Cytocompatibility testing of hydrogels toward bioprinting of

mesenchymal stem cells

Leo Benning,1 Ludwig Gutzweiler,2,3 Kevin Tro€ ndle,2 Julian Riba,2 Roland Zengerle,2,3,4
2 2
Peter Koltay, Stefan Zimmermann, G. Bjo € rn Stark,1 Gu€ nter Finkenzeller1
1
Department of Plastic and Hand Surgery, Faculty of Medicine, Medical Center–University of Freiburg, Freiburg, Germany
2
Laboratory for MEMS Applications, IMTEK-Department of Microsystems Engineering, University of Freiburg, Georges-
Koehler-Allee 103, Freiburg 79110, Germany
3
Hahn-Schickard, Georges-Koehler-Allee 103, Freiburg 79110, Germany
4
FIT-Freiburg Centre for Interactive Materials and Bioinspired Technologies, University of Freiburg, Georges-Koehler-Allee
105, Freiburg 79110, Germany

Received 28 February 2017; revised 7 July 2017; accepted 28 July 2017

Published online 22 August 2017 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jbm.a.36179

Abstract: Mesenchymal stem cells (MSCs) represent a very
attractive cell source for tissue engineering applications aim-
ing at the generation of artificial bone substitutes. The use of
three-dimensional bioprinting technologies has the potential
to improve the classical tissue engineering approach because
bioprinting will allow the generation of hydrogel scaffolds with
high spatial control of MSC allocation within the bioprinted
construct. In this study, we have performed direct comparisons
between commercially available hydrogels in the context of
their cytocompatibility toward MSCs and their physicochemi-
cal parameters with the aim to identify the most suitable
but the latter showed a limited potential to promote MSC pro-
liferation. We concentrated our study on fibrin and collagen
hydrogels and investigated the effect of hydroxyapatite (HA)
inclusion. The inclusion of HA enhanced proliferation and
osteogenic differentiation of MSCs and prevented degradation
of fibrin in vitro. According to viscosity and storage moduli
measurements, HA-blends displayed physicochemical charac-
teristics suitable for DoD printing. In bioprinting experiments,
we confirmed that fibrin and collagen and their respective HA-
blends represent excellent hydrogels for DoD-based printing as
evidenced by high survival rates of printed MSCs. V C 2017 Wiley

hydrogel for drop-on-demand (DoD) printing of MSCs. In this Periodicals, Inc. J Biomed Mater Res Part A: 105A: 3231–3241, 2017.
context, we examined matrigel, fibrin, collagen, gelatin, and
gelatin/alginate at various hydrogel concentrations. Matrigel, Key Words: 3D bioprinting, hydrogel, mesenchymal stem
fibrin, collagen, and gelatin were able to support cell viability, cell, osteogenesis, tissue engineering

How to cite this article: Benning L, Gutzweiler L, Tro€ ndle K, Riba J, Zengerle R, Koltay P, Zimmermann S, Bjo€ rn Stark G,
Finkenzeller G. 2017. Cytocompatibility testing of hydrogels toward bioprinting of mesenchymal stem cells. J Biomed Mater
Res Part A 2017:105A:3231–3241.

INTRODUCTION MSCs are able to support bone regeneration upon implanta-

The treatment of bone defects still remains a great clinical
challenge. In this context, autologous bone graft procedures
are still considered as the gold standard.1 Due to size limita-
tions and donor-site morbidity such as hemorrhage, infection
and chronic pain,2,3 tissue engineering approaches gain on
interest in the recent years. Bone tissue engineering is based
on the use of autologous cells which are ex vivo seeded into
osteoconductive scaffolds and implanted into bone defects. In
this context, mesenchymal stem cells (MSCs) represent a very
attractive cell source, because they have self-renewal poten-
tial and can differentiate in vivo, as well as ex vivo, into several
mesenchymal cell types such as chondrocytes, fat cells and
osteoblasts.4,5 Moreover, these cells also demonstrate a high
ex vivo proliferation capacity and isolation of these cells is not
associated with severe donor-site morbidity.6 It has been
shown in various bone healing models that ex vivo expanded
tion.7–9 The classical approach for tissue engineering of bone
substitutes is based on the use of MSCs seeded into osteocon-
ductive hydrogel scaffolds. These cell-seeded matrices are
first cultivated in vitro for different periods of time, optional
in osteogenic differentiation medium (containing ascorbic
acid, dexamethasone, and b-glycerolphosphate) to promote
osteogenic differentiation of MSCs and then implanted into
bone defects.
However, by using this approach, MSCs are randomly
distributed within the matrices without control of the spa-
tial cell distribution and therefore without the opportunity
to control prospective bone formation within the biomatrix.
Recently, three-dimensional (3D) bioprinting technologies
have been applied enabling the generation of hydrogel scaf-
folds with defined architecture and high spatial control of
cell allocation within the 3D construct during the printing

Correspondence to: G. Finkenzeller; e-mail: guenter.finkenzeller@uniklinik-freiburg.de

C 2017 WILEY PERIODICALS, INC.

