Visual in vivo degradation of injectable hydrogel by real-time and [603304]

Visual in vivo degradation of injectable hydrogel by real-time and
non-invasive tracking using carbon nanodots as fluorescent indicator
Lei Wanga, Baoqiang Lia,*, Feng Xub,c, Ying Lid, Zheheng Xua, Daqing Weia,
Yujie Fenga, Yaming Wanga, Dechang Jiaa, Yu Zhoua
aInstitute for Advanced Ceramics, State Key Laboratory of Urban Water Resource and Environment, Harbin Institute of Technology, Harbin 150001, PR Ch ina
bMOE Key Laboratory of Biomedical Information Engineering, School of Life Science and Technology, Xi'an Jiaotong University, Xi'an, 710049, PR Chin a
cBioinspired Engineering and Biomechanics Center (BEBC), Xi'an Jiaotong University, Xi'an, 710049, PR China
dSino-Russian Institute of Hard Tissue Development and Regeneration, The Second Af filiated Hospital of Harbin Medical University, Heilongjiang Academy
of Medical Sciences, Harbin 150001, PR China
article info
Article history:Received 31 March 2017Received in revised form23 August 2017
Accepted 26 August 2017
Available online 26 August 2017
Keywords:
In vivo degradation
VisualizationInjectable hydrogelReal-time and non-invasiveFluorescence trackingabstract
Visual in vivo degradation of hydrogel by fluorescence-related tracking and monitoring is crucial for
quantitatively depicting the degradation pro file of hydrogel in a real-time and non-invasive manner.
However, the commonly used fluorescent imaging usually encounters limitations, such as intrinsic
photobleaching of organic fluorophores and uncertain perturbation of degradation induced by the
change in molecular structure of hydrogel. To address these problems, we employed photoluminescent
carbon nanodots (CNDs) with low photobleaching, red emission and good biocompatibility as fluorescent
indicator for real-time and non-invasive visual in vitro /in vivo degradation of injectable hydrogels that are
mixed with CNDs. The in vitro /in vivo toxicity results suggested that CNDs were nontoxic. The embedded
CNDs in hydrogels did not diffuse outside in the absence of hydrogel degradation. We had acquired
similar degradation kinetics (PBS-Enzyme) between gravimetric and visual determination, and estab-lished mathematical equation to quantitatively depict in vitro degradation pro file of hydrogels for the
predication of in vivo hydrogel degradation. Based on the in vitro data, we developed a visual platform
that could quantitatively depict in vivo degradation behavior of new injectable biomaterials by real-time
and non-invasive fluorescence tracking. This fluorescence-related visual imaging methodology could be
applied to subcutaneous degradation of injectable hydrogel with down to 7 mm depth in small animal
trials so far. This fluorescence-related visual imaging methodology holds great potentials for rational
design and convenient in vivo screening of biocompatible and biodegradable injectable hydrogels in
tissue engineering.
©2017 Elsevier Ltd. All rights reserved.
1. Introduction
Biocompatible and biodegradable hydrogels with three dimen-
sional polymeric structure have served as water-swollen gels and
received signi ficant attention for biomedical applications such as
controlled drug delivery and tissue engineering during past de-cades [1e5]. As for their tissue engineering applications, a quanti-
tative assessment of in vivo degradation of hydrogels is of great
importance especially for design of customized hydrogel with
controllable degradation rate that can match with the regenerationrate of newly generated tissues [6e8]. However, it is dif ficult to
reveal the in vivo degradation of hydrogels by in vitro degradation
models due to the complex in vivo microenvironment (synergetic
biodegradation by a variety of enzymes and cells). Currently, the
commonly used and reliable technique for quantitatively assessing
thein vivo degradation behavior of hydrogels is mainly based on
gravimetric/volume determination. However, such technique
needs to sacri fice lots of animals, which limits its wide application
in determining the degradation of hydrogels [9e11]. Therefore, to
minimize the uncontrollable parameters and intricate variability
and to reduce the amount of required animals, there is an urgent
need to develop strategies for real-time and non-invasively moni-
toring in vivo degradation. To date, non-invasive imaging tech-
niques, including magnetic resonance imaging [6,12e14], X-ray
*Corresponding author.
E-mail address: libq@hit.edu.cn (B. Li).
Contents lists available at ScienceDirect
Biomaterials
journal homepage: www.elsevier.com/locate/biomaterials
http://dx.doi.org/10.1016/j.biomaterials.2017.08.039
0142-9612/ ©2017 Elsevier Ltd. All rights reserved.Biomaterials 145 (2017) 192 e206

computed tomography [15], ultrasound imaging [16e18] and
fluorescence imaging [8,19e22], provide a reliable and ef ficient
measurement for determining the in vivo degradation of hydrogels
or a platform for continuously and noninvasively monitoring
in vitro cell viability, proliferation and chemosensitivity [23,24] .
Although magnetic resonance imaging and X-ray computed to-
mography are intriguing for real-time and non-invasively moni-
toring in vivo degradation of hydrogels due to their high spatial
resolution, they need radiopaque contrast agents for imaging and
complicated instrument.
Hence fluorescence-related imaging remains the most widely
employed technique for in vivo tracking and monitoring of hydrogel
recently. To achieve this goal, various fluorescent probes such as
organic fluorescent molecules and rare-earth upconversion nano-
particles have been developed to label hydrogels. For example, the
strategies of covalently immobilizing fluorescein to the bio-
materials have been adopted to track and quantify the in vitro or
in vivo degradation of hydrogels [7,8,19,21] . Fluorescein-5-
carboxyamido hexanoic acid was covalently attached to model
materials for in vivo tracking erosion of PEG/dextran hydrogel [8].
Similarly, rhodamine B and IR-Dye 800CW maleimide wereemployed to chemically label PEG hydrogel and hyaluronan
hydrogels, respectively, for monitoring in vivo degradation of
hydrogels [7,19] . However, the strategy based on covalently
immobilized fluorescein to hydrogel suffers from some intractable
issues, including photobleaching when undergoing long exposure
and uncertain perturbation of degradation due to the change in
molecular structure of hydrogels. Recently, silica-coated lantha-
nide-doped rare-earth nanoparticles have been directly embedded
into hydrogel to track hydrogel degradation in live tissues non-
invasively, avoiding the need for chemical bonding with hydro-
gels [22]. However, the potential biotoxicity of rare-earth nano-
particles due to long-term retention in the liver and spleen system
remains uncertain [25e27]. Inspiring by our previous finding that
hydrogel degradation behaviors could be well re flected by the
release of magnetic nanoparticle incorporated within hydrogel
matrix [28], we attempted to explore another strategy of directly
mixing biocompatible and luminescent nanoparticles with low
photobleaching for real-time and non-invasively monitoring the
in vivo degradation of hydrogels.
Serving as novel carbon nanomaterials, carbon nanodots (CNDs)
have offered potential for widespread applications in bioimaging
fields, mainly owing to their intriguing luminescent property, low
photobleaching and good biocompatibility. These outstanding
properties make CNDs promising luminescent nanomaterials for
in vivo imaging compared to conventionally employed organic dyes
and rare-earth nanoparticles [29e35]. Moreover, the large stokes
shift can provide CNDs with red emission wavelengths, allowing
better penetration ability for in vivo fluorescence bioimaging
compared to UV or visible lights that could be largely absorbed by
biomolecules in tissues [36,37] .
Here we explored CNDs as fluorescent indicator for visual
in vitro /in vivo degradation of hydrogel by fluorescence imaging in a
real-time and non-invasive manner. The in vitro /in vivo toxicity of
CNDs was evaluated by in vitro MTT assay and in vivo histopatho-
logical assay. The direct embedding of photoluminescent CNDsallowed low photobleaching, red emission and good biocompati-
bility and avoided the need for chemically bonding. The in vitro
degradation kinetics of hydrogels was investigated by gravimetric
and visual determination, respectively. The CNDs embedded in
hydrogels did not diffuse outside in the absence of hydrogel
degradation. We had established a mathematical equation for
quantitatively depicting hydrogel degradation behavior. With help
of the established mathematical equation we acquired similar
constant kvalue among gravimetric and visual determination,which offered the feasibility in obtaining quantitative degradation
behavior of hydrogels, thus representing conventional gravimetric
determination. We also applied the strategy of fluorescent imaging
forin vivo degradation of injectable hydrogels by real-time and
non-invasively fluorescence monitoring, which quantitatively de-
picts the degradation behavior of biodegradable hydrogel.
2. Experimental section2.1. Chemicals and materials
Chitosan (CS, viscosity average molecular weight M
h¼3.4/C2105,
degree of deacetylation ¼91.4%) was purchased from Qingdao
Hecreat Bio-tech company Ltd (China). Porcine skin gelatin, sodium
alginate, methacrylic anhydride (MA, 94%), photoinitiator (Irga-
cure®2959, I2959), lysozyme (chicken egg white) and fluorescein
isothiocyanate (FITC) were purchased from Sigma-Aldrich (USA).
Citric acid monohydrate and formamide were supplied by Sino-
pharm Chemical Reagent Co (China). Dialysis bags (retained mo-
lecular weight 500 and 8000 e14,000 Da) were supplied by Solarbio
(USA).
2.2. Synthesis and characterization of CNDs
CNDs were synthesized from citric acid and formamide
following a solvothermal method. Brie fly, citric acid monohydrate
(1.19 g) was dissolved in formamide (20 mL), and the mixed solu-
tion was then transferred into a Te flon-lined autoclave (50 mL).
After heating at 180/C14C for 4 h, the obtained dark red solution was
centrifuged at 8000 rpm for 30 min to remove large or agglomer-
ated deposit. The black CNDs powder (~50 mg) was obtained by
dialyzing against deionized water using a dialysis bag and lyophi-
lization. UV evis absorption spectrum of CNDs was exhibited on TU
1901 spectrophotometer and the spectrum was collected from
200 nm to 500 nm. Photoluminescence (PL) spectra were measured
on a Hitachi F-4600 fluorometer equipped with Xe lamp at ambient
conditions. Transmission electron microscopy (TEM) images of
CNDs were pictured from a transmission electron microscope (FEI
Tecnai G2 F30). Fourier transform infrared (FTIR) spectra were ob-
tained on a Perkin-Elmer Spectrum One ranging from 4000 to
500 cm/C01. X-ray photoelectron spectroscopy (XPS) spectra were
performed on an ESCALAB 250Xi X-ray photoelectron spectrometer
with Al/K aas the source, and the energy step size was set as 0.1 eV.
Atomic force microscopy (AFM) was performed on Bruker Dimen-
sion Icon.
2.3. In vitro cytotoxicity assay and cell bioimaging of CNDs
The cytotoxicity was evaluated by tetrazolium-based MTT assay
against NIH/3T3 cells. In 96 well plates, 100 mL suspension of NIH/
3T3 cells (1 /C2104cells/mL) in Dulbecco's Modi fied Eagle's Medium
supplemented with 10% (v/v) fetal bovine serum, penicillin (50 U/
mL)/streptomycin (50 mg/mL) were added to each well and incu-
bated in 5% CO 2humidi fied atmosphere at 37/C14C for 24 h. The CNDs
with different concentrations (200, 400, 600, 800 mg/mL) were
introduced into the wells and incubated for another 24 h, 36 h, 72 h,respectively. At the designated time intervals, the medium was
removed and cells were washed with phosphate-buffered saline.
Then, 20
mL of 5 mg/mL MTT solution was added to each well. The
96-well plates were further incubated for 4 h, followed by
removing the culture medium with MTT, and then 200 mL of DMSO
was added. The optical density of the mixtures at 490 nm was
measured using microplate reader. Cell viability was expressed as
percentage of absorbance relative to control ( i.e., without CNDs).
Experiment was performed in triplicates, with nine replicate wellsL. Wang et al. / Biomaterials 145 (2017) 192 e206 193

