Metal‐Coordinated Adsorption of Nanoparticles to Macrophages for Targeted Cancer Therapy

Abstract

Living cell‐based drug delivery systems (LC‐DDSs) are limited by adverse interactions between drugs and carrier cells, typically drug‐induced toxicity to carrier cells and restriction of carrier cells on drug release. Here, a method is established to adsorb nanocarriers externally to living cells, thereby reducing cytotoxicity caused by drug uptake and realizing improved drug release at the disease site. It is found that a divalent metal ion‐phenolic network (MPN) affords adhesion of poly (lactic‐co‐glycolic acid) nanoparticles onto macrophage (Mφ) surfaces with minimized intracellular uptake and no negative effect on cell proliferation. On this basis, an Mφ‐DDS with doxorubicin‐loaded nanoparticles on cell surface (DOX‐NP@Mφ) is constructed. Compared to intracellular loading via endocytosis, this method well‐maintains bioactivity (viability and migration chemotaxis) of the carrier cell. By virtue of the photothermal effect of MPN at the tumor site, DOX‐NP‐associated vesicles are liberated for improved chemotherapy. This facile, benign, and efficient method (ice bath, 2 min) for extracellular nanoparticle attachment and minimizing intracellular uptake provides a platform technology for LC‐DDS development.

Full text

www.afm-journal.de
Vol. 33 • No. 19 • May 8 • 2023
adfm202370115_OFC_eonly.indd 1 18/04/23 12:35 PM

www.afm-journal.de
©  Wiley-VCH GmbH
2214842 (1 of 13)
Metal-Coordinated Adsorption of Nanoparticles to
Macrophages for Targeted Cancer Therapy
Mao-Hua Zhu, Xin-Di Zhu, Mei Long, Xing Lai, Yihang Yuan, Yanhu Huang,
Lele Zhang, Yuhao Gao, Jiangpei Shi, Qin Lu, Peng Sun, Jonathan F. Lovell,
Hong-Zhuan Chen, and Chao Fang*
Living cell-based drug delivery systems (LC-DDSs) are limited by adverse
interactions between drugs and carrier cells, typically drug-induced toxicity to
carrier cells and restriction of carrier cells on drug release. Here, a method is
established to adsorb nanocarriers externally to living cells, thereby reducing
cytotoxicity caused by drug uptake and realizing improved drug release at
the disease site. It is found that a divalent metal ion-phenolic network (MPN)
aords adhesion of poly (lactic-co-glycolic acid) nanoparticles onto mac-
rophage (Mϕ) surfaces with minimized intracellular uptake and no negative
eect on cell proliferation. On this basis, an Mϕ-DDS with doxorubicin-
loaded nanoparticles on cell surface (DOX-NP@Mϕ) is constructed. Com-
pared to intracellular loading via endocytosis, this method well-maintains
bioactivity (viability and migration chemotaxis) of the carrier cell. By virtue
of the photothermal eect of MPN at the tumor site, DOX-NP-associated
vesicles are liberated for improved chemotherapy. This facile, benign, and
ecient method (ice bath, 2min) for extracellular nanoparticle attachment
and minimizing intracellular uptake provides a platform technology for
LC-DDS development.
DOI: 10.1002/adfm.202214842
macrophages,[2] and natural killer (NK)
cells.[3] Besides, living cells can also be
used as carriers for improved drug delivery.
Living cell-based drug delivery systems (LC-
DDSs) hold potential to overcome some
of the shortcomings of conventional nano-
formulations for therapeutic management
of various diseases.[4,5] Compared to syn-
thetic drug delivery vehicles, living cells are
biocompatible, biomimetic, and safe. With
long circulation, red blood cells (RBCs)
have been investigated as drug carriers
for ≈50 years and Eryaspase (erythrocyte-
encapsulated asparaginase) has entered
Phase III clinical trials.[6,7] Immune cells,
such as macrophages (Mϕ),[8–12] neutro-
phils,[13–15] T cells,[16–18] and natural killer
(NK) cells [19,20] can respond to signaling
molecules at disease sites to achieve tar-
geted drug delivery and treatment.
LC-DDSs are limited mainly by adverse
interactions between drugs and carrier
cells. One major challenge in immune
cell-based drug delivery is to carry the drug without impacting
the cell function. Endocytosis of drug-loaded nanoparticles by
phagocytic immune cells is a common method for intracel-
lular loading.[9,12,14,15] However, premature leakage of the drug
(especially cytotoxic agents) into the cytoplasm will damage the
ReseaRch aRticle
1. Introduction
Cell therapy can replenish deficient cells through blood trans-
fusions or hematopoietic stem cell transplantation or kill
malignant cells via chimeric antigen receptor (CAR) T cells,[1]
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/./adfm..
M.-H. Zhu, X.-D. Zhu, M. Long, X. Lai, Y. Yuan, Y. Huang, L. Zhang,
Y. Gao, J. Shi, Q. Lu, C. Fang
Hongqiao International Institute of Medicine
Tongren Hospital and State Key Laboratory of Oncogenes
and Related Genes
Department of Pharmacology and Chemical Biology
Shanghai Jiao Tong University School of Medicine (SJTU-SM)
Shanghai , China
E-mail: fangchao@sjtu.edu.cn
X.-D. Zhu
Department of Pharmacy
Shanghai Ninth People’s Hospital, SJTU-SM
Shanghai , China
P. Sun
Department of General Surgery
Tongren Hospital, SJTU-SM
Shanghai , China
J. F. Lovell
Department of Biomedical Engineering
University at Bualo
State University of New York
Bualo, NY , USA
H.-Z. Chen
Institute of Interdisciplinary Integrative Biomedical Research
Shuguang Hospital
Shanghai University of Traditional Chinese Medicine
Shanghai , China
C. Fang
Key Laboratory of Basic Pharmacology of Ministry of Education & Joint
International Research Laboratory of Ethnomedicine of Ministry of
Education
Zunyi Medical University
Zunyi , China
Adv. Funct. Mater. 2023, 33, 

