prodrug NIR-NM under anaerobic circum-
stance, compared with 84.6% of normal cells (Figure 6(u)),
which indicated that the prodrug NIR-NM had the ability to
3
.4 Cell imaging and MTT assays
Encouraged by the above results, next, to monitor the drug
release of prodrug NIR-NM in cell, confocal fluorescence
imaging of HeLa cells was performed under normal oxygen
(21% O ) and anaerobic (2% O ) circumstances, respec-
2
2
tively. As shown in Figure 6, compared to the results of cell
imaging under normal oxygen condition (Figure 6(e, h)), the
remarkable fluorescence signal was gathered (Figure 6(n,

506
Li et al. Sci China Chem March (2021) Vol.64 No.3
Figure 6 Fluorescence confocal imaging of HeLa cells under normal (21% O , a–i) and anaerobic (2% O , j–r) condition, respectively. (s) A diagram of
2
2
prodrug NIR-NM activated under anaerobic condition to kill tumor cells. Cell viability of HepG-2 and LO2 cells treated with various concentration (0, 1, 2, 3,
4
, 5, 8, 10, 15, and 20 μM) of prodrug NIR-NM under normal (21% O , t) and anaerobic (2% O , u) condition, respectively. Error bar=±SD (n=6) and scale
2
2
bar: 20 μm (color online).
selectively kill cancer cells, greatly reducing side effects.
.5 In vivo imaging for tracking drug release
3
To study the feasibility of prodrug NIR-NM in vivo, the
images of tumor-bearing BABL/c mice were firstly mon-
itored by intratumor injection. Notably, the fluorescence in-
tensity at the tumor site gradually increased over time
(Figure 7(a1–d1)). In contrast, there was basically no fluor-
escence in normal tissue (Figure 7(a2–d2)) treated the
equivalent prodrug NIR-NM via subcutaneous injection.
These results further demonstrated that prodrug NIR-NM
could be activated efficiently by the over-expressed azo re-
ductase in tumor-microenvironment. That means, the fluor-
escence intensity could indicate how much drug molecules
are released real-time in tumors.
Besides, to enhance tumor uptake and retention, the
commonly used liposome drug delivery system with long
blood circulation time was adopted as the mixed matrix to
encapsulate prodrug NIR-NM. The nanosized FA-lip@NIR-
NM of 170.3 nm (PDI=0.149, Figure 2(b)) was readily
constructed referred to Figure 2(a), which was conducive for
the enrichment of prodrug NIR-NM at the tumor region
owing to EPR effect. To confirm the accumulation effect of
nanoparticles prodrug NIR-NM was used as control. As
Figure 7 Time dependent in vivo fluorescence imaging of tumor-bearing
BABL/c mice before (a) and after (b–d) the intratumor injection (1) and
subcutaneous injection (2) of equivalent prodrug NIR-NM (color online).
demonstrated in Figure 8(f), the tumor location of the mice
treated with FA-lip@NIR-NM by the intravenous injection
showed stronger fluorescence signal than that of injected
with NIR-NM (Figure 8(a)), indicative of more nanoparticle
prodrug enrichment and cleavage of azo bond in tumor site,
further guaranteeing stronger treatment effect. Intriguingly,
the obvious fluorescence signal was still observed after 48 h
(Figure 8(j)), assuring the long-circulation capacity of FA-
lip@NIR-NM in vivo. Whereas, there was no significant
enrichment of the drug from the NIR-NM treatment group
(Figure 8(a–e)). Based on the above results, the nanoparticle
FA-lip@NIR-NM had been shown to be an effective tool for

Li et al. Sci China Chem March (2021) Vol.64 No.3
Res, 2013, 36: 1302–1310
507
6
Yue J, Liu S, Wang R, Hu X, Xie Z, Huang Y, Jing X. Mol Pharm,
2012, 9: 1919–1931
7
Li S, Zhang Y, Wang J, Zhao Y, Ji T, Zhao X, Ding Y, Zhao X, Zhao
R, Li F, Yang X, Liu S, Liu Z, Lai J, Whittaker AK, Anderson GJ, Wei
J, Nie G. Nat Biomed Eng, 2017, 1: 667–679
8
9
0
Xue T, Jia X, Wang J, Xiang J, Wang W, Du J, He Y. Chem Eur J,
2019, 25: 9634–9638
Walsh JC, Lebedev A, Aten E, Madsen K, Marciano L, Kolb HC.
