Colloids and Surfaces B: Biointerfaces 146 (2016) 632–641

Contents lists available at ScienceDirect

Colloids and Surfaces B: Biointerfaces

journal homepage: www.elsevier.com/locate/colsurfb

A comparison of nanoscale and multiscale PCL/gelatin scaffolds

prepared by disc-electrospinning
Dawei Li a,b , Weiming Chen b , Binbin Sun b , Haoxuan Li a , Tong Wu b , Qinfei Ke a,c,∗∗ ,
Chen Huang a , Hany EI-Hamshary d,e , Salem S. Al-Deyab d , Xiumei Mo a,b,∗
a
Engineering Research Center of Technical Textiles, College of Textiles, Donghua University, Shanghai 201620, China
b
State Key Lab. for Modification of Chemical Fibers & Polymer Materials, College of Chemistry, Chemical Engineering and Biotechnology, Donghua
University, Shanghai 201620, China
c
Shanghai Normal University, Shanghai 200234, China
d
Department of Chemistry, College of Science, King Saud University, Riyadh 11451, Saudi Arabia
e
Department of Chemistry, Faculty of Science, Tanta University, Tanta 31527, Egypt

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

Article history: Electrospinning is a versatile and convenient technology to generate nanofibers suitable for tissue engi-
Received 3 February 2016 neering. However, the low production rate of traditional needle electrospinning hinders its applications.
Received in revised form 23 May 2016 Needleless electrospinning is a potential strategy to promote the application of electrospun nanofiber in
Accepted 4 July 2016
various fields. In this study, disc-electrospinning (one kind of needleless electrospinning) was conducted
Available online 7 July 2016
to produce poly(␧-caprolactone)/gelatin (PCL/GT) scaffolds of different structure, namely the nanoscale
structure constructed by nanofiber and multiscale structure consisting of nanofiber and microfiber. It
Keywords:
was found that, due to the inhomogeneity of PCL/GT solution, disc-electrospun PCL-GT scaffold pre-
Disc-electrospinning
Three-dimensional scaffolds sented multiscale structure with larger pores than that of the acid assisted one (PCL-GT-A). Scanning
Acetic acid electron microscopy images indicated the PCL-GT scaffold was constructed by nanofibers and microfibers.
Multiscale structure Mouse fibroblasts and rat bone marrow stromal cells both showed higher proliferation rates on multiscale
Nanofiber scaffold than nanoscale scaffolds. It was proposed that the nanofibers bridged between the microfibers
enhanced cell adhesion and spreading, while the large pores on the three dimensional (3D) PCL-GT scaf-
fold provide more effective space for cells to proliferate and migrate. However, the uniform nanofibers
and densely packed structure in PCL-GT-A scaffold limited the cells on the surface. This study demon-
strated the potential of disc-electrospun PCL-GT scaffold containing nanofiber and microfiber for 3D
tissue regeneration.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction
Tissue engineering is a promising approach for the repair of
tissue defects [1]. An effective scaffold selected for tissue regener-
ation should provide suitable micro-environment for cell adhesion
and proliferation [1]. Various types of scaffolds have been invented
to meet the demand of tissue regeneration, among which fibrous
scaffolds are widely used due to their three dimensional structure
∗ Corresponding author at: State Key Lab. for Modification of Chemical Fibers &
Polymer Materials, College of Chemistry, Chemical Engineering and Biotechnology,
Donghua University, Shanghai 201620, China.
∗∗ Corresponding author at: Engineering Research Center of Technical Textiles,
College of Textiles, Donghua University, Shanghai 201620, China
E-mail address: xmm@dhu.edu.cn (X. Mo).
(3D), high porosity, and interconnected pores which could facilitate
cellular infiltration and transport of nutrition and waste [2,3].
Electrospinning has gained popularity in the fabrication of
nanofibrous scaffolds from diverse natural and synthetic materi-
als, owning to its simplicity and cost-effectiveness. The nanoscale
structure of electrospun nanofiber mimics the structure of native
extracellular matrix (ECM), which is crucial for cell adhesion,
spreading, and proliferation [4]. Thus, electrospun nanofiber scaf-
folds have been widely applied in bone, skin, vascular, nerve, and
cartilage tissue regeneration [2,5–8]. However, it is accepted that
the highly compacted nanofibers may often result in increased cell
spreading and limited cellular infiltration, while cells require full
context of 3D matrix to maintain their morphology and estab-
lish natural behavior [3,9]. Thus, different approaches have been
invented to obtain electrospun structure with large pore [6,10–13].
Kidoaki et al. attempted to remove part of the component in a mul-

http://dx.doi.org/10.1016/j.colsurfb.2016.07.009
0927-7765/© 2016 Elsevier B.V. All rights reserved.
 D. Li et al. / Colloids and Surfaces B: Biointerfaces 146 (2016) 632–641 633

