Multifunctional tri-layer wound dressing containing ZNO nanoparticles and IGF-1 as an efficient biomaterial for healing of full thickness skin injuries

Abstract

Mimicking the hierarchical structure of the skin is one of the most important strategies in skin tissue engineering. Monolayer wound dressings are usually not able to provide several functions at the same time and cannot meet all clinical needs. In order to maximize therapeutic efficiency, herein, we fabricated a Tri-layer wound dressing, where the middle layer was fabricated via 3D-printing and composed of alginate, tragacanth and zinc oxide nanoparticles (ZnO NPs). Both upper and bottom layers were constructed using electrospinning technique; the upper layer was made of hydrophobic polycaprolactone to mimic epidermis, while the bottom layer consisted of Soluplus® and insulin-like growth factor-1 (IGF-1) to promote cell behavior. Swelling, water vapor permeability and tensile properties of the dressings were evaluated and the Tri-layer dressing exhibited impressive antibacterial activity and cell stimulation following by the release of ZnO NPs and IGF-1. Additionally, the Tri-layer dressing led to faster healing of full-thickness wound in rat model compared to monolayer and Bilayer dressings. Overall, the evidence confirmed that the Tri-layer wound dressing is extremely effective for full-thickness wound healing.

Graphical abstract

1. Introduction

The wound is a general term that refers to any disruption in the uniformity of skin tissue. Wounds can result from surgery, cuts, burns, scratches, gunshot wounds, mechanical force, accidents, etc. Wound care affects approximately 5.7 million people in the United States alone and costs the government around 20 billion dollars [1]. The wound healing process begins immediately after the wound is created and consists of four sequential and simultaneous stages: hemostasis, inflammation, proliferation, and remodeling. These four stages occur in a coordinated and consecutive manner. Following the creation of a wound, regardless of its cause, the wound pH becomes similar to that of body fluids and provides a suitable environment for the growth and proliferation of microbes. Infection is one of the main reasons for delays in wound healing, which can lead to the development of chronic, non-healing wounds. Chronic diseases such as diabetes can lead to such wounds [2]. Other factors such as age, gender, hormonal profile, anxiety level, alcoholism, smoking, diet, and body mass can directly impact wound healing [3].

To provide a faster and more efficient wound healing, development of multifunctional biomaterials for controlling bacterial infection, reducing the inflammation and promoting vascular regeneration is crucial [[4], [5], [6]]. Multiple therapeutic methods have been designed to address various types of wounds. Therapeutic approaches that have been implemented so far include autografts, allografts, xenograft tissue, artificial skin, and wound dressings. Autografts require a painful secondary surgery for tissue harvesting [7], and both sites have a high potential for infection. In allografts and xenografts, the body's immune system is a primary concern, as it may reject the transplanted tissue, and the lack of antibacterial activity can lead to the formation of bacterial biofilms and microbial infiltration. Infection is the main reason of delay in wound healing, which may form a chronic non-healing wound [8]. Artificial and synthetic skin comes with issues such as body rejection, unpleasant odor, high cost, and limited materials [9]. It is a fact that in the case of all above-mentioned treatments, a secondary dressing is required, which clearly shows the undeniable need of wound dressing.

Various types of wound dressings, including hydrogels, sponges, films, and foams, have been used so far, most of which are single-layered. Since single-layered wound dressings do not meet all the needs of the wound, the use of multi-layered wound dressings, which simulate the complex structure of skin tissue and have more advantages, is preferred [10,11]. The structure and material of the wound dressing directly impact the wound healing process. Both fabrication and materials can directly affect healing process. Microporosity enhances cell-matrix contact and capillary development whereas macroporosity facilitates the transportation of nutrients and the elimination of cellular waste [12]. A variety of wound dressings, including hydrogels, sponges, films, and foams have been used so far. However, mostly above-mentioned biomaterials are used as a single layer dressing material. Since mono-layer dressings frequently fall short of meeting all therapeutic needs in treatment of full thickness wounds, it appears preferable to use multi-layered dressings, which better mimic the intricate hierarchical structure of skin tissue and offer a number of benefits all at once [13,14]. Hydrophilicity and ability to absorb exudates, ability to loading drugs, maintaining wound moisture, adequate mechanical properties, and mimicking the structure of ECM are the important features of an ideal wound dressing used for perfect wound repair [15,16].

