SILK SERICIN-FUNCTIONALIZED BACTERIAL CELLULOSE AS A

POTENTIAL WOUNDHEALING BIOMATERIAL

Lallepak Lamboni, Ying Li, Jianfeng Liu and Guang Yang

Abstract
Bacterial cellulose (BC) is a polysaccharide known as a suitable matrix for
proper wound healing. To improve this ability, BC was functionalized with silk
sericin (SS) that has cytoprotective and mitogenic effects. The composites
obtained by solution impregnation were stabilized by hydrogen bonds, and SS
could be released in a controlled manner. The constructs were highly porous with
interconnected pores allowing for high water uptake that varied with the SS
concentration used for sample preparation. While SS did not disrupt the stability
of the BC network, soluble SS diffusing from the composites did not influence
keratinocyte growth but enhanced fibroblast proliferation, which would further
optimize the wound healing process and improve extracellular matrix production,
accelerating healing. Further, improved cell viability was observed upon the
composites. Because of their attractive structure and properties, these BC−SS
biomaterials represent potential candidates not only for wound dressing
applications but also for tissue engineering.

INTRODUCTION
Wound dressing is a biomaterial designed to create a suitable
microenvironment for cell adhesion and proper proliferation in order to restore the
physiological and structural properties of wounded skin. It controls cell behavior
by regulating physical, chemical, and biological signals through its surface
microarchitecture. Whether used as temporary coverage or meant for chronic
applications, wound dressings are of capital importance especially for the
treatment of large wounds and may be critical if not life-saving in certain cases
such as extensive burns. In fact, according to the World Health Organization, over
300 000 people die of burn injuries, making skin wounds a major burden both
socially and financially. This high rate is due in part to poor skin healing ability,
additionally to the limitations of current treatments, mostly the organ donor
limitation. Therefore, despite the numerous wound dressing methods available to
date, there is still a pressing need for improved ones, as the former only meet
part of the requirements for an ideal wound dressing. The ideal wound dressing
as currently described represents an effective barrier to micro-organisms, which
absorbs wound exudate and allows gas exchange, while keeping the necessary
moisture at the wound interface without toxicity, allergenic responses, and can
be removed from the wound with minimum pain. This should be made from
readily available and affordable materials that require minimum processing. In
this context, nanofibrous materials, which have attracted significant interest as
scaffolds in current tissue engineering strategies aiming for the repair of various
tissues including skin, have shown to improve biocompatibility. Becaise of their
high porosity, they enable cell adhesion, growth, and differentiation while further
allowing for the delivery of bioactive molecules.
In recent years, bacterial cellulose (BC), a natural polysaccharide, has
shown great potential as a biomaterial for tissue engineering applications
including skin tissue repair. Because of its nanofibrillar structure, BC represents
an adequate matrix for optimal wound healing. In fact, similarity with collagenous
fibers facilitates the interaction between cells and BC fibers, improving its overall
biocompatibility. It is biodegradable and offers good mechanical properties and
high hydrophilicity, all of which contribute to its success in tissue engineering.
Conversely, silk sericin (SS) is a natural hydrophilic protein and a
byproduct of the silk industry that biomedical properties have been well
elucidated. It has cytoprotective and mitogenic effects on mammalian cells with
regenerative abilities on mammalian tissues; particularly, its positive effects on
fibroblasts and keratinocytes, major cell types in skin, has made it very attractive
for the development of skin tissue repair materials. In addition to its potential for
enhancing collagen production and accelerating wound healing, SS has shown
antimicrobial properties necessary for a wound dressing. Nevertheless, due to its
amorphous nature, SS only forms fragile materials unsuitable for biomedical
applications. Interestingly, it has polar side chains with diverse functional groups
(amine, hydroxyl, carboxyl groups) that allow interaction with other compounds
through blending, crosslinking, or copolymerization to yield improved
biomaterials.
This study aims at developing an improved wound dressing material which
would accelerate wound healing, with BC as the niche for optimal wound healing,
and SS as the bioactive molecule that regulates cell behavior and accelerates the
healing process thereby reducing scar formation.