tilayered bicomponent scaffold to increase the pore size [14]. Salt

leaching was also applied to create large pores [15]. However, the
structure of scaffold would be affected by the leaching out pro-
cess. A second method was to increase the diameter of electrospun
fibers to several micrometers, leading to larger pore size and higher
porosity [16], but the method also led to a negative effect on cell
attachment and proliferation.
Recently, combination of nanofibers and microfibers became a
promising alternative for fabrication a 3D scaffold. The microfibers
formed a porous structure large enough for cell migration through
the scaffold, whereas the nanofibers could enhance protein absorp-
tion, improving protein-mediated interaction with the cell surface
so promoting cell adhesion and improving cell viability [17]. Soli-
man et al. electrospun scaffolds of microscale, nanoscale, and
multiscale fibers [18]. In vitro experiment showed the multiscale
scaffold boosted cell motility, survival, and proliferation, indicat-
ing that the 3D multiscale scaffold could allow different cell types
to coexist for the study of cell therapy of complex tissues. Pham
et al. manufactured bilayered constructs consisting of microfiber
scaffold with various thickness of nanofibers [19]. The presence
of nanofiber could enhance cell spreading. However, the layered
structure need a two step process, and the nanofibrous layers
always show a sheet-like structure over microstructure. To solve
this problem, Levorson et al. used a duel spinneret system to elec-
trospin different scale fibers and a cylindrical rotating mandrel to
collect the uniformly mixed multiscale scaffold [20]. It was reported
that the presence of nanoscale fiber could promote the produc-
tion and distribution of glycosaminoglycan, which revealed that
the scaffold was more closely resembling native ECM components.
An alternative strategy is demonstrated in this work in which Fig. 1. PCL/gelatin solution with and without acid. PCL/gelatin and PCL/gelatin/acid
needleless electrospinning is applied to generate 3D scaffold with solution left stand for 0 H (a) and 4 H (b). (c) Schematic of disc-electrospinning
multiscale structure. setup, consisting of a rotating metal disc spinneret, a solution vessel, a high-voltage
To fabricate a scaffold that can properly mimic the native ECM, direct-current power generator supply and a rotary collector.

the components always play a vital role. Synthetic materials pos-

sesse controllable properties, excellent mechanical strength but
poor biocompatibility, while natural materials could make up the
drawbacks and promote cell attachment and proliferation but weak
in mechanical properties [21]. Needle electrospinning was used to
produce Poly(␧-caprolactone) (PCL)/gelatin nanofiber to construct
tissue engineering scaffold [22–24]. The bicomponent nanofiber
possessed the merits of both PCL and gelatin, creating a favorable
environment for cell attachment and proliferation.
In the present study, PCL/gelatin fibrous scaffolds were fabri-
cated via disc-electrospinning. An increased fiber diameter and a
looser structure was achieved when acetic acid was absent in the
electrospinning solution. Fiber morphology, pore size and mechan-
ical properties of the two types of scaffolds were compared. FTIR
and XRD were employed to assess the compositional and crys-
talline characters. Fibroblasts and bMSCs were cultured to study
the biocompatibility.
2. Materials and experiments
DEMO Medical Tech. Co. Ltd. (China). All chemicals were used with-
out further purification.
2.2. Disc-electrospinning
2.5 g PCL and 2.5 g porcine gelatin was dissolved in 50 mL
TFE to generate 10% w/v PCL/gelatin solution. 500 ␮L acetic
acid was added into another 50 mL PCL/gelatin solution to form
10% PCL/gelatin/acid solution. Disc-electrospinning was conducted
using a homemade needleless electrospinning setup described
previously (Fig. 1(c)) [25]. Briefly, the disc-electrospinning setup
consists a rotating metal disc spinneret, a solution vessel, a
high-voltage direct-current power generator supply (Gamma High
Voltage Research, USA) and a rotary collector. The applied voltage,
collecting distance, and the rotating speed of the disc was set as
50 kV, 25 cm, and 10 rpm, respectively. After disc-electrospinning,
the collected nanofiber was kept in vacuum at room temperature
for 72 h to remove the solvent residue.

