Fabrication and evaluation of poly(lactic acid), chitosan, and tricalcium

phosphate biocomposites for guided bone regeneration

Srikanthan Ramesh ,1 Lisa Lungaro,2 Dimitrios Tsikritsis,2 Eric Weflen,1 Iris V. Rivero ,1
Alistair P. D. Elfick2
1
Department of Industrial and Manufacturing Systems Engineering, Iowa State University, Ames,, Iowa 50011
2
Institute for Bioengineering, University of Edinburgh, Mayfield Road, Edinburgh, EH9 3DW, UK
Correspondence to: I. V. Rivero (E - mail: [email protected])

ABSTRACT: This study presents and evaluates an approach for fabricating poly(lactic acid) (PLA)/chitosan (CS)/tricalcium phosphate (TCP)
electrospun scaffolds for guided bone regeneration, a dental procedure that uses membranes to direct and delineate regions of osteogenesis.
Biomaterials were pre-processed using cryomilling, a solid-state grinding technique that facilitates the generation of powdered biocomposites
conducive to electrospinning. X-ray diffraction (XRD) confirmed the generation of cryomilled blends consisting of PLA, CS, and TCP.
Results from the differential scanning calorimetry showed an upward shift in glass transition temperature and an increase in crystallinity
with the inclusion of TCP reinforcing the observations from XRD. Murine macrophages were used to confirm the biocompatibility of the
cryomilled powders and was evaluated using CellTiter-Blue (CTB) cell viability assay and brightfield microscopy. Scanning electron micros-
copy was used to examine the morphology of the fibers produced via electrospinning, while Raman spectroscopy confirmed material homo-
geneity. In vitro studies with MG-63 cells validated the capacity of composite scaffolds to encourage proliferation, while Coherent anti-
Stokes Raman scattering and fluorescence microscopies provided visual evidence of cell proliferation. CTB assay revealed that cells maintain
viability and metabolic activity at 3 and 7 days after seeding, demonstrating the potential of the biocomposite membranes. © 2018 The Authors.
Journal of Applied Polymer Science published by Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2018, 135, 46692.

KEYWORDS: biomaterials; biomedical applications; blends; composites; fibers

Received 13 January 2018; accepted 7 May 2018

DOI: 10.1002/app.46692

INTRODUCTION time of implantation.4 Expanded polytetrafluoroethylene (e-PTFE)

In the recent years, increasing interest in developing dental proce- has dominated as the material for the fabrication of non-resorbable
dures for bone regeneration has been driven by the growing number membranes which needs to be surgically removal.5 However, resorb-
of patients requiring such interventions. Data from National Health able membranes offer an interesting alternative making the process
and Nutrition Examination Survey (NHANES) suggests that 64.7 more patient-friendly.6 Polymers such as chitosan (CS), collagen,
million adults in the USA suffer from periodontitis, a dental disease poly(lactic acid) (PLA), poly(ϵ-caprolactone) (PCL), and
characterized by the destruction of the connective tissue and dental poly(lactic-co-glycolic acid) (PLGA) have been used individually,
bone.1 Guided bone regeneration (GBR), when applied, has been able or in blends, for the fabrication for such membranes.7
to treat bone defects caused by periodontitis.2 In principle, GBR uses In recent years, there has been an increased focus in the develop-
barrier membranes to prevent the entry and proliferation of non- ment and investigation of new polymeric blends that can possess
osteogenic cells in defect sites by selectively promoting the growth of superior mechanical and biological properties in comparison to
osteogenic cells.3 Therefore, the procedure’s success, among other commercially available single biopolymers. For example, Guo et al.8
factors, is also largely dependent on the design and performance of reported the potential of PLGA/PLLA/PDLA blend fibers loaded
the barrier membrane. A suitable barrier membrane is expected to with naringin for GBR. In another study, a mixture of chitosan, colla-
possess biocompatibility, mechanical strength, and pliability at the gen, and poly(ethylene oxide) was used to make up a nanofiber

© 2018 The Authors. Journal of Applied Polymer Science published by Wiley Periodicals, Inc.

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits
use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or
adaptations are made.