EXPERIMENTAL SECTION
Materials. Bombyx mori cocoons from Ningbo Industrial Co., Ltd. (China)
were used. The bacterial strain Acetobacter xylinum (ATCC53582) was
purchased from the American Type Culture Collection (ATCC), while the NIH-
3T3 and HaCaT cell lines were obtained from China Infrastructure of Cell Line
Resources. Yeast extract and peptone were provided from Beijing Shuangxuan
Microbe Culture Medium Products Factory (China) with all other reagents of
chemical grade except for poly(ethylene glycol) (PEG, Biosharp, China)
purchased from Sigma. Dulbecco modified eagle medium (DMEM), fetal bovine
serum (FBS), and trypsin-EDTA (TE) were products of Gibco.
Penicillin/streptomycin mixture was obtained from Beyotime (China). Phosphate-
buffered saline (PBS) solution was supplied from Hyclone, MTT (3-[4,5-
dimethylthiazol-2-yl]-2,5- diphenyl tetrazolium bromide) powder from Sigma, and
alamarBlue from Thermo Fisher Scientific. Fluorescein isothiocyanate labeled
phalloidin (FITC-Phalloidin) was provided from Sigma, Hoechst 33342 from
Invitrogen (Life Technologies), formaldehyde from Solarbio (Beijing, China), and
triton X-100 from Amresco. The BCA (bicinchoninic acid) kit was purchased from
Pierce, Thermoscientific.
Preparation of Bacterial Cellulose Films. BC was produced in Hestrin &
Schramm (HS) medium from Acetobacter xylinum grown in static cultures at 30
°C. The HS medium was composed as follows: citric acid (0.15%, w/v), disodium
phosphate (0.27%, w/v), yeast extract (0.5%, w/v), peptone (0.5%, w/v), and
glucose (2%, w/v). The films were allowed to grow to a thickness of 2−3 mm after
which they were harvested and subjected to cleaning steps to remove the
bacteria and the adsorbed HS medium. They were washed for 3 days in distilled
water, then boiled in NaOH (1 wt %) for 40−45 min, and subsequently washed in
purified water to neutralize the pH. The clean BC membranes were autoclaved in
purified water and stored at 4 °C for further use. For different experimental
purposes, various BC sizes were obtained from different sizes culture flasks or
polystyrene cell culture vessels.
Extraction of Silk Sericin from Bombyx mori Cocoons and
Preparation of BC−SS Composites. SS was extracted from Bombyx mori
cocoons by a method described by Lee et al. with slight modification. The
cocoons were cut into small pieces and boiled under pressure at 120 °C for 1 h
in purified water (1 g of cocoon/10 mL water). Fibroin fibers were removed from
the resulting soup by filtration followed by centrifugation. SS solution was then
concentrated by dialysis against PEG solution and lyophilized. The powder was
stored at −20 °C until use. The absorbance profile of the extract between 700 and
190 nm was scanned by UV spectroscopy (UV-1600 PC spectrophotometer,
MAPADA, coupled to “Mwave − Professional 2” software), from a solution sample
constituted as described further in this section.
BC−SS composites were prepared by solution impregnation as depicted
in Figure 1. To obtain SS solution, SS powder was resuspended in purified water
and dissolved by boiling for 20 min.30 The solution was further filtered through
Minisart high flow syringe filters (pore size 0.2 μm, Sartorius Stedim) for the
removal of particulate impurities. Sterile filters were used when sterilization was
required. Different constructs (BC-SS1:24, BC-SS2:24, BC-SS3:24) were thus
fabricated by varying SS concentration (1, 2, and 3% (w/v), respectively).

Figure 1. Preparation of SS powder, BC pellicles, and BC−SS composites. (c = [SS] in

