Wet electrospun silk fibroin/gold nanoparticle 3D matrices for wound healing applications

Abstract

This study aimed to fabricate 3D silk fibroin (SF) matrices for skin tissue engineering applications. SF/poly(ethylene oxide) solutions were wet electrospun to obtain a fibrous network (0.7–20 μm diameter), which were then lyophilized to obtain 3D porous nanofibrous matrices (SFM-E: ethanol treated silk fibroin matrices). SF matrices were loaded with citrate-capped gold nanoparticles (AuNPs, 14.27 ppm, D_average = 24 nm) (SFM-AuE: ethanol treated silk fibroin matrices incorporated with AuNPs) and investigated for structural and chemical properties, in vitro biocompatibility and in vivo full-thickness dermal wound healing efficacy in a rat model. AuNP incorporation enhanced the degradation profiles and mechanical properties significantly. SFM-E and SFM-AuE showed similar cell attachment and layer by layer proliferation behaviour, but cells had more spread and flattened morphology on SFM-AuE. Both matrix extracts had high cell viability (>90%), indicating good in vitro biocompatibility. Wound closure was statistically more than the untreated skin control (UTSC) in SFM-E and SFM-AuE applied groups. The recovered tensile strength and elastic modulus of SFM-E and SFM-AuE (40–60%) were not as high as the unwounded skin control (UWSC), but they had elongation at break values similar to UWSC. This was attributed to the still ongoing medium to high inflammation levels leading to a low and immature extent of collagen fibrils on postoperative 14th day. There was only a small amount of epithelialization due to scab formation and medium to high level inflammation for both SFM-E and SFM-AuE, but they were better than UTSC in terms of neovascularization and granulation tissue formation. As a whole, inclusion of AuNPs to SF matrices at 14.27 ppm loading brought some enhancement in the matrix properties and did not cause any toxicity in in vitro and in vivo conditions and even had potency to promote wound healing stages.

---

Introduction

Silk fibroin extracted from cocoons of the Bombyx mori silkworm has been utilized as a promising biomaterial for skin tissue engineering due to its advantageous properties such as good biocompatibility¹ adjustable mechanical properties,^2–4 controlled degradation profile^5,6 and positive influence on wound healing performance.^7,8 Silk fibroin has been tailored into various shapes including films,^9–11 sponges,^12,13 hydrogels,^14,15 and fibers fabricated either by electrospinning,¹⁶ wet electrospinning,¹⁷ wet spinning² or microfluidic solution spinning.¹⁸ Among these shapes, in particular, 3D fibrous matrices fabricated by means of wet electrospinning have been stated to be very beneficial in skin tissue engineering¹⁹ because they can mimic the ultrastructural properties of the native extracellular matrix of skin. Therefore, many studies related to skin tissue engineering have focused on developing electrospun matrices made up of pure silk fibroin²⁰ or silk fibroin blended with natural^21,22 or synthetic polymers.²³

It has already been reported that silk fibroin based biomaterials promoted attachment and proliferation of keratinocytes and fibroblasts²⁴ and provided very promising results in wound healing tests.⁸ Many researchers also modified the structure¹¹ or the surface²⁵ of silk fibroin scaffolds or incorporated them with epidermal growth factor,²⁶ silver nanoparticles²⁷ and epidermal growth factor/silver sulfadiazine combinations²⁸ in an aim to enhance the wound healing process by increasing the cell attachment/proliferation (i.e. granulation tissue formation), increasing the rate of neovascularization, reducing the inflammation and preventing the risk of infection.

In the present study, silk fibroin nanofibrous matrices were incorporated with citrate-capped gold nanoparticles as a novel approach to treat skin wounds. Incorporation of AuNPs into tissue scaffolds were already shown to enhance the mechanical stability, the resistance against enzymatic degradation^29,30 and biocompatibility^31,32 of tissue scaffolds. Also, antibodies,³³ growth factors³⁴ and peptides³⁵ were easily incorporated at the gold surface. Furthermore, AuNPs could make the scaffolds gain antioxidant³⁶ and antimicrobial properties.^37–39 Nanocomposite SF/AuNPs structures such as SF nanofiber surfaces modified with AuNPs,⁴⁰ SF nanofiber nerve conduits⁴¹ and SF capped AuNPs⁴² were also investigated recently for different biomedical applications. As a whole, these properties of AuNPs could make them promising candidates for wound healing applications. On the other hand, it was also reported that AuNPs might cause cytotoxicity problems, depending on the shape,⁴³ size,⁴⁴ dose⁴⁵ of AuNPs and type and charge of surface capping/reducing agents used for the synthesis of AuNPs.⁴⁶ The nanotoxicity of AuNPs, due to so many factors being involved, is very complex to understand, and therefore, the results in literature are very controversial. There is now a strong tendency to utilize natural and biocompatible reducing/capping agents such as gelatin,⁴⁷ collagen,⁴⁸ silk fibroin⁴² and plant extracts⁴⁹ for the synthesis of AuNPs. These biomolecules could replace potentially toxic reducing/capping agents such as sodium citrate and increase the stability of AuNPs in aqueous environments and so the potential cytotoxicity problems might be avoided.

Although there is a multitude of studies related to the biomedical use of AuNPs as mentioned previously, only little is known about its in vivo effects such as induction or inhibition of inflammatory immune responses and contribution to skin tissue regeneration. It is hypothesized that the inherent properties of AuNPs, listed above, might bring extra advantages for wound healing when incorporated into silk fibroin scaffolds. Therefore, it was aimed to investigate the effectiveness of this novel SF/AuNPs combination on dermal wound healing in a rat model. To assess the wound healing performance, one of the most common techniques, a full-thickness skin wound model (cutaneous excisional wound), was applied.^50–52 The healing process of such a wound proceeds through angiogenesis, granulation tissue formation, wound closure and re-epithelialization.^53,54 Besides that, biomechanical tests were also conducted on wounded skin specimens to understand the recovered mechanical properties of the wounded skin as it is directly correlated to the wound healing. So, the wound healing performance of the developed matrices was determined by measuring to what extent these stages were succeeded.

Experimental

Materials

Silk cocoon of Bombyx mori origin was bought from Kirman İplik Co. (Bursa, Turkey). Poly(ethylene oxide) (PEO, average M_w ∼ 4 000 000), lithium bromide (ReagentPlus®, ≥99%), sodium carbonate (anhydrous, powder, 99.999%), dialysis tubing cellulose membrane (avg. flat width 25 mm), methanol (CHROMASOLV®, for HPLC, ≥99.9%), ethanol (CHROMASOLV®, for HPLC, ≥99.8%), glutaraldehyde solution (GTA, grade I, 25% in H₂O), collagenase from Clostridium histolyticum (type I-A, lyophilized powder), trisodium citrate dihydrate (Na₃C₆H₅O₇·2H₂), thiazolyl blue tetrazolium bromide (MTT bromide, approx. 98% TLC) were purchased from Sigma-Aldrich (Germany). Hydrogen tetrachloroaurate(III) (HAuCl₄·3H₂O) and dimethyl sulfoxide (DMSO, cell culture grade, min. 99.5%) was obtained from Applichem (Germany). Millex® syringe filter units (disposable, Durapore® 0.1 μm, PVDF) were obtained from Merck Millipore (Germany). Formvar carbon film coated (200 mesh) copper grids for TEM was bought from Electron Microscopy Sciences (USA). All other reagents and solvents used were of reagent grade and obtained from Merck Millipore (Germany), unless otherwise stated in the text.

Synthesis of gold nanoparticles

AuNPs were synthesized with citrate reduction method based on Turkevich method⁵⁵ as depicted in Fig. 1. Firstly, 1 g of HAuCl₄·3H₂O was dissolved in de-ionized water (DI water) to get 1 mM gold salt solution (solution I). Then, 20 mL was taken from solution I and stirred with magnetic stirrer while heating. While stirring at high speed (800 rpm), previously prepared 0.1% trisodium citrate dihydrate (Na₃C₆H₅O₇·2H₂O) solution (solution II) was added onto boiling solution I at a volume of 8 mL. When the color of the solution started to turn from pale blue (nucleation) to deep red (ruby red), the stirring at boiling point was continued for another 15 minutes. Immediately after this step, flask containing the solution was soaked in an ice bath to stop the reaction. Finally, AuNPs suspension was topped to 30 mL and passed through 0.1 μm filter to get AuNPs stock (157 ppm, D_mean = 24 ± 3 nm). High contrast TEM analysis was done using an FEI Tecnai G2 Spirit BioTwin operating at 120 kV to evaluate the size and shape of AuNPs in the Central Laboratory of Middle East Technical University (METU).

Fig. 1 Fabrication of SF/AuNPs matrices with 4 main steps: preparation of spinning solutions (A), wet electrospinning (B), chemical treatment of SF fibers (C) and freeze-drying (D).