for each sample and control per assay.
NIH/3T3 cells with concentration of 1 /C2104cells/mL were
seeded in each well of 96-well plates and cultured at 37/C14C for 24 h.
DMEM medium solution of CNDs (200 mg/mL) was filtered through
a 0.22 mm membrane. The filtered fluorescent culture medium then
added to plates. After an incubation of 6 h, the medium was
removed and the cells were washed three times with PBS and kept
in PBS for bioimaging. The fluorescence images were carried out on
fluorescence microscope (Nikon Eclipse Ti S).
2.4. In vivo toxicity assessment of CNDs
All animal experimental protocols were approved by the animal
care and use regulations (Ethics Committee of Xi'an Jiaotong Uni-
versity), and the experiments were carried out under control of the
University's Guidelines for Animal Experimentation. Kunming mice
were 8 weeks of age, weighed 20 e23 g and acclimatized for 5 days
after arrival. Then mice were subcutaneously injected with CNDs
(500 mLo f1 0 0 0 mg/mL solution for each mouse, i.e., a dose of
~23 mg/kg). Mice treated with normal saline solution (without
CNDs) were used as control group. Mice were weighted every 3days for 3 weeks. Three animals from each group were sacri ficed at
pre-set time points of 1 and 14 days after injection, and various
organs (including liver, spleen, kidney, heart and lung) were
collected and scanned for the fluorescence imaging. The fluores-
cence intensity of each organ was quanti fied using the Carestream
MI software. The histopathological analysis of various organs such
as liver, spleen, kidney, heart and lung was performed by hema-
toxylin and eosin (H &E) staining. Major organs were fixed in 4%
paraformaldehyde buffer solution overnight, followed by dehy-
dration with 70% ethanol, and then paraf fin-embedded. Paraf fin
embedded tissues were cut (5 mm), stained with H &E and exam-
ined under a microscope.
2.5. Visualization of CNDs hybrid hydrogel using CNDs as
fluorescent indicator
Injectable N-methacryloyl chitosan (N-MAC) with degree of
substitution (DS) of 19%, 25% and 28% (N-MAC 19, N-MAC 25, N-
MAC 28) were synthesized according to our previous work [38].T o
optimize the concentration of CNDs in hybrid hydrogel for in vitro
and in vivo visualization, N-MAC phosphate buffer saline (PBS)
solution with different CNDs concentrations (0, 20, 50, 200, 500,
1000, and 1500
mg/mL) were prepared in PDMS mold and irradiated
under Omni Cure®S2000 spot curing system (EXFO Inc, Canada)
with an intensity of 10 mW/cm2for 30 s. Fluorescence imaging
(Pseudo-color images, 590 nm excitation wavelength with 700 nm
emission wavelength) was acquired using a small animal in vivo
fluorescence imaging system (In-Vivo FxPro; Carestream, MI, USA).
The mean fluorescence intensity was quanti fied using the Care-
stream MI software. To assess in vitro visualization of CNDs hybrid
hydrogel, patterned microgels ( i.e., concentric ring and convex)
were prepared by photolithographic method and imaged by fluo-
rescence microscope (Nikon Eclipse Ti S).
To determine the photobleaching, the CNDs hybrid hydrogel or
FITC hybrid hydrogel were prepared and irradiated using a 30 Wxenon excitation source of 365 e405 nm. The fluorescence pseudo-
color images of hybrid hydrogels were acquired using a small ani-
mal in vivo fluorescence imaging system with respect to time. The
fluorescence intensity was quanti fied using the Carestream MI
software.
A laser scanning confocal microscope (PerkinElmer Ultra VIEW
system, USA) was used to examine the homogeneity of CNDs in
hydrogels. In typical experiments, z planes covering 100
mm thick
sections of hydrogels were chosen for imaging. The x-y planeimages at different depth across the thickness of 100 mmw e r e
captured in 10 mm increments and the fluorescence emission in-
tensity was measured. The red field images were excited with a
568 nm Argon/Krypton laser with 400 mW of power (exposure
time: 0.2 s), and emissions were filtered with 605 nm band-pass
filter.
2.6. In vitro degradation of CNDs hybrid hydrogel by gravimetric
determination
In vitro degradation of CNDs hybrid hydrogels was conducted in
10 mL of phosphate buffer saline (PBS, pH ¼7.4) solution with or
without lysozyme (0.2 mg/mL) at 37/C14C. The PBS-lysozyme solution
was refreshed daily to ensure continuous enzyme activity. At pre-
set time intervals, the residual CNDs hybrid hydrogels samples
were removed from medium, gently washed with distilled water
and weighed. The weight loss ( WL) was de fined as:
WL¼ðW0/C0WtȚ=W0/C2100% (1)
where W0and Wtare the weights of samples at initial time and
time t during degradation, respectively.
Synchronously the in vitro accumulative release pro file of CNDs
in hydrogels with or without lysozyme was estimated by using
UVevis spectrophotometer (measuring the absorbance at 270 nm)
from degradation medium.
2.7. Visual in vitro degradation of CNDs hybrid hydrogel by
fluorescence tracking
Visual in vitro degradation of CNDs hybrid hydrogel (N-MAC 19,
N-MAC 25, and N-MAC 28) was performed based on the relative
grayscale change in fluorescent images. Typically, thin cuboid-
shaped (5 /C28/C21 mm) CNDs hybrid hydrogels was immersed in
10 mL of PBS solution with or without lysozyme (0.2 mg/mL) at
37/C14C. At pre-set time intervals, the fluorescent images of CNDs
hybrid hydrogels were captured using a fluorescence microscope
(Nikon Eclipse Ti S) with fixed exposure time (500 ms). The fluo-
rescence reduction (FR) was de fined as:
FR¼ðIOD 0/C0IOD tȚ=IOD 0/C2100% (2)
where IOD 0and IOD tare integrated optical density (analyzed by
Image Pro 6) of the samples at initial time and time t during
degradation, respectively.
2.8. Tissue penetration evaluation of CNDs at wavelength of 590 nm
Fresh slices of chicken chip (breast meat) with different thick-
ness (2 mm, 5 mm and 7 mm) were prepared. The CNDs hybrid
hydrogel was placed on the top of the chicken chip. The excitation
light was located beneath the chicken chip for evaluating tissue
penetration. The fluorescence images (pseudo-color images) were
obtained viasmall animal in vivo fluorescence imaging system. The
fluorescence intensity was quanti fied using the Carestream MI
software. The excitation wavelength, emission wavelength and
exposure time were set as 590 nm, 700 nm and 20 s, respectively.
2.9. Visual in vivo degradation of CNDs hybrid hydrogel by real-
time and non-invasive fluorescence tracking
All animal experimental protocols were approved by the local
animal care and use regulations (Ethics Committee of Xi'an Jiaotong
University), and the experiments were carried out under control of
the University's Guidelines for Animal Experimentation. KunmingL. Wang et al. / Biomaterials 145 (2017) 192 e206 194

