JOURNAL OF TISSUE ENGINEERING AND REGENERATIVE MEDICINE RESEARCH ARTICLE

J Tissue Eng Regen Med 2017; 11: 3469–3480.

Published online 31 May 2017 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/term.2260

Substance P/dexamethasone-encapsulated PLGA

scaffold fabricated using supercritical fluid process for
calvarial bone regeneration
Su Hee Kim1,2, Ji Eun Kim1,2, Soo Hyun Kim1,2,3* and Youngmee Jung2,3*
1
NBIT, KU-KIST Graduate School of Converging Science and Technology, Korea University, 145, Anam-ro, Seongbuk-gu, Seoul 136-701,
Korea
2
Center for Biomaterials, Biomedical Research Institute, Korea Institute of Science and Technology, P.O. Box 131, Cheongryang, Seoul
130-650, Korea
3
Department of Biomedical Engineering, University of Science and Technology (UST), Seoul 136-791, Korea

Abstract
Poly(lactic-co-glycolic acid) (PLGA) scaffolds encapsulated with substance P (SP) and dexamethasone (Dex) by the supercritical
CO2 foaming method were fabricated to treat calvarial bone. We compared the release profiles of SP and Dex according to the
incorporation methods using encapsulation or dipping. Ninety percent of the SP or Dex molecules in the scaffolds prepared
by the encapsulating method were released by day 14 or day 6, respectively. In vivo real-time assays for human mesenchymal
stem cell (hMSC) tracking were performed to confirm the MSC recruitment abilities of the scaffolds. The results showed that
the optical intensity of the SP-encapsulated group was 2.59 times higher than that of the phosphate-buffered saline group
and 1.3 times higher than that of the SP-dipping group. Furthermore, we compared the angiogenesis activity of the scaffolds.
In the SP-encapsulated group, 72.9 ± 2.6% of the vessels showed matured features by 1 week, and it increased to 82.0 ±
4.6% after 4 weeks. We implanted the scaffolds into rat calvarial defects. After 24 weeks, SP- and Dex-encapsulated scaffolds
showed 67.1% and 26.2% higher bone formation than those of the Dex-encapsulated group and SP-encapsulated group, respec-
tively, and they formed 36.1% more bone volume compared with the SP- and Dex-dipped scaffolds. Consequently, the results of
this study suggest that SP- and Dex-encapsulated scaffolds made by the supercritical CO2 foaming method could be a good treat-
ment modality to treat critical bone defects without cell transplantation by recruiting autologous stem cells and forming new
bone tissues. Copyright © 2017 John Wiley & Sons, Ltd.

Received 31 August 2015; Revised 20 April 2016; Accepted 3 July 2016

Keywords substance P; dexamethasone; calvarial defect; bone regeneration; supercritical CO2 foaming; stem cell recruitment

1. Introduction bone regeneration rate was faster by almost twofold com-

Flat bone defects of critical size caused by craniofacial de-
formation, trauma, resection of tumour, and other dis-
eases require transplantation of bone tissue (Szpalski
et al., 2010). Physiologically, bone tissues repeat the cycle
for growing through new bone tissue formation and the
absorption of osteoclasts (Raisz, 1999). However,
critical-sized defects or lack of key cells such as mesenchy-
mal stem cells (MSCs) for bone growth decreases the ca-
pability of bone regeneration. Especially, healing of flat
bones is slower than that of long bone defects due to var-
ious reasons. Ham W. (1930) reported that during devel-
opment, the cranial vault and facial bones with a flat bone
morphology undergo membranous ossification, while
long bones in the limbs are formed by endochondral ossi-
fication. These different types of healing mechanism af-
fect the healing rate. Lim J et al. reported that the tibial
*Correspondence to: Soo Hyun Kim and Youngmee Jung, Center for Biomaterials,
Biomedical Research Institute, Korea Institute of Science and Technology, P.O. Box
131, Cheongryang, Seoul 130-650, Korea. E-mail: soohkim@kist.re.kr;
winnie97@kist.re.kr
pared with the calvarial bone regeneration rate (Lim
et al., 2013). Rapid healing of a long bone defect could
be attributed to the potential of the tibial periosteum. Ad-
ditionally, mechanical loading may also contribute to
rapid bone regeneration in the tibia compared with that
in the calvaria. Mechanical strain is a fundamental physi-
ological factor regulating bone formation and renewal
(Zhang et al., 2013). Lack of mechanical load causes re-
duced production of matrix-protein (Huiskes et al.,
2000; Umemura et al., 2008). Therefore, implantation of
bone substitutes like allografts or osteoinductive biomate-
rials including bioceramics, or polymeric scaffolds is es-
sential to increase the healing efficiency of the calvarial
bone (Bose et al., 2012).
Usually, dexamethasone (Dex) has been used as one of
the components for osteogenic differentiation of MSCs
(Mao et al., 2006; Mikami et al., 2011; Spiro et al., 2010;
Tenenbaum and Heersche, 1985). Furthermore, there
are several studies showing that Dex combined with im-
plants enhanced bone regeneration and the healing ca-
pacity of bones (Guzman-Morales et al., 2009; Tavakoli-
darestani et al., 2014). Mao D et al. reported that Dex

Copyright © 2017 John Wiley & Sons, Ltd.