46692 (1 of 10) J. APPL. POLYM. SCI. 2018, DOI: 10.1002/APP.46692

ARTICLE WILEYONLINELIBRARY.COM/APP

chitosan-collagen membrane to be used in GBR.9 Currently, the maintained at −196  C. The samples were cryomilled for 20 min
enhancement of polyester-based membranes by the addition of inor- (4 milling cycles).22 A cooling time of 1 min was allowed between
ganic and organic materials is also being investigated widely.10–12 In successive cycles and a precool time of 15 min was utilized to
that regard, this study explores and evaluates the preparation ensure homogeneity in temperature at the time of milling.23 The
approach and performance of novel biocomposite membranes con- samples were stored in a silica-filled desiccator for at least 48 h at
sisting of PLA, CS, and tricalcium phosphate (TCP). room temperature before further processing or analysis.
CS, a natural occurring biodegradable polymer, has antimicrobial Characterization of Powder Biocomposite
properties that are desirable in resorbable membranes but inferior Confirmation for the production of a composite blend was
mechanical properties and unsuitable degradation rate have limited achieved using the Rigaku Miniflex 600 XRD analysis unit (Tokyo,
its exclusive use.13,14 On the other hand, PLA, a synthetic polymer Japan) equipped with a Cu-Kα radiation source (λ = 0.154 nm).
known for its suitable mechanical strength and biocompatibility has The voltage and current applied were 30 kV and 15 mA, respec-
been used in the fabrication of commercially available membranes tively. The scan ranged from 3 to 80 with steps of 0.02 .
for two decades now.15 Despite these favorable properties, the release Thermal characterization of the powders produced was performed
of oligomers and acid byproducts during degradation has necessi- using a differential scanning calorimeter (DSC) (Phoenix, NETZSCH
tated the need to tune its properties with ceramic fillers such as Instruments, Burlington, MA). To avoid the degradation of CS, the
TCP.16 Apart from being used to enhance the material strength, TCP method of Sakurai et al.24 was used with modifications. Suyatma
also ensures the presence of an ideal ionic environment for bone et al.25 reported the success of the modified method in characterizing
regeneration. Composite materials thus facilitate the fabrication of PLA/CS biodegradable films. 8 mg of the sample were quenched at
tailor-made membranes that can exhibit positive synergistic effects. 10  C min−1 to −30  C before being heated to 190  C at the same
However, the challenges of attaining uniform dispersions and the rate. The samples were held at 190  C for 1 min before being cooled
possible denaturing of biomolecules remain a hurdle to the success down to −30  C. The samples were then held at −30  C for 3 min
of traditional compatibilization strategies.17 before the second heating cycle in which the samples were heated to
To overcome the above-mentioned challenges associated with mix- 250  C. The second heating scan was used to identify the glass transi-
ing strategies, cryomilling, a solid-state, low temperature grinding tion and melting temperatures along with percent crystallinity. Pro-
process was employed in this study to generate blends of PLA, CS, teus Thermal Analysis version 6.1.0 was used for the analysis.
and TCP. Electrospinning was then utilized to fabricate fibrous
Cytotoxicity Study for Biocomposite Powder
membranes due to its previous success in bone tissue engineering.18
Powders A0, A1, A2, and A3 were weighed and suspended in a
Material properties of cryomilled biocomposites and electrospun
standard DMEM medium (Sigma-Aldrich) at a concentration of
fibers were analyzed using X-ray diffraction (XRD), differential
0.84% w/v, creating four different stock media. Each stock
scanning calorimetry (DSC), scanning electron microscopy (SEM),
medium was then autoclaved at 121  C for 15 min, cooled to
and Raman spectroscopy. Murine macrophages (RAW 264.7) were
room temperature, and diluted using the standard medium to
used to investigate the cytocompatibility of the biocomposite pow-
give particle dose concentrations of 0.3 × 106, 1 × 106, 3 × 106,
ders via CellTiter-Blue (CTB) cell viability assay. The electrospun
and 5 × 106 particles mL−1. The dosing calculations were carried
membranes were also evaluated for their capacity to support the
out assuming a modal particle diameter of 5 μm and a density of
growth of model osteoblasts (MG-6f3 osteosarcoma cells) using
1.2 g cm−3 for the composite material. The new solutions were
CTB cell viability assay as seen elsewhere.19–21 A second viability
named CompA0, CompA1, CompA2, and CompA3 medium.
assay employing fluorescence and coherent anti-Stokes Raman scat-
tering (CARS) microscopies allowed simultaneous imaging of the The RAW264.7 murine macrophage cell line was used to test the
fibers and cells. cytocompatibility of CompA0-A3 medium. The cells, at passage
12, were cultivated in DMEM medium (Sigma-Aldrich) supple-
In this manner, this research validates an effective approach for the
mented with 10% fetal bovine serum (FBS) and 1% penicillin/
fabrication of novel poly(lactide)-based biocomposites membranes
streptomycin (Invitrogen, Paisley, UK). When confluent, cells
through morphological, thermal, and structural characterization. In
were trypsinized, and subsequently seeded into a 96-well plate at
addition, this work also serves as a preliminary study for using novel
the concentration of 3 × 104 cells/well in a volume of 100 μL of
PLA/CS/TCP blends for GBR applications by studying the attach-
DMEM. Cells were then incubated for 24 h at 37  C, 5% CO2,
ment of MG-63 cells on the generated biocomposites scaffolds.
and subsequently exposed to CompA0-A3 media in 100 μL of
medium/well. This yielded an approximate load of 10, 30,
EXPERIMENTAL 100, and 200 particles per cell. Cells were incubated for a further
Fabrication of Powder Biocomposite 24 h and imaged using bright-field microscopy at 200× magnifi-
PLA (Purasorb PL 10; Purac, The Netherlands), CS (448877-50G, cation (Leica Microsystem, Milton Keynes, UK). Next,
Medium Mw; Sigma-Aldrich, Irvine, UK), and TCP CompA0-A3 media was replaced with fresh DMEM and cells
(C5267-100G, 34.0–40.0% Ca basis; Sigma-Aldrich) were cryo- were incubated for further 24 and 48 h, before proceeding with
milled to generate powdered composites. Compositions of the the CTB cell viability assay (Promega, Southampton, UK). Briefly,
blends prepared in this study is as follows: A0: 100% PLA; A1: 20 μL/well of the reagent were added to cells grown in a 96-well
70% PLA, 30% CS; A2: 68% PLA, 30% CS, 2% TCP; A3: 66% plate which were subsequently incubated at 37  C, 5% CO2 for
PLA, 30% CS, 4% TCP. The sample-containing vials were loaded 3 h. At the end of incubation, the supernatant of each well was
into the freezer mill (6870; SPEX, Metuchen, NJ) which was transferred to a fresh 96-well plate and fluorescence was measured

