Bioactive polypeptide hydrogels modified with RGD and N-cadherin mimetic peptide promote chondrogenic differentiation of bone marrow mesenchymal stem cells

Article abstract of DOI:10.1007/s11426-020-9772-0

Cell-material and cell-cell interactions represent two crucial aspects of the regulation of cell behavior. In the present study, poly (L-glutamic acid) (PLG) hydrogels were prepared by catalyst-free click crosslinking via a strain-promoted azide-alkyne cycloaddition (SPAAC) reaction between azido-grafted PLG (PLG-N₃) and azadibenzocyclooctyne-grafted PLG (PLG-ADIBO). The bioactive peptides c(RGDfK) and N-cadherin mimetic peptide (N-Cad) were both conjugated to the PLG hydrogel (denoted PLG+RGD/N-Cad) in order to regulate cell-material and cell-cell interactions. Gelation time and storage modulus of the hydrogels were tunable through variations in the concentration of polypeptide precursors. The hydrogels degraded gradually in the presence of proteinases. The viability of bone marrow mesenchymal stem cells (BMSCs) was maintained when cultured with extracts of the hydrogels or encapsulated within the hydrogels. Degradation was observed within 10 weeks following the subcutaneous injection of hydrogel solution in rats, displaying excellent histocompatibility in vivo. The introduction of RGD into the PLG hydrogel promoted the adhesion of BMSCs onto the hydrogels. Moreover, when encapsulated within the PLG+RGD/N-Cad hydrogel, BMSCs secreted cartilage-specific matrix, in addition to chondrogenic gene and protein expression being significantly enhanced in comparison with BMSCs encapsulated in hydrogels without N-Cad modification. These findings suggest that these biodegradable, bioactive polypeptide hydrogels have great potential for use in 3D cell culture and in cartilage tissue engineering.

Full text of DOI:10.1007/s11426-020-9772-0

SCIENCE CHINA
Chemistry
•
ARTICLES•
https://doi.org/10.1007/s11426-020-9772-0
.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Bioactive polypeptide hydrogels modified with RGD and
N-cadherin mimetic peptide promote chondrogenic differentiation
of bone marrow mesenchymal stem cells
1
,2
1,3
1,3*
1,3
Yan Rong , Zhen Zhang , Chaoliang He & Xuesi Chen
1
Key Laboratory of Polymer Ecomaterials, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences,
Changchun 130022, China;
2
University of Chinese Academy of Sciences, Beijing 100039, China;
3
University of Science and Technology of China, Hefei 230026, China
Received April 7, 2020; accepted May 9, 2020; published online June 8, 2020
Cell-material and cell-cell interactions represent two crucial aspects of the regulation of cell behavior. In the present study, poly
L-glutamic acid) (PLG) hydrogels were prepared by catalyst-free click crosslinking via a strain-promoted azide-alkyne cy-
(
cloaddition (SPAAC) reaction between azido-grafted PLG (PLG-N ) and azadibenzocyclooctyne-grafted PLG (PLG-ADIBO).
3
The bioactive peptides c(RGDfK) and N-cadherin mimetic peptide (N-Cad) were both conjugated to the PLG hydrogel (denoted
PLG+RGD/N-Cad) in order to regulate cell-material and cell-cell interactions. Gelation time and storage modulus of the
hydrogels were tunable through variations in the concentration of polypeptide precursors. The hydrogels degraded gradually in
the presence of proteinases. The viability of bone marrow mesenchymal stem cells (BMSCs) was maintained when cultured with
extracts of the hydrogels or encapsulated within the hydrogels. Degradation was observed within 10 weeks following the
subcutaneous injection of hydrogel solution in rats, displaying excellent histocompatibility in vivo. The introduction of RGD into
the PLG hydrogel promoted the adhesion of BMSCs onto the hydrogels. Moreover, when encapsulated within the PLG+RGD/N-
Cad hydrogel, BMSCs secreted cartilage-specific matrix, in addition to chondrogenic gene and protein expression being
significantly enhanced in comparison with BMSCs encapsulated in hydrogels without N-Cad modification. These findings
suggest that these biodegradable, bioactive polypeptide hydrogels have great potential for use in 3D cell culture and in cartilage
tissue engineering.
polypeptide hydrogel, click crosslinking, biofunctionalization, 3D cell culture, tissue engineering
Citation:
Rong Y, Zhang Z, He C, Chen X. Bioactive polypeptide hydrogels modified with RGD and N-cadherin mimetic peptide promote chondrogenic
differentiation of bone marrow mesenchymal stem cells. Sci China Chem, 2020, 63, https://doi.org/10.1007/s11426-020-9772-0
1
Introduction
(ECM) for tissue engineering applications [1–5]. A number
of naturally-derived polymers and biodegradable synthetic
polymers have been studied as building blocks for hydrogel
formation due to their good biocompatibility and biode-
gradability [6–9]. The advantages of synthetic polymers lie
in their defined composition and chemical structure, in ad-
dition to tunable mechanical properties [8,10]. Synthetic
polypeptides are a type of unique biodegradable synthetic
polymer composed of natural α-amino acid residues, which
Hydrogels are a class of three-dimensional hydrophilic or
amphiphilic polymer network with a large proportion of their
volume consisting of water. Because of their viscoelastic
properties and high water content, hydrogels are considered
potential scaffolds that mimic natural extracellular matrix
*Corresponding author (email: clhe@ciac.ac.cn)
©
Science China Press and Springer-Verlag GmbH Germany, part of Springer Nature 2020
chem.scichina.com link.springer.com

