Biomaterials 140 (2017) 103e114

Contents lists available at ScienceDirect

Biomaterials
journal homepage: www.elsevier.com/locate/biomaterials

Biomaterials that promote cell-cell interactions enhance the paracrine

function of MSCs
Taimoor H. Qazi a, b, David J. Mooney c, Georg N. Duda a, b, Sven Geissler a, b, *
a
Julius Wolff Institute, Charit
€tsmedizin Berlin, Augustenburger Platz 1, 13353 Berlin, Germany
e Universita
b
Berlin-Brandenburg Center for Regenerative Therapies & Berlin-Brandenburg School for Regenerative Therapies, Charit
€tsmedizin Berlin,
e Universita
Augustenburger Platz 1, 13353 Berlin, Germany
c
John A. Paulson School of Engineering and Applied Sciences, Harvard University, 29 Oxford St., Cambridge, MA 02138, USA

a r t i c l e i n f o a b s t r a c t

Article history: Mesenchymal stromal cells (MSCs) secrete paracrine factors that play crucial roles during tissue
Received 6 June 2017 regeneration. Whether this paracrine function is influenced by the properties of biomaterials in general,
Accepted 16 June 2017 and those used for cell delivery in particular, largely remains unexplored. Here, we investigated if three-
Available online 18 June 2017
dimensional culture in distinct microenvironments - nanoporous hydrogels (mean pore size ~5 nm) and
macroporous scaffolds (mean pore size ~120 mm) - affects the secretion pattern of MSCs, and conse-
Keywords:
quently leads to differential paracrine effects on target progenitor cells such as myoblasts. We report that
Trophic secretion
compared to MSCs encapsulated in hydrogels, scaffold seeded MSCs show an enhanced secretion profile
N-cadherin
Muscle regeneration
and exert beneficial paracrine effects on various myoblast functions including migration and prolifera-
Stem cell therapy tion. Additionally, we show that the heightened paracrine effects of scaffold seeded cells can in part be
3D culture attributed to N-cadherin mediated cell-cell interactions during culture. In hydrogels, this physical
Paracrine signaling interaction between cells is prevented by the encapsulating matrix. Functionally blocking N-cadherin
Cryogel negatively affected the secretion profile and paracrine effects of MSCs on myoblasts, with stronger effects
Hydrogel observed for scaffold seeded compared to hydrogel encapsulated cells. Together, these findings
demonstrate that the therapeutic potency of MSCs can be enhanced by biomaterials that promote cell-
cell interactions.
© 2017 Elsevier Ltd. All rights reserved.

1. Introduction
Inefficient natural healing of several debilitating musculoskel-
etal injuries such as volumetric tissue loss and severe muscle
trauma has led to an unmet clinical demand for potent therapies.
Strategies that embody the principles of tissue engineering and
regenerative medicine, such as autologous and allogeneic stem cell
therapies, have gained considerable traction in the past decade
[1e3]. As witnessed by their extensive application in registered
clinical trials, mesenchymal stromal cells (MSCs) in particular have
attracted widespread interest as safe and effective candidates for
the regeneration of injured tissues [4].
In addition to their well-known multi-lineage differentiation
potential, an intriguing characteristic of MSCs is their ability to

Universita
* Corresponding author. Julius Wolff Institute, Charite €tsmedizin Berlin,
Augustenburger Platz 1, 13353 Berlin, Germany.
E-mail address: sven.geissler@charite.de (S. Geissler).
secrete a wide range of bioactive cytokines and growth factors that
can influence nearby cells via paracrine signaling [5,6]. For instance,
MSCs have been documented to be a generous source of bioactive
factors including VEGF, FGF, HGF, IGF, PDGF, ILs, and MMPs [6e9].
These secreted factors can in turn orchestrate various biological
processes that may be desirable for tissue regeneration such as
angiogenesis, immune modulation, and cell recruitment and dif-
ferentiation [10,11]. An increasing body of evidence suggests that
the therapeutic benefits observed following MSC transplantation
may largely be attributed to the paracrine and trophic function of
these cells, rather than their differentiation potential [12]. For this
reason, research efforts by a number of groups have focused on
strategies that can enhance the secretory function of MSCs. These
include preconditioning in hypoxic conditions, genetic manipula-
tion to overexpress certain growth factors, and cytokine stimula-
tion [13,14]. For example, in a previous study we reported that the
paracrine effects of MSCs stimulated with soluble VEGF and IGF-1
enhanced myoblast migration, proliferation, and survival in vitro,
and led to robust muscle regeneration in vivo [15]. Recent studies