2.1. Materials
PCL (average Mw 80,000) was purchased from Sigma-Aldrich
(USA). 2,2,2-Trifluoroethanol (TFE) was purchased from Shanghai
Darui Finechemical Co., Ltd. (China). Porcine gelatin was pur-
chased from MP Biomedicals, LLC (France). Acetic acid of analytical
reagent was supplied by Kunshan JingKe Microelectronics Mate-
rial Co., Ltd. (China). The culture medium and related reagents
were purchased from Gibco Life Technologies Co., Ltd. (USA). 1-(3-
Dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC)
and N-Hydroxysuccinimide (NHS) was purchased from Shanghai
2.3. EDC/NHS crosslinking
Before in vitro experiments, EDC/NHS solution was used to
crosslink the gelatin in PCL/gelatin scaffolds. 5.76 g EDC and 3.75 g
NHS was dissolved in 200 mL of 95% ethanol solution. The final
concentration of EDC and NHS was 0.15 mol−1 and 0.16 mol−1 ,
respectively. The crosslinking process was conducted by immers-
ing the PCL/gelatin scaffolds in the EDC/NHS solution at 4 ◦ C for
12 h. After the crosslinking process, the scaffolds were rinsed with
75% ethanol solution for three time, and dried for 48 h at 37 ◦ C.
634 D. Li et al. / Colloids and Surfaces B: Biointerfaces 146 (2016) 632–641
2.4. Characterizations
Fiber morphology of the disc-electrospun PCL/Gelatin nanofiber
was examined by scanning electron microscopy (SEM, TM 3000,
Hitachi, Japan). The disc-electrospun nanofiber was briefly sputter-
coated with gold for 30 s. Images were acquired using an SEM
operating at an accelerating voltage of 15 kV. Diameter of fibers
were measured based on the SEM figures by image visualization
software Image J (National Institutes of Health, USA). 100 fibers
were measured for each sample.
The pore size of the scaffolds was determined as described
previously [25]. Briefly, the PCL/gelatin scaffolds was cut into
30 mm × 30 mm squares, which were wetted by the wetting agent,
calwick (PMI Porous Materials Int.), with a surface tension of
21 dyn cm−1 . Subsequently, a CFP-1100-AI capillary flow porom-
eter (PMI Porous Materials Int.) was used to determine the pore
size and pore distribution.
Tensile properties of disc-electrospun nanofiber were ana-
lysed by stressing uniaxially until the scaffolds failed. Various
disc-electrospun PCL/gelatin nanofiber scaffolds (before and after
crosslinking) were tested (n = 5). All samples were cut into same
size of 50 mm × 10 mm and tested using a universal materials tester
(H5 K-S, Hounsfield, UK) with a 50 N load cell at ambient temper-
ature of 20 ◦ C and humidity of 65%. The effective length of the test
was 30 mm with a stretching speed of 10 mm/min.
Chemical analysis of PCL/gelatin nanofiber scaffolds (before and
after crosslinking) was carried out by Fourier transform infrared
spectroscopy (FTIR). A FTIR spectrophotometer (Avatar380, USA)
was used to obtain FTIR spectra of the disc-electrospun PCL/gelatin
nanofiber scaffolds over the range of 500–4000 cm−1 at scanning
resolution of 2 cm−1 .
X-ray diffraction spectroscopy was conducted on a D/max-2550
PC (Rigaku, Japan) with Cu K␣ radiation. The operating voltage
and current was set as 40 kV and 200 mA. The disc-electrospun
PCL/gelatin nanofiber scaffolds were examined between 0 and
60◦ (2␪) at a scanning rate of 1◦ per minute.
2.5. In vitro cell culture
Mouse fibroblasts (L929) were supplied by the Institute of Bio-
chemistry and Cell Biology of the Chinese Academy of Sciences
(Shanghai, China). L929 were maintained and expanded in high
glucose Dulbecco’s modified eagle medium (DMEM) medium con-
taining 10% fetal serum, 100 units/mL penicillin and 100 units/mL
streptomycin, incubated in humidified atmosphere with 5% CO2 at
37 ◦ C. The culture medium was refreshed every other day.
Rat bone marrow stromal cells (bMSCs) were isolated using
density gradient centrifugation as described previously [26]. The
isolated bMSCs were maintained and expanded in DMEM/F12
medium containing 10% fetal serum, 100 units/mL penicillin and
100 units/mL streptomycin. The culture medium was refreshed
every other day.
The in vitro biocompatibility assessment was performed as
described previously [25]. L929 and bMSCs were seeded on the pre-
pared scaffolds. For L929, 2 × 104 cells were seeded on the scaffolds
with tissue culture plates (TCPs) as control. The culture medium
was refreshed every other day. After 1, 3, 5 and 7 days of culture,
the culture medium was removed and the samples were rinsed
with PBS buffer for three times to remove the unattached cells and
the culture medium. Subsequently, the amount of cells was deter-
mined by the standard MTT assay described previously [25]. For
bMSCs, 10 × 104 cells were seeded on the scaffolds, and the MTT
assay was performed on 3, 7, 14 days post seeding. The experiment
was conducted with 6 parallel samples.
Cell morphology on the scaffolds was determined followed the
method reported by Wu et al. [27]. Briefly, cells cultured on the scaf-
folds at predetermined time were fixed, dehydrated, and followed
by freeze-drying for 24 h. Then the samples were sputter-coated
with gold and visualized by SEM.
Confocal laser microscopy imaging was applied to observe the
distribution and morphology of cells on the scaffolds [25]. Cells on
the scaffolds were fixed and the cytoskeletons and nuclei of cells
were stained with 25 ␮g mL−1 rhodamine-conjugated phalloidin
(Invitrogen, USA) and 10 ␮g mL−1 4 ,6 -diamidino-2-phenylindole
hydrochloride (DAPI, Invitrogen, USA) for 30 min and 5 min, respec-
tively. Subsequently, the cells were observed using confocal laser
scanning microscopy (CLSM) (Carl Zeiss, LSM 700, Germany).
Cell infiltration was analysed with z-stacked confocal
microscopy. For each sample, 20 slices were photoed at one
field of vision by CLSM. Thereafter, imaging software ZEN 2008
(Carl Zeiss, Germany) was employed to provide 3D views of the
stained cells growing on the scaffolds using depth coding.
2.6. Statistical analysis
All the data were obtained at least in triplicate and all values
were presented as the mean and standard deviation (SD). Statistical
analysis was performed by the one-way analysis of variance using
Origin 8.0 (OriginLab Inc., USA). The statistical difference between
two sets of data was considered when * p < 0.05, ** p < 0.01, and ***
p < 0.001.
3. Results
3.1. Phase separation and disc-electrospinning
Although both PCL and gelatin could be dissolved in TFE to
form transparent solution, when being mixed together, the mixed
PCL/gelatin solution transformed to emulsion and became opaque.
Phase separation occurred and two different layer of solution was
formed after left to stand for hours [28]. The addition of acetic
acid could transform the opaque solution clear again. The com-
parison of the two kinds of solution was shown in Fig. 1(a). After
4 h of standing, the PCL/gelatin solution became layered, while
the PCL/gelatin/acid solution was still transparent (Fig. 1(b)). Feng
et al. confirmed that in the bottom layer, gelatin was dominant,
while PCL was mainly in the top layer [28]. In the initial stage of
the phase separation, PCL phase was continuous in the PCL/gelatin
solution while the gelatin was micro balloon sphere dispersed in
the PCL phase. The gelatin sphere gradually increased and aggre-
gated which resulted in layered solution after hours. As traditional
electrospinning always lasts for several hours, phase separation
affects the morphology and composition of generated fiber. Disc-
electrospinning always last for only several minutes, during which
the properties of the solution will not change evidently. The prop-
erties of obtained fiber would keep same during the process. During
disc-electrospinning, the production for PCL-GT is 9.47 ± 0.56 g h−1
and for PCL-GT A is 9.80 ± 0.55 g h−1 .
3.2. Morphology of disc-electrospun nanofiber
Fig. 2 illustrates the morphology and micro structure of the disc-
electrospun scaffolds. It could be seen that the PCL-GT scaffold
was fluffy and loosely packed while the PCL-GT-A scaffold showed
a dense appearance (Fig. 2(a–c)). The average diameter of fibers
for PCL-GT and PCL-GT-A was 592.3 ± 598.0 and 316.1 ± 163.7 nm,
respectively. The diameter of fibers in PCL-GT scaffolds was ran-
domly ranged from ∼100 nm to several micron (Fig. 2(d), (e) and
(h)). The pores between fibers were much larger than those of
PCL-GT-A. The microfibers built up the backbone, constructing the
support of large pore. The nanofibers formed webs and bridged
between the microfibers, filling into the large pores (Fig. 2(e)). In the
 D. Li et al. / Colloids and Surfaces B: Biointerfaces 146 (2016) 632–641 635

Fig. 2. (a) Appearance of disc-electrospun PCL-GT and PCL-GT-A scaffold. Side view of the PCL-GT (b) and PCL-GT-A (c) scaffold. SEM images of PCL-GT (d), (e) and PCL-GT-A
(f), (g) scaffold. Fiber size distribution of PCL-GT (h) and PCL-GT-A (i) scaffold.