mice were 8 weeks of age, weighed 20 e25 g and acclimatized for 5
days after arrival. The CNDs (1000 mg/mL) and N-MAC (N-MAC 19,
N-MAC 25, and N-MAC 28) was directly dissolved in PBS to form a
homogeneous solution (sterilized by filtration; 0.22 mm). The
Kunming mice were randomly distributed in three groups treated
with CNDs hybrid hydrogels (N-MAC 19, N-MAC 25 and N-MAC 28).
The transdermal curing hydrogels were carried out viasubcu-
taneous injection of mice. After anesthetization, 500 mL of CNDs
hybrid solution was injected into subcutaneous space of mice back
through syringe with 25G needle and then sequentially crosslinked
by UV irradiation for 30 s. Mice treated with free CNDs (CNDs so-
lution without hydrogel) were used as control group. At pre-set
time intervals, the mice were anesthetized and the fluorescence
images (pseudo-color images) were obtained viasmall animal
in vivo fluorescence imaging system. The fluorescence intensity was
quanti fied using the Carestream MI software. The excitation
wavelength, emission wavelength and fixed exposure time were set
as 590 nm, 700 nm and 20s, respectively. In order to prove the
reliability of the visual determination method, the mice in the
parallel control group were euthanized at the designed time in-
tervals and the remained hydrogels were collected, washed withPBS and then weighted to estimate the percentage of degradation.
To con firm the feasibility that this visual determination method can
be applicable to various biomaterial systems, CNDs hybrid gelatin
hydrogel and alginate hydrogel were injected into subcutaneous
space of mice back using the same procedure for visual in vivo
degradation by real-time and non-invasive fluorescence tracking.
2.10. Histological observation of CNDs hybrid hydrogel
To assess the biocompatibility of CNDs hybrid hydrogel, the mice
were sacri ficed by intraperitoneal injection of excess chloral hy-
drate at the designated time intervals of 72, 120, 192, and 288 h. The
hydrogel samples adjacent with surrounding skin was resected and
fixed immediately in 4% paraformaldehyde buffer solution. The
samples were dehydrated with 70% ethanol, and then paraf fin-
embedded. Paraf fin embedded tissues were cut (5 mm), stained
with H &E and examined under a digital microscope.2.11. Statistical analysis
All the data were expressed as means ±standard deviation of at
least triplicate samples. The statistically signi ficant difference was
evaluated by Student's T-test, and statistical signi ficance was
considered for p value <0.05, n ¼3.
3. Results and discussion
To design and apply biodegradable injectable hydrogels ratio-
nally in tissue engineering, the detailed in vivo degradation pro file
should be well known. However, conventional gravimetric and
volume determination by periodically weighting hydrogels neces-
sitates a large number of sacri fices and involves the issues of un-
controllable parameters and intricate variability [39e41].S oa n
alternative and reliable strategy, fluorescence imaging technique is
used for visual in vivo degradation of hydrogels in a real-time and
non-invasive manner. Scheme 1 illustrates the preparation of
fluorescent CNDs hybrid chitosan hydrogel (
lex¼590 nm,
lem¼700 nm) by mixing with CNDs and visual in vitro /in vivo
degradation of hybrid hydrogel by real-time and non-invasive
fluorescence tracking. We directly mixed CNDs into the UV-
crosslinkable and injectable chitosan (N-MAC) solution with
different DS, and acquired fluorescence CNDs hybrid hydrogel after
UV irradiation. Low photobleaching, red emission and goodbiocompatibility enable the CNDs to be used as promising fluo-
rescent indicator for visual in vivo degradation of hydrogel. When
applying the strategy of fluorescent imaging for in vivo quantitative
degradation of hydrogels, the fluorescence images of CNDs hybrid
hydrogels during degradation were obtained using fluorescence
microscope. A quantitative relationship between fluorescence
reduction and degradation time was established. We further
investigated the correlation between gravimetric and visual
determination in vitro and found similar degradation pro file, which
demonstrates the feasibility that strategy of fluorescent imaging for
quantitative degradation could represent conventional gravimetric
determination and avoid extensive animal sacri fices. Hence the
strategy of fluorescent imaging can also be expanded to in vivo
degradation of hydrogels by real-time and non-invasive
Scheme 1. Schematic illustration of low photobleaching, red fluorescence emission and good biocompability CNDs for visual in vitro /in vivo degradation of injectable hydrogel by
real-time and non-invasive fluorescence tracking. The visual determination (replacing conventional gravimetric determination) using CNDs as fluorescent indicator was performed
for monitoring in vitro degradation of hydrogel. This visual determination could also be expanded to quantitatively assessment of in vivo degradation of hydrogels by real-time and
non-invasive fluorescence tracking.L. Wang et al. / Biomaterials 145 (2017) 192 e206 195

fluorescence tracking, which will provide a reliable visual platform
that quantitatively assess the degradation pro file of hydrogel.
3.1. Morphology, optical and in vitro/in vivo toxicity of CNDs
We synthesized CNDs viasolvothermal method and character-
ized the morphology, surface properties and optical properties via
TEM, AFM, XPS and PL spectra. As illustrated in Fig. 1 A, the
morphology of CNDs was examined using HR-TEM. The mean
diameter of CNDs was 5.1 ±1.2 nm without aggregation, supporting
that CNDs were uniformly spherical morphology. Fast Fourier
transform revealed characteristic hexagonal diffraction pattern of
graphite, further supporting the formation of graphitic structure[42]. As shown in AFM images, the densely and well-dispersed
CNDs appeared on the silicon substrate with particle heights in-
formation about 3 e6 nm, which also consistent with particle size
from HR-TEM image. As illustrated in Fig. 1 B, the CNDs absorption
spectrum exhibited a characteristic absorption peak at 270 nm,
which are assigned to typical absorption of
p/p*electronic tran-
sition of aromatic system (suggestive of sp2carbon network). The
well-dispersed suspension showed transparent under day light and
exhibited bright blue luminescence under UV excitation (insert in
Fig. 1 B). Three dominant peaks at 286, 400 and 531 eV appeared in
XPS survey spectrum, suggesting composition of carbon/C1s, ni-
trogen/N1s, and oxygen/O1s elements in CNDs ( Fig. S1A ). The high-
resolution of C1s spectrum ( Fig. S1B ) could be fitted to three peaks
at 284.6, 285.8 and 287.9 eV, corresponding to C ]C, CeN/CeOH
and C]O bonds. The FTIR spectrum showed characteristic peaks at
3210, 1680 and 1600 which correspond to s(stretching vibration)
OeH/NeH,sC]O and NH 2band, respectively ( Fig. S2 ), indicating
the presence of hydroxyl, carboxyl and amino groups. The presenceof abundant functional groups imparted excellent solubility in
water without further chemical modi fication. The PL emission
wavelength shifted to longer wavelength as excitation wavelengths
increased from 300 to 460 nm. The strongest fluorescence emission
was observed with a peak at 460 nm when using 375 nm excitation
wavelengths ( Fig. 1 C). The absolute fluorescence quantum yield of
CNDs was measured to be 17.2%, which was comparable to previous
reports [43,44] . Notably, the large stokes shift provided CNDs with
significant bene fits (red emission) for in vivo fluorescence bio-
imaging, as red emission wavelengths would provide a deeper
penetration ability, minimal auto fluorescence and high signal-to-
noise ratios [45,46] .
The desired fluorescent indicator for visual in vivo degradation
of hydrogel should be non-toxic. Therefore, the in vitro cytotoxicity
of CNDs with different concentrations (0, 200, 400, 600 and 800 mg/
mL) was evaluated by culturing with NIH/3T3 cells for 24, 36 and
72 h using MTT method, respectively. The NIH/3T3 cells viability
with various concentrations maintained all above 90% even at 72 h
(Fig. 1 D). In addition, no statistically signi ficant differences were
observed between CNDs groups and control group in cell viability,
which indicated that CNDs exhibited good biocompatibility,potentially for live cell imaging and visual in vivo degradation of
hydrogel. The application of CNDs as live cell imaging probe was
demonstrated. As depicted in Fig. 1 E, cell uptake of CNDs was
clearly observed with bright red fluorescence at fluorescence
(540e560 nm) and bright field after incubation with 200
mg/mL
CNDs for 6 h. The result implied that CNDs can serve as fluorescent
probes for live cell bioimaging. Moreover, no auto fluorescence
emerged from cells of control group in fluorescence field, which
confirmed that bright red fluorescence was ascribed to the cell
uptake of CNDs and being internalized into the cells. More
Fig. 1. Characterization and in vitro cytotoxicity of CNDs. (A) TEM images of the CNDs (mean diameter ¼5.1±1.2 nm) with high-magni fication HR-TEM images, the insert
displays the fast-Fourier-transformed diffraction pattern. AFM topography image of CNDs on a silicon substrate, with the height pro file along the line in the topographic image; (B)
UVevis absorbance spectra of CNDs diluted suspension. Inset shows photographs of CNDs suspension under day light (left) and UV light at 365 nm (right); (C)Excitation-dependent
photoluminescence spectra of CNDs. PL emission wavelength shifts to longer wavelength (from 430 to 580 nm) as excitation wavelength increased from 3 00 to 460 nm. The
maximum emission is 450 nm when using 375 nm excitation; (D)Cell viability of NIH/3T3 cells after incubation with different concentrations CNDs for 24, 36 and 48 h, determined
by MTT assay. The cells viability with various concentrations maintains all above 90%; (E)Merge image ( fluorescence and bright field) of NIH/3T3 cells treated with CNDs.L. Wang et al. / Biomaterials 145 (2017) 192 e206 196

importantly, no morphological damage of the cells was observed
upon incubation with the CNDs further demonstrating their good
biocompatibility. The primary challenge of any fluorescent indica-
tor is to achieve high photostability and possess low cytotoxicity. So
in this regard, organic fluorescence dyes, semiconducting quantum
dots and upconversion nanoparticles were extensively studied as
bioimaging probes. However, the inherent limitations include poor
photostability of organic fluorescence dyes and potential toxicity of
heavy metal containing semiconducting quantum dots/upconver-
sion nanoparticles [47e49]. Hence, for in vivo applications, CNDs
could serve as an exciting alternative as they were nontoxic.
To determine the clearance and in vivo biocompability of CNDs
asfluorescent indicator, mice were sacri ficed after 1 and 14 days of
injection and isolated liver, spleen, kidney, heart and lung were
scanned for the fluorescence imaging. As depicted in Fig. 2 A, the
observed fluorescence from control group was so low and typical
autofluorescence of tissues [50]. Importantly, the same marginal
values of fluorescence were also observed after 1 and 14 days of
CNDs injection. No statistically signi ficant differences were
observed between CNDs group and control group in fluorescence
intensity of various organs ( Fig. 2 B). Based on the results of fluo-
rescence measurements, we could infer that the CNDs using as
fluorescent indicator would be completely removed from the
body after 14 days of CNDs injection, and long-term retention in
various organs could not occur. Over a period of 3 weeks, neither
death nor signi ficant body weight drop was noted in the CNDs
group ( Fig. S3 ), indicating that the mice of CNDs group could
continue to mature without any signi ficant toxic effects. As theinjection of CNDs may induce subsequent damage in the organs
related to nanoparticle clearance, in vivo toxicity of CNDs was
further examined in various organs (including liver, spleen, kid-
ney, heart and lung) after 1 and 14 days of CNDs injection, using
CNDs-untreated rats as the control group. The key organs from
both CNDs injection and control groups had integrated tissue
structure without edema, in flammation and abnormal defects
(Fig. 2 C), similar with previous researches about in vivo toxicity of
CNDs [51,52] . Histopathologically, no signi ficant alterations
occurred in these tissues relative to the control group, demon-
strating that no signi ficant in vivo toxicity was observed when
employing CNDs as fluorescent indicator. In both CNDs injection
and control groups, cardiac myocytes were clear and arrayed in
order without any in flammatory exudate, hemorrhage, hyper-
trophy or necrosis. The structure of liver lobules with central vein
was clearly delineated, and there were no in flammatory in-
filtrates. Additionally, the splenic corpuscle structure of spleen
was normal and clearly delineated. No pathological changes were
found in spleen sinus, the white pulp, and red pulp of spleen. The
tissue structure of lung was normal, and there were no bronchi-
oles and alveoli ectasia or collapse, alveolar epithelial denatur-ation, interstitial hyperemia, or in flammatory cell in filtration
surrounding the bronchus. From micrograph of mice kidney, the
renal glomerulus and various kidney tubes displayed normal
shape, without degeneration, bleeding or in flammatory exudate.
All above histopathological r e s u l t ss u g g e s tt h a tC N D sw e r e
nontoxic in vivo . Thus CNDs could serve as a promising fluorescent
indicator for visual in vivo degradation of hydrogel.
Fig. 2. In vivo toxicity assessment of CNDs. (A) Fluorescence imaging of various organs (including liver, spleen, kidney, heart and lung) at 1 and 14 days. Mice treated with normal
saline solution (without CNDs) were control group; (B)Quanti fication of the fluorescence intensity from various organs (including liver, spleen, kidney, heart and lung); (C)H&E
stained tissue slices (liver, spleen, kidney, heart and lung) of mice injected with CNDs solution (dose of 23 mg/kg) at 1 and 14 days.L. Wang et al. / Biomaterials 145 (2017) 192 e206 197