3470
induced osteogenesis through the regulation of hedgehog
signalling molecules in rat MSCs (Mao et al., 2006). Addi-
tionally, Spiro AS et al. showed that short-term applica-
tion of Dex improved bone morphogenetic protein-7-
induced ectopic bone formation in vivo, and Tavakoli-
darestani R et al. reported that Dex-loaded hydroxyapatite
enhanced bone regeneration in rat calvarial defects (Spiro
et al., 2010; Tavakoli-darestani et al., 2014).
However, even though Dex enhances bone formation
in vivo environment, it has limitations in clinical use.
Long-term delivery of Dex, which is a kind of glucocorti-
coid, could lead to bone loss and osteoporosis as reported
in previous experimental and clinical studies (Liu et al.,
2011; Mazziotti et al., 2007). Therefore, the appropriate
dose and period for delivering Dex are important factors
in the successful regeneration of ectopic bone tissues.
Recently, several studies have reported bone regenera-
tion through stem cell recruitment. Catarina R et al. inves-
tigated NK-cell-mediated MSC recruitment by fibrinogen
adsorption. Designing novel biomaterials leading to ratio-
nal modulation of inflammatory responses was proposed
as an alternative to traditional bone regeneration strate-
gies such as stem cell therapy or the implantation of tissue
engineered bone (Catarina et al., 2012). Substance P (SP)
has been applied to calvarial bone regeneration. Kim et al.
showed that dual delivery of SP and bone morphogenetic
protein-2 (BMP-2) enhanced bone formation (Noh SS
et al., 2015; La WG et al., 2014). They delivered SP and
BMP-2 using graphene oxide-coated titanium to further
promote bone formation. This approach improved the
osteointegration of titanium in dental or orthopaedic
implants.
Substance P is an 11-amino acid neuropeptide secreted
from the peripheral terminals of sensory nerve fibres and
an injury-inducible factor that acts to induce mobilization
of CD29+ stromal-like cells in wound-healing processes
(Hong et al., 2009). It was also confirmed that controlled
release of SP induced the recruitment of circulating angio-
genic cells from the blood to the injury site, resulting in
enhanced angiogenesis (Andersson et al., 2011; Kohara
et al., 2010; Seegers et al., 2003). In our previous work,
we used KLD12-SP, which was conjugated with SP at the
end of the KLD peptide chain, to treat calvarial bone de-
fects. In that study, it was confirmed that SP released from
peptide hydrogels recruited circulating MSCs, which con-
sequently led to accelerate bone tissue formation, and
repaired bone defects by the addition of new bone tissue
smoothly integrated with the host tissue (Kim et al.,
2015a).
There have been studies on the biomedical applications
of supercritical CO2 due to its biocompatibility, non-toxic
properties and good solvent power. Additionally, because
supercritical CO2 has relatively mild critical points, it has
advantages for use with biomolecules like growth factors
or small molecules (Ginty et al., 2008; Lee and Shin,
2013; Tai et al., 2007; Tomasko et al., 2009). In our previ-
ous work, we fabricated highly elastic poly (lactide-co-
caprolactone) (PLCL) scaffolds using a supercritical CO2
foaming method. They had many large and opened-pores
Copyright © 2017 John Wiley & Sons, Ltd.
S. H. Kim et al.
and improved interconnectivities between the pores. Fur-
thermore, the PLCL scaffolds showed no cytotoxic effects,
but did show enhanced tissue formation (Kim et al.,
2013b). Additionally, we could fabricate TGF-β3-encapsu-
lated PLCL scaffolds with the supercritical CO2 foaming
method. Because supercritical CO2 can loosen polymer
chains during the fabrication process, biomolecules that
are mixed with the polymer can be trapped in the polymer
matrix when the CO2 molecules are released from the re-
actor. We confirmed that TGF-β3 was homogeneously dis-
tributed in the polymer matrix and that they were
sustainably released over 8 weeks and, eventually, they
enhanced hyaline cartilaginous tissue formation (Kim
et al., 2015b).
In this study, we hypothesized that a dual biomolecule
(SP and Dex)-encapsulated poly(lactic-co-glycolic acid)
(PLGA) scaffold would have the ability to recruit MSCs,
induce angiogenesis for forming matured bone structures
and, subsequently, improve bone regeneration. Moreover,
we hypothesized that controlled delivery of SP would ac-
celerate recruitment of key cells, and controlled delivery
of Dex would increase the ALP activity during tissue re-
generation and finally enhance osteogenic differentiation
of the recruited MSCs by release of SP. To confirm this,
we fabricated SP- and Dex-encapsulated PLGA scaffolds
using the supercritical CO2 foaming method and im-
planted them into rat calvarial defect models. We tried
to control the sustained release of Dex and SP to improve
the efficiency of bone tissue regeneration.
2. Materials and methods
2.1. Fabrication and characterization of scaffolds
One-hundred milligrams of PLGA (Mn = 49 500, Mw =
98 700; Resomer, USA) was mixed with 1 mg SP
(Calbiochem, USA) and 1.5 mg Dex (Sigma, USA). After
the mold containing the mixture of PLGA, SP and Dex
was put into a reactor, it was heated to 37 °C and filled
with CO2 at 270 bars. After the conditions of the reactor
reached the saturation pressure and temperature, the
mixture was left to incubate for 1 h. The reactor was
depressurized back to ambient pressure over a period of
3 h as a venting step. The process of producing a porous
scaffold with the scCO2 was controlled with a mass flow
controller (MFC, TSC-100, MKP nano high-tech, Korea;
Kim et al., 2013b). The scaffolds for the phosphate-
buffered saline (PBS), SP-dipping, Dex-dipping or SP +
Dex-dipping groups were fabricated with only PLGA pow-
der. Then, they were soaked in Dulbecco’’s phosphate-
buffered saline (DPBS; Corning, USA) or SP solution
(500 μg mL
1) for the PBS- or SP-dipping group. For the
Dex-dipping and SP + Dex-dipping groups, Dex (3 mg
mL
1) and SP + Dex solution (500 μg SP and 3 mg Dex
mL
1) were prepared. Four milligrams of Dex was dis-
solved in anhydrous ethanol (Sigma, USA) at 37 °C and,
then, PBS or SP solution was added to the Dex solution.
SP-encapsulated scaffolds, Dex-encapsulated scaffolds
J Tissue Eng Regen Med 2017; 11: 3469–3480.
DOI: 10.1002/term
Calvarial bone regeneration with SP/Dex encapsulated PLGA scaffold
and SP + Dex-encapsulated scaffolds were dipped for
wetting. SP-encapsulated scaffolds were dipped in DPBS
for 1 h; and Dex-encapsulated scaffolds and SP + Dex-
encapsulated scaffolds were dipped in DPBS containing
0.1 v/v% ethanol for 1 h at 37 °C. Ethanol was used to sol-
ubilize Dex because it has low water solubility.
Poly(lactic-co-glycolic acid) scaffolds were fabricated
with fluorescence-labelled SP (SP-TAMRA, Peptron, Ko-
rea) and Dex (dexamethasone fluorescein, Molecular
Probes, USA) to confirm the distribution of biomolecules
in porous scaffolds.
The morphologies of the scaffolds were examined with
scanning electron microscopy (SEM; Hitachi, Japan) oper-
ating at 15 kV. Porosity was measured with a mercury in-
trusion porosimeter (AutoPore IV, MicroMeritics, USA;
Kim et al., 2013b). Distribution of fluorescence-labelled
SP and Dex was observed by confocal microscope