Fabrication of scaffolds

Preparation of spinning solutions. Cocoons (5 g) were boiled for 30 min in an aqueous solution of 0.02 M Na₂CO₃ (2 L) and then rinsed thoroughly with DI water 3 times (20 min each) for 1 hour to extract the glue-like sericin proteins. The extracted SF (3.5 g) was then dissolved in 9.3 M LiBr solution at 60 °C, yielding a highly concentrated gel-like solution. This solution was dialyzed against DI water using a dialysis membrane for 3 days and changing the DI water two times every day. The final concentration of SF in aqueous solution was about 10%, which was determined by weighing the remaining solid after drying a particular volume of this solution in an oven (60 °C) for 2–3 hours. During preliminary studies, at first the pure SF solutions at different concentrations (1–10%, in d-water) and then the groups of SF/PEO blend solutions were prepared to generate stable, continuous spinning (Table 1). For instance, SF7.2P0.3 solution was prepared by adding 1 mL of PEO solution (2% in DI water) to 5 mL SF solution (8.6% in DI water). Homogeneous solutions were obtained by mildly stirring for at least 2 hours at RT. The solutions could be waited at 4 °C as long as 2 weeks without any observable changes (premature β-sheet formation) in the blends. SF was incorporated with AuNPs in a similar fashion. SF7.2P0.3-Au was prepared by adding 1 mL of PEO solution (2% DI water) containing AuNPs (85.62 ppm) to 5 mL of SF solution (8.6% in DI water) to get final concentration of AuNPs as 14.27 ppm in the polymer blend solution (Fig. 1(A)).

Table 1 Optimization of SF/PEO wet electrospinning blend solutions

Groups	SF concentration in final solution (% w/v)	PEO concentration in final solution (% w/v)	Polymer ratio, SF : PEO (w/w)	Total polymer concentration (% w/v)
SF1	1	—	1 : 0	1
SF5	5	—	5 : 0	5
SF10	10	—	10 : 0	10
SF8.3P0.3	8.3	0.3	25 : 1	8.7
SF7.2P0.3	7.2	0.3	21.5 : 1	7.5
SF7.2P0.3-Au	7.2	0.3	21.5 : 1	7.5

---

Wet electrospinning. The flow rate of the SF solutions dripping out from the syringes (blunt needle, 18 G) was adjusted as 0.8–2 mL h⁻¹ with a syringe pump (New Era, USA). The fibers were collected in ethanol bath (96%) placed on metal collector rotating at 30–60 rpm speed and located perpendicular at a 13–23 cm distance away from the syringe needle. By adjusting the voltage (12–19 kV) of the high voltage power supply (Gamma, USA) connected to the needle tip, the continuity of the solution jet flow and the fiber accumulation inside the ethanol bath was achieved (Fig. 1(B)). The electrospun SF fibers in ethanol were then transferred into glass beakers and chemically treated.

Chemical treatment of silk fibroin fibers. Different chemical treatment methods such as ethanol (96%), methanol (100%) and GTA (3% in DI water) were tried to enhance the stability of the SF matrix (SF7.2P0.3 group) in aqueous environment. In order to form β-sheets, SF fiber suspension was placed in ethanol overnight at 4 °C (SFM-E), or put into methanol for 5 min (SFM-M), or treated with 3% GTA solution for 2 hours (SFM-G) at RT. All treated groups were rinsed with DI water afterwards (Fig. 1(C)).

Freeze-drying. After water was drained, SF fibers were poured to 6 well plates. After that, it was frozen at −20 °C for 1 day and the frozen fibrous mat was freeze-dried (0.050 mBar and −80 °C) for 2 days. Finally, SF matrices with sponge-like (about 1–2 mm in thickness) porous nanofibrous structure with silky soft surface were obtained (Fig. 1(D)). Eventually, silk fibroin matrices treated with ethanol (SFM-E), methanol (SFM-M) or GTA (SFM-G) and ethanol treated silk fibroin matrices incorporated with AuNPs (SFM-AuE) were put in a desiccator until they were used in the experiments.

Structural and chemical characterization

Scanning electron microscopy. Electrospun SF matrices were sputter-coated with gold–palladium (E1010, Ion Sputter, Hitachi, Japan) before analysis by scanning electron microscopy (SEM; S-3000H, Hitachi, Japan) at the Department of Metallurgical and Materials Engineering, METU. For each matrix, three random micrographs were taken and diameters of at least 100 randomly selected fibers from each micrograph were measured using the ImageJ program. Fiber diameter and the distribution of fiber sizes of each matrix were averaged from at least 300 fibers.

Fourier transform infrared spectrophotometer analysis. Infrared spectra of SF film, sericin extracted SF fibers, and chemically treated groups (SFM-E, SFM-M, SFM-G and SFM-AuE) were investigated with an attenuated total reflectance Fourier transform infrared (ATR-FTIR; IFS/66S, Hyperion 1000) spectrophotometer in the Central Laboratory of METU. Each spectrum was acquired in transmittance mode on a ZnSe ATR crystal cell by accumulation of 256 scans with a resolution of 4 cm⁻¹ and a spectral range of 4000–400 cm⁻¹.

Porosity and pore size distribution. For measuring the pore size distribution of SF matrices, mercury porosimeter (Quantachrome Corporation, Poremaster 60) low pressure analysis (0–50 psi) was conducted. In brief, the pores of scaffolds are filled with mercury by changing the pressure from 0 (large pores) to 50 psi (relatively small pores) values. Data showing the pore volume (mercury volume inclusion in the pores/the weight of scaffold) vs. the corresponding pore sizes is acquired by the porosimeter. Cumulative pore volumes divided by the total pore volume normalized to scaffold weight (% per g) vs. the determined pore size interval was drawn to get pore size distribution histogram figures. Porosity of the SF matrices was measured by liquid displacement method. Hexane was used as the displacement liquid. A pycnometer was used for porosity calculation. First, the empty (W₀), and the full weight (W₁) of pycnometer with n-hexane was measured. Dry sample (W_s) was immersed in pycnometer to top and air bubbles were removed by vacuuming for 10 minutes. Afterwards, the weight of pycnometer with wet sample (W₂) was measured. Finally, the wet weight of sample (W₃) was obtained. The following equations were used for the porosity calculations (n = 4):

W_W1 = W₁ − W₀

W_W2 = W₂ − W₀ − W_s

V_S, V_P and P are skeletal volume, pore volume and porosity of sample, respectively and D_hexane (0.6548 g mL⁻¹) is the density of n-hexane.

Hydrolytic and enzymatic degradation. The degradation of SF matrices in aqueous environment with or without enzyme was analyzed. For the hydrolytic degradation (HD), the samples (1 × 1 cm²) were put into PBS (pH: 7.2, 0.01 M) and incubated at 37 °C for 2 weeks. For the enzymatic degradation (ED), the scaffolds were incubated in PBS (pH: 7.2, 0.01 M, 5 mL) added with 60 μL of type I collagenase solution (1 mg mL⁻¹) until complete degradation. Gravimetrical analysis was carried out by removing out the matrices at the specified incubation periods and measuring their dry weights to get the % degradation results using the equation (n = 4):
W_d denotes the initial dry weight of the test samples, while W_t represents the dry weight at the end of incubation time.

Mechanical tests. The mechanical properties of SF matrices were evaluated by applying stretch test with Lloyd LS500 Material Testing Machine (Lloyd, England) using Nexygen computer software. Rectangular samples (5 × 30 mm) were cut from scaffolds. The gauge length and width were 10 and 5 mm, respectively. The thickness of the test samples were measured by Vernier caliper and found to be about 1 mm. The crosshead speed of the system was adjusted to 10 mm min⁻¹ to get a constant strain rate of 100% per min. The results of tests were obtained as load versus deflection curves which were then converted into stress–strain data by the computer program. Tensile strength (TS), modulus of elasticity (E), and percent elongation at break (EAB) were calculated from the stress–strain curves (n ≥ 6). The tests were performed in either dry or wet testing conditions. For wet testing, the test samples were first equilibrated in PBS (pH: 7.2, 0.01 M) and then tested.

In vitro biocompatibility tests

Cell culture conditions. L929 mouse fibroblast (ATCC, USA) cell lines were cultured in complete cell culture medium, i.e. Dulbecco's Modified Eagle's Medium (DMEM, high glucose-glutamine) supplemented with fetal bovine serum (FBS, 10%, v/v) and penicillin/streptomycin (10 U mL⁻¹) at 37 °C under humidified atmosphere of 5% CO₂ and 95% air in an incubator (5215, SHEL LAB, USA). When the cells reached at least 80–90% confluency, they were trypsinized with trypsin/EDTA solution (0.1% in PBS) and counted with a cell counter (Nucleocounter, LICOR, USA).