V 3231
process.10,11 Due to the possibility of tuning the spatial dis-
tribution of MSCs during ex vivo printing, optimization of
bone formation upon implantation should be achieved. Bio-
printing of cells can be realized by different printing meth-
ods like inkjet printing (or more generally drop-on-demand
[DoD] methods), microextrusion or laser assisted bioprint-
ing.12 However, in all 3D bioprinting processes, hydrogels
are used for cell printing. These hydrogels have to be com-
patible with the encapsulated cells, the printing process and
must also provide the desired mechanical stability of the
printed 3D construct. Hydrogels currently used in the field
of 3D bioprinting are often based on naturally derived poly-
mers such as matrigel, fibrin, collagen, gelatin, or alginate.13
However, the optimal hydrogel has to be determined for
each cell type and each printing technology. Concerning bio-
printing of MSCs, the capability of certain hydrogels to sup-
port printability, viability, proliferation, and osteogenic
differentiation of MSCs has not been studied systematically
so far. Therefore, in this study, we performed side by side
comparisons of the above mentioned hydrogels in terms
of their cytocompatibility toward MSCs. Those hydrogels
that supported viability and proliferation of MSCs were fur-
ther characterized for their potential to support osteogenic
differentiation of MSCs. By this selection process, we identi-
fied fibrin and collagen as suitable hydrogels for bioprinting
of MSCs.
It is well known that hydroxyapatite (HA), a natural
component of bone tissue, shows many positive effects on
osteoblasts and MSCs, supports bone regeneration and
thereby represents an excellent scaffold material for bone
tissue engineering applications.14 In this context, it was
shown in the literature that HA supports proliferation of
osteoblasts and MSCs, either by its own or in combination
with other osteoinductive materials.15–18 Moreover, it is also
well established that HA supports osteogenic differentiation
of MSCs in vitro, as well as in vivo.15,19–21 For these reasons,
we investigated the effects of HA inclusion into collagen and
fibrin on the above mentioned biological parameters and on
the physicochemical characteristics of the hydrogels. The
goal of this study was to identify a hydrogel or a HA-tuned
hydrogel with physicochemical characteristics suitable for
DoD printing, with good long-term stability, with MSC cyto-
compatibility and with the ability to support osteogenic dif-
ferentiation of MSCs.
MATERIALS AND METHODS
Preparation of hydrogels
Matrigel. Growth factor reduced matrigel (354230, Corning)
was kept on ice during hydrogel preparation. Our experi-
mental setups include samples with 100, 66, and 33%
matrigel. Dilutions were prepared with phosphate buffered
saline (PBS) (L182-10, Biochrom).
Fibrin. Fibrinogen (341576, Calbiochem) and thrombin
(605190, Calbiochem) are the well-known components of in
vivo coagulation processes. Serin protease thrombin allows
proteolytic processing of fibrinogen and the subsequent for-
mation of an enzymatically cross-linked protein complex,
3232 BENNING ET AL.
the fibrin. Due to the temperature-dependent kinetics of
thrombin, all preparations were kept on ice. Since the fibrin
structure is mainly determined by the thrombin concentra-
tion, we prepared all samples with 10 mg/mL fibrinogen
and evaluated the effects of 10, 5, and 2 U/mL thrombin. In
some of the hydrogels, fibrinogen was tuned with 1% HA
(677418, Sigma) before mixing with thrombin. Fibrinogen
and thrombin solutions were mixed 1:1. Dilutions from
stock solution were prepared with PBS.
Collagen. Collagen-I from rat tail (345236, Corning) was
kept acidic and could be handled in a highly viscous state.
Samples were prepared to final concentrations of 1, 2, and
3 mg/mL according to the manufacturer’s instructions. In
some of the hydrogels, collagen was tuned with 1% HA.
Gelation was induced by neutralization with equimolar 1M
NaOH.
Gelatin. Gelatin Type-A (G2500-100G, Sigma), withdrawn by
acidic drying out from porcine skin was dissolved in PBS.
Due to gelatin’s gel-sol transition at about 378C, we
required a covalent stabilization procedure to obtain a
hydrogel that remained in its semi-solid state under the
incubation circumstances. Therefore, we kept each gelatin
sample at room temperature and induced a covalent cross-
linking by adding 0.3% glutaraldehyde (G6257, Sigma).
After 12 h, all samples were washed thoroughly with PBS
and possible aldehyde residues were inactivated through
the addition of 50 mM glycin (357002, Calbiochem). After
another washing step with PBS, samples were ready to be
seeded with cells. Samples were prepared with final concen-
trations of 5, 10, and 15% gelatin. Dilutions were prepared
with PBS.
Alginate/gelatin. Alginic acid sodium salt (180947, Sigma)
was dissolved in PBS and mixed with gelatin. Preparations
were kept on a rocking table at 378C. Cross-linking of algi-
nate requires bivalent cations (for example, calcium), which
were added as 0.2M CaCl2 (21114, Fluka). After 5 min,
remaining CaCl2 was taken out of each well and each sam-
ple was rinsed thoroughly with PBS. Samples were prepared
with 5% gelatin and 1% alginate, 10% gelatin and 1% algi-
nate and 15% gelatin and 1% alginate. The gelatin compo-
nents in these samples were not covalently cross-linked by
glutaraldehyde treatment.
Determination of viscosity
Viscosity measurements of hydrogel components were con-
ducted at room temperature using a MCR 52 rotational rhe-
ometer from Anton Paar GmbH, Austria. Most biopolymer
solutions are non-Newtonian fluids exhibiting a non-linear
relation between shear rate and shear stress. Hence, shear
rates acting on the liquid may vary dependent on the used
printing system. As a result shear stress and thus viscosity
is dynamic for the same liquid printed with different sys-
tems. For that reason dynamic viscosity was measured
between 1 and 2000/s.
CYTOCOMPATIBILITY TESTING OF HYDROGELS