www.afm-journal.dewww.advancedsciencenews.com
2214842 (2 of 13) ©  Wiley-VCH GmbH
bioactivity and chemotaxis of carrier cells.[11,21] Therefore, to
avoid substantial loss of cell function, carrier cells usually need
to be infused back shortly (8–12 h) after drug loading,[14] which
brings logistical diculties to their application. In turn, carrier
cells may not readily release drugs or alternatively could inacti-
vate drugs within intracellular microenvironments such as lys-
osomes,[8,11,21] leading to a diminished therapeutic ecacy. In
contrast, attaching drug-loaded nanoparticles or backpacks onto
cell surfaces has the potential advantages of less cell damage and
easier drug release at the disease site. Extracellular loading has
mainly been achieved through chemical conjugation,[16,17,20,22]
physical adhesion,[8,23] or antibody-receptor recognition.[24,25]
However, these methods involve generally complex chemical
and engineering procedures, and in some cases, nanoparticles
are still internalized by cells in a short time after they attach to
cell surface.[20,25] Therefore, there is a need for more facile and
benign extracellular LC-DDS loading methods, which can mini-
mize nanoparticle uptake to obtain an extended bioactivity main-
taining of carrier cells for improved therapeutic application.
Metal-phenolic networks (MPNs) are an emerging class of
supramolecular structures formed by metal ions coordinated
to phenolic ligands, typically tannic acid (TA), a food additive
recognized as safe by the United States Food and Drug Adminis-
tration (FDA).[26] By virtue of the adherent properties of phenolic
molecules, MPNs were recently demonstrated for the attach-
ment of nanoparticles on the surfaces of living cells.[27] Never-
theless, whether dierent metal ions would aect MPN-medi-
ated nanoparticle adsorption and how long surface adhesion
can be maintained remains to be determined. Elucidating these
issues is important for the generation of ecient LC-DDSs.
Here, we reveal that compared to trivalent ions (Fe3+ and
Cr3+), divalent ions (Cu2+, Mn2+, Zn2+, and Co2+)-contained MPN-
endowed nanoparticles with much better surface adhesion and
minimized cellular uptake over 10d. Based on this finding, we
developed an Mϕ-DDS with doxorubicin (DOX)-loaded polymeric
nanoparticles (NPs) on cell surface, named as DOX-NP@Mϕ. In
this study, macrophages were used as the model cells to evaluate
the advantage of metal-coordinated adsorption of nanoparti-
cles over the common method of intracellular loading through
endocytosis. Compared to traditional intracellular loading via
endocytosis, this method maintained the bioactivity (viability
and migration chemotaxis) of the carrier cell (Scheme 1A). By
virtue of the photothermal eect of MPN, liberation of DOX-NP-
associated vesicles was achieved for improved tumor cell uptake
and chemotherapy (Scheme 1B). Moreover, the construction of
this Mϕ-DDS is facile, benign, and ecient (ice bath, 2 min).
These integrated merits together exhibit a promising potential of
this platform technology for translational application in cancer
clinics. DOX-NP@Mϕ was optimally fabricated and its superior
performance in vitro and in vivo was demonstrated.
2. Results and Discussion
2.1. Metal Ion Screening and Construction of Macrophages with
Surface-Adsorbed Nanoparticles
Murine Mϕ cells (RAW264.7) with surface-adsorbed PLGA
nanoparticles were constructed as illustrated in Scheme1A and
Figure S1A (Supporting Information). As the essential trace
elements in humans,[28] six metal ions (Cr3+, Mn2+, Fe3+, Co2+,
Cu2+, and Zn2+) were investigated for their eects on nanopar-
ticle adhesion on cell surface. For visual distinction, coumarin
6-labeled PLGA nanoparticles (coumarin 6 loading of 0.64%)
were set to dierent pseudo-colors according to the apparent
colors of the metal salts and their aqueous solutions (Figure 1A;
FiguresS2 and S3, Supporting Information). Compared to an
average of ≈25% nanoparticle internalization (based on the
≈75% proportion of extracellular fluorescent intensity) in the
use of Cr3+ and Fe3+, MPNs comprising Mn2+, Co2+, Cu2+, or
Zn2+ conferred a better surface adhesion with less than ≈10%
nanoparticle uptake inside the cells (Figure 1B). A 3D recon-
struction of confocal images further confirmed this high degree
of surface adherence (Figure S4 and Movie S1, Supporting
Information). This observation may be ascribed to the number
of coordination bonds formed between metal ions and tannic
acid (TA). Divalent ions commonly form 4 coordination bonds
with ligands, whereas trivalent ions (Fe3+and Cr3+) form six.
The latter would result in more occupation of phenolic hydroxyl
groups on TA and thus less was left for the TA-mediated adhe-
sion between PLGA nanoparticles and Mϕ surface. Also of
note, low temperature (ice bath), a condition that blocks active
uptake, was necessary for this high degree of surface adhe-
sion; in contrast, a considerable number of nanoparticles was
internalized when this experiment was performed at room
temperature (FigureS5, Supporting Information). A proposed
schematic image showing the interactions between metal-
coordinated nanoparticles and macrophage surface was shown
in Figure S1B (Supporting Information). The interactions
included the metal ion-mediated coordination and the TA-medi-
ated ionic interactions, hydrogen bonding, and hydrophobic
interactions with the amino acids of the membrane proteins.[29]
The underlying mechanism responsible for the reduced non-
specific uptake by Mϕ needs to be investigated further. As
elemental copper is also included in Vyxeos (an FDA-approved
liposomal formulation for leukemia treatment),[30] in following
study, Cu2+-phenolic network (CuPN) was used for the Mϕ-
DDS construction.
Ultra-high resolution scanning electron microscopy (SEM)
images visually confirmed the CuPN-mediated nanoparticle
attachment on Mϕ surface, in contrast to the intact cells
(Figure 1C; Figure S6, Supporting Information). It is noted
that the cell surface was only partially covered by the adhered
nanoparticles (Figure1C; FiguresS4, S6, and MovieS1, Sup-
porting Information), ensuring the maintenance of cell bio-
activity. The uneven distribution of the nanoparticles on Mϕ
surface may be caused by the inadequate mixing of mac-
rophages and NP@MPN. The short mixing duration (2min)
would also aect the even adsorption on cell surface. CuPN-
mediated nanoparticle (coumarin 6 labeled) surface adhesion
was well maintained even after 10 d in culture (Figure 1D;
Figure S7, Supporting Information). To our knowledge, this
is the longest surface adhesion duration ever reported. Such a
long stability would enable sucient time from cytopharma-
ceutical preparation to bedside use in eventual clinical appli-
cations. Of note, the surface fluorescence intensity declined
on days 6 and 10, which may be partially ascribed to the
proliferative cell division (Figure 1E; Figure S8, Supporting
Adv. Funct. Mater. 2023, 33, 
16163028, 2023, 19, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202214842 by Shanghai Jiao Tong University, Wiley Online Library on [09/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

www.afm-journal.dewww.advancedsciencenews.com
2214842 (3 of 13) ©  Wiley-VCH GmbH
Information) or nanoparticle shedding due to the disas-
sembly of MPN,[27,31] which led to a reduction in the number
of nanoparticles on the cell surface. Adsorbed nanoparti-
cles were observed on the telophase Mϕ membrane in the
cleavage furrow (Figure S9, Supporting Information). The
influence of surface adhesion on cell viability and prolifera-
tion was real-time monitored using a IncuCyte live cell anal-
ysis system (Figure1E,F). From the 3-day observation before
cell confluency, NP@Mϕ exhibited overlapping growth pro-
file to Mϕ alone at three dierent, original cell concentra-
tions, indicating that both PLGA nanoparticles and CuPNs
were well biocompatible and non-toxic.
In this study, doxorubicin (DOX) for breast cancer therapy
was chosen as the chemotherapeutic agent, which was loaded
in PLGA nanoparticles (DOX-NP) using the emulsification-
solvent evaporation method. DOX-NP (DOX loading of 4.6%)
was spherical as revealed by Cryo-EM (FigureS10, Supporting
Information), and had a size of 124 nm and zeta potential of
−38 mV (Figure 1G). DOX-NP was first coated with an iron
phenolic network (FePN) to confer photothermal eect for
nanoparticle liberation from the carrier cell and then coated
with CuPN for cell surface adhesion. The viabilities of mac-
rophages decorated with Fe nanoparticles (NP@FePN) were
well maintained (Figure S11, Supporting Information), indi-
cating that similar to NP@CuPN (Figure1E,F), NP@FePN was
also non-toxic. After MPN coating, the size of the nanoparticles
(DOX-NP@MPN) increased to 178nm, and zeta potential was
−46 mV (Figure 1G). MPN coating conferred DOX-NP with
UV absorption at ≈300nm and dark color, which were mainly
derived from FePN (FigureS12, Supporting Information). The
metal elements were identified in high-angle annular dark-field
(HAADF) TEM for Fe (Figure S13, Supporting Information)
and in an energy-dispersive X-ray spectroscopy (EDS) assay for
Fe and Cu (FigureS14, Supporting Information), confirming
the MPN coating.
Images of confocal laser scanning microscope (CLSM)
demonstrated that the red fluorescent DOX-NP was adsorbed
on the surface of RAW264.7 cell via MPN-assisted adhesion
Adv. Funct. Mater. 2023, 33, 
Scheme 1. A) Schematic illustration of the construction of macrophages with surface-attached DOX-NP via MPN. High bioactivities are maintained
compared to the conventional intracellular loading via endocytosis. B) The therapeutic performance of DOX-NP@Mϕ in tumors.
16163028, 2023, 19, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202214842 by Shanghai Jiao Tong University, Wiley Online Library on [09/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