46692 (2 of 10) J. APPL. POLYM. SCI. 2018, DOI: 10.1002/APP.46692

ARTICLE WILEYONLINELIBRARY.COM/APP

with a microplate reader at 560/690 nm (Modulus II Microplate 10 min. Cell attachment was encouraged by incubating scaffolds
Multimode Reader, Turner Biosystems, Sunnyvale, CA). Cells at 37  C, 5% CO2 for 15 min and then 500 μL of DMEM were
grown in DMEM were used as negative controls and all experi- added to each scaffold. The media was changed every 3 days,
ments were conducted in triplicate. Data are expressed as mean ± with cell viability being investigated after 3 and 7 days using the
standard deviation (SD). CTB assay. At the end of the incubation period, 100 μL of CTB
solution were added to scaffolds, which were incubated at 37  C,
Fabrication of Resorbable Membranes 5% CO2 for 3 h. Then, the supernatant from each well, together
Electrospinning was employed to generate fibrous membranes with the scaffolds, were transferred into a fresh microvial which
using the cryomilled biocomposites. Trifluoroacetic acid (TFA) was centrifuged at 1000 rpm for a minute. The supernatants were
(O4902-100; Fisher Scientific) was used as the solvent. In a typi- put into a fresh microvial and vortexed. 100 μL of the superna-
cal process, 18 w/v % of the cryomilled powder was added to tants were transferred into a glass bottom microplate, and fluo-
5 mL of TFA, and stirred using a magnetic stirrer at 25  C for at rescence was measured at 560/690 nm. Samples were investigated
least 24 h. A stationary copper plate covered with aluminum foil in triplicate. All data are expressed as mean ± SD.
was used as the collector. The needle tip to collector distance was
Imaging was also used to assess cell viability using LIVE/DEAD
set to 13 cm, and the potential difference was adjusted between
staining (Invitrogen). The cell-laden scaffolds were incubated with
15 and 17.5 kV as needed, with a flow rate of 0.05 mL min−1.
Calcein acetoxymethyl (Calcein-AM) 2 μM plus ethidium
Characterization of Resorbable Membranes homodimer-1(EthD-1) 4 μM for 15 min at 37  C, 5% CO2, in the
SEM (JCM-6000Plus NeoScope JEOL, Peabody, MA) was used to dark. Scaffolds were gently washed with 1 mL of PBS post-
confirm the successful production of electrospun fibrous mem- incubation. Fluorescent dyes were excited and imaged using two-
branes. Fiber diameter was measured using JCM-6000 software photon emission fluorescence (TPEF), while fibrous scaffolds were
version 1.1. The mean diameters of the membranes were calcu- resolved using CARS. TPEF was read using two bandpass filters;
lated using 30 measurements from three independent samples. for green fluorescence, a 515/42 nm combined with a 535/40 nm
All diameters are represented as mean ± SD. and for red fluorescent marker, the 609/54 and 640/14 nm (all
Semrock). For CARS, a 1064 nm (Stokes beam) and 810.3 nm
Further quality assurance for fiber production was conducted (pump beam) were used to excite the 2940 cm−1 CH stretch
with NIR Raman spectroscopy. Electrospun membranes of A0, vibration, signal was selected using a 660/13 nm bandpass filter.
A1, A2, and A3 were cut into four different pieces and sterilized
by UV irradiation for 15 min. Raman spectra were recorded at Statistical Analysis
20 different locations across each sample (4 × 5 grid with Results of the CTB cell viability assays were analyzed using
200 μm spacing) using a Renishaw InVia microscope (785 nm PRISM version 7.0 (GraphPad Software, San Diego, CA) with
excitation, 65 mW at the sample, ×40 objective giving a 10 μm 95% confidence intervals (CI) of the difference. Two-way
focal spot diameter). Spectra were dispersed by a 600 lines/mm ANOVA was performed to evaluate statistical significance with a
grating onto a thermoelectrically cooled CCD, yielding a spectral designated Type I error rate of 0.05%.
resolution of 7 cm−1 across a sampling range 400–3200 cm−1
wavenumbers of Stokes shift.
RESULTS AND DISCUSSION
Performance Assessment of Resorbable Membranes Morphology of Cryomilled Composite Powders
MG-63 human osteosarcoma cell line was cultivated in DMEM SEM was utilized to evaluate the morphology of the cryomilled
medium (Sigma-Aldrich) supplemented with 10% FBS and 1% biocomposite particles in comparison to as-received, non-
penicillin/streptomycin (Invitrogen). When confluent, cells were cryomilled PLA. The composite powders were composed of
trypsinized, and suspended into 30 μL of medium and subse- sharp-edged particles ranging in size (across all the composite
quently seeded onto the scaffolds (7.5 × 104 cells/scaffold). Before blends) from 15 to 20 μm, significantly smaller than the particle
cell seeding, scaffolds were placed one per well in a sterile 48-well size of the materials before cryomilling [Figure 1(a,b)]. Addition-
plate, UV irradiated for 15 min and pre-soaked into DMEM for ally, dry agglomeration of particles was also observed.