http://dx.doi.org/10.1016/j.biomaterials.2017.06.019
0142-9612/© 2017 Elsevier Ltd. All rights reserved.
104 T.H. Qazi et al. / Biomaterials 140 (2017) 103e114
have highlighted the importance of cell-cell interactions in regu-
lating the paracrine activity of MSCs by reporting that 3D multi-
cellular spheroids show superior secretion profiles compared to
similar number of MSCs in monolayer culture. For instance, Lee
et al. reported that MSC spheroid formation via calcium dependent
E-cadherin interactions led to improved paracrine activity and
therapeutic efficacy in a rat myocardial infarction model [16].
Similar paracrine effects have been reported for spheroids formed
using adipose derived [17], umbilical cord derived [18], and bone
marrow derived MSCs [19].
Researchers have also explored the feasibility of applying MSC
derived conditioned medium (MSC-CM) to stimulate tissue
regeneration. For instance, Osugi et al. reported enhanced in vivo
bone formation in rat calvarial defects after the application of MSC-
CM that contained high quantities of IGF-1 and VEGF [20]. Others
have used this cell-free approach in spinal cord and cardiac injury
models [21,22]. While the transplantation of such a cytokine rich
cocktail avoids the cost and labor intensive process of cell har-
vesting and subsequent expansion, these cytokines may quickly
lose bioactivity in vivo due to short half-lives. Moreover, complete
recovery from severe injuries may require a prolonged delivery of
paracrine factors which necessitates MSC transplantation. In the
vast majority of pre-clinical and clinical studies, MSCs are admin-
istered in a bolus manner either systemically (via the circulatory
system) or locally at the site of injury [23,24]. A number of draw-
backs are associated with these modes of cell delivery. These
include rapid cell death [25], accumulation in non-targeted tissues
[26], and diffusion of the cell suspension away from the site of
injury due to anoikis [27]. By providing a temporary matrix that can
facilitate cell attachment, viability, and growth, biomaterials can
potentially be used as local delivery depots secreting paracrine
signaling molecules at sites of tissue damage. Tuning the physical
properties of such biomaterial depots may offer the possibility to
elicit desirable MSC activity and thereby steer the regenerative
function of target progenitor cells.
A wide range of biomaterials with varying structures, chemis-
tries, and mechanical properties have been used for in vivo cell
delivery [28,29]. From a clinical translation perspective, nano-
porous hydrogels may be considered attractive materials for cell
delivery because they can be injected near the site of injury in a
minimally invasive manner and adapt to the size and shape of the
defect [30,31]. Commonly used hydrogels such as alginate, poly
ethylene glycol (PEG), and hyaluronic acid offer the advantage of
tunable chemistries and mechanical properties, but prevent
changes in the morphology of encapsulated cells. On the other
hand, pre-formed 3D scaffolds fabricated via 3D printing, electro-
spinning, or freeze drying, may not be conveniently injectable but
offer macroporous microenvironments that facilitate cell
spreading, migration, and establishment of cell-cell contacts [32].
While a great deal of research effort has focused on using bio-
materials to direct stem cell fate, there is a lack of sufficient
knowledge on how biomaterial properties influence the paracrine
function of MSCs.
We hypothesized that compared to nanoporous hydrogels that
are used for cell encapsulation and have pores in the range of
5e70 nm [33,34], macroporous scaffolds with pore sizes in the
range of 10e200 mm may facilitate both cell-cell and cell-matrix
interactions and lead to enhanced paracrine effects of MSCs. To
address this hypothesis, we evaluated the secretion profile of MSCs
cultured in 3D alginate based substrates that had similar mechan-
ical and chemical properties but distinct structural microenviron-
ments. The secreted factors were characterized, and their potential
implication for muscle regeneration was investigated by analyzing
the behavior and function of myoblasts cultured in conditioned
media. Functional blocking experiments were used to determine
the role of substrate microenvironment mediated cell-cell in-
teractions in regulating the paracrine effects of MSCs. Our data
indicate that changes in biomaterial microenvironment can elicit
strong paracrine responses from MSCs by promoting local cell-cell
interactions.
2. Materials and methods
2.1. Alginate processing and substrate fabrication
Ultrapure low and high molecular weight alginates (LVG, MVG)
were purchased from Pronova. A 1:1 (w/w) mixture of the alginates
was reconstituted to a concentration of 1% w/v, and functionalized
with the cell adhesive peptide motif G4RGDSP (Peptide 2.0, USA)
using standard carbodiimide chemistry to a final degree of substi-
tution of 20, as previously described [35]. Briefly, NHS, EDC, and the
peptide motifs were added sequentially to the reconstituted algi-
nate, and allowed to react for 20 h. The reaction was stopped using
hydroxyl amine hydrochloride, and the alginates were dialyzed
(MWCO 10000, Spectra/Por®) over three days in distilled water
with a decreasing salt concentration. The dialyzed alginate was
mixed with activated charcoal, sterile filtered, frozen, and lyophi-
lized until use.
Hydrogels containing encapsulated cells were prepared as fol-
lows. A 2.5% w/v solution of alginate was mixed with a cell sus-
pension containing 2
106 freshly trypsinized MSCs and 4% v/v of
calcium sulfate slurry (1.22 M in dH2O) using two syringes con-
nected with a Luer-lock syringe connector. The components were
mixed rapidly to a yield a cell containing alginate mixture (conc. 2%
w/v) and cast between two glass plates separated by a 1 mm spacer.
The mixture was allowed to ionically crosslink for 45 min before
gels were punched out (diameter: 8 mm) and placed in 24 well
plates with full cell culture media. Scaffolds were prepared in a
similar manner, but without mixing with cells. The punched out
gels (~50 mm3) were frozen at 80  C and lyophilized to create a
porous structure. The interconnected porous structure was
confirmed with scanning electron microscopy, and pore size (short
axis) was manually determined using ImageJ. Freshly trypsinized
MSCs were then seeded onto the 3D porous scaffolds (4
106 cells/
cm3) and incubated for 20 min before adding full culture media.
2.2. Mechanical characterization
Hydrogel and scaffold stiffness was determined in the wet state
using a bench-top PIUMA nanoindenter (Optics 11, Amsterdam). A
spherical indentation tip (radius 48.5 mm) was used to probe
samples immersed in PBS up to an indentation depth of 10 mm
(sample thickness ~1000 mm). Matrix scans (3
3 points, step size
in each direction ¼ 100 mm) were carried out at several different
locations on a number of hydrogel and scaffold samples. Using the
Oliver and Pharr model, the PIUMA nanoindenter estimated the
samples' Young's modulus from the slope of the unloading curve in
the region 65%e85% of maximum load.
2.3. Cell culture
MSCs isolated from the tibiae of Sprague Dawley rats via bone
marrow biopsies were used in this study. Cell culture was carried
out in full culture media i.e. low glucose DMEM containing fetal calf
serum (10%), penicillin/streptomycin (1%), and Glutamax™ (1%).
Cells were passaged at approximately 70% confluency, and were not
used beyond passage 5. C2C12 myoblasts (ATCC) were cultured in
full culture media and were not used beyond passage 10. To induce
myogenic differentiation, the serum concentration was reduced to
5%.
 T.H. Qazi et al. / Biomaterials 140 (2017) 103e114
2.4. MSC viability
MSCs were isolated from hydrogels or scaffolds by alginate
dissolution using 50 mM EDTA in PBS. Cells were stained with
calcein AM (1:1000) and propidium iodide (1:2500) in culture
media for 15 min. The stained cell suspension was imaged using an
inverted fluorescent microscope. Viability was calculated by
quantifying the number of live and dead cells in at least 5 regions of
interest.
2.5. Conditioned media (CM)
To generate CM from tissue culture plastic (TCP), MSCs were
seeded at a density of 1
104 cells/cm2 on 6 well plates. To generate
Hydrogel-CM, freshly trypsinized MSCs were mixed with alginate
at a density of 2
106 cells/mL alginate and the mixture was
subsequently crosslinked. To generate Scaffold-CM, trypsinized
MSCs were seeded onto scaffolds at a density of 4
106 cells/cm3. A
high cell density was used for the scaffolds to account for the cell
leakage (~40e60%) associated with the macroporosity. MSCs
cultured on TCP, hydrogels, or scaffolds were washed with PBS
before adding 1.5 mL serum-free media. For hydrogel encapsulated
and scaffold seeded MSCs, collection of CM started 3 days after
initial cell seeding/encapsulation. After 24 h, the conditioned media
was collected, centrifuged to pellet cellular debris, and the super-
natants were sterile filtered and stored at 80  C until use. At the
time of CM collection, cells on the different substrates were coun-
ted using the CASY® cell counter. Cells on TCP were trypsinized,
whereas those on alginate substrates were isolated by using 50 mM
EDTA to dissociate the scaffolds/gels. Cell numbers were then used
to normalize concentrations and intensities obtained in ELISA and
cytokine array experiments.
2.6. N-cadherin blocking
Functional blockade of cell-cell interactions was carried out
according to methods described in previously published reports
[36,37]. Briefly, trypsinized MSCs were pelleted and re-suspended
in media containing 40 mg/mL anti-N cadherin antibody (GC-4,
Sigma-Aldrich) and incubated for 45 min. The cells were then
washed twice with PBS and seeded onto the scaffolds or encapsu-
lated in hydrogels. After 3 days in culture, MSCs were treated with
4 mM EDTA in PBS for 5 min to disrupt any calcium dependent cell-
cell interactions, washed twice with PBS, and placed in fresh serum
free media supplemented with an additional 10 mg/mL of anti-N
cadherin antibody. The additional application of N-cadherin anti-
body was carried out to prevent the re-establishment of cell-cell
contacts that were disrupted using EDTA.
2.7. ELISA and cytokine arrays
Rat specific ELISAs for VEGF (cat. # RRV00), FGF (cat. # MFB00),
