Diatom-like silica–protein nanocomposites for sustained drug delivery of ruthenium polypyridyl complexes

Article abstract of DOI:10.1016/j.jinorgbio.2021.111489

Inspired by the unique glass cell wall of diatom, we design a new nanostructure of human serum albumin nanoparticle (HSANP) coated with silica (HSA/SiO₂), which consists of a core–satellite assembly of small silica nanoparticles on a single HSA

Full text of DOI:10.1016/j.jinorgbio.2021.111489

Journal of Inorganic Biochemistry 221 (2021) 111489
Contents lists available at ScienceDirect
Journal of Inorganic Biochemistry
journal homepage: www.elsevier.com/locate/jinorgbio
Diatom-like silica–protein nanocomposites for sustained drug delivery of
ruthenium polypyridyl complexes
Hongdong Shia, Jingxue Loua, Simin Lina, Yi Wanga, Yatao Hua, Pingyu Zhanga,
,
*
Yangzhong Liub, Qianling Zhanga
a Graphene Composite Research Center, College of Chemistry and Environmental Engineering, Shenzhen University, Shenzhen, Guangdong 518060, China
b CAS Key Laboratory of Soft Matter Chemistry, CAS High Magnetic Field Laboratory, Department of Chemistry, University of Science and Technology of China, Hefei,
Anhui 230026, China
A R T I C L E I N F O
A B S T R A C T
Keywords:
Inspired by the unique glass cell wall of diatom, we design a new nanostructure of human serum albumin
nanoparticle (HSANP) coated with silica (HSA/SiO2), which consists of a core–satellite assembly of small silica
nanoparticles on a single HSANP. The HSA/SiO2 nanoparticles are used for delivering ruthenium polypyridyl
complexes into cells. The silica coating increases the Ru loading efficiency, and prevents the burst release of Ru
from HSA/SiO2. The Ru release rate can be controlled by adjusting the amount of coated silica on HSANP,
affording a drug delivery system with controlled drug release rate. The Ru-HSA/SiO2 nanoparticles show high
stability in physiological condition, and significantly increase the Ru uptake into cells, which proceeds via
clathrin-mediated endocytosis into the lysosomes. The silica coating takes no effect on the fluorescence intensity
and ROS generation of loaded Ru in HSA/SiO2. Furthermore, Ru4-HSA/SiO2 exhibit weak cytotoxicity in dark,
however, the nanodrug can be activated by light irradiation and generate ROS to damage cells, thus achieving an
excellent photodynamic therapy efficiency. Therefore, the diatom-like nanostructure can function as sustained
drug delivery nanocarrier of ruthenium polypyridyl complex and can be used for bioimaging and photodynamic
therapy.
Diatom-like nanostructure
Silica–protein nanocomposite
Ruthenium polypyridyl complexes
Sustained drug delivery
Controlled drug release
Photodynamic therapy
1. Introduction
toxicity still need to be resolved [12,13].
Recently, various nanoparticles have been widely used in drug de-
Platinum-based metallodrugs such as cisplatin are successful anti-
cancer agents. However, their application is limited by their severe side
effects and drug resistance [1,2]. In the search for other more effective
anticancer metallodrugs, ruthenium complexes have emerged as prom-
ising candidates that show excellent properties including low toxicity,
overcoming of drug resistance, and antimetastasis activity [3–6]. In
addition, ruthenium polypyridyl complexes have special photophysical
and photochemical properties, and can be used in many fields including
catalysis, solar energy, sensor, and biological applications [7–9]. In
recent years, ruthenium polypyridyl complexes have been widely used
in biological systems as luminescent probes, luminescent imaging
agents, photosensitizers, therapeutics, and theranostics [10,11]. Unfor-
tunately, a number of limitations regarding short circulation time,
nonspecific binding with biomolecules, slow cellular uptake, and
livery to increase drug circulation time, enhance drug efficacy, and
reduce side effects [14–16]. Moreover, nanoparticles are found to
accumulate at the tumor site by enhanced permeability and retention
(EPR) effects [17,18]. Many nanodelivery systems are used for deliv-
ering ruthenium polypyridyl complexes [19], such as liposomes [20],
polymers [21], inorganic nanoparticles [22–24], metal–organic frame-
works [25], and protein nanoparticles [26,28]. In these cases, the
ruthenium complexes are loaded in the nanoparticles through covalent
or noncovalent bonds. For covalent conjugation, the presence of active
groups (carboxyl, amino, hydroxyl) in the ruthenium polypyridyl com-
plexes is required. However, most of these complexes, such as [Ru
(bpy)3]2+ (bpy = 2,2′-bipyridine), [Ru(bpy)2(dppz)]2+(dppz = dipyr-
idophenazine), and [Ru(phen)2(dppz)]2+ (phen = 1,10-phenanthro-
line), do not contain active groups; therefore, they only can be loaded in
Abbreviations: TEOS, tetraethyl orthosilicate; PBS, phosphate buffer saline; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetra-zolium bromide; DMEM, Dul-
becco’s Modified Eagle Medium; HepG, human hepatoellular carcinoma; PDT, photodynamic therapy; DNA, deoxyribonucleic acid.
* Corresponding author.
E-mail addresses: liuyz@ustc.edu.cn (Y. Liu), zhql@szu.edu.cn (Q. Zhang).
https://doi.org/10.1016/j.jinorgbio.2021.111489
Received 22 December 2020; Received in revised form 28 April 2021; Accepted 8 May 2021
Available online 13 May 2021
0162-0134/© 2021 Elsevier Inc. All rights reserved.

