Three-dimensional fabrication of cell-laden biodegradable poly(ethylene glycol-co-depsipeptide) hydrogels by visible light stereolithography

Article abstract of DOI:10.1039/c5tb01468a

Stereolithography (SLA) holds great promise in the fabrication of cell-laden hydrogels with biomimetic complexity for use in tissue engineering and pharmaceutics. However, the availability of biodegradable photocrosslinkable hydrogel polymers for SLA is v

Full text of DOI:10.1039/c5tb01468a

Journal of
Materials Chemistry B
View Article Online
View Journal | View Issue
PAPER
Three-dimensional fabrication of cell-laden
biodegradable poly(ethylene glycol-co-depsipeptide)
hydrogels by visible light stereolithography
Cite this: J. Mater. Chem. B, 2015,
, 8348
3
ab
ac
a
d
Laura Elomaa, Chi-Chun Pan, Yaser Shanjani, Andrey Malkovskiy,
b
aef
Jukka V. Seppala and Yunzhi Yang*
¨
¨
Stereolithography (SLA) holds great promise in the fabrication of cell-laden hydrogels with biomimetic
complexity for use in tissue engineering and pharmaceutics. However, the availability of biodegradable
photocrosslinkable hydrogel polymers for SLA is very limited. In this study, a water-soluble methacrylated
poly(ethylene glycol-co-depsipeptide) was synthesized to yield
a biodegradable photocrosslinkable
macromer for SLA. Structural analysis confirmed the inclusion of biodegradable peptide and ester groups
and photocrosslinkable methacrylate groups into the polymer backbone. The new macromer combined
with the RGDS peptide was used for the SLA fabrication of hydrogels in the absence and the presence of
cells. With the increasing light exposure time in SLA, the mechanical stiffness of the hydrogels increased
from 3 ꢀ 1 kPa to 38 ꢀ 13 kPa. The total mass loss of the samples within 7 days in PBS was 13–21%
and within 24 days was 35–66%. Due to degradation, the mechanical stiffness decreased by one order
magnitude within 7-day incubation in PBS. Encapsulated endothelial cells proliferated in the hydrogels
during the 10-day in vitro cell culturing study. The macromer was further used in SLA to fabricate
bifurcating tubular structures as preliminary vessel grafts. The new biodegradable, photocrosslinkable
polymer is a significant addition to the very limited material selection currently available for the SLA-based
additive manufacturing of cell-laden tissue engineering constructs.
Received 22nd July 2015,
Accepted 9th September 2015
DOI: 10.1039/c5tb01468a
www.rsc.org/MaterialsB
reducing the potential risk of harming DNA of encapsulated
cells. The recent progress of biomaterials and 3D fabrication
1
. Introduction
4
Stereolithography (SLA) is increasingly used for the fabrication of technology makes SLA increasingly attractive for the fabrication
biomimetic complex scaffolds and hydrogels for tissue engineering of hydrogel constructs with complex designs and enhanced
(TE) and pharmaceutics. It enables high-resolution and relatively functionality that cannot be achieved via more traditional
fast scaffold fabrication in the absence of harsh chemicals and high fabrication methods.
1–3
fabrication temperature. Of various SLA techniques, digital light
To fabricate cell-laden hydrogel constructs by SLA, a bio-
processing (DLP)-based SLA is particularly suitable for the three- compatible, biodegradable hydrogel polymer with appropriate
dimensional (3D) fabrication of cell-laden hydrogels due to its photocrosslinking kinetics and cell attachment properties is
shorter production times compared with traditional laser-based crucially needed. However, the availability of such polymers for SLA
1
SLA techniques and the use of visible light instead of UV light, is currently very limited. Previously, naturally derived, biodegradable
methacrylated gelatin has been used for the SLA-based fabrication
of cell-free hydrogels, and synthetic (meth)acrylated poly(ethylene
a
5,6
Department of Orthopaedic Surgery, Stanford University School of Medicine,
00 Pasteur Drive, Stanford, CA 94305, USA. E-mail: ypyang@stanford.edu;
3
glycol) (PEG) has been applied in several cell-laden, non-degradable
Fax: +1 650 721 5404; Tel: +1 650 723 0772
7–11
hydrogels.
Even though naturally derived polymers demonstrate
b
c
d
e
f
Department of Biotechnology and Chemical Technology, Aalto University School of
Chemical Technology, Kemistintie 1, 02150 Espoo, Finland
biodegradability and good cell adhesion properties, they do not
typically allow for controlled tuning of their properties without
Department of Mechanical Engineering, Stanford University School of Engineering,
Building 380, Sloan Mathematical Center, Stanford, CA 94305, USA
Biomaterials and Advanced Drug Delivery Laboratory, Stanford University School
of Medicine, 269 Campus Drive, Stanford, CA 94305, USA
1
2
simultaneously interfering their bioactivity. Instead, synthetic
PEG allows for systematic modification of its physical and
mechanical properties to satisfy the requirements of a variety
Department of Materials Science and Engineering, Stanford University School of
Engineering, 496 Lomita Mall, Stanford, CA 94305, USA
1
2,13
of target applications.