www.afm-journal.dewww.advancedsciencenews.com
2214842 (4 of 13) ©  Wiley-VCH GmbH
(Figure 1H; Figure S15, Supporting Information). Compared
to other cell-membrane-attached methods that commonly need
30 min–4 h at 37 °C,[6] it took only 2 min for the completion
of MPN-mediated surface adhesion without other engineering
steps, exhibiting a more facile and ecient advantage.
2.2. Eects of Surface-Attached DOX-NP on Macrophage
Viability and Migration
Next, we investigated whether this surface attachment of DOX-NP
can maintain a higher and extended bioactivity of the carrier
Adv. Funct. Mater. 2023, 33, 
Figure 1. Metal ion screening and construction of macrophages with surface-adsorbed nanoparticles. A) Influence of dierent metal ions on nanopar-
ticle surface adhesion. Coumarin -labeled PLGA nanoparticles were set to dierent pseudo colors according to the apparent colors of the metal salts
and their aqueous solutions. Approximately µg nanoparticles adhered to the surface of ×Mϕ cells. B) Proportions of extracellular fluorescent
intensity in panel (A). C) SEM image of the Mϕ (RAW.) with CuPN-mediated, surface-adsorbed nanoparticles. The cells (pink) and nanoparticles
(green) were painted with pseudo colors. D) Confocal microscopy images of Mϕ with CuPN-mediated, surface-adsorbed coumarin -labeled nano-
particles after , , , and day in vitro culture. E) Eect of CuPN-mediated, surface-adsorbed PLGA nanoparticles on Mϕ proliferation. Images were
photographed using the IncuCyte live cell analysis system. The initial Mϕ number was ×  per well. F) Mϕ proliferation rates in panel (E) were
shown as phase object confluence (%). G) Particle size and zeta potentials (ζ-pot.) of DOX-NP and DOX-NP@MPN were measured through DLS. H)
Fluorescent images of DOX-NP@Mϕ under CLSM. DOX was detected at Ex nm and Em nm. In panels (G) and (H), DOX-NP was first coated
with FePN to confer photothermal eect and then with CuPN for cell surface adhesion. Data are expressed as mean± s.d. n=– in panel (B).
n= in panel (E). n= in panel (G).
16163028, 2023, 19, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202214842 by Shanghai Jiao Tong University, Wiley Online Library on [09/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

www.afm-journal.dewww.advancedsciencenews.com
2214842 (5 of 13) ©  Wiley-VCH GmbH
cells, compared to the conventional method of loading nanoparti-
cles intracellularly via endocytosis. For both intracellular loading
and surface attachment, DOX loading on cells is well propor-
tional to the added drug dosage (Figure 2A). This good linear
correlation can be used for the preparation of Mϕ with specific
DOX loading. Notably, the steeper slope for the surface adhesion
method demonstrated that this strategy can achieve more drug
loading compared to intracellular loading through 3 h cellular
uptake when same amount of drug was added, exhibiting a supe-
rior drug loading capacity. Moreover, the DOX loadings of the
surface attachment method (≈1 µg 10−6 cells) were comparable
to previous reports about living cells with cargo-loaded nanopar-
ticles on their surface.[5,9,17]
We first examined the cell viability after DOX loading. Com-
pared to intracellular loading, MPN-assisted DOX-NP surface
attachment maintained the cell viability at the tested drug
loadings. Specifically, at DOX loading of 0.6µg per 106 cells,
80% of the cells with the surface-attached nanoparticles were
viable after 48h, whereas there were only 20% viable cells left
in the circumstance of intracellular loading (Figure 2B). Lac-
tate dehydrogenase (LDH) examination is another sensitive
measure for cytotoxicity evaluation. LDH levels released into
Adv. Funct. Mater. 2023, 33, 
Figure 2. Comparison of intracellular loading and surface adhesion of DOX-NP on Mϕ viability and migration toward conditioned T culture medium.
A) Linear relation between added DOX and the drug loading on Mϕ. Eects of intracellular and extracellular DOX loading on viabilities B) and LDH
release C) of Mϕ cells. D) Live cells (green) and dead cells (red) identified by calcein-AM/PI staining. E) Mϕ apoptosis detected by flow cytometry.
F) Eect of intracellular and extracellular DOX loading on Mϕ migration toward conditioned T culture medium. G) Quantified migration ratios in
panel (F). Data are expressed as mean ± s.d. n= in panels (A) and (G). n= in panels (B) and (C). **p<., ***p<..
16163028, 2023, 19, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202214842 by Shanghai Jiao Tong University, Wiley Online Library on [09/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

www.afm-journal.dewww.advancedsciencenews.com
2214842 (6 of 13) ©  Wiley-VCH GmbH
the extracellular medium reflect the drug-caused damage to the
cell membrane. Surface attachment of DOX-NP had moderate
eect on cell membranes. In contrast, intracellular DOX-NP
loading resulted in more LDH release at all tested drug load-
ings and in a dose-dependent manner, indicating higher cyto-
toxicity (Figure2C). The decreased damage to the carrier cells
through surface attachment was also confirmed using Live/
Dead dual fluorophore staining (Figure2D) and apoptosis assay
through flow cytometry (Figure 2E). Significantly decreased
dead and apoptotic cells were generated when DOX-NP was
loaded on the cell surface, reflecting the pronounceably protec-
tive eect on cell viability.
A cell migration assay was performed to examine whether
DOX-NP-loaded cells retained the tumor-tropic capability.
Mϕ with surface-attached nanoparticles also formed pseudo-
pods, which are necessary for their locomotion (Figure S16,
Supporting Information).[32] The cells with surface-attached
DOX-NP transmigrated more eciently upon the chemotaxis
of 4T1 cell conditioned medium, compared to those with the
nanoparticle loaded intracellularly (Figure2F,G). This observa-
tion was consistent with the better protection on Mϕ viability
via the surface adhesion (Figure2B–E).
2.3. Light Irradiation-Induced Enhanced Nanoparticle Uptake
and Cytotoxicity in 4T1 Cells
Ecient unloading of the drug at disease sites for high avail-
ability to the targeted cells is an important concern for LC-DDS.
Drug-induced carrier cell death and follow-up release may be
poorly ecient.[6] Drug release via response to disease sig-
nals has been investigated.[14,20,33] However, such endogenous
stimulus-triggered drug release would be confined by the het-
erogenous condition of the disease site. We hypothesized that
photothermal eect may assist collapse the cell-nanoparticle
composites and liberate DOX-NP for improved cellular uptake
and chemotherapy.
FePN, either alone or covered on DOX-NP, aorded an
absorption in the NIR window (the 650–900 nm wavelength
range) (FigureS17, Supporting Information), consistent with a
previous observation.[34] After 10min light irradiation (808nm,
2.0Wcm−2), the temperatures of FePN, DOX-NP@FePN, and
DOX-NP@Mϕ in PBS increased to 50–51 °C (Figure 3A,B).
Consecutively cycled heating of DOX-NP@Mϕ for four times
achieved the highly similar and comparable temperature
change profiles, indicating the photothermal stability and struc-
tural integrity of the MPN component (Figure3C).
To examine the influence of light irradiation on cell-nano-
particle composites, Mϕ with attached NP (coumarin 6 labeled)
was first irradiated (808nm, 2W cm−2) for 10 min. Then, the
cells were observed under CLSM. It showed that Mϕ cells
were a destroyed and considerable number of nanoparticle-
associated vesicles detached (Figure3D). In this scenario, Mϕ
destruction by light irradiation can avoid the possible adverse
eects of their active forms with specific phenotypes on tumor
promotion.[2]
To further identify liberated smaller vesicles or particles,
centrifugation (1000 rpm, 5 min) was performed to remove
cells and large debris, and particle size distribution of the
supernatant was assayed using dynamic light scattering (DLS).
Two groups of particles with dierent sizes were identified
(Figure3E). The particles with size between 100–200nm may
be partially free liberated PLGA nanoparticles, and the ones
with size between 300–700nm would be the smaller nanopar-
ticle-associated vesicles that cannot be observed under CLSM.
The detached nanoparticle-associated vesicles were also identi-
fied using flow cytometry (Figure 3F). In this test, the nano-
particles were labeled with DiD and Mϕ was stained with
Hoechst 33342. Light irradiation led to liberation of DiD-labeled
nanoparticle-associated vesicles or free nanoparticles, and
this eect increased with irradiation time from 10 to 20 min
(Figure3F,G). Accordingly, the percentage of Mϕ with attached
nanoparticles in all particles declined (Figure 3F; Figure S18,
Supporting Information).
We used photothermal, a clinically relevant, external stim-
ulus for nanoparticle liberation and follow-up ecient utiliza-
tion by tumor cells. Without light irradiation, few nanoparticles
were engulfed by tumor cells. In contrast, after previous light
irradiation, high nanoparticle-associated fluorescence in tumor
cells was observed (Figure 3H), indicating the enhanced cel-
lular uptake resulted from the liberated nanoparticle-associated
vesicles or even the free nanoparticles. Flow cytometry assay
further confirmed light irradiation facilitated 8.5-fold increased
uptake of nanoparticles (green) by 4T1 tumor cells (Figure3I).
As an extraneous stimulus, NIR light irradiation is still con-
stricted in clinical settings to local disease sites which may
be dicult to access, especially in deeper regions of body. To
address this issue, interventional methods with optical fiber
and endoscopy might be feasible.[35] As expected, the improved
nanoparticle uptake led to enhanced cytotoxicity to 4T1 tumor
cells at all tested DOX concentrations (Figure3J,K; FigureS19,
Supporting Information).
2.4. Tumor Targeting and Photothermal Heating of
NP@Mϕ in Vivo
We next examined the tumor targeting and biodistribution of
NP@Mϕ in vivo. For imaging observation, Mϕ and surface-
attached PLGA nanoparticles were labeled with DiR and DiD,
respectively, to fabricate a dual-fluorophore labeled NP@Mϕ.
The fluorescent signal of both nanoparticles (DiD labeled)
and carrier cells (DiR labeled) reached the peak 2h after injec-
tion (Figure 4A,B). Thus, 2h after i.v. injection was chosen as
the time point for light irradiation treatment. Similar time-
dependent distribution profiles of the carrier cell and nanoparti-
cles in tumors and other major organs (FigureS20, Supporting
Information) reflected the structural integrity of NP@Mϕ
in vivo, which synchronized the pharmacokinetics of the two
components. The spatiotemporally synchronous distribution
in tumors was also confirmed from the multi-angle (coronal,
sagittal, transaxial, and perspective) monitoring in a Spectrum
CT instrument (Figure4C). Nanoparticles attached on Mϕ dis-
tributed more in tumor sites than PEGylated ones (FigureS21,
Supporting Information), indicating the improved tumor tar-
geting conferred by Mϕ. This superior tumor targeting led to
3.7-fold higher DOX contents in tumors at 2 h after injection
(Figure4D,E). Correspondently, the temperature at the tumor
Adv. Funct. Mater. 2023, 33, 
16163028, 2023, 19, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202214842 by Shanghai Jiao Tong University, Wiley Online Library on [09/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