Cytotoxicity of scaffold extracts. The cytotoxicity test of SF matrix extracts (SFM-E and SFM-AuE) was performed according to ISO 10993-5 standards. Briefly, matrices were initially exposed to UV light for 30 min and then held in 70% ethanol supplemented with antibiotics (penicillin/streptomycin, 10 U mL⁻¹) overnight for sterilization. After sterilization, the samples were rinsed with sterile DI water completely and held in complete cell culture medium (DMEM with 10% FBS) temporarily. Finally, they were incubated in extract medium (complete cell culture medium) at 37 °C for 24 hours while shaking at 200 rpm. Extraction was done in accordance with ISO 10993-12 standards. Complete cell culture medium without any scaffolds was used as control. After seeding L929 cells on 96 well plates (10 000 cells per well) for 1 day, the cells were checked for subconfluency. Then, medium was removed and refreshed with extract medium of samples. Cell viability was evaluated with methylthiazolyldiphenyl tetrazolium bromide (MTT) cell viability assay after 1 day. Briefly, MTT stock solution (5 mg mL⁻¹) was prepared by dissolving in PBS buffer. The MTT test solution was prepared by diluting the MTT stock solution in 1/10 ratio with DMEM (without phenol red). At the end of 24 hours, the medium of cells was removed and washed with PBS. Subsequently, 100 μL MTT test solution was added to each well. The test wells were incubated at 37 °C in a dark environment for 4 hours. Afterwards, MTT test solution was removed and 100 μL DMSO was added and wells were shaken at 200 rpm for 15 minutes. Spectrophotometric measurements were done at 570 nm primary wavelength and 650 nm reference wavelength (n = 6). The % cell viability was calculated by normalizing the extract sample OD results to that of control.

Cell attachment and proliferation on scaffolds. The sterilized matrices (SFM-E and SFM-AuE) were cut into rectangular shapes (1 × 1 cm) and put onto sterile 24-well suspension plates. Before the cell seeding, the samples were kept in DMEM briefly. Afterwards, L929 fibroblasts were seeded onto scaffolds at a seeding density of 100 cells per μL (100 000 cells per well) and held in a CO₂ incubator for 1 and 3 days. The cell culture medium inside the wells was refreshed every other day. Eventually, the medium was removed and the scaffolds seeded with cells were fixed in 3% GTA (in PB) solution for 1 h at RT. They were completely dehydrated in a freeze-dryer, then were coated with gold and analyzed by SEM (JSM-6400 Electron Microscope, JEOL Ltd., Japan) at the Department of Metallurgical and Materials Engineering, METU.

In vivo full-thickness skin wound healing tests

Animals. Thirty healthy adult male Wistar albino rats weighing 200–250 g were used. All the rats were obtained from Kobay Co. (Ankara, Turkey). The rats were housed in stainless steel cages in an animal room maintained at 22 ± 2 °C with a 12 hour alternating dark–light cycle. All were fed with the same amount of a laboratory pellet diet and water ad libitum and fasted for 12 hours before the procedures. The procedures in this experimental study were performed in accordance with the National Guidelines for the Use and Care of Laboratory Animals and approved by the Animal Ethics Committee of Ankara Training and Research Hospital.

Full thickness wound model. Three groups were randomly constituted of 10 rats each. The rats were anaesthetized with intramuscular injections of 50 mg kg⁻¹ ketamine hydrochloride (Ketalar, Parke-Davis-Eczacibasi, Istanbul, Turkey). The operation sites were shaved and disinfected with povidone-iodine. A 2 × 1 cm rectangular-shaped incision was made on the back of the rats centered on the midline, and then a standard full-thickness skin defect, including panniculus carnosus, was created on this site.

Treatment groups. All SF matrices were cut off into the same size and subsequently each skin wound of rats was softly covered with SFM-E or SFM-AuE (n = 10). In the untreated skin control group (UTSC), defects moisturized with physiological saline were covered with only surgical drapes (Ioban® incise drape, 3M Health Care, USA). In the study groups, SFM-E or SFM-AuE were applied onto previously moisturized defect sites and covered with surgical drapes. All wounds were followed for 14 days and no complications developed during this period. All rats were euthanized with high-dose ketamine hydrochloride on postoperative 14th day.

Histopathological analysis. The wound area was excised together with the scar tissue. All specimens were fixed in 10% phosphate-buffered formaldehyde solution for 24 hours at room temperature. Histopathological assays were performed in a blind manner by a pathologist. Specimens were washed in tap water and dehydrated through graded alcohol series. After passing through the routine histological series, tissues were embedded in paraffin blocks. Sections of 5 μm were cut, deparaffinized and rehydrated. Sections were counterstained with hematoxylin and eosin (H&E). The histopathological scores were given with respect to re-epithelization (Table 1 in the ESI†), inflammation (Table 2 in the ESI†), neovascularization (Table 3 in the ESI†) and granulation tissue formation (Table 4 in the ESI†) degrees. Inflammation, neovascularization and granulation tissue formation were scored on a numerical scale from 0 to 3 and the degree of re-epithelialization was scored on a numerical scale from 0 to 2.

The examination of wound closure. Open wounds were drawn on graph acetate paper with a marker pen on day 0 and 14. The surface area of the wounds was measured with a planimetric programme on computer by scanning the acetate sheets. The percentage of closure was calculated by the following formula:
where A₁₄ and A₀ denotes total wound areas on day 14 and 0.

Biomechanical tests. Biomechanical testing was performed on postoperative 14th day. The mechanical properties of the harvested biopsy specimens were evaluated by applying tension test with Lloyd LS500 Material Testing Machine (Lloyd, England) using Nexygen computer software. Rectangular samples (5 × 50 mm) were cut from the wound sites. After harvesting the skin samples, they were kept in sterile PBS solution (pH: 7.2, 0.01 M) and then the wet skin samples underwent biomechanical testing in the same day. Cranial–caudal orientation was preserved and the healed wound, if present, was centrally located within the testing unit. The gauge length and width were 20 and 5 mm, respectively. The thickness of the test samples were calculated by means of Vernier caliper and found to be about 1–2 mm. The crosshead speed of the system was adjusted to 10 mm min⁻¹ to get a constant strain rate of 50% per min. Before tests, all specimens were preloaded at 5 N to flatten the skin samples. To prevent slippage of the scaffolds from the grips they were covered with sand paper (60 grids) at the attachment site. The results of tests were obtained as load versus deflection curves which were then converted into stress–strain data by the computer program. Modulus of elasticity, ultimate tensile strength and percent elongation at break was calculated (n ≥ 6) from the stress–strain curves of samples. The biomechanical test results were expressed in terms of recovered percentage of modulus of elasticity (RE), tensile strength (RTS) and percent elongation at break (REB) by normalizing the mechanical properties of specimens treated with study groups by the corresponding values of the unwounded skin specimens.

Statistical analysis

Multiple comparisons between the groups were performed with one-way analysis of variance (ANOVA) and post hoc tests. Differences between the groups were analyzed with the Tukey's tests (HSD or LSD). Statistical analysis was performed with the SPSS-22 Software Programme (SPSS Inc., USA). Differences were considered significant for p < 0.05.

Results and discussion

Morphology analysis

According to the electrospinning experiments, fibrous structures could not be formed with only SF solutions. Therefore, obtaining nanofibers from only SF solutions (SF1, SF5 and SF10) was thought to be difficult. In order to be able to successfully electrospin SF, an additional polymer, PEO which dissolves in water was used. SF at 10% concentration would form very large diameter fibers greater than 50 μm, but 7–8% SF concentrations were found to be optimum to be able to form fibers with diameters at nano range. The operating parameters (flow rate, voltage and distance) of SF8.3P0.3 and SF7.2P0.3 groups were adjusted to get fibers with various sizes (0.7–20 μm) (Fig. 2). The most successful matrices (with homogenous fiber size distribution, fiber size in nanometer range, and smooth fibers) were obtained in SF7.2P0.3 group (Fig. 2(D1) and (D2)). SF fibers were connected to each other and created numerous pores whose sizes reached up to several tens of microns. AuNPs were also added into SF solution (SF7.2P0.3-Au) before the electrospinning process. AuNPs concentration in the SF solution was 14.27 ppm. In comparison to the SF group containing no AuNPs (SF7.2P0.3), SF7.2P0.3-Au had slightly smaller diameter nanofibers, but the size distribution range was a little bit wider with smaller and larger nanofibers (Fig. 2(E1) and (E2)). The original morphology of sericin extracted SF fiber was also compared with the reconstituted silk fibroin fibers fabricated in this study (Fig. 2(F1) and (F2)).

Fig. 2 SEM micrographs and fiber diameter histograms of SF8.3P0.3 group with different operation parameters: flow rate = 2 mL h⁻¹, voltage = 12 kV, distance = 13 cm (A1 and A2), flow rate = 1 mL h⁻¹, voltage = 12 kV, distance = 20 cm (B1 and B2), flow rate = 0.8 ml h⁻¹, voltage = 19 kV, distance = 23 cm (C1 and C2), SF7.2P0.3 (D1 and D2), SF7.2P0.3-Au groups (E1 and E2) obtained with the optimized operational parameters (flow rate = 0.8 mL h⁻¹, voltage = 19 kV and distance = 23 cm) and sericin extracted pure silk fibroin fibers (F1 and F2). Small figure on the top right corner of (E1) is the TEM image of AuNPs.