However, PEG is known to be resistant
1
4
to protein adsorption and cell attachment and is inherently
non-degradable, which restricts the migration and proliferation
Department of Bioengineering, 443 Via Ortega, Stanford University School of
Engineering, Stanford, CA 94305, USA
8
348 | J. Mater. Chem. B, 2015, 3, 8348--8358
This journal is ©The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry B
9
ꢁ1
of encapsulated cells inside the crosslinked hydrogel. The lack of Star-shaped four-arm PEG (10 000 g mol , Creative PEGworks)
controlled degradation of PEG has been previously addressed by was used after drying under vacuum at 100 1C for 3 h. Acryl-PEG-SVA
1
2,15
ꢁ1
incorporating enzymatically cleavable peptides
labile ester bonds
or hydrolytically (3400 g mol , Laysan Bio, Inc.) and the RGDS peptide (Bachem)
3
,16
into its backbone. However, the photo- were used as received. Immortalized HUVECs expressing green
crosslinking properties and fabrication parameters of these fluorescent protein (GFP) were a generous gift from the late
biodegradable polymers have not been optimized for use in Dr J. Folkman, Children’s Hospital, Boston. Fetal bovine serum
the SLA-based fabrication of cell-laden hydrogels. To success- (FBS), antibiotic solution, and trypsin-EDTA solution were purchased
fully 3D fabricate cell-laden hydrogels, it is essential to under- from Invitrogen Co. EBMt endothelial basal medium containing
stand how parameters of both the material and the SLA process and EGMt Single Quotst kits were purchased from Lonza, Inc.
s
affect the resulting hydrogel properties. In a photocrosslinked AlamarBlue cell viability assay was from Bio-Rad Laboratories.
hydrogel, properties like swelling capacity, material degrada-
tion, and mechanical stiffness depend directly on the cross-
2.2 Photoinitiator synthesis
linking density of the gel, which in turn is known to depend on A lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP) photo-
the parameters like polymer concentration and molecular initiator was synthesized as previously described by Fairbanks
24
weight as well as photoinitiator concentration and crosslinking et al. Briefly, 2,4,6-trimethylbenzoyl chloride was added dropwise
13,17,18
energy.
The understanding of correlation between fabrication to an equimolar amount of dimethyl phenylphosphonite at RT
parameters and hydrogel properties will allow for better control under a nitrogen atmosphere. After mixing for 18 h, a fourfold
over the characteristics of cell-laden hydrogels and will help to excess of lithium bromide dissolved in 2-butanone was added to
regulate and promote the intended functions of encapsulated cells the mixture, and the reaction was continued for 10 min at 50 1C.
1
5,19,20
in the hydrogel matrix.
After heating, the mixture was kept at RT for 4 h; the resulting
21
In our previous study, we synthesized a biodegradable photo- precipitate was filtered and washed once with 2-butanone, and
crosslinkable poly(e-caprolactone-co-depsipeptide) for use in DLP- then twice with diethyl ether, and finally dried under vacuum.
based SLA to fabricate porous hydrophobic TE scaffolds. Previously,
depsipeptides have been used to synthesize hydrophilic PEG-based
copolymers for use as non-photocrosslinkable, temperature
2.3 Synthesis of photocrosslinkable macromers and adhesion
peptides
2
2,23
responsive gels and drug delivery micelles.
In our current A biodegradable, water-soluble PEG-co-PDP oligomer was
work, we used the L-alanine-derived depsipeptide to synthesize a synthesized by the ring-opening polymerization (ROP) of the
new biodegradable, photocrosslinkable poly(ethylene glycol-co- L-alanine-derived depsipeptide using 4-arm PEG as a macro-
depsipeptide) (PEG-co-PDP) macromer for the SLA-based fabrica- initiator as shown in Scheme 1A. Before ROP, a 3-methyl-
tion of cell-laden hydrogel constructs. We hypothesized that morpholine-2,5-dione (MMD) monomer was synthesized as
2
1
the new photocrosslinkable macromer, which combines both described in our previous study. Briefly, L-alanine was reacted
naturally derived and synthetic building blocks, will allow for the with an equimolar amount of chloroacetyl chloride in the
3
D fabrication of biodegradable tissue engineering grafts that diethyl ether/water mixture at ꢁ5 1C by simultaneously adding
support cell proliferation within the hydrogel matrix. Subse- 4 M NaOH solution into the flask to maintain the pH value at
quently, we studied the effect of light exposure time in SLA on 11. After adding dropwise chloroacetyl chloride and NaOH for
the hydrogel characteristics and hypothesized that by controlling 20 min, the aqueous layer was acidified to pH 1 with 4 M HCl
the light exposure time, we can tune the hydrogel properties and was extracted twice with ethyl acetate. The combined
without the need for changing the parameters of the hydrogel organic layers were washed twice with saturated NaCl solution,
4
solution during the fabrication process. We evaluated the suitability dried with MgSO and concentrated under vacuum. The resulting
of our new polymer for the fabrication of cell-laden hydrogels by chloroacetyl alanine was then dissolved in dry N,N-dimethyl-
first characterizing its chemical structure and gelling properties as formamide (DMF) and mixed dropwise with triethylamine. After
well as its hydrolytic mass loss and mechanical properties, and next 8 h at 90 1C under nitrogen and overnight at room temperature, the
we studied its cell encapsulation capacity with human umbilical reaction solution was filtered and most of the solvent was removed
vein endothelial cells (HUVECs). Finally, we demonstrated the use under reduced pressure. The resulting cyclic MMD in a small
of our new polymer for the SLA-based fabrication of tubular amount of DMF was then purified with chloroform and diethyl
structures as preliminary vessel graft models.
ether and dried under vacuum. To polymerize the hydroxyl-ended
PEG-co-PDP oligomer with a designated molecular weight of
ꢁ
1
1
1 000 g mol , the MMD monomer was added with the dried
PEG macroinitiator and the Sn(Oct) catalyst to a flask, which was
then evacuated and flushed several times with nitrogen. ROP was
continued for 6 days at 130 1C, after which the resulting oligomer
2
2
. Experimental
2.1 Materials
L-Alanine and chloroacetyl chloride (Acros Organics), tin(II) was purified by dialysis and freeze-drying. To obtain a photo-
2
-ethylhexanoate (Sn(Oct) ) and methacrylic anhydride (Sigma- crosslinkable macromer, the oligomer was further reacted with
2
Aldrich), 2,4,6-trimethylbenzoyl chloride (Chem-Impex Int’l Inc.), an excess of methacrylic anhydride at 60 1C for 24 h, after which
dimethyl phenylphosphonite (Alfa Aesar), lithium bromide (Strem it was precipitated in cold diethyl ether and dried under vacuum.