HGF (cat. # MHG00), and IGF (cat. # MG100) were purchased from
R&D systems, and used according to the manufacturer's in-
structions to detect concentrations in conditioned media. Rat spe-
cific ELISA for LIF was purchased from LSBio (cat. # LS-F13295), and
used according to the manufacturer's instructions. Rat specific
cytokine arrays were purchased from Ray Biotech (cat. # AAR-BLM-
1-4-RB and cat. # AAR-CYT-2-8-RB), and used according to the
manufacturer's instructions. Images were acquired using a G:BOX
imager (Syngene). Intensities were determined using the Protein
Array Analyzer plugin in ImageJ. To account for inter-array differ-
ences, all detected intensities were normalized to positive controls
within each array. The intensities were then normalized to cell
105
numbers recorded earlier for each substrate condition.
2.8. Myoblast migration
Migration was determined using a scratch wound healing assay,
as previously described [15]. Briefly, C2C12 myoblasts were seeded
in 24 well plates at a density of 1.4
105 per well and allowed to
reach confluency. A 200 mL pipette tip was used to make a scratch
across the monolayer. Wells were washed with PBS and appropriate
conditioned media was then added. Selected regions of interest
were imaged every 30 min using a live cell microscope (Leica,
Germany). Scratch area was determined using T-Scratch software
[38]. Cell velocity, distance, displacement, and directionality were
determined by using the Manual Tracking tool in ImageJ to track
single cells in independent wells, and the data was analyzed using
the Chemotaxis and Migration Tool (ibidi®).
2.9. Myoblast proliferation
Proliferation was determined using the CyQUANT® cell prolif-
eration assay kit, as previously described [15]. Briefly, C2C12
myoblasts were seeded in 48 well plates at a density of 2.5
104
per well and allowed to attach overnight. After washing with PBS,
appropriate serum free conditioned media was then added to the
wells, and cultured over two days. At t ¼ 0, 24 h, and 48 h, media
was aspirated, cells were washed with PBS, and frozen at 80  C for
up to 7 days. For quantification, cells were thawed, and CyQUANT
dye (400 mM) and lysis buffer (20 mM) in distilled water were
added to each well followed by incubation for 7.5 min in the dark.
Fluorescence values were measured using a plate reader.
2.10. Myoblast survival
Cell survival was determined using the Trypan Blue dye to
calculate cell viability. After 24 h culture in appropriate media
(control, or CM), cells were washed with PBS, trypsinized, and the
cell suspension was mixed with Trypan blue dye (1:1). The number
of live and dead cells were quantified by counting using a
hemocytometer.
2.11. Myoblast differentiation
To assess myotube formation, C2C12 myoblasts were seeded in
24 well plates at a density of 1
105 per well and allowed to attach
overnight. To induce differentiation, appropriate serum free
conditioned media was exogenously supplemented with 5% v/v FCS
before being added to the cells. Media was refreshed on alternate
days. Myogenic differentiation was evaluated using immunofluo-
rescence imaging as described earlier. Myotube lengths and density
were quantified using ImageJ. Further quantification was carried
out using Western Blot analysis. Differentiated myoblasts were
lysed using SDS lysis buffer. Protein content was determined using
the DC protein assay (Bio-Rad). The Novex NuPAGE® system (Invi-
trogen) was used according to manufacturer's instructions followed
by semi-dry transfer. Primary antibodies for Myosin heavy chain
(Clone # MF20, R&D Systems, Catalog # MAB4470) (1:500), Myo-
genin (Clone #F5D, Abcam, Catalog # ab1835) (1:500), and the
house keeping gene GAPDH (Clone # 6C5, Abcam, Catalog #
ab8245) (1:10000) were used with TBSTþ5% non-fat milk. Anti-
mouse horseradish peroxidase (GE Healthcare) (1:7000) was uti-
lized as the secondary antibody. Using ECL substrate (GE Health-
care) pictures were acquired using the G:BOX imager (Syngene).
Analysis was performed using band intensities normalized to
GAPDH and quantified by ImageJ. Myoblast fusion index was
106 T.H. Qazi et al. / Biomaterials 140 (2017) 103e114
quantified by expressing as a percentage the number of nuclei in
myotubes divided by the total number of nuclei in myotubes and
myoblasts per region of interest (ROI).
2.12. Immunofluorescent staining
Cells were fixed with 4% paraformaldehyde for 10 min, per-
meabilized with 0.1% saponin (myotubes) or 0.1% Triton-X and
Tween-20 (MSCs), blocked in 1% BSA-PBS containing 3% FCS, and
incubated with the appropriate primary antibody (N-Cadherin:
1:100, cat. # H-63, Santa Cruz Biotechnology (a kind gift from Prof.
Sigmar Stricker); Myosin heavy chain: 1:100, cat. # MAB4470,
Thermo Scientific) overnight on a shaker at 4  C. Cells were then
incubated with an appropriate secondary antibody (for N-Cadherin:
Goat anti-rabbit IgG Alexa Fluor® 546, cat. # A11035, Invitrogen; for
Myosin heavy chain: Goat anti-mouse IgG Alexa Fluor® 488, cat. #
A11029, Invitrogen) for 1 h at room temperature. DAPI (1:1000 of
1 mg/mL initial concentration in PBS) was added for 10 min at room
temperature. Cells were washed with PBS, and visualized using an
inverted microscope (DMI6000B, Leica, Germany). Actin filaments
were stained with Alexa Fluor® 488 Phalloidin (cat. # A12379,
Thermo Scientific) according to the manufacturer's instructions.
2.13. qPCR
For gene expression analysis, cell containing scaffolds or gels
were mechanically homogenized in a tube containing Trizol® re-
agent (Life Technologies), and RNA was isolated according to the
manufacturer's instructions. RNA concentration was determined
using a NanoDrop spectrophotometer. The iScript™ Reverse Tran-
scription Supermix was used to transcribe RNA into cDNA. SYBR
Green dye was used to detect fluorescence. The amplification pro-
file was assessed using a LightCycler® 480 (Roche, Germany). Gene
expression was quantified using the DDCt method and fold change
was calculated using the formula 2-DDCt. Values for the genes of
interest were normalized to the housekeeping gene (Eef1a). The
following primers were used for qPCR analysis:
Gene
CDH1 (E-Cadherin)
CDH2 (N-Cadherin)
Hgf (Hepatocyte growth factor)
Vegfa (Vascular endothelial growth factor A)
Fgf2 (Fibroblast growth factor 2)
Mmp13 (Matrix metalloproteinase 13)
Tgfb1 (Transforming growth factor beta-1)
Eef1a (Elongation factor) [Housekeeping gene]
GAPDH (Glyceraldehyde 3-phosphate dehydrogenase)
2.14. Statistical analysis
All values are depicted as mean ± standard deviation. Experi-
ments were performed with MSCs from four biological replicates,
and were repeated independently in some cases. This information
is included in figure legends. Statistical analysis was carried out in
GraphPad Prism 7.0 (GraphPad Software Inc., USA). The Shapiro-
Wilk test was used to test normality. If the data was normally
distributed, two-tailed student's t-test (two groups) or one-way
ANOVA with Tukey's post-hoc test (>two groups) was used. If
normality could not be confirmed, the Mann Whitney U test (two
groups) or the Kruskal-Wallis test with Tukey's post-hoc test (>two
groups) was used. Levels of statistical significance were set at
*p < 0.05, **p < 0.01, ***p < 0.001.
3. Results
3.1. MSC culture in distinct substrate microenvironments
In addition to TCP wells, primary rat bone marrow derived MSCs
were cultured in alginate based 3D substrates. RGD modified algi-
nate was ionically crosslinked to fabricate nanoporous hydrogels
and freeze-dried macroporous (mean pore size: 122 ± 29 mm)
scaffolds (Fig. S1). Although both are hydrogels, we will refer to the
nanoporous and macroporous alginates as hydrogels and scaffolds,
respectively. MSCs on TCP exhibited a typical spread morphology
with prominent actin filaments (Fig. 1 a). When encapsulated in
hydrogels, MSCs were homogenously distributed but adopted a
round morphology (Fig. 1 b). On scaffolds, MSCs were observed to
spread on and across the interconnected pore walls where they
established physical contacts with neighboring cells (Fig. 1 c).
Despite the changes in morphology, MSCs maintained high viability
for at least seven days in static in vitro culture conditions in all
groups (Fig. 1 d). While TCP is known to be non-physiologically stiff,
the alginate based hydrogel and scaffold substrates had similar
mean stiffness values of ~20 kPa (Fig. 1 e).
3.2. Modifying biomaterial microenvironment alters MSCs'
secretory profile
The conditioned media of MSCs cultured on the three substrates
(hereafter referred to as: TCP-CM, Hydrogel-CM, and Scaffold-CM)
were characterized using cytokine arrays and ELISA kits to assess
whether changes in substrate microenvironment influenced MSC
secretion profile. Cytokine array analysis (Fig. S3) using MSCs from
four biological donors showed striking differences in the secretion
profile depending on the culture substrate (Fig. 2 a). In general, high
concentrations of almost all cytokines were detected in Scaffold-
CM, whereas lower concentrations (represented by darker in-
tensities on the heatmap) were observed for TCP-CM and Hydrogel-
Forward Primer (50 /30 ) Reverse Primer (50 /30 )
ACCCCCTGTTGGCGTTTTCA CATCACGGAGGTTCCTGGAAGAG
GGAGCCGATGAAGGAACCACA TGAAGATGCCCGTTGGAGGC
CCCCCATGAACACAGCTTTTTG GCTTTCACCGTTGCAGGTCA
CTTCAAGCCGTCCTGTGTGC GGCTCACAGTGATTTTCTGGCT
AAGAGCGACCCACACGTCAAA CTGCCCAGTTCGTTTCAGTGC
ACAAGCAGCTCCAAAGGCTACA GCTGGGTCACACTTCTCTGGT
TGGACCGCAACAACGCAAT ACTCAGGCGTATCAGTGGGG
CCCTGTGGAAGTTTGAGACC CTGCCCGTTCTTGGAGATAC
ATGGGAAGCTGGTCATCAAC GTGGTTCACACCCATCACAA
CM. Some cytokines such as brain-derived neurotrophic factor
(BDNF) and matrix metalloproteinase-13 (MMP-13) were present
in high concentrations in all CM without a dependency on culture
substrate. Interestingly, platelet derived growth factor (PDGF-AA)
and interleukin-2 (IL-2) were highly secreted only in TCP-CM,
whereas high concentrations of interleukin-13 (IL-13) was detec-
ted only in Hydrogel-CM. In stark contrast, relatively higher con-
centrations of most cytokines were detected in Scaffold-CM. These
included, among others, basic fibroblast growth factor (b-FGF),
interleukin-1 (IL-1), and transforming growth factor (TGF-b).
ELISA kits were used to more precisely quantify the secretion of
bioactive factors that are desirable during tissue regeneration
 T.H. Qazi et al. / Biomaterials 140 (2017) 103e114 107