Chemical treatment of SF

Non-treated SF nanofibrous matrices are easily swollen and are readily soluble in water unless stabilized by treating with certain organic solvents. In order to prepare more water resistant SF, different chemical treatments were tried. It is known that SF is insoluble in water while β-sheet conformation prevails. Chemical treatment of SF matrix with different chemicals (EtOH, MetOH and GTA) showed that the fibers became more brittle after MetOH and GTA treatment and they were prone to mechanical breakage. Furthermore, SFM-M and SFM-G were brittle with fracturing tendency in DI water. It is known that MetOH-treated SF matrices typically have a high content of crystalline regions, and are often very brittle.²⁰ However, the integrity of the scaffolds was more robust when only EtOH was used. Among these groups, only SFM-E was extensible without any rapid breakage. As distinct from these reasons related to the solubility or brittleness issues, residues of organic solvents such as MetOH and GTA could pose toxicity problems when the SF fibers are exposed to cells. Although chemical treatment of SF fiber with MetOH^56,57 and GTA⁵⁸ was reported in the literature, the biocompatibility concerns were also noted. Hence, it was supposed that avoiding the use of such organic solvents could enhance the potential biocompatibility of the electrospun fibers. In addition, the physical effects of these chemical treatment methods were studied with SEM analysis and no clear morphology change was observed for any of the chemically treated groups (data not shown). A recent study also encourages the use of ethanol treatment in enhancement of mechanical and biological characteristics of SF nanofibrous scaffolds.⁵⁹ As a whole, it was decided to use only EtOH to induce a β-sheet conformation transition. The mechanism of conformation transition of SF with ethanol treatment was explained in a previous study.⁶⁰ Essentially, the conformation transition is a process of hydrogen bond rearrangement. When the electrospun matrix is immersed in ethanol, it swells and its free volume increases and there is enough space for the molecule chains to rearrange.

Chemical characterization

Fourier transform infrared spectroscopy analysis. Infrared spectroscopy, frequently used to investigate the conformational changes of silk fibroin^20,59 was applied to study the molecular conformation of SF matrices fabricated in this study. Random coil showed strong absorption bands at 1647 cm⁻¹ (amide I), 1541 cm⁻¹ (amide II) and 1235 cm⁻¹ (amide III), while the β-sheet showed absorption bands at 1624 cm⁻¹ (amide I), 1519 cm⁻¹ (amide II).²⁰ As shown in Fig. 3(A), the structure of all chemically treated SF matrices (SFM-E, SFM-M, SFM-G and SFM-AuE) and SF film obtained after drying in an oven at 60 °C was characterized by absorption bands at 1621–1634 cm⁻¹ (amide I), 1513–1515 cm⁻¹ (amide II), 1229–1230 cm⁻¹ (amide III), attributed to the β-sheet conformation. The characteristic β-sheet absorptions at wavenumbers; 1620 cm⁻¹, 1503 cm⁻¹ and 1221 cm⁻¹ (sericin extracted SF fibers) were present in all of the groups (SFM-E, SFM-M, SFM-G, SFM-AuE and SF film). However, the intensity of SF film was too low compared to other groups, showing weak β-sheet conformation. As a result, to create β-sheet conformation similar to the natural fibroin, EtOH chemical treatment was determined to be sufficient. Formation of β-sheet conformation was required to fabricate water stable silk fibroin scaffolds. In addition, the washing step of chemicals used for the treatment of scaffolds resulted in loss of typical absorption peaks of PEO occurring at 839 cm⁻¹, 944 cm⁻¹, and 959 cm⁻¹ (=CH₂ bending); 1090 cm⁻¹, 1240 cm⁻¹, and 1278 cm⁻¹ (O–C stretching); 1359 cm⁻¹ and 1466 cm⁻¹ (O–H bending); and 2859 cm⁻¹ (O–H stretching),⁶¹ thus, PEO was thought to be washed away from the SF matrices.

Fig. 3 FTIR-ATR spectra of untreated (SF film, sericin extracted SF fibers), and chemically treated (SFM-E, SFM-M, SFM-G, SFM-AuE) SF scaffolds (A), porosity and pore size distribution histograms of SFM-E (B) and SFM-AuE (C), hydrolytic (D) and enzymatic (E) degradation profiles of these matrices. *: significant difference between SFM-E and SFM-AuE in the same time point (n = 4 ± SD, p < 0.05). Statistical method: ANOVA, post hoc, multiple comparisons (Tukey HSD). Ethanol, methanol and GTA treated SF matrices and ethanol treated SF matrices incorporated with AuNPs were denoted with SFM-E, SFM-M, SFM-G and SFM-AuE, respectively.

Porosity and pore size distribution analysis. Porosity results of SF matrices (SFM-E and SFM-AuE) were similar (67–69%). The pore size distribution analysis showed that the scaffolds had a wide range of pore size distribution (0–200 μm), concentrating around 0–50 μm (Fig. 3(B) and (C)). A tissue engineering scaffold requires a porous structure with a high porosity and interconnected pores which allow cell migration, mass transport, growth, and new tissue formation in 3D.⁶² The porous nature of the SF matrices will likely to help in absorbing wound fluids, keeping the wound dry and also providing oxygen supply to the wound. The size range of L929 cells is 5–10 μm.⁶³ Thus, it might be inferred that average pore sizes greater than 10 μm are necessary to allow cell passage and migration. In addition, PEO used in this study to aid the electrospinning is readily soluble in water and might have acted like a porogen inside the SF matrices after washing step. The use of PEO seems to be a common strategy to create a porous environment suitable for cell penetration into the scaffold.^64,65 PEO extracted SF matrices could readily allow the migration of cells towards inside of the scaffold.⁶⁶

Hydrolytic and enzymatic degradation. Both SF matrices (SFM-E and SFM-AuE) had a very high resistance against hydrolytic and enzymatic degradation (HD and ED) within 2 weeks of incubation period (Fig. 3(D) and (E)). HD of SFM-E was barely detectable in 1 day incubation. Yet, after 1 week, HD of SFM-E increased slightly and was constant during the next 2 weeks (18%). On the other hand, HD of SFM-Au was only at 3% levels during 2 weeks (Fig. 3(D)). SFM-E and SFM-AuE had very low enzymatic degradation (16–27%) during 2 weeks of incubation period (Fig. 3(E)). In comparison, HD and ED of SFM-AuE was lower than SFM-E, indicating that AuNPs incorporation into matrices had a decreasing influence on both the hydrolytic or enzymatic degradation profiles. It was reported before that biodegradation rate of scaffolds made up of SF differed with respect to the type of enzyme used.⁶⁷ The strength of some enzymes on degradation of SF was classified as follows: protease XIV > collagenase IA > α-chymotrypsin. Thus, the low degradation profile of SF matrices could also be related to the strength of collagenase type I. In natural skin tissue, the concentration of the enzymes differ with respect to wound level or to the type of the wounded tissue; therefore, faster or slower rates of degradation could be encountered in in vivo conditions. It should also be taken into account that the results of enzymatic degradation are not only related to the type and concentration of the enzyme but also to the intrinsic properties of the fabricated material. The results reported in the present study suggest that ethanol treated SF fibers offer a stable structure highly resistant against biodegradation due to its robust β-sheet formation. While transmittance peaks indicative of characteristic β-sheet formation are not significantly different between SFM-E and SFM-AuE, the distribution and sizes of the crystals may vary, contributing to the different rates of degradation.⁶⁸ It was reported that in vivo complete degradation of SF scaffolds could range between 2 months and 1 year according to processing differences,⁶⁸ that encompassed most of the polymeric scaffolds in common use today such as collagen,⁶⁹ poly(lactic-co-glycolic acid)⁷⁰ and polycaprolactone.⁷¹ In native fiber form, the highly crystalline silk fibers (>50% beta sheet content) require months to years to fully degrade, due to a surface enzymatic hydrolysis reaction.^72,73 In a study with mice it was reported that three-dimensional porous silk scaffolds implanted intramuscularly showed degradation changing from weeks (aqueous processing, low content of beta sheet) to over one year (solvent processing, high content of beta sheet).⁶⁸

Mechanical tests. Tensile strength (TS), elongation at break (EAB) and elastic modulus (E) of scaffolds were within the range of 0.032–0.073 MPa, 22–301% and 0.0052–0.27 MPa, respectively (Table 2). TS of SF matrices increased significantly after AuNPs incorporation. Significant differences of TS after AuNPs incorporation were observed for both wet and dry testing. However the significant increase for E after AuNPs incorporation was measured only in wet testing conditions. AuNPs incorporation did not have a clear effect on EAB of matrices. In contradiction to the increasing effect of AuNPs incorporation on TS and E, performing the tests in wet conditions decreased TS and E, but increased EAB significantly for all groups. The same increase effect of AuNPs on strength of scaffold was also encountered in other studies such as chitosan–AuNPs nanocomposite⁷⁴ and poly(L-lactide)–AuNPs composite scaffolds.⁷⁵ Thus, it was concluded that AuNPs loading into scaffolds tended to make SF matrices more strong.