Determination of viscoelastic properties—Storage and
loss modulus
To make a statement about the mechanical stability of
hydrogels the dynamic shear modulus (G*), composed of
storage (G0 ) and loss modulus (G00 ) is measured using a
MCR 52 rotational rheometer from Anton Paar GmbH, Aus-
tria (plate diameter: 25 mm; constant torque: 2.5 mN/m).
Storage and loss moduli are measured at 1 Hz and 378C
within the plateau regime that was previously determined
by frequency sweeps between 1 and 10 Hz.
Swelling/degradation assays
Hydrogels were immersed in alpha minimum essential
medium (a-MEM), 1% penicillin/streptomycin, 10% fetal
calf serum (FCS) and incubated for 7 days to evaluate possi-
ble swelling or degradation. 600 mL of every hydrogel were
plated in tissue culture dishes (35–3001, Falcon). Blank
measurements were carried out for all dishes used. 600 mL
of growth medium was added on top of each hydrogel and
measurements were conducted on day 0, 1, 4, and 7. Prior
to each measurement, the entire supernatant was removed
from the dish to allow the precise measurement of the
hydrogel. After every weighting procedure, fresh media was
pipetted onto each gel. Weight alterations were monitored
over the full course of the experiment.
Printing of MSCs
A piezo-driven DoD dispenser (PipeJet P9, Biofluidix GmbH,
Germany) was used to dispense MSCs between hydrogel
layers to investigate cell viability subsequent to printing.
The dispenser exhibits a protruded polymeric nozzle (inner
diameter: 500 mm) connected to a reservoir via silicone tub-
ing. MSC concentration was adapted to 2 3 106 cells/mL
and suspended in alpha MEM, 1% penicillin/streptomycin,
10% FCS, 10 mg/mL fibrinogen, with or without 1% HA
(fibrin hydrogel 6 HA) or in alpha MEM, 1% penicilin/strep-
tomycin, 10% FCS, 3 mg/mL collagen with or without 1%
HA (collagen hydrogel 6 HA). The PipeJet reservoir was
filled with respective MSC suspensions. Following, MSC sus-
pensions were printed (single droplet volume: approxi-
mately 50 nL) on the bottom of 24 well plates coated with
150 mL fibrin (10 mg/mL fibrinogen crosslinked with 10 U/
mL thrombin) in the presence or absence of 1% HA or 150
mL collagen (3 mg/mL) with or without 1% HA.
After printing, an additional layer of respective hydro-
gels was applied (150 mL) and filled with 500 mL alpha
MEM, 1% penicillin/streptomycin, 10% FCS. MSCs were
incubated for 24 h and analyzed for cell viability by using
the Live/Dead Viability Kit (L3224, Life Technologies).
Cell culture
Human MSCs were isolated and expanded as described
before.22 In brief, mononuclear cells (MNCs) were purified
by density gradient centrifugation with Biocoll Separating
Solution (Biochrom AG, Berlin, Germany) from human bone
marrow. MNCs were seeded in culture flasks at a density of
5 3 105 cells/cm2 in expansion medium [alpha-MEM, 10%
FCS, 50 mg/mL gentamicin, 5 ng/mL basic fibroblast growth
factor (bFGF)] at 378C, 5% CO2. The medium was changed
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | DEC 2017 VOL 105A, ISSUE 12
ORIGINAL ARTICLE
three times weekly, removing all nonadherent cells. Once
adherent cells had grown to confluence, they were
detached and reseeded at a density of approximately 6000
cells/cm2 and cultivated for two further passages. Flow
cytometric analysis were performed for the characterization
of surface protein expression patterns. Cell surface markers
CD105, CD90, CD73, CD45, and CD14 were analyzed using
phycoerythrin-labeled monoclonal antibodies (abcam, Cam-
bridge, UK) according to previously published protocols.23
MSCs expressed CD105, CD90 and CD73, while CD45 and
CD14 were not expressed. For osteogenic differentiation,
MSC-seeded hydrogels were incubated for 21 days in osteo-
genic differentiation medium (DMEM supplemented with
10% FCS, 50 mg/mL gentamicin, 10 mM b-
glycerophosphate, 0.1 mM dexamethasone and 50 mM ascor-
bic acid). Medium was changed every 3 days.
Viability assay
To determine the hydrogels’ cytocompatibility and the viabil-
ity of MSCs after the printing process, a Live/Dead Viability/
Cytotoxicity Kit (L3224, Life Technologies) was used in
accordance to the supplier’s instructions. MSCs were either
seeded manually on the hydrogels (1 3 104 cells per hydro-
gel) or dispensed between the respective hydrogel layers by
our PipeJet dispenser and incubated for a period of 24 h. Fol-
lowing the removal of the growth medium and a washing
step with PBS, the freshly prepared Live/Dead solution was
added to each sample. Samples are kept at room temperature
for 30 min before pictures were obtained from an Axiovert
100 microscope (Carl Zeiss Jena GmbH, Jena) after fluorescent
excitation of the probes. Fluorescent signals are based on the
intracellular processing of the cell-permeant calcein AM to
calcein, which emits fluorescence at 515 nm. Due to the
demand of an active processing, calcein AM is used to identify
live cells. Ethidium homodimer 1 (EthD1) emits fluorescence
at 635 nm when intercalating DNA. Since EthD1 can only per-
meate damaged cell membranes, it visualizes dead cells.
Cell proliferation assay
Cell proliferation on hydrogels was evaluated over the
course of 7 days by using the cell activity assay CellTiter96
Aqueous One Solution (G3580, Promega). The addition of
the [3-(4,5-dimethyl-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-
sulfophenyl)-2H-tetrazolium, inner salt (MTS) derivate
allows a colorimetric determination of the cells’ activity
seeded onto the hydrogels. Samples of each hydrogel were
prepared in triplicates for measurements on day 1, 4, and 7.
Per well, 215 mL hydrogel were given in 12-well plates. 1 3
104 MSCs were plated onto every hydrogel. Media were
changed on day 3 and 6. Prior to the addition of the react-
ing agent, all samples were washed thoroughly with PBS.
After 3 h of incubation at 378C, the evaluation was carried
out on a colorimetric Tecan SpectraFluor Plus reader (Tecan
Group AG, M€annedorf, CH). Increasing optical densities
reflect an increasing cell number.
Alkaline phosphatase assay
Alkaline phosphatase activity in cellular extracts was deter-
mined as described.24 In brief, cells on the hydrogels were
3233
FIGURE 1. Cytotoxicity testing of hydrogels. MSCs were plated on hydrogels and viability of cells was evaluated after 24 h by using a Live/Dead
assay. Representative images were taken using a fluorescence microscope. Viable cells show green fluorescence, dead cells show red fluores-
cence. (A) Viability of MSCs on pure hydrogels. (B) Viability of MSCs on HA-tuned fibrin and collagen.