Chemicals), and 2-butanone (Acros Organics) were used as received. The macromer was then purified by dialysis and freeze-drying.
This journal is ©The Royal Society of Chemistry 2015
J. Mater. Chem. B, 2015, 3, 8348--8358 | 8349

View Article Online
Journal of Materials Chemistry B
Paper
3 2
chloroform (CDCl ) or in deuterium oxide (D O), and the
spectrum was acquired in a 5 mm NMR tube at RT. The
presence of peptide and ester bonds in the PEG-co-PDP copolymer
was analyzed using a Bruker Vertex 70 FTIR spectrometer with a
diamond ATR. The thermal properties of the oligomers and macro-
mers were determined using a Q2000 differential scanning calori-
meter (DSC, TA Instruments). Samples were first heated from 25 1C
ꢁ1
to 100 1C at a rate of 20 1C min , after which the temperature was
ꢁ1
decreased to ꢁ100 1C at 10 1C min . After 2 min at ꢁ100 1C, the
ꢁ
1
sample was heated from ꢁ100 1C to 100 1C at 10 1C min . The
melting point (T ) and the melting enthalpy (DH) were determined
m
from the second heating scan. Gel content (G) of the photocros-
slinked hydrogels was measured by extracting the pre-weighed
hydrogel in distilled water for 10 h and subsequently drying the
sample to obtain its dry weight. G was then calculated by dividing
d
the dry weight of the sample after extraction (m ) by its dry weight
25
0
before extraction (m ) according to eqn (1):
G = m /m
d
0
ꢂ 100%
(1)
To measure the initial swelling degree (Q) of hydrogels, the
samples were immersed in distilled water for 24 h to reach the
swelling equilibrium, after which the weighed samples were
dried and weighed again. By using the wet weight (msw) and dry
Scheme 1 (A) Synthesis of the cyclic MMD monomer and further ROP
and methacrylation of the PEG-co-PDP copolymer. (B) Synthesis of an
acryl-PEG-RGDS adhesion peptide.
ꢁ1
weight (m
for water (r
As a control polymer, the PEG macromer was synthesized by based polymer (r
methacrylating the four-arm PEG macroinitiator with methacrylic eqn (2):
anhydride at 60 1C for 24 h. In addition, to enable covalent linking
of the RGDS adhesion peptide to the PEG-co-PDP hydrogel polymer,
d
) of the samples and the density values of 1.0 g ml
ꢁ1
s
) and 1.2 g ml for the photocrosslinked PEG-
2
6
p
), the swelling degree was calculated using
2
5
Q = 1 + r
p
/r
s
ꢂ (msw/m
d
ꢁ 1)
(2)
RGDS was coupled with an acryl-PEG linker as shown in Scheme 1B. The in vitro mass loss of hydrogels was monitored by measuring
Acryl-PEG-SVA and RGDS were dissolved in 50 mM NaHCO
solution the dry weight of the samples after incubation in PBS (pH 7.4,
pH 8.3) and mixed for 4 h at RT. The resulting acryl-PEG-RGDS was with 0.02% sodium azide) at 37 1C. At days 7, 14, and 24,
3
(
then purified by dialysis and freeze-drying.
triplicate samples were removed from PBS, weighed, and dried
first at ambient pressure for 2 days and then in a freeze-dryer
for 3 days. After drying, the remaining mass was divided by the
initial mass to obtain the remaining mass percentage. The
swelling degree during the degradation study was calculated
using eqn (2) as described before. The viscoelastic behavior
of 3D fabricated hydrogels was analyzed using an Ares G2
rheometer (TA Instruments) using an 8 mm parallel plate
geometry. The viscoelastic regime of the samples was first
determined with an amplitude sweep from 0.1 to 50% strain,
after which the angular frequency sweep was run from 0.1 to
2.4 Fabrication of cell-free hydrogels by stereolithography
To prepare a photocrosslinkable hydrogel solution for use in
SLA, the photocrosslinkable PEG-co-PDP macromer (10% w/v)
and the LAP photoinitiator (0.25% w/v) were dissolved in EBM-2
medium. The viscosity of the filtered solution was increased by
s
10
adding 7% w/v of Percoll medium to the polymer solution.
For degradation and mechanical studies, cell-free PEG-co-PDP
hydrogels (d = 8.5 mm, h = 0.6 mm) were fabricated using an
in-house built visible light projection SLA (3200 lumens) in a
layer-by-layer manner with a layer thickness of 200 mm. The
layer exposure time was 100 s, 120 s, or 160 s. For SLA
fabrication, the digital 3D model of the hydrogels was designed
using SolidWorks CAD software and was converted into printing
instructions using free Slic3r software (Slic3r.org). The light
exposure time was controlled using free Pronterface software
ꢁ1
10 rad s
at 0.5% strain and 25 1C. AFM force–distance
measurements were carried out using an Agilent 5500 AFM in
a liquid cell with PBS as the liquid phase. Tip sensitivity
calibration was performed on the glass surface after the end
of each set of experiments with the respective tip. The tip used
ꢁ1
for the experiments (NanoAndMore, k = 0.08 N m , R =
50 nm) had a SiO sphere on its apex. The stiffness of tips
9
2
(pronterface.com).
from this batch was confirmed by the Sader method and was
found to be o20% different from the nominal reported value.
FDS curve data analysis was carried out using commercially
available SPIP software from Image Metrology A/S. The baseline
2.5 Physicochemical characterization of the polymers and the
cell-free hydrogels
1
H NMR spectra were recorded on a Varian Inova 300 MHz correction from the laser reflection and the determination of
NMR spectrometer. Samples were dissolved in deuterated the Young’s modulus using Hertz sphere approximation were
8
350 | J. Mater. Chem. B, 2015, 3, 8348--8358
This journal is ©The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry B
performed in batch processing regimes, with the results being described for the disc samples using 200 mm as a layer thick-
checked visually for mistakes. ness and 120 s as a light exposure time.