Table 2 Mechanical test results of SF matrices in wet or dry testing conditions

Groups	Tensile strength TS (MPa)	Elongation at break EAB (%)	Elastic modulus E (MPa)
Data are expressed as mean ± SD (n ≥ 6). Statistical analysis method: ANOVA, post hoc tests, multiple comparisons (Tukey HSD).a Significant difference from SFM-E^dry (p < 0.05).b Significant difference from SFM-E^wet (p < 0.05).c Significant difference from SFM-AuE^dry (p < 0.05).d Significant difference from SFM-AuE^wet (p < 0.05). SFM-E and SFM-AuE indicates ethanol treated silk fibroin matrices without and with AuNPs, respectively.
SFM-E^dry	0.032 ± 0.0091^c	21.58 ± 10.96^b^,^d	0.26 ± 0.19^b^,^d
SFM-E^wet	0.011 ± 0.0029^c^,^d	168.35 ± 29.16^a^,^c^,^d	0.0052 ± 0.0013^a^,^c^,^d
SFM-AuE^dry	0.073 ± 0.025^a^,^b^,^d	29.53 ± 4.84^b^,^d	0.27 ± 0.15^b^,^d
SFM-AuE^wet	0.041 ± 0.015^b^,^c	301.40 ± 81.26^a^,^b^,^c	0.013 ± 0.0047^a^,^b^,^c

---

In vitro biocompatibility tests

Cytotoxicity of scaffold extracts. The L929 cell viability in extract medium of SF matrices (SFM-E and SFM-AuE) was studied. As shown in Fig. 4(A), the cell viability values of both SF matrices exceeded 90% values, reaching to very high values close to control (only cell culture medium) group. Hence, both groups were defined as noncytotoxic according to protocols of ISO 10993-5 standards. The non-cytotoxicity of the SF matrices could be attributed to four aspects. The two of them, the size and concentration of AuNPs, have recently been confirmed in our previous study that AuNPs larger than 20 nm size and with a concentration below 20 ppm were nontoxic on HaCat keratinocytes and 3T3 fibroblasts for at least 1 week period.⁶¹ Also, AuNPs (1.5 nm) with different surface charge (positive, negative and neutral) demonstrated a dose-dependent toxicity on HaCat keratinocytes.⁷⁶ So, the size and loading of AuNPs into SF matrices were adjusted as 24 nm and 14.27 ppm. Small AuNPs can be endocytosed by cells and form internal aggregates, which result in cytotoxicity.⁷⁷ The increased cytotoxicity of the AuNPs is caused by aggregation; therefore, preventing the aggregation of the particles in the internal environment of the cells can greatly decrease the toxicity of the AuNPs. The other important parameters are the inherent biocompatibility of silk fibroin and the good aqueous stability of SF matrices.

Fig. 4 Cell viability results of test groups (A) and the SEM micrographs showing the attachment of L929 fibroblasts on SFM-E (B) and SFM-AuE (C) in 1 day incubation and the proliferation of these cells on SFM-E (D) and SFM-AuE (E) in 3 days incubation. Small figures are the false-colored SEM micrographs showing the scaffold fibers in red and the attached cells in green-blue colours. SFM-E and SFM-AuE indicates ethanol treated silk fibroin matrices without and with AuNPs, respectively.

Cell attachment and proliferation on scaffolds. Cell morphology, attachment, and interaction between L929 fibroblasts and SF fibers at 1st and 3rd days were all observed by SEM micrographs (Fig. 4(B–E)). The tests demonstrated that the cells could adhere to SF fibers gaining a globular morphology in one day and overcrowded the surface of SF fibers with their characteristic spindle-like elongated morphology in 3 days as a clear indication of biocompatibility of both SF matrices (SFM-E and SFM-AuE). Fig. 4(D and E) also provided an evidence of cell orientation on the surface or in deeper regions of matrices. The cells seemed to grow along the SF fibers in a layer by layer orientation extending towards the deeper regions. The surface wettability of biomaterials has a very strong influence on the attachment, proliferation, migration, and viability of many different cells.²³ Pure SF nanofibrous matrices are said to be ultra-hydrophilic due to their hydrophilic groups.²³ Here, the hydrophilicity of pure SF matrices (SFM-E) is thought be increased after AuNPs incorporation, because it was reported recently that loading of AuNPs into electrospun PVA nanofibers increased its hydrophilicity.⁴⁹ So, SFM-AuE showed better cell attachment and more spread morphology than SFM-E in 1 day incubation period, probably due to the differences in surface hydrophilicity (Fig. 4(B and C)). These hydrophilic surfaces generally displayed better affinity for cells, but the adhesion of cell attachment proteins are hindered on such surfaces that might cause to an adverse effect on cell proliferation in long term incubation periods. Therefore, in order to improve both the protein and cell attachment it was suggested to create combined hydrophilic and hydrophobic surfaces.⁷⁸ However, the increased adhesion of proteins on AuNPs surface was proved previously,³⁵ so it is expected that the adhesion of proteins on SFM-AuE will increase, thereby improving the proliferation of cells in long term incubation periods. It can be qualitatively evaluated from the number and morphology of cells in Fig. 4(D and E) that cell spreading and coverage on SFM-AuE surface was more compared to SFM-E after 3 days. These differences might be made clearer by increasing the AuNPs ratio incorporated into SF matrix or bonding the nanoparticles covalently on the matrix surface to create a more active AuNPs-cell interaction. The ratio of AuNPs incorporation into SF was deliberately kept at low levels (14.27 ppm in the electrospinning solution) to evade the cytotoxicity since it was reported previously in our study that citrate capped AuNPs had a dose-dependent toxic effect on cells.⁶¹ Using natural and biocompatible reducing/capping agents such as gelatin,⁴⁷ collagen,⁴⁸ silk fibroin⁴² and plant extracts⁴⁹ for the green synthesis of AuNPs might alleviate the cytotoxicity and so the incorporation extent of AuNPs into scaffolds could be increased more safely. There are also different studies showing the good biocompatibility (good attachment and proliferation of cells) of AuNPs incorporated nanofibers made up of poly L-lactide,⁷⁵ polyvinyl alcohol⁴⁹ and polymethylglutarimide nanofibers.⁷⁹

In vivo full-thickness skin wound healing tests

Histopathological results.
Re-epithelialization. According to histopathology analyses, 14 day after the surgery, very slight re-epithelialization (re-epithelialization score of 0–1) was observed for any of the groups and there was no statistical difference between any of the test groups (Fig. 5(A)). The desiccation of the wounds leaded to the formation of scab (Fig. 5(D2)), and as a consequence, the epithelialization process was slowed down. Deterioration of the epithelialization process would have been avoided provided that the wound was able to be kept moist. As shown in a previous study, the epithelialization was accelerated on moist wound surfaces,⁸⁰ since keratinocytes migrated more easily over a moist wound surface than underneath a scab.⁸¹ Yet, the start of the re-epithelialization of the study groups was confirmed as depicted in Fig. 5(B and C), signifying that there were no foreign material reaction to the SF matrices towards which epithelial cells had grown.

Fig. 5 Histopathological analysis of skin biopsies taken on postoperative 14th day (A) and representative light micrographs of sections taken from SFM-E (B1 and B2), SFM-AuE (C1 and C2) and UTSC (D1 and D2) stained with H&E (100× magnification). *: significant difference from UTSC (n = 10 ± SD, p < 0.05). Statistical method: ANOVA, post hoc, multiple comparisons (LSD). The arrows indicate re-epithelialization (keratinocytes: K), granulation (fibroblasts: F, collagen: C), neovascularization (vascularization: V), inflammatory cells (macrophages: M, lymphocytes: L), scab (S) and SF matrices (SFM-E and SFM-AuE). The extent of re-epithelialization, granulation, neovascularization and inflammation was shown with asterisk symbol (***: intense, **: moderate and *: low). SFM-E and SFM-AuE indicates ethanol treated silk fibroin matrices without and with AuNPs, respectively and UTSC is untreated skin control group.