3234 BENNING ET AL. CYTOCOMPATIBILITY TESTING OF HYDROGELS

 ORIGINAL ARTICLE

FIGURE 2. Effect of hydrogels on MSC proliferation. MSCs were seeded onto the hydrogels and proliferation was determined by using a
MTS assay at days 1, 4, and 7. (A) Proliferation on pure hydrogels. (B) Proliferation of MSCs on HA-tuned fibrin and collagen. Shown are
means 6 SD. n 5 3.

washed once with PBS and then lysed by the addition of
25 mM Tris-HCl, pH 8.5; 0,5% Triton X-100. Lysates were
transferred to Eppendorf tubes and subjected to three freeze-
thaw cycles. Samples (20 mL) were mixed with 100 mL of CSPD
substrate (Applied Biosystems, Forster City, CA) and incubated
at room temperature for 15 min. Thereafter, light output was
measured by a plate luminometer (Spectrafluor Plus, Tecan) in
relative light units (RLU). Cellular alkaline phosphatase (ALP)
activities were normalized to total protein content with the
bicinchoninic acid (BCA) assay kit (Pierce, Rockford).

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | DEC 2017 VOL 105A, ISSUE 12 3235
FIGURE 3. Effect of hydrogels on osteogenic differentiation of MSCs. MSCs were plated on fibrin (A and B) and collagen (C and D) in the
absence (-HA) or presence of HA (1HA). MSCs were incubated for 21 days in normal growth medium (norm-Med., A and C) or osteogenic differ-
entiation medium (diff-Med., B and D). Osteogenic differentiation was assessed by measuring alkaline phosphatase activity, which is indicated
by RLU. Shown are means 6 SD. n 5 3. Asterisks indicate statistically significant differences between the indicated groups (*p < 0.05; **p < 0.005;
***p < 0.0005).

Statistical analysis excellent results on MSC viability at all concentrations tested,

Statistically significant differences between groups were
determined by using an unpaired Student’s t test. Statistical
significance was defined when p < 0.05.
RESULTS
In this study, we intended to perform side by side compari-
sons between various popular bioprinting hydrogels in
terms of their cytocompatibility toward MSCs and their
rheological characteristics. In our experiments, hydrogels
were prepared and investigated in three different concentra-
tions, covering a concentration range corresponding to typi-
cal concentrations published by other groups in bioprinting
applications previously. Concerning fibrin, fibrinogen was
used at a fixed concentration of 10 mg/mL only, but instead
fibrin crosslinking was induced with three different throm-
bin concentrations of 2, 5 and 10 U/mL.
Viability of MSCs
First, we performed Live/Dead assays to investigate the
effects of the hydrogels on MSC viability. As shown in
Figure 1, matrigel, fibrin, collagen, and gelatin showed
indicated by the fact that no dead cells could be detected. In
contrast, dead MSCs indicated by red fluorescence, could be
detected in the alginate/gelatin group at alginate/gelatin
concentration ratios of 1/10% and 1/5%. Moreover, poor
cell attachment, indicated by lower cell numbers and the
rounded phenotype of the MSCs, could be detected in the
(1%/5%) alginate/gelatin group. Interestingly, MSCs attach-
ment and viability could be improved by increasing the gela-
tin concentration to 15% in the alginate/gelatin blend. Fibrin
and collagen formulations were also tuned with 1% HA [Fig.
1(B)]. In comparison to fibrin and collagen alone, we
observed that HA inclusion into these hydrogels did not
influence MSC viability or MSC attachment.
Proliferation of MSCs
Next, we measured MSC proliferation by MTS assays per-
formed on days 1, 4, and 7 after seeding the MSCs onto the
hydrogels (Fig. 2). No cell proliferation was detectable on
alginate/gelatin hydrogels. On pure gelatin hydrogels, MSC
proliferation was only detectable at the 5% gelatin concen-
tration, whereas at higher concentrations, MSCs fail to

3236 BENNING ET AL. CYTOCOMPATIBILITY TESTING OF HYDROGELS

 ORIGINAL ARTICLE

FIGURE 4. Swelling/degradation properties of HA-tuned fibrin and collagen. HA-tuned collagen at different concentrations or HA-tuned fibrin
generated with different thrombin concentrations were plated in tissue culture dishes and after gelation overlaid with growth medium. The
weight of the gels after removal of the growth medium was determined at day 0, 1, 4, and 7. Swelling is indicated by an increase in weight,
whereas degradation of the hydrogels is indicated by a decrease in weight. Shown are means 6 SD. n 5 3.

proliferate. In contrast, MSCs proliferated very well on
matrigel, fibrin, and collagen. Inclusion of HA into collagen
revealed that HA inclusion had only marginal positive
effects on MCS proliferation, whereas HA inclusion dramati-
cally improved MSC proliferation on fibrin [Fig. 2(B)].
Osteogenic differentiation of MSCs
Based on the results from the viability and proliferation
experiments, we focused our interest on fibrin and collagen
and the respective HA blends. In this context, we investi-
gated the osteogenic differentiation potential of MSCs, culti-
vated on fibrin and collagen hydrogels or the respective HA
blends. As shown in Figure 3, cultivation of MSCs in osteo-
genic differentiation medium for 3 weeks dramatically
increased alkaline phosphatase expression, which served as
a marker for osteogenesis, in comparison to MSCs cultivated
in normal growth medium. This holds true for MSCs culti-
vated on fibrin [Fig. 3(A,B)] as well as for MSCs cultivated
on collagen [Fig. 3(C,D)]. Interestingly, we also observed a
positive effect of HA inclusion in fibrin gels on osteogenic
differentiation irrespective of whether MSCs were cultivated
in normal growth medium or osteogenic differentiation
medium [Fig. 3(A,B)]. In fibrin gels, we also observed a
gradual increase in ALP expression at lower thrombin con-
centrations when MSCs were cultivated in osteogenic differ-
entiation medium [Fig. 3(B)]. In collagen hydrogels, a
positive effect of HA on osteogenic differentiation was only
seen in the 1 mg/mL collagen gel when MSCs were culti-
vated under osteogenic differentiation conditions [Fig. 3(D)].
Moreover, osteogenic differentiation showed a dependency
on collagen concentrations. ALP activity declines by increas-
ing the collagen concentrations when MSCs were cultivated
in the presence of HA in normal, as well as in osteogenic
differentiation medium [Fig. 3(C,D)] and also in the absence
of HA, when MSCs were cultivated under osteogenic differ-
entiation conditions [Fig. 3(D)].
Swelling/degradation behavior of hydrogels
Next, we explored the mechanical stability of the HA-tuned
hydrogels in an aqueous environment. For this purpose, we
performed swelling/degradation experiments in growth
medium, in which hydrogels were incubated for up to 7
days. To measure swelling (weight increase) or degradation
(weight decrease), hydrogels were weighed upon gelation
(day 0) and after 1, 4, and 7 days (Fig. 4). HA-blends of
fibrin and collagen showed remarkable stability over the 7
days time course at all tested hydrogel concentrations (Fig.
4). This means neither swelling, nor degradation of hydro-
gels was detectable in vitro in an aqueous environment.
Physicochemical characteristics of hydrogels
Next, we investigated the physicochemical parameters of the
hydrogels. Viscosity of HA-tuned collagen and the HA-tuned
fibrin component fibrinogen can be seen in Figure 5.