2
Rong et al. Sci China Chem
show excellent biocompatibility, biodegradability, and bio-
mimetic secondary structure similar to that of natural pro-
teins. In addition, the inherent carboxyl, amino or hydroxyl
side groups of some polypeptides facilitate additional in-
corporation of functional moieties [11–14]. Thus, polypep-
tide hydrogels have recently been investigated as emerging
scaffolds for 3D cell culture and tissue engineering appli-
cations [15–20]. For instance, in previous studies, we and
other groups, have developed poly(L-glutamic acid) (PLG)-
based hydrogels with adjustable mechanical properties for
encapsulation of bone marrow mesenchymal stem cells
typic interaction of N-cadherin on adjacent mesenchymal
cells. N-cadherin contains an extracellular domain re-
sponsible for its homotypic binding as well as a cytosolic
domain to bind the actin cytoskeleton through the catenin
family [32,34]. Previous studies have shown that synthetic
mimetic peptides containing the His-Ala-Val (HAV) domain,
an evolutionarily conserved motif in the first extracellular
domain (ECD1) of N-cadherin, have similar binding prop-
erties to N-cadherin [35,36]. Bian et al. [37] and Burdick et
al. [38–40] have reported that a self-assembled peptide hy-
drogel and a hyaluronic acid (HA) hydrogel grafted with the
N-cadherin mimetic peptide promoted chondrogenic differ-
entiation of mesenchymal stem cells (MSCs). Hydrogels that
incorporate cues for both matrix-cell and cell-cell interac-
tions may represent a new paradigm in scaffolds that regulate
MSC differentiation. Nevertheless, to date, hydrogels cap-
able of both promoting cell adhesion on materials and reg-
ulating cell-cell interactions are rarely reported.
In the present work, an injectable PLG hydrogel cross-
linked via a catalyst-free SPAAC reaction between azido and
dibenzocyclooctyne groups was developed under physiolo-
gical conditions. Both c(RGDfK) and a N-cadherin mimetic
peptide (N-Cad) were conjugated to the PLG network in
order that the hydrogel could regulate cell proliferation and
differentiation (Figure 1). Gelation time, mechanical prop-
erties and degradation properties of the hydrogels were in-
vestigated. The biocompatibility of hydrogels was evaluated
both in vitro and in vivo and the proliferation of cells en-
capsulated in the hydrogels measured. Furthermore, the ac-
cumulation of cartilage-specific matrix and the expression of
chondrogenic markers on BMSCs encapsulated in the bio-
functionalized hydrogels were studied.
(BMSCs) and chondrocytes [16,21].
Additionally, many crosslinking approaches have been
utilized to fabricate covalently-crosslinked networks of hy-
drogels. In previous studies, polypeptide hydrogels cross-
linked via a horseradish/hydrogen peroxidase (HRP/H O )
2
2
system, hydrazone formation reaction, and an N-(3-di-
methylaminopropyl)-N′-ethylcarbodiimide hydrochloride/N-
hydroxysuccinimide (EDC/NHS) system, have been devel-
oped as tissue engineering scaffolds [13,16,21]. However,
crosslinking systems involving H O , free radicals, alde-
2
2
hydes or small-molecule chemical crosslinking agents may
cause potential toxicity towards the encapsulated cells. Re-
cently, bioorthogonal reactions, especially catalyst-free click
reactions, have attracted considerable attention as emerging
crosslinking strategies due to their speed, efficiency and high
selectivity, in addition to their excellent biocompatibility
both in vitro and in vivo [22–24]. For example, it has been
reported that bone marrow stromal cells encapsulated in poly
(
ethylene glycol)-co-polycarbonate hydrogels crosslinked
using strain-promoted azide-alkyne cycloaddition
SPAAC) reaction displayed greater cellular viability than
a
(
those encapsulated in photo-crosslinked poly(ethylene gly-
col) dimethacrylate (PEG-DMA) hydrogels [25].
The biofunctionality of hydrogels is also a crucial factor
that significantly affects tissue regeneration in addition to the
method used in cross-linking. Based on the interactions be-
tween materials and cells, considerable effort has been de-
voted to the design of materials that actively communicate
with complex biological environments [15,26–28]. The
conjugation of bioactive factors is a facile and effective way
to impart biological function to synthetic materials. RGD, a
cell binding domain that is found in fibronectin (FN), lami-
nin (LN) and collagen, is the most commonly used cell-
adhesive peptide. A total of 24 integrins bind to ECM mo-
lecules in an RGD-mediated manner [29,30]. It has been
reported that, compared with linear RGD, cyclic RGD
2 Materials and methods
2
.1 Materials
γ-Benzyl-L-glutamate-N-carboxyanhydride (BLG NCA)
was synthesized in accordance with a previously published
method [41]. EDC (Aladdin, 98%), 4-dimethylaminopyr-
idine (DMAP, Aladdin, 99%, USA), NHS (Aladdin, 98%),
c(RGDfK) (GL Biochem, 95%), N-cadherin mimetic peptide
(Ac-HAVDIGGGK, GL Biochem, 95%, China), Scrambled
peptide (Ac-AGVGDHIGK, GL Biochem, 95%) and azadi-
benzocyclooctyne-ethylamine•TFA
(ADIBO-NH •TFA,
2
Biomatrik, China) were used as received. Azide-substituted
diethylene glycol (EG N ) was prepared by the reaction of
2
3
(cRGD) exhibits stronger affinity to integrin with greater
sodium azide (NaN , Multipoint Chem, 98%) with 2-(2-
3
biological activity [31]. Furthermore, neural cadherin (N-
cadherin) is considered an important factor for regulating cell
differentiation by directing cell-cell interactions during me-
senchymal condensation and chondrogenesis [32,33]. The
intercellular communication is mediated through the homo-
chloroethoxy)ethanol (Adamas, 98%), in accordance with a
previously-published method [42]. Tetrahydrofuran (THF)
and N,N-dimethylformamide (DMF) were dried over mag-
nesium sulfate prior to use. Dulbecco’s modified eagle
medium: nutrient mixture F-12 (DMEM/F12) was purchased