Fig. 1. Schematic depiction and corresponding morphological evaluation of MSCs cultured in different substrate microenvironments: (a) TCP well plate, (b) encapsulated in an
alginate hydrogel, and (c) seeded on a macroporous alginate scaffold. MSCs were stained with DAPI (nuclei ¼ blue) and Phalloidin (F-Actin ¼ green) [Scale bar: a-b ¼ 100 mm,
c ¼ 50 mm]. (d) MSCs maintained long-term viability on the three substrates, as determined by LIVE/DEAD staining on days 1, 3, and 7. (e) Alginate hydrogels and scaffolds showed
similar mechanical properties, as determined by micro-indentation [n ¼ 15, two-tailed student's t-test]. (For interpretation of the references to colour in this figure legend, the
reader is referred to the web version of this article.)

Fig. 2. Substrate microenvironment alters MSC paracrine factor secretion. (a) Heatmap showing differences in the secretion profile of MSCs cultured on TCP (TCP-CM), in hydrogels
(Hydrogel-CM), and on scaffolds (Scaffold-CM) [mean intensities of n ¼ 4 biological replicates]. (b) Of the three substrates, Scaffold-CM contained significantly higher concentrations
of growth factors desirable for tissue regeneration such as VEGF, FGF2, IGF, HGF, and LIF [n ¼ 4 biological replicates, one-way ANOVA with Tukey's post-hoc test].

(Fig. 2 b). Vascular endothelial growth factor (VEGF) is a potent cells secreted significantly lower amounts. Concentrations of
stimulator of angiogenesis and is known to be secreted by many cell secreted basic fibroblast growth factor (FGF2), insulin-like growth
types including MSCs. We found that MSCs on TCP and scaffolds factor (IGF), hepatocyte growth factor (HGF), and leukemia inhibi-
secreted the highest concentrations of VEGF, whereas encapsulated tory factor (LIF) were significantly higher in Scaffold-CM compared
108 T.H. Qazi et al. / Biomaterials 140 (2017) 103e114
to the other groups. To exclude the possibility of secreted factors
getting entrapped within the nanoporous hydrogel matrix, mRNA
expression was performed for multiple cytokines, and growth
factor concentrations were determined after hydrogel lysis.
Together, these investigations revealed that MSCs in scaffolds show
a several fold higher mRNA expression of multiple cytokines
(Fig. S4a), and that insignificant amounts of cytokines were
detected after hydrogel and scaffold lysis (Fig. S4b). These results
indicate that biomaterial microenvironment can strongly modulate
MSC secretion, and that despite chemical and mechanical similar-
ities, 3D macroporous scaffolds promoted significantly enhanced
cytokine and growth factor secretion by MSCs compared to
hydrogels.
3.3. Secreted factors differentially influence myoblast functions
The impact of altered MSC secretion was investigated in the
context of muscle regeneration by analyzing the behavior and
function of C2C12 myoblasts in the presence of CM from different
substrates. Successful skeletal muscle regeneration after trauma
requires muscle progenitor cells to survive the harsh injury envi-
ronment, undergo proliferation, migrate towards the site of tissue
damage, and at later stages undergo myogenic differentiation to
form contractile fibers. Therefore, in vitro assays were used to study
how MSC derived paracrine factors could influence these functions
in myoblasts.
In a survival assay designed to evaluate the anti-apoptotic ef-
fects of MSC-CM, we found that Scaffold-CM promoted significantly
higher viability of C2C12 myoblasts in injury mimicking conditions,
compared to CM from other substrates (Fig. 3 a). In the absence of
CM, cell viability dropped to ~50%, which was the lowest among all
tested groups. Next, we tested the ability of myoblasts to proliferate
in serum free conditions over 48 h (Fig. 3 b). Despite the lack of
serum, which negatively affected cell proliferation (control),
serum-free CM stimulated cell growth. After 48 h, significantly
greater cell numbers were detected in the presence of Scaffold-CM
compared to CM from TCP and hydrogel substrates. The normalized
metabolic activity was similar in all groups after 24 h, but was
significantly higher in the scaffold and hydrogel CM groups after
48 h (Fig. 3 c).
A scratch wound healing assay was used to quantify the
migratory behavior of myoblasts exposed to CM from different
substrates. The analysis revealed that myoblasts exposed to
Scaffold-CM showed faster kinetics of scratch closure during the
experiment and had repopulated a significantly greater scratch area
at the end of the experiment (Fig. 3 def). No differences in repo-
pulated scratch area or the rates of scratch closure were observed
between TCP-CM and Hydrogel-CM. These readouts provide infor-
mation on collective cell migration. To investigate single cell
behavior, we performed single cell tracking and analyzed mean
migration velocity, total migration distance, displacement, and
directionality of a large number of individual myoblasts. The ana-
lyses revealed that Scaffold-CM stimulated myoblasts to migrate
faster (Fig. 3 g) and farther (Fig. 3 h,i) compared to TCP-CM and
Hydrogel-CM. The directionality remained unaffected by the CM
(Fig. 3 j).
Next, the influence of MSC secreted factors on myoblast fusion
and in vitro myotube formation was studied (Fig. 4). Under differ-
entiation inducing conditions (low serum concentrations), myo-
blasts fuse together to form long, multinucleated myotubes and
express differentiation markers such as myosin heavy chain (MHC)
and myogenin (MyoG). Differentiation was first assessed by
immunofluorescent staining of MHC in myoblasts cultured in CM
supplemented with 5% serum (Fig. 4 a). Over the same time dura-
tion, myotube formation in the presence of TCP-CM and Hydrogel-
CM was comparable to the positive control (non-conditioned
medium þ 5% serum), whereas Scaffold-CM inhibited myotube
formation. Quantification of the acquired images revealed signifi-
cantly lower mean myotube lengths and densities in the Scaffold-
CM group compared to all others (Fig. 4 b,c). These results were
confirmed by Western blot analysis (Fig. 4 d). The normalized
protein expression levels of MHC and MyoG were significantly
lower in Scaffold-CM myoblasts cultures compared to all other
groups (Fig. 4 e,f). In summary, our in vitro data provide evidence
that the ability of MSCs to modulate myoblasts functions via
paracrine mechanisms may strongly depend on the biomaterial
microenvironment.
3.4. N-cadherin mediated cell-cell interactions enhance MSC
paracrine effects
We tested the hypothesis that enhanced cell-cell interactions
facilitated by the interconnected porous structure of scaffolds
played a causal role in the observed modulation of secretion profile
and paracrine effects of MSCs. Cadherins are a family of cell mem-
brane proteins that promote cell-cell adhesion. E-Cadherin and N-
Cadherin in particular have been implicated in the formation of
multicellular spheroids that lead to enhanced secretion by MSCs.
Therefore, we first performed immunofluorescent staining on
hydrogel encapsulated and scaffold seeded MSCs with an anti-N-
Cadherin antibody. Whereas only a proportion of isolated MSCs in
hydrogels expressed N-Cadherin, a much more intense staining was
observed for MSCs cultured in scaffolds (Fig. 5 a). This difference was
confirmed by performing qPCR analysis, which revealed that scaf-
fold seeded MSCs showed an almost 10-fold higher expression of
the N-Cadherin gene CDH2 relative to encapsulated MSCs (Fig. 5 b).
To determine if selectively blocking N-Cadherin in MSCs would alter
their secretion profile, MSCs were treated with an N-Cadherin
blocking antibody (þnAb) prior to encapsulation or seeding on the
3D substrates (Fig. S1). Conditioned media from untreated and nAb
treated MSCs were collected and characterized using representative
cytokine arrays. The fold difference in intensities with nAb treat-
ment relative to untreated MSCs was quantified (Fig. 5 c). The results
showed that blocking N-Cadherin in scaffold seeded MSCs down-
regulated the secretion of all cytokines on the array (red striped
bars). However, blocking N-Cadherin in hydrogel encapsulated
MSCs led to a downregulation in a fraction of the cytokines (green
striped bars), and fold-changes in secretion were in general, less
pronounced compared to the scaffold group.
Next, we investigated the effects of CM from nAb treated MSCs
on myoblast proliferation, migration, and differentiation. Interest-
ingly, while after 24 h a significantly lower number of proliferated
cells was observed in Hydrogel-CM þ nAb compared to Hydrogel-
CM and no difference was observed in the scaffold groups (þ/
nAb), the proliferation of myoblasts after 48 h in the presence of
CM þ nAb dropped significantly in the scaffold group, but was not
significantly affected in the hydrogel group (Fig. 5 d). In terms of
collective cell migration, the kinetics of scratch closure and the
repopulated scratch area were significantly reduced by N-Cadherin
blocking in the scaffold group, but remained unaffected in Hydro-
gel-CM þ nAb cultures (Fig. 5 eeg). Analysis of single cell tracking
during the scratch migration assay revealed that N-Cadherin
blocking reverses the positive effects of Scaffold-CM on myoblasts'
migration velocity and their total traversed distance to the values of
Hydrogel-CM, which remained unaffected by nAb treatment (Fig. 5
h,i). The displacement was more strongly affected in Scaffold-
CM þ nAb, whereas directionality was adversely influenced in both
substrate groups (Fig. 5 j,k). The presence of nAb in control media
alone did not significantly affect myoblast function (Fig. S5).
Multinucleated myotube formation by C2C12 myoblasts in the
 T.H. Qazi et al. / Biomaterials 140 (2017) 103e114 109