FIGURE 5. Viscosity measurements of hydrogels at room temperature in relation to various shear rates. Collagen (1, 2 and 3 mg/mL) and fibrino-
gen (10 mg/mL) tuned with (1% HA) show pseudoplastic behavior. Shown are mean values.

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | DEC 2017 VOL 105A, ISSUE 12 3237

FIGURE 6. Viscoelastic properties of collagen (1, 2 and 3 mg/mL) and
fibrin prepared by crosslinking fibrinogen (10 mg/mL) with 2, 5 and 10
U/mL thrombin. Hydrogels were tuned with 1% HA. The dynamic
shear modulus (G*) of hydrogels, which is composed of storage (G0 )
and loss modulus (G00 ) was determined at 1 Hz. Measurements were
conducted at 378C. Shown are means 6 SD. n 5 3. Asterisks indicate
statistically significant differences between the indicated groups
(*p < 0.05; ***p < 0.0005).
HA-tuned fibrinogen and collagen showed pseudoplastic
behaviors typical for biopolymer solutions. This means the
viscosity decreases with increasing shear rates. Collagen
showed a higher viscosity than fibrinogen. However, viscos-
ities ranging from 2 to 15 mPa s (for shear rates between 1
and 2000/s) are within the range covered by most DoD dis-
pensers and even inkjet printers.25
Viscoelastic properties were measured to determine the
hydrogels ability for 3D printing. Storage and loss moduli for
the HA-tuned hydrogels are given in Figure 6. For collagen,
the storage moduli increase for increasing concentrations.
Concerning fibrin, we observed an inverse relationship
between thrombin concentration, used to crosslink fibrino-
gen, and the storage modulus. This means, the lower the
thrombin concentration, the higher the storage modulus of
the hydrogel. We also noticed that HA-tuned fibrin displayed
significantly higher storage moduli than HA-tuned collagen at
all tested concentrations, suggesting that fibrin is stiffer than
collagen and therefore 3D bioprinted fibrin scaffolds may be
more stable than collagen based constructs.
Bioprinting of MSCs
We investigated whether fibrin and collagen and their
respective HA-modified blends can be used as bioinks
printed with our DoD-based dispenser setup and whether
MSCs are intact and viable after the printing process. There-
fore, MSCs were resuspended in respective hydrogels and
printed onto basal layers of fibrin and collagen (both with
or without 1% HA), using the PipeJet P9 dispenser as
before. After printing, MSCs were incubated for 24 h and
then analyzed for viability by using a Live/Dead assay (Fig.
7). As can be seen in Figure 7(A), the majority of MSCs
remained intact after printing, as indicated by green fluores-
cence. Quantification of viability revealed survival rates of
3238 BENNING ET AL.
82 6 8% in fibrin, 89 6 8% in HA-tuned fibrin, 81 6 17% in
collagen and 89 6 12% in HA-tuned collagen [Fig. 7(B)].
DISCUSSION
In this study, we evaluated five hydrogels or hydrogel blends
in various final concentrations in terms of their ability to sup-
port MSC viability and proliferation for prospective inkjet-
based bioprinting applications. All of the tested hydrogels
have already been used for bioprinting of various cell
types,26–33 however, a direct side by side comparison of the
hydrogels in terms of their cytocompatibility toward human
MSCs has not been published so far. In prospective studies,
we intend to use bioprinted MSC-containing hydrogels for
implantation studies in critical sized bone defects for the
development of artificial bone substitutes. According to the
Live/Dead assays, MSCs showed high cell viability on matri-
gel, fibrin, collagen, and gelatin at all tested hydrogel concen-
trations. However, attachment and viability on alginate/
gelatin blends were strongly impaired at low gelatin concen-
tration of 5%, but could be improved by increasing the
amount of gelatin. Alginate/gelatin blends with high propor-
tions of gelatin have already been used for bioprinting of
MSCs and showed acceptable results on viability.34 The poor
attachment and viability of cells to alginate/gelatin blends
with low gelatin concentrations can be explained by the fact
that alginate does not contain arginine/glycin/aspartate
(RGD) motifs and therefore development of focal adhesion
contacts between alginate and MSCs is strongly impaired.35–37
As this study focuses on basic and non-chemical modifications
of the hydrogels, increasing the gelatin component in the
blends can partly compensate this disadvantage of alginate at
the expense of the hydrogel’s stability.
Proliferation of MSCs on hydrogels was monitored by a
MTS assay over a time period of up to 7 days. Matrigel,
fibrin, and collagen hydrogels were able to support prolifer-
ation of MSCs at all tested concentrations, but this was not
the case for the alginate/gelatin blends, presumably because
of the impaired cell attachment to alginate. On pure gelatin
hydrogels, MSCs proliferation was only detectable at the
lowest concentration. At higher concentrations, MSC prolif-
eration was not supported, probably because residues of
the used cross-linker glutaraldehyde, which is known to act
negative on cell proliferation,38,39 could not be completely
withdrawn by our washing procedure.
Based on the viability and proliferation experiments, we
decided to focus our interest on fibrin and collagen in our
further studies. Matrigel was not followed up because in
terms of a putative prospective clinical application, one has
to consider that matrigel is not clinically approved, since it
is produced from murine Engelbreth-Holm-Swarm sarcoma.
Moreover, cooling of matrigel is necessary for DoD printing.
Therefore, in a future bioprinting application, the respective
dispensers must be cooled below 108C making the process
more difficult. Gelatin was not followed up because of the
negative effects of the cross-linking procedure on cell prolif-
eration at higher hydrogel concentrations.
CYTOCOMPATIBILITY TESTING OF HYDROGELS
 ORIGINAL ARTICLE