% w/v, 1, 2, 3; A. xylinum = Acetobacter xylinum; HS = Hestrin and Schramm).
 Scanning Electron Microscope (SEM), Brunauer−Emmett− Teller
(BET), and Fourier Transform Infrared (FTIR) Analyses. SEM (NanoSEM 450)
observation was carried out to examine the morphology of the constructs and
confirm the incorporation of SS into the BC network. Cross-section imaging was
performed after platinum coating (Precision etching coating system, Gatan,
Model 682). Samples were prepared by freeze-drying (48 h freezing at −20 °C,
and 48 h lyophilization at −50 °C), and cross sections were achieved after
steeping in liquid nitrogen. For SS sample, the lyophilized powder was gold-
coated and observed similarly to the scaffolds.
The microstructure of the membranes was further examined by BET
analysis using Micromeritics ASAP 2020 (ASAP 2020 V4.00 (V4.00 H)). For this
purpose, 6-well plates of BC membranes were composited with SS and frozen-
dried as indicated above. The membranes were weighed, placed in the automate,
and subjected to a dehydrating phase where they were heated up to 120 °C to
get rid of resident moisture. After cooling the samples back to ambient
temperature, the surface area and porosity measurements where performed
using nitrogen as the adsorptive gas with an equilibration time of 10 s.
The chemical interaction between BC and SS was studied by FTIR (Vertex
70, Bruker). The samples were prepared by freeze-drying, and analyzed between
4000 and 450 cm−1 with the spectra obtained at a resolution of 4 cm−1 and a
number of accumulated scans of 16.
In Vitro SS Release from the Samples. The physical integrity of the
composites was assessed by measuring the release of SS at different time
intervals while the never-dried samples were immersed in a known volume of
PBS (1X, pH 7.4) containing sodium azide (0.02 wt %) to avoid contamination.
For this purpose, BC was synthesized in 6- well cell culture plates (ø 34.8 mm;
9.5 cm2 ). The experiments were conducted at 37 °C over 7 time-points including
0, 1, 15, and 30 min and 1, 2, and 3 days, respectively. The cumulated SS content
at each time interval was determined by colorimetric test using the BCA protein
assay kit. The absorbance was measured in a microplate reader (318C
Microplate Reader). For each time-point, three replicates were used, and the
experiment was repeated three times.
Thermal and Mechanical Stability of the Composites. The thermal
behavior of the composites was determined by thermogrametric analysis (TGA)
(PerkinElmer Instruments). The samples were heated from room temperature to
600 °C, at a rate of 10 °C/min under nitrogen atmosphere.
A tensile test was performed on the samples in wet state, using the model
CMT6503 of SANS Universal Testing Machines (SANS Test Machine Co. Ltd.
Shenzhen, China). The test specimens, dumbbell shaped, were cut out of the
never-dried films using a mold of similar shape. For each sample, three
specimens were tested. The tensile strength and Young’s modulus were
thereafter determined.
Water Uptake Ability. The swelling behavior of the composites in distilled
water was studied by gravimetric method, as reported by Mandal et al.,31 with
slight modification. After freeze-drying, the dry weight (Wd) of each sample was
measured. The sample was then placed in distilled water and left to swell at 37
°C in a closed container, after which the weight of the swollen sample (Ws) was
measured. The swelling ratio was calculated according to the equation below