2.7 Cell proliferation
2
.6 Cell cultures and 3D fabrication of cell-laden hydrogels
The metabolic activity of seeded or encapsulated HUVECs was
s
HUVECs were cultured in endothelial basal medium (EBM-2) measured using a colorimetric AlamarBlue cell viability assay.
mixed with a Single Quots (EGM-2, excluding hydrocortisone) At predetermined time points, the hydrogel samples were
endothelial growth supplement, 10% FBS, and 1% antibiotic transferred to a new well-plate, and 500 mL of fresh medium
s
2
solution under standard conditions (5% CO , 37 1C, and 95% and 50 mL of AlamarBlue solution were added into the wells.
humidity). To study the material cytocompatibility, two cell-free After incubation for 4 h, the medium was aliquoted to three
hydrogel groups were fabricated by SLA using the hydrogel parallel 100 mL samples, and the optical density was measured
s
solutions consisting of the neat PEG-co-PDP macromer or PEG- using a SpectraMax
M2e microplate reader (Molecular
co-PDP macromer mixed with 2 mM acryl-PEG-RGDS. HUVECs Devices) at 555 nm/585 nm excision/emission.
3
(
1 ꢂ 10 cells) were seeded on top of the hydrogels (d = 8.5 mm
2.8 Statistical analysis
and h = 0.6 mm). For cell encapsulation studies, cell-laden
hydrogel discs were prepared by SLA using the filtered PEG-co- For the statistical analysis, samples were used in triplicate, and
6
PDP/RGDS solution mixed with HUVECs (5 ꢂ 10 cells per ml) the statistical significance was determined using the IBM SPSS
and the same fabrication parameters as for the cell-free hydro- Statistics software using a one-way ANOVA followed by Tukey’s
gels (Fig. 1). Before hydrogel fabrication, the container for post hoc test ( p o 0.05).
hydrogel solution was sterilized with 70% ethanol followed by
thorough rinsing with PBS, and the air in the SLA system was
continuously filtered with a HEPA filter. To further demon-
strate the use of the PEG-co-PDP polymer for the 3D fabrication
3
. Results
3.1 The macromer synthesis
of vascular tissue engineering grafts, cell-laden hydrogel rings The water-soluble photocrosslinkable PEG-co-PDP macromer
and bifurcating hydrogel vessels were fabricated by SLA as for use in the 3D fabrication of cell-laden hydrogels was
Fig. 1 A schematic presentation of the SLA-based fabrication of cell-laden hydrogels: (A) light is projected to photocrosslinkable hydrogel solution to
form the designed pattern, and the hydrogel is crosslinked in a layer-by-layer manner; (B) after crosslinking the last layer, the platform is moved up and
the resulting hydrogel is removed from the SLA.
This journal is ©The Royal Society of Chemistry 2015
J. Mater. Chem. B, 2015, 3, 8348--8358 | 8351

View Article Online
Journal of Materials Chemistry B
Paper
1
Fig. 2 H NMR spectra of (A) MMD monomer and PEG before ROP, (B) PEG-co-PDP oligomer after ROP, and (C) photocrosslinkable PEG-co-PDP
3
macromer after methacrylation, all in CDCl .
synthesized by ROP of the L-alanine-derived depsipeptide and a appearance of the peaks i, j, and k in Fig. 2C attributed to the
-arm PEG macroinitiator followed by the methacrylation. The methacrylate groups of the macromer. No trace of oligomeric
4
1
reactions were monitored by H NMR characterization as peak 1 was left in the macromer, showing the complete
shown in Fig. 2. The cyclic MMD monomer that was synthe- methacrylation. The dialyzed and freeze-dried macromer was
1
sized from L-alanine and chloroacetyl chloride gave the H NMR a light-brown powder showing great water-solubility at room
peaks 1–3 in Fig. 2A. The MMD monomer was then copolymerized temperature. The yield of the purified macromer was 97%.
with PEG, resulting in a water-soluble, hydroxyl-terminated Some residual water remained in the polymer despite the
1
oligomer. In the H NMR spectrum (Fig. 2B), the monomer extended freeze-drying, giving a broad peak near 2 ppm in
1
peaks transferred to the corresponding oligomer peaks 1–3, the H NMR spectrum.
0
while new peak b appeared and was attributed to the last CH
2
RGDS conjugation with acryl-PEG-SVA was confirmed using
H NMR. After 4 h of mixing, the proton peak g of the acryl-PEG-
1
protons of the PEG arms before depsipeptide units. 2D COSY
H NMR revealed a strong coupling between peaks 2 and 3 of SVA end-group shown in Fig. 3A disappeared and new peaks
the oligomer. As calculated from the H NMR spectrum, the assigned to the RGDS peptide appeared as shown in Fig. 3B.
1
1
polymerization degree of the MMD monomer was around 85%, After dialysis and freeze-drying, acryl-PEG-RGDS was a water-
ꢁ
1
when the molecular weight of the oligomer was approximately soluble white powder with a molecular weight of 3700 g mol
.
ꢁ
1
10 850 g mol . The yield of the oligomer was 98%. The
Copolymerization of depsipeptide units into PEG was mon-
oligomer was further functionalized with methacrylate end- itored by FTIR. Fig. 4 shows the spectra of hydroxyl-ended PEG
groups to obtain the macromer with photocrosslinkable double before copolymerization and the resulting PEG-co-PDP oligomer.
bonds. The successful methacrylation was observed as a dis- The FTIR analysis revealed that the ROP of PEG with MMD
ꢁ
1
ꢁ1
appearance of peak 1 in Fig. 2B attributed to the last CH
2
resulted in new peaks at 1750 cm and 1670 cm attributed to
protons before the hydroxyl groups of the oligomer and as an the CQO ester and CQO amide stretches, respectively, and a
8
352 | J. Mater. Chem. B, 2015, 3, 8348--8358
This journal is ©The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry B
1
2
Fig. 3 H NMR spectra of (A) acryl-PEG-SVA and (B) acryl-PEG-RGDS in D O.