FIGURE 7. Viability of MSCs after printing with a DoD-based bioprinter. MSCs were suspended in fibrinogen solution (with or without 1% HA)
and printed onto a fibrin layer (with or without 1% HA) or suspended in collagen solution (with or without 1% HA) and printed onto a collagen
layer (with or without 1% HA). After printing and gelation of the MSC containing hydrogels, the arrays were covered by an additional layer of
the respective hydrogels and incubated in alpha MEM, 1% pen/strep, 10% FCS at 378C. Viability of cells was evaluated after 24 h by using a Live/
Dead assay. (A) Representative images, which were taken by using a fluorescence microscope, are shown. Viable cells show green fluorescence,
dead cells show red fluorescence in their nuclei. Dead cells are exemplary indicated by white arrows. (B) Quantification of cell viability by deter-
mination of the ratio of viable cells versus total cell numbers per region of interest. Shown are means 6 SD. n 5 9.

Since we intend to use 3D-bioprinting for prospective
bone tissue engineering applications, we tuned fibrin and
collagen with HA, which represents an excellent bone sub-
stitute material.14 In the Live/Dead assay, we observed
that MSC attachment and viability was not altered upon
inclusion of HA. In contrast, we observed a positive effect
of HA on MSC proliferation on collagen and fibrin hydro-
gels, whereby this effect was most pronounced on fibrin
gels. A positive influence of HA on MSC proliferation was
previously reported on biphasic calcium phosphate
ceramics.15
We also observed that addition of HA to fibrin strongly
improved the osteogenic differentiation of MSCs, as evi-
denced by an increase in alkaline phosphatase activity. This
was also partially seen in collagen, but only in 1 mg/ml col-
lagen hydrogels when MSCs were grown under osteogenic
growth conditions. A positive effect of HA on osteogenic dif-
ferentiation of MSCs was previously reported.15,19,20 Con-
cerning the fibrin hydrogel, we also noticed that there was
an inverse relationship between the thrombin concentra-
tions, used to induce gelling of the fibrinogen, and the ALP
activity. This means, that stiffer fibrin gels support