The measurements were carried out on independent samples for different

time intervals, 6, 12, 24, 48, and 72 h, using 24-well plates BC membranes (ø
15.6 mm; 1.9 cm2 ).
Biocompatibility Testing of the Samples by Use of Extracts. To study
the effect of the films on the cell behavior, the NIH-3T3 (mouse embryonic
fibroblast cell line) and HaCaT (human skin keratinocytes) cell lines were used.
Cells were routinely cultured in culture flasks in high glucose (4.5 g L−1 ) DMEM
containing Lglutamin and pyruvate (110 g L−1 ), supplemented with 10% FBS
(Gibco, U.S.A.) and 1% penicillin/streptomycin, at 37 °C under an atmosphere of
5% CO2. While fibroblasts were passaged at 70−80% confluence using
trypsin/EDTA, keratinocytes were subcultured at a confluence of 80−90%. For
both cell lines, culture media were changed every 2−3 days. For in vitro tests,
cells from passages 6−9 were used.
The cytotoxicity of the samples’ extracts was tested according to the
international standard ISO 10993-5-2009 dedicated to in vitro evaluation of
medical devices (Reference number. ISO 10993- 5:2009(E)). Prior to extraction,
unmodified BC and BC−SS samples were washed 3 times in PBS (1X, pH 7.4)
and conditioned in culture media for 30 min at room temperature. The extracts
were then obtained by incubating the scaffolds (BC and composites, 2−3 mm
thick) at 37 °C in fresh complete culture media at the ratio of 1.25 cm2 /mL.
Before testing, cells were seeded at 10 000 cells/well directly into 96-well
cell culture plate’s wells and cultured for 24 h. Thereafter, the culture media were
fully replaced with the extracts, and the cells were incubated further for 1 and 3
days. Normal complete cell culture medium was used as control. The cell viability
was measured by MTT assay. Briefly, the yellow MTT powder was solubilized in
PBS (5 mg mL−1 ), microfilter-sterilized (Minisart high flow syringe filters, pore
size 0.2 μm, Sartorius Stedim), and used as indicated by the manufacturer. The
MTT solution was aseptically added to ongoing cultures at a ratio of 1/10 relatively
to the volume of culture media in individual wells, after replenishment of the latter.
Cultures were put back to incubation for 4 h, after which the MTT was converted
to purple formazan crystals by mitochondrial dehydrogenases in viable cells. A
syringe (1 mL) was used to gently remove the supernatant medium, as the
crystals were formed on the bottom of the wells. Formazan was dissolved at 37
°C, using DMSO as solvent. The optical density (OD) in different wells was read
at 492 nm, using a microplate reader (318C Microplate Reader). The cell viability
which directly correlates with the absorbance of solubilized formazan solutions
was expressed as the OD of representative wells, after blank corrections. The
experiments were performed in triplicate.
AlamarBlue Assay. Fibroblast and keratinocyte cell proliferation on the
constructs was quantified by alamarBlue assay, respectively. The test works on
the principle that metabolizing cells reduce the blue oxidized form of the dye into
a red form, which absorbance or fluorescence may be monitored. Cells were
seeded onto the scaffolds (10 000 cells/scaffold) in 96-well plates, after
confluence on culture flasks. Before cell seeding, the samples were prepared as
described above. They were washed 3 times in PBS and incubated in culture
media for 30 min at room temperature. They were then laid in new wells and kept
in the cell incubator for about 10−15 min to level their temperature to 37 °C. The
cell suspension was then seeded onto the scaffolds in a dropwise manner, and
put to incubation for 2 h to allow the cells to settle and attach to the membranes.
Culture media were thereafter gently added into the wells, and the plates were
incubated as routinely done. Empty polystyrene cell culture wells (designated as
PS) were used as control, while two types of blanks, medium only and BC +
medium (without SS modification), were employed. The culture was stopped at
different time-points to assess the cell response to the scaffolds with different
specimens being used for the different timepoints. For each sample type, the
experiment was done in triplicate.
For our scaffolds, the alamarBlue assay was performed as follows: At
different time-points (1, 3, and 5 days), exhausted media from the cultures were
replaced with fresh ones, and the cultures put back to incubation for at least 1 h
before addition of alamarBlue solution at a ratio of 1:10 relative to the volume of
culture media per well. The test plates were incubated for 3 h in the cell incubator
after which aliquots of the supernatant mixture were taken for measurements.
The absorbance was read at 570 nm with the reference wavelength of 630 nm in
a microplate reader (318C Microplate Reader), and the percentage of reduced
alamarBlue that reflects the cell growth was then calculated for each sample.
Confocal Microscopy. To visualize the cells onto the tested samples, the
seeded cells were stained for actin filaments. Coverslips (CS) were used as
controls. The cell-laden scaffolds and CS were washed in PBS and fixed with 4%
formaldehyde for 20 min, followed by a permeabilization step with 0.1% triton X-
100 after a second washing step. They were then washed and stained for 20 min
at room temperature with FITC-Phalloidin solution (15 μg mL−1 ) made up in
DMSO and diluted in PBS. The constructs were washed again and
counterstained for 30 min at room temperature with Hoechst 33342 diluted in
purified water (5 μg mL−1 ). The constructs then went through a final washing
step with purified water before being kept in PBS until mounting for microscopy.
CS were mounted using mounting medium, while the cell-loaded scaffolds were
mounted in PBS. Confocal images were captured using Ti Nikon Eclipse confocal
microscope.

STATISTICAL ANALYSIS
For comparison among groups, One-way Analysis of Variance (ANOVA)
test was used with Tukey’s Multiple Comparison test for the posthoc pairwise
comparisons in OriginPro 8 software. Statistical significance was obtained at P <
0.05.
RESULTS AND DISCUSSION
Characteristics of Hot Water-Extracted SS. For the purpose of this
study, SS was extracted from Bombyx mori silk cocoons in water, by high
pressure and temperature technique (Figure 1). Contrary to the enzymatic and
chemical extractions, the hot water extraction produces purer SS protein devoid
of impurities and byproducts that may be derived from sample processing when
using enzymes or chemicals. Additionally, this method has the advantage of
yielding higher methionine and cysteine contents in SS, two amino acids that
support the most important aspects of wound healing, that is, cell growth and
collagen production. On the basis of the ability of proteinaceous compounds to
absorb UV light due to peptide bonds (absorb between 215 and 230 nm) and
aromatic amino acids (absorb between 260 and 290 nm), UV spectroscopy was
employed to characterize the SS extract. Two absorption peaks at 276 and 216
nm were distinguished in the absorbance profile of the sample (Figure 2b),
corresponding to aromatic amino acids and peptide bonds, respectively. Similar
to previous reports, peptide bonds presented the highest absorbance, confirming
the minor representation of aromatic amino acids in SS. Interestingly, the
absorption peak for aromatic amino acids was registered at a wavelength very
close to that reported for a commercial SS sample (275.4 nm), ensuring the good
quality of the SS sample.