Rong et al. Sci China Chem
3
Figure 1 (a) Schematic diagram of the formation of click-crosslinked polypeptide hydrogels from PLG-N and PLG-ADIBO for use as a cell delivery
3
vehicle. (b) Bioactive c(RGDfK) and N-cadherin mimetic peptides introduced into the PLG backbone to construct bioactive polypeptide hydrogels (color
online).
from Gibco. Elastase (4.6 U/mg) was purchased from Wor-
thington and transforming growth factor-β1 (TGF-β1) was
obtained from PeproTech (USA). Insulin-transferrin-sodium
selenite media supplement (ITS), papain (30 U/mg), protei-
measuring the absorbance at 420 nm with an ultraviolet-visi-
ble (UV-Vis) spectrophotometer (UV-2401PC, Shimadzu,
Japan). Scanning electron microscopy (SEM) was performed
on a scanning electron microscope (Merlin, Zeiss, Germany).
nase
K
(30 U/mg) and 1,9-dimethyl-methylene blue
(DMMB) were acquired from Sigma-Aldrich (USA).
2
.3 Synthesis of PLG-ADIBO, PLG-N , and peptide-
3
modified PLG-N3
2
.2 Characterization
PLG was prepared in accordance with a method we pre-
viously reported [41]. Briefly, BLG NCA was dissolved in
anhydrous DMF under an atmosphere of nitrogen and an
appropriate quantity of n-hexylamine added. The mixture
was stirred at room temperature for 3 d and the resultant poly
(γ-benzyl-L-glutamate) (PBLG) was purified by precipita-
tion into diethyl ether, filtration then vacuum drying. The
PBLG obtained in this way (6 g) was dissolved in di-
chloroacetic acid (60 mL) then 33% hydrobromic acid in
acetic acid (20 mL) was added to the solution. The depro-
tection reaction of PBLG was allowed to proceed whilst
stirring at room temperature for 2 h, and the product of the
reaction obtained by precipitation into diethyl ether and fil-
tration. Crude PLG was purified by dissolution in DMF,
dialysis against deionized water and lyophilization.
1
Proton nuclear magnetic resonance ( H NMR) spectra were
recorded using a Bruker AV 500 NMR spectrometer (Ger-
many) in deuterium oxide (D O), deuterated trifluoroacetic
2
acid (CF COOD) or deuterated chloroform (CDCl ). The
3
3
molecular weights and polydispersity indexes (PDI) were
measured by gel permeation chromatography (GPC) on a
Waters 1515 Isocratic HPLC pump equipped with Ultra-
hydrogel™ linear column and a Waters 2414 refractive index
detector (USA). The eluant was 0.2 M phosphate buffer (PB)
containing 0.1 M NaNO and 0.2 g/L NaN . The flow rate was
3
3
0.6 mL/min and the testing temperature was 35 °C. Mono-
disperse polyethylene glycol (PEG, Waters Co. Ltd., USA)
with molecular weights ranging from 1,000 to 2.18×10 were
5
used as the standards for generating the calibration curve. The
grafting ratio of c(RGDfK) was measured by using a Perkin-
Elmer LS50B luminescence spectrometer (USA). The fluor-
escence excitation spectra were recorded at the detection
wavelength of 395 nm. The grafting ratio of N-cadherin mi-
metic peptide and scrambled peptide was determined by
ADIBO-modified PLG (PLG-ADIBO) was synthesized by
coupling PLG with ADIBO-NH •TFA via an EDC/NHS-
2
mediated amidation reaction. PLG was dissolved in DMF
and EDC then NHS was added (EDC:NHS:–COOH=2:2:10,
molar ratio) to activate a proportion of the carboxyl groups.

4
Rong et al. Sci China Chem
ADIBO-NH was then added (ADIBO:–COOH=1:10, molar
Scram.
Gelation time was measured using a vial tilting method.
When no flow was observed 30 s after inversion of the vial,
the sample was regarded as having formed a gel.
2
ratio) and the mixture stirred at room temperature in the dark
for 3 d. The resultant PLG-ADIBO was dialyzed against
deionized water then lyophilized.
Azido-modified PLG (PLG-N ) was synthesized by cou-
3
pling PLG with 2-(2-azidoethoxy)ethanol via an EDC/
DMAP-mediated coupling reaction. PLG was dissolved in
DMF and 2-(2-azidoethoxy)ethanol was added (N3:
2
.5 Morphology of hydrogels
SEM was employed for observation of morphology of the
PLG hydrogel. The samples were frozen in liquid nitrogen
and subsequently lyophilized for 3 d. The specimens were
sectioned and cross sections were coated with gold. Mor-
phology of the PLG hydrogel was measured on a scanning
electron microscope.
–
COOH=1:10, molar ratio). After several minutes, EDC and
DMAP were added (EDC:DMAP:–COOH=2:0.2:10, molar
ratio), after which the mixture was stirred at room tem-
perature for 3 d. The resultant PLG-N was dialyzed against
3
deionized water then lyophilized.
RGD and N-cadherin mimetic peptide (N-Cad)-modified
PLG-N (denoted PLG-N /RGD/N-Cad) was obtained by
3
3
2
.6 Mechanical characterization
successively coupling PLG-N with c(RGDfK) and N-Cad
3
using an EDC/NHS-mediated amidation reaction. Following
Rheological tests were conducted on an MCR302 Rheometer
containing a 25 mm parallel plate at 37 °C. Briefly, once
the dissolution of PLG-N in DMF, EDC and NHS were
3
added (EDC:NHS:–COOH=4:4:100, molar ratio) and the
solution stirred at room temperature for 1 d. The product was
precipitated in diethyl ether, washed with ethanol, collected
by centrifugation and dried under vacuum. The resultant
product was then dissolved in DMF. c(RGDfK) and N-Cad
were added and the mixture was stirred at room temperature
for 3 d (c(RGDfK):N-Cad:–COOH=1:1:100, molar ratio).
mixtures of PLG-N and PLG-ADIBO at different con-
3
centrations had been prepared, samples were immediately
deposited into a 500 μm gap between two parallel plates of a
rheometer. Time-dependent storage modulus (G′) and loss
modulus (G″) were measured at a constant frequency of 1 Hz
and strain of 1%.
The resultant PLG-N /RGD/N-Cad was dialyzed against
deionized water then lyophilized. Finally, a control sample
was fabricated by synthesis of RGD and scrambled peptide
3
2
.7 In vitro hydrogel degradation
Degradation of the hydrogels was evaluated by incubating
0.1 mL hydrogel in 1 mL 0.01 M PBS with or without an
enzyme at 37 °C. To assess enzymatic degradation, protei-
nase K (5 U/mL) or elastase (5 U/mL) was added to samples
then refreshed every day. The mass of the gel was weighed
daily until the samples were completely degraded.
(
Scram)-modified PLG-N (denoted PLG-N /RGD/Scram)
3
3
using a similar method, but with Scram substituted in place
of N-Cad.
1
The grafting ratio of each molecule was ascertained by H
NMR. The content of RGD was rechecked by measuring
arginine content, as previously reported [43]. The con-
centration of N-cadherin mimetic peptide and scrambled
peptide were rechecked using the Pauly reaction [44] (Figure
S1(a, b), Supporting Information online).
2
.8 In vitro cytocompatibility
Rabbit BMSCs were expanded to passage 3 using growth
media (DMEM/F12 with 10% FBS, 50 U/mL penicillin and
5
0 μg/mL streptomycin). BMSCs were seeded into the wells
2
.4 Hydrogel formation
4
of 96-well plates at a density of 1×10 cells/well in 0.1 mL of
media for 24 h. The media was subsequently replaced with
0.1 mL of extract (DMEM/F12, 37 °C, 72 h) of the PLG-N3/
PLG-ADIBO hydrogel. Fresh media and PEI 25K
(0.125 mg/mL) were used as negative and positive controls,
respectively. After 24 h of incubation, cells were incubated
with media containing 3-(4,5-dimethylthiazol-2-yl)-2,5-di-
phenyltetrazolium bromide (MTT) (1 mg/mL) for 4 h after
which the absorbance at 490 nm was measured using a mi-
croplate reader (Tecan Spark, Switzerland).
PLG-ADIBO, PLG-N , PLG-N /RGD/N-Cad and PLG-N /
3
3
3
RGD/Scram were dissolved separately in phosphate buffered
saline (PBS, 0.01 M) at predetermined concentrations. PLG
hydrogels with and without peptide modification were then
formed by mixing PLG-ADIBO with PLG-N , PLG-N /
3
3
RGD/N-Cad or PLG-N /RGD/Scram at an N /ADIBO molar
3
3
ratio of 1:1. Specifically, the PLG hydrogels without peptide
modification were prepared by mixing PLG-ADIBO with
PLG-N . The PLG hydrogels modified with RGD and N-Cad
3
(
denoted PLG+RGD/N-Cad) were prepared by mixing PLG-
The viability of BMSCs within hydrogels was evaluated as
follows. BMSC-encapsulated hydrogels were obtained by
ADIBO with PLG-N /RGD/N-Cad. The PLG hydrogels
3
modified with RGD and Scram (denoted PLG+RGD/Scram)
mixing PLG-N and PLG-ADIBO solutions with a cell
3
were prepared by mixing PLG-ADIBO with PLG-N /RGD/
suspension under sterile conditions at 37 °C. The final con-
3