Antioxid Redox Sign, 2014, 21: 1516–1554
Apte S, T. Chin F, E. Graves E. Curr Org Synth, 2011, 8: 593–603
Brown JM, Wilson WR. Nat Rev Cancer, 2004, 4: 437–447
1
11
Figure 8 Long-time tracking drug release in tumor-bearing BABL/c mice
4, 6, 8, 24, and 48 h) after intravenous injection of NIR-NM (a–e) and FA-
lip@NIR-NM (f–j). The dotted red line indicates the location of the tumor
color online).
(
12 Huang J, Wu Y, Zeng F, Wu S. Theranostics, 2019, 9: 7313–7324
13 Yoon J, Li H, Kim D, Yao Q, Ge H, Chung J, Fan J, Wang J, Peng X.
Angew Chem Int Ed, 2020, anie.202009796
(
14
He H, Du L, Tan M, Chen Y, Lu L, An Y, Wang Y, Li X, Li B, Shen J,
Wu J, Shuai X. Sci China Chem, 2020, 63: 936–945
1
5
Verwilst P, Han J, Lee J, Mun S, Kang HG, Kim JS. Biomaterials,
monitoring drug release in real time in situ and improving
treatment outcomes.
2017, 115: 104–114
1
1
1
6
7
8
Liao D, Johnson RS. Cancer Metastasis Rev, 2007, 26: 281–290
Vaupel P. Oncol, 2008, 13: 21–26
Chang Q, Jurisica I, Do T, Hedley DW. Cancer Res, 2011, 71: 3110–
4
Conclusions
3120
1
2
2
9
0
1
Yan C, Guo Z, Shen Y, Chen Y, Tian H, Zhu WH. Chem Sci, 2018, 9:
4959–4969
Sharma A, Lee MG, Won M, Koo S, Arambula JF, Sessler JL, Chi SG,
Kim JS. J Am Chem Soc, 2019, 141: 15611–15618
Xue T, Shen J, Shao K, Wang W, Wu B, He Y. Chem Eur J, 2020, 26:
In summary, using molecular engineering strategy, we have
constructed a smart prodrug NIR-NM by integration of ni-
trogen mustard into an NIR fluorophore. Prodrug NIR-NM
could be activated by overexpressed azo reductase in anae-
2521–2528
robic cells (2% O ), visually releasing the active anticancer
22 Li Q, Cao J, Wang Q, Zhang J, Zhu S, Guo Z, Zhu WH. J Mater Chem
B, 2019, 7: 1503–1509
2
drugs, which displayed the ability to selectively kill tumor
cells, greatly reducing side effects. The supposed recognition
mechanism was confirmed by HRMS analysis. Notably, the
nanoparticle FA-lip@NIR-NM with high encapsulation ef-
ficiency and stability, demonstrated the advantages of tumor
enrichment and long circulation in vivo owing to the EPR
effect, which was used visually long-time to track the level of
drug release and improve the therapeutic effect. Thereof, we
believe that the nanoparticle FA-lip@NIR-NM could be an
ideal platform for prodrug delivery to provide strong tech-
nical support for cancer treatment in the future.
2
3
Denmeade SR, Jakobsen CM, Janssen S, Khan SR, Garrett ES, Lilja
H, Christensen SB, Isaacs JT. J Natl Cancer Institute, 2003, 95: 990–
1000
24 Denmeade SR, Mhaka AM, Rosen DM, Brennen WN, Dalrymple S,
Dach I, Olesen C, Gurel B, DeMarzo AM, Wilding G, Carducci MA,
Dionne CA, Moller JV, Nissen P, Christensen SB, Isaacs JT. Sci
Translational Med, 2012, 4: 140ra86
25
26
27
Liu HW, Hu XX, Li K, Liu Y, Rong Q, Zhu L, Yuan L, Qu FL, Zhang
XB, Tan W. Chem Sci, 2017, 8: 7689–7695
Pennacchietti S, Michieli P, Galluzzo M, Mazzone M, Giordano S,
Comoglio PM. Cancer Cell, 2003, 3: 347–361
Ballinger JR. Seminars Nucl Med, 2001, 31: 321–329
28 Vaupel P, Schlenger K, Knoop C, Höckel M. Cancer Res, 1991, 51:
316–3322
3
29
30
31
32
33
34
35
Yang Z, Cao J, He Y, Yang JH, Kim T, Peng X, Kim JS. Chem Soc
Rev, 2014, 43: 4563–4601
Silvers WC, Prasai B, Burk DH, Brown ML, McCarley RL. J Am
Chem Soc, 2013, 135: 309–314
Kwon N, Cho MK, Park SJ, Kim D, Nam SJ, Cui L, Kim HM, Yoon J.