Figure 2. Sample characterization by FTIR (a) and UV spectrophotometry (b) and release profile
of SS from the composites scaffolds (c). n = 3.

Morphology and Structure of BC−SS composites. Various BC−SS

composites were obtained by immersing BC membranes in SS solutions (1, 2
and 3%, w/v), yielding 3 types of samples, the BC-SS1:24, BC-SS2:24 and BC-
SS3:24 scaffolds, respectively. The incorporation of SS within the BC network
was verified by morphological analysis using SEM imaging (Figure 3). The BC
matrix was formed by randomly organized fiber network typical of this material,
while SS presented as an amorphous and smooth structure with a few particles
as previously reported. In the composites, SS was visualized as coverage upon
the BC fibers, assembling them to form larger ribbons in the matrix, which
confirms its gluelike character.15 Additionally, SS could be distinguished within
the void spaces of BC matrix, leading to an overall increased density within the
network with increased SS content. However, the randomly entangled fiber
network of BC remained well preserved despite the presence of SS,
demonstrating the good compatibility between the two components and the good
integration of SS within the BC network without disruptive effects upon the latter.
The top of the composites was made by a smooth layer of SS (data not shown)
comparable to that observed for earlier reported BC− collagen composites.39 All
structures proved to be homogeneous, presenting numerous interconnected
pores that would enable gas exchange and easy absorption of wound exudate,
important parameters for a wound dressing. Moreover, these features would
contribute to successful applications in tissue engineering, allowing for nutrient
and signaling molecules circulation, while facilitating waste removal, for overall
enhanced cell attachment and cell migration.

Figure 3. SEM images of the samples. Scale bar = 10 μm.

The microstructure of the scaffolds was further evaluated by BET analysis.
The data are summarized in Table 1. As a result of the increased density
observed by SEM imaging, a decrease in the specific surface area was verified
within the composites as compared to BC, which was more pronounced at higher
SS concentrations. Similar trends were reported when BC was modified with
hyaluronic acid. Consequently, the same tendency was featured in the total pore
volume which directly correlates to the specific surface area of individual
scaffolds. However, the average pore diameter was not significantly different
among scaffolds, indicating a decrease in the overall number of pores within the
scaffolds in addition to the evidenced decrease in the total pore volume,
consistent with the morphology observed in the SEM images. Overall, while
confirming the nanoporosity of BC structure, these results attest to the successful
modification of BC with SS by solution impregnation without cross-linking.
Table 1. Porosity and Swelling Ratio of BC and Composites (n = 3).

Chemical Interactions between BC and SS within the Composites.