3.2. Visualization of CNDs hybrid hydrogel using CNDs as
fluorescent indicator
Thefluorescence pseudo-color images of CNDs hybrid hydrogel
with different concentrations of CNDs were acquired on a small
animal in vivo fluorescence imaging system. The quantitative rela-
tionship of fluorescence signal intensity versus CNDs concentration
and the pseudo-color images were shown in Fig. 3 A. The fluores-
cence signal intensity was positively correlated to CNDs concen-
tration. When CNDs concentration increased to 1000 mg/mL,
fluorescence signal intensity does not increase anymore. It was
suggested that CNDs concentration of 1000 mg/mL could give a
detectable and strong enough fluorescence signal for visualization
in vitro /in vivo . The evaluation of CNDs hybrid hydrogel for in vivo
bioimaging was investigated using subcutaneous injection into
mice back and then sequentially applied 30s UV irradiation. The
fluorescent region signals of CNDs hybrid hydrogel could be obvi-
ously observed with a distinct boundary between CNDs hybrid
hydrogel and surrounding skin tissue, which showed that CNDs
hybrid hydrogel are capable of in vivo bioimaging with long emis-
sion wavelengths ( lem: 700 nm) as shown in Fig. S4 . Importantly,
the excitation-dependent PL behavior allowed CNDs with crucial
benefits for in vivo fluorescence bioimaging owing to red emission
wavelengths with stronger penetration ability. The patterning
CNDs hybrid microgels were fabricated viaUV lithography with
help of photomask. The images of patterned microgel, such as
concentric ring pattern (left) and convex pattern (right), exhibited
green and red fluorescence under different excitation (460 e490
and 540 e560 nm), and highly tallied with photomask which
demonstrated that CNDs could be dispersed homogeneously in
patterned microgels ( Fig. 3 B).In order to take one picture, it includes several procedures (such
as pre-scan, ROI position adjustment and taking picture) and the
time required to irradiate is about 120s. Speci fically, it usually takes
several quarters required to irradiate for long-term in vivo hydrogel
tracking. So the photostability of fluorescent indicator indeed was a
non-negligible problem, especially for the biomaterial with slow
degradation. The photobleaching of CNDs was compared with FITC
under a continuous 365 nm UV lamp illumination. The images of
CNDs hybrid hydrogel ( Fig. 3 C, up) and FITC hybrid hydrogel
(Fig. 3 C, down) showed a clearly fluorescence signal. An obvious
descending of fluorescence intensity of FITC hybrid hydrogel could
be observed, while the CNDs hybrid hydrogel's fluorescence in-
tensity changed slightly. According to the vividly contrast ( Fig. 3 D),
the decay curve of FITC descended (decrease by 50%) with a much
larger amplitude than the curve of CNDs (decrease by less than
10%). It was elucidated that CNDs as fluorescent indicator were
much more photostable than fluorescent dye (such as FITC or RhB)
[53,54] .
To demonstrate the homogeneity of CNDs in hydrogels, a cuboid
region (100 /C2300/C2300mm) was chosen in hydrogels, and the x-y
plane images at different depth in the vertical cross-sections of theregions were captured. The fluorescence intensity pro files were
examined using confocal microscopy ( Fig. 3 E). The representative
x-z plane image exhibited homogeneous and bright red fluores-
cence, suggesting that the CNDs were evenly distributed in hori-
zontal cross-section. The fluorescence intensity at different depth
displays relatively consistent, further con firming the homogeneity
of CNDs in horizontal cross-sections. Additionally the representa-
tive x-y plane images at different depth also exhibited homoge-
neous and bright red fluorescence, demonstrating that the CNDs
were evenly distributed in vertical cross-sections. Thus the
Fig. 3. CNDs serving as fluorescent indicator for visualization of CNDs hybrid hydrogel. (A) Quantitative fluorescence intensity of CNDs hybrid hydrogel versus CNDs con-
centration. Inset shows fluorescent pseudo-colored image of hydrogel with different CNDs concentrations; (B)Fluorescence images of patterning CNDs hybrid microgels of
concentric ring pattern (left) and convex pattern (right), exhibiting green and red fluorescence under different excitation (460 e490 and 540 e560 nm); (C)Photobleaching
experiment of CNDs hybrid hydrogel. Fluorescent pseudo-colored image taken in every 5 min interval showed photobleaching of CNDs hybrid hydrogel or FITC hybrid hydrogel
when exposed to 100 W Xenon excitation source; (D)A speci fic region of interest (ROI) was selected for quantitative calculation of fluorescence intensity. CNDs as the fluorescent
indicator were much more photostable than fluorescent dye (such as FITC and RhB) and much more suitable for in vivo visualization research; (E)Confocal microscopy images of
CNDs hybrid hydrogel: the representative x-z plane image, the representative x-y plane images and fluorescence intensity pro file at different depth. The red field was excited with a
568 nm laser and emissions were filtered with a 605 nm band-pass filter. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version
of this article.)L. Wang et al. / Biomaterials 145 (2017) 192 e206 198

homogeneity of CNDs in vertical cross-sections and horizontal
cross-section proved the homogeneity in three-dimensional space.
Fluorescence-related imaging technique has seen revolutionary
advancements in both sensitivity and resolution over the past
decade. However, photoinduced degradation (photobleaching) of
the commonly used organic fluorescent molecule remains a key
obstacle that limits the temporal and spatial resolutions of imaging
[55]. In addition, photoinduced fluorophore toxicity (phototoxicity)
may also lead to unwanted perturbations to the biological system
that can obscure the signal of interest [56]. Conversely, CNDs
exhibit high resistance to photobleaching and possess better bio-
compability compared to organic fluorescent molecule and tradi-
tional semiconductor quantum dots and much more suitable for
visualization research [57].
3.3. Visual In vitro degradation of CNDs hybrid hydrogel
To explore the correlation between CNDs release and hydrogel
degradation by gravimetric determination in vitro , we evaluated
the degradation of CNDs hybrid hydrogels in PBS and PBS-lysozyme
solution over 500 h by synchronously recording the weight loss of
hydrogel and amount of released CNDs at regular intervals. Almost
no weight loss (no degradation) was observed during the 500 h for
all CNDs hybrid hydrogels (N-MAC 19, N-MAC 25, and N-MAC 28) in
absence of lysozyme ( Fig. 4 A), which could speci fically break
glycosidic bonds of chitosan backbone linkages [58]. More impor-
tantly, the CNDs release behavior followed a similar trend: CNDs(1%e3% releases) were hardly detected during 500 h. It was sug-
gested that CNDs release was not controlled by diffusion but is to be
related to hydrogel degradation ( Fig. 4 B). That was similar with our
previous research about magnetic nanoparticles release pro file:
only degradation of hydrogels would allow the release of the
magnetic nanoparticles [28]. It had been con firmed that there were
abundant functional groups (including hydroxyl, carboxyl and
amino groups) on the surface of CNDs ( Fig. S1 &Fig. S2 ). And chi-
tosan is a biodegradable and natural polymer with amino groups.
Thus, CNDs in hydrogels would be immobilized to the chitosan
polymer chain due to the hydrogen bonds [59,60] formed between
carboxyl groups ( eCOOH) of CNDs and amino groups ( eNH
2)o f
chitosan in the absence of hydrogel degradation. The CNDs are
retained in the hydrogel by non-covalent interactions (partially
ascribed to hydrogen bonds) [61]. The CNDs immobilized to the
chitosan polymer chain could diffuse outside from hydrogel when
degradation of chitosan hydrogel occurred. The degradation ofhydrogels is mostly ascribed to breakage of polymer backbone in
the presence of lysozyme [62,63] . The degradation of CNDs hybrid
hydrogels by gravimetric determination was followed for 500 h.
The CNDs hybrid hydrogels (N-MAC 19, N-MAC 25, and N-MAC 28)
ultimately degraded with 48%, 59% and 85% after 500 h in the
presence of lysozyme, indicating that increase of chemical cross-
linked networks density could result in a slower in vitro degrada-
tion pro file with an identical enzyme concentration ( Fig. 4 A). On
the other hand, the release behaviors of CNDs embedded in
hydrogels also exhibited a very similar trend: the accumulative
release amount of CNDs in hybrid hydrogels (N-MAC 19, N-MAC 25,
and N-MAC 28) ultimately reached 43%, 57% and 82% after 500 h in
the presence of lysozyme. Furthermore, following the first-order
enzymatic hydrolysis kinetics [64e66], a mathematical equation
forin vitro degradation of hybrid hydrogel was developed via
exponential fitting for evaluating the correlation between CNDs
release and hydrogel degradation in vitro in the presence of
enzyme: -dm E/dt¼kmE, upon integration and deformation
WL%¼/C18
1/C0WE
WE0/C19
/C2100% ¼1/C0expð/C0ktȚ (3)
Where WLis the weight loss, WEand WE0is the undegraded
weight and initial weight of hydrogel, kis the enzymatic hydrolysis
constant related to the polymeric structure, tis degradation time.
Thus, 1- mE/mE0represents the fraction of degraded hydrogel. Using
Eq.(3)tofit the degradation curves shown in Fig. 4 A, we had ac-
quired the following fitted equations, respectively: N-MAC 19
WL¼1/C0exp(-0.00527t) (R2¼0.98), N-MAC 25 WL¼1/C0exp(-
0.00245t) (R2¼0.96), N-MAC 28 WL¼1/C0exp(-0.001t) (R2¼0.94).
Furthermore, we also found that there existed an analogous and
classical Lagergren pseudo first order enzyme adsorption kinetics
[67,68] :dq/dt¼ka(qe-q), upon integration and deformation
q
qe¼1/C0expð/C0ktȚ (4)
Where qeand qis the equilibrium adsorption mass and
adsorption mass of enzyme, kais the pseudo first order adsorption
rate coef ficient. Based on the close correlation between hydrogel
degradation and enzyme adsorption, the degradation mechanism
of CNDs hybrid hydrogels could be explained that the enzyme
adsorption and contact with hydrogel internal/external surface
gradually triggered the breakage of chitosan polymer backbone. It
was demonstrated that the embedded CNDs were released from
Fig. 4. In vitro degradation of CNDs hybrid hydrogel. (A) Gravimetric degradation of CNDs hybrid hydrogel and (B)release behavior of CNDs in the absence of lysozyme or in the
presence of lysozyme over a period of 500 h. The CNDs encapsulated inside hydrogels did not diffuse outside in the absence of hydrogel degradation (NO d egradation and NO CND
release). In vitro mathematical equation was established to quantitatively depict degradation pro file.L. Wang et al. / Biomaterials 145 (2017) 192 e206 199