Repair of the dermis. Histological evaluation revealed that the inflammatory phase of healing was still continuing on postoperative 14th day with vascular proliferation, eosinophils, plasma cells, neutrophils and giant cells in dermis. The prolonged medium to high degree inflammation (score of 2–3) of SF matrices could be due to their stable undegraded structure (Fig. 5(B2) and (C2)), but still the inflammatory reaction for them was slightly milder than UTSC (Fig. 5(A)). There were few lymphocytes, plasma cells and giant cells for SF matrices (Fig. 5(B) and (C)); on the contrary, there were many inflammatory cells and micro-abscess formations for UTSC (Fig. 5(D)). Wound healing process is known to start with an inflammatory response characterized by the infiltration of neutrophils and macrophages to the wound bed.⁸² After neutrophils act to destroy all the foreign particles and bacteria, they are phagocytosed by macrophages. The presence of macrophages is an indication of the fact that the inflammatory phase is about to end and the proliferative phase is beginning.⁸³ As seen in the histological sections (Fig. 5(B) and (C)), macrophages dominated the wound bed rather than neutrophils in SF matrices indicating that proliferative phase has started in these groups earlier than UTSC. On the superficial part of the dermis, a necrotic tissue (scab) was observed as a result of the damage and desiccation of the wound, especially in UTSC group (Fig. 5(D2)). Following the inflammatory phase of wound healing the proliferative phase starts. This phase is marked by the accumulation and proliferation of fibroblasts and the subsequent formation of granulation tissue.⁵⁴ Proliferation of fibroblasts and new endothelial cells with characteristic circular nuclei, which forms granulation tissue, seemed to be more in SF matrix groups than UTSC (Fig. 5(A)). It was realized that the inflammatory response components was started to be replaced with capillary and collagen fiber formation, especially for SF matrix groups (Fig. 5(B) and (C)). As a consequence, the highest score for granulation tissue was given for SF matrices, which were higher than UTSC (Fig. 5(A)). These findings indicated that the SF matrices developed in this study (SFM-E and SFM-AuE) could promote the wound healing process via medium to high level prolonged inflammation followed by increased neovascularization and granulation tissue formation. Again no significant differences were observed between SFM-E and SFM-AuE. As explained in the previous parts, these differences could be made clearer by increasing the AuNPs content in SF matrix either by passively embedding them inside the matrix or bonding them to the matrix surface to create a more bioactive surface. SF nanofibers were shown to have a notable effect on cell attachment and spreading of normal human keratinocytes and fibroblasts in vitro⁸⁴ and SF from Bombyx mori domestic silkworm exhibited as high attachment and growth of fibroblast cells as collagen did.⁸⁵ Furthermore, it was also reported that SF film treatment produced less inflammation and neutrophil–lymphocyte infiltration of the wounds than clinically used collagen based DuoActive dressing.⁸ The good biocompatibility and high stability of SF even in harsh in vivo conditions⁶⁸ make it an appropriate candidate as AuNP carrier since immobilization of AuNPs in a slow degrading 3D nanofibrous matrix will tend to decrease high or burst AuNP release rates. Thus, the potential cause of toxicity related with internalization of AuNP aggregates in cells⁷⁷ can be eliminated with such carriers. The size and dose dependent toxicity of citrate capped AuNPs on keratinocytes and fibroblasts was encountered in our previous study as well.⁶¹ However, the results of this study are still encouraging the use of SF matrices as efficiently robust structures for wound healing applications since they were still available under wound bed even after 14 days of surgery (Fig. 5(B2) and (C2)). Furthermore, degraded AuNPs immobilized SF matrix fragments are expected to have larger sizes which are difficult for the cells to internalize than citrate capped AuNPs only. Future trend for the synthesis of AuNPs for medical use is now switching from using potentially toxic reagents, such as sodium citrate, citrate capped AuNPs being vulnerable to aggregation even in mild electrolyte conditions, towards more biocompatible biomolecules rendering more stable and biofunctional AuNPs biocomplexes.^42,47–49
The examination of wound closure. Wound healing results were depicted representatively with photos taken on postoperative 14th day (Fig. 6(A–D)). Wound closure rates of all the test groups on postoperative 14th day were given in Fig. 6(E). As seen, the wound closure in the case of the wounds treated with SF matrices (SFM-E and SFM-AuE) was statistically more than the untreated group at 14th day (p < 0.05, ANOVA, post hoc, multiple comparisons LSD). These results indicated that all the fabricated SF matrices could accelerate the closure of full-thickness wounds in rats better than UTSC.

Fig. 6 Representative images showing the wound closure extent of SFM-E (A), SFM-AuE (B), UTSC (C) and the original 2 × 1 cm full-thickness wounds (D), the corresponding wound closure of these groups on postoperative 14th day (E), representative stress strain curves of test groups (F), figure showing the successful tensile breakage (G) and biomechanical test results of these groups on postoperative 14th day (H). #: significant difference from UTSC. *: significant difference from UWSC group (n ≥ 6 ± SD, p < 0.05). Statistical method: ANOVA, post hoc, multiple comparisons (LSD). RTS, REAB and RE indicate recovered biomechanical properties, i.e. tensile strength, elongation at break and elastic modulus of test groups normalized to their unwounded skin controls, respectively. SFM-E and SFM-AuE indicates ethanol treated silk fibroin matrices without and with AuNPs, respectively. Untreated and unwounded skin control groups were denoted with UTSC and UWSC, respectively.

Biomechanical tests. Representative stress–strain curves of skin samples from all groups were presented in Fig. 6(F). The tensile strength of an incisional wound is an assessment of the total wound strength. Here, the recovered percentages of these mechanical properties (RTS, REAB and RE) were used to understand to what degree the wounds returned to the unwounded state and to normalize the differences in the physical conditions of the rats. The tests were accepted successful when the rupture occurred in the middle of gauge length and the skin samples did not slip from the grip sites Fig. 6(G). So, unsuccessful tests were discarded from the results data. Biomechanical tests showed that EAB of all study groups (SFM-E and SFM-AuE) and UTSC recovered at least 100% of their unwounded state on postoperative 14th day (Fig. 6(H)). However, RTS and RE values were at the 40–60% levels, being significantly lower than UWSC. There were also no significant differences between any of the study groups and UTSC for these mechanical properties (Fig. 6(H)). The mechanical properties of the skin are dominated by the dermis in normal conditions.⁸⁶ Reduced mechanical properties of the test samples than UWSC might be related to the decrease of collagen synthesis or functional properties of collagen fibers synthesized by fibroblasts.^87,88 In our study, the granulation tissue scores in the wounded dermis of SFM-E and SFM-AuE groups was more than UTSC during the 14 days period, and an intense neovascularization was seen in both SF matrices during this period (Fig. 5(A)). Also, there were medium to high degree inflammation with SFM-E and SFM-AuE, still staying undegraded in the wound bed (Fig. 5(B2) and (C2)). Opposing to the TS increasing effect of fibroblasts, the density of inflammatory cells such as polymorphonuclear leukocytes and their long presence during wound healing might decelerate the dermal repair process and decrease wound TS, because they degrade proteins like collagen with enzymes.⁸⁹ Thus, the collagen synthesized by fibroblasts within the proliferative phase, could be lowered in case of a severe acute inflammatory response.⁸⁹ Therefore, it was supposed that the significant difference of RTS and RE of test groups than unwounded skin might be caused by the still ongoing medium to high inflammation levels leading to high unmatured extent of collagen fibrils. Also, lower tensile modulus values (softer skin samples) than normal skin observed for all the test samples could be related to this unmatured loose structure of the new collagen and fibrin fibers. All the test samples elongated as much as the normal skin, probably due to this extensible and loose structure. Since the biomechanical tests are mainly related to the healing extent of the dermis, the insignificant difference between results of SFM-E and SFM-AuE could be based on the previous explanations related to the incorporation ratio of AuNPs into these scaffolds.

Conclusions

The present study demonstrates a novel design of 3D SF nanofibrous matrices incorporated with AuNPs as a potential wound healing biomaterial. SF matrices with or without AuNPs were both found to be suitable for wound healing applications regarding their undisturbed, toxic reagent free, highly porous structure enabling cell attachment, proliferation and penetration, high stability against hydrolytic and enzymatic degradation, and good mechanical properties. AuNPs incorporation into SF matrices slowed down the degradation of SF matrices and enhanced the scaffolds mechanical properties. It was found that the healing process of SF matrices with or without AuNPs had similar tendencies initially mediated by medium to high degree inflammation followed by enhanced granulation tissue formation, neovascularization and wound closure compared to UTSC, indicating good in vivo biocompatibility and better healing success in 14 day treatment durations than UTSC. However, AuNPs inclusion into SF matrices did not have a notable effect on these healing stages of wound due to the low and passive loading of AuNPs inside SF matrix. Low loading extent of AuNPs (14.27 ppm) was deliberately chosen in this study since citrate capped AuNPs had a tendency to cause potential toxicity above 20 ppm levels, as observed in our previous study. As a consequence, future studies will focus on loading green-synthesized AuNPs on the surface of 3D SF matrices to obtain significant differences in both in vitro and in vivo test results. Although the wound healing effect of SF based scaffolds have been reported repeatedly, the knowledge about in vivo effect of AuNPs in combination with SF has not been presented before. These results will contribute to literature on the biomedical applications of AuNPs which have gained high interest recently. We may suggest that AuNPs hold a great promise in the field of skin wound healing applications due to their unique properties such as good biocompatibility, proven antibacterial and antioxidant effect, and easy incorporation of healing agents such as drugs and growth factors on the gold surface. Based on these findings, 3D SF nanofibrous matrices incorporated with AuNPs might be proposed as a clinically useful biomaterial for skin wound treatment in the future.