Figure 1. Representative SEM images of (a) as-received non-cryomilled PLA and (b) cryomilled PLA/CS/TCP (A3).

46692 (3 of 10) J. APPL. POLYM. SCI. 2018, DOI: 10.1002/APP.46692

ARTICLE WILEYONLINELIBRARY.COM/APP

The primary step in the membrane fabrication process was to uti- 2θ = 25.66 , 2θ = 31.58 , and 2θ = 32.68 . All of the results for
lize cryomilling to generate a fine powder blend of PLA, CS, and composite blends A1, A2, and A3 showed similar PLA diffraction
TCP. Cryomilling, being a mechanical attrition process, subjects peaks at 2θ = 16.38 and 19.44 . However, the diffractograms of
the materials to repeated fracturing and welding by means of col- all the CS-containing blends showed an increase in intensity at
lisions with a high-energy impactor bar. Other than being a sol- 2θ ≈ 20 when compared to the profile of pure PLA. For blends
ventless process, the use of cryomilling provided two benefits A2 and A3, an additional peak was observed at 2θ = 32 which
(a) it reduced the size disparity between the as-received PLA, CS, appeared to grow with increasing TCP content in the polymer
and TCP particles thus avoiding granular convective effects,26 matrix.
and (b) it generated a fine powder blend of dissimilar materials
PLA showed the typical profile of a semi-crystalline polymer consist-
making it easier for dispersion in solvents. It was possible to ing of both crystalline and amorphous regions. The positions of the
attain a considerable reduction in particle size in short period of peaks aligned with results reported previously.32 The strong crystal-
processing because the polymers were milled below their glass line peak of PLA corresponded to the (110) and/or (200) plane of a
transition temperature which facilitated the occurrence of brittle typical orthorhombic crystal.33 The two peaks shown by CS were
fractures.27,28 The cryomilled powders showed signs of dry indicative of the hydrated crystalline nature of the material.34 The CS
agglomeration due to electrostatic and van der Waals forces peaks at angles 2θ = 9.34 and 19.44 corresponded to d-spacing
between the particles.29–31 values of 0.95 and 0.45 nm which matched previous findings.35 Also,
TCP peaks at 2θ= 25.66 , 2θ = 31.58 , 2θ = 32.68 , 2θ = 33.86 ,
X-Ray Diffraction Analysis 2θ = 39.60 , and 46.54 confirmed the presence of TCP.36 For all
XRD was utilized to confirm the successful production of powder three composite blends, the increase in intensity around 2θ ≈ 20
blends consisting of PLA, CS, and TCP. Figure 2 shows the XRD was indicative of the presence of CS and the growing intensity of the
profiles for as-received PLA, CS, and TCP along with the profiles peak around 2θ ≈ 32 with increasing TCP content reflected a suc-
for the composite blends prepared by cryomilling. The cessful inclusion of the ceramic component in the polymer matrix.37
as-received PLA exhibited a very strong crystalline peak centered The profiles of the composite blends were observed to be influenced
at 2θ= 16.38 and a relatively weak peak at 2θ = 18.72 . For the by those of the individual materials indicating that the materials
CS, a sharp diffraction peak was observed at 2θ = 19.44 and a blended only on a physical-level. On the whole, the XRD data sub-
weak peak centered around 2θ = 9.34 . The as-received TCP stantiated the ability of cryomilling to generate a blend of dissimilar
showed a lot of distinct peaks with high intensity peaks at materials.

Figure 2. XRD profiles of composite blends along with their individual components.