Rong et al. Sci China Chem
5
centration of cells and hydrogels was adjusted as required.
The morphology of cells within the hydrogels was observed
using actin and nuclear staining, as follows. BMSC-
encapsulated hydrogels were prepared in glass bottom cell
culture dishes (1.5×10 cells/mL). After 48 h, the samples
were fixed using paraformaldehyde and stained firstly with
Alexa Fluor 488 Phalloidin (30 min) then DAPI (1 μg/mL,
Chondrogenesis of BMSCs in 2D and 3D culture was
achieved by incubation in induction media (DMEM/F12, 50
U/mL penicillin, 50 μg/mL streptomycin, 10 ng/mL TGF-
β1, 1% ITS, 50 μg/mL ascorbate, 50 μg/mL L-proline,
0.1 μM dexamethasone, 0.9 mM sodium pyruvate). The
media was replaced every three days.
5
5
min). Cell morphology was imaged using confocal laser
2
.11 Cell proliferation within hydrogels
scanning microscopy (LSM 700, Zeiss). The viability of
BMSCs in the hydrogels was assessed using live-dead cell
staining. Briefly, the hydrogels (4×10 cells/mL) were
stained at predetermined time points with 2 μM calcein-AM
and 4.5 μM propidium iodine (PI) in 0.01 M PBS. The
staining was performed at 37 °C in the dark for 45 min and
the results were observed then imaged using a fluorescence
microscope. For 3D views of confocal laser scanning mi-
Hydrogels containing cells were prepared using the same
procedure as described above. The proliferation of BMSCs
or chondrocytes within hydrogels (2×10 cells/mL) was
measured using a cell counting kit-8 (CCK-8) assay. CCK-8
solution (10% v/v) was added to each sample at pre-
determined time points then incubated for 3 h. Absorbance at
450 and 630 nm (baseline correction) was then measured.
Cells were cultured with proliferation medium (DMEM/F12,
antibiotics and 10% FBS) at 37 °C and in an atmosphere
containing 5% CO2.
5
5
5
croscopy images, the hydrogels (4×10 cells/mL) were
stained at 48 h with 2 μM calcein-AM and 4.5 μM PI, and a
2
00 μm z-stack was acquired.
2
.9 In vivo hydrogel degradation and biocompatibility
2.12 Glycosaminoglycan (GAG) production
In vivo hydrogel degradation and biocompatibility were
tested using a Sprague Dawley (SD) rat model (~200 g body
weight). A 0.2 mL volume of 3% (w/v) hydrogel precursor
solution was subcutaneously injected into the dorsal region
of each rat which then underwent gelation. At predetermined
time intervals (0, 1, 2, 4, 6, 8, 10 weeks), the animals were
sacrificed and the remaining hydrogel collected then pho-
tographed. Tissue surrounding the hydrogel was harvested
and fixed in 4% paraformaldehyde. Histopathological ex-
amination was performed after hematoxylin and eosin
After culturing for 3 weeks, hydrogels containing BMSCs
(6×10 cells/mL) were lyophilized, the dry weight measured
6
then digested with papain solution (5 U/mL) at 60 °C for
20 h. GAG production was quantified using a di-
methylmethylene blue dye assay (DMMB) in accordance
with a method described previously [15]. Chondroitin sul-
fate-A was used to generate a standard curve. dsDNA was
quantified using PicoGreen reagent, in accordance with the
manufacturer’s instructions (Figure S1(c, d)).
(
H&E) staining of the tissue.
All animal experiments were approved and conducted in
2
.13 RNA extraction and gene expression analysis
accordance with the guidelines of the Animal Experimental
Center of Jilin University Institutional Animal Care and Use
Committee.
BMSC-laden hydrogels (1, 3, 7 d) were transferred into
TRIzol reagent (Invitrogen, USA) then evenly ground to
extract the RNA in accordance with the manufacturer’s in-
structions, and then quantified using a microplate reader.
RNA samples were reverse-transcribed into cDNA using
HiScript III RT SuperMix for qPCR (+gDNA wiper) (Va-
zyme). Quantitative real-time polymerase chain reaction
(qRT-PCR) was performed using a Mxpro3005P QPCR
system (Stratagene, USA) with ChamQ SYBR Color qPCR
Master Mix. Primer sequences are listed in Table S1 (Sup-
porting Information online). Levels of gene expression were
2
.10 Cell extraction, culture and induction
Rabbit BMSCs were isolated from 4-week-old New Zealand
White rabbits using a whole bone marrow adherence method,
as previously described [16]. Briefly, bone marrow from
rabbit femurs was collected, filtered using 200 mesh, washed
in PBS, then cultured in DMEM/F12 (50 U/mL penicillin,
–
ΔΔCt
5
0 μg/mL streptomycin, 10% FBS) medium for approxi-
calculated using the 2
method and normalized against
mately 2 weeks. At an appropriate level of cell expansion,
the cells were digested (0.25% trypsin) and passaged for
subsequent use (Figure S2). Meanwhile, cartilage was ob-
tained from rabbit knee joints and chondrocytes isolated by
sequential trypsin and collagenase digestion. Chondrocyte
media consisted of DMEM/F12, antibiotics (100 U/mL pe-
nicillin and 100 μg/mL streptomycin) and 10% FBS.
GAPDH (housekeeping gene).
2
.14 Protein extraction and expression assessment
After 10 days of incubation, BMSC-laden hydrogels were
ground and lysed in RIPA buffer (containing PMSF) at 4 °C.
After quantification of protein concentration in samples