PCL-GT-A scaffold, the diameter of fibers was thinner and showed
a smaller distribution, resulted in a flat and densely packed film
(Fig. 2(f), (g) and (i)).
Feng et al. reported that the addition of acetic acid could change
the solution properties by reducing the pH value, surface tension
and viscosity, and increasing the conductivity [28]. Higher surface
tension and viscosity would work against the electric field force
and thus restrict the stretching, resulting in thicker fiber diameter
in PCL-GT scaffold. And the increase of conductivity also facilitated
the stretching to avoid the microfibers [29]. Moreover, the inhomo-
geneity and phase separation of the solution caused the uniformity
of the surface of polymer jets. As the PCL phase was continuous and
636 D. Li et al. / Colloids and Surfaces B: Biointerfaces 146 (2016) 632–641
Fig. 3. Pore size distribution of PCL-GT (a) and PCL-GT-A (b) scaffold. The average pore diameter for PCL-GT and PCL-GT-A scaffolds was 13.39 ␮m and 2.40 ␮m, respectively.
the gelatin formed particles dispersed in the PCL phase. Current
experimental phenomena suggested that PCL occupied the body
of the polymer jet generated from the disc, while the gelatin part
formed spherical droplets inserted in the PCL polymer jets, causing
the undulating surface of the electrified jet. Forced by the surface
tension, electrostatic and coulombic repulsion force, gelatin might
be generated into branches from the jet body. After stretching,
the branches were collected and form nanofiber web between the
microfibers. It was reported that gelatin in TFE solution was neg-
atively charged, which may also caused the formation of ultrafine
fiber of ∼100 nm.
3.3. Physical property of disc-electrospun nanofiber
Fig. 3 illustrates the pore size distribution in two scaffolds. It
could be seen in the PCL-GT scaffold, the diameter of pores under-
went a wider distribution ranging from 9 ␮m to 19 ␮m and the
average diameter was 13.39 ␮m. In the PCL-GT-A scaffold, the aver-
age pore diameter was 2.40 ␮m, concentrating between 1.7 ␮m and
2.9 ␮m. The large pores were constructed by the loosely distributed
microfibers, indicated in Fig. 2(d) and (e). While the densely packed
nanofibers formed a compact structure, reducing the pore size as
in Fig. 2(f) and (g).
The mechanical property was characterized by stretching 5
specimens of each scaffold. Fig. 4(a) show the stress-strain behavior
of the disc-electrospun PCL-GT and PCL-GT-A scaffold. The aver-
age tensile strength, strain at break, and Young’s modulus were
illustrated in Fig. 4(b)–(d), respectively. It could be seen from these
results that the PCL-GT-A scaffold showed higher tensile strength
and Young’s modulus than those of PCL-GT scaffold. However, it
was not statistically different for their strain at break. In the PCL-
GT scaffold, the density of fibers was less than that of PCL-GT-A
scaffold. Therefore, the contact points for a single fiber with other
fibers were also less in PCL-GT scaffold, which reduced the mechan-
ical strength and increased the motion facility. Thus, the PCL-GT
scaffold showed weaker strength and Young’s modulus.
ATR-FTIR and XRD results were shown in Fig. 4e and f. FTIR
spectra for the two scaffolds were just the same. The PCL related
stretching modes were observed in PCL-GT and PCL-GT-A scaffolds,
including 2946 cm−1 (asymmetric CH2 stretching), 2868 cm−1
(symmetric CH2 stretching), 1727 cm−1 (carbonyl stretching), and
1293 cm−1 (C O and C C stretching in the crystalline phase) [30].
Typical protein related bonds were found at 1652 cm− 1 (amide I)
and 1549 cm−1 (amide II), corresponding to C O stretching vibra-
tion, and coupling of bending of N H bond and stretching of C N
bond, respectively [31]. The addition of acetic acid showed no influ-
ence in the components in the disc-electrospun PCL/gelatin fibers.
XRD results suggested that, compared to the PCL-GT scaffold,
the characteristic diffraction peaks of PCL at 21.56 and 23.84◦ were
significantly sharpened in the PCL-GT-A scaffold, indicating a much
higher crystallinity. During needle electrospinning, the addition of
acetic acid could improve the miscibility of PCL and gelatin solu-
tion. Thus the better miscibility of blended nanofibers decrease
the crystallinity [28]. However, in the disc-electrospun PCL-GT and
PCL-GT-A scaffolds a reverse result was obtained. The fibers in PCL-
GT-A were thinner than that in PCL-GT scaffold, demonstrating
a stronger stretching was performed during disc-electrospinning.
Tim et al. found that the smaller diameter fibers was resulted from
sufficient crystallization of polymer chain before the polymer jets
reaches the collector, along with extra elongation of the jets [32].
3.4. Cell proliferation on scaffolds
Biocompatibility such as cell proliferation, adhesion, spreading,
and migration on the scaffolds is an important aspect to evaluate
the fitness for tissue regeneration. For tissue engineering scaffold,
surface properties and structure are major factors in regulating cell
behavior [9]. L929 cells and BMSCs were cultured on two types
of scaffolds to determine the biocompatibility. As the proliferation
rate of BMSCs was much slower than that of L929, the culturing
periods were set as 1, 3, 7 days and 3, 7, 14 days for L929 and
BMSCs, respectively. MTT assay was conducted at predetermined
time point and the result was illustrated in Fig. 5. In Fig. 5(a), it
could be found that during the 7 day long culturing, cells on PCL-
GT-A scaffold were less than that on PCL-GT. After 3 days of culture,
L929 on TCP were higher than on PCL-GT scaffold. Interestingly, the
trend was reversed after 7 days of culture. Similar trend occurred
to the BMSCs during the 14 days culturing period (Fig. 5(b)). BMSCs
cultured on the PCL-GT scaffolds showed significant increase over
time and performed the fastest growth rate throughout the cultur-
ing time. After 14 days, BMSCs on the PCL-GT scaffolds increased
from 100 thousand to over 500 thousand, which was twice as many
as that on PCL-GT-A (about 250 thousand).
These results indicated that PCL-GT scaffold with large pores
performed better than the PCL-GT-A scaffold with small pore in
long period cell proliferation. In our previous study, L929 showed
high proliferation rate on the disc-electrospun PCL scaffold with 3D
structure [25]. It was concluded that the large pores in the PCL-GT
scaffold provided another direction for cell spreading and migra-
tion [33]. The interconnected pores could promote the nutrition
transportation and the removal of metabolic waste.
 D. Li et al. / Colloids and Surfaces B: Biointerfaces 146 (2016) 632–641 637