(LSM700, Zeiss, Germany).
2.2. In vitro SP and Dex release test
The quantities of SP molecules released from the PLGA
scaffolds were measured using an ELISA kit (Substance
P Parameter Assay Kit, R&D Systems, USA). Additionally,
the concentrations of Dex were measured with UV spec-
troscopy at 239 nm (Pargaonkar et al., 2005). In this
study, the release profiles of six groups were compared:
Group 1 = SP dip.; Group 2 = SP encap.; Group
3 = Dex dip.; Group 4 = Dex encap.; Group 5 = SP +
Dex dip.; and Group 6 = SP + Dex encap. The SP and
Dex molecules were extracted in 10 mL DPBS containing
0.1% bovine serum albumin (Sigma, USA). The experi-
ment was performed at 37 °C with shaking at 100 rpm
for dynamic conditions. The extracts were collected every
2–3 days, and DPBS was exchanged with a fresh solution
at the time of collection.
2.3. Isolation and fluorescence-activated cell sorting
(FACS) of human adipose-derived MSCs
Human subcutaneous adipose tissue samples were ob-
tained from the abdomens of seven female donors be-
tween 35 and 54 years old with the approval of the
Catholic University of Korea Institutional Review Board.
The tissues were processed to isolate human adipose-
derived stem cells (hADSCs) as previously described
(Guzman-Morales et al., 2009). The hADSCs obtained
from different patients were expanded between passages
2 and 3 and randomly used in the experiments. To con-
firm the MSC markers of the isolated human adipose-
derived MSCs, FACS analysis was performed. Expressions
of CD29, CD90, CD105, CD166, CD31, CD34 and CD45
were evaluated. Cells were immunolabelled with 1 mg
of fluorescence-conjugated mouse antihuman monoclo-
nal antibodies for 30 min at 4 °C. The fluorescence-
labelled cells were analysed with a MoFlo cell sorter
(DakoCytomation, USA).
Copyright © 2017 John Wiley & Sons, Ltd.
3471
2.4. In vivo MSC recruitment and real-time human (h)MSC
tracking assay
To investigate hMSC recruitment to the scaffolds contain-
ing SP, an in vivo real-time hMSC tracking assay was per-
formed (Kim et al., 2015a). All animals (balb/c, 7 weeks
old, male, 20–30 g; Nara Biotech, Korea) were cared for
according to the methods approved by the Institutional
Animal Care and Use Committee (IACUC, 2009-01-5254-
002) at the Korea Institute of Science and Technology.
For subcutaneous implantation, temporary anaesthesia
was performed by intra-muscular (IM) injection with a
cocktail of zolazepam and tiletamine (0.3 mL kg
1;
Zoletil®, Virbac, reconstFrance) and xylazine (0.1 mL
kg
1; Rompun®, Bayer, Germany).
Human MSCs were labelled with near-infrared (NIR)
nanoparticles (NEO-LIVE Magnoxide 675, ex/em =
675/700; Biterials, Korea) according to the manufac-
turer’’s instructions.
Three kinds of scaffolds (Group 1 = PBS-dipped PLGA
scaffolds; Group 2 = SP-dipped PLGA scaffolds; and
Group 3 = SP-encapsulated PLGA scaffolds) were pre-
pared for this study. The method in detail is given in Sec-
tion 2.1. The scaffolds were implanted subcutaneously
into nude mice (n = 3), and then hMSCs (1 × 106 cells
per 200 μL of PBS) were injected into the tail vein of the
mice. After surgery, the movement of the injected cells
was tracked every 2–3 days for 8 weeks with an in vivo
fluorescence imaging system (I-VIS spectrum, Caliper Life
Sciences, USA). After 8 weeks, the scaffolds were har-
vested and analysed. The acquired images from the I-VIS
spectrum were processed with an image analysis program
(Image J, NIH, USA) to calculate the optical density, and
harvested samples (n = 3) were sectioned at 7 μm thick-
nesses and then stained with 40 ,6-diamidino-2-
phenylindole (DAPI; Molecular Probes). To confirm the
recruitment of MSCs at the initial stage, non-labelled
MSCs were injected into the tail vein of mice instead of
NIR-labelled MSCs. Two weeks after surgery, the three
kinds of scaffolds were explanted and sectioned with a mi-
crotome. The samples were stained with CD29 and CD90
antibodies (1 : 1000 and 1 : 500; Abcam, Cambridge, MA,
USA) as MSC markers.
2.5. Calvarial defect model in SD-RATs
All animals (SD-RAT, male, 7 weeks, 200–300 g;
Samtako, Korea) were cared for according to the methods
approved by the Institutional Animal Care and Use Com-
mittee at Korea Institute of Science and Technology. For
the surgery, temporary anaesthesia was performed by
IM injection with a cocktail of zolazepam and tiletamine
(0.3 mL kg
1; Zoletil®, Virbac, France) and xylazine
(0.1 mL kg
1; Rompun®, Bayer, Germany). A defect 8
mm in diameter was made in the calvarial bone using a
trephine bur (TPHB-B8, OSUNG, Korea) after the perios-
teum was removed. The prepared samples were im-
planted to investigate bone regeneration. Each implant
J Tissue Eng Regen Med 2017; 11: 3469–3480.
DOI: 10.1002/term
3472
was observed after 24 weeks by micro-computed tomog-
raphy (CT) and histological staining analysis. For the
calvarial defect surgery, the rats were randomly allocated
into eight groups (n = 6) as follows: Group 1 = defect;
Group 2 = PBS; Group 3 = SP dip.; Group 4 = SP
encap.; Group 5 = Dex dip.; Group 6 = Dex encap.;
Group 7 = SP + Dex dip.; and Group 8 = SP +
Dex encap, and three kinds of scaffolds (n = 3, Group
1 = PBS, Group 2 = SP dip., Group 3 = SP encap.) were
implanted in the defect sites and then harvested at 1 and
4 weeks after implantation to evaluate in situ MSC
recruitment and angiogenesis.
2.6. Micro-CT
Quantitative three-dimensional analysis of the calvarial
defect samples (n = 6) was analysed using a
Skyscan1172 CT (Bruker, USA). At 24 weeks, the animals
were killed, and the bone defects were harvested. Briefly,
the specimens were fixed in 10% buffered neutral forma-
lin (v/v) for 1 day. Samples were held securely in a 5-mL
conical tube and centred in the middle of the micro-CT
machine. Samples were focused and calibrated. Following
calibration, samples were scanned using the following set-
tings: 49 kV, 200 μA and Al 0.5 mm filter. Raw data were
collected and reconstructed with the NRecon software
(Bruker, USA). New bone volume, bone surface/bone vol-
ume and bone volume/total volume were calculated using
the CTAn software (Bruker, USA; Kanczler et al., 2008).
2.7. Histological analysis by haematoxylin & eosin (H&E)
and Masson’’s trichrome (MT) staining
After micro-CT analysis, samples were decalcified in a
decalcifying solution (Sigma, USA) for 6 h. The specimens
were dehydrated in a graded ethanol series (80–100%),
and embedded in paraffin. The sampled tissues (n = 6)
were sectioned at 7 μm thickness and then stained with
H&E or MT stains. The sections were examined under
light microscopy (Nickon, Japan). The stained area was
calculated with the MT-stained images (n = 6) using
the image analysis program Image J.
2.8. Immunofluorescence staining
Residual SP in the scaffolds and recruitment of MSCs
were investigated by immunofluorescence staining. Resid-
ual SP was observed with anti-SP antibody (1 : 50; Santa
Cruz, USA). Migrated MSCs to the injury site during bone
regeneration were analysed by staining with the MSC
markers CD29 and CD90 (1 : 1000 and 1 : 500; Abcam,
Cambridge, MA, USA; Kim et al., 2013a). To evaluate the
density of vascular cells in the scaffolds 1 week and 4