the hydrogels as CNDs hybrid hydrogels degraded and a positive
correlation between CNDs release and hydrogel degradation was
achieved. All these results indicated that it was possible to predict
in vitro degradation kinetics of injectable biomaterials through
employing CNDs release pro file. To con firm this, we also fit the
CNDs release curve using the established mathematical equation
and acquired the following fitted equations, respectively: N-MAC 19
AR¼1/C0exp(-0.00518t) (R2¼0.97), N-MAC 25 AR¼1/C0exp(-
0.0025t) (R2¼0.95), N-MAC 28 AR¼1/C0exp(-0.00127t) (R2¼0.93).
It should be noted that we acquired similar kvalue betweengravimetric determination (5.27 /C210/C03, 2.45 /C2/C2 10/C03and
1/C210/C03) and CNDs release (5.18 /C210/C03, 2.5 /C210/C03and
1.27/C210/C03), which further con firmed that the positive correlation
between CNDs release and hydrogel degradation. Only by degra-
dation of CNDs hybrid hydrogels can CNDs embedded in hydrogel
be released.
To test the feasibility of CNDs serving as fluorescent indicator for
monitoring in vitro degradation of hydrogel viathe strategy of
fluorescent imaging, the fluorescence images acquired with
700 nm emission ( Fig. 5 A) and quanti fiedfluorescence reduction at
Fig. 5. In vitro degradation of CNDs hybrid hydrogel by visual determination. (A) Representative fluorescent image of in vitro degradation of CNDs hybrid hydrogel (N-MAC 19,
N-MAC 25 and N-MAC 28) in the absence of lysozyme or in the presence of lysozyme over a period of 500 h; (B)Quantitative fluorescence reduction of hybrid hydrogels (N-MAC 19,
N-MAC 25, and N-MAC 28) in vitro degradation in the absence of lysozyme or in the presence of lysozyme as a function of time; (C)Correlation graphs of degradation of CNDs hybrid
hydrogel by weight loss, CNDs release and visual determination.L. Wang et al. / Biomaterials 145 (2017) 192 e206 200

0, 54, 140, 180, 300, 504 h was shown in Fig. 5 B. In absence of
enzyme, all CNDs hybrid hydrogels maintained their initial shape
andfluorescence intensity during the whole degradation process.
Slight fluorescence reduction (FR) (2% e5%) occurred ( Fig. 5 B),
indicating that almost no degradation occurred, and the CNDs did
not diffuse outside of the hydrogels. On the other hand, in the
presence of lysozyme the shape of CNDs hybrid hydrogels dimin-
ished with different degree of enzymatic degradation, indicating
the decrease of chemical cross-linked networks density could result
in faster degradation. It was quanti fied that FL of CNDs hybrid
hydrogels (N-MAC 19, N-MAC 25, and N-MAC 28) ultimately
reached a different extent: 39%, 52% and 79% after 500 h as shown
inFig. 5 B. An exponential fitting was also established for evaluating
the dependence between gravimetric degradation and visual
in vitro degradation in the presence of enzyme. We have acquired
the following fitted equations, respectively: N-MAC 19 FR¼1/C0
exp(-0.00455t) (R2¼0.97), N-MAC 25 FR¼1/C0exp(-0.00201t)
(R2¼0.96), N-MAC 28 FR¼1/C0exp(-0.00125t) (R2¼0.94). We ac-
quired similar kvalue between gravimetric determination
(5.27 /C210/C03, 2.45 /C2/C2 10/C03and 1 /C210/C03) and visual determination
(4.55 /C210/C03, 2.01 /C2/C2 10/C03and 1.25 /C210/C03), which con firmed a
close correlation between the strategy of fluorescent imaging and
gravimetric degradation. The strategy of fluorescent imaging would
potentially provide a reliable platform to predict degradation ki-
netics of injectable biomaterials.
To demonstrate the feasibility that the strategy of fluorescent
imaging indeed re flected real hydrogel degradation pro file, we
compared the correlation of classical gravimetric degradation,
CNDs release and visual determination ( Fig. 5 C). It was noteworthy
that hydrogel degradation from gravimetric degradation coincided
with CNDs release pro file and visual determination for all time
points (from 54 to 504 h). No statistically signi ficant differences
were observed in gravimetric degradation, CNDs release and visualdetermination for all time points. These matching trends between
gravimetric degradation and CNDs release suggested that CNDs
release could be used as an indication of hydrogel degradation. The
close correlation between visual determination and gravimetric
degradation further demonstrate that the strategy of fluorescent
imaging would potentially provide a reliable platform to monitor
hydrogel degradation in vivo by real-time and non-invasive fluo-
rescence tracking.
3.4. Visual in vivo degradation of CNDs hybrid hydrogel by real-
time and non-invasive fluorescence tracking
Fluorescence is a direct method that allows the imaging of a
fluorescent probe in tissue in vivo .In vivo fluorescence imaging is
possible up to a depth of several centimeters, limited by photon
absorption, scattering and diffusion. The use of red emission
wavelengths allows deeper penetration than other wavelengths.
Our experiments were performed in shaved mice to further mini-
mize auto fluorescence background from fur (including keratin,
porphyrins, collagen and elastin) [69]. To evaluate the tissue
penetration ability of CNDs at wavelength of 590 nm, the fluores-
cent images of CNDs hybrid hydrogel on the top of the chicken chip
were obtained by allowing 590 nm light pass though chicken chip
with different thickness ( Fig. 6 A). Bright fluorescent derived from
CNDs hybrid hydrogel was observed before placing chicken chip
(0 mm) under the excitation of 590 nm light. Strong fluorescent
emission could still be observed with the tissue coverage of 2 or
5 mm thickness (reduced by 15 e30%), demonstrating relatively
deep tissue penetration capability of CNDs at wavelength of
590 nm. However the fluorescent sharply reduced by 60% with the
tissue coverage of 7 mm thickness ( Fig. 6 B). The extinction coef fi-
cient ( ε) of CNDs was determined as 1.36 cm2g/C01according to Beer-
Lambert Law ( Fig. S5 ). Overall, the good biocompatibility, low
Fig. 6. Tissue penetration evaluation of CNDs at wavelength of 590 nm for visual in vivo degradation of hydrogel. (A) Illustration of the experimental setup used to estimate the
tissue penetration ability of CNDs at wavelength of 590 nm. The CNDs hybrid hydrogel was placed on the top of chicken chip with varying thickness (2 e7 mm); (B)Thefluorescence
images (pseudo-color images) of CNDs hybrid hydrogel and quantitative fluorescence intensity. The excitation wavelength, emission wavelength and fixed exposure time were set as
590 nm, 700 nm and 20s, respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)L. Wang et al. / Biomaterials 145 (2017) 192 e206 201

photobleaching and relatively deep tissue penetration would
enable CNDs to serve as promising fluorescent indicator for long-
term in vivo hydrogel tracking.
To investigate visualization and quantitation of in vivo degra-
dation, CNDs hybrid hydrogels different DS were subcutaneously
injected and transdermally UV cured to distinguish in vivo fate of
hydrogels. The CNDs provided necessary contrast between
embedded hybrid hydrogels and surrounding tissues to document
shape and location of hydrogel implant during degradation process.
Mice subcutaneously injected with CNDs hybrid hydrogels were
analyzed viaqualitative (visualization) and quantitative ( fluores-
cence reduction) strategy. To evaluate the effect of CNDs on UV light
penetration (UV crosslinking ability), hydrogels with or without
CNDs were prepared. UV curing depth of hydrogel with or without
CNDs was 1.4 cm and 1.8 cm, respectively ( Fig. S6 ). It suggested that
although the UV light penetration was indeed inhibited, the UV
curing depth of 1.4 cm could meet the requirement of transdermal
curing. Based on the above in vitro visual determination thatprovided a reliable and quantitative relationship between fluores-
cence reduction and hydrogel degradation, we further applied this
visual determination in degradation of hydrogel in vivo . As depicted
inFig. 7 A, all CNDs hybrid hydrogels remained localized to the site
of subcutaneous injection. A distinct decay of the fluorescence
signal in vivo was observed from the hydrogels over degradation
time. Since the CNDs embedded in hydrogels as the fluorescent
indicator were low photobleaching (almost no fluorescence
reduction during irradiation), the signal attenuation in area of in-
terest (AOI) should be ascribe to CNDs release induced by hydrogel
degradation and rapid diffusion away from the site of AOI. Quali-
tative analysis of in vivo fluorescence images with different CNDs
hybrid hydrogels indicated the differences in degradation kinetics
and more signal attenuation in AOI with lower DS: the decrease of
chemical cross-linked networks density resulted in faster degra-
dation. However the fluorescence signal from mice treated with
CNDs solution sharply reduced and even disappeared after short
time (72 h) of CNDs injection, suggesting that free CNDs or released
Fig. 7. Visual in vivo degradation of CNDs hybrid hydrogel by real-time and non-invasive fluorescence tracking. (A) Representative in vivo pseudo-colored images of CNDs
hybrid hydrogel (N-MAC 19, N-MAC 25, and N-MAC 28) viasubcutaneous injection over 288 h. A gradual attenuation of fluorescence signal was observed in all hybrid hydrogels
with different rates. Mice treated with free CNDs (CNDs solution without hydrogel) were control group; (B)Quantitative in vivo degradation viafluorescence reduction of CNDs
hybrid hydrogels (N-MAC 19, N-MAC 25, and N-MAC 28) as a function of time; (C)The gross appearance (up) and corresponding fluorescence pseudo-colored images (down) of
CNDs hybrid hydrogel with adjacent skin after 0 h, 72 h and 288 h. The CNDs hybrid hydrogels were visible and did not reveal redness, swelling and bleedin g. Fluorescence signal
demonstrated a well overlap with optical imaging and exhibited an obvious distinction with surrounding skin tissue; (D)Correlation of CNDs hybrid hydrogel degradation
quanti fied by applying non-invasive visual determination and invasive gravimetric degradation. (For interpretation of the references to colour in this figure legend, the reader is
referred to the web version of this article.)L. Wang et al. / Biomaterials 145 (2017) 192 e206 202