46692 (4 of 10) J. APPL. POLYM. SCI. 2018, DOI: 10.1002/APP.46692

ARTICLE WILEYONLINELIBRARY.COM/APP

Table I. Second Heating Scan Data from DSC Analysis

X∘c Hm
 
Material Tg( C) Tm( C) (%) (J/g)
A0 57.6 177.5 15.2 14.5
A1 60.9 177 15.0 14.3
A2 62.7 177.6 23.9 22.3
A3 64 177.4 27.8 26.2

at 2 and 4 wt % content, respectively. The bioceramic containing

fibers experienced branching and seemed to have a rougher surface
even though the presence of TCP on the surface on the fibers was
Figure 3. DSC second heating scans of the cryomilled composites. not evident from the micrographs.
It was difficult to identify a common solvent to dissolve PLA and
Differential Scanning Calorimetry Analysis chitosan for electrospinning because of the inherent dissimilarity
DSC was used to quantify the glass transition and melting tem- of these polymers. The benefit of using TFA was twofold (a) it
peratures as well as crystallinity of PLA and the composite pow- facilitated the electrospinning of PLA nanofibers by increasing the
der blends. The thermograms from the second heating scan are solution conductivity and (b) destroyed the strong associations
shown in Figure 3 while associated thermal parameters are between chitosan molecules and thereby facilitating the continu-
reported in Table I. DSC curve of pure PLA revealed an average ous production of fibers.41,42 The addition of chitosan increased
glass transition temperature of 57.6  C followed by an endother- the fiber diameter by enhancing the viscosity of the solution, and
mic melting peak at 177.5  C with an enthalpy of fusion of 14.5 J led to a broader distribution that was likely a consequence of
g−1. It was observed that the introduction of CS into the PLA increased charge density.43 The increase of fiber diameter and dis-
matrix produced an increase in the glass transition temperature tribution was reported in a previous work where similar concen-
while decreasing crystallinity. The percent crystallinity was trations of TCP were responsible for the inability of the polymer
enhanced by the inclusion of the bioceramic, increasing to 23.9% chains to stretch.44 It is also likely that modifications to the visco-
and 27.8% with the addition of 2 and 4 wt % TCP, respectively, in elastic component of the rheological behavior of the solutions
comparison to the 15.2% computed for pure PLA. changed the degree of swell experienced after extrusion from the
The second heating scan was used to conduct the analysis as it needle. It is possible that the diameter of the fibers could have
allowed for the direct comparison of materials. The glass transition influenced cell growth.45 Nevertheless, all the scaffolds showed
temperature of PLA obtained in this study was similar to values promise of being structurally similar to the collagen fibers in the
reported in other studies.38 The thermogram obtained for pure bone as they fell within the range 100–2000 nm and the roughness
PLA was indicative of its semi-crystalline nature, reinforcing the of the composite scaffolds were seen as a factor that may positively
result from the diffraction study. The inclusion of CS into the PLA influence cell attachment.46
matrix shifted the glass transition temperature of the polymer Homogeneity of Electrospun Barrier Membranes
blend (A1) upward. The semi-crystalline nature of CS, as shown in The average Raman spectra of 20 measurements captured for the
its diffraction profile, was responsible for the shift being modest. three raw materials and four electrospun composite membranes
Nevertheless, the observed increase was attributed to CS obstruct- is shown in Figure 5. These measurements were carried to con-
ing the movement of PLA chains.39 The increase in crystallinity firm material homogeneity after membrane fabrication.
with the addition of TCP into the polymer matrix was explained
by the former acting as a nucleation agent for crystallization.40 The results of the Raman spectroscopy reinforce the findings of
Since crystallinity has an influence on the rate of degradation, the the powder investigation regarding the homogeneity of the mate-
degradability of scaffolds can possibly be tailored by altering the rial produced, and confirms this even distribution of materials is
TCP content in the polymer matrix. maintained in its electrospun form. As the proportion of PLA in
the material falls from A0 to A3 the contribution of the PLA
component of the spectrum can be seen to reduce. Additionally,
Morphology of Electrospun Resorbable Membranes it can be seen that the deviation from the mean spectrum is mod-
Figure 4 shows SEM micrographs of randomly oriented electro- est, indicating that the composition of the material was similar at
spun nano/microfibers of powder blends prepared in this study. all 20 sampling locations.
As can be seen, the PLA nanofibers exhibited a round and smooth
surface morphology with an average diameter of 265.83 nm. The Cytotoxicity Study for Biocomposite Powder Blend
fibers generated from PLA/CS blends displayed a similar morphol- The cytocompatibility of the composite cryomilled powder blends
ogy, however, with approximately a 30% increase in mean diame- was tested on a murine macrophage cell line (RAW 264.7).
ter with a broader distribution. The inclusion of TCP lead to the Results of the CTB assay [Figure 6(a,b)] indicate that 24 h after
formation of significantly broader fibers of 891.83 and 1325.7 nm returning to cultivation in standard medium, cell viability is

46692 (5 of 10) J. APPL. POLYM. SCI. 2018, DOI: 10.1002/APP.46692

ARTICLE WILEYONLINELIBRARY.COM/APP

Figure 4. SEM images of fibrous membranes with fiber diameters represented as mean ± SD.