www.afm-journal.dewww.advancedsciencenews.com
2214842 (7 of 13) ©  Wiley-VCH GmbH
Adv. Funct. Mater. 2023, 33, 
Figure 3. Photothermal-induced heating, enhanced cellular uptake, and cytotoxicity in vitro. For photothermal test in vitro, each group contained FePN of
µgmL−. A) Photothermal heating curve at power density of Wcm−. B) Photothermal images in Eppendorf tubes were captured by thermal imaging
camera (Testo ). C) Photothermal stability of DOX-NP@Mϕ. D) Light irradiation-induced Mϕ destruction and liberation of PLGA nanoparticle-asso-
ciated vesicles observed under CLSM. PLGA nanoparticles were labeled with coumarin . E) DLS assay of the detached particles with size below µm.
F) Flow cytometry assay of NP@Mϕ after light irradiation. Mϕ cells were labeled with Hoechst  and PLGA nanoparticles were labeled with DiD. G) The
percentages of the detached particles (DiD positive) were statistically quantified. H) Influence of light irradiation on T cellular uptake of coumarin-labeled
PLGA NPs. Light irradiation on Mϕ was performed before the co-incubation of Mϕ and T cells. I) Quantified cellular uptake in panel (H) by flow cytometry.
J) Schematic illustration of T-luc cell cytotoxicity test through bioluminescence imaging. High cytotoxicity is reflected in the low bioluminescence signal.
K) Influence of light irradiation on T cytotoxicity. Light irradiation on DOX-NP@Mϕ with the indicated DOX concentrations was performed before the
co-incubation of the two cells. Data are expressed as mean ± s.d., n= in panels (A) and (G). n= in panel (I). n= in panel (K). ***p<..
16163028, 2023, 19, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202214842 by Shanghai Jiao Tong University, Wiley Online Library on [09/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

www.afm-journal.dewww.advancedsciencenews.com
2214842 (8 of 13) ©  Wiley-VCH GmbH
site can be elevated with time to ≈50°C after 5min light irra-
diation (808nm, 2Wcm−2) (Figure4F; FigureS22, Supporting
Information), which was favorable for the nanoparticle libera-
tion from Mϕ and enhanced chemotherapy.
2.5. Antitumor Eect of DOX-NP@Mϕ In Vivo
Based on the in vivo tumor targeting and photothermal heating
of tumor site, we next evaluated the in vivo antitumor eect of
DOX-NP@Mϕ in an orthotopic 4T1 breast cancer model. This
animal test was performed as shown in Figure 5A. When the
tumor grew to ≈50mm3 (7 d after tumor cell inoculation), ani-
mals were randomly allocated into 6 groups, and i.v. injected
with DOX-NP@Mϕ or other controls on days 0, 2, and 4 for
three times. The total DOX dose was 2mgkg−1 when involved,
i.e., for each injection of DOX, the dose was 0.67mgkg−1. Irra-
diation (808nm, 2Wcm−2, 10min) on tumors was performed
2h after each injection. Tumor growth and mouse body weight
were monitored until day 28. DOX-NP@Mϕ eectively delayed
the tumor growth compared to NP@Mϕ (empty)+L (light), free
DOX, and DOX-NP (PEGylated). The strongest antitumor eect
was achieved in the group of DOX-NP@Mϕ+L, exhibiting
≈93% inhibition of tumor volume compared to the saline group
Adv. Funct. Mater. 2023, 33, 
Figure 4. In vivo tumor targeting and photothermal heating. A) PLGA nanoparticles were labeled with DiD, and Mϕ cells were labeled with DiR to
construct the dual-fluorescence labeled NP@Mϕ. Time-dependent, ex vivo imaging of the tumors was performed under IVIS Spectrum/CT system
after i.v. injection of dual-labeled NP@Mϕ (× Mϕ cells). B) Quantified fluorescence signals of DiD and DiR in panel (A). C) Imaging of the tumor
site h after i.v. injection of the dual-labeled NP@Mϕ. Tumor sites from the coronal, sagittal, transaxial, and perspective viewing angles were shown.
D,E) DOX-NP@Mϕ (× Mϕ cells, .mgkg− DOX) was i.v. injected into the mice. After h, DOX fluorescence of the excised tumors was imaged
D) and DOX contents in tumors were determined E). F) Temperature curves of the tumor sites after light irradiation. Data are expressed as mean ±
s.d., n=– in panels (B), (E), and (F). *p<..
16163028, 2023, 19, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202214842 by Shanghai Jiao Tong University, Wiley Online Library on [09/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

www.afm-journal.dewww.advancedsciencenews.com
2214842 (9 of 13) ©  Wiley-VCH GmbH
Adv. Funct. Mater. 2023, 33, 
Figure 5. Antitumor ecacy of DOX-NP@Mϕ in vivo. A) Time schedule of the treatment and related examination. B) Tumor volume curves of each
experimental group. C) Mouse survival curves. Median survivals were noted. D) Mouse body weight. E–H) In a separate study, tumors were excised on
day , photographed E), and weighed F). The tumor of one mouse in the group of DOX-NP@Mϕ+L disappeared. G) Lung metastases were detected
using bioluminescence imaging. The metastasis frequencies are summarized in a heatmap. H) Identification of lung metastasis (red arrow) in H&E
staining of the lung sections. I,J) h after the last injection, tumor tissues were processed for H&E staining I). The necrotic regions were indicated
using the dotty red line. J) The proliferative tumor cells were stained with PCNA antibodies. Data are expressed as means ± s.d. n=- in panels (B),
(C), and (D). n= in panels (F) and (G). ***p<..
16163028, 2023, 19, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202214842 by Shanghai Jiao Tong University, Wiley Online Library on [09/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