CNDs from hybrid hydrogels could rapidly diffuse away from the
subcutaneous injection site and could not make noise signal for
long-term visual in vivo degradation of hydrogel. Furthermore
quantitative fluorescence reduction over time would provide reli-
able assessment of hydrogels degradation as shown in Fig. 7 B. It
was clearly demonstrated that N-MAC 19 exhibited a faster
degradation compared with N-MAC 25 and N-MAC 28, which
consistent with the trends of in vitro degradation. After 288 h, about
87%, 63% and 58% of the initial hydrogels was biodegraded in the
case of N-MAC 19, N-MAC 25 and N-MAC 28, respectively. It was
well known that in body environment, the implantation of bio-
materials would activate foreign body response, which is correlated
to the release of various enzymes (including collagenase, lipases
and lysozyme) and other bio-reactive intermediates that could
accelerate degradation of the implanted biomaterial [70,71] .I t
should be noted that the quantitative in vivo fluorescence reduction
demonstrated a close correlation with in vitro degradation behavior
(including gravimetric and visual determination) but with faster
rate of degradation probably due to synergetic degradation with a
variety of enzymes in vivo . These degradation results were consis-
tent with previous literatures [8]that a higher chemical cross-
linked networks density limits the in vivo bioresorption and
degradation rate and lower cross-linked networks density may
account for disproportionately fast bioresorption kinetics.
To further con firm the feasibility and veracity of visual method
for assessment of in vivo degradation, mice were sacri ficed after 0 h,
72 h and 288 h and injection sites were surgically dissected to
exposure the CNDs hybrid hydrogel with adjacent skin ( Fig. 7 C, up).
The gross appearance demonstrated that CNDs hybrid hydrogels
were signi ficantly visible and did not reveal a series of alterations in
the microvasculature including redness, swelling and bleeding [72],
indicating serious in flammatory response was hardly observed
near the hybrid hydrogels injection site. The fluorescence pseudo-
color images of CNDs hybrid hydrogel with adjacent skin were
shown in ( Fig. 7 C, down). It was found that the fluorescence signal
gradually attenuated over the time of in vivo degradation, which is
consistent with the in vivo optical imaging results shown in ( Fig. 7 C,
up). It should be noted that the intense fluorescence signal
demonstrated a well overlap with optical imaging and exhibited an
obvious distinction with the surrounding skin tissue.
In order to verify that the visual determination method indeed
reflected real in vivo hydrogel degradation pro file, the correlation
between classical gravimetric degradation and visual determina-
tion was compared parallelly ( Fig. 7 D). The weight loss results were
plotted as percentage of the hydrogel degradation versus time and
correlated with the fluorescence reduction results obtained from
the visual determination. It was noteworthy that results from
hydrogel weight loss coincided with that from visual determination
for all time points (72, 192 and 288 h). No statistically signi ficant
differences were observed between gravimetric degradation and
visual determination, further con firmed the reliability and accuracy
using CNDs as fluorescent indicator for long-term visual in vivo
degradation of hydrogel.
To verify that this visual determination method could be
applicable to various biomaterial, CNDs hybrid gelatin hydrogel and
alginate hydrogel was injected into subcutaneous space using theabove procedure for visual in vivo degradation. As depicted in
Fig. S7 , the CNDs hybrid gelatin hydrogel could almost completely
degrade after 150 h, however, the CNDs hybrid alginate hydrogel
could completely degrade after 350 h. It was clearly demonstrated
that different fluorescence reduction time between gelatin hydro-
gel and alginate hydrogel indeed depended on the property of
material itself.
Recently fluorescence-related imaging technique has been
employed for non-invasive in vivo tracking of biomaterialdegradation ( Table S1 )[7,8,19,22,73 e75]. For example, the strate-
gies of covalently bonding organic fluorescent molecule (such as
Fluorescein-5-carboxyamido hexanoic acid [8], FITC [75], Rhoda-
mine B [7], Alexa Fluor 546 [74]and IR-Dye 800CW maleimide [19])
to the biomaterials have been adopted to track and quantify the
in vivo degradation of hydrogels or fibrin sealant. However, the
strategy based on covalently immobilized fluorescent molecule to
biomaterials suffers from some intractable issues, including high
photobleaching for long-term in vivo hydrogel tracking and un-
certain perturbation of degradation pro file due to the change in
molecular structure of hydrogels. There were some reports of
tracking hydrogel degradation non-invasively though directly
embedding lanthanide-doped rare-earth upconversion nano-
particles (such as silica-coated LiYF
4: Yb/Tm UCNPs [22] and pol-
yacrylic acid-coated NaYF 4: Yb/Tm UCNPs [73]) into hydrogel,
avoiding the need for chemical bonding with hydrogels matrix.
However, the synthesized UCNPs had to be further decorated with
oleic acid and silica (or polyacrylic acid) to improve dispersion and
biocompatibility, respectively. And the potential in vivo biotoxicity
of UCNPs nanoparticles due to long-term retention in the liver and
spleen system remains uncertain and more systematic in-vestigations are still required [48,76] . Thus, these outstanding
properties of low photobleaching and good biocompatibility make
CNDs promising luminescent nanomaterials for in vivo imaging
compared to conventionally employed organic dyes and rare-earth
nanoparticles.
In order to evaluate the in vivo degradation and biocompatibility
of the CNDs hybrid hydrogel, histopathological analysis of the host
tissues surrounding hydrogels was performed. For gross observa-
tion of CNDs hybrid hydrogel in vivo , the injection sites were sur-
gically dissected to exposure the CNDs hybrid hydrogel with
adjacent skin. The gross appearance of CNDs hybrid hydrogel
demonstrated that the CNDs hybrid hydrogel was tightly adhered
to the injection regional skin and did not show acute in flammation,
redness, bleeding and swelling on the skin. The results of H &E
staining of the surrounding skin tissues at 72, 120, 192 and 288 h
after subcutaneous injection were shown in Fig. 8 and CNDs hybrid
hydrogel was indicated by eosin-staining (marked with asterisks).
At the first 72 and 120 h post-injection, increased number of in-
flammatory cells was observed, indicating an acute in flammatory
reaction surrounding the eosinophilic hydrogel in the initial stage.
Hence, the results indicated that the CNDs hybrid hydrogel dis-
played acceptable biocompatibility in vivo , suggesting that the
degradation of hydrogel was consistent with visual determination
result. Overall the CNDs can be used as indicator for the monitoring
of degradation of hydrogels, because the CNDs were characterized
by red fluorescence emission, low photobleaching and good
biocompatibility. In addition, good homogeneity of CNDs in
hydrogels and the feature that the embedded CNDs in hydrogels
did not diffuse outside in the absence of hydrogel degradation, also
enable CNDs to be used as an indicator for the monitoring of
degradation of hydrogels by real-time and non-invasive tracking
(Fig. S8 ).
In vivo fluorescence imaging of hydrogel degradation would
enable the reduction of animal numbers in comparison with con-
ventional invasive methods in which tissue sample at each time-point depended on the sacri fice of larger numbers of animals.
Therefore, in vivo fluorescence imaging was in accordance with the
‘3Rs ’principle of replacement, reduction and re finement in animal
studies. So a universal method was established to follow and
quantify degradation of injectable hydrogel by non-invasive
tracking using CNDs as fluorescent indicator. The use of in vivo
fluorescence imaging was a simple and cost-effective method
which allowed continuous analysis of the process of hydrogel
degradation and further allowed the selection and optimization ofL. Wang et al. / Biomaterials 145 (2017) 192 e206 203

injectable hydrogel in tissue engineering, including the addition of
cells as well as growth factors or the combination with other
biomaterials.
4. Conclusion
We explored CNDs as fluorescent indicator for real-time and
non-invasively tracking in vitro /in vivo degradation of hydrogels by
fluorescence imaging. The direct embedding of photoluminescent
CNDs allowed low photobleaching, red emission and good
biocompatibility and avoid the need for chemically bonding. The
CNDs embedded in hydrogels did not diffuse outside in the absence
of hydrogel degradation. We established mathematical equation of
in vitro degradation for quantitatively depicting degradation pro file
of hydrogel and a close correlation (similar degradation kinetics)
between visual determination and gravimetric determination was
acquired. This in vitro visual determination could also be expanded
toin vivo degradation of hydrogels by real-time and non-invasive
fluorescence monitoring. A visual platform which quantitatively
depicts the degradation pro file of biodegradable hydrogel has been
developed and can be applied to subcutaneous degradation of
injectable hydrogel with down to 7 mm depth in small animal
trials. This fluorescence-related visual imaging methodology would
hold great potential for tracking and location in live tissue,
providing a multifunctional theranostic platform.
Conflict of interest
The authors declare no con flicts of interest.
Statement of signi ficance
As to the field of tissue engineering, a quantitative assessment of
in vivo degradation of engineered hydrogels is of great importance
especially for advisable design of molecule structure and matching
up with regeneration rate of newly generated tissue. However,
fluorescence-related visual imaging usually encounters some
challenge such as intrinsic photobleaching of fluorophores and
uncertain perturbation of degradation induced by change inmolecular structure. The present work is devoted to employed
photoluminescent carbon nanodots (CNDs) as fluorescent indicator
with low photobleaching, red emission and good biocompatibility
for visualization and quantitation of in vitro /in vivo degradation of
hydrogels by real-time and non-invasive tracking. Also in vitro
mathematical equation of hydrogel degradation kinetics had been
established which would open the possibility to develop predictivein vivo models for tissue engineered hydrogel scaffold.
Acknowledgments
The authors thank the financial support from National Science
Foundation of China (51372051, 51621091), State Key Laboratory of
Urban Water Resource and Environment of Harbin Institute of
Technology (2016TS03), HIT Environment and Ecology InnovationSpecial Funds (HSCJ201623), Innovation Talents of Harbin Science
and Engineering (2013RFLXJ023) and Fundamental Research Funds
for Central Universities (HIT.IBRSEM.201302). F.X. was supported by
the National Natural Science Foundation of China (11372243,
11522219, and 11532009).
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.biomaterials.2017.08.039 .
References
[1]M. Liu, Y. Ishida, Y. Ebina, T. Sasaki, T. Hikima, M. Takata, T. Aida, An aniso-
tropic hydrogel with electrostatic repulsion between cofacially alignednanosheets, Nature 517 (2015) 68 e72.
[2]Q. Peng, X. Sun, T. Gong, C.Y. Wu, T. Zhang, J. Tan, Z.R. Zhang, Injectable and
biodegradable thermosensitive hydrogels loaded with PHBHHx nanoparticles
for the sustained and controlled release of insulin, Acta Biomater. 9 (2013)
5063e5069 .
[3]K.A. Mosiewicz, L. Kolb, A.J. van der Vlies, M.M. Martino, P.S. Lienemann,
J.A. Hubbell, M. Ehrbar, M.P. Lutolf, In situ cell manipulation through enzy-
matic hydrogel photopatterning, Nat. Mater 12 (2013) 1072 e1078 .
[4]J.A. Burdick, W.L. Murphy, Moving from static to dynamic complexity in
hydrogel design, Nat. Commun. 3 (2012) 1269 .
[5]J. Kopecek, Hydrogel biomaterials: a smart future? Biomaterials 28 (2007)
5185e5192 .
[6]A. Berdichevski, H. Simaan Yameen, H. Dafni, M. Neeman, D. Seliktar, Using
Fig. 8. The Gross appearance and H &E stained histological analysis of subcutaneous tissue with the CNDs hybrid hydrogel at 72, 120, 192 and 288 h post-injection and the CNDs
hybrid hydrogel was indicated by eosin-staining (marked with asterisks).L. Wang et al. / Biomaterials 145 (2017) 192 e206 204