Fig. 3. Substrate microenvironment dependent MSC paracrine effects on myoblast function. (a) Myoblast survival in injury mimicking conditions [n ¼ 8, one-way ANOVA with
Tukey's post-hoc test]. Assessment of myoblast (b) proliferation, and (c) metabolic activity over 48 h [n ¼ 12, one-way ANOVA with Tukey's post-hoc test]. Migration behavior of
myoblasts was analyzed in a scratch wound healing assay; graphs show (d) repopulated scratch area after 20 h, and (e) the corresponding kinetics of scratch closure [n ¼ 4 biological
replicates, one-way ANOVA with Tukey's post-hoc test]. (f) Representative images of myoblasts migrating into empty scratch area; dashed lines indicate scratch contours at t ¼ 0.
Single cell tracking provided insights into (g) migration velocity, (h) total migrated distance, (i) displacement, and (j) directionality of myoblasts [n ¼ 68e120, Kruskal-Wallis test
with Dunn's multiple comparison test].

presence of Hydrogel-CM (þ/-nAb) was similar to the positive
control (differentiation media). The inhibition of myotube forma-
tion observed in the presence of Scaffold-CM group was apparently
reversed in the presence of Scaffold-CM þ nAb, although this did
not reach statistical significance (p ¼ 0.06) (Fig. 5 l,m).
Together, these results provide evidence to support the hy-
pothesis that differences in N-Cadherin mediated cell-cell in-
teractions between hydrogel encapsulated and scaffold seeded
MSCs regulates the paracrine effects on myogenic progenitor cells.
4. Discussion
A significant body of work has demonstrated that biomaterial
properties can be used to control MSC differentiation [39,40]. In a
similar manner, gaining the ability to control or enhance the
paracrine properties of these cells using cues presented from bio-
materials can be a promising strategy to improve the therapeutic
outcomes of MSC based cell therapies.
Although it is known that dynamic, 3D culture significantly al-
ters the secretion profile of MSCs [41,42], the vast majority of
studies investigating MSC secretome have been carried out with
conditioned media from cells seeded on tissue culture plastic
substrates under static conditions [43]. A discrepancy may there-
fore arise between the secretion profile characterized in vitro on
non-physiologically stiff 2D culture substrates and what the MSCs
may secrete in vivo. This is especially plausible because MSCs are
110 T.H. Qazi et al. / Biomaterials 140 (2017) 103e114

Fig. 4. Substrate microenvironment dependent MSC paracrine effects on myoblast differentiation. (a) Representative immunofluorescent images of differentiated myoblasts stained
with DAPI (nuclei ¼ blue) and the differentiation marker Myosin heavy chain (MHC ¼ green) [Scale bar ¼ 100 mm]. Images from several regions of interest were used to determine
(b) mean myotube length [n ¼ 60, Kruskal-Wallis test with Dunn's multiple comparison test], and (c) mean myotube density [n ¼ 5, one-way ANOVA with Tukey's post-hoc test]. (d)
Western blot analysis was carried out to quantify expression of the differentiation markers MHC and Myogenin; protein loading was determined by the detection of GAPDH. Blots
were quantified using densitometry and corresponding graphs show expression of (e) MHC, and (f) Myogenin [n ¼ 3 biological replicates, one-way ANOVA with Tukey's post-hoc
test]. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

known to be responsive to environmental stimuli [44], and could
encounter a wide spectrum of in vivo environments with different
stiffness (e.g. fibrotic tissues), chemical composition (e.g. variation
in ECM composition), and shear stresses (e.g. after systemic de-
livery). Similarly, cells may encounter unique 3D microenviron-
ments in biomaterials used for transplantation, which may affect
their function [45,46]. In agreement with this line of thought, our
results show a remarkable variation between the secretion profile
of TCP cultured and 3D substrate cultured MSCs.
Previous studies carried out in 3D substrates report that the
secretion behavior of MSCs can vary substantially with changes in
the substrate's chemical (conjugated peptides, material composi-
tion) and mechanical (stiffness) properties [47e49]. Therefore in
this study, similar ionic crosslinker concentrations were used with
a common batch of RGD modified alginate to yield structurally
different 3D substrates with similar chemical and mechanical
properties. Although the freezing and subsequent lyophilization
processes employed to produce macroporous scaffolds may lead to
a variability of local stiffness values (Fig. S2), our data indicate that
this did not lead to significant differences compared to hydrogels.
High concentrations of secreted VEGF were detected in TCP-CM,
which is an expected outcome as some studies have correlated
VEGF secretion with increased substrate stiffness [46]. However,
comparably high concentrations of VEGF were also detected in
Scaffold-CM where the MSCs were cultured on relatively compliant
(~20 kPa) substrates. Other factors including HGF, FGF, IGF, and LIF
were highly secreted by scaffold seeded MSCs compared to corre-
sponding cultures on TCP and in hydrogels. These growth factors
may especially be desirable for skeletal muscle regeneration where
their utility is well documented [50,51]. For example, LIF is known
for its role in maintaining the pluripotency of embryonic stem cells,
but has also been implicated in the modulation of myoblast pro-
liferation and differentiation [52]. Similarly, IGF is a potent
myogenic factor that stimulates satellite cells, promotes their
myogenic differentiation, increases muscle mass and strength, and
has been implicated in the timely onset of the anti-inflammatory
phase after injury [53]. Although we confirmed that secreted sol-
uble factors were not entrapped inside hydrogels in significant
concentrations (Fig. S4), the diffusivity of exosomes or other cell-
derived vesicles (in the nm size range) through the nanoporous
hydrogels has not been investigated, and may potentially be
affected by the pore size of the encapsulating hydrogel. However,
this seems not to have influenced the paracrine effects of MSCs on
myoblast function, as TCP-CM and Hydrogel-CM induce similar
migratory, proliferative, and differentiation behavior in C2C12
myoblasts. Diffusion of cell derived vesicles through matrices of
various pore sizes could be a point of investigation in future studies.
The secreted factors are likely to synergistically contribute
 T.H. Qazi et al. / Biomaterials 140 (2017) 103e114 111

Fig. 5. The expression of cell-cell adhesion protein N-Cadherin by MSCs is regulated by 3D substrate microenvironment and plays a key role in MSC paracrine effects on myoblasts.
(a) Representative immunofluorescent images of hydrogel-encapsulated and scaffold-seeded MSCs stained with DAPI (nuclei ¼ blue) and anti-N Cadherin antibody (N-
cadherin ¼ red) [Scale bar: 50 mm]. (b) MSCs in scaffolds showed a significantly higher expression of the N-cadherin gene CDH2, as determined by qPCR [n ¼ 3, two-tailed student's
t-test]. (c) Functional blocking of N-Cadherin in MSCs using a neutralization antibody (nAb) adversely affected the secretion profile of MSCs [CM pooled from n ¼ 3 biological
replicates]. Paracrine effects of nAb-treated and untreated MSCs on myoblast function were compared using migration, proliferation, and differentiation assays. (d) Myoblast
proliferation in the presence of CM from 3D cultured MSCs with (striped bars) or without (solid bars) N-cadherin blocking [n ¼ 4, one-way ANOVA with Tukey's post-hoc test]. (eek)
Myoblast migration was negatively influenced by N-Cadherin blocking of scaffold-seeded MSCs (Scaffold-CM vs. Scaffold-CM þ nAb) but remained unaffected by N-Cadheirn
blocking of hydrogel-encapsulated MSCs (Hydrogel-CM vs. Hydrogel-CM þ nAb). (eeg) Repopulated scratch area and corresponding kinetics of scratch closure [n ¼ 4, one-way
ANOVA with Tukey's post-hoc test]. (hek) Single cell tracking readouts [n ¼ 90e120, one-way ANOVA with Dunn's post-hoc test]. The effects of CM from anti-N cadherin anti-
body treated MSCs on myoblast differentiation was assessed by quantifying myoblast fusion index. (l) Representative immunofluorescent images of multinucleated myotubes
stained with myosin heavy chain (MHC) and DAPI [Scale bar: 100 mm], and (m) the corresponding fusion index values [n ¼ 8, one-way ANOVA with Tukey's post-hoc test]. (For
interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