H. Shi et al.
Journal of Inorganic Biochemistry 221 (2021) 111489
nanoparticles by noncovalent interactions including porous absorption,
electrostatic interaction, hydrophobic interaction, and encapsulation,
which results in low drug loading efficiency and drug burst release under
physiological condition [13].
mmol), bpy (936 mg, 6 mmol), and LiCl (840 mg, 20 mmol) were dis-
solved in 5 mL dimethylformamide. After removing the oxygen by filling
with argon, the mixture was stirred and heated to 140 ◦C at reflux. After
8 h reaction, the solution was cooled down to room temperature, and
excess acetone was added. The mixture was placed at 4 ◦C overnight to
crystallize the product, which was then collected by centrifugation,
washed several times with water, acetone, and diethyl ether, and dried
by lyophilization.
Diatoms are major group of phytoplankton in the world oceans,
which are central in aquatic ecosystems and the global carbon cycle. A
striking feature of diatoms is their beautiful cell wall (or frustule), which
is made of hydrate glass (SiO2.nH2O) [29]. The frustule has many
functions such as controlling the sinking rate, giving mechanical pro-
tection from predators, and protecting the DNA from ultraviolet light
[30–32]. In this work, inspired by the glass wall of diatoms, we con-
structed a new nanostructure consisting of a core–satellite assembly of
small silica nanoparticles (SiNP) on a single Ru-human serum albumin
nanoparticle (Ru-HSANP), which afforded silica-coated Ru-HSA/SiO2
nanoparticles, as shown in Scheme 1. The silica coating could effectively
prevent the burst release of hydrophilic ruthenium polypyridyl com-
plexes from the HSA nanoparticles in physiological condition compared
with the noncoated system. The amount and size of silica nanoparticles
on HSANP were controlled by adding different amount of tetraethyl
orthosilicate (TEOS) and ammonia during the silica coating process. In
addition, the release rate of ruthenium complexes could be tuned by
controlling the amount of silica. Thus, we developed a drug delivery
system in which the drug release rate can be controlled.
[Ru(bpy)3]Cl2 (Ru1): Complex [Ru(bpy)3](PF6)2 was synthesized
according to methods described in our previous work [26]. Briefly, Ru
(bpy)2Cl2⋅2H2O (100 mg, 0.192 mmol) was dissolved in 4 mL ethylene
glycol and bpy (45 mg, 0.288 mmol). After filling with argon to remove
the oxygen, the mixture was heated to 120 ◦C at reflux and stirred for 6
h. After cooling to room temperature, 23 mL distilled water was added
excess saturated KPF6 solution. A dark red precipitate was produced and
collected by centrifugation. The solid was washed three times with
water, diethyl ether, and purified by chromatography (aluminum oxide,
acetonitrile: ethanol = 1:1). The solvent was evaporated, and the
resultant residue was washed by saturated KPF6 solution. The product
was washed several times with water, and was then collected by
centrifugation. Finally, the solid was dissolved in acetone and precipi-
tated by adding diethyl ether. The procedure was repeated three times.
The as-obtained pure [Ru(bpy)3](PF6)2 was dried by lyophilization.
[Ru(bpy)3]Cl2 was prepared as follows: [Ru(bpy)3](PF6)2 was dis-
solved in 5 mL acetone, and an excess amount of tetrabutylammonium
chloride dissolved in 3 mL acetone was added. A red precipitate was
produced, and the mixture was stirred for 30 min. The solid was
collected by centrifugation and washed with acetone and diethyl ether
three times. The product (Ru1) was dried by lyophilization and char-
acterized by 1H NMR (Fig. S1). 1H NMR (400 MHz, D2O) δ 8.50 (d, 6H),
8.00 (t, 6H), 7.79 (d, 6H), 7.37–7.28 (t, 6H).
2. Experimental section
2.1. Materials
2,2′-bipyridine (bpy), 6,6′-dimethyl-2,2′-bipyridine (dmbpy),
lithium chloride (LiCl), and potassium hexafluorophosphate (KPF6)
were purchased from Shanghai Aladdin Industrial Corporation (China),
Dulbecco’s modified eagle medium (DMEM), Roswell Park Memorial
Institute (RPMI 1640), trypsin, fetal bovine serum (FBS),
Methylthiazolyldiphenyl-tetrazolium bromide (MTT), and herring
sperm DNA were obtained from Sangon Biotech (Shanghai) Co. Ltd.
(China). All other chemicals and solvents were obtained commercially
and used without further purification.
[Ru(bpy)2(HPIP)]Cl2 (Ru2): 2-(4-Hydroxyphenyl)imidazo[4,5-f]
[1,10] phenanthroline (HPIP): The HPIP ligand was prepared using a
similar method to that reported in the literature [34,35]. Briefly, 1,10-
phenanthroline 5,6-dione (210 mg, 1 mmol), 4-hydroxybenzaldehyde
(146.4 mg, 1.2 mmol), ammonium acetate (1540 mg, 20 mmol), and
acetic acid (25 mL) were added to a 100 mL round-bottomed flask, and
the mixture was heated to 130 ◦C at reflux and stirred for 3 h. The so-
lution was cooled down to room temperature, and the pH was adjusted
to 7.0 with 1 M ammonia to precipitate the product. The precipitate was
collected by vacuum filtration and washed with water, methanol, and
diethyl ether. The product was dried by lyophilization.
2.2. Synthesis of Ru1, Ru2, Ru3 and Ru4
Ru(bpy)2Cl2⋅2H2O was prepared according to the literature method
[27,33]. ruthenium(III) chloride hydrate (RuCl3⋅3H2O) (780 mg, 2.98
Scheme 1. Schematic illustration of preparation and intracellular delivery of Ru-HSA/SiO2 for fluorescence imaging and photodynamic therapy.
2

H. Shi et al.
Journal of Inorganic Biochemistry 221 (2021) 111489
[Ru(bpy)2(HPIP)]Cl2 was synthesized according to the reported
15000 rpm for 10 min. The supernatant was collected, and the fluo-
rescence intensity of Ru1 was measured.
work [36], and prepared similarly to Ru1 except using HPIP instead of
bpy, and characterized by 1H NMR (Fig. S2). 1H NMR (500 MHz, d6-
DMSO) δ 10.13 (s, 1H), 9.09 (d, 2H), 8.87 (dd, 4H), 8.22 (t, 2H),
8.19–8.09 (m, 4H), 8.06 (d, 2H), 7.97–7.80 (m, 4H), 7.60 (dd, 4H), 7.35
(t, 2H), 7.03 (d, 2H).
2.5. Measurement of reactive oxygen species (ROS)
Quantification of total ROS: Samples of Ru4 and Ru4-HSA/SiO2 were
[Ru(bpy)2(dmbpy)]Cl2 (Ru3): Ru3 was prepared as previous work
[27,37], and synthesized by similar method as Ru1 using 6,6′-dimethyl-
2,2′-bipyridine (dmbpy) instead of bpy. The Ru3 was characterized by
1H NMR (Fig. S3). 1H NMR (300 MHz, D2O) δ 8.53 (d, 2H), 8.46 (d, 2H),
8.31 (d, 2H), 8.07 (m, 4H), 7.92 (m, 4H), 7.76 (d, 2H), 7.44 (t, 2H), 7.28
(d, 2H), 7.20 (t, 2H), 1.72 (s, 6H).
placed in a 96-well plate at a Ru concentration of 50 μM. After incu-
bation for 4 h, the samples were irradiated with 450 nm light (25 mW/
cm2) for different time (0, 2.5, 5, 10, 15, 20, 25, 30, 40 min). The amount
of ROS generated from irradiated Ru4 and Ru4-HSA/SiO2 was measured
using a hydroperoxide quantitative assay kit.
Measurement of singlet oxygen: Singlet oxygen was measured by
monitoring the degrading of p-nitrosodimethylaniline (RNO). Briefly,
[Ru(bpy)2(APIP)]Cl2 (Ru4): 2-(4-nitrophenyl)-1H-imidazo[4,5-f]
[1,10]-phenanthroline (NPIP) was synthesized following a similar pro-
cedure to that of HPIP except using 4-nitrobenzaldehyde instead of 4-
Ru4 or Ru4-HSA/SiO2 (OD450 = 0.1), RNO (20 μM) and histidine (10
mM) were mixed in phosphate buffer saline (PBS) solution, and irradi-
ated with 450 nm light (25 mW/cm2) for different time (0, 5, 10, 15, 20,
25, 30 min). The absorbance at 450 nm was recorded by a Bio-Rad 680
microplate reader.
hydroxybenzaldehyde.
2-(4-aminophenyl)-1H-imidazo[4,5-f][1,10]-
phenanthroline (APIP) was prepared following the reported method
[38]. Briefly, 1 g NPIP was suspended in 30 mL 1,4-dioxane, and sodium
sulfide (20 eq) was dissolved in 30 mL water. Both solutions were heated
to 80 ◦C and then mixed. The mixture was stirred continually at 80 ◦C for
24 h, then the dioxane was evaporated to precipitate the product, which
was filtered off and washed with water and ethyl acetate several times.
The pure product was obtained by recrystallization from methanol. Ru4
was synthesized similarly to Ru1 using APIP instead of bpy [38], and
characterized by 1H NMR (Fig. S4). 1H NMR (400 MHz, d6-DMSO) δ 8.89
(q, 4H), 8.22 (q, 4H), 8.10 (t, 2H), 8.00 (d, 2H), 7.86 (m, 6H), 7.59 (m,
4H), 7.35 (t, 2H), 6.73 (d, 2H), 5.74 (s, 2H).
Measurement of hydroxyl radical: The hydroxyl radical generation
from Ru4 or Ru4-HSA/SiO2 was detected by bleaching of methylene
blue. Ru4 or Ru4-HSA/SiO2 solution (Ru concentration: 50 μM) were
mixed with methylene blue solution (25 μM), and irradiated with 450
nm light (25 mW/cm2) for different time (0, 2.5, 5, 10, 15, 20, 30, 40
min). The absorbance of methylene blue was measured by a Bio-Rad 680
microplate reader.
2.6. Cellular uptake
2.3. Synthesis of Ru-HSA/SiO2 nanoparticles
HepG 2 cells were seeded in a 3.5 cm dish at a density of 3 × 105 cells
HSA (400
μ
L, 40 mg/mL), H2O (240
μ
L), Ru1, Ru2, Ru3 or Ru4 (160
per well. After incubation at 37 ◦C in 5% CO2 for 12 h, Ru1, Ru2, Ru1-
μ
L, 10 mM), 1 M sodium chloride (8
μL), and 1 M sodium hydroxide
HSA/SiO2_10-5, and Ru2-HSA/SiO2_10-5 (Ru concentration: 25
μM)
(9–10 μL) were added to a 10 mL round-bottomed flask and stirred
were added. After incubation for different time (1, 4, 8 h), the medium
was removed, and the cells were washed several times with PBS. Fluo-
rescence microscopy was used to obtain the fluorescence image of the
cells.
vigorously for 15 min. Then, 4 mL ethanol was added dropwise, and the
mixture was stirred vigorously for further 2 min. The emulsion was
dispersed by ultrasound for 1 min. Finally, 9.4 μL 8% glutaraldehyde
was added, and the solution was stirred at 500 rpm for 12 h to obtain
mixture 1.
Endocytosis inhibition: HepG2 cells were pretreated with different
endocytosis inhibitors, including chloroproamzine (20
μg/mL), ami-
Ru-HSANP: 1 mL mixture 1 was centrifuged at 15000 rpm for 10 min
to collect the Ru-HSA nanoparticles (Ru-HSANP), which were dispersed
in 1 mL ultrapure water by sonication. Ru-HSANP was washed with
water twice and finally dispersed in 1 mL ultrapure water.
loride (2 mM), or methyl-β-cyclodextrin (5 mM) for 30 min at 37 ◦C.
Cells were incubated at 4 ◦C for inhibiting endocytosis process. Then,
Ru1-HSA/SiO2_10-5 (25 μ
M) was added and incubated for 4 h at 37 ◦C.
After that, cells were harvested and analyzed by flow cytometry.
Ru-HSA/SiO2_15–5: 15 μL TEOS was injected in 1 mL mixture 1 and
stirred at 500 rpm for 5 min. After brief sonication for 30 s, 5
μL
ammonium hydroxide (NH3⋅H2O, 37%) was slowly added and stirred
vigorously at 800 rpm for 4 h. The particles were collected by centri-
fugation at 10000 rpm for 5 min and dispersed in 1 mL ultrapure water
with sonication. The Ru-HSA/SiO2_15-5 nanoparticles were washed
with water twice and then dispersed in 1 mL ultrapure water.
2.7. Subcellular colocalization
HepG 2 cells were seeded on glass-bottom culture dishes at a density
of 2 × 105 cells per well for 24 h. Ru1-HSA/SiO2_10-5 (Ru: 25
μM) was
added and incubated for 4 h. Then, the medium was removed. The cells
were incubated with LysoTracker® Green (1 mM) for 1 h, then washed
with PBS three times. Fluorescence imaging was performed by confocal
laser scanning microscopy.
Ru-HSA/SiO2_5-5, Ru-HSA/SiO2_10-5, Ru-HSA/SiO2_20-5, Ru-HSA/
SiO2_2.5-10, Ru-HSA/SiO2_5-10, Ru-HSA/SiO2_10-10, Ru-HSA/SiO2_15-
10, and Ru-HSA/SiO2_20-10 were synthesized as Ru-HSA/SiO2_15-5 by
adding 5
μ
L TEOS and 5
μ
L NH3⋅H2O, 10
μ
L TEOS and 5
L NH3⋅H2O, 5
L TEOS and 10 L NH3⋅H2O, 15 L TEOS and
L TEOS and 10 μL NH3⋅H2O, respectively.
μ
L NH3⋅H2O, 20
μ
L TEOS and 5
μ
L NH3⋅H2O, 2.5
μ
L TEOS and 10
μ
μ
L TEOS
and 10
μ
L NH3⋅H2O, 10
μ
μ
μ
μ
2.8. In-cell ROS measurement
10
μ
L NH3⋅H2O, and 20
HepG 2 cells were seeded in a 48-well plate at a density of 1.5 × 104
2.4. Ru1 release study
cells per well and cultured for 24 h. The cells were treated with Ru4 and
Ru4-HSA/SiO2 (Ru concentration: 10 μM). The Rosup (ROS generation
A Ru1 release kinetics study was conducted by measuring the fluo-
rescence intensity of Ru1 released from Ru1-HSANP, Ru1-HSA/SiO2_5-
5, Ru1-HSA/SiO2_10-5, Ru1-HSA/SiO2_15-5, Ru1-HSA/SiO2_20-5, Ru1-
HSA/SiO2_5-10, Ru1-HSA/SiO2_10-10, Ru1-HSA/SiO2_15-10, and Ru1-
agent, supplied in a ROS assay kit) was also added as a positive control.
After incubation for 2 h, the cells were irradiated by 450 nm light (25
mW/cm2) for 10 min. Then, the medium was removed, and 2,7-dichlor-
odihydrofluorescein diacetate (DCFH-DA, 20 μM) was added. After in-
HSA/SiO2_20-10. Briefly, 20
were dispersed in 200
different time (0, 1, 2, 4, 8, 12, 24, 48 h), the samples were centrifuged at
μ
L aliquots of Ru1-HSA/SiO2 solutions
cubation for 20 min, the medium was removed and washed with PBS
three times. Fluorescence imaging was performed by fluorescence
microscopy.
μ
L DMEM/FBS medium. After incubation for
3