Table 1 Melting point and melting enthalpy of the oligomers and
macromers
Oligomer
(1C)
Macromer
(1C)
ꢁ1
ꢁ1
Sample
T
m
DH (J g
)
T
m
DH (J g )
PEG
PEG-co-PDP
55.82
45.43
160.3
81.38
43.98
42.59
100.9
76.60
m
The methacrylation decreased the T and DH values both for the
PEG control polymer and for the PEG-co-PDP copolymer.
3.2 Cytocompatibility of photocrosslinked hydrogels
To study the in vitro cytocompatibility of the methacrylated
PEG-co-PDP polymer and to study the effect of the RGDS
adhesion peptide on cell proliferation, PEG-co-PDP and PEG-
Fig. 4 FTIR spectra of the PEG homopolymer and the PEG-co-PDP co-PDP/RGDS hydrogels were photocrosslinked and HUVECs
copolymer showing the new ester and amide bonds of the copolymer.
were seeded on the hydrogels. The metabolic activity of
HUVECs, which is an indicator of the number of live cells,
new peak at 1540 cm attributed to the N–H bend of the was monitored using the colorimetric AlamarBlue assay.
depsipeptide units. The FTIR data together with the H NMR Fig. 5 shows that on both hydrogel groups, the metabolic
ꢁ1
s
1
data thereby confirmed that the designated poly(ether ester activity of cells increased from day 4 to day 14. The RGDS
amide) copolymer was successfully synthesized.
peptide increased significantly the cell activity on PEG-co-PDP/
The effect of the new depsipeptide units and the subsequent RGDS hydrogels compared to neat PEG-co-PDP hydrogels.
methacrylation on the thermal properties of PEG was characterized
using the DSC analysis. The DSC curves revealed a clear melting
peak for all the oligomers and macromers. Table 1 shows that the
3
.3 Fabrication of hydrogels with adjusted light exposure
times
m
copolymerization of depsipeptides into PEG decreased T of the The cell-free PEG-co-PDP hydrogels were 3D fabricated using
oligomer by around 10 1C and its DH to the half of the initial value. the visible light projection SLA in a layer-by-layer manner at
This journal is ©The Royal Society of Chemistry 2015
J. Mater. Chem. B, 2015, 3, 8348--8358 | 8353

View Article Online
Journal of Materials Chemistry B
Paper
Fig. 5 Metabolic activity of HUVECs seeded on PEG-co-PDP and PEG-
s
co-PDP/RGDS hydrogels as measured using a colorimetric AlamarBlue
cell viability assay.
Table 2 Gel content and swelling degree of the hydrogels prepared by
SLA
Light exposure time/layer (s)
Gel content (%)
Swelling degree
100
120
160
85 ꢀ 2
86 ꢀ 1
88 ꢀ 1
21 ꢀ 1
18 ꢀ 1
16 ꢀ 1
Fig. 6 (A) In vitro mass loss and (B) swelling degree of PEG-co-PDP
hydrogels with various crosslinking times during the 24-day degradation
RT. To study the effect of crosslinking time on the swelling
capacity, mass loss, and mechanical properties of hydrogels, study in PBS (pH 7.4).
three different light exposure times, 100 s, 120 s, and 160 s per
layer, were used in SLA. Table 2 shows that the gel content of
the hydrogels increased from 85% to 88% with the increasing
9
1
.4 ꢀ 1.2 kPa, and 37.9 ꢀ 13.3 kPa for the 100 s, 120 s, and
60 s samples, respectively. Within 7 days in PBS (Fig. 7C), the
crosslinking time, while the swelling degree decreased from
21 to 16.
Young’s modulus decreased to 0.24 ꢀ 0.02 kPa, 0.96 ꢀ 0.50 kPa,
and 1.52 ꢀ 0.55 kPa, respectively. Rheometric analysis of the
hydrogels revealed that the bulk storage modulus of the sam-
ples increased with increasing crosslinking time. The storage
modulus was initially in the order of magnitude of 1–10 kPa
3.4 In vitro mass loss and swelling of the 3D hydrogels
The in vitro hydrolytic mass loss of photocrosslinked PEG-co-
PDP hydrogels was monitored by weighing the remaining dry
mass of cell-free hydrogels after predetermined times in PBS.
Fig. 6A shows that the total mass loss of hydrogels decreased
with increasing crosslinking time. During 24 days in PBS, the
(
Fig. 7B), while it decreased to 0.1–1 kPa after 7 days in
PBS (Fig. 7D). The one order magnitude decrease within the
-day hydrolysis period was consistent with both the AFM and
rheometric studies.
7
100 s samples lost 66% of their initial dry mass, while the 120 s
samples lost 45% and the 160 s samples lost 35% of their initial
mass. Fig. 6B shows that the swelling degree of hydrogels
increased during the 24-day degradation study from 21 to 58
for the 100 s samples, from 18 to 45 for the 120 s samples, and
from 16 to 40 for the 160 s samples.
3
.6 Proliferation of encapsulated HUVECs in cell-laden
hydrogels and vessel graft fabrication
Cell-laden hydrogels were fabricated by photocrosslinking the
mixture of a filtered PEG-co-PDP/RGDS hydrogel solution and
HUVEC suspension using a visible light projection SLA. The
effect of crosslinking time on cell proliferation was tested by
3.5 Mechanical and viscoelastic properties of 3D hydrogels
To evaluate the effect of light exposure time and polymer fabricating cell-laden hydrogels using the same light exposure
degradation on the local mechanical stiffness of 3D fabricated times as for the cell-free hydrogels and by monitoring the
hydrogels, AFM was used to measure the Young’s modulus of metabolic activity of encapsulated cells using a colorimetric
s
wet samples right after SLA fabrication and after 7 days in PBS. AlamarBlue
assay. Fig. 8 shows that cells significantly
Fig. 7A shows that the stiffness increased with crosslinking increased their metabolic activity in all sample groups from
time, and the initial Young’s modulus was 3.0 ꢀ 1.0 kPa, day 1 to day 10. At each designated time point, there
8
354 | J. Mater. Chem. B, 2015, 3, 8348--8358
This journal is ©The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry B
Fig. 7 Local Young’s modulus measured using AFM and bulk storage and loss modulus of PEG-co-PDP hydrogels measured using a rheometric analysis
A and B) right after SLA fabrication and (C and D) after 7 days in PBS.