The hydrophilic hetero-polysaccharide tragacanth (TG) has a 1,4 linked-d-galacturonic acid backbone with side chains of xylose, arabinose, galactose and fucose. Bassorin (water swellable) and water-soluble arabinogalactan (tragacanthin), two components of which are exuded by the plant Astragalus gummifer, are both present in TG [17]. It has been widely reported that TG has wide range of antibacterial activity and has a positive impact on wound healing [18,19]. Alginate (Alg) is also a natural polymer that is extracted from different sources of macroalgae (seaweeds). This biopolymer has found a special place in wound healing applications due to its excellent biocompatibility and ability to absorb wound exudates, which minimizes the risk of adverse allergic effects and bacterial infection [20]. Therefore, by blending Alg and TG, a biocompatible, bioactive, antibacterial and absorbent cross-linkable structure can be fabricated. Zinc oxide (ZnO) is considered as one of the metallic oxides that plays a significant part in the wound healing. Along with direct medicinal uses including anti-inflammatory and anti-bacterial properties, zinc also serves as a micronutrient, which is vital for the healing process [21]. Additionally, it serves as a co-factor for the encoding of around 10% of the body's proteins, which play a significant regulatory function in transcriptional control, DNA repair, and extracellular matrix (ECM) synthesis [22]. More importantly, zinc is involved in every stages of wound healing, including the homeostatic, inflammatory, proliferative, and remodeling. Zinc is very vital for wound healing since it can encourage re-epithelialization and vascularization [23]. Growth factors (GFs) play an integral role in wound healing process. The multifunctional GFs such as insulin-like growth factor-1 (IGF-1) are thought to be valuable candidates for wound treatment because these GFs accelerate wound healing through different mechanism. IGF-1 functions as a chemotactic agent for endothelial cells, promotes keratinocyte and fibroblast migration, proliferation and motility, and speeds up the healing [24].

The goal of this study was to fabricate a multifunctional, Tri-layer wound dressing that will heal the wound effectively and mimic the natural skin structure. The 3D printed middle-layer was composed of Alg, TG, and ZnO, which was sandwiched between two layer of electrospinning nanofibers. Upper layer was made of hydrophobic polycaprolactone (PCL) nanofibers aim to avoid bacterial and fluid penetration. The composition of bottom nanofibers were Soluplus® and IGF-1. Polyvinyl caprolactam-polyvinyl acetate-polyethylene glycol graft copolymer (PCL-PVAc-PEG) is a hydrophilic and non-cytotoxic drug carrier known as Soluplus® [25]. The bottom nanofibrous layer attempted to provide microporous structure and it was assumed that nanofibers could offer the best circumstances for cell attachment and proliferation by simulating the ECM. To the best of our knowledge, no previous research has been done on this biomimetic Tri-layer 3D printed-electrospun wound dressing carrying both antibacterial agent and growth factor. Morphological, physical, chemical, mechanical, and biological assessments, as well as animal research, were carefully carried out and studied in order to evaluate the performance of the as-prepared Tri-layer wound dressing.

2. Materials and methods

2.1. Materials

Soluplus® (Sol) was provided by BASF Company (Ludwigshafen, Germany). Sodium alginate (Alg-Na) was bought from Sinopharm Chemical Reagent Co. Ltd. (Shanghai, China), in which the viscosity of an aqueous solution of 10 g/l at 20 °C was > 0.02 Pa.s, M_w = 153,300 g/mol, and the ratio of mannuronic acid (M blocks) to guluronic acid (G blocks), i.e., M/G ratio = 0.8. Polycaprolactone (PCL, MW = 80,000) was purchased from Shanghai Macklin Biochemical Co., Ltd (Shanghai, China). Tragacanth (TG) gel was created by dissolving powdered TG gum from Astragalus gummifer (available in the Iranian market) in deionized water. ZnO nanoparticles (ZnO NPs, nanopowder, <100 nm particle size), insulin like growth factor (IGF-1), fetal bovine serum (FBS), bovine serum albumin (BSA) and phosphate buffer saline (PBS) were purchased from Sigma-Aldrich (United States) and all the chemical and reagents were utilized as received. The Pasteur Institute in Iran donated L929 mouse fibroblast cells and the bacterial strains of Escherichia coli (E. coli, ATCC 25,922) and Staphylococcus aureus (S. aureus, ATCC 25,923).

2.2. Fabrication of the tri-layer wound dressing

Middle layer: The intermediate layer was prepared with a 3D printer. Initially, Alg and a 10% (w/v) solution of TG were prepared and mixed in a ratio of 75:25 (w/w). This ratio was selected based on printability in pilot experiments and the structure of the 3D printed scaffold. The 3D printed scaffolds with Alg:TG solution ratios of 100:0, 50:50 and 25:75 (w/w) are shown in Fig. S1, which had irregular and messy structures. Then, 3% ZnO NPs were added to the solution and stirred for 12 h [26,27]. The mixture of Alg-TG-ZnO was 3D printed using an extrusion bioprinter (Abitin 3, Iran) with CAD printing method. A scaffold was prepared at a print speed of 6 mm/s with an inner nozzle diameter of 0.25 mm. The scaffold was fabricated in 6 alternating layers of 0 and 90° with a height of 0.25 mm. It was immediately placed in a 0.25 molar calcium chloride solution for ion cross-linking. After 24 h, the scaffold was removed from the solution, washed with distilled water and freeze-dried. In the same procedure three types of scaffolds including; Alg, Alg-TG and Alg-TG-ZnOwere printed for comparing their properties.