H. Shi et al.
Journal of Inorganic Biochemistry 221 (2021) 111489
2.9. Live/dead viability assay
3. Results and discussion
HepG 2 cells were seeded in a 48-well plate at a density of 3 × 104
3.1. Preparation of Ru-loaded HSA/SiO2 nanoparticles
cells per well. After cultured for 24 h, the cells were preincubated with
Ru4 and Ru4-HSA/SiO2 (Ru concentration: 5
μ
M) for 4 h. Then, the cells
Silica-coated Ru-HSANP (Ru-HSA/SiO2) were obtained by hydro-
lyzing TEOS on the surface of Ru-HSANP. Ru-HSANP was prepared by
crosslinking an Ru-HSA mixture using glutaraldehyde in an ethanol/
water solution (5:1) (see Experimental details). Four ruthenium com-
plexes (Ru1, Ru2, Ru3 and Ru4) were synthesized as model compounds
of ruthenium polypyridyl complexes. Ru1 and Ru2 emit red
were irradiated with 450 nm light (12.5 mW/cm2) for 20 min. After
incubation for 24 h, the medium was removed, and the cells were
stained with calcein-AM (4 μM) and propidium iodide (PI, 6 μM) for 15
min. The cell fluorescence was measured by fluorescence microscopy.
Fig. 1. TEM images of Ru1-HSA/SiO2_x-y and photos of Ru1-HSA/SiO2_x-y solutions. (A) TEM images of Ru1-HSA/SiO2_x-y. x and y represent the volume (
μ
L) of
TEOS and NH3⋅H2O (37%) added in the silica coating process, respectively. x μL TEOS and y μL NH3⋅H2O were added into 1 mL Ru1-HSANP mixture 1 (see
Experimental section). After reaction for 4 h, the Ru1-HSA/SiO2 x-y nanoparticles were collected by centrifugation, washed with water three times, and finally
dispersed in 1 mL water. (B) Picture of Ru1, Ru1-HSANP, Ru1-HSA/SiO2_5-5, Ru1-HSA/SiO2_10-5, Ru1-HSA/SiO2_15-5, and Ru1-HSA/SiO2_20-5 solution. The Ru1
concentration of all samples was 200 μM.
4