(
structure as required for functional vascular tissue engineering
grafts.
4. Discussion
Our goal in this work was to develop a new biodegradable
photocrosslinkable polymer to expand the limited repertoire of
polymers suitable for the SLA-based 3D fabrication of cell-laden
hydrogels. Subsequently, we used this polymer to increase the
understanding of how light exposure time in SLA affects the key
characteristics of the resulting hydrogel. As PEG is a non-degradable
polymer, it is not ideal for use in cell-laden hydrogels where cell
growth requires increasing volume of void space inside the polymer
9
,10
matrix.
To address the lack of degradation, we synthesized
Fig. 8 Metabolic activity of encapsulated cells in PEG-co-PDP/RGDS a biodegradable PEG-co-PDP oligomer by copolymerizing the
hydrogels within 10 days of cell culturing as measured using a colorimetric
L-alanine-derived depsipeptide with a four-arm PEG polymer.
The resulting oligomer was successfully functionalized with
s
AlamarBlue cell viability assay.
methacrylate end groups to obtain a photocrosslinkable macro-
was no significant difference in cell activity among the sample mer suitable for SLA.
1
groups.
H NMR and FTIR analysis confirmed that the depsipeptide
To demonstrate the capacity of the new PEG-co-PDP macro- units introduced biodegradable ester and peptide bonds to the PEG
mer in the SLA-based fabrication of tubular hydrogel constructs chains and the methacrylate groups were successfully attached to
for vascular tissue engineering, ring-like cell-laden hydrogels the end of the copolymer chains. Thermal characterization revealed
and bifurcating vascular tubes were designed and 3D fabricated that the copolymerization of PEG with the depsipeptide
using SLA. The fluorescence images in Fig. 9A and B show that decreased its melting temperature and enthalpy, indicating that
the cells were homogenously distributed in the hydrogel ring the depsipeptide units interfered with the highly crystalline
after 1 day of cell culturing. The uniform and round shape of structure of PEG. For SLA-based hydrogel fabrication, a visible
the hydrogel ring is shown in Fig. 9C. Fig. 9D shows that the 3D light sensitive photoinitiator, LAP, was synthesized, and a low
fabricated bifurcating hydrogel constructs closely resembled concentration of LAP (0.25% w/v) was mixed with the aqueous
their digital models and had a uniform and smooth wall macromer solution to generate a readily photocrosslinkable
This journal is ©The Royal Society of Chemistry 2015
J. Mater. Chem. B, 2015, 3, 8348--8358 | 8355

View Article Online
Journal of Materials Chemistry B
Paper
Fig. 9 (A and B) Fluorescence images and (C) a photograph of a cell-laden hydrogel construct; (D) CAD models of bifurcating vascular tubes and the
resulting hydrogel vessels visualized by their cross-section and top view.
hydrogel solution. 3D modeled hydrogels were successfully with increasing crosslinking time. As more double bonds reacted
photocrosslinked in a layer-by-layer manner leading to homo- within the continued light exposure time, the crosslinking density
genously crosslinked hydrogel constructs. The cytocompatibility and gel content of the hydrogels increased, resulting in decreased
of the PEG-co-PDP copolymer was evaluated by seeding HUVECs mesh size and more restricted water penetration into the hydrogel
on the 3D fabricated hydrogels. Cells proliferated from day 4 to network. A similar decrease in the swelling capacity has been
day 14 on both PEG-co-PDP and PEG-co-PDP/RGDS hydrogels. previously reported for PEG hydrogels, where the crosslinking
27
Immobilizing RGDS adhesion peptides to the hydrogels density was increased by increasing polymer concentration or
9
enhanced cell proliferation by providing them with integrin- by decreasing the molecular weight of the polymer.
specific attachment sites. Based on the encouraging prolifera-
Besides measuring the gel content and initial swelling
tion results, we decided to combine the PEG-co-PDP macromer capacity, we evaluated the hydrogel mass loss and changes in
with 2 mM acryl-PEG-RGDS in further 3D fabrication of cell- swelling degree and mechanical stiffness as a measure of
laden hydrogels.
polymer degradation. The mass loss of PEG-co-PDP hydrogels
The microenvironment of cells encapsulated in 3D hydro- followed a zero order curve trend within the 24-day incubation
gels comprises biophysical and biomechanical cues, including in PBS. Such a similar linear mass loss has been previously
swelling capacity, material mass loss, and mechanical stiffness reported for photocrosslinked PEG-co-polylactide (PEG-co-PLA)
28
of the hydrogel. We aimed at studying how these properties hydrogels. The total mass loss of our samples within 24 days in
depend on the crosslinking density of hydrogels, which is PBS ranged from 66% for 100 s samples to 35% for 160 s samples.