towards tissue repair and regeneration by modulating the function associated with the long-term ex-vivo culture of satellite cells [56].
of endogenous progenitor cells such as satellite cells [54,55]. In this Moreover, because they are widely used in basic research, C2C12
study, we used C2C12 myoblasts as a representative myogenic myoblasts allow us to compare observations and results with those
progenitor cell primarily due to the unresolved challenges reported in literature [57,58]. Successful skeletal muscle
112 T.H. Qazi et al. / Biomaterials 140 (2017) 103e114
regeneration is characterized by efficient migration, proliferation,
and fusion of satellite cells and their myoblast progenitors to form
new muscle fibers [59,60]. Our results indicate that MSCs can have
wide-ranging paracrine effects on various aspects of myoblast
function. Compared to other CM groups, Scaffold-CM stimulated
greater proliferation, survival, metabolic activity, and migration of
myoblasts. Fusion of individual progenitor myoblasts into elon-
gated, multinucleated myotubes marks the beginning of a process
that leads to the formation of new myofibers [61]. Factors such as
FGF, HGF, and VEGF have been implicated in the regulation of
myoblast proliferation and terminal differentiation [62]. Our
observation that Scaffold-CM inhibited myotube formation may
likely be due to the presence of high concentrations of several
growth factors that regulate myoblast cell-cycle withdrawal [63]. In
the wake of muscle injury, the primary requirements for regener-
ation involve survival of myoblasts, their migration and sufficient
proliferation at the site of injury. Once an adequate number of cells
are present, differentiation is initiated by the fusion of myoblasts
which leads to the formation of contractile myofibers. Accordingly,
previous studies have reported that delayed differentiation in
conjunction with increased proliferation and migration of myogenic
progenitor cells leads to in vivo muscle regeneration [64].
We propose that the strong paracrine effects of scaffold seeded
MSCs may be attributed to cell-cell interactions that are supported
by the macroporous structure of the scaffold, but inhibited by the
nanoporous structure of the hydrogel (Fig. S6). In mesenchymal
cells, cell-cell interactions via cadherins are developmentally rele-
vant because their condensation leads to the onset of endochondral
ossification [65]. In a biomaterials based regeneration context, Bian
and colleagues have shown that recapitulating mesenchymal
condensation in hydrogel encapsulated MSCs by using matrix
presented N-cadherin mimicking peptide fragments can enhance
chondrogenesis [66]. Further, they demonstrated that blocking N-
cadherin with an antibody similar to the one used in this study
abrogated the improved chondrogenic effects. In terms of paracrine
effects, researchers have reported cadherin mediated spheroid/
aggregate formation that show enhanced secretion of several cy-
tokines and growth factors [67]. Although we did not observe
spontaneous spheroid formation in scaffolds seeded with MSCs,
physical interaction between the cells was qualitatively evident.
Using qPCR and immunofluorescence staining, we confirmed that
scaffold seeded MSCs showed a higher expression of N-cadherin.
Next, we asked whether blocking N-cadherin would negatively
affect the secretion profile of scaffold seeded MSCs, and whether
the paracrine effects on myoblasts would only be as effective as
those of Hydrogel-CM. Indeed, using a neutralization antibody led
to a downregulation of all detected cytokines in scaffold seeded
MSCs, but the effects were not as strongly apparent for hydrogel
encapsulated MSCs. The substantially better paracrine effects of
Scaffold-CM on myoblast migration were abrogated to levels that
were comparable to Hydrogel-CM. It is unclear why myoblast
proliferation was reduced in Hydrogel-CM þ nAb, but we speculate
that this might be due to a differential modulation in the secretion
of factors that are primary stimulators of proliferation and migra-
tion. Furthermore, an insignificant yet apparent reversal of the in-
hibition of myotube formation was observed in Scaffold-CM þ nAb.
Whether the differences in the ability of the two CM groups to
support (Hydrogel-CM) or inhibit (Scaffold-CM) myogenic differ-
entiation and myotube formation has a tangible impact on the
regeneration of severely injured skeletal muscle should be a point
of investigation in future studies.
Cell encapsulating hydrogels have been suggested to act as a
protective barrier against the potentially harmful effects of in-
flammatory cytokines that transplanted cells may encounter at the
site of injury [68,69]. While a change in encapsulated cell
morphology does not seem to be required for MSC differentiation
[70,71], our results show that hydrogels that prevent cell-cell in-
teractions and/or changes in morphology leading to cell-cell in-
teractions may alter or suppress the paracrine properties of MSCs,
indicating that this may be a potentially important variable to
examine in different MSC-based applications. This does not imply
that hydrogels in general would suppress the therapeutic potential
of MSCs. Optimization of mechanical and physicochemical prop-
erties of the encapsulating biomaterial may alter MSC function.
Some work in this area was recently described by Cai et al. who
showed that engineered hydrogels with a stiffness range of
200e400 Pa enhance the paracrine secretion of pro-angiogenic
factors from adipose derived MSCs which were able to spread out
and proliferate within the gel, possibly also leading to cell-cell
contacts [72]. Chaudhuri et al. reported that tuning the viscoelas-
ticity of alginate hydrogels promoted dramatic changes in encap-
sulated cell morphology and proliferation behavior which
regulated stem cell fate [73]. Similarly, Follin et al. encapsulated
adipose derived MSCs in alginates with very low molecular weights
that permitted cell spreading, and led to increased paracrine ac-
tivity [74]. Together, these examples indicate that cell-cell in-
teractions, and morphological changes that may lead to them, could
be promoted in engineered hydrogel systems while also retaining
their injectability.
Taken together, the results described in this study provide
strong support for the use of macroporous scaffolds to promote
cell-cell interactions and thereby enhance the paracrine function of
transplanted MSCs. While previously, most work on cell-cell in-
teractions and cytokine secretion has utilized self-organizing
spheroid cultures that require minimal matrix contact, our work
opens new avenues for exploring the synergistic effects of cell-cell
and cell-matrix interactions. Future studies could potentially
investigate the influence of various physical properties of the
biomaterial matrix on the ability of MSCs to form cell-cell contacts.
Additionally, the impact of cell-cell interactions on the immuno-
modulatory function of MSCs should be investigated before vali-
dating these in vitro findings in a relevant in vivo tissue injury
model.
5. Conclusions
Paracrine mechanisms allow MSCs to orchestrate a regenerative
response by instructing desirable functions in progenitor cell
populations. That the ability of MSCs to accomplish this effect can
be substantially improved using biomaterial properties such as
porosity is an exciting avenue to improve the outcomes of MSC
based cell therapies. The work reported here provides new insights
into the role of a biomaterial's structural microenvironment in
promoting cell-cell interactions and thereby enhancing the para-
crine effects of MSCs. These findings are likely to be relevant in
designing biomaterial based stem cell therapies for a wide range of
tissue engineering applications.
Acknowledgements
The authors acknowledge Dag Wulsten for assistance with