H. Shi et al.
Journal of Inorganic Biochemistry 221 (2021) 111489
fluorescence, which can be used to study the cellular uptake of Ru1- or
Ru2-loaded HSA/SiO2 nanoparticles by fluorescent imaging. Mean-
while, Ru3 is a photoactive complex that exhibits dissociation under
light irradiation. The Ru3-HSA/SiO2 nanoparticles were used to study
the effect of silica coating on the photoactivation of Ru3. Ru4 has high
efficiency of singlet oxygen generation under light irradiation. The
photodynamic therapy (PDT) effect of Ru4-HSA/SiO2 was also
investigated.
0.10, Fig. 2A). The stability of Ru1-HSA/SiO2_10-5 was analyzed in
DMEM containing 10% FBS (Fig. 2B). Slight changes of particle size
were observed over 48 h, which indicated the high stability of the Ru1-
HSA/SiO2_10-5 particles in DMEM/FBS medium. The DMEM/FBS me-
dium is a mimic of plasma, and thus widely used in cell culture. The high
stability in DMEM/FBS medium indicates that the silica-coated HSANP
nanoparticles are suitable for application in drug delivery. The effects of
different amounts of silica coating on the size distribution and ζ-po-
tential of Ru1-HSA/SiO2 were analyzed (Fig. 2C, Fig. S5). The results
showed that the size of Ru-HSANP before coating was 180 nm. The silica
coating on Ru1-HSANP resulted in a size increase to 300 nm for Ru1-
HSA/SiO2_20-10, and a further increase was observed upon adding more
silica. The ζ-potential was found to increase from ꢀ 32 mV for Ru1-
HSANP to ꢀ 22 mV for Ru1-HSA/SiO2 due to SiO2 coating. Further
increasing the amount of silica had little effect on the ζ-potential of the
nanoparticles. The silica content in Ru1-HSA/SiO2 was measured by
thermogravimetric analysis (TGA), heating to 700 ◦C under air atmo-
sphere (Fig. 2D). Ru1-HSANP and SiO2 were used as control, showing
97.7% and 14.3% total mass loss, respectively. The mass loss of Ru1-
HSA/SiO2_5-5, Ru1-HSA/SiO2_10-5, Ru1-HSA/SiO2_15-5, and Ru1-
HSA/SiO2_20-5 was determined to be 83.5%, 71%, 61.7%, and 55.2%,
respectively, which indicates that the silica content in Ru1-HSA/SiO2_5-
5, Ru1-HSA/SiO2_10-5, Ru1-HSA/SiO2_15-5, and Ru1-HSA/SiO2_20-5
was 19.2%, 33.7%, 44.5%, and 52.2%, respectively.
3.2. Characterization of Ru-HSA/SiO2
The morphology of Ru1-HSA/SiO2 was analyzed by transmission
electron microscopy (TEM). The results showed that Ru1-HSANP con-
sisted of spheres with a particle size of about 100 nm. As shown in
Fig. 1A, the size of Ru1-HSA/SiO2 increased with the amount of silica.
Ru1-HSA/SiO2 presented a granular surface covered by individual small
spherical SiO2 nanoparticles, which formed core–satellite structures. By
adjusting the amount of TEOS and ammonia added in the silica coating
process, we could efficiently control the content of SiO2 on the surface of
Ru1-HSANP. Thus, the amount of SiO2 coated on Ru1-HSANP was found
to increase with the addition of TEOS. Meanwhile, at the same TEOS
concentration, the addition of ammonia increased the size and number
of silica nanoparticles. The loading of ruthenium complex in HSANP and
HSA/SiO2 was evidenced by the deep yellow colour of the particles
(Fig. 1B). Centrifugation of a Ru1-HSA/SiO2 solution led to colorless and
clear supernatants, and the Ru1 complex was completely coprecipitated
with the HSA/SiO2 nanoparticles. The Ru1-HSA/SiO2 precipitate could
also be resuspended and well dispersed in water. These results indicate
that the Ru1 complex was stably loaded in HSA/SiO2 in water solution.
Dynamic light scattering (DLS) measurements were used to deter-
mine the hydrodynamic diameter and zeta potential (ζ-potential) of
Ru1-HSA/SiO2. The Ru1-HSA/SiO2_10–5 nanoparticles were used for
cellular uptake study; therefore, we investigated their size distribution
and stability in DMEM/FBS solution. The results showed that the
average hydrodynamic radius of Ru1-HSA/SiO2_10-5 in solution was
200 nm with a narrow size dispersion (polydispersion index (PDI) =
3.3. Ru loading and release study of Ru1-HSA/SiO2
The effects of Ru loading and release efficiency from Ru-HSA/SiO2
by silica coating were analyzed. The Ru1 loading results showed that
52% Ru1 was loaded in Ru1-HSANP. Interestingly, the loading effi-
ciency increased to nearly 90% with a small amount silica coating (few
silica nanoparticles were observed on the HSANP surface as shown in the
TEM images of Ru1-HSA/SiO2_2.5-10). As the amount of silica
increased, the Ru1 loading efficiency could reach nearly 100% (Fig. 3A).
In the absence of silica coating, Ru1 complexes were easily leaked out
from Ru1-HSANP; an 82% Ru leakage was observed within 15 min in
Fig. 2. Characterization of Ru1-HSA/SiO2. (A) Size
distribution of Ru1-HSA/SiO2_10-5 measured by DLS.
(B) Time-dependent size variation of Ru1-HSA/
SiO2_10-5 measured in DMEM medium containing
10% FBS. (C) Size distribution of Ru1-HSANP, Ru1-
HSA/SiO2_5-5,
Ru1-HSA/SiO2_10-5,
Ru1-HSA/
SiO2_15-5, Ru1-HSA/SiO2_20-5, Ru1-HSA/SiO2_5-10,
Ru1-HSA/SiO2_10-10, Ru1-HSA/SiO2_15-10, and
Ru1-HSA/SiO2_20-10. (D) Thermogravimetric anal-
ysis (TGA) of Ru1-HSANP, Ru1-HSA/SiO2_5-5, Ru1-
HSA/SiO2_10-5,
SiO2_20-5.
Ru1-HSA/SiO2_15-5,
Ru1-HSA/
5

H. Shi et al.
Journal of Inorganic Biochemistry 221 (2021) 111489
Fig. 3. Ru1 loading and release study. (A) Effect of different amounts of silica coating on Ru1 loading efficiency in HSA/SiO2 (black line). Different volumes of TEOS
and 10 L NH3⋅H2O (37%) were added in 1 mL mixture 1 (see Experimental section). After reacting for 4 h, the supernatant was collected by centrifugation, and the
absorbance of the residual Ru1 was measured at 450 nm. Ru1 leakage from Ru1-HSA/SiO2 (red line). 50 L Ru1-HSANP, Ru1-HSA/SiO2_5-10, Ru1-HSA/SiO2_10-10,
Ru1-HSA/SiO2_15-10, and Ru1-HSA/SiO2_20-10 were dispersed in 200 L PBS solution and incubated for 15 min. The supernatants were collected by centrifugation,
and the absorbance of the released Ru1 was measured at 450 nm. (B) Ru1 release kinetics from Ru1-HSA/SiO2. 25 L Ru1-HSA/SiO2 with different amounts of silica
μ
μ
μ
μ
coating was dispersed in 200 μL DMEM/FBS solution and incubated for different periods of time. The supernatants were collected by centrifugation, and the
fluorescence of the released Ru1 was measured. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of
this article.)
Fig. 4. (A) UV–vis absorption spectra of Ru3 in the dark or irradiated for 5, 10, 20, and 30 min with a LED white light. (B) Fluorescence measurement of DNA
reactions of Ru3 using EtBr as probe. Reactions were performed with herring sperm DNA (0.01 mg) at a Pt/nucleotide or Ru/nucleotide ratio of 1:1 in 10 mM
phosphate buffer (pH 7.4) containing 10 mm NaClO4 at 37 ◦C. EtBr (0.04 mg) was added before the fluorescence measurements. (C) Fluorescence emission spectra of
Ru1-HSA/SiO2. The fluorescence of Ru1-HSANP, Ru1-HSA/SiO2_5-5, Ru1-HSA/SiO2_10-5, Ru1-HSA/SiO2_15-5, and Ru1-HSA/SiO2_20-5 was measured at a Ru
concentration of 60 μM. (D) Effect of Ru3 loaded in HSA/SiO2 on the photoactivation of Ru3. The absorbance of Ru3, Ru3-HSANP, Ru3-HSA/SiO2_5-5, Ru3-HSA/
SiO2_10-5, Ru3-HSA/SiO2_15-5, and Ru3-HSA/SiO2_20-5 at 450 nm was measured after LED white light irradiation (15 mW/cm2).
6