bimodal MRI/ fluorescence imaging to identify host angiogenic response to
implants, Proc. Natl. Acad. Sci. U. S. A. 112 (2015) 5147 e5152 .
[7]W. Wang, J. Liu, C. Li, J. Zhang, J. Liu, A. Dong, D. Kong, Real-time and non-
invasive fluorescence tracking of in vivo degradation of the thermosensitive
PEGlyated polyester hydrogel, J. Mater. Chem. B 2 (2014) 4185 e4192 .
[8]N. Artzi, N. Oliva, C. Puron, S. Shitreet, S. Artzi, A. Bon Ramos, A. Groothuis,
G. Sahagian, E.R. Edelman, In vivo and in vitro tracking of erosion in biode-
gradable materials using non-invasive fluorescence imaging, Nat. Mater 10
(2011) 704 e709.
[9]A. Mahdavi, L. Ferreira, C. Sundback, J.W. Nichol, E.P. Chan, D.J. Carter,
C.J. Bettinger, S. Patanavanich, L. Chignozha, E. Ben-Joseph, A. Galakatos,
H. Pryor, I. Pomerantseva, P.T. Masiakos, W. Faquin, A. Zumbuehl, S. Hong,
J. Borenstein, J. Vacanti, R. Langer, J.M. Karp, A biodegradable and biocom-patible gecko-inspired tissue adhesive, Proc. Natl. Acad. Sci. U. S. A. 105 (2008)
2307e2312 .
[10] C.J. Bettinger, J.P. Bruggeman, J.T. Borenstein, R.S. Langer, Amino alcohol-based
degradable poly(ester amide) elastomers, Biomaterials 29 (2008) 2315 e2325 .
[11] A.C. Grayson, G. Voskerician, A. Lynn, J.M. Anderson, M.J. Cima, R. Langer,
Differential degradation rates in vivo and in vitro of biocompatible poly(lactic
acid) and poly(glycolic acid) homo- and co-polymers for a polymeric drug-
delivery microchip, J. Biomater. Sci. Polym. Ed. 15 (2004) 1281 e1304 .
[12] Y. Liang, A. Bar-Shir, X. Song, A.A. Gilad, P. Walczak, J.W. Bulte, Label-free
imaging of gelatin-containing hydrogel scaffolds, Biomaterials 42 (2015)144e150.
[13] S.M. Dorsey, M. Haris, A. Singh, W.R. Witschey, C.B. Rodell, F. Kogan, R. Reddy,
J.A. Burdick, Visualization of injectable hydrogels using chemical exchange
saturation transfer MRI, ACS Biomater. Sci. Eng. 1 (2015) 227 e237.
[14] J. Liu, K. Wang, J. Luan, Z. Wen, L. Wang, Z. Liu, G. Wu, R. Zhuo, Visualization of
in situ hydrogels by MRI in vivo, J. Mater. Chem. B 4 (2016) 1343 e1353 .
[15] G.H. van Lenthe, H. Hagenmuller, M. Bohner, S.J. Hollister, L. Meinel, R. Muller,
Nondestructive micro-computed tomography for biological imaging and
quanti fication of scaffold-bone interaction in vivo, Biomaterials 28 (2007)
2479e2490 .
[16] A.J. Neumann, T. Quinn, S.J. Bryant, Nondestructive evaluation of a new hy-
drolytically degradable and photo-clickable PEG hydrogel for cartilage tissueengineering, Acta Biomater. 39 (2016) 1 e11.
[17] M.A. Rice, K.R. Waters, K.S. Anseth, Ultrasound monitoring of cartilaginous
matrix evolution in degradable PEG hydrogels, Acta Biomater. 5 (2009)
152e161.
[18] K. Kim, C.G. Jeong, S.J. Hollister, Non-invasive monitoring of tissue scaffold
degradation using ultrasound elasticity imaging, Acta Biomater. 4 (2008)
783e790.
[19] Y. Zhang, F. Rossi, S. Papa, M.B. Violatto, P. Bigini, M. Sorbona, F. Redaelli,
P. Veglianese, J. Hilborn, D.A. Ossipov, Non-invasive in vitro and in vivo
monitoring of degradation of fluorescently labeled hyaluronan hydrogels for
tissue engineering applications, Acta Biomater. 30 (2016) 188 e198.
[20] X. Ma, X. Sun, D. Hargrove, J. Chen, D. Song, Q. Dong, X. Lu, T.H. Fan, Y. Fu,
Y. Lei, A biocompatible and biodegradable protein hydrogel with green and
red auto fluorescence: preparation, characterization and in vivo biodegrada-
tion tracking and modeling, Sci. Rep. 6 (2016) 19370 .
[21] X. Dong, C. Wei, T. Liu, F. Lv, Z. Qian, Real-time fluorescence tracking of pro-
toporphyrin incorporated thermosensitive hydrogel and its drug release
in vivo, ACS Appl. Mater. Interfaces 8 (2016) 5104 e5113 .
[22] G. Jalani, R. Naccache, D.H. Rosenzweig, S. Lerouge, L. Haglund, F. Vetrone,
M. Cerruti, Real-time, non-invasive monitoring of hydrogel degradation using
LiYF 4:Yb(3ț)/Tm(3ț) NIR-to-NIR upconverting nanoparticles, Nanoscale 7
(2015) 11255 e11262 .
[23] J. Fu, C. Wiraja, R. Chong, C. Xu, D.A. Wang, Real-time and non-invasive
monitoring of embryonic stem cell survival during the development ofembryoid bodies with smart nanosensor, Acta Biomater. 49 (2017) 358 e367.
[24] K.F. Lei, M.H. Wu, C.W. Hsu, Y.D. Chen, Real-time and non-invasive impedi-
metric monitoring of cell proliferation and chemosensitivity in a perfusion 3D
cell culture micro fluidic chip, Biosens. Bioelectron. 51 (2014) 16 e21.
[25] Z. Gu, L. Yan, G. Tian, S. Li, Z. Chai, Y. Zhao, Recent advances in design and
fabrication of upconversion nanoparticles and their safe theranostic applica-
tions, Adv. Mater 25 (2013) 3758 e3779 .
[26] L. Cheng, K. Yang, M. Shao, X. Lu, Z. Liu, In vivo pharmacokinetics, long-term
biodistribution and toxicology study of functionalized upconversion nano-
particles in mice, Nanomedicine 6 (2011) 1327 e1340 .
[27] L. Xiong, T. Yang, Y. Yang, C. Xu, F. Li, Long-term in vivo biodistribution im-
aging and toxicity of polyacrylic acid-coated upconversion nanophosphors,Biomaterials 31 (2010) 7078 e7085 .
[28] F. Xu, F. Inci, O. Mullick, U.A. Gurkan, Y. Sung, D. Kavaz, B. Li, E.B. Denkbas,
U. Demirci, Release of magnetic nanoparticles from cell-encapsulating
biodegradable nanobiomaterials, ACS Nano 6 (2012) 6640 e6649 .
[29] J. Bartelmess, S.J. Quinn, S. Giordani, Carbon nanomaterials: multi-functional
agents for biomedical fluorescence and Raman imaging, Chem. Soc. Rev. 44
(2015) 4672 e4698 .
[30] S.N. Baker, G.A. Baker, Luminescent carbon nanodots: emergent nanolights,
Angew. Chem. Int. Ed. 49 (2010) 6726 e6744 .
[31] Y.P. Sun, B. Zhou, Y. Lin, W. Wang, K.A. Fernando, P. Pathak, M.J. Meziani,
B.A. Harruff, X. Wang, H. Wang, P.G. Luo, H. Yang, M.E. Kose, B. Chen,
L.M. Veca, S.Y. Xie, Quantum-sized carbon dots for bright and colorful pho-toluminescence, J. Am. Chem. Soc. 128 (2006) 7756 e7757 .
[32] P. Miao, K. Han, Y. Tang, B. Wang, T. Lin, W. Cheng, Recent advances in carbonnanodots: synthesis, properties and biomedical applications, Nanoscale 7
(2015) 1586 e1595 .
[33] K. Jiang, S. Sun, L. Zhang, Y. Lu, A. Wu, C. Cai, H. Lin, Red, green, and blue
luminescence by carbon dots: full-color emission tuning and multicolor
cellular imaging, Angew. Chem. Int. Ed. 54 (2015) 5360 e5363 .
[34] M.M. Titirici, R.J. White, N. Brun, V.L. Budarin, D.S. Su, F. del Monte, J.H. Clark,
M.J. MacLachlan, Sustainable carbon materials, Chem. Soc. Rev. 44 (2015)
250e290.
[35] S. Hu, A. Trinchi, P. Atkin, I. Cole, Tunable photoluminescence across the entire
visible spectrum from carbon dots excited by white light, Angew. Chem. Int.
Ed. 54 (2015) 2970 e2974 .
[36] S.L. Jacques, Optical properties of biological tissues: a review, Phys. Med. Biol.
58 (2013) R37 e
R61.
[37] J.V. Frangioni, In vivo near-infrared fluorescence imaging, Curr. Opin. Chem.
Biol. 7 (2003) 626 e634.
[38] B. Li, L. Wang, F. Xu, X. Gang, U. Demirci, D. Wei, Y. Li, Y. Feng, D. Jia, Y. Zhou,
Hydrosoluble, UV-crosslinkable and injectable chitosan for patterned cell-
laden microgel and rapid transdermal curing hydrogel in vivo, Acta Bio-mater. 22 (2015) 59 e69.
[39] J.K. Kim, K.W. Lee, T.E. Hefferan, B.L. Currier, M.J. Yaszemski, L.C. Lu, Synthesis
and evaluation of novel biodegradable hydrogels based on poly(ethylene
glycol) and sebacic acid as tissue engineering scaffolds, Biomacromolecules 9
(2008) 149 e157.
[40] O. Benny, S.K. Kim, K. Gvili, I.S. Radzishevsky, A. Mor, L. Verduzco, L.G. Menon,
P.M. Black, M. Machluf, R.S. Carroll, In vivo fate and therapeutic ef ficacy of PF-
4/CTF microspheres in an orthotopic human glioblastoma model, FASEB J. 22
(2008) 488 e499.
[41] C. Witt, K. Mader, T. Kissel, The degradation, swelling and erosion properties
of biodegradable implants prepared by extrusion or compression moulding of
poly(lactide-co-glycolide) and ABA triblock copolymers, Biomaterials 21(2000) 931 e938.
[42] S.H. Jin, D.H. Kim, G.H. Jun, S.H. Hong, S. Jeon, Tuning the photoluminescence
of graphene quantum dots through the charge transfer effect of functional
groups, ACS Nano 7 (2013) 1239 e1245 .
[43] X.F. Jia, J. Li, E.K. Wang, One-pot green synthesis of optically pH-sensitive
carbon dots with upconversion luminescence, Nanoscale 4 (2012)
5572e5575 .
[44] J. Lu, J.X. Yang, J.Z. Wang, A.L. Lim, S. Wang, K.P. Loh, One-Pot synthesis of
fluorescent carbon nanoribbons, nanoparticles, and graphene by the exfolia-
tion of graphite in ionic liquids, ACS Nano 3 (2009) 2367 e2375 .
[45] D.E. Hudson, D.O. Hudson, J.M. Wininger, B.D. Richardson, Penetration of laser
light at 808 and 980 nm in bovine tissue samples, Photomed. Laser Surg. 31(2013) 163 e168.
[46] P. Juzenas, A. Juzeniene, O. Kaalhus, V. Iani, J. Moan, Noninvasive fluorescence
excitation spectroscopy during application of 5-aminolevulinic acid in vivo,
Photochem. Photobiol. Sci. 1 (2002) 745 e748.
[47] J.H. Yu, S.H. Kwon, Z. Petrasek, O.K. Park, S.W. Jun, K. Shin, M. Choi, Y.I. Park,
K. Park, H.B. Na, N. Lee, D.W. Lee, J.H. Kim, P. Schwille, T. Hyeon, High-reso-
lution three-photon biomedical imaging using doped ZnS nanocrystals, Nat.
Mater 12 (2013) 359 e366.
[48] A. Gnach, A. Bednarkiewicz, Lanthanide-doped up-converting nanoparticles:
merits and challenges, Nano Today 7 (2012) 532 e563.
[49] L.M. Maestro, E.M. Rodriguez, F. Vetrone, R. Naccache, H.L. Ramirez, D. Jaque,
J.A. Capobianco, J.G. Sole, Nanoparticles for highly ef ficient multiphoton
fluorescence bioimaging, Opt. Express 18 (2010) 23544 e
23553 .
[50] B.Y. Wu, H.F. Wang, J.T. Chen, X.P. Yan, Fluorescence resonance energy
transfer inhibition assay for alpha-fetoprotein excreted during cancer cell
growth using functionalized persistent luminescence nanoparticles, J. Am.
Chem. Soc. 133 (2011) 686 e688.
[51] Y. Chong, Y.F. Ma, H. Shen, X.L. Tu, X. Zhou, J.Y. Xu, J.W. Dai, S.J. Fan, Z.J. Zhang,
The in vitro and in vivo toxicity of graphene quantum dots, Biomaterials 35
(2014) 5041 e5048 .
[52] H.Q. Tao, K. Yang, Z. Ma, J.M. Wan, Y.J. Zhang, Z.H. Kang, Z. Liu, In Vivo NIR
fluorescence imaging, biodistribution, and toxicology of photoluminescent
carbon dots produced from carbon nanotubes and graphite, Small 8 (2012)
281e290.
[53] Y.X. Wu, X.B. Zhang, D.L. Zhang, C.C. Zhang, J.B. Li, Y. Wu, Z.L. Song, R.Q. Yu,
W.H. Tan, Quench-shield ratiometric upconversion luminescence nanoplat-
form for biosensing, Anal. Chem. 88 (2016) 1639 e1646 .
[54] Z.Y. Liao, S.A. Bogh, M. Santella, C. Rein, T.J. Sorensen, B.W. Laursen, T. Vosch,
Emissive photoconversion products of an amino-triangulenium dye, J. Phys.Chem. A 120 (2016) 3554 e3561 .
[55] T. Ha, P. Tinnefeld, Photophysics of fluorescent probes for single-molecule
biophysics and super-resolution imaging, Annu. Rev. Phys. Chem. 63 (2012)
595e617.
[56] J.W. Dobrucki, D. Feret, A. Noatynska, Scattering of exciting light by live cells
influorescence Confocal imaging: phototoxic effects and relevance for FRAP
studies, Biophys. J. 93 (2007) 1778 e1786 .
[57] K. Wang, Z.C. Gao, G. Gao, Y. Wo, Y.X. Wang, G.X. Shen, D.X. Cui, Systematic
safety evaluation on photoluminescent carbon dots, Nanoscale Res. Lett. 8
(2013) .
[58] C.E. Kast, A. Bernkop-Schnurch, Thiolated polymers ethiomers: development
and in vitro evaluation of chitosan-thioglycolic acid conjugates, Biomaterials22 (2001) 2345 e2352 .
[59] A. Konwar, N. Gogoi, G. Majumdar, D. Chowdhury, Green chitosan ecarbonL. Wang et al. / Biomaterials 145 (2017) 192 e206 205