Fig. 4. Mechanical properties of PCL-GT and PCL-GT-A scaffold. (a) Stress-strain curve, (b) Tensile strength, (c) Stress at break, and (d) Young’s modulus. FTIR spectra (a) and
X-ray diffraction (b) of PCL-GT and PCL-GT-A scaffolds.

Fig. 5. Cell proliferation of L929 (a) and BMSCs (b) on PCL-GT, PCL-GT-A scaffolds and TCP quantitatively measured by MTT assay. * indicates statistically difference for
p < 0.05; ** indicates statistical difference for p < 0.01; *** indicates statistical difference for p < 0.001.

3.5. Morphology of cells on scaffolds
The structure of scaffolds was treated as a priority for cellular
colonization in tissue regeneration, which would affect the inter-
action of cells with the scaffolds. The morphology of cells on two
scaffolds was observed by SEM (Fig. 6). L929 on both scaffolds
showed a significantly increase over time from day 1 to day 7. On the
first day after cell seeding, cells adhered to the fibers and formed a
typical spindle shape (Fig. 6(a), (a’), (d), (d’)). On the PCL-GT scaffold,
cells were thinner than that on PCL-GT-A scaffold. L929 cells was
found not only on the surface, but also inside the scaffold. Three
days later, number of cells in the scaffolds increased. Some cells
have migrated into the PCL-GT scaffolds, as labeled by the white
arrows in Fig. 6(b), (b’). However, for the PCL-GT-A scaffolds, the
pore size between fiber was much smaller than cell, restricting the
cells on the surface (Fig. 6(e), (e’)).
After 7 days of culture, L929 on PCL-GT-A scaffolds further
spread and connected with each other, forming a thin layer
(Fig. 6(f), (f’)). While for the PCL-GT scaffold, cells could spread in
the thickness direction. Most cells kept single and distributed dis-
638 D. Li et al. / Colloids and Surfaces B: Biointerfaces 146 (2016) 632–641

Fig. 6. SEM images of L929 on two scaffolds. L929 cultured on PCL-GT scaffolds after 1 day (a), (a’), 3 days (b), (b’), and 7 days (c), (c’). L929 cultured on PCL-GT-A scaffolds
after 1 day (d), (d’), 3 days (e), (e’), and 7 days (f), (f’) (Bar: 100 ␮m).

persedly (Fig. 6(c), (c’)). Only part of the surface was covered by
cells, leaving efficient space for further proliferation. Round cells
were observed on PCL-GT-A scaffolds in Fig. 6(f), (f’) (indicated by
white circles). This suggested that the insufficient living space lim-
ited the spreading of cells. In contrast, most cells on the PCL-GT
scaffolds kept the spindle shaped or starlike, which was considered
as a signal of health.
The morphology of cells was also observed by LSCM (Fig. 7).
On the first day, cells on both scaffolds showed spindle or starlike
shapes, indicating normal spreading (Fig. 7(a), (d)). After 3 days,
the whole surface of the PCL-GT-A scaffold was occupied by cells
(Fig. 7(e)). On the 7th day, part of the cells turned round (Fig. 7(f)),
which was caused by the lack of living space. In contrast, after
7 days, gaps between cells could still be observed on the PCL-GT
scaffold (Fig. 7(c)). The spindle shaped cells were dominant, demon-
strating the potential for further proliferation.
Fig. 7(g–j) illustrates the depth of fibroblasts cultured on PCL-GT
and PCL-GT-A scaffolds, respectively. The stained cells were pho-
tographed by LSCM in Z-stacked mode. The obtained images was
stacked up and translated into depth code where different colors
represent an increasing depth from red to blue. It could be seen that
cells could be observed in a depth of 140 ␮m from the upper sur-
face in the PCL-GT scaffold (Fig. 7(g) and (i)), which indicated that
cells had infiltrated into the scaffold for a 140 ␮m long distance.
While on the PCL-GT-A scaffold, only a thin layer of cells, ranging
from 20 ␮m to 70 ␮m (Fig. 7(h) and (j)), could be seen, meaning
that cells were restricted on the surface of the scaffold.
4. Discussion
Electrospinning has been widely used in the fabrication of
scaffolds with nanoscale structure, while the low production rat
 D. Li et al. / Colloids and Surfaces B: Biointerfaces 146 (2016) 632–641 639

Fig. 7. Confocal images of L929 cells on scaffolds. L929 on PCL-GT scaffolds after (a) 1 day, (b) 3 days, (c) 7 days of growth, on PCL-GT-A after (d) 1 day, (e) 3 days, (f) 7 days
of culture. Cell penetration on (g) PCL-GT scaffold, (h) PCL-GT-A scaffold, and side view of the cells distribution in PCL-GT (i) and PCL-GT-A scaffold (j). visualized by confocal
images in z-stack mode. Different colors represent an increasing depth from red to blue. (For interpretation of the references to colour in this figure legend, the reader is
referred to the web version of this article.)