H. Shi et al.
Journal of Inorganic Biochemistry 221 (2021) 111489
PBS. The Ru1 leakage could be efficiently prevented by silica coating;
the higher the amount of silica, the lower the amount of Ru1 leakage
from the carriers (Fig. 3A). The kinetics of Ru1 release from Ru1-HSA/
SiO2 were investigated in DMEM/FBS medium (Fig. 3B). The results
showed that 96% Ru1 was burst released from Ru1-HSANP within 1 h.
Noteworthily, we could efficiently control the Ru release rate by surface
modification with silica, specifically, by controlling the amount of silica.
The similar drug loading and release results were found for Ru2-HSA/
SiO2 and Ru4-HSA/SiO2 (Fig. S6), demonstrating a drug delivery system
with a desired drug release rate.
was found to increase in a time-dependent manner. The cellular uptake
of Ru1 was also analyzed by flow cytometry to find that the cellular
uptake increased with the concentration of Ru1-HSA/SiO2_10-5 (Fig. 5B)
and with incubation time within 8 h (Fig. 5C). The efficient cellular
uptake of Ru2-HSA/SiO2 and Ru4-HSA/SiO2 were also observed by flow
cytometry (Fig. S7B, S7C). The cellular uptake of Ru1-HSA/SiO2, Ru2-
HSA/SiO2, Ru3-HSA/SiO2, and Ru4-HSA/SiO2 were also investigated by
inductively coupled plasma mass spectrometry (ICP-MS) (Fig. S7D).
These results further demonstrated that the Ru1-HSA/SiO2, Ru2-HSA/
SiO2, Ru3-HSA/SiO2, and Ru4-HSA/SiO2 could be internalized by cells.
The cellular distribution of Ru1-HSA/SiO2_10-5 was investigated by
confocal laser scanning microscopy (CLSM). Cells were treated with
Ru1-HSA/SiO2_10-5 and then stained with commercial LysoTracker®
Green (LTG). A bright yellow fluorescence was observed from the
merged image of HepG2 cells (Fig. 5F). The colocalization coefficient of
Ru1-HSA/SiO2_10-5 and LTG was as high as 0.85. These results indicate
that the Ru1-HSA/SiO2_10-5 nanoparticles in the cells were mainly
located in the lysosomes. The cellular internalization of Ru1-HSA/
SiO2_10–5 suggests the potential use of the HSA/SiO2 nanoparticles for
ruthenium polypyridyl complexes delivery.
3.4. Effect of silica coating on Ru complexes
Photoactive ruthenium polypyridyl complex Ru3 was loaded in
HSA/SiO2. The light-induced activation of Ru3 was measured by UV–vis
absorption spectra. As shown in Fig. 4A, the peak at 450 nm decreased
gradually upon light irradiation, and a new peak at 485 nm increased.
Light irradiation induced the dissociation of the dmbpy ligand from Ru3
as previously reported [25], affording a [Ru(bpy)2(H2O)2]2+ product
that could be expected to react with DNA. To verify the hypothesis, we
studied the reaction of Ru3 with fish sperm DNA, which was monitored
by fluorescence spectroscopy using ethidium bromide (EtBr) as a probe
(Fig. 4B). EtBr is a DNA intercalator, and the EtBr–DNA complex exhibits
a strong fluorescence at 592 nm. However, a DNA crosslinking agent
such as cisplatin can prevent intercalation of EtBr into DNA, which re-
sults in quenching of the fluorescence of the EtBr–DNA complex. The
control experiment using cisplatin clearly showed that DNA platination
induced a slow decrease in the fluorescence. In contrast, in the dark, Ru3
did not cause any change in the fluorescence of EtBr–DNA complex,
indicating that Ru3 was unreactive toward DNA. However, upon light
irradiation, Ru3 quickly induced a fluorescence decrease, which sug-
gests that Ru3 was activated by light irradiation and reacted with DNA
much more rapidly than cisplatin.
The cellular uptake mechanism of Ru1-HSA/SiO2 was investigated at
low incubation temperature and by treatment with three endocytosis
inhibitors, such as amiloride, chlorpromazine (CPZ), or methyl-
β-cyclodextrin (m-β-CD), which inhibited the formation of micro-
pinocytosis, caveolae, and clathrin vesicles, respectively [39,40]. The
uptake of Ru1-HSA/SiO2 was significantly reduced at low temperature
(4 ◦C) in comparison with that at 37 ◦C (Fig. 5D, E), which indicates that
it was an energy-dependent endocytosis process [41]. When the cells
were pretreated with the endocytosis inhibitors, the extent of cellular
uptake of the particles was reduced about 97% for CPZ treatment,
whereas treatment with amiloride and m-β-CD had little effect on the
cellular uptake (Fig. 5D, E). The results suggest that the internalization
of Ru-HSA/SiO2 proceeded mainly through clathrin-mediated
endocytosis.
Fluorescent Ru1 was loaded in HSA/SiO2, and the fluorescence of the
resulting Ru-HSA/SiO2 nanoparticles was investigated. The fluorescence
of Ru1-HSA/SiO2_5-5, Ru1-HSA/SiO2_10-5, Ru1-HSA/SiO2_15-5, and
Ru1-HSA/SiO2_20-5 changed only slightly compared with that of Ru1-
HSANP (Fig. 4C), indicating that the silica coating did not affect the
fluorescence of the ruthenium complexes. The effect of loading Ru3 in
HSA/SiO2 on the photoactivation of Ru3 was also investigated. Ru3-
HSA/SiO2 having different amounts of silica coating were prepared.
Then, the absorbance of Ru3-HSA/SiO2 was measured at 450 nm upon
light irradiation at different periods of time. The results (Fig. 4D) show
that the half-life (t1/2) of the photoinduced dissociation of Ru3, Ru3-
HSANP, Ru3-HSA/SiO2_5-5, Ru3-HSA/SiO2_10-5, Ru3-HSA/SiO2_15-5,
and Ru3-HSA/SiO2_20-5 was about 0.94, 3.63, 4.15, 9.22, 18.94, and
24.62 h, respectively. These results indicate that the photosensitivity of
Ru3 decreased by loading in HSANP and HSA/SiO2; the higher amount
of silica coating, the weaker photosensitivity of Ru3 in Ru3-HSANP.
From another perspective, the silica coating strategy can be used to
protect the photosensitive complexes from light-induced degradation.
This result is consistent with previously reported work showing that
diatom SiO2 frustules can decrease the rate of ultraviolet light-induced
degradation of DNA inside the cells [29].
3.6. ROS generation from Ru4-HSA/SiO2 in vitro and in cells
To investigate whether the Ru complexes loaded in HSA/SiO2 can
retain their function of ROS generation, the Ru4 complex was chosen for
analysis due to its high efficiency of singlet oxygen (1O2) generation
[38]. In the present work, Ru4 generated 1O2 when the Ru4-HSA/SiO2
nanodrugs were irradiated with 450 nm light. Due to its extremely high
reactivity, the generated 1O2 can react with the amino acid residues of
the HSA mixed with Ru4 in the HSA/SiO2 nanoparticles. It has been
reported that tryptophan, tyrosine, and histidine could react with 1O2 to
produce other ROS such as radicals or peroxides through an electron
transfer mechanism [42,43]. Therefore, we presumed that other ROS
species apart from 1O2 could be generated from irradiated Ru4-HSA/
SiO2. To verify this, a hydrogen peroxide assay kit was used to detect
total ROS generation from Ru4-HSA/SiO2. The kit generates Fe3+ by
oxidizing Fe2+, which then reacts with xylenol orange to form a purple
product, thereby enabling the determination of the hydrogen peroxide
concentration. So the total ROS species generated from the irradiated
Ru4-HSA/SiO2 can be measured by the kit. The results (Fig. 6A, B) show
that Ru4 and Ru4-HSA/SiO2 did not generate ROS in the dark. After light
irradiation, ROS were detected, and were found to increase with irra-
diation time and Ru concentration. Interestingly, Ru4 produced lower
levels of ROS than Ru4-HSA/SiO2, which was probably caused by the
3.5. Cellular uptake of Ru-HSA/SiO2
The cellular uptake of Ru-HSA/SiO2_10-5 on HepG2 cells was
investigated by fluorescent imaging (Fig. 5). The fluorescence emission
of Ru1 and Ru2 was not affected by loading in HSA/SiO2_10-5 nano-
particles. Therefore, the cellular internalization of Ru and Ru-HSA/SiO2
could be monitored using fluorescence microscopy (Fig. 5A). The results
clearly show that free Ru1 and Ru2 had weak red fluorescence within 8
h, indicating the low cellular uptake efficacy for both complexes. The
red fluorescence was clearly observed within 8 h by using the HSA/SiO2
carrier to load the Ru complexes, and the cellular fluorescence intensity
inactivity of 1O2 due to its short lifetime (4
μs in water). In contrast, the
1O2 produced from Ru4-HSA/SiO2 can react with the amino acid resi-
dues of HSA to generate other ROS species with longer lifetime, thereby
oxidizing more Fe2+ to Fe3+ than free Ru4 under light irradiation.
To confirm the generation of other ROS species from the reaction of
1O2 with HSA, the ROS produced from Ru4 mixed with different con-
centrations of HSA was analyzed. Fig. 6C shows that the amount of ROS
increased with the HSA concentration, which fully demonstrates that the
7