www.afm-journal.dewww.advancedsciencenews.com
2214842 (10 of 13) ©  Wiley-VCH GmbH
Adv. Funct. Mater. 2023, 33, 
(Figure 5B,E,F; Figure S23, Supporting Information). The
tumor in one mouse from the group of DOX-NP@Mϕ+L even-
tually disappeared (Figure 5E). Accordingly, DOX-NP@Mϕ+L
treatment significantly extended the mouse median survival to
95d with a longest increase in life span (ILS) of 201.6% com-
pared to NP@Mϕ (empty)+L (36.5d, 15.9%), free DOX (42d,
33.3%), DOX-NP (PEGylated) (43d, 36.5%), and DOX-NP@Mϕ
(50d, 58.7%) (Figure5C; TableS1, Supporting Information).
Lung metastasis is common for 4T1 tumor-bearing mice.
Bioluminescence imaging was used to identify 4T1 tumor
metastasis. On day 28, 2–4 mice in saline and other control
groups had metastasis in the lung (40–80% metastasis fre-
quency), while no mice with lung metastasis were observed
in DOX-NP@Mϕ+L group (Figure5G,H). The suppression of
lung metastasis would be attributed to the stronger inhibition
of the orthotopic tumor. Slight loss of body weight occurred
in the groups of DOX-NP@Mϕ and DOX-NP@Mϕ+L after
3 injections, while the body weight quickly recovered after day
6 (Figure 5D), indicating an acceptable acute toxicity of the
treatment.
Histopathological assay indicated that DOX-NP@Mϕ+L
treatment resulted in generally more necrosis (Figure 5I;
Figure S24, Supporting Information) and fewer PCNA-pos-
itive proliferative cells (Figure 5J; Figure S25, Supporting
Information) in tumors relative to saline and other controls.
No obvious histological toxicity to major organs (heart, liver,
spleen, lung, and kidney) was observed in all groups as shown
in the sections for hematoxylin and eosin staining (FigureS26,
Supporting Information). Blood analysis, including biochem-
ical parameters for the evaluation of hepatoxicity and nephro-
toxicity and routine CBC (complete blood count) and WBC
(white blood cell) dierential assay, was further performed to
evaluate the treatment-associated toxicity. No significant dif-
ferences in the multiple parameters of DOX-NP@Mϕ group
were observed when compared to the saline group, indicating
a good tolerance of the therapy (Figures S27 and S28, Sup-
porting Information).
3. Conclusion
In summary, the approach developed here has multiple poten-
tial advantages for LC-DDSs. First, for the protection of carrier
cell bioactivity, divalent metal ions-based MPN can aord excel-
lent adhesion of nanoparticles to Mϕ surface with minimized
uptake (5–10%). Compared to the conventional intracellular
loading via endocytosis, this superior surface attachment well
maintained the bioactivity of carrier cells, aording extended
time window for clinical application. Second, for ecient drug
release and utilization by targeted cells, improved uptake of
nanoparticle-associated vesicles by tumor cells and enhanced
chemotherapy can be achieved through MPN-mediated pho-
tothermal eect, a clinically relevant, controllable stimulus.
Third, this method for LC-DDS construction is facile, benign,
and ecient (ice bath, 2 min), avoiding the generally compli-
cated chemical and engineering procedures. This platform
technology may also be tailored for other cell types and nan-
oparticles with diverse cargos according to specific disease
conditions.
4. Experimental Section
Materials, Cells, and Animal: Tannic acid (TA, %) was purchased
from Yuanye Biotechnology (Shanghai, China). Chromium (III) chloride
hexahydrate (CrCl·HO), Manganese (II) chloride (MnCl), Iron (III)
chloride hexahydrate (FeCl·HO), Cobalt (II) chloride (CoCl), Copper
(II) chloride dihydrate (CuCl·HO), and Zinc chloride (ZnCl) were
purchased from Macklin Biochemical (Shanghai, China). PLGA (acid
terminated, lactide: glycolide :, MW -), coumarin
,-morpholinepropanesulfonic acid (MOPS, %), and sodium
cholate were purchased from Merck. Doxorubicin hydrochloride was
obtained from HVSF United Chemical Materials (Beijing, China).
DiR (,-dioctadecyl-,,,-tetramethylindotricarbocyaine iodide)
and DiD (,″-dioctadecyl-,,″,′-tetramethylindodicarbocyanine,
-chlorobenzenesulfonate salt) were purchased from Yeasen
Biotechnology (Shanghai, China). Cell counting kit- (CCK-) and
Annexin V-FITC Apoptosis Detection Kit were purchased from Dojindo
Laboratories (Kumamoto, Japan). Lactate dehydrogenase (LDH) kit
and LIVE/DEAD cytotoxicity assay kit were obtained from Beyotime
Biotechnology (Shanghai, China). D-luciferin was purchased from J&K
Scientific Company (Shanghai, China). Dulbecco’s modified Eagle’s
medium (DMEM), fetal bovine serum (FBS), penicillin, streptomycin,
and trypsin were obtained from Thermo Fisher Scientific (Shanghai,
China).
Mouse macrophage RAW. and T breast cancer cell line was
purchased from ATCC (Manassas, VA). T-luc cells (T transfected
with luciferase) were established by Shanghai Model Organisms Center
(China). T and RAW. cells were cultured in DMEM medium with
% FBS, UL− penicillin, and mgL− streptomycin. The culture
was maintained at °C in a humidified atmosphere containing % CO.
Female BALB/c mice (≈ g) were provided by Shanghai Laboratory
Animal Center (Chinese Academy of Sciences, Shanghai, China).Female
BALB/c mice (~  g) were provided by Shanghai Laboratory Animal
Center (Chinese Academy of Sciences, Shanghai, China). The animal
experiment in this study was approved by the IACUC of Shanghai Jiao
Tong University School of Medicine.
Preparation and Characterization of DOX-NP@MPN:  mg
doxorubicin hydrochloride and  µL triethylamine were added into
 mL dichloromethane (DCM). Stirring overnight was performed for
dehydrochlorination. Then, µL DOX in DCM (mgmL−) was mixed
with mL PLGA in DCM (mgmL−) and mL of % sodium cholate
in water. After probe ultrasonication at W for s, the emulsion was
poured into mL .% sodium cholate solution and stirred for h in
the dark to evaporate DCM. DOX-NP was collected after centrifugation
at  g for min. DOX loading was examined through fluorescence
detection (Ex  nm, Em  nm). Coumarin -labeled PLGA
nanoparticles were also prepared using the similar protocol in which
DOX in DCM was replaced with the fluorescent probes.
For metal ion-phenolic network (MPN) coating, µL tannic acid (TA)
in water (m) was added to  mL DOX-NP solution (mgmL−) for
s vortex. Then, µL metal chloride solution ( m) was added to
the mixture under water bath ultrasound for MPN coating. Next, µL
-morpholinepropanesulfonic acid (MOPS) solution ( m, pH .)
was added for s vortex. Free TA and metal ions were then removed by
centrifugation (g,  min) to purify the resulting DOX-NP@MPN.
To acquire enough photothermal eect, more amount of TA (µL) and
FeCl ( µL) solutions were used in the production of FePN-coated
DOX-NP. The UV–vis–NIR spectra of metal ions-contained MPNs
and their coated DOX-NP was measured by a microplate reader
(SpectraMax M/Me Multimode Microplate Readers, Molecular
Devices). The metal chloride solutions were prepared by solving their
salts, including CrCl·HO, MnCl, FeCl·HO, CoCl, CuCl·HO, and
ZnCl, in water. The particle size and zeta potential of the nanoparticles
were characterized by ZetaSizer Nano ZS instrument (Malvern,
Worcestershire, UK).
Metal Ion Screening and Establishment of DOX-NP@Mϕ: As the trace
elements essential for humans,[] six metal ions (Cr+, Mn+, Fe+, Co+,
Cu+, and Zn+) were screened for optimal Mϕ cell surface attachment.
16163028, 2023, 19, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202214842 by Shanghai Jiao Tong University, Wiley Online Library on [09/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