Chem Commun, 2017, 53: 525–528
Peng X, Gao J, Yuan Y, Liu H, Lei W, Li S, Zhang J, Wang S.
Bioconjugate Chem, 2019, 30: 2828–2843
Acknowledgements This work was supported by the National Creative
Research Initiative programs of the National Research Foundation of Korea
(NRF), the Korean Government (MSIP) (2012R1A3A2048814), the Na-
tional Natural Science Foundation of China (21421005, 21808028) and the
Natural Science Foundation of Liaoning United Fund (U1608222,
U1908202).
Conflict of interest The authors declare no conflict of interest.
Li Y, Sun Y, Li J, Su Q, Yuan W, Dai Y, Han C, Wang Q, Feng W, Li
F. J Am Chem Soc, 2015, 137: 6407–6416
Xu F, Li H, Yao Q, Ge H, Fan J, Sun W, Wang J, Peng X. Chem Sci,
Supporting information The supporting information is available online at
http://chem.scichina.com and http://link.springer.com/journal/11426. The
supporting materials are published as submitted, without typesetting or
editing. The responsibility for scientific accuracy and content remains en-
tirely with the authors.
2019, 10: 10586–10594
Zhou Y, Maiti M, Sharma A, Won M, Yu L, Miao LX, Shin J, Podder
A, Bobba KN, Han J, Bhuniya S, Kim JS. J Control Release, 2018,
288: 14–22
36
Ding N, Li Z, Tian X, Zhang J, Guo K, Wang P. Chem Commun, 2019,
1
2
3
4
Boulikas T, Vougiouka M. Oncol Rep, 2004, 11: 559–595
Eckford PDW, Sharom FJ. Chem Rev, 2009, 109: 2989–3011
Oun R, Moussa YE, Wheate NJ. Dalton Trans, 2018, 47: 6645–6653
Huang CY, Ju DT, Chang CF, Reddy PM, Velmurugan BK. BioMe-
dicine, 2017, 7: 12–23
55: 13172–13175
37 Biswas S, Rajesh Y, Barman S, Bera M, Paul A, Mandal M, Pradeep
Singh ND. Chem Commun, 2018, 54: 7940–7943
38 Kumar R, Kim EJ, Han J, Lee H, Shin WS, Kim HM, Bhuniya S, Kim
JS, Hong KS. Biomaterials, 2016, 104: 119–128
5
Yang QZ, Wang C, Lang L, Zhou Y, Wang H, Shang DJ. Arch Pharm
39 Zhao X, Ha W, Gao K, Shi Y. Anal Chem, 2020, 92: 9039–9047

5
08
Li et al. Sci China Chem March (2021) Vol.64 No.3
Feng W, Gao C, Liu W, Ren H, Wang C, Ge K, Li S, Zhou G, Li H,
Du J, Li Y, Wang J, Yoon J, Peng X. J Am Chem Soc, 2020, 142:
4
4
4
4
4
4
0
Wang S, Jia G, Li Z, Zhang J. Chem Commun, 2016, 52: 9434–9437
Li B, Liu P, Yan D, Zeng F, Wu S. J Mater Chem B, 2017, 5: 2635–
6381–6389
1
2
3
4
5
46 Li H, Yao Q, Xu F, Xu N, Duan R, Long S, Fan J, Du J, Wang J, Peng
X. Biomaterials, 2018, 179: 1–14
2643
Jangili P, Won M, Kim SJ, Chun J, Shim I, Kang C, Ren WX, Kim JS.
ACS Appl Bio Mater, 2019, 2: 3532–3539
Lindahl LM, Fenger-Gron M, Iversen L. J Eur Acad Dermatol Ve-
nereol, 2013, 27: 163–168
47 Li H, Li Y, Yao Q, Fan J, Sun W, Long S, Shao K, Du J, Wang J, Peng
X. Chem Sci, 2019, 10: 1619–1625
48 Li H, Yao Q, Xu F, Xu N, Sun W, Long S, Du J, Fan J, Wang J, Peng
X. Front Chem, 2018, 6: 485
49 Shi C, Li M, Zhang Z, Yao Q, Shao K, Xu F, Xu N, Li H, Fan J, Sun
W, Du J, Long S, Wang J, Peng X. Biomaterials, 2020, 233: 119755
50 Kohn KW, Spears CL, Doty P. J Mol Biol, 1966, 19: 266–288
Kim YH, Martinez G, Varghese A, Hoppe RT. Arch Dermatol, 2003,
139: 165–173
Li H, Yao Q, Sun W, Shao K, Lu Y, Chung J, Kim D, Fan J, Long S,