H. Shi et al.
Journal of Inorganic Biochemistry 221 (2021) 111489
Fig. 5. Cellular uptake of Ru-HSA/SiO2. (A)
Cellular uptake of Ru-HSA/SiO2 measured by
fluorescence microscopy. HepG2 cells were
treated with Ru1, Ru2, Ru1-HSA/SiO2_10-5, and
Ru2-HSA/SiO2_10-5 (Ru concentration: 25
μM)
for 1, 4, 8 h. (B) Cellular uptake of Ru1-HSA/
SiO2 analyzed by flow cytometry. HepG2 cells
were treated with different concentrations of
Ru1-HSA/SiO2 (Ru1 concentration: 6.25, 12.5,
25, 50
μM) for 4 h. (C) Cellular uptake of Ru1-
HSA/SiO2 (Ru1 concentration: 25
μ
M) at 1, 2, 4,
8 h measured by flow cytometry. (D) Cellular
fluorescence intensity of HepG2 cells after
treatment with Ru1-HSA/SiO2 (Ru concentra-
tion: 25 μ
M) for 4 h at 37 ◦C as determined by
flow cytometry. The cells were preincubated
with endocytosis inhibitors amiloride, CPZ, or m-
β-CD, and were also treated at 4 ◦C for compar-
ison. (E) Bar diagram showing the uptake of Ru1-
HSA/SiO2 as mean fluorescence intensity from
the results of flow cytometric analysis. (F)
Colocalization experiment of Ru1-HSA/SiO2.
HepG2 cells were incubated with Ru1-HSA/SiO2
(Ru: 25
μM) for 4 h and then stained with
LysoTracker® Green (LTG). The fluorescence
imaging was measured by confocal laser scan-
ning microscopy (CLSM). (For interpretation of
the references to colour in this figure legend, the
reader is referred to the web version of this
article.)
8

H. Shi et al.
Journal of Inorganic Biochemistry 221 (2021) 111489
Fig. 6. ROS generation from Ru4-HSA/
SiO2. (A) Measurement of ROS generated
from Ru4 and Ru4-HSA/SiO2 (Ru concen-
tration: 50 μM) under 450 nm light irra-
diation (25 mW/cm2) for 0, 2.5, 5, 10, 15,
20, 25, 30, 40 min. (B) ROS measurement
from Ru4 and Ru4-HSA/SiO2 under 450
nm light irradiation (25 mW/cm2) for 30
min at different Ru concentrations (0.8,
1.6, 3.2, 6.2, 12.5, 25, 50, 100, 200
μM).
(C) ROS generation from Ru4-HSA
mixture. Ru4 (50
μM) was mixed with
different amounts of HSA (0, 0.5, 1, 2 mg/
mL). The mixtures were irradiated with
450 nm light (25 mW/cm2) for 30 min,
and the ROS was analyzed using
a
hydrogen peroxide assay kit. (D) Estima-
tion of 1O2 generation from Ru4 and Ru4-
HSA/SiO2 (Ru concentration: 6.7
μM)
based on the changes of A/A0 ratio of the
p-nitrosodimethylaniline/histidine (RNO/
His) system at 440 nm under light irradi-
ation. (E) The generation of hydroxyl
radical from Ru4 and Ru4-HSA/SiO2 in
dark was monitored by the bleaching of
methylene blue. (F) The generation of hy-
droxyl radical from Ru4 and Ru4-HSA/
SiO2 upon light irradiation was analyzed
by the bleaching of methylene blue. (G)
ROS detection of Ru4-HSA/SiO2 in HepG2
cells by fluorescence imaging. Cells were
treated with Ru4 and Ru4-HSA/SiO2 (Ru
concentration: 10 μM) for 2 h, and were
then irradiated with 450 nm light for 10
min. The cells were further incubated with
DCFH-DA (20 μM) for 20 min. (For inter-
pretation of the references to colour in this
figure legend, the reader is referred to the
web version of this article.)
9

H. Shi et al.
Journal of Inorganic Biochemistry 221 (2021) 111489
1O2 from Ru4 could react with HSA to produce other ROS species. Then,
p-nitrosodimethylaniline/histidine (RNO/His) was used as 1O2 scav-
enger to determine the 1O2 generated from Ru4 and Ru4-HSA/SiO2. This
method is based on the spectrophotometrical monitoring of the
bleaching of RNO at 440 nm, which is induced by the reaction of 1O2
with histidine. This reaction produces a peroxide intermediate that is
capable of bleaching RNO [44]. The result (Fig. 6D) shows that both Ru4
and Ru4-HSA/SiO2 could induce a decrease in the absorbance of RNO/
His at 440 nm, being the decrease induced by Ru4 larger than that by
Ru4-HSA/SiO2. This result suggests that Ru4-HSA/SiO2 produced less
1O2 than Ru4, probably due to the reaction of 1O2 with HSA.
suggesting that Ru4-HSA/SiO2 could generate hydroxyl radical under
light irradiation. These studies indicated that 1O2 and hydroxyl radical
were produced from Ru4-HSA/SiO2 upon irradiation.
The cellular ROS generation was analyzed by DCFH-DA assay. DCFH-
DA is a cell permeable dye that can be activated by ROS to produce a
highly fluorescent agent in cells. After incubation with Rosup, Ru4, and
Ru4-HSA/SiO2, the cells were treated with the DCFH-DA dye and sub-
jected to fluorescence microscopy measurement. The cells treated with
Rosup as a ROS generation agent showed an intense green fluorescence,
whereas a weak green fluorescence was observed in the negative control
or in cells treated with Ru4 and Ru4-HSA/SiO2 in the dark (Fig. 6G),
indicating a low ROS level in cells. However, irradiation with 450 nm
light resulted in a strong green fluorescence in cells as positive control,
indicating the generation of ROS. These results suggest that the Ru
complexes loaded in HSA/SiO2 retained the function of ROS generation.
To investigate whether there are other ROS species apart from 1O2
that could be generated from the irradiated Ru4-HSA/SiO2, the gener-
ation of hydroxyl radical from Ru4-HSA/SiO2 was studied by methylene
blue (MB) degradation experiment. MB is usually used as a chemical
probe for hydroxyl radical, because it can be specially oxidated by hy-
droxyl radical, resulting in the decrease in the absorbance of MB at 654
nm. The result showed that the Ru4 in dark or upon light irradiation
could not oxidize the MB (Fig. 6E), while Ru4-HSA/SiO2 upon light
irradiation induced the decrease of absorption peak of MB (Fig. 6F),
3.7. Photodynamic effect of Ru4-HSA/SiO2
The photodynamic therapy (PDT) effect of Ru4-HSA/SiO2 on HepG2
and A549 cells was assessed with MTT assay. We tested the dark toxicity
Fig. 7. Photodynamic therapy effect of Ru4-HSA/SiO2. MTT assay of cancer cells incubated with Ru4 and Ru4-HSA/SiO2 in the dark and under light irradiation
(12.5 mW/cm2) for 20 min. (A) HepG2 cells. (B) A549 cells. (C) Live/dead assay of HepG2 cells stained with calcein-AM (green)/PI (red) after incubation with Ru4
and Ru4-HSA/SiO2 (Ru concentration: 5
μ
M) in the dark and under light irradiation (12.5 mW/cm2) for 20 min. (For interpretation of the references to colour in this
figure legend, the reader is referred to the web version of this article.)
10