The modification of BC with SS was explored by FTIR analysis (Figure 2a).
Pristine BC exhibited a typical spectrum of bacterial cellulose with distinctive
peaks of cellulose chains at 3347 cm−1 (hydroxyl groups and interand
intramolecular hydrogen bonds), 1635 cm−1 (carbonyl groups of the glucose),
1053 cm−1 (C O C stretching), and 899 cm−1 β-glycosidic linkages). In SS
spectrum, the random coils were represented by the amide II and amide I peaks
at 1512 and 1635 cm−1 , respectively. The amide III peak reflecting the presence
of β-sheets was detected at 1240 cm−1 . Meanwhile, a broad peak centered at
3274 cm−1 was registered, which was assigned to the overlapping of free
hydroxyl groups and the vibration of −NH groups. For the composite samples,
both the characteristic peaks of BC and SS could be observed, demonstrating
the successful modification of BC by impregnation in SS solutions. More
precisely, the peak of BC at 3347 cm−1 merged with that of SS at lower
wavenumber around 3274 cm−1 , suggesting a decrease in hydrogen bonding
within BC. In the composites, this peak also decreased in intensity, suggesting a
lower number of free hydroxyl groups probably due to their involvement in
hydrogen bonds formation with other functional groups in SS. This interaction
was further verified by the shrinking of the BC peak at 2897 cm−1 related to
aliphatic -CH groups. Besides, a fainting of the BC peak at 1425 cm−1 (vibrations
of -CH2-) was detected, implying the involvement of the carbon on the position 6
of the glucose, while that of the peak at 1053 cm−1 suggests a contribution from
the C-O-C groups in BC. The changes were more pronounced with increased SS
concentrations.
Structural Stability of the Composites. To verify the physical stability of
the scaffolds and to anticipate their behavior in physiological conditions, the
release profile of SS from the constructs was determined in PBS (pH 7.4, 37 °C)
by BCA protein assay. The results plotted in Figure 2c showed a burst release
for all samples during the first 15 min following immersion in PBS solution due to
a diffusion of excess SS that remained on the surface of the BC membranes
without involvement in molecular interactions with the latter. This was followed by
a gradual release of SS with a stabilization phase indicating the successful
preparation of stable BC−SS biocomposites without cross-linking. Furthermore,
the amount of SS released from the composites depended on the concentration
of SS used for sample preparation, revealing a controllable behavior. Hence, at
physiological pH and temperature, soluble SS could be released in a controlled
manner from the constructs. This would allow the bioactive molecule to exert its
beneficial effects on cell growth by reaching the neighboring cells and tissues
through diffusion.
 Thermal and Mechanical Stability. Because of the need for sterilization
at high temperatures before use, thermal stability constitutes an important
parameter for biomaterials. Additionally, for wound dressing applications in
severe burns that often become hyperthermic due to extended inflammation
(body temperature around 40 °C), biomaterials are required to be thermally
stable, considering the large extent to which the body temperature may vary in
such patients. The thermal stability of our BC−SS biomaterials was thus
evaluated by TGA. The spectra of the samples are presented in Figure 4a.
Respectively to the thermal degradation of BC (which constitutes the main chains
in the composites) that encompasses dehydration, depolymerization,
decomposition of the glycosyl units, and ends by the formation of a charred
residue, the thermal degradation of the samples occurred in 2 steps. A
dehydration phase was registered from room temperature to around 230 °C, and
a main-degradation phase that corresponds to the degradation of the main
components (BC and SS) came about between 240 and 400 °C. SS presented
the highest water loss in comparison with BC, indicating its higher ability to absorb
moisture from the ambient medium. In wound dressing applications, this would
limit the drying of the wound thereby providing the appropriate moisture at the
wound bed, favoring tissue regeneration. Interestingly, this ability was slightly
improved in the composites when considered with unmodified BC membranes.
Also, in the main degradation phase no significant difference was observed
between the composites and pristine BC, which is in agreement with the SEM
results that suggest a good compatibility between BC and SS allowing for the
incorporation of SS within the BC network without disruption of BC’s main
structure. Pure SS however showed a rather irregular spectrum, confirming its
amorphous structure.

Figure 4. Physical properties of the samples. (a) = TGA, (b) = water uptake ability; (c) = Young’s
modulus; (d) = stress−strain curve by tensile testing. n = 3.
 Because physical stability is one of the leading criteria in the choice of a
material for biomedical applications particularly tissue engineering, the scaffolds
were further subjected to mechanical testing, namely, tensile testing. The
stress−strain curves and the Young’s modulus of the samples are shown in
Figure 4d,c, respectively. Unsurprisingly, BC exhibited the highest mechanical
properties, which were subsequently slightly altered upon SS incorporation.
Because of its amorphous nature, SS is a material devoid of mechanical strength
often leading to the decrease of the overall mechanical strength of the composite
after blending with polymers of higher mechanical stability such as BC, as
reported for SS-collagen composite films. Nevertheless, the Young’s modulus of
BC samples was found not significantly higher when compared with the
composites’ (Figure 4c).
Swelling Ability. Figure 4b represents the profiles of water uptake ability
measured from the scaffolds in distilled water. Similar trends were observed in all
samples, with a rapid swelling after 6 h and a decreasing swelling ratio with higher
SS content, as reported for previously developed SS-carboxymethyl cellulose
composites. As suggested by the FTIR spectra, with increased protein content,
the incorporation of SS between the molecular chains of BC leads to a weakening
of the hydrogen bonds within the network, diminishing the ability of the latter to
lock water molecules. Moreover, as discussed in the previous section, increased
SS amounts result in decreased sample porosity, restricting the free space
available for water uptake. Hence, by varying the SS concentration, different
levels of swelling ratio can be obtained, which would be interesting for the
treatment of wounds with different exudate levels. Overall, all samples
demonstrated high swelling abilities, as the smallest swelling ratio (22.92 in BC-
SS3:24 samples) represented more than 20-fold the weight of the dry samples
and remained several folds higher than the values reported elsewhere for
potential material candidates for biomedical and tissue engineering applications.
Thus, consistently with the SEM results, all scaffolds remained highly porous after
incorporation of SS, allowing for the high swelling observed in water.
Effects of BC−SS Composites on Cell Behavior. To test the
biocompatibility of the samples, NIH-3T3 fibroblast cell line was cultured in the
samples extracts, using normal culture media as control (Figure 5a). Fibroblasts
are key players in wound healing. Aside from the production of extracellular
matrix proteins (mainly collagen necessary for wound closure), they coordinate
the phases of wound healing by producing cytokines and growth factors. In our
study, in all samples no cytotoxicity was detected at either D1 or D3. Furthermore,
the extracts obtained from the composites significantly enhanced the cell viability
when compared with the control at D1. Similarly, higher cell viability was observed
for the cells cultured in composites’ extracts at D3 in comparison with those
cultured in normal culture media with a significant increase for the BCSS3:24
samples. Importantly, the cell viability with the composites’ extracts was
enhanced relatively to BC at D1 with a very significant difference at D3. A dose-
dependent effect was observed, based on the concentration of SS solution used
for composite preparation. Additionally, there was no significant difference
between the effects of BC extracts and normal cell culture media. These data
thus suggest that the positive effect of BC−SS samples’ extracts on fibroblast cell
viability resulted from soluble SS released from the scaffolds, which promoted
cell proliferation without inducing any cytotoxicity. Hence, beneficial behaviors
may be expected from these composites both in in vitro and in vivo applications,
as active SS molecules would diffuse into the milieu, assisting cell growth thanks
to sericin’ s cytoprotective and mitogenic abilities. When the extracts were tested
on keratinocytes (Figure 5c), higher cell viability was observed with the
composites relatively to pristine BC at D1, which was found statistically
insignificant except for the cells cultured in extract from BC-SS3:24 samples.
Moreover, at D3 no significant difference was noted between scaffolds. Hence,
in accordance with the research by Akturket al., SS did not have a significant
effect upon keratinocyte cell growth. However, interestingly, the cell viability with
all scaffolds’ extracts was found significantly higher as compared to that with
normal cell culture media, indicating a good biocompatibility.