weeks after implantation, endothelial cells (ECs) and vas-
cular smooth muscle cells (SMCs) were stained with poly-
clonal rabbit anti-human von Willebrand factor antibody
(vWF; 1 : 100; Abcam, USA) and monoclonal mouse
Copyright © 2017 John Wiley & Sons, Ltd.
S. H. Kim et al.
anti-human α-smooth muscle actin antibody (α-SMA; 1 :
100; Santa Cruz Biotech, USA), respectively. vWF- and
α-SMA-positive vessels in three random fields were evalu-
ated, and vWF+ cell density and α-SMA+ vessel density
were quantified with the image analysis program Image
J (n = 3 in each group). The maturation index was quan-
tified as the ratio of α-SMA-positive vessels to the total
number of vessels (Kim et al. 2011).
2.9. Statistical analysis
All samples were assayed at least in triplicate, and the re-
sults obtained were expressed as the standard deviations
(SDs) above and below the mean. All statistical analyses
were performed with the ANOVA test (GraphPad Prism
6, USA). Results were considered to be statistically signif-
icant when the P-value was less than 0.05.
3. Results
3.1. Characterization of the Dex/SP-encapsulated
scaffolds
The SP- and Dex-encapsulated PLGA scaffolds were fabri-
cated with the supercritical CO2 foaming method. Addi-
tionally, PLGA scaffolds without any biomolecules were
fabricated with the same method for the control and dip-
ping groups. For the experiments, the scaffolds were cut
into disks after the skin layer of the constructs was
removed (diameter = 8.2 mm, thickness = 1.3 mm)
(Table 1). From the SEM results on the morphology of
the scaffolds, all the scaffolds had many opened pores
that were 214 ± 23.5 μm in pore size. Additionally, the
average porosity of the PLGA scaffolds was 69.4 ±
4.2%. There were no significant differences in the pore
size and porosity of the pure PLGA scaffolds and the
biomolecule-encapsulated PLGA scaffolds (Figure 1a
and b). We confirmed the distribution of the biomole-
cules in the polymer matrix after fabrication of the scaf-
folds with supercritical CO2. Fluorescence-labelled SP
and Dex molecules were homogeneously distributed in
the upper part and lower part of the scaffolds (Figure 1c
and d). From the result, we concluded that the SP and
Dex molecules were homogeneously mixed and distrib-
uted in the PLGA polymer matrix.
3.2. SP and Dex release studies in vitro
Substance P and Dex release profiles were investigated
with in vitro release tests. In Figure 1, 90% of the SP mol-
ecules in the dipping groups (SP dip. And SP + Dex dip.)
were released by day 1. In contrast, 90% of the SP mole-
cules in the encapsulated groups (SP encap. And SP +
Dex encap.) were released by day 14 and continuously de-
tected over 56 days. The release behaviour of SP in dual
biomolecule delivery was similar to that in single biomol-
ecule delivery. In the dipping groups, 90% of the Dex
J Tissue Eng Regen Med 2017; 11: 3469–3480.
DOI: 10.1002/term
Calvarial bone regeneration with SP/Dex encapsulated PLGA scaffold
Figure 1. (a) Characterization of the poly(lactic-co-glycolic acid) (PLGA) scaffolds. A PLGA scaffold. (b) Scanning electron microscopy (SEM) images of the PLGA scaffold. Scale
bar: 100 μm. (c) Fluorescence images of substance P (SP; red) and dexamethasone (Dex; green) in the upper part of the PLGA scaffold. Scale bar: 200 μm. (d) Fluorescence images
of SP (red) and Dex (green) in the lower part of the PLGA scaffold. Scale bar: 200 μm. In vitro release tests: (e) release profile of SP; (f) release profile of Dex. [Colour figure can be
viewed at wileyonlinelibrary.com]
molecules were released by day 1, and 90% of the Dex
molecules in the encapsulated groups were released by
day 6. There was no significant difference in the releasing
rates of Dex between single biomolecule delivery and dual
biomolecule delivery just like for SP.
3.3. In vivo MSC recruitment and real-time MSC tracking
assay
To confirm MSC recruitment to the scaffolds containing
SP, an in vivo real-time hMSC tracking assay was per-
formed. hMSCs were used for this experiment. After isola-
tion of the hMSCs from adipose tissue, MSC markers were
evaluated with FACS analysis. CD29, CD90, CD105 and
CD166 were used as MSC markers. Additionally, CD31,
CD34 and CD45 were observed as EC lymphocyte or
haematopoietic cell markers. Among the cells, 99.2%
expressed CD29 and CD166, and 97.9% of the cells
expressed CD105 and CD90. On the other hand, 99.2%
of the cells did not express CD31 and CD45, and 98.2%
of the cells did not express CD31 and CD34 (Figure 2a).
From these results, we could confirm that almost all of
the isolated cells (> 97.9%) expressed the MSC markers.
These hMSCs from adipose tissues were used to investi-
gate the MSC recruitment ability of the scaffolds contain-
ing SP in a subcutaneous implantation model. The
samples were harvested 2 weeks after surgery to confirm
recruitment at the initial period and stained with hMSC
marker CD29 and CD90 antibodies. As seen in Figure 2b
and c, MSCs rarely migrated to the scaffold containing
PBS. In contrast, many MSCs migrated to the scaffolds
containing SP. Especially, the stained area of the SP-
encapsulated group was about 35% higher in MSCs than
that of the SP-dipping group. Furthermore, a real-time
hMSC tracking assay was performed to observe the long-
Copyright © 2017 John Wiley & Sons, Ltd.
3473
term effects of the MSC recruitment. hMSCs labelled with
NIR-nanoparticles, which were injected into nude mice
that had the scaffolds subcutaneously implanted by intra-
venous injection, were tracked every 2 or 3 days for 8
weeks with an NIR-imaging machine. The optical signal-
ling of the NIR-labelled hMSCs in the scaffold was first de-
tected 2 weeks after implantation. The injected hMSCs
had migrated to the scaffolds over 8 weeks. Figure 2d
shows that the optical intensity of the SP-encapsulated
group was 2.59 times higher than that of the PBS group
and 1.3 times higher than that of the SP-dipping group
at 8 weeks. Additionally, after explantation at 8 weeks
and identifying the labelled, recruited MSCs in the scaf-
folds, more than 30% of the MSCs were observed in the
SP-encapsulated scaffold compared with the SP-dipping
scaffold (Figure 2e and f). We believe this result is be-
cause 90% of the SP molecules were released over 14 days
continuously from the scaffold in the SP-encapsulated
group based on the in vitro results of the SP release tests,
which means SP was sustainably released compared with
the SP-dipping group. Thus, the results suggest that the
long-term delivery of SP improved the efficiency of MSC
recruitment to the scaffolds.
3.4. In situ MSC recruitment in a rat calvarial defect
We evaluated in situ MSC recruitment for three kinds of
scaffolds at the calvarial defect sites at weeks 1 and 4 after
implantation (PBS, SP dip. And SP encap. Groups). As
seen by the DAPI staining data in Figure 3, a similar
amount of cells migrated to the scaffolds in all the groups
at week 1. However, in the case of the PBS group, only a
small number of cells were observed in the scaffold at 4
weeks compared with the other groups. In contrast, many
cells were present in the scaffolds in both the SP-dipping
J Tissue Eng Regen Med 2017; 11: 3469–3480.
DOI: 10.1002/term
3474 S. H. Kim et al.