H. Shi et al.
Journal of Inorganic Biochemistry 221 (2021) 111489
of Ru4 and Ru4-HSA/SiO2 at a concentration up to 40
μ
M. Only a weak
of silica. Thus, the HSA/SiO2 nanoparticles can be potentially used for
controlled delivery of hydrophobic and hydrophilic drugs.
cytotoxic effect was found, which was weaker for the Ru4-HSA/SiO2
system than for Ru4 (Fig. 7A, B). The results suggest that the Ru4-HSA/
SiO2 system had good biocompatibility in the dark. In contrast, the
survival rate of cells upon treatment with Ru4 and Ru4-HSA/SiO2
decreased significantly under light irradiation. The IC50 values of Ru4-
HSA/SiO2 against HepG2 and A549 cells under light irradiation were
In summary, we developed a new nanostructure consisting of a
core–satellite assembly of small silica nanoparticles on a single HSANP
for loading ruthenium polypyridyl complexes. The silica coating im-
proves the Ru loading efficiency, and prevents Ru burst release from Ru-
HSANP. Meanwhile, the Ru release rate can be controlled by adjusting
the amount of silica coating, which can in turn be controlled by
adjusting the amount of TEOS and ammonia added in the silica coating
process. Therefore, a delivery system for ruthenium polypyridyl com-
plexes in which the release rate can be controlled is presented. The
fluorescent intensity of Ru was not affected when Ru was loaded in HSA/
SiO2. The cellular red fluorescence of Ru was clearly observed, which
indicates that the HSA/SiO2 carrier can efficiently deliver Ru into cells.
The internalization of Ru-HSA/SiO2 was mainly clathrin-mediated
endocytosis, and the nanosystem was mostly located in the lysosomes.
The loading of a photosensitive Ru3 in HSA/SiO2 induced a decrease in
the photosensitivity of Ru3, which was more pronounced as the amount
of silica coating increased. Therefore, the strategy of silica coating for
encapsulating photosensitive complexes can be used to protect them
against light-induced degradation. The Ru4 complexes loaded in HSA/
SiO2 retained their function of ROS generation in vitro and in cells.
Interestingly, both 1O2 and other ROS species were detected from irra-
diated Ru4-HSA/SiO2. Ru4-HSA/SiO2 had weak dark toxicity, whereas
Ru4-HSA/SiO2 showed excellent PDT efficiency under light irradiation.
Taken these results together, it can be concluded that the HSA/SiO2
nanosystem can efficiently deliver Ru complexes into cells for bio-
imaging and photodynamic therapy. It also enables the control of the
drug release rate by adjusting the amount of silica, which is suitable for
delivery of various drugs for cancer therapy. Furthermore, HSA and SiO2
are highly biocompatible, which renders the diatom-like HSA/SiO2
nanostructure ideal as a versatile nanocarrier for drug delivery.
2.18 and 1.76
μM, and those of Ru4 were 1.60 μM for HepG2 and 1.04
μ
M for A549. Although the Ru4-HSA/SiO2 showed the similar photo-
toxicity to that of Ru4 in cell lines, nanodrug can increase drug circu-
lation time, enhance drug efficacy, and reduce side effects. Importantly,
nanodrug can passively accumulate to tumor site by the enhanced
permeation and retention (EPR) effect. Thus, the Ru4-HSA/SiO2 prob-
ably showed much better photodynamic effect in vivo. The light-induced
cell death and low dark toxicity of Ru4-HSA/SiO2 suggest its potential as
an effective PDT agent.
To confirm these results, the PDT effect of Ru4-HSA/SiO2 was also
evaluated on HepG2 cells by a live/dead assay. Cells were stained by a
green fluorescence dye (calcein-AM) and a red fluorescence dye (pro-
pidium iodide) to visualize the live and dead cells via fluorescence im-
aging. As shown in Fig. 7C, almost all cells treated with Ru4 and Ru4-
HSA/SiO2 in the dark were alive and showed an intense green fluores-
cence, which is indicative of the biocompatibility of Ru4 and Ru4-HSA/
SiO2 in this condition. In contrast, under light irradiation, the Ru4 and
Ru4-HSA/SiO2 groups displayed obvious red fluorescence, suggesting
that Ru4-HSA/SiO2 can efficiently kill cancer cells by PDT. The live/
dead assay results were consistent with those of the MTT assay, both
confirming that Ru4-HSA/SiO2 exhibited excellent PDT effects.
4. Discussion and conclusion
Silica nanoparticles (SiNPs) are widely investigated at the preclinical
stage and are used in therapeutic and diagnostic applications including
bioimaging, biosensing, target drug delivery, stimuli-responsive de-
livery, and gene transfection [45,46]. In addition, SiNPs are attractive
materials due to a variety of physiochemical properties such as
biocompatibility, biodegradability, and size, shape, and porosity
tunability, which render them suitable for biomedical application [47].
Diatom microalgae consisted of SiO2 has been widely used as carrier for
delivery of drugs [48]. Ruthenium complexes were also reported to be
delivered by diatom microalgae [49]. Moreover, surface modification of
SiNP, which has a significant influence in circulation, targeting, drug
release, and cellular internalization, is easily achieved by reaction with
commercially available alkoxysilane agents. SiNP can be decorated with
various molecules, such as polyethylene glycol (PEG), polymers, pep-
tides, proteins, aptamers, saccharides, and small molecules [50]. In this
work, Ru-HSA/SiO2 was modified with PEG and target molecules for its
use in theranostic application.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Acknowledgements
This work was supported by the National Natural Science Foundation
of China (21671137, 21877103, 81602344, 21907069), the Project of
Natural
Science
Foundation
of
Guangdong
Province
(2019A1515011958), National Key R&D Program of China
(2017YFA0505400), China Postdoctoral Science Foundation
(2017M612746), the PhD Start-up Fund of Natural Science Foundation
of Guangdong Province (2017A030310445) and Collaborative Innova-
tion Center of Suzhou Nano Science and Technology for support.
HSA is the most abundant plasma protein, and is essential for
transporting numerous endogenous molecules. Due to its high solubility,
stability, biocompatibility, and biodegradability, HSA has been widely
used as a versatile nanocarrier for delivery of a large variety of drugs or
diagnostic agents toward biomedical application [51,52]. Moreover, an
Food and Drug Administration (FDA)-approved albumin paclitaxel
nanoformulation (Abraxane), in which paclitaxel is bound to HSA via a
hydrophobic interaction, has been successfully applied for treating
various types of cancer such as metastatic breast cancer, non-small cell
lung cancer, and pancreatic cancer [51]. Albumin-based nanoparticles
are normally used to deliver lipophilic drugs because the latter can be
stably and efficiently bound to the hydrophobic cavity of albumin.
However, the encapsulation efficiency of albumin nanoparticles for
hydrophilic drugs is relatively low. In addition, hydrophilic drugs tend
to undergo burst release from HSA nanoparticles. In this work, the
diatom-inspired nanostructure of silica-coated HSA nanoparticles can
prevent the burst release of Ru-based hydrophilic drugs. Especially, the
drug release rate can be accurately controlled by adjusting the amount
Author statement
Hongdong Shi: Conceptualization, Investigation, Writing - original
draft. Jingxue Lou: Data curation, Investigation. Simin Lin: Visualiza-
tion, Software. Yi Wang: Software, Formal analysis. Yatao Hu: Software,
Validation. Pingyu Zhang: Validation. Yangzhong Liu: Supervision,
Writing - review & editing. Qianling Zhang: Project administration,
Writing - review & editing.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.jinorgbio.2021.111489.
11