correlated with the light exposure time used for photocrosslinking When all the crosslinks of the multi-arm copolymer had broken,
the hydrogel. When the light exposure time in SLA was increased the copolymer unit was free to diffuse out of the hydrogel network,
from 100 s to 160 s per layer, the gel content of hydrogels increased leading to the hydrogel mass loss. The difference in mass loss rates
from 85% to 88%. The swelling degree decreased from 21 to 16 of our SLA fabricated hydrogels can be explained by the difference in
8
356 | J. Mater. Chem. B, 2015, 3, 8348--8358
This journal is ©The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry B
their crosslinking densities. Longer light exposure time hydrogels. Cell activity studies revealed that the initially non-
increased the crosslinking degree, when more polymer arms porous PEG-co-PDP/RGDS hydrogels supported the prolifera-
became covalently bound to the hydrogel network, making the tion of encapsulated HUVECs within the 10-day cell culturing
release of a polymer chain more difficult and slowing down the period regardless of crosslinking time. Between days 7 and 10,
mass loss. The increased crosslinking degree significantly cell metabolic activity slightly decreased in the hydrogel group
decreased the initial swelling degree from 21 to 16, and conse- with the longest crosslinking time. This most likely resulted
quently, the increased water uptake of the hydrogels with shorter from the high crosslinking degree and slow mass loss of the
crosslinking time accelerated their hydrolytic mass loss. The hydrogels so that the dense polymer network restricted cell
degradation of polymer networks was also evidenced as an proliferation within the hydrogels. However, no statistically
increase in the swelling degree of hydrogels during incubation significant difference was observed between days 7 and 10 in
in PBS. The increase in swelling degree was slow in the early any group. A previous study elsewhere showed that non-
stage of degradation, but accelerated in the later stage, indicating degradable, non-porous PEG hydrogels prepared by SLA had
that the decrease in crosslinking density occurred in a non-linear no cell proliferation within 7-day cell culturing, while the
fashion. This result was consistent with a similar exponential addition of small channels to the hydrogel improved cell
1
0
increase in swelling degree of previous biodegradable PEG-co-PLA proliferation. In our study, we attributed successful cell
28
hydrogels.
proliferation to polymer degradation and subsequent mass loss
In addition to the swelling capacity and mass loss, the and decreased mechanical stiffness of the hydrogel matrix. This
control over the mechanical properties of cell-laden hydrogels is consistent with previous studies showing that the high
1
9
is important for cell viability and function. We used oscillatory stiffness of PEG hydrogels tends to decrease the proliferation
rheometry to determine the viscoelastic properties of bulk of encapsulated cells due to the physical barrier caused by its
9
,15
hydrogels and AFM to study the local mechanical stiffness of non-degradable polymer network.
By using the new bio-
the hydrogels both right after SLA fabrication and after 7-day degradable polymer, we achieved hydrogels that degrade
incubation in PBS. The storage modulus describing the elastic simultaneously with cell growth and can continuously provide
properties of hydrogels increased with light exposure time and the cells with more space and a softer environment to grow
thus with the increased crosslinking density, being initially in inside the hydrogel matrix.
the order of magnitude of 1–10 kPa. The Young’s modulus of
Our long-term goal is to engineer biodegradable cell-laden
hydrogels ranged initially from 3 kPa to 38 kPa, being in the hydrogel-based vasculature using visible light SLA. In our
stiffness range of natural soft tissues regardless of the cross- current work, we successfully demonstrated the use of our
2
9
linking time. Elsewhere, hydrogels composed of 10% w/v of new PEG-co-PDP macromer for the SLA-based fabrication of
PEG diacrylates with comparable chain lengths have shown bulk 3D modeled vascular grafts. Though HUVECs proliferated
3
0
20
stiffnesses of 50 kPa and 170 kPa, indicating that our within the hydrogels, we understand that the initial cell den-
hydrogels were in the lower range of mechanical stiffness of sity, cell culture period, and biomimetic incubation conditions
corresponding PEG hydrogels. Within 7 days in PBS, the storage such as pulsatile perfusion may need to be optimized to achieve
modulus of the gels decreased to the range of 0.1–1 kPa and the endothelium and cell network formation within the hydrogel.
Young’s modulus decreased to the range of 0.2 kPa to 1.5 kPa. The polymer degradation rate can be further tuned to fit in with
Thus, the one order magnitude decrease was consistent both in the rate of cell growth and appropriate mechanical change over
the bulk storage modulus and the local stiffness of hydrogels. As time by changing the amount of depsipeptide monomer in the
the mechanical elasticity and the strength of photocrosslinked biodegradable PEG-co-PDP copolymer. In the future, more
hydrogels are caused by retractive forces of the crosslinked studies are needed to evaluate in vivo biodegradation and
2
8
polymer chains, the cleavage of biodegradable PEG-co-PDP biocompatibility of the biodegradable hydrogels developed in
chains decreased the total forces, together with the increased this study.
swelling of the gels, resulting in decreased mechanical stiffness
of degraded hydrogels. The hydrolysis of degradable bonds in
PEG-co-PDP chains resulted in broken crosslinks, first weaken-
ing the stiffness of the hydrogel and subsequently leading to
5
. Conclusions
mass loss after every crosslink of the multi-arm macromer had We successfully synthesized a biodegradable, photocrosslink-
broken. Even though the mechanical stiffness of the hydrogels able PEG-co-PDP macromer and demonstrated its use in the
decreased by one order magnitude within 7-day incubation in SLA-based fabrication of cell-laden hydrogels for vascular appli-
PBS, the samples retained their structural integrity.
cations. The depsipeptide units introduced biodegradable
3D fabrication of cell-laden hydrogels with biomimetic bonds to the PEG backbone, and by adjusting the light expo-
complexity and enhanced functionality is particularly promis- sure time in the SLA, we were able to tune the swelling capacity,
ing in tissue engineering and drug screening; it provides more degradation rate, and mechanical stiffness of the resulting
physiologically relevant data on cell behavior and response to hydrogels without the need for changing the parameters of
stimuli than can be obtained by seeding cells on a surface of hydrogel solution. The good printability and cell encapsulation
prefabricated hydrogels. We fabricated cell-laden hydrogels by capacity of our new biodegradable photocrosslinkable polymer
SLA using the same crosslinking parameters as for the cell-free make it an attractive extension to the extremely limited
This journal is ©The Royal Society of Chemistry 2015
J. Mater. Chem. B, 2015, 3, 8348--8358 | 8357

View Article Online
Journal of Materials Chemistry B
Paper
repertoire of biodegradable polymers currently available for the 11 K. Arcaute, L. Ochoa, F. Medina, C. Elkins, B. Mann and
SLA-based fabrication of cell-laden tissue engineering grafts.