Figure 5. Baselined and normalized (to unit area) Raman spectra of each raw
material (PLA, CHI, TCP) and each type of electrospun scaffold (A0, A1, A2,
A3). Each spectrum represents the mean of 20 individual spectra with ±1SD.

statistically reduced with respect to the control for all the concen- Figure 6. CTB assay performed on RAW 264.7 cell line incubated with
trations tested. However, 48 h after returning to standard composite cryomilled powders blend A0, A1, A2, and A3, considering dif-
medium cultivation, the difference in cell viability between cells ferent particles/cell loads, after (a) 24 h and (b) 48 h particle exposure in
incubated with different cryomilled powders and controls is standard medium conditions. Results are expressed as fluorescence intensity
(relative fluorescence unit-RFU) and are indicated as mean ± SD.

46692 (6 of 10) J. APPL. POLYM. SCI. 2018, DOI: 10.1002/APP.46692

ARTICLE WILEYONLINELIBRARY.COM/APP

Figure 7. Representative bright-field micrographs of RAW 264.7 macrophages incubated for 24 h with different composite cryomilled powders. (a) 10 and
(b) 30 particles/cell loads. (c) control; magnification = 200×; Scale bar = 20 μm.

46692 (7 of 10) J. APPL. POLYM. SCI. 2018, DOI: 10.1002/APP.46692

ARTICLE WILEYONLINELIBRARY.COM/APP

Figure 9. CTB assay performed on MG-63 cells seeded into electrospun

scaffolds, after 3 and 7 days from cell seeding. Results expressed as
fluorescence intensity in Relative Fluorescence Units (RFU) and are indicated
as mean ± SD.

activation, and the lack of a sustained cell viability challenge

(i.e., viability recovery at 48 h) indicate that the cells may not be
ingesting significant numbers of particles. Importantly, neither
are the cells significantly perturbed by their extracellular interac-
tion with the biocomposite powders after an initial retardation of
their proliferation. Our results suggest that particle release should
not be a significant source of biocompatibility challenges.

Performance Assessment of PLA/CS/TCP Biocomposite

Membranes
The viability of MG-63 cells was qualitatively examined by fluo-
rescence and CARS microscopy at day 3 and 7 after cell seeding
onto scaffolds. All the scaffolds show high cell viability and low
number of dead cells, as shown in Figure 8. MG-63 cell viability
was evaluated also by CTB assay, at day 3 and 7 from cell seeding
into scaffolds (Figure 9). The results confirm the observations
from the fluorescence investigation, showing that cells are viable
at both time points investigated.
Cells were able to colonize the scaffolds, and proliferate on them.
Figure 8. Fluorescence microscopy on MG-63 cells seeded into electrospun
There was evidence that the scaffolds containing CS (A1, A2, and
scaffolds. Scaffolds fibers are depicted using CARS; Scale bar = 20 μm.
A3) appeared more suitable for cell attaching than scaffold A0, as
indicated by the pictures from fluorescence and CARS micros-
reduced, with no statistical difference seen between controls and copy, which show a higher number of cells. Scaffolds A1 and A3,
cells incubated with particles. Additionally, cell morphology was in particular, were noticed to have a good cell attachment along
investigated using bright-field microscopy [Figure 7(a–c)]. No scaffold fibers. The results of the CTB assay reflected the observa-
morphological changes were seen in macrophages incubated with tions made from the qualitative analysis. There was an increase
composite cryomilled powders when compared to control macro- of registered fluorescence intensity, corresponding to an increase
phages grown in standard medium suggesting a low induction of of cell viability, at day 7 with respect to day 3, indicating that
phagocytic response. In addition, these micrographs reveal that cells may be able to integrate inside the scaffolds and to grow on
particles, when suspended in culture medium, form agglomera- them with the passing of days.
tions, to which macrophages can be seen to adhere with subse-
The aim of this study was to determine if osteoblast-like cells are
quent cell proliferation.
able to attach and proliferate on electrospun scaffolds. Although
The high particle dose per cell in biocompatibility test represents no statistical difference was noticed among the different type of
an aggressive test of phagocytic response. The particle size scaffolds investigated, the promising results obtained on scaffolds
formed by cryomilling lies at the upper end of the phagocytic A1–A3 encourage additional investigations. Further study will
range of macrophages.47 Particle agglomeration, as observed investigate induction and maintenance of osteoblastic phenotype
in vitro, will act to further decrease the particles’ bioavailability. by following appropriate osteoblastic markers over extended
The lack of morphological cues suggesting macrophagic periods; such as, the alkaline phosphatase activity, the formation