Acknowledgements

This investigation was financially supported by the Scientific and Technological Research Council of Turkey, TUBITAK (Project no. 2120189) in the framework of 1512- Entrepreneurship and Multistep R&D Funding Program. The commercialization potential of the SF matrices developed here was also investigated with the cooperation of METU and Remoderm Medical Ltd. Co. The authors express their gratitude to Tufan Emiroglu for his helps in construction of wet electrospinning system and Dr Temel Bilici for his valuable advices in the design of SF matrices. The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Notes and references

1. L. Meinel and D. L. Kaplan, Adv. Drug Delivery Rev., 2012, 64, 1111–1122.
2. S. W. Ha, A. E. Tonelli and S. M. Hudson, Biomacromolecules, 2005, 6, 1722–1731.
3. B. Zuo, L. Dai and Z. Wu, J. Mater. Sci., 2006, 41, 3357–3361.
4. C. Jiang, X. Wang, R. Gunawidjaja, Y.-H. Lin, M. K. Gupta, D. L. Kaplan, R. R. Naik and V. V. Tsukruk, Adv. Funct. Mater., 2007, 17, 2229–2237.
5. U. J. Kim, J. Park, H. J. Kim, M. Wada and D. L. Kaplan, Biomaterials, 2005, 26, 2775–2785.
6. N. Minoura, M. Tsukada and M. Nagura, Biomaterials, 1990, 11, 430–434.
7. J. H. Yeo, K. G. Lee, H. C. Kim, H. Y. L. Oh, A. J. Kim and S. Y. Kim, Biol. Pharm. Bull., 2000, 23, 1220–1223.
8. A. Sugihara, K. Sugiura, H. Morita, T. Ninagawa, K. Tubouchi, R. Tobe, M. Izumiya, T. Horio, N. G. Abraham and S. Ikehara, Proc. Soc. Exp. Biol. Med., 2000, 225, 58–64.
9. H.-J. Jin, J. Park, R. Valluzzi, P. Cebe and D. L. Kaplan, Biomacromolecules, 2004, 5, 711–717.
10. K. Hu, Q. Lv, F. Cui, Q. Feng, X. Kong, H. L. Wang, L. Y. Huang and T. Li, J. Bioact. Compat. Polym., 2006, 21, 23–37.
11. L. Huang, C. Li, W. Yuan and G. Shi, Nanoscale, 2013, 5, 3780–3786.
12. L. Meinel, V. Karageorgiou, R. Fajardo, B. Snyder, V. Shinde-Patil, L. Zichner, D. Kaplan, R. Langer and G. Vunjak-Novakovic, Ann. Biomed. Eng., 2004, 32, 112–122.
13. Y. Wang, D. J. Blasioli, H.-J. Kim, H. S. Kim and D. L. Kaplan, Biomaterials, 2006, 27, 4434–4442.
14. A. Motta, C. Migliaresi, F. Faccioni, P. Torricelli, M. Fini and R. Giardino, J. Biomater. Sci., Polym. Ed., 2004, 15, 851–864.
15. M. Fini, A. Motta, P. Torricelli, G. Giavaresi, N. Nicoli Aldini, M. Tschon, R. Giardino and C. Migliaresi, Biomaterials, 2005, 26, 3527–3536.
16. X. Zhang, M. R. Reagan and D. L. Kaplan, Adv. Drug Delivery Rev., 2009, 61, 988–1006.
17. S. Y. Park, C. S. Ki, Y. H. Park, H. M. Jung, K. M. Woo and H. J. Kim, Tissue Eng., Part A, 2010, 16, 1271–1279.
18. M. E. Kinahan, E. Filippidi, S. Koster, X. Hu, H. M. Evans, T. Pfohl, D. L. Kaplan and J. Wong, Biomacromolecules, 2011, 12, 1504–1511.
19. D. Atila, D. Keskin and A. Tezcaner, Carbohydr. Polym., 2015, 133, 251–261.
20. B.-M. Min, L. Jeong, K. Y. Lee and W. H. Park, Macromol. Biosci., 2006, 6, 285–292.
21. I.-S. Yeo, J.-E. Oh, L. Jeong, T. S. Lee, S. J. Lee, W. H. Park and B.-M. Min, Biomacromolecules, 2008, 9, 1106–1116.
22. C. R. Yoo, I.-S. Yeo, K. E. Park, J. H. Park, S. J. Lee, W. H. Park and B.-M. Min, Int. J. Biol. Macromol., 2008, 42, 324–334.
23. S. Shahverdi, M. Hajimiri, M. A. Esfandiari, B. Larijani, F. Atyabi, A. Rajabiani, A. R. Dehpour, A. A. Gharahaghaji and R. Dinarvand, Int. J. Pharmacol., 2014, 473, 345–355.
24. R. E. Unger, M. Wolf, K. Peters, A. Motta, C. Migliaresi and C. J. Kirkpatrick, Biomaterials, 2004, 25, 1069–1075.
25. Z. Karahaliloğlu, B. Ercan, E. B. Denkbaş and T. J. Webster, J. Biomed. Mater. Res., Part A, 2015, 103, 135–144.
26. A. Schneider, X. Y. Wang, D. L. Kaplan, J. A. Garlick and C. Egles, Acta Biomater., 2009, 5, 2570–2578.
27. P. Uttayarat, S. Jetawattana, P. Suwanmala, J. Eamsiri, T. Tangthong and S. Pongpat, Fibers Polym., 2012, 13, 999–1006.
28. E. S. Gil, B. Panilaitis, E. Bellas and D. L. Kaplan, Adv. Healthcare Mater., 2013, 2, 206–217.
29. S. A. Grant, C. S. Spradling, D. N. Grant, D. B. Fox, L. Jimenez, D. A. Grant and R. J. Rone, J. Biomed. Mater. Res., Part A, 2014, 102, 332–339.
30. C. R. Deeken, D. B. Fox, S. L. Bachman, B. J. Ramshaw and S. A. Grant, J. Biomed. Mater. Res., Part B, 2011, 97, 334–344.
31. Q. Cui, A. Yashchenok, L. Zhang, L. Li, A. Masic, G. Wienskol, H. Möhwald and M. Bargheer, ACS Appl. Mater. Interfaces, 2014, 6, 1999–2002.
32. H.-S. Hung, C.-H. Chang, C.-J. Chang, C.-M. Tang, W.-C. Kao, S.-Z. Lin, H.-H. Hsieh, M.-Y. Chu, W.-S. Sun and S.-H. Hsu, PLoS One, 2014, 9, e104019,  DOI:10.1371/journal.pone.0104019.
33. G. Obaid, I. Chambrier, M. J. Cook and D. A. Russell, Photochem. Photobiol. Sci., 2015, 14, 737–747.
34. C. Paviolo, J. W. M. Chon and A. H. A. Clayton, Small, 2015, 11, 1638–1643.
35. R. Gopalakrishnan, E. R. A. Singam, J. V. Sundar and V. Subramanian, Phys. Chem. Chem. Phys., 2015, 17, 5172–5186.
36. V. Sanna, N. Pala, G. Dessì, P. Manconi, A. Mariani, S. Dedola, M. Rassu, C. Crosio, C. Iaccarino and M. Sechi, Int. J. Nanomed., 2014, 9, 4935–4951.
37. Y. Zheng, M. Xiao, S. Jiang, F. Ding and J. Wang, Nanoscale, 2013, 5, 788–795.
38. A. Regiel-Futyra, M. Kus-Liśkiewicz, V. Sebastian, S. Irusta, M. Arruebo, G. Stochel and A. Kyzioł, ACS Appl. Mater. Interfaces, 2015, 7, 1087–1099.
39. H. M. A. Ehmann, D. Breitwieser, S. Winter, C. Gspan, G. Koraimann, U. Maver, M. Sega, S. Köstler, K. Stana-Kleinschek, S. Spirk and V. Ribitsch, Carbohydr. Polym., 2015, 117, 34–42.
40. T. Cohen-Karni, K. J. Jeong, J. H. Tsui, G. Reznor, M. Mustata, M. Wanunu, A. Graham, C. Marks, D. C. Bell, R. S. Langer and D. S. Kohane, Nano Lett., 2012, 12, 5403–5406.
41. S. Das, M. Sharma, D. Saharia, K. K. Sarma, M. G. Sarma, B. B. Borthakur and U. Bora, Biomaterials, 2015, 62, 66–75.
42. L. Jia, L. Guo, J. Zhu and Y. Ma, Mater. Sci. Eng., C, 2014, 43, 231–236.
43. S. Wang, W. Lu, O. Tovmachenko, U. S. Rai, H. Yu and P. C. Ray, Chem. Phys. Lett., 2008, 463, 145–149.