R. Wicker, Mater. Res. Soc. Symp. Proc., 2005, 874,
91–198.
1
12 L. Lin, J. Zhu, K. Kottke-Marchant and R. E. Marchant,
Tissue Eng., Part A, 2014, 20, 864–873.
Acknowledgements
13 H. Liao, D. Munoz-Pinto, X. Qu, Y. Hou, M. A. Grunlan and
M. S. Hahn, Acta Biomater., 2008, 4, 1161–1171.
Graduate School in Chemical Engineering (LE), American-
Scandinavian Foundation (LE), Research Foundation of Helsinki 14 W. R. Gombotz, W. Guanghui, T. A. Horbett and A. S. Hoffman,
University of Technology (LE), Emil Aaltonen Foundation (LE),
J. Biomed. Mater. Res., 1991, 25, 1547–1562.
Alfred Kordelin Foundation (LE), NIH R01AR057837 (NIAMS, 15 K. Bott, Z. Upton, K. Schrobback, M. Ehrbar, J. A. Hubbell,
YY), NIH R01DE021468 (NIDCR, YY), DOD W911NF-14-1-0545
YY), DOD W81XWH-10-1-0966 (PRORP, YY), Coulter Transla-
M. P. Lutolf and S. C. Rizzi, Biomaterials, 2010, 31,
8454–8464.
(
tional Research Award (YY), and Stanford Nano/Biomedical 16 S. J. Bryant, R. J. Bender, K. L. Durand and K. S. Anseth,
Seed Grant (YY) are gratefully acknowledged for the funding
Biotechnol. Bioeng., 2004, 86, 747–755.
of this work. Alexander Stahl (Stanford University) is acknowl- 17 I. Mironi-Harpaz, D. Y. Wang, S. Venkatraman and
1
edged for running the H NMR of acryl-PEG samples. Stanford
D. Seliktar, Acta Biomater., 2012, 8, 1838–1848.
Nano Shared Facilities were used for part of the experiments in 18 G. D. Nicodemus and S. J. Bryant, Tissue Eng., Part B, 2008,
this work.
14, 149–165.
19 D. Dikovsky, H. Bianco-Peled and D. Seliktar, Biophys. J.,
2008, 94, 2914–2925.
2
0 S. R. Peyton, C. B. Raub, V. P. Keschrumrus and A. J. Putnam,
Biomaterials, 2006, 27, 4881–4893.
References
1
2
3
F. P. W. Melchels, J. Feijen and D. W. Grijpma, Biomaterials, 21 L. Elomaa, Y. Kang, J. V. Sepp ¨a l ¨a and Y. Yang, J. Polym. Sci.,
010, 31, 6121–6130. Part A: Polym. Chem., 2014, 52, 3307–3315.
L. Elomaa, A. Kokkari, T. N ¨a rhi and J. V. Sepp ¨a l ¨a , Compos. 22 Y. Ohya, H. Yamamoto, K. Nagahama and T. Ouchi,
Sci. Technol., 2013, 74, 99–106. J. Polym. Sci., Part A: Polym. Chem., 2009, 47, 3892–3903.
T. M. Seck, F. P. W. Melchels, J. Feijen and D. W. Grijpma, 23 Y. Zhao, J. Li, H. Yu, G. Wang and W. Liu, Int. J. Pharm.,
2
J. Controlled Release, 2010, 148, 34–41.
2012, 430, 282–291.
4
5
H. Ikehata and T. Ono, J. Radiat. Res., 2011, 52, 115–125.
S. P. Grogan, P. H. Chung, P. Soman, P. Chen, M. K. Lotz,
24 B. D. Fairbanks, M. P. Schwartz, C. N. Bowman and
K. S. Anseth, Biomaterials, 2009, 30, 6702–6707.
S. Chen and D. D. D’Lima, Acta Biomater., 2013, 9, 7218–7226. 25 A. Lendlein, A. M. Schmidt, M. Schroeter and R. Langer,
R. Gauvin, Y. C. Chen, J. W. Lee, P. Soman, P. Zorlutuna, J. Polym. Sci., Part A: Polym. Chem., 2005, 43, 1369–1381.
J. W. Nichol, H. Bae, S. Chen and A. Khademhosseini, 26 H. Lin, T. Kai, B. D. Freeman, S. Kalakkunnath and
Biomaterials, 2012, 33, 3824–3834. D. S. Kalika, Macromolecules, 2005, 38, 8381–8393.
K. Arcaute, B. K. Mann and R. B. Wicker, Ann. Biomed. Eng., 27 F. M. Andreopoulos, E. J. Beckman and A. J. Russell,
006, 34, 1429–1441.
Biomaterials, 1998, 19, 1343–1352.
B. Dhariwala, E. Hunt and T. Boland, Tissue Eng., 2004, 10, 28 A. T. Metters, K. S. Anseth and C. N. Bowman, Polymer, 2000,
316–1322. 41, 3993–4004.
V. Chan, P. Zorlutuna, J. H. Jeong, H. Kong and R. Bashir, 29 I. Levental, P. C. Georges and P. A. Janmey, Soft Matter,
Lab Chip, 2010, 10, 2062–2070. 2007, 3, 299–306.
0 H. Lin, D. Zhang, P. G. Alexander, G. Yang, J. Tan, 30 S. Lee, X. Tong and F. Yang, Acta Biomater., 2014, 10,
A. W. Cheng and R. S. Tuan, Biomaterials, 2013, 34, 331–339. 4167–4174.
6
7
8
9
2
1
1
8
358 | J. Mater. Chem. B, 2015, 3, 8348--8358
This journal is ©The Royal Society of Chemistry 2015