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | DEC 2017 VOL 105A, ISSUE 12 3239
osteogenic differentiation of native MSCs, at least when cul-
tivated in osteogenic differentiation medium. However, a
contrary relationship could be detected on collagen gels. In
this case, osteogenic differentiation of MSCs was better at
low collagen concentrations, resulting in less stiffer gels.
The swelling/degradation properties of hydrogels repre-
sents an important point in the context of the long-term sta-
bility of the graft after implantation. The ideal hydrogel for
our prospective application should show only minimal swel-
ling or degradation. In a 1 week swelling/degradation in
vitro study, where the hydrogels have been submerged in
growth medium, we have seen that both HA-tuned hydro-
gels showed neither swelling, nor degradation at all tested
concentrations. Therefore, HA-tuned fibrin and collagen rep-
resent excellent hydrogels for bioprinting of MSCs.
Viscosity has been measured regarding inkjet-based
printing ability. HA-tuned fibrinogen and collagen do not
show a Newtonian behavior of shear rate versus shear
stress, instead they are shear thinning, exhibiting pseudo-
plastic behaviors typical for biopolymer solutions. This
means the viscosity decreases with increasing shear rates.
However, HA-tuned fibrin as well as collagen exhibit viscos-
ities ranging from 2 to 15 mPa s (for shear rates between 1
and 2000/s) which can be usually dispensed by most inkjet
printers.25
Viscoelastic properties are indicating the hydrogels
potential to resist deformations. This is especially important
for the printing process itself and for potential handling
steps during graft implantation. For this reason mechanical
stability can be determined by measuring the storage and
loss modulus of hydrogels. It is evident that hydrogels
exhibiting high storage moduli will be preferably used for
bioprinting as they are less prone against deformations.
For HA-tuned collagen, the storage moduli increase with
increasing concentrations. Concerning HA-tuned fibrin, there
was an inverse correlation between the storage modulus
and the thrombin concentration, used to crosslink fibrino-
gen. This means that crosslinking of fibrinogen with low
concentrations of thrombin results in stiffer hydrogels. This
effect may be explained by a more homogeneous crosslink-
ing process at lower thrombin concentrations.
Finally, we investigated the printability of fibrin and colla-
gen, as well as viability of printed MSCs. As expected, we were
able to demonstrate printability of both hydrogels with our
printing system based on a PipeJet P 9 dispenser (BioFluidix
GmbH, Germany) as previously reported by other groups
working with other DoD printers.30,40,41 Moreover, tuning of
fibrin and collagen with HA had no negative effect, neither on
printability, nor on survival of bioprinted MSCs. In our experi-
ments it became evident that there was also no difference
detectable between fibrin and collagen in terms of viability of
bioprinted MSCs. Printing of collagen-hydoxyapatite compos-
ite bioinks has already been demonstrated by means of low-
temperature additive manufacturing.42 However, to our
knowledge, DoD-based bioprinting of HA-tuned collagen or
HA-tuned fibrin has not been demonstrated so far.
In this study, MSCs were printed in two-dimensional
arrays by our PipeJet dispenser to prove that the selected
3240 BENNING ET AL.
setup and the resulting shear force generated by our DoD-
based printing device are compatible with the survival of
MSCs. However, the spatial allocation of MSCs in a 3D
printed construct was not investigated so far, but this issue
will be addressed in prospective studies. Nevertheless, it
has already been reported in the literature that printing of
MSCs in fibrin-collagen gels in multiple layers, thereby pro-
ducing a 3D MSC-laden construct, is feasible.30
In summary, taking into account all of the results gener-
ated in this study and also in the context of regulatory
affairs, among the hydrogels investigated in this study, fibrin
and collagen turned out to be most suitable for bioprinting
of MSCs to produce bone tissue equivalents. Modification of
collagen and fibrin with HA further supported the biological
parameters of MSC-seeded hydrogels. Furthermore, HA-
tuning of these hydrogels did neither compromise printabil-
ity, nor viability of bioprinted MSCs. This study paves the
way for further exploration of 3D printing technology for
bioprinting of MSCs to produce bone tissue equivalents for
various medical applications.
ACKNOWLEDGMENTS
We thank Brunhilde Baumer for excellent technical assistance.
This work was supported by funding through the Deutsche
Forschungsgemeinschaft (FI 790/10–1 and KO 3910/1–1).
REFERENCES
1. Khan SN, Cammisa FP, Jr., Sandhu HS, Diwan AD, Girardi FP,
Lane JM. The biology of bone grafting. J Am Acad Orthop Surg
2005;13:77–86.
2. Silber JS, Anderson DG, Daffner SD, Brislin BT, Leland JM,
Hilibrand AS, Vaccaro AR, Albert TJ. Donor site morbidity after
anterior iliac crest bone harvest for single-level anterior cervical
discectomy and fusion. Spine (Phila Pa 1976) 2003;28:134–139.
3. Kneser U, Schaefer DJ, Polykandriotis E, Horch RE. Tissue engi-
neering of bone: the reconstructive surgeon’s point of view. J Cell
Mol Med 2006;10:7–19.
4. Sotiropoulou PA, Perez SA, Salagianni M, Baxevanis CN,
Papamichail M. Characterization of the optimal culture conditions
for clinical scale production of human mesenchymal stem cells.
Stem Cells 2006;24:462–471.
5. Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R,
Mosca JD, Moorman MA, Simonetti DW, Craig S, Marshak DR.
Multilineage potential of adult human mesenchymal stem cells.
Science 1999;284:143–147.
6. Wakitani S, Goto T, Pineda SJ, Young RG, Mansour JM, Caplan
AI, Goldberg VM. Mesenchymal cell-based repair of large, full-
thickness defects of articular cartilage. J Bone Joint Surg Am
1994;76:579–592.
7. Bruder SP, Kurth AA, Shea M, Hayes WC, Jaiswal N, Kadiyala S.
Bone regeneration by implantation of purified, culture-expanded
human mesenchymal stem cells. J Orthop Res 1998;16:155–162.
8. Kon E, Muraglia A, Corsi A, Bianco P, Marcacci M, Martin I, Boyde A,
Ruspantini I, Chistolini P, Rocca M, Giardino R, Cancedda R, Quarto
R. Autologous bone marrow stromal cells loaded onto porous
hydroxyapatite ceramic accelerate bone repair in critical-size defects
of sheep long bones. J Biomed Mater Res 2000;49:328–337.
9. Gazit D, Turgeman G, Kelley P, Wang E, Jalenak M, Zilberman Y,
Moutsatsos I. Engineered pluripotent mesenchymal cells integrate
and differentiate in regenerating bone: A novel cell-mediated
gene therapy. J Gene Med 1999;1:121–133.
10. Derby B. Printing and prototyping of tissues and scaffolds. Sci-
ence 2012;338:921–926.
11. Ozbolat IT, Yu Y. Bioprinting toward organ fabrication: Challenges
and future trends. IEEE Trans Biomed Eng 2013;60:691–699.
CYTOCOMPATIBILITY TESTING OF HYDROGELS