www.afm-journal.dewww.advancedsciencenews.com
2214842 (11 of 13) ©  Wiley-VCH GmbH
Adv. Funct. Mater. 2023, 33, 
For this test, × Mϕ cells (in mL % glucose) were mixed with mg
coumarin -labeled NP@MPN (in mL % glucose) in ice bath. Then,
µL metal chloride solution (m) and µL MOPS (m) were
added sequentially. After stirring for min, the cell suspensions were
centrifuged ( rpm,  min,  °C) to remove free nanoparticles and
obtain purified NP@Mϕ. The coumarin -labeled nanoparticle surface
adhesion was observed under confocal laser scanning microscopy (Ex
nm, Em nm). The proportion of extracellular to total fluorescence
intensity was assayed using Image J software (NIH, USA). The ions
with the best surface-attached ability were adopted for DOX-NP@Mϕ
preparation. Mϕ cells with surface-adsorbed nanoparticles were also
observed using an ultra-high resolution scanning electron microscope
(Hitachi Regulus ).
DOX-NP@Mϕ was fabricated by mixing × Mϕ cells with  mg
DOX-NP@MPN in ice bath. Other procedures are same to those
described above. DOX loadings per  cells were determined by
quantifying the cell-associated DOX contents via fluorescence assay (Ex
nm, Em  nm). For comparison, Mϕ with intracellular loading of
DOX-NP was prepared via the endocytosis. Briefly, mL macrophage
suspension (×mL−) in serum-free medium was placed in a -well
plate for overnight incubation. Then,  mg DOX-NP in  µL serum-
free medium was added to the well. After h incubation, Mϕ cells with
intracellularly loaded DOX-NP were obtained.
Proliferation and Viability Detection of Mϕ Cells with Surface-Attached
DOX-NP: Proliferation of Mϕ cells with surface-attached PLGA
nanoparticles was real-time detected using an IncuCyte ZOOM Live-Cell
Analysis System (Essen Bioscience). For this assay, NP@Mϕ (empty)
or Mϕ alone were seeded in a -well plate at a cell density of × ,
× , and × per well, respectively. Then, the plate was placed at
°C in the instrument for continuous monitoring and photographing.
Cell proliferation curves presented as phase object confluence were
plotted using IncuCyte A software.
Multiple methods were used to assay the viability of Mϕ cells with
surface-attached or intracellularly loaded DOX-NP. DOX loadings of
., ., and . µg − cells were used in the tests. Mϕ cells with
the loaded DOX-NP were seeded in -well plates at a cell density of
×  per well. After h incubation, the cell viabilities were examined
using Cell Counting Kit- (CCK-, Dojindo Laboratories) assay. In
another test, lactate dehydrogenase (LDH) release, an index reflecting
the cell membrane integrity, was examined using a LDH kit (Beyotime
Biotechnology) h after the cell seeding.
Cell viabilities were also examined using the LIVE/DEAD cell viability
assay kit (Beyotime Biotechnology) and observed under confocal laser
scanning microscopy (CLSM, Leica TCS SP). In addition, Annexin
V-FITC apoptosis detection using flow cytometry (Attune NxT Flow
Cytometer, Thermo Fisher Scientific) was performed. For these two tests,
Mϕ cells with the surface-attached or intracellularly loaded DOX-NP
were seeded in a -well plate at a cell density of × per well, and the
assays were carried out after h incubation.
Migration Assay of Mϕ Cells with Surface-Attached DOX-NP: The
migration ability of DOX-NP@Mϕ was evaluated using Transwell
chamber. × Mϕ cells with the specific DOX loadings (., ., and
. µg − cells) were added to the upper chamber. The conditioned
culture medium of T cells were added to the lower chamber as a
chemotaxis stimulus. After h incubation, the chamber was taken out,
washed with PBS, and fixed with methanol. The migrated cells were
then stained with crystal violet, photographed under a microscope, and
counted using Image J software. Migration ratio was calculated relative
to the cells with no treatments. Mϕ cells with intracellularly loaded
DOX-NP were included as a control.
Photothermal Heating of DOX-NP@Mϕ In Vitro: FePN,
DOX-NP@FePN, and DOX-NP@Mϕ suspensions ( µg mL− FePN
in each group) prepared as described above were added to a -well
plate or Eppendorf tubes, respectively. Samples in the wells and tubes
were continuously irradiated with an  nm laser (Changchun New
Industries Optoelectronics Technology, Jilin, China) at  W cm− for
 min. The temperature increases in the wells were monitored using
a Pt temperature probe (Testo). The temperature changes in the
photothermal images were recorded using a Testo  thermal imaging
camera. Photothermal stability was examined through cyclically min
heating DOX-NP@Mϕ suspension for  times.
Identification of Photothermal-Induced Liberation of Nanoparticle-
Associated Vesicles: Mϕ cells with surface-attached PLGA nanoparticles
(coumarin  labeled) were prepared as described above. After min light
irradiation ( nm,  W cm−), the liberated nanoparticle-associated
vesicles (Ex  nm, Em nm) were observed under CLSM. Smaller
detached particles were further analyzed using DLS after removing the
cells and debris via centrifugation (rpm, min) at °C.
In another test, PLGA nanoparticles were labeled with DiD, and Mϕ
cells were labeled with Hoechst  to generate the dual-fluorescence
labeled NP@Mϕ. After  or  min light irradiation, flow cytometry
assay was performed to identify the detached vesicles or nanoparticles,
which were DiD positive and Hoechst negative.
Cellular Uptake and Cytotoxicity in 4T1 Cells: T cells were seeded in
confocal dishes at a cell density of ×  per well. After h, the cells
were incubated with NP@Mϕ pre-treated with light irradiation (nm,
Wcm−, min) or not, for h. PLGA nanoparticles were labeled with
coumarin  and the final concentration in this test was  ng mL− in
the medium. Then the culture medium was replaced with fresh medium
and cellular uptake of the photothermal-induced nanoparticle-associated
vesicles was observed under CLSM (Ex nm, Em nm) and further
quantified using flow cytometry.
To assay the cytotoxicity of DOX-NP@Mϕ to tumor cells, T-luc cells
were seeded at × per well in -well plates. After h, the cells were
incubated with DOX-NP@Mϕ pre-treated with light irradiation (nm,
 W cm−,  min) or not for h, and the final DOX contents in the
medium were ., , , and µgmL−, respectively. Then, the cells were
incubated in µL fresh medium containing D-luciferin (J&K Chemical,
mgmL−) for min. T cell viabilities were estimated by comparing
the bioluminescence signal intensity under an IVIS Spectrum/CT
imaging system (PerkinElmer). Mϕ cells alone were included as control.
Tumor Targeting and Biodistribution In Vivo: PLGA nanoparticles were
labeled with DiD, and Mϕ cells were labeled with DiR to construct the
dual-fluorescence labeled NP@Mϕ. When the orthotopic T tumors
grew to ≈ mm, NP@Mϕ ( ×  Mϕ cells) in µL PBS was i.v.
injected into the mice. After , , , and h, the mice were sacrificed and
the tumors were excised for ex vivo imaging under an IVIS Spectrum/
CT imaging system. The signals of DiD (Ex nm, Em nm) and
DiR (Ex nm, Em nm) were identified for the PLGA nanoparticles
and Mϕ cells, respectively. The distribution of the two components
in other major organs (heart, liver, spleen, liver, and kidney) was also
examined. Tumor targeting was also assayed using the D imaging from
multiple angles (coronal, sagittal, transaxial, and perspective) h after
i.v. injection. For this examination, imaging of the tumor site only was
performed under the IVIS Spectrum/CT system.
To identify the drug distribution in tumors, DOX-NP@Mϕ ( × 
Mϕ cells, . mgkg− DOX) was i.v. injected into the mice.  h after
injection, mouse tumors were excised for ex vivo imaging (Ex nm,
Em  nm). The tumors were then homogenized for DOX content
measurement by fluorescence detection (Ex  nm, Em  nm).
DOX-NP (PEGylated) was included as control.
Antitumor Therapy of DOX-NP@Mϕ In Vivo: When the orthotopic
tumor grew to ≈mm ( d after inoculation of .×  tumor cells),
the mice were randomly allocated to  groups for various treatments:
) Saline, ) NP@Mϕ (empty)+L (light), ) free DOX, ) DOX-NP
(PEGylated), ) DOX-NP@Mϕ and ) DOX-NP@Mϕ+L. Intravenous
injections were arranged on days , , and , respectively. For each
injection, . mg kg− DOX in the formulation was administered
to the mice when involved. Thus, the total DOX dose was  mg kg−.
Light irradiation ( nm,  W cm−,  min) on the tumor site was
performed  h after each injection. Tumor volume was calculated as
(length)× (width)/ and mouse bodies were recorded every two days.
Survival observation was recorded till tumor volume of the mouse
reached the ethical limit (mm). The increase in life span (ILS) was
calculated by comparing the median survival time of the mice from the
treatment group and the saline group.[]
16163028, 2023, 19, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202214842 by Shanghai Jiao Tong University, Wiley Online Library on [09/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