46692 (8 of 10) J. APPL. POLYM. SCI. 2018, DOI: 10.1002/APP.46692

ARTICLE WILEYONLINELIBRARY.COM/APP

of calcium phosphate and the expression of osteogenic genes such 12. Arunjaroensuk, S.; Panmekiate, S.; Pimkhaokham, A. Int.
as osteopontin and osteocalcin.48 J. Oral Maxillofac. Implants 2018, 33, 206.
13. Kuo, S. M.; Chang, S. J.; Chen, T. W.; Kuan, T. C.
CONCLUSIONS J. Biomed. Mater. Res. Part A 2006, 76, 408.
Cytotoxicity testing on precursor powders used to manufacture 14. Bavariya, A. J.; Norowski, A. P.; Anderson, M. K.;
the scaffolds reveals a response from macrophages 24 h after Pradeep, A. C.; Garcia-Godoy, F.; Stein, S. H.; Bumgardner, J. D.
exposure to the powders, with their viability reduced when com- J. Biomed. Mater. Res. Part B: Appl. Biomater. 2014, 102, 1084.
pared to untreated controls. Biocompatibility of micron-sized 15. Galgut, P.; Pitrola, R.; Waite, I.; Doyle, C.; Smith, R. J. Clin.
particles of material is known to diverge from the same material Periodontol. 1991, 18, 581.
in the bulk; all the materials that have been utilized have estab-
16. Cao, L.; Duan, P. G.; Wang, H. R.; Li, X. L.; Yuan, F. L.;
lished biocompatibility. While the particles lie at the upper bound
Fan, Z. Y.; Li, S. M.; Dong, J. Int. J. Nanomed. 2012, 7, 5881.
of the phagocytosable range, it is likely that at the high particle
loads used there will be sufficient bio-available material to 17. Allaf, R. M.; Rivero, I. V. J. Mater. Sci. Mater. Med. 2011, 22,
impinge on cell viability. This effect could persist despite the ten- 1843.
dency of the powders to form large agglomerates when in sus- 18. Ramachandran, K.; Gouma, P. I. Recent Pat. Nanotechnol.
pension. When nanoparticles aggregate, cytotoxicity may 2008, 2, 1.
potentially arise from many factors, hence it is difficult to define 19. Yoo, C. K.; Jeon, J. Y.; Kim, Y. J.; Kim, S. G.; Hwang, K. G.
the causal factors of macrophage activation and the ultimate Maxillofac. Plast. Reconstr. Surg. 2016, 38, 17.
cytotoxicity of the particles. However, after 48 h recovery period 20. Kim, M. J.; Choi, H. J.; Cho, J.; Lee, J. B.; Sung, H. J.;
of cultivation in standard medium the viability of the macro- Kim, J. K. Int. J. Polym. Sci. 2017, 2017, 1. DO:
phages that were challenged with particles are comparable to that 10.1155/2017/1286109.
of controls. This could be interpreted as a transient cell inhibition
21. Sui, G.; Yang, X.; Mei, F.; Hu, X.; Chen, G.; Deng, X.;
by particles, allowing the cells to restore their viability and prolif-
Ryu, S. J. Biomed. Mater. Res. Part A 2007, 82, 445.
eration activity. Importantly, electrospun fibers scaffolds formed
by these powders show good biocompatibility and appear suitable 22. Lim, J.; Chong, M. S. K.; Chan, J. K. Y.; Teoh, S. H. Small
scaffolds for osteoblast-like cells, promoting adhesion and prolif- 2014, 10, 2495.
eration. MG-63 cell morphology, being stellate not spherical, 23. Jonnalagadda, J. B.; Rivero, I. V.; Warzywoda, J.
shows good attachment to individual fibers of the scaffolds, with J. Biomater. Appl. 2015, 30, 472.
the cells’ growth increasing over time. These preliminary material 24. Sakurai, K.; Maegawa, T.; Takahashi, T. Polymer (Guildf )
quality and biocompatibility experiments justify further investiga- 2000, 41, 7051.
tion into the mechanical performance of the membranes and 25. Suyatma, N. E.; Copinet, A.; Tighzer, L.; Coma, V.
longer-term cell response, to include osteogenesis and biodegra- J. Polym. Environ. 2004, 12, 1.
dation studies.
26. György, M.; György, B.; Péter, A.; Istvan, R. J. Polym. Eng.
1993, 12, 33.
REFERENCES 27. Zhu, Y. G.; Li, Z. Q.; Zhang, D.; Tanimoto, T. J. Polym. Sci.
1. Eke, P. I.; Dye, B. A.; Wei, L.; Thornton-Evans, G. O.; Part B: Polym. Phys. 2006, 44, 1161.
Genco, R. J. J. Dent. Res. 2012, 91, 914. 28. Smith, A. P.; Harald, A.; Koch, C. C.; Smith, S. D.;
2. Lee, J. Y.; Lee, J.; Kim, Y. K. J. Dent. Sci. 2013, 8, 281. Spontak, R. J. Macromolecules 2000, 33, 1163.
3. Retzepi, M.; Donos, N. Clin. Oral Implants Res. 2010, 29. Ye, J.; Han, B. Q.; Schoenung, J. M. Philos. Mag. Lett. 2006,
21, 567. 86, 721.
4. Scantlebury, T. V. J. Periodontol. 1993, 64, 1129. 30. Zhu, Y. G.; Li, Z. Q.; Zhang, D.; Tanimoto, T. J. Appl.
Polym. Sci. 2006, 99, 501.
5. Aaboe, M.; Pinholt, E. M.; Hjorting-Hansen, E. Br. J. Oral
Maxillofac. Surg. 1995, 33, 312. 31. Jiang, J.; Oberdörster, G.; Biswas, P. J. Nanopart. Res. 2009,
11, 77.
6. Jung, R. E.; Fenner, N.; Hämmerle, C. H. F.;
Zitzmann, N. U. Clin. Oral Implants Res. 2013, 24, 1065. 32. Naga, N.; Yoshida, Y.; Noguchi, K.; Murase, S. Open
7. Wang, J.; Wang, L.; Zhou, Z.; Lai, H.; Xu, P.; Liao, L.; J. Polym. Chem. 2013, 3, 29.
Wei, J. Polymers 2016, 8, 115. 33. Wu, T. M.; Wu, C. Y. Polym. Degrad. Stab. 2006, 91, 2198.
8. Guo, Z.; Wu, S.; Li, H.; Li, Q.; Wu, G.; Zhou, C. Dent. 34. Wang, S. F.; Shen, L.; Zhang, W. D.; Tong, Y. J. Biomacro-
Mater. 2018, 37, 317. molecules 2005, 6, 3067.
9. Gao, B.; Li, X. J.; Lin, M.; Li, Y. Y.; Dong, Y. Chin. 35. Muzzarelli, C.; Francescangeli, O.; Tosi, G.;
J. Stomatol. 2018, 53, 85. Muzzarelli, R. A. A. Carbohydr. Polym. 2004, 56, 137.
10. Castro, A. G. B.; Diba, M.; Kersten, M.; Jansen, J. A.; van den 36. Siddiqui, N.; Pramanik, K. J. Appl. Polym. Sci. 2014, 131,
Beucken, J. J. J. P.; Yang, F. Mater. Sci. Eng. C 2018, 85, 154. DOI 10.1002/app.41025. doi:10.1002/app.41025
11. Fu, L.; Wang, Z.; Dong, S.; Cai, Y.; Ni, Y.; Zhang, T.; 37. Spiridon, I.; Paduraru, O. M.; Zaltariov, M. F.; Darie, R. N.
Wang, L.; Zhou, Y. Materials (Basel) 2017, 10, 257. Ind. Eng. Chem. Res. 2013, 52, 9822.