6
Rong et al. Sci China Chem
using a BCA kit (Thermo, USA), equal quantities of protein
per sample were electrophoresed in 10% polyacrylamide gel
then transferred to a polyvinylidene fluoride (PVDF) mem-
brane. The membranes were blocked in Tris-buffered saline/
Tween-20 (TBST) with 5% bovine serum albumin at room
temperature for 1 h. The membranes were then incubated in
fresh blocking buffer overnight at 4 °C with one of the fol-
lowing primary antibodies: ACAN Rabbit pAb (1:1,000,
ABclonal), COL2A1 Rabbit pAb (1:1,000, ABclonal) or β-
actin monoclonal antibody (1:2,000, rabbit, CST). Subse-
quently, the membranes were washed with TBST and in-
cubated with goat anti rabbit IgG-HRP secondary antibody
by ring opening polymerization of BLG NCA, which was
confirmed by H NMR.
Based on the gel permeation chromatography test, the
1
number-average molecular weight (M ) of PLG was 18,900,
n
with a PDI of 1.19. PLG-N or PLG-ADIBO was then syn-
3
thesized by grafting N - or ADIBO-bearing molecules to the
3
carboxyl side groups of PLG via carbodiimide chemistry. To
obtain PLG-N /RGD/N-Cad (or PLG-N /RGD/Scram),
3
3
PLG-N3 was additionally modified with amino-bearing
c(RGDfK) and N-cadherin mimetic peptide (or scrambled
peptide) using an EDC/NHS-mediated amidation reaction.
The grafting ratios of target molecules were calculated as the
number of grafted target groups per 100 L-glutamic acid
(1:2,000, ABclonal) at room temperature for 40 min. The
1
membranes were then washed in TBST and bands were vi-
sualized by enhanced chemiluminescence (ECL) using a
Western blotting substrate (Beyotime). The density of each
protein band was measured using Image J software. β-actin
represented the control.
residues based on H NMR results (Figure 2, Table 1).
ADIBO was grafted to PLG at a grafting ratio of approxi-
mately 6%, as determined by comparison of the integral of
the phenyl peak at around 7.2 ppm with that of the methylene
peak of the L-glutamate unit at around 4.3 ppm (Figure 2).
Similarly, the grafting ratios of N , RGD, N-cadherin mi-
3
metic peptide (N-Cad) and scrambled peptide (Scram) were
calculated as approximately 9%, 0.5%, 0.3%, and 0.3%, re-
spectively (Table 1). Furthermore, the concentration of RGD
was re-evaluated by measuring arginine content, resulting in
a grafting ratio of 0.54%. N-cadherin mimetic peptide and
scrambled peptide were retested using the Pauly reaction,
which indicated grafting ratios of 0.33% and 0.35% (Figure
3
Results and discussion
3
.1 Synthesis and characterization of PLG-ADIBO,
PLG-N , and peptide-modified PLG-N
3
3
The synthetic routes of PLG modification with N , ADIBO
and peptides are displayed in Scheme 1. PLG was obtained
3
Scheme 1 Routes of synthesis of PLG-ADIBO, PLG-N and PLG-N /RGD/N-Cad (color online).
3
3

Rong et al. Sci China Chem
7
1
Figure 2 H NMR spectra of PLG (a), PLG-ADIBO (b), PLG-N /RGD/N-Cad (c) in D O (color online).
3
2
Table 1 Reaction feeds, grafting ratios and grafting efficiencies of the PLG graft copolymers
Molecules to be
a)
grafted (X)
Reaction feed ratio:
X/–COOH (mol/mol)
b)
b)
Sample
Grafting ratio
Grafting efficiency
PLG-ADIBO
PLG-N3
ADIBO
N3
1:10
6%
9%
60%
90%
1:10
PLG-N /RGD/Scram
RGD, Scram
RGD, N-Cad
1:100, 1:100
1:100, 1:100
0.52%, 0.32%
0.49%, 0.3%
52%, 32%
49%, 30%
3
PLG-N /RGD/N-Cad
3
1
a) X represents the molecule to be grafted; b) calculated by H NMR.
S1(a, b)). Thus, the re-evaluated figures are consistent with
the H NMR data.
to study the viscoelastic properties of the hydrogels, the
storage modulus (G′) and loss modulus (G″) during the ge-
lation process were measured over time (Figure 3(d)). It was
observed that, where the crosspoint of G′ and G′′ was con-
sidered the gelation point, the gelation time reduced with
increasing polymer concentration from 3% to 6% (w/v). The
results agreed well with those of the vial inversion test.
Furthermore, G′ at the plateau increased markedly from 300
to 9,400 Pa as polymer concentration increased from 3% to
6% (w/v), likely due to the increase in crosslinking density. It
has been reported that hydrogels with relatively low cross-
linking density are more conducive to maintaining stem cell
viability and function [45–47], and so 3% (w/v) hydrogel was
selected for the remaining cell studies. Additionally, the
viscoelastic properties of hydrogels formed from PLG-
ADIBO with PLG-N /RGD/N-Cad or PLG-N /RGD/Scram
1
3
.2 Preparation and characterization of the hydrogels
Transparent PLG-based hydrogels were formed by simply
mixing PLG-N and PLG-ADIBO in PBS at an N /ADIBO
3
3
molar ratio of 1:1 without external stimulus or catalyst
Figure 3(a)). The rate of crosslinking was positively corre-
(
lated with polymer concentration. When the concentrations
of each were increased from 3% to 6% (w/v), gelation time of
the hydrogels reduced markedly from 35 to 1.5 min, as
measured using the vial inversion method (Figure 3(b)). In-
creased rate of gelation with increasing polymer concentra-
tion can be attributed to the greater concentrations of N and
3
ADIBO groups, which contributed to the crosslinking reac-
tion [25]. Furthermore, the result of SEM showed that the
PLG hydrogel had porous structure, which was conducive to
the circulation of bioactive molecules (Figure 3(c)). In order
3
3
were also tested (Figure S3). The results indicate that PLG
hydrogels with peptide modification exhibited comparable
G′ and G′′ to the PLG hydrogel without peptide modification,