Figure 5. In vitro biocompatibility of the constructs. (a,c) Viability of fibroblasts and keratinocytes
cultured in samples’ extracts, respectively: MTT assay. (b,d) Viability of fibroblasts and
keratinocytes cultured onto the scaffolds, respectively: alamarBlue assay. (*) Comparison
between scaffolds and control (PS) for the same time-point; (#) comparison of the composites’
extracts with BC extract for the same time-point; (+) comparison with D3, for scaffolds. (***) P <
0.001; (**) P < 0.01; (*) P < 0.05; (###) P < 0.001; (##) P < 0.01; (#) P < 0.05. n = 3.

The cell attachment and proliferation on the substrates were further

examined by alamarBlue assay, which monitors the live cell’s overall reducing
environment, based on the activity of mitochondrial dehydrogenases and
cytochromes. It has the advantage of being simple in execution without requiring
cell killing, which allows for the use of the same samples for different sets of
experiments. At D1, the fibroblast cells attached to the scaffolds independently of
the SS content, as similar cell viability values were obtained for all samples
including the pristine BC membranes, inferring that cell attachment in all scaffolds
was exclusively regulated by BC’s surface properties that supplanted the effects
of soluble SS observed in the MTT assays (Figure 5b). Hence, incorporating SS
within the BC network did not significantly alter the interfacial features of the
resulting biomaterials, which is a determinant factor in cell adhesion. AlamarBlue
reduction levels in the scaffolds groups were comparable to that observed in the
PS group (polystyrene plate, control group), confirming the biocompatibility of the
prepared scaffolds. Monitoring the cell growth at D3 and D5 indicated that no
significant cell proliferation occurred on both scaffolds, suggesting a slow initial
cell growth onto the scaffolds. This was further confirmed by confocal microscope
images that exhibited round and nonspread cells on the scaffolds after 3 days of
culture (Figure 6). More accurately, on the CS control the cells were well spread,
exhibiting their characteristic spindle shape with central nuclei, while on BC and
the composites a lower extent cell spreading was observed, which was better in
BC-SS1:24 and BC-SS2:24 as compared to the other scaffolds. This behavior
may be due to the reported surface tension and stiffness of BC, which may limit
cell spreading and migration, and thus cell division. Similarly, Favi et al., reported
that due to the higher stiffness of PS substrates and the softness of BC,
mesenchymal stem cells displayed flat and spread-out morphology on PS while
presenting a round shape on BC membranes up to 14 days in culture. What is
more, it has been demonstrated with fibroblasts that cells spread more on stiffer
surfaces such as the polystyrene plate or the coverslips than they do on softer
surfaces such as BC. Hence, longer cell culturing times might be needed in the
context of this work to allow for successful fibroblast cell spreading, growth, and
proliferation. However, it is important to notify that same level of cell viability was
registered on different scaffolds from D1 to D5 in the alamarBlue assays,
underlining the absence of any cell toxicity. Meanwhile, relatively higher cell
numbers were observed on the composites as compared to BC membranes in
the confocal images. Same observation applied to keratinocyte cells from D1 to
D3 (Figure 5d), with no significant cell growth occurring upon the scaffolds
(including BC) within the aforementioned time period. However, at D5, the cell
viability in each group of the scaffolds was significantly higher than that at D3,
indicating an active cell proliferation between the two time-points. Nevertheless,
no significant difference was observed between BC and BC−SS composites at
all time-points, confirming that SS did not significantly influence the cell behavior
of keratinocytes cultured onto the composite materials in the context of this work.
These results were in agreement with the confocal images (Figure 7). On the
coverslips, keratinocyte cells developed their typical “cobblestone” pattern made
of cells closely joined one to the other. Meanwhile, on the BC and its composites
the cells remained round with higher cell numbers in the composites as indicated
by bigger spherelike structures showing higher nuclei numbers.