46692 (9 of 10) J. APPL. POLYM. SCI. 2018, DOI: 10.1002/APP.46692

ARTICLE WILEYONLINELIBRARY.COM/APP

38. Correlo, V. M.; Boesel, L. F.; Bhattacharya, M.; 44. Siqueira, L.; Passador, F. R.; Costa, M. M.; Lobo, A. O.;
Mano, J. F.; Neves, N. M.; Reis, R. L. Mater. Sci. Eng. A Sousa, E. Mater. Sci. Eng. C 2015, 52, 135.
2005, 403, 57. 45. Sisson, K.; Zhang, C.; Farach-Carson, M. C.; Chase, D. B.;
39. Tanase, C. E.; Spiridon, I. Mater. Sci. Eng. C 2014, 40, 242. Rabolt, J. F. J. Biomed. Mater. Res. Part A 2010, 94, 1312.
40. Mathieu, L. M.; Bourban, P. E.; Månson, J. A. E. Compos. 46. Murugan, R.; Ramakrishna, S. Compos. Sci. Technol. 2005,
Sci. Technol. 2006, 66, 1606. 65, 2385.
41. Hadjizadeh, A.; Savoji, H.; Ajji, A. Biomed Res. Int. 2016, 47. Elfick, A. P. D.; Green, S. M.; McCaskie, A. W.;
2016, 1. DOI 10.1155/2016/8921316. Birch, M. A. J. Biomed. Mater. Res. Part B: Appl. Biomater.
42. Lee, K. Y.; Jeong, L.; Kang, Y. O.; Lee, S. J.; Park, W. H. 2004, 71, 244.
Adv. Drug Deliv. Rev. 2009, 61, 1020. 48. Tsai, S. W.; Liou, H. M.; Lin, C. J.; Kuo, K. L.; Hung, Y. S.;
43. Meng, Z. X.; Zheng, W.; Li, L.; Zheng, Y. F. Mater. Chem. Weng, R. C.; Hsu, F. Y. PLoS One 2012, 7, e31200. DOI
Phys. 2011, 125, 606. 10.1371/journal.pone.0031200.

46692 (10 of 10) J. APPL. POLYM. SCI. 2018, DOI: 10.1002/APP.46692