Figure 2. (a) Fluorescence-activated cell sorting (FACS) analysis of isolated human mesenchymal stem cells (hMSCs). (b) Mesenchymal stem cell (MSC) recruitment 2 weeks af-
ter subcutaneous implantation (non-labelled hMSCs were used, Groups = SP encap., SP dip. And PBS scaffolds). PBS, phosphate-buffered saline; SP, substance P. (c) Quantitative
analysis by the image J program. (d) In vivo real-time hMSC tracking assay. Movement of labelled hMSCs to the scaffolds containing SP or PBS over 8 weeks. Optical images of the
scaffolds explanted after 8 weeks and quantitative analysis of the optical intensity. (e) Neo-live labelled hMSCs 8 weeks after subcutaneous implantation of the scaffolds. (f)
Quantitative analysis by the image J program (*P < 0.05). Scale bar: 200 μm. [Colour figure can be viewed at wileyonlinelibrary.com]

and SP-encapsulated groups. Especially, we could observe
many cells expressing MSC markers CD29 and CD90 in
the SP-encapsulated scaffold at week 1, and we could con-
tinuously see the signalling at week 4. However, the
stained area of the SP-dipping group decreased by week
4. In the case of the centre part of the scaffolds
(Figure 3c and d), it had a similar tendency for staining
as the edge part of the scaffolds (Figure 3a and b).
To further prove our hypothesis, we confirmed the pres-
ence of residual SP in the scaffolds with SP antibodies at
weeks 1 and 4 after implantation (Figure 3e and f). The
stained areas of the SP-encapsulated group were 2.53
times larger at week 1 and 2.52 times larger at week 4
than those of the SP-dipping group. We believe that the
SP in the SP-dipping scaffolds is adsorbed onto the surface
of the scaffolds rapidly and then easily washed out, while
the SP trapped in the polymer mass remains for a long
time inside the SP-encapsulated scaffolds.
3.5. Angiogenesis (capillary density and maturation index)
analysis in a rat calvarial defect
We evaluated angiogenesis and microvessel formation at
the calvarial defect sites where the scaffolds were im-
planted. In Figure 4a and b, the vWF+ cell density of the
SP-encapsulated group was 3.21 times higher than that
of the PBS group and 2.2 times higher than that of the
SP-dipping group at week 1 after implantation. Further-
more, the α-SMA+ vessel density for the SP-encapsulated
group was 10.1 times higher than that of the PBS group
and 2.1 times higher than that of the SP-dipping group
at week 1 after implantation (Figure 4a and c). The mat-
uration index was calculated based on double staining re-
sults at weeks 1 and 4 to determine the percentage of
matured vessels. The SP-encapsulated group had the
highest index value among all the groups, with 72.9 ±
2.6% of the vessels showing matured features at week 1,
and it increased up to 82.0 ± 4.6% at week 4. The SP-
dipping group showed 70.7 ± 5.0% of matured vessels,
and it decreased to 56.9 ± 6.6% by week 4 (Figure 4a
and d).
3.6. Micro-CT analysis for bone formation in rat calvarial
defects
Seven kinds of scaffolds were implanted into rat calvarial
defects to evaluate the bone regeneration efficiency of the
Dex- and SP-encapsulated scaffolds (Group 1 = Defect;
Group 2 = PBS; Group 3 = SP dip.; Group 4 = SP
encap.; Group 5 = Dex dip.; Group 6 = Dex encap.;
Group 7 = SP + Dex dip.; and Group 8 = SP +
Dex encap.). We evaluated the formation of newly formed
calcified bone tissue for bone regeneration by micro-CT
analysis 24 weeks after the surgery (Figure 5).
New bone volume of the Dex-encapsulated group was
15.6% higher than that of the defect group and 24.1%
higher than that of the PBS group. However, there was
no significant difference between the Dex-dipping and