of traditional needle electrospinning limit the application Gen-
erally, the production rate for the conventional electrospinning
varies from 0.1 to 1.0 g/h. The development of needleless elec-
trospinning raised an valid solution to scale up the production of
nanofiber. In this study, disc-electrospinning was used to fabri-
cate PCL/gelatin scaffold with homogeneous and inhomogeneous
PCL/gelatin blends. The properties of both scaffolds was char-
acterized and in vitro experiment was conducted to study the
biocompatibility.
Electrospinning PCL/gelatin blends into nanofiber has been
extensively studied in previous literatures [22–24,28]. Nanofiber
constructed by PCL and gelatin possessed the merits of both
synthetic and natural materials. Gelatin could improve the bio-
compatibility, while the PCL strengthen the mechanical properties.
However, the obtained PCL/gelatin scaffold was constructed by
fibers of nanoscale, which reduced the pore size to several microm-
eters between fibers in the scaffold, inhibiting the cells on the very
640 D. Li et al. / Colloids and Surfaces B: Biointerfaces 146 (2016) 632–641
Fig. 8. The architecture of electrospun nanofibrous scaffold affects the distribution and morphology of cells. (a) Cells adhere and along a single microfiber. (b) and (d) Cells
spread on the surface of nanofiber scaffold. (c) and (e) Cells adhere on both microfiber and nanofiber inserted in the microstructure.
surface. Thus the cells could only form a thin layer without migra-
tion into the scaffold.
On the other hand, to generate the nanofiber into a 3D scaffold
allowing cell infiltration, several methods were invented, includ-
ing salt leaching, freeze drying, water bath collecting. All these
methods generated large pore un-uniformly distributed in the scaf-
fold. Some parts formed voids, while the rest was still densely
packed nanofibers. Recently, the combination of nanofiber and
microfiber provide a promising method to solve the existing prob-
lems. Microfiber formed the backbone of the scaffold while the
nanofibers bridged between microfibers. Experiments had verified
that the multiscale scaffold could boost cell motility, survival, and
proliferation. However, the existing methods were based on needle
electrospinning, which always need a dual needle system.
Here, we compared the nanoscale and multiscale PCL/Gelatin
scaffold prepared via disc-electrospinning. Experiments has indi-
cated that phase separation would happen when PCL and gelatin
dissolved in TFE simultaneously. Gelatin would form micro spheres
distributed in the PCL solution. In contrast, the addition of acetic
acid could reduce the pH of the solution and thus eliminate the
phase separation. During disc-electrospinning, phase separation
would help generate multiscale fibers form the inhomogeneous
solution (Figs. 1 and 2). Compared with the uniform nanofiber
of PCL-GT-A scaffold (Fig. 2(f) and (g)), multiscale PCL-GT scaf-
fold shows larger pore size. The nanofiber of hundreds nanometer
could be found bridged between the microfibers, forming nets in
the pores (Fig. 2(d) and (e)). The pore size distribution verified the
difference (Fig. 3). It could be found that the pore size was signifi-
cantly raised in the PCL-GT scaffold. The pore size is quite important
for the cell colonization in the scaffold. Migration of cells into the
scaffold is crucial for the success of the graft and the overall healing
of defect. Greater pore size always leads to enhanced cell infiltra-
tion. However, the multiscale PCL-GT scaffold is loosely packed,
which reduces the mechanical strength and increased the motion
facility (Fig. 4). The addition of acetic acid have no influence on the
components but increased the crystallinity of the obtained fiber
(Fig. 4).
To study the biocompatibility, L929 and BMSCs were cultured
on the multiscale scaffold. Results indicated that cells culture on
PCL-GT scaffold showed a higher proliferation rate during the cul-
ture period, even better than the TCP (Fig. 5). As the culturing
period extended, the superiority of the multiscale scaffold became
more obvious. Which could attribute to the more space provided
by the large pore structure of the PCL-GT scaffold. Moreover, the
interconnected pore structure in the PCL-GT scaffold facilitated the
transport of nutrition and metabolic waste.
Scaffolds constructed by nanofiber with nanoscale architectures
were believed to have larger surface areas to absorb proteins, pre-
senting more binding sites for cell attachment [9]. Electrospun
nanofiber was widely applied in the fabrication of various tissue
engineering scaffolds. In present, it has been verified that cells
culture on microfiber scaffold always adhere and spread along a
single fiber (Fig. 8(a)) [18,34]. Without support, cells could not col-
onize in the space between fibers. Compared with the microfiber,
electrospun nanofiber could promote cells attachment, spreading,
and proliferation [19]. The high porosity would facilitate nutri-
tion transport and removal of the waste. However, as illustrated
in Fig. 8(b) and (d), the small pore size in the scaffold limited the
cell migration into the scaffold [5,35,36].
Nanofibers and microfibers were combined to form 3D struc-