www.afm-journal.dewww.advancedsciencenews.com
2214842 (12 of 13) ©  Wiley-VCH GmbH
Adv. Funct. Mater. 2023, 33, 
 h after the last injection, tumors and major organs (heart, liver,
spleen, lung, and kidney) from  mice per group were excised, fixed, and
processed for paran section and H&E staining for histopathological
evaluation. The proliferative tumor cells were identified using a PCNA cell
proliferation kit (Servicebio, China). Microscopy images were taken using
Nikon Eclipse E photomicroscope and analyzed using Image J software.
In a separate study, mice from each group (n=) were sacrificed
on day  for isolated tumor photographing and weighting. The mouse
lungs were excised for ex vivo bioluminescent imaging to detect the
metastasis as previously described [] and processed for H&E staining
and pathological assay.
Blood Analysis for Safety Evaluation: Healthy female BALB/c mice were
i.v. administered with saline, NP@Mϕ (empty), free DOX, DOX-NP
(PEGylated), and DOX-NP@Mϕ, respectively. The doses of Mϕ and
DOX were the same as those in the therapy protocol. h after the rd
injection, blood collected from the retro-orbital vein was analyzed in
Drug Safety Evaluation Research Center (Shanghai Institute of Materia
Medica, Shanghai, China). Main serum biochemistry parameters relative
to the liver and kidney toxicity were examined. Routine complete blood
count and white blood cell dierential were also monitored.
Statistical Analysis: GraphPad Prism . software (La Jolla, CA) was
used for the statistical analysis in this study. Dierences between groups
were examined using Student’s test or ANOVA with follow-up Tukey’s
multiple comparisons. p-value below . was considered significant.
Supporting Information
Supporting Information is available from the Wiley Online Library or
from the author.
Acknowledgements
M.-H.Z and X.-D.Z contributed equally to this work. The authors thank
the Core Facility of Basic Medical Sciences (SJTU-SM) for the help in
imaging and flow cytometry assay. This work was supported by National
Natural Science Foundation of China ( and ), and
Program of Shanghai Academic Research Leader (Shanghai Municipal
Science and Technology Commission) (XD), and Shanghai
Municipal Health Commission (XD).
Conflict of Interest
The authors declare no conflict of interest.
Data Availability Statement
The data that support the findings of this study are available from the
corresponding author upon reasonable request.
Keywords
doxorubicin, living cell-based drug delivery systems, macrophages,
metal ion-phenolic networks, photothermal therapies
Received: December , 
Revised: January , 
Published online: February , 
[] P. Agarwalla, E. A. Ogunnaike, S.Ahn, K. A.Froehlich, A.Jansson,
F. S.Ligler, G.Dotti, Y.Brudno, Nat. Biotechnol. 2022, 40, .
[] M. Klichinsky, M. Ruella, O. Shestova, X. M. Lu, A. Best,
M. Zeeman, M. Schmierer, K. Gabrusiewicz, N. R. Anderson,
N. E.Petty, K. D. Cummins, F.Shen, X. Shan, K. Veliz, K.Blouch,
Y. Yashiro-Ohtani, S. S. Kenderian, M. Y. Kim, R. S. O’Connor,
S. R. Wallace, M. S. Kozlowski, D. M. Marchione, M. Shestov,
B. A.Garcia, C. H.June, S.Gill, Nat. Biotechnol. 2020, 38, .
[] M.Daher, K.Rezvani, Cancer Discov. 2021, 11, .
[] a) A. C.Anselmo, S. Mitragotri, J. Controlled Release 2014, 190, ;
b) L. A. L.Fliervoet, E.Mastrobattista, Adv. Drug Delivery Rev. 2016,
106, ; c) M.Ayer, H. A. Klok, J. Controlled Release 2017, 259, ;
d) L. M.Bush, C. P.Healy, S. B.Javdan, J. C.Emmons, T. L. Deans,
Trends Pharmacol. Sci. 2021, 42, .
[] Q.Hu, W.Sun, J.Wang, H.Ruan, X.Zhang, Y.Ye, S.Shen, C.Wang,
W.Lu, K.Cheng, G.Dotti, J. F.Zeidner, J.Wang, Z.Gu, Nat. Biomed.
Eng. 2018, 2, .
[] W. Li, Z. Su, M. Hao, C. Ju, C. Zhang, J. Controlled Release 2020,
328, .
[] a) P. Hammel, P. Fabienne, L. Mineur, J. P. Metges, T. Andre,
C.De La Fouchardiere, C.Louvet, F.El Hajbi, R.Faroux, R.Guimbaud,
D.Tougeron, O.Bouche, T.Lecomte, C.Rebischung, C.Tournigand,
J. Cros, R. Kay, A. Hamm, A. Gupta, J. B. Bachet, I. El Hariry,
Eur J Cancer 2020, 124, ; b) J. B. Bachet, H. Blons, P. Hammel,
I. E.Hariry, F.Portales, L.Mineur, J. P.Metges, C.Mulot, C.Bourreau,
J. Cain, J. Cros, P. Laurent-Puig, Clin. Cancer Res. 2020, 26, ;
c) L. S.Lynggaard, G.Vaitkeviciene, C.Langenskiöld, A. K.Lehmann,
P. M. Lähteenmäki, K. Lepik, I. El Hariry, K. Schmiegelow,
B. K.Albertsen, Br. J. Haematol. 2022, 197, .
[] N. Doshi, A. J. Swiston, J. B. Gilbert, M. L. Alcaraz, R. E. Cohen,
M. F.Rubner, S.Mitragotri, Adv. Mater. 2011, 23, H.
[] W. Zhang, M. Wang, W. Tang, R.Wen, S. Zhou, C. Lee, H. Wang,
W. Jiang, I. M.Delahunty, Z. Zhen, H. Chen, M.Chapman, Z. Wu,
E. W.Howerth, H.Cai, Z.Li, J.Xie, Adv. Mater. 2018, 30, .
[] a) Y. Xia, L. Rao, H. Yao, Z. Wang, P.Ning, X. Chen, Adv. Mater.
2020, 32, ; b) V. D. Nguyen, H. K. Min, D. H. Kim,
C. S. Kim, J. Han, J. O. Park, E. Choi, ACS Appl. Mater. Inter-
faces 2020, 12, ; c) C. Gao, Q. Cheng, J. Wei, C. Sun, S. Lu,
C. H. T.Kwong, S. M. Y.Lee, Z.Zhong, R.Wang, Mater. Today 2020,
40, ; d) L.Liu, H.Li, J.Wang, J.Zhang, X. J.Liang, W.Guo, Z.Gu,
Adv. Drug Delivery Rev. 2022, 183, ; e) Y.Chen, D.Qin, J.Zou,
X. Li, X. D. Guo, Y. Tang, C.Liu, W. Chen, N. Kong, C. Y. Zhang,
W.Tao, Adv. Mater. 2022, .
[] P.Sun, Q.Deng, L.Kang, Y.Sun, J.Ren, X.Qu, ACS Nano 2020, 14,
.
[] M. R. Choi, K. J. Stanton-Maxey, J. K. Stanley, C. S. Levin,
R.Bardhan, D. Akin, S. Badve, J.Sturgis, J. P.Robinson, R. Bashir,
N. J.Halas, S. E.Clare, Nano Lett. 2007, 7, .
[] a) D. Chu, J.Gao, Z.Wang, ACS Nano 2015, 9, ; b) D.Chu,
X.Dong, X. Shi, C.Zhang, Z. Wang, Adv. Mater. 2018, 30, ;
c) H. Zhang, Z. Li, C. Gao, X. Fan, Y. Pang, T.Li, Z. Wu, H. Xie,
Q.He, Sci Robot 2021, 6, aaz
[] J.Xue, Z.Zhao, L.Zhang, L.Xue, S.Shen, Y.Wen, Z.Wei, L.Wang,
L.Kong, H.Sun, Q.Ping, R. Mo, C.Zhang, Nat. Nanotechnol. 2017,
12, .
[] J.Shao, M.Xuan, H. Zhang, X. Lin, Z. Wu, Q.He, Angew. Chem.,
Int. Ed. 2017, 56, .
[] a) M. T. Stephan, J. J. Moon, S. H.Um, A.Bershteyn, D. J. Irvine,
Nat. Med. 2010, 16, ; b) M.Hao, S. Hou, W.Li, K. Li, L. Xue,
Q.Hu, L.Zhu, Y.Chen, H. Sun, C. Ju, C. Zhang, Sci. Transl. Med.
2020, 12, aaz.
[] B. Huang, W. D. Abraham, Y. Zheng, S. C. Bustamante López,
S. S.Luo, D. J.Irvine, Sci. Transl. Med. 2015, 7, ra.
[] a) M.Ayer, M. Schuster, I.Gruber, C.Blatti, E.Kaba, G. Enzmann,
O.Burri, R.Guiet, A.Seitz, B. Engelhardt, H. A. Klok, Adv. Health-
care Mater. 2021, 10, ; b) T.Thomsen, R.Reissmann, E.Kaba,
B.Engelhardt, H. A.Klok, Biomacromolecules 2021, 22, .
16163028, 2023, 19, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202214842 by Shanghai Jiao Tong University, Wiley Online Library on [09/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