nanoindentation experiments. This work was supported by grants
from the BCRT and BSRT through funding by the German Federal
Ministry of Education and Research (BMBF) and the German
Research Foundation, respectively. T.H.Q acknowledges funding
from the Friede Springer Stiftung and the BSRT. D.J.M was sup-
ported by the Einstein fellowship award. G.N.D and S.G were
partially supported by the BMBF (DIMEOS, 01EC1402B) and the
DFG Research Unit (FOR 2165; grant GE2512/2-1). The funders had
no role in study design, data collection and analysis, decision to
 T.H. Qazi et al. / Biomaterials 140 (2017) 103e114
publish, or preparation of the manuscript.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.biomaterials.2017.06.019.
References
[1] A.F. Steinert, L. Rackwitz, F. Gilbert, U. Noth, R.S. Tuan, Concise review: the
clinical application of mesenchymal stem cells for musculoskeletal regener-
ation: current status and perspectives, Stem Cells Transl. Med. 1 (2012)
237e247.
[2] T.H. Qazi, D.J. Mooney, M. Pumberger, S. Geißler, G.N. Duda, Biomaterials
based strategies for skeletal muscle tissue engineering: existing technologies
and future trends, Biomaterials 53 (2015) 502e521.
[3] J. Bashir, A. Sherman, H. Lee, L. Kaplan, J.M. Hare, Mesenchymal stem cell
therapies in the treatment of musculoskeletal diseases, PM R J. Inj. Funct.
Rehabilation 6 (2014) 61e69.
[4] A. Trounson, C. McDonald, Stem cell therapies in clinical trials: progress and
challenges, Cell Stem Cell. 17 (2015) 11e22.
[5] M.F. Pittenger, A.M. Mackay, S.C. Beck, R.K. Jaiswal, R. Douglas, J.D. Mosca, et
al., Multilineage potential of adult human mesenchymal stem cells, Science
284 (1999) 143e147.
[6] A.I. Caplan, D. Correa, The MSC: an injury drugstore, Cell Stem Cell. 9 (2011)
11e15.
[7] P.R. Baraniak, T.C. McDevitt, Stem cell paracrine actions and tissue regenera-
tion, Regen. Med. 5 (2010) 121e143.
[8] S.H. Ranganath, O. Levy, M.S. Inamdar, J.M. Karp, Harnessing the mesenchymal
stem cell secretome for the treatment of cardiovascular disease, Cell Stem Cell.
10 (2012) 244e258.
[9] A.J. Salgado, R.L. Reis, N.J. Sousa, J.M. Gimble, Adipose tissue derived stem cells
secretome: soluble factors and their roles in regenerative medicine, Curr.
Stem Cell Res. Ther. 5 (2010) 103e110.
[10] F.G. Teixeira, M.M. Carvalho, N. Sousa, A.J. Salgado, Mesenchymal stem cells
secretome: a new paradigm for central nervous system regeneration? Cell.
Mol. Life Sci. 70 (2013) 3871e3882.
[11] S. Meirelles Lda, A.M. Fontes, D.T. Covas, A.I. Caplan, Mechanisms involved in
the therapeutic properties of mesenchymal stem cells, Cytokine Growth
Factor Rev. 20 (2009) 419e427.
[12] H.M. Wobma, D. Liu, G. Vunjak-Novakovic, Paracrine Effects of Mesenchymal
Stromal Cells Cultured in Three-dimensional Settings on Tissue Repair, ACS
Biomaterials Science & Engineering, 2017.
[13] C.P. Chang, C.C. Chio, C.U. Cheong, C.M. Chao, B.C. Cheng, M.T. Lin, Hypoxic
preconditioning enhances the therapeutic potential of the secretome from
cultured human mesenchymal stem cells in experimental traumatic brain
injury, Clin. Sci. (London, England : 1979) 124 (2013) 165e176.
[14] A. Saparov, V. Ogay, T. Nurgozhin, M. Jumabay, W.C.W. Chen, Preconditioning
of human mesenchymal stem cells to enhance their regulation of the immune
response, Stem Cells Int. 2016 (2016) 10.
[15] M. Pumberger, T.H. Qazi, M.C. Ehrentraut, M. Textor, J. Kueper, G. Stoltenburg-
Didinger, et al., Synthetic niche to modulate regenerative potential of MSCs
and enhance skeletal muscle regeneration, Biomaterials 99 (2016) 95e108.
[16] E.J. Lee, S.J. Park, S.K. Kang, G.-H. Kim, H.-J. Kang, S.-W. Lee, et al., Spherical
bullet formation via e-cadherin promotes therapeutic potency of mesen-
chymal stem cells derived from human umbilical cord blood for myocardial
infarction, Mol. Ther. 20 (2012) 1424e1433.
[17] C.L. Rettinger, A.B. Fourcaudot, S.J. Hong, T.A. Mustoe, R.G. Hale, K.P. Leung,
In vitro characterization of scaffold-free three-dimensional mesenchymal
stem cell aggregates, Cell Tissue Res. 358 (2014) 395e405.
[18] J.M. Santos, S.P. Camoes, E. Filipe, M. Cipriano, R.N. Barcia, M. Filipe, et al.,
Three-dimensional spheroid cell culture of umbilical cord tissue-derived
mesenchymal stromal cells leads to enhanced paracrine induction of wound
healing, Stem Cell Res. Ther. 6 (2015) 90.
[19] S.S. Ho, K.C. Murphy, B.Y. Binder, C.B. Vissers, J.K. Leach, Increased survival and
function of mesenchymal stem cell spheroids entrapped in instructive algi-
nate hydrogels, Stem Cells Transl. Med. 5 (2016) 773e781.
[20] M. Osugi, W. Katagiri, R. Yoshimi, T. Inukai, H. Hibi, M. Ueda, Conditioned
media from mesenchymal stem cells enhanced bone regeneration in rat cal-
varial bone defects, Tissue Eng. Part A 18 (2012) 1479e1489.
[21] D. Cantinieaux, R. Quertainmont, S. Blacher, L. Rossi, T. Wanet, A. Noel, et al.,
Conditioned medium from bone marrow-derived mesenchymal stem cells
improves recovery after spinal cord injury in rats: an original strategy to avoid
cell transplantation, PLoS One 8 (2013) e69515.
[22] L. Timmers, S.K. Lim, I.E. Hoefer, F. Arslan, R.C. Lai, A.A.M. van Oorschot, et al.,
Human mesenchymal stem cell-conditioned medium improves cardiac
function following myocardial infarction, Stem Cell Res. 6 (2011) 206e214.
[23] F. Wang, T. Yasuhara, T. Shingo, M. Kameda, N. Tajiri, W.J. Yuan, et al., Intra-
venous administration of mesenchymal stem cells exerts therapeutic effects
on parkinsonian model of rats: focusing on neuroprotective effects of stromal
cell-derived factor-1alpha, BMC Neurosci. 11 (2010) 52.
[24] M.F. Berry, A.J. Engler, Y.J. Woo, T.J. Pirolli, L.T. Bish, V. Jayasankar, et al.,
113
Mesenchymal stem cell injection after myocardial infarction improves
myocardial compliance, Am. J. Physiol. Heart Circ. Physiol. 290 (2006)
H2196eH2203.
[25] Y. Fan, M. Maley, M. Beilharz, M. Grounds, Rapid death of injected myoblasts
in myoblast transfer therapy, Muscle & Nerve 19 (1996) 853e860.
[26] U.M. Fischer, M.T. Harting, F. Jimenez, W.O. Monzon-Posadas, H. Xue,
S.I. Savitz, et al., Pulmonary passage is a major obstacle for intravenous stem
cell delivery: the pulmonary first-pass effect, Stem Cells Dev. 18 (2009)
683e692.
[27] S. Lee, E. Choi, M.-J. Cha, K.-C. Hwang, Cell adhesion and long-term survival of
transplanted mesenchymal stem cells: a prerequisite for cell therapy,
Oxidative Med. Cell. Longev. 2015 (2015) 9.
[28] C.A. Cezar, D.J. Mooney, Biomaterial-based delivery for skeletal muscle repair,
Adv. drug Deliv. Rev. 84 (2015) 188e197.
[29] S.J. Bidarra, C.C. Barrias, P.L. Granja, Injectable alginate hydrogels for cell de-
livery in tissue engineering, Acta biomater. 10 (2014) 1646e1662.
[30] L.M. Marquardt, S.C. Heilshorn, Design of injectable materials to improve stem
cell transplantation, Curr. Stem Cell Rep. 2 (2016) 207e220.
[31] J.A. Hunt, R. Chen, T. van Veen, N. Bryan, Hydrogels for tissue engineering and
regenerative medicine, J. Mater. Chem. B 2 (2014) 5319e5338.
[32] N. Annabi, J.W. Nichol, X. Zhong, C. Ji, S. Koshy, A. Khademhosseini, et al.,
Controlling the porosity and microarchitecture of hydrogels for tissue engi-
neering, Tissue Eng. Part B Rev. 16 (2010) 371e383.
[33] T. Boontheekul, H.-J. Kong, D.J. Mooney, Controlling alginate gel degradation
utilizing partial oxidation and bimodal molecular weight distribution, Bio-
materials 26 (2005) 2455e2465.
[34] A. Leal-Egana, U.D. Braumann, A. Diaz-Cuenca, M. Nowicki, A. Bader, Deter-
mination of pore size distribution at the cell-hydrogel interface,
J. Nanobiotechnol. 9 (2011) 24.
[35] J.A. Rowley, G. Madlambayan, D.J. Mooney, Alginate hydrogels as synthetic
extracellular matrix materials, Biomaterials 20 (1999) 45e53.
[36] B.D. Cosgrove, K.L. Mui, T.P. Driscoll, S.R. Caliari, K.D. Mehta, R.K. Assoian, et al.,
N-cadherin adhesive interactions modulate matrix mechanosensing and fate