dots nanocomposite hydrogel film with superior properties, Carbohydr.
Polym. 115 (2015) 238 e245.
[60] B.H. Honig, W.L. Hubbell, Stability of “salt bridges ”in membrane proteins,
Proc. Natl. Acad. Sci. U. S. A. 81 (1984) 5412 e5416 .
[61] P. Thoniyot, M.J. Tan, A.A. Karim, D.J. Young, X.J. Loh, Nanoparticle-hydrogel
composites: concept, design, and applications of these promising, multi-functional materials, Adv. Sci. 2 (2015) 1400010 .
[62] H.S. Chou, M. Larsson, M.H. Hsiao, Y.C. Chen, M. Roding, M. Nyden, D.M. Liu,
Injectable insulin-lysozyme-loaded nanogels with enzymatically-controlled
degradation and release for basal insulin treatment: in vitro characteriza-
tion and in vivo observation, J. Control. Release 224 (2016) 33 e42.
[63] H. Zhang, S.H. Neau, In vitro degradation of chitosan by bacterial enzymes
from rat cecal and colonic contents, Biomaterials 23 (2002) 2761 e2766 .
[64] P. Martens, T. Holland, K.S. Anseth, Synthesis and characterization of
degradable hydrogels formed from acrylate modi fied poly(vinyl alcohol)
macromers, Polymer 43 (2002) 6093 e6100 .
[65] A.T. Metters, K.S. Anseth, C.N. Bowman, Fundamental studies of a novel,
biodegradable PEG-b-PLA hydrogel, Polymer 41 (2000) 3993 e4004 .
[66] D. Cekmecelioglu, O.N. Uncu, Kinetic modeling of enzymatic hydrolysis of
pretreated kitchen wastes for enhancing bioethanol production, Waste
Manage 33 (2013) 735 e739.
[67] J.K. Ko, E. Ximenes, Y. Kim, M.R. Ladisch, Adsorption of enzyme onto lignins of
liquid hot water pretreated hardwoods, Biotechnol. Bioeng. 112 (2015)447e456.[68] X.H. Ju, M. Engelhard, X. Zhang, An advanced understanding of the speci fic
effects of xylan and surface lignin contents on enzymatic hydrolysis of
lignocellulosic biomass, Bioresour. Technol. 132 (2013) 137 e145.
[69] R. Weissleder, A clearer vision for in vivo imaging, Nat. Biotechnol. 19 (2001)
316e317.
[70] D.M. Higgins, R.J. Basaraba, A.C. Hohnbaum, E.J. Lee, D.W. Grainger,
M. Gonzalez-Juarrero, Localized immunosuppressive environment in the
foreign body response to implanted biomaterials, Am. J. Pathol. 175 (2009)
161e170.
[71] J. Hilborn, L.M. Bjursten, A new and evolving paradigm for biocompatibility,
J. Tissue Eng. Regen. Med. 1 (2007) 110 e119.
[72] B. Rihova, Immunocompatibility and biocompatibility of cell delivery systems,
Adv. Drug Deliv. Rev. 42 (2000) 65 e80.
[73] Y.Q. Dong, G.R. Jin, C.C. Ji, R.Y. He, M. Lin, X. Zhao, A. Li, T.J. Lu, F. Xu, Non-
invasive tracking of hydrogel degradation using upconversion nanoparticles,
Acta Biomater. 55 (2017) 410 e419.
[74] S. Wolbank, V. Pichler, J.C. Ferguson, A. Meinl, M. van Griensven, A. Goppelt,
H. Redl, Non-invasive in vivo tracking of fibrin degradation by fluorescence
imaging, J. Tissue Eng. Regen. Med. 9 (2015) 973 e976.
[75] R.J. Wade, E.J. Bassin, C.B. Rodell, J.A. Burdick, Protease-degradable electrospun
fibrous hydrogels, Nat. Commun. 6 (2015) .
[76] A. Gnach, T. Lipinski, A. Bednarkiewicz, J. Rybka, J.A. Capobianco, Upconverting
nanoparticles: assessing the toxicity, Chem. Soc. Rev. 44 (2015) 1561 e1584 .L. Wang et al. / Biomaterials 145 (2017) 192 e206 206