8
Rong et al. Sci China Chem
Figure 3 (a) Photographs of the sol-gel transition of 3% (w/v) PLG. (b) Gelation time of different concentrations of PLG hydrogels at 37 °C. (c) SEM
image of lyophilized 3% (w/v) PLG hydrogels. Scale bar: 100 μm. (d) Storage modulus (G′) and loss modulus (G″) of PLG hydrogels with different
concentrations as a function of time. (e) In vitro degradation profiles for the 3% (w/v) hydrogels incubated in 0.01 M PBS (pH 7.4) containing 5 U/mL
proteinase K, 5 U/mL elastase, and PBS without any proteinase as control, respectively (n=3) (color online).
and so such differences are unlikely to cause additional in-
terference to subsequent cell tests [48].
measured by MTT, using PEI 25K (0.125 mg/mL) as a po-
sitive control. Cell viability in the presence of hydrogel ex-
tract was maintained above 90% (Figure 4(a)), indicating
that the hydrogel exhibited no detectable cytotoxicity.
BMSCs were additionally encapsulated in PLG hydrogels,
and cell morphology and survival monitored. The hydrogels
retained transparent during the experimental period of cell
culture, which was beneficial to the observation of cell status
within the hydrogels. Compared with the spreading observed
on tissue culture plastic (TCP), BMSCs instead retained a
spherical morphology when dispersed inside the hydrogels
(Figure 4(b, c)), consistent with previously-published lit-
erature [49]. Moreover, live-dead staining of the cells in-
dicated that the majority remained alive (stained green) after
14 days in culture (Figure 4(d)). The hydrogel demonstrated
excellent biocompatibility in vitro, as shown by the results of
cytotoxicity and 3D culture assay. The excellent cyto-
compatibility of the polypeptide hydrogel was likely due to
the low cytotoxicity of PLG and the mild gelation process of
the SPAAC reaction [24,50].
Degradation of the hydrogels in vitro was measured in the
presence of proteinases at 37 °C. As shown in Figure 3(e),
the hydrogels were relatively stable in PBS for 2 weeks,
whereas degradation of the hydrogels occurred after addition
of various proteinases. The gel samples degraded completely
in approximately 7 days in the presence of elastase, but more
rapidly, within 2 days, in the presence of protease K. No-
tably, the weight of the gels increased during the initial stages
then decreased in solutions containing enzymes. This can
reasonably be explained by the gradual cleavage of PLG-
based networks leading to a decrease in crosslinking density
and hydrogel swelling during the initial stage [17]. After
integrity of the hydrogels could not be maintained during the
later stages, the samples decomposed completely.
3
.3 In vitro and in vivo biocompatibility
In vitro biocompatibility of the PLG hydrogel was evaluated
by survival of the BMSCs. The cells were cultured with
hydrogel extract (72 h, 37 °C) for 24 h then cell viability was
In vivo degradation and biocompatibility of the hydrogels
were tested in a rat model. After mixing PLG-ADIBO and

Rong et al. Sci China Chem
9
Figure 4 (a) Viability of BMSCs after exposure to the supernatant of 3% (w/v) PLG hydrogels for 24 h. PEI 25K was used as a positive control (n=3). (b, c)
Confocal images of BMSCs after encapsulation in 3% (w/v) PLG hydrogels for 48 h: (b) cells were stained with Alexa Fluor 488-Phalloidin (F-actin, green)
and DAPI (cell nuclei, blue). Scale bar: 50 μm; (c) cells were stained with calcein-AM and PI. Scale bar: 100 μm. (d) Live-dead cell staining of BMSCs in 3%
(w/v) PLG hydrogels after incubation for 7 or 14 days. Cells were stained with calcein-AM (green, live) and PI (red, dead). Scale bar: 100 μm (color online).
PLG-N with or without peptide modification for 10 min,
(Figure 6(b)). It is noteworthy that, even though BMSCs
maintained high survival rate, the proliferation of BMSCs
within the hydrogels was slower than that of chondrocytes.
This is probably attributed to the fact that BMSCs were more
sensitive to the variation of microenvironment, leading to
relatively slower proliferation rate. Similar results have been
reported in some other systems, in which cells kept alive over
long-term culture in hydrogels, but no significant prolifera-
tion was observed [53,54]. The adhesion promoting effect of
RGD has been widely recognized. Song et al. [55] also es-
tablished that hydrogels modified with RGD promoted the
proliferation of MSCs within the gels. In the present study,
the bioactivity of hydrogels also improved by the introduc-
tion of RGD. In addition, the surface adhesion of BMSCs on
the hydrogels was measured using gradient RGD modified
gels (Figure S6). It was found that the modification of PLG
hydrogels with RGD (≥0.5 mM) promoted cell adhesion,
with cell proliferation being significantly enhanced (Figure
S7).
3
mixtures of the 3% (w/v) precursor solutions were injected
subcutaneously into the backs of rats for in situ gelation.
Stable hydrogels were formed within several minutes, and
the subcutaneous hydrogels degraded completely within 10
weeks.
The inflammatory reactions of the tissue surrounding the
hydrogels were evaluated using H&E staining. A slightly
increased accumulation of inflammatory cells was observed
within 2 weeks following injection of the hydrogel, implying
that a mild foreign body reaction had occurred [15]. It is
noteworthy that the number of inflammatory cells in the
vicinity decreased as gradual degradation of the hydrogels
occurred, suggesting a gradual alleviation of the in-
flammatory reaction (Figure 5 and Figures S4, S5). After the
complete degradation of the hydrogels, the inflammatory
reaction disappeared. Furthermore, the presence of hydrogels
caused no apparent necrosis, edema, hyperemia or hemor-
rhaging at the site of injection. The results above indicate that
the PLG hydrogel was biodegradable and exhibited good
biocompatibility in vivo.
3
.5 Differentiation of BMSCs into chondrocytes within
the hydrogels
3
.4 Cell proliferation within hydrogels
Chondrogenesis of BMSCs within the polypeptide hydrogel
was induced using chondrogenic induction medium. The
secretion of glycosaminoglycans (GAGs) is a characteristic
of chondrocytes, so measurement of the accumulation of
GAGs within the hydrogels is evidence of the differentiation
of BMSCs into chondrocytes [56]. After incubation of
BMSCs in hydrogels in the presence of chondrogenic in-
duction medium for 3 weeks, the yield of GAGs was mea-
sured by DMMB and normalized to DNA content. As shown
The 3% (w/v) PLG hydrogel formed from PLG-N and PLG-
3
ADIBO was used as the scaffold for 3D cell culture, with cell
proliferation measured using a CCK-8 assay. Chondrocytes
that had been encapsulated in the hydrogels and cultured for
a week, had proliferated continuously, as displayed in Figure
6
(a). Notably, the BMSCs did not proliferate within the PLG
hydrogels, as expected [51,52]. However, cell proliferation
within c(RGDfK)-modified hydrogels increased somewhat