44. Y. Pan, S. Neuss, A. Leifert, M. Fischler, F. Wen, U. Simon, G. Schmid, W. Brandau and W. Jahnen-Dechent, Size-dependent cytotoxicity of gold nanoparticles, Small, 2007, 3, 1941–1949.
45. S. Lu, D. Xia, G. Huang, H. Jing, Y. Wang and H. Gu, Concentration effect of gold nanoparticles on proliferation of keratinocytes, Colloids Surf., B, 2010, 81, 406–411.
46. C. M. Goodman, C. D. McCusker, T. Yilmaz and V. M. Rotello, Toxicity of gold nanoparticles functionalized with cationic and anionic side chains, Bioconjugate Chem., 2004, 157, 897–900.
47. M. P. Neupane, S. J. Lee, I. S. Park, M. H. Lee, T. S. Bae, Y. Kuboki, M. Uo and F. Watari, J. Nanopart. Res., 2011, 13, 491–498.
48. J. Spadavecchia, E. Apchain, M. Alberic, E. Fontan and I. Reiche, Angew. Chem., Int. Ed., 2014, 53, 8363–8366.
49. R. Manjumeena, T. Elakkiya, D. Duraibabu, A. Feroze Ahamed, P. Kalaichelvan and R. Venkatesan, J. Biomater. Appl., 2015, 29, 1039–1055.
50. D.-H. Roh, S.-Y. Kang, J.-Y. Kim, Y.-B. Kwon, H. Y. Kweon, K.-G. Lee, Y.-H. Park, R.-M. Baek, C.-H. Heo, J. Choe and J.-H. Lee, J. Mater. Sci.: Mater. Med., 2006, 17, 547–552.
51. C. Akalin, S. Kuru, A. M. Barlas, K. Kismet, B. Kaptanoglu, A. Demir, H. M. Astarci, H. Ustun and E. Ertas, International Wound Journal, 2014, 11, 64–68.
52. C. Lin, C. Mao, J. Zhang, Y. Li and X. Chen, Biomed. Mater. (Bristol, U. K.), 2012, 7, 045017,  DOI:10.1088/1748-6041/7/4/045017.
53. K. S. Midwood, L. V. Williams and J. E. Schwarzbauer, Int. J. Biochem. Cell Biol., 2004, 36, 1031–1037.
54. A. J. Singer and R. A. F. Clark, N. Engl. J. Med., 1999, 341, 738–746.
55. J. Kimling, M. Maier, B. Okenve, V. Kotaidis, H. Ballot and A. Plech, J. Phys. Chem. B, 2006, 110, 15700–15707.
56. C. Li, C. Vepari, H.-J. Jin, H. J. Kim and D. L. Kaplan, Biomaterials, 2006, 27, 3115–3124.
57. A. J. Meinel, K. E. Kubow, E. Klotzsch, M. Garcia-Fuentes, M. L. Smith, V. Vogel, H. P. Merkle and L. Meinel, Biomaterials, 2009, 30, 3058–3067.
58. C. S. Ki, S. Y. Park, H. J. Kim, H.-M. Jung, K. M. Woo, J. W. Lee and Y. H. Park, Biotechnol. Lett., 2008, 30, 405–410.
59. M. Gholipourmalekabadi, M. Mozafari, M. Bandehpour, M. Sameni and H. Ghanbarian, Adv. Mater. Lett., 2015, 6, 391–394.
60. H. Cao, X. Chen, L. Huang and Z. Shao, Mater. Sci. Eng., C, 2009, 29, 2270–2274.
61. Ö. Aktürk and D. Keskin, Turk. J. Biol., 2015, 39 DOI:10.3906/biy-1502-49.
62. T. Muthukumar, P. Prabu, K. Ghosh and T. P. Sastry, Colloids Surf., B, 2014, 113, 207–212.
63. A. Higuchi and Y. Tsukamoto, J. Biomed. Mater. Res., Part A, 2004, 71, 470–479.
64. B. M. Baker, A. O. Gee, R. B. Metter, A. S. Nathan, R. L. Marklein, J. A. Burdick and R. L. Mauck, Biomaterials, 2008, 29, 2348–2358.
65. J. L. Lowery, N. Datta and G. C. Rutledge, Biomaterials, 2010, 31, 491–504.
66. H.-J. Jin, J. Chen, V. Karageorgiou, G. H. Altman and D. L. Kaplan, Biomaterials, 2004, 25, 1039–1047.
67. M. Li, M. Ogiso and N. Minoura, Biomaterials, 2003, 24, 357–365.
68. Y. Wang, D. D. Rudym, A. Walsh, L. Abrahamsen, H.-J. Kim, H. S. Kim, C. Kirker-Head and D. L. Kaplan, Biomaterials, 2008, 29, 3415–3428.
69. H. Ueda, L. Hong, M. Yamamoto, K. Shigeno, M. Inoue, T. Toba, M. Yoshitani, T. Nakamura, Y. Tabata and Y. Shimizu, Biomaterials, 2002, 23, 1003–1010.
70. M. Ramchandani, M. Pankaskie and D. Robinson, J. Controlled Release, 1997, 43, 161–173.
71. H. Sun, L. Mei, C. Song, X. Cui and P. Wang, Biomaterials, 2006, 27, 1735–1740.
72. G. H. Altman, F. Diaz, C. Jakuba, T. Calabro, R. L. Horan, J. Chen, H. Lu, J. Richmond and D. L. Kaplan, Biomaterials, 2003, 24, 401–416.
73. R. L. Horan, K. Antle, A. L. Collette, Y. Wang, J. Huang, J. E. Moreau, V. Volloch, D. L. Kaplan and G. H. Altman, Biomaterials, 2005, 26, 3385–3393.
74. Y.-L. Lin, J.-C. Jen, S.-H. Hsu and I.-M. Chiu, Surg. Neurol., 2008, 70, S9–S18.
75. K. D. McKeon-Fischer and J. W. Freeman, J. Tissue Eng. Regener. Med., 2011, 5, 560–568.
76. N. M. Schaeublin, L. K. Braydich-Stolle, A. M. Schrand, J. M. Miller, J. Hutchison, J. J. Schlager and S. M. Hussain, Nanoscale, 2011, 3, 410–420.
77. W. Cui, J. Li, Y. Zhang, H. Rong, W. Lu and L. Jiang, Nanomedicine, 2012, 8, 46–53.
78. K. Zhang, H. Wang, C. Huang, Y. Su, X. Mo and Y. Ikada, J. Biomed. Mater. Res., Part A, 2010, 93, 984–993.
79. D. Jung, I. Minami, S. Patel, J. Lee, B. Jiang, Q. Yuan, L. Li, S. Kobayashi, Y. Chen, K. B. Lee and N. Nakatsuji, J. Nanobiotechnol., 2012, 10, 23,  DOI:10.1186/1477-3155-10-23.
80. G. D. Winter, Nature, 1962, 193, 293–294.
81. G. D. Winter and J. T. Scales, Nature, 1963, 197, 91–92.
82. A. K. Tsirogianni, N. M. Moutsopoulos and H. M. Moutsopoulos, Injury, 2006, 37, S5–S12.
83. T. Stipcevic, A. Piljac and G. Piljac, Burns, 2006, 32, 24–34.
84. B.-M. Min, L. Jeong, Y. S. Nam, J.-M. Kim, J. Y. Kim and W. H. Park, Int. J. Biol. Macromol., 2004, 34, 223–230.
85. N. Minoura, S. Aiba, Y. Gotoh, M. Tsukada and Y. Imai, J. Biomed. Mater. Res., 1995, 29, 1215–1221.
86. F. H. Silver, L. M. Siperko and G. P. Seehra, Skin Res. Tech., 2003, 9, 3–23.
87. T. Quan, Y. Shao, T. He, J. J. Voorhees and G. Fisher, J. Invest. Dermatol., 2010, 130, 415–424.
88. G. J. Fisher, T. Quan, T. Purohit, Y. Shao, M. K. Cho, T. He, J. Varani, S. Kang and J. J. Vorohees, Am. J. Pathol., 2009, 174, 101–114.
89. T. Muehlberger, J. M. Moresi, H. Schwarze, G. Hristopoulos, F. Laenger and L. Wong, J. Am. Acad. Dermatol., 2005, 52, 583–588.

---

Footnotes

† Electronic supplementary information (ESI) available: The scoring system of epithelialization; the inflammation scoring system; the neovascularization scoring system; the granulation tissue formation scoring system. See DOI: 10.1039/c5ra24225h

‡ Current address: Department of Bioengineering, Kırıkkale University, Kırıkkale, Turkey. E-mail: E-mail: omerakturk@kku.edu.tr; Tel: +90-0318-3574242.

---