Figure 6. Confocal imaging of NIH-3T3 cells cultured onto the scaffolds (10 000 cells each) for 3
days. (Green, actin filaments; blue: nuclei, scale bar, 100 μm).
Figure 7. Confocal imaging of HaCaT cells cultured onto the scaffolds (10 000
cells each) for 7 days. (Green, actin filaments; blue, nuclei; scale bar, 100 μm).

CONCLUSIONS
In this study, bacterial cellulose−silk sericin composite biomaterials were
successfully obtained by solution impregnation method without cross-linking. A
good compatibility was observed between BC and SS, as the latter integrated
perfectly the BC network, covering the random fibers and maintaining the native
BC’s nonwoven fiber structure that is attractive for wound dressing application.
Consequently, no significant alteration was observed in the thermal and
mechanical stability of the BC network upon SS incorporation. Within the
composites, the two macromolecules were stabilized by hydrogen bonding, as
shown by the FTIR spectra. All structures were highly porous and presented
interconnected pores necessary for gas exchange and absorbing of wound
exudate. Besides, SS has the ability to enhance oxygen permeability, which is an
important factor for proper wound healing. These structures may also be good
candidates for tissue engineering applications, as they would enable easy
nutrient and metabolites circulation, enhancing the overall cell viability. The
swelling ability of the constructs varied with the SS content, with high capacity for
water absorption in all scaffolds. Thus, they may be used for different types of
wounds according to their exudate level. The SS release assay confirmed the
stability of the samples, as SS effectively remained in the structures while being
released in a SS content (initial SS concentration used for the preparation of the
composite)- dependent manner. Thereby, the released SS whose amount can be
controlled would interact with cells and exercise its benefial biological effects,
mainly its mitogenic effect, thus accelerating the healing process and resulting in
reduced scar formation. This was further confirmed in vitro, as extract solutions
from the composites significantly enhanced the viability of fibroblast cells, in
comparison with both BC and normal cell culture media. However, the viability of
keratinocytes was not significantly influenced by the presence of SS. AlamarBlue
assay performed on cells cultured directly onto the scaffolds indicated no
difference between pristine BC and BC−SS composites, while inferring the need
for longer time periods to effectively allow fibroblast cell profileration upon the
constructs in the experimental conditions employed in this work. Overall, while
further in vivo investigations are required to fully validate these materials, it can
be concluded that they represent potential candidates for wound healing and
tissue engineering applications.

ABBREVIATIONS
BC: bacterial cellulose; SS: silk sericin; HS medium: Hestrin and Schramm
medium; PEG: poly(ethylene glycol); SEM: scanning electron microscope; FTIR:
Fourier transform infrared; PBS: phosphate-buffered saline; TGA:
thermogravimetric analysis; BCA: bicinchoninic acid; BET: Brunauer−
Emmett−Teller; Ws: swollen weight; Wd: dry weight; ATCC: American Type
Culture Collection; DMEM: Dulbecco’s Modified Eagle Medium; FBS: fetal bovine
serum; EDTA: ethylene diamine tetra acetic acid; MTT: 3-[4,5-dimethylthiazol-2-
yl]-2,5-diphenyl tetrazolium bromide; DMSO: dimethyl sulfoxide; OD: optical
density; PS: polystyrene cell culture plate; CS: coverslips; FITC: Fluorescein
isothiocyanate.