www.afm-journal.dewww.advancedsciencenews.com
2214842 (13 of 13) ©  Wiley-VCH GmbH
Adv. Funct. Mater. 2023, 33, 
[] E. L. Siegler, Y. J. Kim, X. Chen, N. Siriwon, J. Mac, J. A. Rohrs,
P. D.Bryson, P.Wang, Mol. Ther. 2017, 25, .
[] S.Im, D.Jang, G.Saravanakumar, J.Lee, Y. Kang, Y. M.Lee, J.Lee,
J. Doh, Z. Y.Yang, M. H. Jang, W. J. Kim, Adv. Mater. 2020, 32,
.
[] Z.Zhang, H.Wang, T. Tan, J.Li, Z. Wang, Y.Li, Adv. Funct. Mater.
2018, 28, .
[] H.Kim, K. Shin, O. K.Park, D.Choi, H. D.Kim, S.Baik, S. H. Lee,
S. H.Kwon, K. J.Yarema, J.Hong, T.Hyeon, N. S. Hwang, J. Am.
Chem. Soc. 2018, 140, .
[] C. W. t. Shields, M. A.Evans, L. L.Wang, N. Baugh, S.Iyer, D.Wu,
Z.Zhao, A.Pusuluri, A.Ukidve, D. C. Pan, S. Mitragotri, Sci. Adv.
2020, 6, aaz.
[] N. L.Klyachko, R.Polak, M. J.Haney, Y. Zhao, R. J. Gomes Neto,
M. C.Hill, A. V.Kabanov, R. E.Cohen, M. F.Rubner, E. V.Batrakova,
Biomaterials 2017, 140, .
[] L.Tang, Y.Zheng, M. B. Melo, L.Mabardi, A. P.Castaño, Y. Q.Xie,
N. Li, S. B. Kudchodkar, H. C. Wong, E. K. Jeng, M. V. Maus,
D. J.Irvine, Nat. Biotechnol. 2018, 36, .
[] a) J.Guo, B. L.Tardy, A. J.Christoerson, Y. Dai, J. J. Richardson,
W. Zhu, M.Hu, Y.Ju, J.Cui, R. R.Dagastine, I. Yarovsky, F. Caruso,
Nat. Nanotechnol. 2016, 11, ; b) W. Zhang, Q. A. Besford,
A. J. Christoerson, P. Charchar, J. J. Richardson, A. Elbourne,
K. Kempe, C. E. Hagemeyer, M. R. Field, C. F. McConville,
I. Yarovsky, F.Caruso, Nano Lett. 2020, 20, ; c) L. Xie, J. Li,
G.Wang, W.Sang, M. Xu, W.Li, J. Yan, B.Li, Z. Zhang, Q. Zhao,
Z.Yuan, Q.Fan, Y.Dai, J. Am. Chem. Soc. 2022, 144, ; d) J.Yan,
G. Wang, L. Xie, H. Tian, J. Li, B. Li, W. Sang, W. Li, Z. Zhang,
Y.Dai, Adv. Mater. 2022, 34, .
[] Z. Zhao, D. C. Pan, Q. M. Qi, J. Kim, N. Kapate, T. Sun,
C. W. t. Shields, L. L. Wang, D. Wu, C. J. Kwon, W. He, J. Guo,
S.Mitragotri, Adv. Mater. 2020, 32, .
[] W.Mertz, Science 1981, 213, .
[] a) Y. Han, Z. Lin, J. Zhou, G. Yun, R. Guo, J. J. Richardson,
F.Caruso, Angew. Chem., Int. Ed. 2020, 59, ; b) J.Chen, S.Pan,
J.Zhou, Z.Lin, Y.Qu, A.Glab, Y.Han, J. J.Richardson, F.Caruso,
Adv. Mater. 2022, 34, .
[] J. E.Lancet, G. L.Uy, L. F.Newell, T. L.Lin, E. K.Ritchie, R. K.Stuart,
S. A.Strickland, D.Hogge, S. R. Solomon, D. L.Bixby, J. E.Kolitz,
G. J. Schiller, M. J.Wieduwilt, D. H. Ryan, S. Faderl, J. E.Cortes,
Lancet Haematol. 2021, 8, .
[] J. Guo, Y. Ping, H. Ejima, K. Alt, M. Meissner, J. J. Richardson,
Y. Yan, K. Peter, D. von Elverfeldt, C. E. Hagemeyer, F. Caruso,
Angew. Chem., Int. Ed. 2014, 53, .
[] S. Mylvaganam, S. A. Freeman, S. Grinstein, Curr. Biol. 2021, 31,
R.
[] X.He, H. Cao, H.Wang, T.Tan, H.Yu, P.Zhang, Q.Yin, Z.Zhang,
Y.Li, Nano Lett. 2017, 17, .
[] T. Liu, M. Zhang, W. Liu, X. Zeng, X. Song, X. Yang, X. Zhang,
J.Feng, ACS Nano 2018, 12, .
[] X.Li, J. F. Lovell, J. Yoon, X. Chen, Nat. Rev. Clin. Oncol. 2020, 17,
.
[] H. J.Li, J. Z.Du, X. J.Du, C. F.Xu, C. Y.Sun, H. X.Wang, Z. T.Cao,
X. Z. Yang, Y. H.Zhu, S. Nie, J. Wang, Proc. Natl. Acad. Sci. USA
2016, 113, .
[] H. Y. Feng, Y. Yuan, Y. Zhang, H. J. Liu, X. Dong, S. C. Yang,
X. L. Liu, X. Lai, M. H. Zhu, J. Wang, Q.Lu, Q.Lin, H. Z.Chen,
J. F.Lovell, P.Sun, C.Fang, Nano-Micro Lett. 2021, 13, .
16163028, 2023, 19, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202214842 by Shanghai Jiao Tong University, Wiley Online Library on [09/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License