10
Rong et al. Sci China Chem
Figure 5 (a) In vivo implantation of 3% (w/v) PLG hydrogels after various time intervals. (b) H&E staining images of tissues surrounding the injection sites
at different time periods. Scale bar: 100 μm (color online).
Figure 6 (a) Proliferation of chondrocytes encapsulated in 3% (w/v) PLG hydrogels as a function of time. (b) Proliferation of BMSCs within the three types
of hydrogel: 3% (w/v) PLG hydrogels, 3% (w/v) PLG+RGD/Scram hydrogels, 3% (w/v) PLG+RGD/N-Cad hydrogels (n=3), (**p<0.01, ***p<0.001) (color
online).
in Figure 7(a), the quantity of GAG generated by the cells
within the PLG+RGD/N-Cad hydrogel was significantly
higher than that secreted by cells in PLG and PLG+RGD/
Scram hydrogels. The result indicates that the N-cadherin
mimetic peptide promoted chondrogenic differentiation of
BMSCs. Furthermore, similar results of gene and protein
expression were obtained. It was found that the levels of
aggrecan and type II collagen (collagen II) gene expression
increased in all three groups during 1 week incubation in
chondrogenic induction media (Figure 7(b, c)). However, the
gene expressions of these markers of chondrogenic differ-
entiation following culture in the PLG+RGD/N-Cad hydro-
gel were significantly higher than those in other groups. The
Western blot (WB) bands indicated that protein expression of
aggrecan and collagen II were also correspondingly en-
hanced in the hydrogel modified with N-cadherin mimetic

Rong et al. Sci China Chem
11
Figure 7 Chondrogenic differentiation of BMSCs in response to N-Cad-modified 3D hydrogels. (a) Quantification of GAG in hydrogels containing
BMSCs after induction of chondrogenesis in culture for 3 weeks (normalized to DNA content). (b, c) Quantitative RT-PCR of aggrecan (b) and collagen II
(c) in BMSCs within hydrogels during the first week of induced differentiation (normalized to the housekeeping gene GAPDH). (d, e) Representative Western
blots for aggrecan and collagen II of BMSCs after induction of differentiation for 10 days (n=3), (*p<0.05, **p<0.01) (color online).
peptide (Figure 7(d, e)). These results indicate that the N-
cadherin mimetic peptide in the PLG hydrogel promotes
chondrogenic differentiation of BMSCs, likely due to the
promotion of cell-cell interactions caused by the peptide
variety of proteinases. From the cell tests and in vivo gelation
experiments, it was found that the hydrogels exhibited ex-
cellent cytocompatibility in vitro and displayed good bio-
compatibility in vivo. Conjugation of RGD to the hydrogels
promoted the adhesion and proliferation of BMSCs on the
surface of the PLG hydrogels, and also increased the pro-
liferation of BMSCs within the hydrogels in a 3D culture
model. In particular, PLG hydrogels modified with both
RGD and N-Cad were more conducive to the promotion of
chondrogenic differentiation of BMSCs encapsulated within
the hydrogels. Therefore, bioactive polypeptide hydrogels
represent promising materials for use as scaffolds for cell
culture and cartilage tissue engineering.
[38,39]. Furthermore, the results also established the feasi-
bility of using PLG hydrogels grafted with active peptides to
regulate cell behavior.
4
Conclusions
In the present study, PLG hydrogels were formed by copper-
free click reaction chemistry between PLG-N and PLG-
3
ADIBO under physiological conditions. The gelation time
and mechanical properties of the polypeptide hydrogels were
adjusted by changing the concentration of the precursor
polymer. The PLG hydrogels degraded in the presence of a
Acknowledgements This work was supported by the National Natural
Science Foundation of China (51973218, 51622307, 21574127,
51520105004) and the Youth Innovation Promotion Association, Chinese
Academy of Sciences.