ture with large pores and efficient nanoscale binding sites. Pham
et al. manufactured bilayered constructs consisting of microfiber
scaffold with various thickness of nanofibers [19]. The presence
of nanofiber could enhance cell spreading, while increasing the
thickness of nanofiber layer reduced cell infiltration. Soliman et al.
produced the multiscale 3D scaffolds containing fibers with aver-
age diameter of 3.3 ␮m and 0.6 ␮m via multimodal electrospinning
[18]. Results indicated that nanofibers properly inserted in the
microfiber scaffold could crucially enhance the interaction between
cells and scaffold (Fig. 8(c)).
In this study, multiscale scaffold constructed by PCL and
Gelatin was prepared via disc-electrospinning. In this scaffold, the
microfibers acted as the backbone and supported the 3D structure.
Nanofibers inserted among the microfiber framework, bridging
adjacent microfibers. As shown in Fig. 8(e), fibroblasts cultured on
the PCL-GT scaffold were dispersedly distributed in a 3D manner.
Some cells adhered and spread along a single microfiber (white
lines). Some cells partly attached to a microfiber and spread out
along nanofibers two types of adhesion points (white arrows). The
 D. Li et al. / Colloids and Surfaces B: Biointerfaces 146 (2016) 632–641
others adhered on the nanoweb constructed by nanofibers among
microfibers (hollow arrows).
5. Conclusion
We prepared multiscale PCL-GT scaffold containing nanofiber
and microfiber via disc-electrospinning. Due to the inhomogene-
ity of the PCL-GT in TFE solution, microfibers and nanofibers were
generated simultaneously and formed multiscale scaffolds with
large pores. Compared with the acetic acid assisted PCL-GT-A scaf-
fold constructed by uniform nanofibers, PCL-GT scaffold possessed
larger pores with a diameter of 13.39 ␮m. In vitro experiments indi-
cated that the L929 and BMSCs cultured on PCL-GT scaffold showed
higher proliferation rate. PCL-GT also enhanced the cell infiltration
into the scaffold while the nanofibrous structure of PCL-GT-A scaf-
fold inhibited the cell on the surface. In the PCL-GT scaffold, cells
seemed to settle and align on the microfibers but bridge between
microfibers through the nanoweb constructed by the nanoscale
fibers, distributing throughout the whole scaffold in a 3D manner.
Results in this study demonstrated the potential of disc-electrospun
PCL-GT scaffold with multiscale structure in the application of tis-
sue engineering.
Acknowledgements
This research was supported by National Natural Science Foun-
dation of China (31470941, 31271035, and 51403033), Science
and Technology Commission of Shanghai Municipality Program
(14JC1492100), Ph.D. Programs Foundation of Ministry of Edu-
cation of China (20130075110005) and light of textile project
(J201404), The authors would like to extend their sincere apprecia-
tion to the Deanship of Scientific Research at King Saud University
for its funding of this research through the research group project
NO. RGP-201, and the Fundamental Research Funds for the Central
Universities (CUSF-DH-D-2014011, and 2232014D3-15).
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.colsurfb.2016.07.
009.
References
[1] S. Yang, K.-F. Leong, Z. Du, C.-K. Chua, The design of scaffolds for use in tissue
engineering. Part I. Traditional factors, Tissue Eng. 7 (2001) 679–689.
[2] Q.P. Pham, U. Sharma, A.G. Mikos, Electrospinning of polymeric nanofibers for
tissue engineering applications: a review, Tissue Eng. 12 (2006) 1197–1211.
[3] J. Rnjak-Kovacina, A.S. Weiss, Increasing the pore size of electrospun scaffolds,
Tissue Eng. Part B: Rev. 17 (2011) 365–372.
[4] D.R. Nisbet, J.S. Forsythe, W. Shen, D.I. Finkelstein, M.K. Horne, Review paper:
a review of the cellular response on electrospun nanofibers for tissue
engineering, J. Biomater. Appl. 24 (2009) 7–29.
[5] W. He, Z. Ma, W.E. Teo, Y.X. Dong, P.A. Robless, T.C. Lim, S. Ramakrishna,
Tubular nanofiber scaffolds for tissue engineered small-diameter vascular
grafts, J. Biomed. Mater. Res. Part A 90 (2009) 205–216.
[6] P. Causin, R. Sacco, M. Verri, A multiscale approach in the computational
modeling of the biophysical environment in artificial cartilage tissue
regeneration, Biomech. Model. Mechanobiol. 12 (2013) 763–780.
[7] M.P. Prabhakaran, J.R. Venugopal, T.T. Chyan, L.B. Hai, C.K. Chan, A.Y. Lim, S.
Ramakrishna, Electrospun biocomposite nanofibrous scaffolds for neural
tissue engineering, Tissue Eng. Part A 14 (2008) 1787–1797.
[8] R.A. Franco, Y.K. Min, H.M. Yang, B.T. Lee, Fabrication and biocompatibility of
novel bilayer scaffold for skin tissue engineering applications, J. Biomater.
Appl. 27 (2013) 605–615.
[9] M.M. Stevens, J.H. George, Exploring and engineering the cell surface
interface, Science 310 (2005) 1135–1138.
[10] I.K. Shim, M.R. Jung, K.H. Kim, Y.J. Seol, Y.J. Park, W.H. Park, S.J. Lee, Novel
three-dimensional scaffolds of poly(l-lactic acid) microfibers using
electrospinning and mechanical expansion: fabrication and bone
regeneration, J. Biomed. Mater. Res. Part B Appl. Biomater. 95 (2010) 150–160.
641
[11] C. Vaquette, J. Cooper-White, A simple method for fabricating 3-D
multilayered composite scaffolds, Acta Biomater. 9 (2013) 4599–4608.
[12] C.R. Wittmer, A. Hébraud, S. Nedjari, G. Schlatter, Well-organized 3D