H. Shi et al.
Journal of Inorganic Biochemistry 221 (2021) 111489
[29] E.V. Armbrust, Nature 459 (2009) 185–192.
References
[30] J.A. Raven, A.M. Waite, New Phytol. 162 (2004) 45–61.
[31] C.E. Hamm, R. Merkel, O. Springer, P. Jurkojc, C. Maier, K. Prechtel, V. Smetacek,
Nature 421 (2003) 841–843.
[1] C.A. Rabik, M.E. Dolan, Cancer Treat. Rev. 33 (2007) 9–23.
[2] L. Kelland, Nat. Rev. Cancer 7 (2007) 573–584.
[32] L.E. Aguirre, L.Q. Ouyang, A. Elfwing, M. Hedblom, A. Wulff, O. Inganas, Sci. Rep.
8 (2018) 5138.
[3] J.M. Rademaker-Lakhai, D. van den Bongard, D. Pluim, J.H. Beijnen, J.H.
M. Schellens, Clin. Cancer Res. 10 (2004) 3717–3727.
[33] B.P. Sullivan, D.J. Salmon, T.J. Meyer, Inorg. Chem. 17 (1978) 3334–3341.
[34] J.Q. Wang, P.Y. Zhang, C. Qian, X.J. Hou, L.N. Ji, H. Chao, J. Biol. Inorg. Chem. 19
(2014) 335–348.
[4] V. Brabec, O. Novakova, Drug Resist. Update 9 (2006) 111–122.
[5] C. Scolaro, A. Bergamo, L. Brescacin, R. Delfino, M. Cocchietto, G. Laurenczy, T.
J. Geldbach, G. Sava, P.J. Dyson, J. Med. Chem. 48 (2005) 4161–4171.
[6] A. Amin, M.A. Buratovich, Mini-Rev. Med. Chem. 9 (2009) 1489–1503.
[7] A.O. Adeloye, P.A. Ajibade, Molecules 19 (2014) 12421–12460.
[8] K. Ocakoglu, S. Okur, Sensor Actuat. B Chem. 151 (2010) 223–228.
[9] I. Dragutan, V. Dragutan, A. Demonceau, Molecules 20 (2015) 17244–17274.
[10] F.E. Poynton, S.A. Bright, S. Blasco, D.C. Williams, J.M. Kelly, T. Gunnlaugsson,
Chem. Soc. Rev. 46 (2017) 7706–7756.
[35] A. Byrne, C. Dolan, R.D. Moriarty, A. Martin, U. Neugebauer, R.J. Forster,
A. Davies, Y. Volkov, T.E. Keyes, Dalton Trans. 44 (2015) 14323–14332.
[36] W.J. Mei, J. Liu, K.C. Zheng, L.J. Lin, H. Chao, A.X. Li, F.C. Yun, L.N. Ji, Dalton
Trans. 7 (2003) 1352–1359.
[37] B.S. Howerton, D.K. Heidary, E.C. Glazer, J. Am. Chem. Soc. 134 (2012)
8324–8327.
[38] S. Chakrabortty, B.K. Agrawalla, A. Stumper, N.M. Vegi, S. Fischer, C. Reichardt,
M. Kogler, B. Dietzek, M. Feuring-Buske, C. Buske, S. Rau, T. Weil, J. Am. Chem.
Soc. 139 (2017) 2512–2519.
[11] J.D. Knoll, C. Turro, Coordin. Chem. Rev. 282 (2015) 110–126.
[12] M. Dickerson, B. Howerton, Y. Bae, E.C. Glazer, J. Mater. Chem. B 4 (2016)
394–408.
[39] H.D. Shi, S.Z. Liu, J.J. Cheng, S.M. Yuan, Y. Yang, T.T. Fang, K.M. Cao, K.J. Wei, Q.
L. Zhang, Y.Z. Liu, ACS Appl. Mater. Inter. 11 (2019) 3645–3653.
[40] Z. Zhang, M.K.G. Jayakumar, X. Zheng, S. Shikha, Y. Zhang, A. Bansal, D.J.J. Poon,
P.L. Chu, E.L.L. Yeo, M.L.K. Chua, S.K. Chee, Y. Zhang, Nat. Commun. 10 (2019)
4586.
[13] B.M. Blunden, M.H. Stenzel, J. Chem. Technol. Biot. 90 (2015) 1177–1195.
[14] Q.Q. Cheng, Y.Z. Liu, Wires Nanomed. Nanobi. 9 (2017), e1410.
[15] Y. Chen, Y. Gao, Y. Li, K. Wang, J. Zhu, J. Mater. Chem. B 7 (2019) 460–468.
[16] P. Wang, W. Wu, R. Gao, H. Zhu, J. Wang, R. Du, X. Li, C. Zhang, S. Cao, R. Xiang,
ACS Appl. Mater. Interfaces 11 (2019) 13935–13944.
[41] S. Behzadi, V. Serpooshan, W. Tao, M.A. Hamaly, M.Y. Alkawareek, E.C. Dreaden,
D. Brown, A.M. Alkilany, O.C. Farokhzad, M. Mahmoudi, Chem. Soc. Rev. 46
(2017) 4218–4244.
[17] H. Maeda, Bioconjug. Chem. 21 (2010) 797–802.
[18] H. Maeda, Y. Matsumura, Adv. Drug Deliver. Rev. 63 (2011) 129–130.
[19] N. Soliman, G. Gasser, C.M. Thomas, Adv. Mater. 32 (2020) 2003294.
[20] J.L. Shen, H.C. Kim, J. Wolfram, C.F. Mu, W. Zhan, H.R. Liu, Y. Xie, J.H. Mai,
H. Zhang, Z. Li, M. Guevara, Z.W. Mao, H.F. Shen, Nano Lett. 17 (2017)
2913–2920.
[42] M. Gracanin, C.L. Hawkins, D.I. Pattison, M.J. Davies, Free Radic. Biol. Med. 47
(2009) 92–102.
[43] M. Lan, S. Zhao, W. Liu, C.S. Lee, W. Zhang, P. Wang, Adv. Healthcare Mater. 8
(2019), e1900132.
[21] W. Sun, S. Li, B. Haupler, J. Liu, S. Jin, W. Steffen, U.S. Schubert, H.J. Butt, X.
J. Liang, S. Wu, Adv. Mater. 29 (2017) 1603702.
[44] P.Y. Zhang, Y. Wang, K.Q. Qiu, Z.Q. Zhao, R.T. Hu, C.X. He, Q.L. Zhang, H. Chao,
Chem. Commun. 53 (2017) 12341–12344.
[22] P.Y. Zhang, J.Q. Wang, H.Y. Huang, B.L. Yu, K.Q. Qiu, J.J. Huang, S.T. Wang,
L. Jiang, G. Gasser, L.N. Ji, H. Chao, Biomaterials 63 (2015) 102–114.
[23] L.Z. He, Y.Y. Huang, H.L. Zhu, G.H. Pang, W.J. Zheng, Y.S. Wong, T.F. Chen, Adv.
Funct. Mater. 24 (2014) 2754–2763.
[45] A.C. Anselmo, S. Mitragotri, AAPS J. 17 (2015) 1041–1054.
[46] A. Liberman, N. Mendez, W.C. Trogler, A.C. Kummel, Surf. Sci. Rep. 69 (2014)
132–158.
[47] Z.X. Li, J.C. Barnes, A. Bosoy, J.F. Stoddart, J.I. Zink, Chem. Soc. Rev. 41 (2012)
2590–2605.
[24] P.Y. Zhang, H.Y. Huang, J.J. Huang, H.M. Chen, J.Q. Wang, K.Q. Qiu, D.L. Zhao, L.
N. Ji, H. Chao, ACS Appl. Mater. Inter. 7 (2015) 23278–23290.
[25] R. Chen, J.F. Zhang, J. Chelora, Y. Xiong, S.V. Kershaw, K.F. Li, P.K. Lo, K.
W. Cheah, A.L. Rogach, J.A. Zapien, C.S. Lee, ACS Appl. Mater. Inter. 9 (2017)
5699–5708.
[48] J. Delasoie, P. Schiel, S. Vojnovic, J. Nikodinovic-Runic, F. Zobi, Pharmaceutics 12
(2020) 480.
[49] J. Delasoie, J. Rossier, L. Haeni, B. Rothen-Rutishauser, F. Zobi, Dalton Trans. 47
(2018) 17221–17232.
[26] X. Li, Y.J. Zhang, H. Chen, J. Sun, F.D. Feng, ACS Appl. Mater. Inter. 8 (2016)
22756–22761.
[50] M. Vallet-Regi, M. Colilla, I. Izquierdo-Barba, M. Manzano, Molecules 23 (2018)
47.
[27] H.D. Shi, T.T. Fang, Y. Tian, H. Huang, Y.Z. Liu, J. Mater. Chem. B 4 (2016)
4746–4753.
[51] F. Liu, J. Mu, B.G. Xing, Curr. Pharm. Des. 21 (2015) 1866–1888.
[52] Q. Chen, Z. Liu, Adv. Mater. 28 (2016) 10557–10566.
[28] M. Jiang, S. Li, J. Wu, W. Li, X.A. Wen, H. Liang, F. Yang, J. Inorg. Biochem. 215
(2021) 111318.
12