12. Murphy SV, Atala A. 3D bioprinting of tissues and organs. Nat
Biotechnol 2014;32:773–785.
13. Malda J, Visser J, Melchels FP, Jungst T, Hennink WE, Dhert WJ,
Groll J, Hutmacher DW. 25th anniversary article: Engineering
hydrogels for biofabrication. Adv Mater 2013;25:5011–5028.
14. Tadic D, Epple M. A thorough physicochemical characterisation
of 14 calcium phosphate-based bone substitution materials in
comparison to natural bone. Biomaterials 2004;25:987–994.
15. Hu J, Zhou Y, Huang L, Liu J, Lu H. Effect of nano-hydroxyapatite
coating on the osteoinductivity of porous biphasic calcium phos-
phate ceramics. BMC Musculoskelet Disord 2014;15:114.
16. Zhang C, Cao M, Lan J, Wei P, Cai Q, Yang X. Regulating prolifer-
ation and differentiation of osteoblasts on poly(l-lactide)/gelatin
composite nanofibers via timed biomineralization. J Biomed
Mater Res A 2016;104:1968–1980.
17. Tesavibul P, Chantaweroad S, Laohaprapanon A, Channasanon S,
Uppanan P, Tanodekaew S, Chalermkarnnon P, Sitthiseripratip K.
Biocompatibility of hydroxyapatite scaffolds processed by
lithography-based additive manufacturing. Biomed Mater Eng
2015;26:31–38.
18. Neuss S, Stainforth R, Salber J, Schenck P, Bovi M, Knuchel R,
Perez-Bouza A. Long-term survival and bipotent terminal differen-
tiation of human mesenchymal stem cells (hMSC) in combination
with a commercially available three-dimensional collagen scaf-
fold. Cell Transplant 2008;17:977–986.
19. Bassi G, Guilloton F, Menard C, Di Trapani M, Deschaseaux F,
Sensebe L, Schrezenmeier H, Giordano R, Bourin P, Dominici M,
and, others. Effects of a ceramic biomaterial on immune modula-
tory properties and differentiation potential of human mesenchy-
mal stromal cells of different origin. Tissue Eng Part A 2015;21:
767–781.
20. He J, Jiang B, Dai Y, Hao J, Zhou Z, Tian Z, Wu F, Gu Z. Regula-
tion of the osteoblastic and chondrocytic differentiation of stem
cells by the extracellular matrix and subsequent bone formation
modes. Biomaterials 2013;34:6580–6588.
21. Matsushima A, Kotobuki N, Tadokoro M, Kawate K, Yajima H,
Takakura Y, Ohgushi H. In vivo osteogenic capability of human
mesenchymal cells cultured on hydroxyapatite and on beta-
tricalcium phosphate. Artif Organs 2009;33:474–481.
22. Mehlhorn AT, Schmal H, Kaiser S, Lepski G, Finkenzeller G, Stark
GB, Sudkamp NP. Mesenchymal stem cells maintain TGF-beta-
mediated chondrogenic phenotype in alginate bead culture. Tis-
sue Eng 2006;12:1393–1403.
23. Dominici M, Le Blanc K, Mueller I, Slaper-Cortenbach I, Marini F,
Krause D, Deans R, Keating A, Prockop D, Horwitz E. Minimal cri-
teria for defining multipotent mesenchymal stromal cells. The
International Society for Cellular Therapy position statement.
Cytotherapy 2006;8:315–317.
24. Blum JS, Li RH, Mikos AG, Barry MA. An optimized method for
the chemiluminescent detection of alkaline phosphatase levels
during osteodifferentiation by bone morphogenetic protein 2.
J Cell Biochem 2001;80:532–537.
25. Calvert P. Inkjet Printing for materials and devices. Chem Mater
2001; 13:3299–3305.
26. Norotte C, Marga FS, Niklason LE, Forgacs G. Scaffold-free vascu-
lar tissue engineering using bioprinting. Biomaterials 2009;30:
5910–5917.
27. Boland T, Mironov V, Gutowska A, Roth EA, Markwald RR. Cell
and organ printing 2: Fusion of cell aggregates in three-
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | DEC 2017 VOL 105A, ISSUE 12
ORIGINAL ARTICLE
dimensional gels. Anat Rec A Discov Mol Cell Evol Biol 2003;272:
497–502.
28. Cui X, Boland T. Human microvasculature fabrication using ther-
mal inkjet printing technology. Biomaterials 2009;30:6221–6227.
29. Sakai S, Ito S, Inagaki H, Hirose K, Matsuyama T, Taya M,
Kawakami K. Cell-enclosing gelatin-based microcapsule produc-
tion for tissue engineering using a microfluidic flow-focusing sys-
tem. Biomicrofluidics 2011;5:13402.
30. Skardal A, Mack D, Kapetanovic E, Atala A, Jackson JD, Yoo J,
Soker S. Bioprinted amniotic fluid-derived stem cells accelerate
healing of large skin wounds. Stem Cells. Transl Med 2012;1:792–
802.
31. Snyder JE, Hamid Q, Wang C, Chang R, Emami K, Wu H, Sun W.
Bioprinting cell-laden matrigel for radioprotection study of liver
by pro-drug conversion in a dual-tissue microfluidic chip. Biofab-
rication 2011;3:034112.
32. Duarte Campos DF, Blaeser A, Weber M, Jakel J, Neuss S,
Jahnen-Dechent W, Fischer H. Three-dimensional printing of stem
cell-laden hydrogels submerged in a hydrophobic high-density
fluid. Biofabrication 2013;5:015003.
33. Fedorovich NE, Schuurman W, Wijnberg HM, Prins HJ, van Weeren
PR, Malda J, Alblas J, Dhert WJ. Biofabrication of osteochondral
tissue equivalents by printing topologically defined, cell-laden
hydrogel scaffolds. Tissue Eng Part C Methods 2012;18:33–44.
34. Wust S, Godla ME, Muller R, Hofmann S. Tunable hydrogel com-
posite with two-step processing in combination with innovative
hardware upgrade for cell-based three-dimensional bioprinting.
Acta Biomater 2014;10:630–640.
35. Bidarra SJ, Barrias CC, Fonseca KB, Barbosa MA, Soares RA,
Granja PL. Injectable in situ crosslinkable RGD-modified alginate
matrix for endothelial cells delivery. Biomaterials 2011;32:7897–
7904.
36. Wang W, Liu Z, Sun P, Fang C, Fang H, Wang Y, Ji J, Chen J.
RGD peptides-conjugated pluronic triblock copolymers encapsu-
lated with AP-2alpha expression plasmid for targeting gastric can-
cer therapy in vitro and in vivo. Int J Mol Sci 2015;16:16263–
16274.
37. Kopf M, Campos DF, Blaeser A, Sen KS, Fischer H. A tailored
three-dimensionally printable agarose-collagen blend allows
encapsulation, spreading, and attachment of human umbilical
artery smooth muscle cells. Biofabrication 2016;8:025011.
38. Ceylan S, Gokturk D, Bolgen N. Effect of crosslinking methods on
the structure and biocompatibility of polyvinyl alcohol/gelatin cry-
ogels. Biomed Mater Eng 2016;27:327–340.
39. Lai JY, Ma DH. Glutaraldehyde cross-linking of amniotic mem-
branes affects their nanofibrous structures and limbal epithelial
cell culture characteristics. Int J Nanomedicine 2013;8:4157–4168.
40. Moon S, Hasan SK, Song YS, Xu F, Keles HO, Manzur F,
Mikkilineni S, Hong JW, Nagatomi J, Haeggstrom E,
Khademhosseini A, Demirci U. Layer by layer three-dimensional
tissue epitaxy by cell-laden hydrogel droplets. Tissue. Eng Part C
Methods 2010;16:157–166.
41. Xu T, Binder KW, Albanna MZ, Dice D, Zhao W, Yoo JJ, Atala A.
Hybrid printing of mechanically and biologically improved con-
structs for cartilage tissue engineering applications. Biofabrication
2013; 5:015001.
42. Lin KF, He S, Song Y, Wang CM, Gao Y, Li JQ, Tang P, Wang Z,
Bi L, Pei GX. Low-temperature additive manufacturing of biomi-
mic three-dimensional hydroxyapatite/collagen scaffolds for bone
regeneration. ACS Appl Mater Interfaces 2016;8:6905–6916.
3241