Copyright © 2017 John Wiley & Sons, Ltd. J Tissue Eng Regen Med 2017; 11: 3469–3480.
DOI: 10.1002/term
Calvarial bone regeneration with SP/Dex encapsulated PLGA scaffold 3475

Figure 3. Mesenchymal stem cell (MSC) recruitment of in vivo samples. (a) Edge of the scaffolds. (b) Quantification of the stained area. (c) Centre of the scaffolds. (d) Quanti-
+
fication of the stained area. The edge at 24 weeks (arrow: interface between host and new tissue) [ P < 0.05 with SP encap. (1w), *P < 0.05 with SP encap. (4w)]. SP, substance
+
P. Scale bar: 200 μm. (e) Residual SP in the scaffolds was observed by anti-SP. (f) Stained area was measured by the Image J program [ P < 0.05 with SP encap. (1w), *P < 0.05
#
with SP encap. (4w), P < 0.05]. Scale bar: 200 μm. [Colour figure can be viewed at wileyonlinelibrary.com]

Dex-encapsulated groups for the micro-CT data. Although
in Figure 2b, 90% of the Dex molecules from the Dex-
dipping scaffolds were released by day 1 and those from
the Dex-encapsulated scaffolds were released by day 6 in
the in vitro release tests, the results of the bone regenera-
tion showed that the release kinetics rarely affected new
bone tissue formation.
Furthermore, we could confirm that the scaffolds con-
taining SP showed a higher bone regeneration ability than
the scaffolds containing Dex when we compared the sin-
gle molecule delivery groups (SP dip., SP encap., Dex
dip. And Dex encap. Groups). The SP-encapsulated scaf-
folds had 53.1% higher mineralization than that of the de-
fect group, 64.4% higher mineralization than that of the
PBS group, and 12.8% higher mineralization than that
of the SP-dipped group. Additionally, the dual molecule
delivered scaffolds of SP and Dex fabricated with the dip-
ping method had a 25.3% higher mineralization than that

Copyright © 2017 John Wiley & Sons, Ltd. J Tissue Eng Regen Med 2017; 11: 3469–3480.
DOI: 10.1002/term
3476 S. H. Kim et al.

Figure 4. Angiogenesis in a rat calvarial defect. (a) Representative immunostaining of endothelial cells (ECs) and smooth muscle cells (SMCs) in each group 1 and 4 weeks after
+ 2
implantation, respectively. Groups = SP encap., SP dip. And PBS scaffolds. PBS, phosphate-buffered saline; SP, substance P. (b) Quantification of the vWF cell density (μm

2 + 2
2
mm ). vWF, von Willebrand factor antibody. (c) Quantification of the α-SMA vessel density (μm mm ). α-SMA, α-smooth muscle actin antibody. (d) Maturation index (%).
+ +
Quantification of the maturation index was based on a comparison of the percentage of α-SMA vessels to the total number of vessels [ P < 0.05 with SP encap. (1w),
#
*P < 0.05 with SP encap. (4w), P < 0.05]. Scale bar: 200 μm. [Colour figure can be viewed at wileyonlinelibrary.com]

of the Dex-only delivered scaffolds fabricated with dip-
ping and was similar to the SP-only delivered scaffolds
fabricated with dipping. The SP- and Dex-encapsulated
scaffolds showed a 67.1% higher bone formation than that
of the Dex-only encapsulated scaffolds, and 26.2% higher
bone formation than that of the SP-only encapsulated
scaffolds. Moreover, they showed 36.1% higher new bone
volume than that of the SP and Dex delivered scaffolds
fabricated with the dipping method. In the case of the
scaffolds containing SP and Dex fabricated with the dip-
ping method, the SP and Dex molecules were almost all
released from the scaffolds by day 1, and they showed
similar bone regeneration capacities to that of the SP-only
delivered scaffolds, while the SP- and Dex-encapsulated
scaffolds fabricated with the supercritical CO2 foaming
method showed an enhanced bone formation capacity
compared with the SP- or Dex-only delivered scaffolds.
Therefore, the micro-CT results suggest that the long-term
delivery of SP by encapsulation into the scaffolds with the
supercritical CO2 foaming method enhanced MSC recruit-
ment, and the Dex released from the scaffolds facilitated
osteogenic differentiation of the recruited MSCs in the
calvarial defect sites.
3.7. Histological studies of the in vivo samples
We performed a histological analysis of the in vivo sam-
ples by H&E and MT staining. In Figure 6a and b, brown
bony tissues (black arrows = bony tissue) were observed
at the edge parts of the implanted sites in the MT-stained
images for the PBS and defect groups, and most of the
stained area showed fibrous tissues that were stained blue
by the methyl blue dye in the two groups. In contrast, cal-
cified bony tissues formed at both the edge and centre
parts of the SP or Dex delivered scaffolds. The area of fi-
brotic tissue decreased in the SP- or Dex-treated groups.
Especially, the portion of collagenous tissue was the low-
est in the SP- and Dex-encapsulated group. Figure 6c
shows the quantitative analysis of the stained area of col-
lagenous tissue and bony tissue. The SP-containing
groups had higher bony tissue than those of the Dex-