commitment of mesenchymal stem cells, Nat. Mater. 15 (2016) 1297e1306.
[37] E.J. Lee, E.K. Choi, S.K. Kang, G.H. Kim, J.Y. Park, H.J. Kang, et al., N-cadherin
determines individual variations in the therapeutic efficacy of human um-
bilical cord blood-derived mesenchymal stem cells in a rat model of
myocardial infarction, Mol. Ther. J. Am. Soc. Gene Ther. 20 (2012) 155e167.
[38] T. Geback, M.M. Schulz, P. Koumoutsakos, M. Detmar, TScratch: a novel and
simple software tool for automated analysis of monolayer wound healing
assays, BioTechniques 46 (2009) 265e274.
[39] S.W. Lane, D.A. Williams, F.M. Watt, Modulating the stem cell niche for tissue
regeneration, Nat. Biotechnol. 32 (2014) 795e803.
[40] H.J. Anderson, J.K. Sahoo, R.V. Ulijn, M.J. Dalby, Mesenchymal stem cell fate:
applying biomaterials for control of stem cell behavior, Front. Bioeng. Bio-
technol. 4 (2016) 38.
[41] F.G. Teixeira, K.M. Panchalingam, R. Assuncao-Silva, S.C. Serra, B. Mendes-
Pinheiro, P. Patricio, et al., Modulation of the mesenchymal stem cell secre-
tome using computer-controlled bioreactors: impact on neuronal cell prolif-
eration, survival and differentiation, Sci. Rep. 6 (2016) 27791.
[42] J.D. Glaeser, S. Geissler, A. Ode, C.J. Schipp, G. Matziolis, W.R. Taylor, et al.,
Modulation of matrix metalloprotease-2 levels by mechanical loading of
three-dimensional mesenchymal stem cell constructs: impact on in vitro tube
formation, Tissue Eng. Part A 16 (2010) 3139e3148.
[43] T. Schinkothe, W. Bloch, A. Schmidt, In vitro secreting profile of human
mesenchymal stem cells, Stem Cells Dev. 17 (2008) 199e206.
[44] M.B. Murphy, K. Moncivais, A.I. Caplan, Mesenchymal stem cells: environ-
mentally responsive therapeutics for regenerative medicine, Exp. Mol. Med.
45 (2013) e54.
[45] C. Fischbach, H.J. Kong, S.X. Hsiong, M.B. Evangelista, W. Yuen, D.J. Mooney,
Cancer cell angiogenic capability is regulated by 3D culture and integrin
engagement, Proc. Natl. Acad. Sci. U. S. A. 106 (2009) 399e404.
[46] A.A. Abdeen, J.B. Weiss, J. Lee, K.A. Kilian, Matrix composition and mechanics
direct proangiogenic signaling from mesenchymal stem cells, Tissue Eng. Part
A 20 (2014) 2737e2745.
[47] R.A. Marklein, D.E. Soranno, J.A. Burdick, Magnitude and presentation of
mechanical signals influence adult stem cell behavior in 3-dimensional
macroporous hydrogels, Soft Matter 8 (2012) 8113e8120.
[48] D. Thomas, G. Fontana, X. Chen, C. Sanz-Nogue
s, D.I. Zeugolis, P. Dockery, et
al., A shape-controlled tuneable microgel platform to modulate angiogenic
paracrine responses in stem cells, Biomaterials 35 (2014) 8757e8766.
[49] S. Jose, M.L. Hughbanks, B.Y. Binder, G.C. Ingavle, J.K. Leach, Enhanced trophic
factor secretion by mesenchymal stem/stromal cells with Glycine-Histidine-
Lysine (GHK)-modified alginate hydrogels, Acta biomater. 10 (2014)
1955e1964.
[50] C. Borselli, H. Storrie, F. Benesch-Lee, D. Shvartsman, C. Cezar, J.W. Lichtman,
et al., Functional muscle regeneration with combined delivery of angiogenesis
and myogenesis factors, Proc. Natl. Acad. Sci. 107 (2010) 3287e3292.
[51] J. Menetrey, C. Kasemkijwattana, C.S. Day, P. Bosch, M. Vogt, F.H. Fu, et al.,
Growth factors improve muscle healing in vivo, British Volume, J. Bone Jt.
Surg. 82-B (2000) 131e137.
[52] L.C. Hunt, J. White, The role of leukemia inhibitory factor receptor signaling in
skeletal muscle growth, injury and disease, Adv. Exp. Med. Biol. 900 (2016)
45e59.
[53] Y.H. Song, J.L. Song, P. Delafontaine, M.P. Godard, The therapeutic potential of
114 T.H. Qazi et al. / Biomaterials 140 (2017) 103e114
IGF-I in skeletal muscle repair, Trends Endocrinol. Metabol. 24 (2013)
310e319.
[54] Fry CS, Kirby TJ, Kosmac K, McCarthy JJ, Peterson CA. Myogenic progenitor
cells control extracellular matrix production by fibroblasts during skeletal
muscle hypertrophy. Cell Stem Cell.20:56e69.
[55] S. Hauerslev, J. Vissing, T.O. Krag, Muscle atrophy reversed by growth factor
activation of satellite cells in a mouse muscle atrophy model, PLoS One 9
(2014), e100594.
[56] B.D. Cosgrove, A. Sacco, P.M. Gilbert, H.M. Blau, A home away from home:
challenges and opportunities in engineering in vitro muscle satellite cell
niches, Differ. Res. Biol. Divers. 78 (2009) 185e194.
[57] A.J. Engler, M.A. Griffin, S. Sen, C.G. Bo €nnemann, H.L. Sweeney, D.E. Discher,
Myotubes differentiate optimally on substrates with tissue-like stiffness,
Pathol. Implic. Soft Stiff Microenviron. 166 (2004) 877e887.
[58] C. Sassoli, A. Pini, F. Chellini, B. Mazzanti, S. Nistri, D. Nosi, et al., Bone marrow
mesenchymal stromal cells stimulate skeletal myoblast proliferation through
the paracrine release of VEGF, PLoS One 7 (2012), e37512.
[59] C.A. Griffin, K.A. Kafadar, G.K. Pavlath, MOR23 promotes muscle regeneration
and regulates cell adhesion and migration, Dev. Cell. 17 (2009) 649e661.
[60] F. Alwes, C. Enjolras, M. Averof, Live imaging reveals the progenitors and cell
dynamics of limb regeneration, eLife 5 (2016), e19766.
[61] S.M. Abmayr, G.K. Pavlath, Myoblast fusion: lessons from flies and mice,
Development 139 (2012) 641e656.
[62] T.A. Linkhart, C.H. Clegg, S.D. Hauschka, Control of mouse myoblast commit-
ment to terminal differentiation by mitogens, J. Supramol. Struct. 14 (1980)
483e498.
[63] C.H. Clegg, T.A. Linkhart, B.B. Olwin, S.D. Hauschka, Growth factor control of
skeletal muscle differentiation: commitment to terminal differentiation oc-
curs in G1 phase and is repressed by fibroblast growth factor, J. Cell Biol. 105
(1987) 949e956.
[64] M. Bencze, E. Negroni, D. Vallese, H. Yacoub-Youssef, S. Chaouch, A. Wolff, et
al., Proinflammatory macrophages enhance the regenerative capacity of hu-
man myoblasts by modifying their kinetics of proliferation and differentiation,
Mol. Ther. 20 (2012) 2168e2179.
[65] F. Long, D.M. Ornitz, Development of the endochondral skeleton, Cold Spring
Harb. Perspect. Biol. (2013) 5.
[66] L. Bian, M. Guvendiren, R.L. Mauck, J.A. Burdick, Hydrogels that mimic
developmentally relevant matrix and N-cadherin interactions enhance MSC
chondrogenesis, Proc. Natl. Acad. Sci. 110 (2013) 10117e10122.
[67] T.J. Bartosh, J.H. Ylostalo, A. Mohammadipoor, N. Bazhanov, K. Coble,
K. Claypool, et al., Aggregation of human mesenchymal stromal cells (MSCs)
into 3D spheroids enhances their antiinflammatory properties, Proc. Natl.
Acad. Sci. U. S. A. 107 (2010) 13724e13729.
[68] A. Moshaverinia, C. Chen, X. Xu, S. Ansari, H.H. Zadeh, S.R. Schricker, et al.,
Regulation of the stem cellehost immune system interplay using hydrogel
coencapsulation system with an anti-inflammatory drug, Adv. Funct. Mater.
25 (2015) 2296e2307.
[69] A. Burdick Jason, L. Mauck Robert, S. Gerecht, To serve and protect: hydrogels
to improve stem cell-based therapies, Cell Stem Cell. 18 (2016) 13e15.
[70] N. Huebsch, P.R. Arany, A.S. Mao, D. Shvartsman, O.A. Ali, S.A. Bencherif, et al.,
Harnessing traction-mediated manipulation of the cell/matrix interface to
control stem-cell fate, Nat. Mater. 9 (2010) 518e526.
[71] S. Khetan, M. Guvendiren, W.R. Legant, D.M. Cohen, C.S. Chen, J.A. Burdick,
Degradation-mediated cellular traction directs stem cell fate in covalently
crosslinked three-dimensional hydrogels, Nat. Mater. 12 (2013) 458e465.
[72] L. Cai, R.E. Dewi, A.B. Goldstone, J.E. Cohen, A.N. Steele, Y.J. Woo, et al.,
Regulating stem cell secretome using injectable hydrogels with in situ
network formation, Adv. Healthc. Mater. 5 (2016) 2758e2764.
[73] O. Chaudhuri, L. Gu, D. Klumpers, M. Darnell, S.A. Bencherif, J.C. Weaver, et al.,
Hydrogels with tunable stress relaxation regulate stem cell fate and activity,
Nat. Mater. 15 (2016) 326e334.
[74] B. Follin, M. Juhl, S. Cohen, A.E. Pedersen, M. Gad, J. Kastrup, et al., Human
adipose-derived stromal cells in a clinically applicable injectable alginate
hydrogel: phenotypic and immunomodulatory evaluation, Cytotherapy 17
(2015) 1104e1118.