12
Rong et al. Sci China Chem
Conflict of interest The authors declare no conflict of interest.
444
2
7
8
Tibbitt MW, Rodell CB, Burdick JA, Anseth KS. Proc Natl Acad Sci
USA, 2015, 112: 14444–14451
Shi S, Yu SJ, Li G, He CL, Chen XS. Sci China Technol Sci, 2020,
https://doi.org/10.1007/s11431-019-1466-1
Supporting information The supporting information is available online
at http://chem.scichina.com and http://link.springer.com/journal/11426. The
supporting materials are published as submitted, without typesetting or
editing. The responsibility for scientific accuracy and content remains en-
tirely with the authors.
2
2
3
3
3
9
0
1
2
Hersel U, Dahmen C, Kessler H. Biomaterials, 2003, 24: 4385–4415
Leahy DJ, Aukhil I, Erickson HP. Cell, 1996, 84: 155–164
Yao X, Peng R, Ding J. Adv Mater, 2013, 25: 5257–5286
DeLise AM, Tuan RS. J Cell Biochem, 2002, 87: 342–359
1
2
3
Langer R. Mol Ther, 2000, 1: 12–15
Lee KY, Mooney DJ. Chem Rev, 2001, 101: 1869–1880
Moon HJ, Ko DY, Park MH, Joo MK, Jeong B. Chem Soc Rev, 2012,
33 Delise AM, Tuan RS. Dev Dyn, 2002, 225: 195–204
34 Gumbiner BM. Nat Rev Mol Cell Biol, 2005, 6: 622–634
35 Blaschuk OW, Sullivan R, David S, Pouliot Y. Dev Biol, 1990, 139:
227–229
36 Williams E, Williams G, Gour BJ, Blaschuk OW, Doherty P. J Biol
Chem, 2000, 275: 4007–4012
4
1: 4860
Tsou YH, Khoneisser J, Huang PC, Xu X. Bioactive Mater, 2016, 1:
9–55
4
5
3
Zhong L, Qu Y, Shi K, Chu B, Lei M, Huang K, Gu Y, Qian Z. Sci
China Chem, 2018, 61: 1553–1567
37 Li R, Xu J, Wong DSH, Li J, Zhao P, Bian L. Biomaterials, 2017, 145:
33–43
6
7
Yu L, Ding J. Chem Soc Rev, 2008, 37: 1473
Rice JJ, Martino MM, De Laporte L, Tortelli F, Briquez PS, Hubbell
JA. Adv Healthc Mater, 2013, 2: 57–71
Annabi N, Tamayol A, Uquillas JA, Akbari M, Bertassoni LE, Cha C,
Camci-Unal G, Dokmeci MR, Peppas NA, Khademhosseini A. Adv
Mater, 2015, 26: 85–124
38 Bian L, Guvendiren M, Mauck RL, Burdick JA. Proc Natl Acad Sci
USA, 2013, 110: 10117–10122
39 Cosgrove BD, Mui KL, Driscoll TP, Caliari SR, Mehta KD, Assoian
RK, Burdick JA, Mauck RL. Nat Mater, 2016, 15: 1297–1306
40 Kwon MY, Vega SL, Gramlich WM, Kim M, Mauck RL, Burdick JA.
Adv Healthc Mater, 2018, 7: 1701199
8
9
0
Xue K, Wang X, Yong PW, Young DJ, Wu YL, Li Z, Loh XJ. Adv
Therap, 2019, 2: 1800088
Lv J, Wu G, Liu Y, Li C, Huang F, Zhang Y, Liu J, An Y, Ma R, Shi
L. Sci China Chem, 2019, 62: 637–648
41 Zhao C, Zhuang X, He C, Chen X, Jing X. Macromol Rapid Commun,
2008, 29: 1810–1816
42 Cheng Y, He C, Xiao C, Ding J, Zhuang X, Chen X. Polym Chem,
2011, 2: 2627
1
1
1
Deming TJ. Prog Polym Sci, 2007, 32: 858–875
He C, Zhuang X, Tang Z, Tian H, Chen X. Adv Healthc Mater, 2012,
43 Graf N, Bielenberg DR, Kolishetti N, Muus C, Banyard J, Farokhzad
OC, Lippard SJ. ACS Nano, 2012, 6: 4530–4539
44 Mann T, Leone E. Biochem J, 1953, 53: 140–148
45 Chung C, Anderson E, Pera RR, Pruitt BL, Heilshorn SC. Soft Matter,
2012, 8: 10141
12
13
14
1: 48–78
Shirbin SJ, Karimi F, Chan NJA, Heath DE, Qiao GG. Biomacro-
molecules, 2016, 17: 2981–2991
Hu C, Liu X, Ran W, Meng J, Zhai Y, Zhang P, Yin Q, Yu H, Zhang
Z, Li Y. Biomaterials, 2017, 144: 60–72
46 Bian L, Hou C, Tous E, Rai R, Mauck RL, Burdick JA. Biomaterials,
2013, 34: 413–421
1
1
5
6
Ren K, He C, Xiao C, Li G, Chen X. Biomaterials, 2015, 51: 238–249
Ren K, Cui H, Xu Q, He C, Li G, Chen X. Biomacromolecules, 2016,
47 M. Jonker A, A. Bode S, H. Kusters A, van Hest JCM, Löwik DWPM.
Macromol Biosci, 2015, 15: 1338–1347
1
7: 3862–3871
Xu Q, He C, Zhang Z, Ren K, Chen X. ACS Appl Mater Interfaces,
016, 8: 30692–30702
48 Even-Ram S, Artym V, Yamada KM. Cell, 2006, 126: 645–647
49 Sridhar BV, Brock JL, Silver JS, Leight JL, Randolph MA, Anseth
KS. Adv Healthc Mater, 2015, 4: 702–713
1
7
2
1
1
8
9
Zhou X, Li Z. Adv Healthc Mater, 2018, 7: 1800020
Patel M, Park S, Lee HJ, Jeong B. Tissue Eng Regen Med, 2018, 15:
50 Ren K, He C, Cheng Y, Li G, Chen X. Polym Chem, 2014, 5: 5069–
5076
5
21–530
51 Park H, Choi B, Hu J, Lee M. Acta Biomater, 2013, 9: 4779–4786
52 Qu C, Bao Z, Zhang X, Wang Z, Ren J, Zhou Z, Tian M, Cheng X,
Chen X, Feng C. Int J Biol Macromolecules, 2019, 125: 78–86
53 Lueckgen A, Garske DS, Ellinghaus A, Mooney DJ, Duda GN, Ci-
pitria A. Biomaterials, 2019, 217: 119294
2
2
0
1
Liu R, Shi Z, Sun J, Li Z. Sci China Chem, 2018, 61: 1414–1319
Yan S, Zhang X, Zhang K, Di H, Feng L, Li G, Fang J, Cui L, Chen X,
Yin J. J Mater Chem B, 2016, 4: 947–961
2
2
Agard NJ, Prescher JA, Bertozzi CR. J Am Chem Soc, 2004, 126:
1
5046–15047
54 Ansari S, Chen C, Xu X, Annabi N, Zadeh HH, Wu BM, Kha-
demhosseini A, Shi S, Moshaverinia A. Ann Biomed Eng, 2016, 44:
1908–1920
2
2
3
4
Kim E, Koo H. Chem Sci, 2019, 10: 7835–7851
Madl CM, Katz LM, Heilshorn SC. Adv Funct Mater, 2016, 26: 3612–
3
620
55 Hong KH, Song SC. Biomaterials, 2019, 218: 119338
56 Yang J, Zhang YS, Yue K, Khademhosseini A. Acta Biomater, 2017,
57: 1–25
2
2
5
6
Xu J, Filion TM, Prifti F, Song J. Chem Asian J, 2011, 6: 2730–2737
DeForest CA, Anseth KS. Annu Rev Chem Biomol Eng, 2012, 3: 421–