Copyright © 2017 John Wiley & Sons, Ltd. J Tissue Eng Regen Med 2017; 11: 3469–3480.
DOI: 10.1002/term
Calvarial bone regeneration with SP/Dex encapsulated PLGA scaffold 3477

Figure 5. Micro-computed tomography (CT) analysis of calvarial bone defect. Samples were explanted 24 weeks after implantation. Group 1 = defect; Group 2 = PBS; Group
3 = SP dip.; Group 4 = SP encap.; Group 5 = Dex dip.; Group 6 = Dex encap.; Group 7 = SP + Dex dip.; and Group 8 = SP + Dex encap. Dex, dexamethasone; PBS,
phosphate-buffered saline; SP, substance P. (a) Micro-CT images of the samples. (b–d) New bone volume, bone surface, and new bone volume/total volume were calculated
+ # $
by the CTAn program (*P < 0.05 with defect, P < 0.05 with SP encap., P < 0.05 with Dex encap. And P < 0.05 with SP + Dex encap.). Scale bar: 8.2 mm

containing groups. Especially, the SP- and Dex-
encapsulated scaffolds had the largest brown area com-
pared with the other groups, which is consistent with
the micro-CT results.
4. Discussion
We have evaluated SP- and Dex-encapsulated PLGA scaf-
folds on a rat calvarial defect model. We created bone de-
fects and implanted them into the defect sites. According
to the reference, for the rat calvarial defect, 8 mm is gen-
erally accepted as a critical size; however, smaller defects
have been investigated in models with two defects per an-
imal. The critical sized calvarial bone defect cannot be
completely filled with calcified bone without any treat-
ment (Spicer et al., 2012). In this study, 8-mm-diameter
calvarial defects were created in the centre of the calvarial
bone, and we implanted bioactive scaffolds into the defect
site.
We encapsulated SP and Dex for long-term delivery sys-
tem of SP and Dex. SP and Dex release profiles were inves-
tigated with in vitro release tests. Ninety percent of the SP
molecules in the encapsulated groups (SP encap. And SP
+ Dex encap.) were released by day 14 and continuously
detected over 56 days. In contrast, 90% of the Dex mole-
cules in the encapsulated groups were released by day 6.
Regarding the release competition of Dex and SP, Dex
molecules were released faster than that of SP in our
Dex and SP delivery system. The reason for this we think
is that the Dex molecules diffuse more easily from the
scaffolds because they have a lower molecular weight
than that of SP. The molecular weights of SP and Dex
are 1.392 kDa and 0.392 kDa, respectively. Kim CJ et al.
reported that the release of macromolecules was mainly
dependent on the diffusion of the macromolecules at the
surface of the matrix initially, and then dependent on
the degradation of the matrix at the later stages, while
small drugs seemed to depend totally on diffusion for
the duration of the release study (Kim, 1998). Thus, com-
paring the small molecules of Dex and SP, Dex had a
faster diffusion rate than that of SP because of the differ-
ence in molecular weight.
When tissue has damage, SP is released from peripheral
blood vessels, and one of the important roles of the SP is
to induce wound-healing processes by recruiting CD29+
cells (Catarina et al., 2012). Wound-healing processes in-
clude MSC recruitment, cytokine secretion and angiogen-
esis. Therefore, we expected that local delivery of SP from
the scaffolds would induce MSC recruitment and angio-
genesis to the scaffolds implanted in the calvarial defect.
We investigated MSC homing via real-time MSC tracking
assay. We compared the efficiency of MSC recruitment be-
tween SP-dipped scaffolds and SP-encapsulated scaffolds.
More than 30% of the MSCs were observed in the SP-
encapsulated scaffold compared with the SP-dipping scaf-
fold (Figure 2e and f). Additionally, we evaluated in situ

Copyright © 2017 John Wiley & Sons, Ltd. J Tissue Eng Regen Med 2017; 11: 3469–3480.
DOI: 10.1002/term
3478 S. H. Kim et al.

Figure 6. Bone reconstruction in each group in a calvarial model at 24 weeks. (a and b) Haematoxylin & eosin (H&E) and Masson’’s trichrome (MT) staining for each group
(Group 1 = defect; Group 2 = PBS; Group 3 = SP dip.; Group 4 = SP encap.; Group 5 = Dex dip.; Group 6 = Dex encap.; Group 7 = SP + Dex dip.; and Group 8 =
SP + Dex encap.) in the calvarial model at 24 weeks (black arrows = bony tissue). Dex, dexamethasone; PBS, phosphate-buffered saline; SP, substance P. (c) Quantitative anal-
#
ysis of the stained area (collagenous and non-collagenous tissue) ( P > 0.05 with Dex encap ). Scale bar: 1 mm (a); 200 μm (b). [Colour figure can be viewed at
wileyonlinelibrary.com]

MSC recruitment on a rat calvarial defect model. At 1
week and 4 week after implantation of scaffolds, we
harvested them and evaluated via immunofluorescence
staining. From the results, we concluded that the SP-
encapsulated scaffolds had a better MSC recruitment abil-
ity than that of the SP-dipping group. We believe these
results are due to the duration of the SP release from
the scaffolds. Ninety percent of the SP molecules were re-
leased over 14 days continuously from the scaffold in the
SP-encapsulated group based on the in vitro results of
the SP release tests, which means SP was sustainably re-
leased compared with the SP-dipping group. Thus, the re-
sults suggest that the long-term delivery of SP improved
the efficiency of MSC recruitment to the scaffolds.
Angiogenesis is crucial during tissue regeneration be-
angiogenesis into the damaged tissue. We tried to com-
pare the effects of release rate of SP. Therefore, we ob-
served the influence between short-term delivery and
long-term delivery systems of SP in the present study.
We have tried long-term delivery of SP via encapsulation
of SP in the polymer matrix to enhance angiogenesis.
Kohara et al. (2010) reported that the local delivery of
SP achieved the recruitment of circulating cells from the
blood flow around the delivered sites. Moreover, they
showed that controlled release of SP enhanced angiogen-
esis, and SP treatment with a high dose (5 μg per mouse)
increased blood vessel density. In this study, 90% of the
Table 1. Characterization of the PLGA scaffolds: size, porosity and pore size of the
PLGA scaffolds

cause it supplies nutrients and oxygen to damaged tissues PLGA scaffolds

as well as supplying a source of homing, circulating MSCs.
Especially, for successful bone regeneration, vessel forma- Size D = 8.2 mm, T = 1.3 mm
Porosity* 69.4 ± 4.2%
tion in newly formed bone tissues is important in terms of Pore size† 214 ± 23.5 μm
mimicking native bone tissues that include microvessels
*Mercury intrusion porosimeter (MIP).
and form complex tissues (Kim et al., 2011; Madeddu, †
Scanning electron microscopy (SEM).
2005). According to the references, SP induces PLGA, poly(lactic-co-glycolic acid).