nanofibrous composite constructs using cooperative effects between
electrospinning and electrospraying, Polymer 55 (2014) 5781–5787.
[13] S.J. Kim, D.H. Jang, W.H. Park, B.-M. Min, Fabrication and characterization of
3-dimensional PLGA nanofiber/microfiber composite scaffolds, Polymer 51
(2010) 1320–1327.
[14] S. Kidoaki, I.K. Kwon, T. Matsuda, Mesoscopic spatial designs of nano-and
microfiber meshes for tissue-engineering matrix and scaffold based on newly
devised multilayering and mixing electrospinning techniques, Biomaterials
26 (2005) 37–46.
[15] J. Nam, Y. Huang, S. Agarwal, J. Lannutti, Improved cellular infiltration in
electrospun fiber via engineered porosity, Tissue Eng. 13 (2007) 2249–2257.
[16] Q.P. Pham, U. Sharma, A.G. Mikos, Electrospun poly(␧-caprolactone)
microfiber and multilayer nanofiber/microfiber scaffolds: characterization of
scaffolds and measurement of cellular infiltration, Biomacromolecules 7
(2006) 2796–2805.
[17] K. Shalumon, N. Binulal, M. Deepthy, R. Jayakumar, K. Manzoor, S. Nair,
Preparation, characterization and cell attachment studies of electrospun
multi-scale poly(caprolactone) fibrous scaffolds for tissue engineering, J.
Macromol. Sci. Part A 48 (2010) 21–30.
[18] S. Soliman, S. Pagliari, A. Rinaldi, G. Forte, R. Fiaccavento, F. Pagliari, O.
Franzese, M. Minieri, P. Di Nardo, S. Licoccia, E. Traversa, Multiscale
three-dimensional scaffolds for soft tissue engineering via multimodal
electrospinning, Acta Biomater. 6 (2010) 1227–1237.
[19] Quynh P. Pham, Upma Sharma, Antonios G. Mikos, Electrospun
poly(␧-caprolactone) microfiber and multilayer nanofiber/microfiber
scaffolds: characterization of scaffolds and measurement of cellular
infiltration, Biomacromolecule 7 (10) (2006) 2796–2805.
[20] E.J. Levorson, P. Raman Sreerekha, K.P. Chennazhi, F.K. Kasper, S.V. Nair, A.G.
Mikos, Fabrication and characterization of multiscale electrospun scaffolds
for cartilage regeneration, Biomed. Mater. 8 (2013) 014103.
[21] X. Wu, L. Black, G. Santacana-Laffitte, C.W. Patrick Jr., Preparation and
assessment of glutaraldehyde-crosslinked collagen-chitosan hydrogels for
adipose tissue engineering, J. Biomed. Mater. Res. Part A 81 (2007) 59–65.
[22] Y. Zhang, H. Ouyang, C.T. Lim, S. Ramakrishna, Z.M. Huang, Electrospinning of
gelatin fibers and gelatin/PCL composite fibrous scaffolds, J. Biomed. Mater.
Res. Part B: Appl. Biomater. 72 (2005) 156–165.
[23] W. Ji, F. Yang, J. Ma, M.J. Bouma, O.C. Boerman, Z. Chen, J.J. van den Beucken,
J.A. Jansen, Incorporation of stromal cell-derived factor-1alpha in PCL/gelatin
electrospun membranes for guided bone regeneration, Biomaterials 34 (2013)
735–745.
[24] L. Ghasemi-Mobarakeh, M.P. Prabhakaran, M. Morshed, M.-H. Nasr-Esfahani,
S. Ramakrishna, Electrospun poly(␧-caprolactone)/gelatin nanofibrous
scaffolds for nerve tissue engineering, Biomaterials (29 2008) 4532–4539.
[25] D. Li, T. Wu, N. He, J. Wang, W. Chen, L. He, C. Huang, H.A. Ei-Hamshary, S.S.
Al-Deyab, Q. Ke, X. Mo, Three-dimensional polycaprolactone scaffold via
needleless electrospinning promotes cell proliferation and infiltration,
Colloids Surf. B: Biointerfaces 121 (2014) 432–443.
[26] X. Li, Y. Zhang, G. Qi, Evaluation of isolation methods and culture conditions
for rat bone marrow mesenchymal stem cells, Cytotechnology 65 (2013)
323–334.
[27] J. Wu, S. Liu, L. He, H. Wang, C. He, C. Fan, X. Mo, Electrospun nanoyarn scaffold
and its application in tissue engineering, Mater. Lett. 89 (2012) 146–149.
[28] B. Feng, H. Tu, H. Yuan, H. Peng, Y. Zhang, Acetic-acid-mediated miscibility
toward electrospinning homogeneous composite nanofibers of GT/PCL,
Biomacromolecules 13 (2012) 3917–3925.
[29] D.H. Reneker, A.L. Yarin, Electrospinning jets and polymer nanofibers,
Polymer (Guildf) 49 (2008) 2387–2425.
[30] S.A. Catledge, W.C. Clem, N. Shrikishen, S. Chowdhury, A.V. Stanishevsky, M.
Koopman, Y.K. Vohra, An electrospun triphasic nanofibrous scaffold for bone
tissue engineering, Biomed. Mater. 2 (2007) 142–150.
[31] C.S. Ki, D.H. Baek, K.D. Gang, K.H. Lee, I.C. Um, Y.H. Park, Characterization of
gelatin nanofiber prepared from gelatin-formic acid solution, Polymer 46
(2005) 5094–5102.
[32] C.T. Lim, E.P.S. Tan, S.Y. Ng, Effects of crystalline morphology on the tensile
properties of electrospun polymer nanofibers, Appl. Phys. Lett. 92 (2008)
141908.
[33] K. Sisson, C. Zhang, M.C. Farach-Carson, D.B. Chase, J.F. Rabolt, Fiber diameters
control osteoblastic cell migration and differentiation in electrospun gelatin,
J. Biomed. Mater. Res. Part A 94 (2010) 1312–1320.
[34] Y. Tang, C. Wong, H. Wang, A. Sutti, M. Kirkland, X. Wang, T. Lin,
Three-dimensional tissue scaffolds from interbonded
poly(epsilon-caprolactone) fibrous matrices with controlled porosity, Tissue
Eng. Part C: Methods 17 (2011) 209–218.
[35] C. Vaquette, J.J. Cooper-White, Increasing electrospun scaffold pore size with
tailored collectors for improved cell penetration, Acta Biomater. 7 (2011)
2544–2557.
[36] A. Yin, K. Zhang, M.J. McClure, C. Huang, J. Wu, J. Fang, X. Mo, G.L. Bowlin, S.S.
Al-Deyab, M. El-Newehy, Electrospinning collagen/chitosan/poly(l-lactic
acid-co-epsilon-caprolactone) to form a vascular graft: mechanical and
biological characterization, J. Biomed. Mater. Res. Part A 101 (2013)
1292–1301.