Copyright © 2017 John Wiley & Sons, Ltd. J Tissue Eng Regen Med 2017; 11: 3469–3480.
DOI: 10.1002/term
Calvarial bone regeneration with SP/Dex encapsulated PLGA scaffold
SP molecules in the dipping group were released by day 1,
while 90% of the SP molecules in the encapsulated groups
were released by day 14. We think that the released SP
from the scaffolds induced angiogenesis at the defect site
by recruiting cells related to wound-healing processes in
both the SP-dipping and SP-encapsulated groups at week
1 after implantation. However, the sustained release of SP
from the SP-encapsulated group we think helped to main-
tain the matured vessels over the 4 weeks after
implantation.
Numerous studies have reported that Dex induces ec-
topic bone formation. Mikami Y et al. reported that Dex
enhanced differentiation of osteoblastic cells in vitro
(Mikami et al., 2011). Moreover, Dex increased the level
of ALP activity and the expression of osteocalcin.
Tavakoli-darestani R et al. showed that Dex-loaded hy-
droxyapatite enhances bone regeneration about 20%
compared with hydroxyapatite (Tavakoli-darestani et al.,
2014). Especially, Spiro AS et al. showed that BMP-7-
induced ectopic bone formation was significantly en-
hanced by systemic short-term application of Dex (Spiro
et al., 2010). Short-term delivery of Dex significantly in-
creased the bone volume and the osteoblast population
in the ectopic bone nodules compared with the untreated
controls. Therefore, we think that the released Dex from
the Dex-dipping or -encapsulated scaffolds within 1 week
played an important role in bone regeneration in the
calvarial defects.
Various studies have reported many kinds of roles for
SP, and they have revealed that SP stimulates differen-
tiation of osteoblast by upregulating osterix expression
(Fu et al., 2014; Goto et al., 2007; Tampieri et al.,
2008). According to Goto T et al., SP stimulated the ex-
pression of osteocalcin mRNA in osteoblasts at day 14
or day 21, whereas SP did not simulate the expression
of runx2 or type I collagen mRNA at day 7 but stimu-
lated them at day 14. These results indicate that SP
stimulates bone formation through osteoblastic cells
via neurokinin-1 receptors at the late-stage of bone
formation (Goto et al., 2007). Thus, we think that our
SP- and Dex-encapsulated scaffolds fabricated with
the supercritical CO2 foaming method would improve
bone formation in calvarial defect sites through the
synergetic effects of SP and Dex inducing the secretion
of bone extracellular matrices (ECMs) and osteogenic
differentiation.
Regarding SP, besides recruiting CD29+ cells, SP is
involved in immune responses for tissue regeneration
and wound healing. When tissue is damaged, many
kinds of cells migrate to the injured site. To regenerate
damaged tissue, not only do blood vessels, cytokines
and key cells such as MSCs have to gather at the defect,
they must also be balanced for successful wound
healing. Otherwise, fibrous tissue would form at the in-
jured site. It was confirmed that long-term delivery of SP
induced angiogenesis and maturation of the blood ves-
sels, and key cells, nutrients and oxygen would be sup-
plied by the blood vessels that would accelerate tissue
regeneration. Furthermore, when an injury occurs, an
Copyright © 2017 John Wiley & Sons, Ltd.
3479
immune response takes place at the injury site to treat
the tissue. Overexpression of pro-inflammatory factors,
such as tumour necrosis factor-α (TNF-α), interleukin
(IL)-1, IL-6, IL-8, IL-12 and IL-17, causes further infiltra-
tion of a variety of inflammatory cells and immune cells
into the tissue, and cellular transformation of fibroblasts
and finally tissue destruction. Hong HS and Son et al.
reported that SP played an anti-inflammatory role in a
collagen II-induced arthritis model (Hong HS et al.,
2014). Additionally, SP induces M2-type macrophages
after spinal cord injury (Jiang et al., 2012). SP skewed
the systemic cytokine profiles toward anti-inflammation
by reducing the levels of TNF-α and IL-17 and increasing
that of IL-10. Moreover, SP could increase the popula-
tion ratio of regulatory T-cells and M2-type macro-
phages in the circulatory blood cells. These studies
have shown that SP accelerates tissue regeneration by
anti-inflammatory modulation.
In this study, we fabricated SP- and Dex-encapsulated
PLGA scaffolds with the supercritical CO2 foaming
method. Controlled release of SP and Dex improved
MSC recruitment, angiogenesis and osteogenesis, and,
subsequently, they stimulated osteoblastic bone formation
and accelerated tissue regeneration in calvarial defects.
From the results, we could conclude that SP- and Dex-
encapsulated scaffolds enhanced calvarial bone formation
without cell transplantation.
4.1. Conclusion
We developed PLGA scaffolds encapsulating SP and Dex
using the supercritical CO2 foaming method. They had a
porous structure with a pore size and porosity of
214 ± 23.5 μm and 69.4 ± 4.2%, respectively. From
the SP- and Dex-encapsulated PLGA scaffolds, 90% of
the SP molecules were released by day 14, and 90% of
the Dex molecules were released by day 6. From the
in vivo real-time hMSC tracking assay, 30% more hMSCs
migrated to the SP-encapsulated scaffolds than that of
the SP-dipped scaffolds. When we implanted the scaf-
folds into rat calvarial defects to evaluate the effects
on bone regeneration, it was confirmed that the SP-
and Dex-encapsulated scaffolds enhanced the formation
of calcified new bone tissue and bone regeneration.
Consequently, our results show that a SP- and Dex-
encapsulated scaffold fabricated with the supercritical
CO2 foaming method is a good strategy to treat calvarial
bone defects without cell transplantation by recruiting
autologous stem cells and differentiation of the re-
cruited cells.
Acknowledgments
This work was supported by a grant of the Korea Health tech-
nology R&D Project through the Korea Health Industry Devel-
opment Institute (KHIDI), funded by the Ministry of Health &
Welfare (HI15C3060-010115), by a grant of Basic Science
J Tissue Eng Regen Med 2017; 11: 3469–3480.
DOI: 10.1002/term
3480
Research Program through the National Research Foundation
of Korea (NRF) funded by the Ministry of Science, ICT and fu-
ture Planning (2016R1A2B2009550), Republic of Korea, and
by a KU-KIST program of Korea University (R1309541).
References
Andersson G, Backman LJ, Scott A et al. 2011; Substance P
accelerates hypercellularity and angiogenesis in tendon
tissue and enhances paratendinitis in response to Achil-
les tendon overuse in a tendinopathy model. Br J Sports
Med 45: 1017–1022.
Bose S, Roy M, Bandyopadhyay A et al. 2012; Recent ad-
vances in bone tissue engineering scaffolds. Trends
Biotechnol 30: 546–554.
Catarina R, Daniela P, Raquel M et al. 2012; Enhanced mes-
enchymal stromal cell recruitment via natural killer cells
by incorporation of inflammatory signals in biomaterials.
J R Soc Interface 9: 261–271.
Fu S, Mei G, Wang Z et al. 2014; Neuropeptide substance P
improves osteoblastic and angiogenic differentiation ca-
pacity of bone marrow stem cells in vitro. Biomed Res
Int 2014: 596023.
Ginty PJ, Barry JJ, White LJ et al. 2008; Controlling protein
release from scaffolds using polymer blends and compos-
ites. Eur J Pharm Biopharm 68: 82–89.
Goto T, Nakao K, Gunjigake KK et al. 2007; Substance P
stimulates late-stage rat osteoblastic bone formation
through neurokinin-1 receptors. Neuropeptides 41:
25–31.
Guzman-Morales J, El-Gabalawy H, Pham MH et al. 2009;
Effect of chitosan particles and dexamethasone on hu-
man bone marrow stromal cell osteogenesis and angio-
genic factor secretion. Bone 45: 617–626.
Ham W. 1930; A histological study of the early phases of
bone repair. J Bone Joint Surg Am 12: 827–844.
Hong HS, Lee J, Lee E et al. 2009; A new role of substance P
as an injury-inducible messenger for mobilization of
CD29(+) stromal-like cells. Nat Med 15: 425–435.
Hong HS and Son Y. 2014; Substance P ameliorates collagen
II-induced arthritis in mice via suppression of the inflam-
matory response. Biochem Biophys Res Commun 453:
179–184.
Huiskes R, Ruimerman R, Van Lenthe GH et al. 2000; Effects
of mechanical forces on maintenance and adaptation of
form in trabecular bone. Nature 405: 704–706.
Jiang MH, Chung E, Chi GF et al. 2012; Substance P induces
M2-type macrophages after spinal cord injury.
Neuroreport 23: 786–792.
Kanczler JM, Ginty PJ, Barry JJ et al. 2008; The effect of mes-
enchymal populations and vascular endothelial growth fac-
tor delivered from biodegradable polymer scaffolds on
bone formation. Biomaterials 29: 1892–1900.
Kim CJ. 1998; Effects of drug solubility, drug loading, and
polymer molecular weight on drug release from Polyox
tablets. Drug Dev Ind Pharm 24: 645–651.
Copyright © 2017 John Wiley & Sons, Ltd.
Conflict of interest
We declare that we have no conflicts of interest in the
authorship or publication of this contribution.
Kim JH, Jung Y, Kim SH et al. 2011; The enhancement of
mature vessel formation and cardiac function in in-
farcted hearts using dual growth factor delivery with
self-assembling peptides. Biomaterials 32: 6080–6088.
Kim JH, Jung Y, Kim BS et al. 2013a; Stem cell recruitment
and angiogenesis of neuropeptide substance P coupled
with self-assembling peptide nanofiber in a mouse hind
limb ischemia model. Biomaterials 34: 1657–1668.
Kim SH, Jung Y, Kim SH et al. 2013b; A biocompatible tissue
scaffold produced by supercritical fluid processing for
cartilage tissue engineering. Tissue Eng Part C Meth 19:
181–188.
Kim SH, Hur W, Kim JE et al. 2015a; Self-assembling pep-
tide nanofibers coupled with neuropeptide substance P
for bone tissue engineering. Tissue Eng Part A 21:
1237–1246.
Kim SH, Jung Y, Kim SH et al. 2015b; TGF-beta encapsu-
lated PLCL scaffold by a supercritical CO-HFIP co-
solvent system for cartilage tissue engineering. J Control
Release 206: 101–107.
Kohara H, Tajima S, Yamamoto M et al. 2010; Angiogenesis
induced by controlled release of neuropeptide substance
P. Biomaterials 31: 8617–8625.
La WG, Jin M, Park S et al. 2014; Delivery of bone mor-
phogenetic protein-2 and substance P using graphene
oxide for bone regeneration. Int J Nanomedicine 9:
107–116.
Lee SH and Shin H. 2013; Matrices and scaffolds for deliv-
ery of bioactive molecules in bone and cartilage tissue
engineering. Adv Drug Deliv Rev 59: 339–359.
Lim J, Lee J, Yun H-S et al. 2013; Comparison of bone regen-
eration rate in flat and long bone defects: calvarial and
tibial bone. J Tissue Eng Regen Med 10: 336–340.
Liu Y, Chen Y, Zhao H et al. 2011; Effects of different doses
of dexamethasone on bone qualities in rats. J Biomed
Eng, 28: 737–743, 747.
Madeddu P. 2005; Therapeutic angiogenesis and
vasculogenesis for tissue regeneration. Exp Physiol 90:
315–326.
Mao D, Edwards JR, Harvey A et al. 2006; Prediction of
foam growth and its nucleation in free and limited ex-
pansion. Chem Eng Sci, 61: 1836–1845.
Mazziotti G, Giustina A, Canalis E et al. 2007; Glucocorti-
coid-induced osteoporosis: clinical and therapeutic as-
pects. Arq Bras Endocrinol Metabol 51: 1404–1412.
Mikami Y, Lee M, Irie S et al. 2011; Dexamethasone modu-
lates osteogenesis and adipogenesis with regulation of
osterix expression in rat calvaria-derived cells. J Cell Phys-
iol 226: 739–748.
S. H. Kim et al.
Pargaonkar N, Lvov YM, Li N et al. 2005; Controlled release
of dexamethasone from microcapsules produced by poly-
electrolyte layer-by-layer nanoassembly. Pharm Res 22:
826–835.
Raisz LG. 1999; Physiology and pathophysiology of bone re-
modeling. Clin Chem 45: 1353–1358.
Seegers HC, Hood VC, Kidd BL et al. 2003; Enhancement of
angiogenesis by endogenous substance P release and
neurokinin-1 receptors during neurogenic inflammation.
J Pharmacol Exp Ther 306: 8–12.
Noh SS, Bhang SH, La WG et al. 2015; A dual delivery of
substance P and bone morphogenetic protein-2 for mes-
enchymal stem cell recruitment and bone regeneration.
Tiss Eng Part A 21: 1275–1287.
Spicer PP, Kretlow JD, Young S et al., 2012; Evaluation of
bone regeneration using the rat critical size calvarial de-
fect. Nat Protoc 7: 1918–1929.
Spiro AS, Beil FT, Schinke T et al. 2010; Short-term applica-
tion of dexamethasone enhances bone morphogenetic
protein-7-induced ectopic bone formation in vivo. J
Trauma 69: 1473–1480.
Szpalski C, Barr J, Wetterau M et al. 2010; Cranial bone de-
fects: current and future strategies. Neurosurg Focus 29:
E8.
Tai H, Mather ML, Howard D et al. 2007; Control of pore
size and structure of tissue engineering scaffolds pro-
duced by supercritical fluid processing. Eur Cell Mater
14: 64–77.
Tampieri A, Sandri M, Landi E et al. 2008; Design of graded
biomimetic osteochondral composite scaffolds. Biomate-
rials 29: 3539–3546.
Tavakoli-darestani R, Manafi-rasi A, Kamrani-rad A et al.
2014; Dexamethasone-loaded hydroxyapatite enhances
bone regeneration in rat calvarial defects. Mol Biol Rep
41: 423–428.
Tenenbaum HC, Heersche JN. 1985; Dexamethasone stimu-
lates osteogenesis in chick periosteum in vitro. Endocri-
nology 117: 2211–2217.
Tomasko DL, Burley A, Feng L et al. 2009; Development of
CO2 for polymer foam applications. J Supercrit Fluids
47: 493–499.
Umemura Y, Nagasawa S, Honda A et al. 2008; High-impact
exercise frequency per week or day for osteogenic re-
sponse in rats. J Bone Miner Metab 26: 456–460.
Zhang WYP, Dai Q, Fang B et al. 2013; p38-MAPK signaling
pathway is not involved in osteogenic differentiation
during early response of mesenchymal stem cells to con-
tinuous mechanical strain. Mol Cell Biochem 378:
19–28.
J Tissue Eng Regen Med 2017; 11: 3469–3480.
DOI: 10.1002/term