Upper layer: The top layer was obtained by electrospinning a 10% (w/v) solution of PCL in hexafluoroisopropanol (HFIP) [28]. In summary, 1 ml of the prepared solution was loaded into a syringe with a 0.6 mm needle and fibers were electrospun onto the surface of the Alg-TG-ZnO printed scaffold at a rate of 0.1 ml/h. The voltage was 22 kV and the distance between the collector plate and the needle was 15 cm. The fabricated 3D printed Alg-TG-ZnO containing upper layer of PCL nanofibers was named as Bilayer dressing.

The bottom layer: The bottom layer consisted of nanofibers placed on the lower surface of the Bilayer scaffold. For the solution preparation, 33% (w/v) of Sol was dissolved in distilled water at 50 °C for 12 h using a magnetic stirrer [29]. Then, 5 µg/ml IGF-1 was added to the Sol solution at room temperature and stirring was continued for another 1 h [30]. Finally, 1 ml Sol-IGF1 solution was electrospun (according to the conditions mentioned above) below the Alg-TG-ZnO scaffold. Scheme 1 presents the fabrication steps of Tri-layer dressing (g/h/cm³/mmHg).

Scheme 1.

Schematic representation of constructing a Tri-layer wound dressing.

2.3. Physical assessment

Before the surface of the samples were investigated with a scanning electron microscope (SEM, Quanta FEG 250, Spain), a tiny coating of gold was applied on samples top layer. The surface of Alg, Alg-TG, Alg-TG-ZnO NPs, the top surface (PCL nanofibers), and the top surface of Bilayer wound dressings, both the top and bottom surfaces of the Tri-layer wound dressings, Sol/IGF-1 nanofibers, and the cross-section surface were observed with SEM with an accelerating voltage of 15 kV. To determine the elemental structure distribution, an energy dispersive spectrometer (EDS) mapping equipment (Shimadzu, Japan) was connected to SEM, and an image of zinc distribution was obtained. The chemical fingerprint of the samples was taken with fourier transform infrared (FTIR) (FTIR-8400, Shimadzu Co. Japan) spectroscopy at a resolution of 4 cm⁻¹ and a wavelength of 4000–500 cm⁻¹. The hydrophilicity of Alg, Alg-TG, Alg-TG-ZnO, PCL nanofibers, and Sol/IGF-1 was calculated by measuring the contact angle. The contact angle value was determined using images obtained from a camera connected to the device and software (Wettability Pro Classic, version 2.0.0 from the Czech Republic). In this way, 4-µl drops of water were placed on three different points of each sample and after 5 s imaging was carried out, and the contact angle was reported as an average.

The scaffolds' tensile profile was determined using a universal testing machine (H5K-S, Houns field Test Equipment Ltd., England) in accordance with ASTM-D882 [9]. The samples were cut into 3 cm × 1 cm dimensions and placed vertically between the grips of the testing machine, and stress-strain curves were obtained by applying a force equivalent to 50 N and a pulling speed of 5 mm/min at a temperature of 25 °C and 40% humidity.

2.4. Swelling ratio, weight loss, water vapor permeability, and release of ZINC and IGF-1

The swelling ratio was evaluated for Alg, Alg-TG, Alg-TG-ZnO, Bilayer and Tri-layer wound dressings by immersing in 20 ml PBS solution with a pH of 7.4 ± 0.4 for 24 h at 37 °C. After this period, the sample was removed from the solution and held with forceps for 10 s to remove excess PBS. The swelling ratio was calculated using Eq. (1) [31]:

$$Swellingratio(\%)=\frac{{\mathbf{W}}_{\mathbf{w}}-{\mathbf{W}}_{\mathbf{d}}}{{\mathbf{W}}_{\mathbf{d}}}\times 100\%$$ (1)

Where, W_w and W_d in this equation stand for the sample's weight in the swollen condition and dry state, respectively.

The weight loss of Alg, Alg-TG, Alg-TG-ZnO, Bilayer and Tri-layer wound dressings was calculated in vitro to assess the degradation of the samples. In summary, 10 mm × 10 mm samples were immersed in PBS at 37 °C for 1, 3, 5 and 7 d, and at each time point, the sample was removed with forceps and held for 10 s to remove excess PBS, weighed at room temperature and the actual weight loss was calculated. The percentage of degradation was calculated using Eq. (2) [32]:

$$\begin{array}{c}\hfill \mathrm{Degradation}(\%)=\frac{{\mathrm{W}}_0-{\mathrm{W}}_{\mathrm{t}}}{{\mathrm{W}}_0}\times 100\%\end{array}$$ (2)

Where, W_t refers to the sample's final dry weight and W₀ is the same samples original dry weight before deterioration.

Water vapor permeability (WVP) is evaluated using the ASTM E 96 desiccant technique [33]. For this purpose, the samples were cut into circles and placed on the mouth of a beaker containing 50 ml water. Then the beaker was placed in a desiccator containing silica gel and tightly closed with parafilm strip and placed in an incubator at 30 °C with a relative humidity of 75%. After 24 h, the weight change was determined, and the WVP was calculated using the Eq. (3):

$$\text{WVP}\text{rate}=\frac{\mathrm{W}}{A\times \Delta \mathrm{p}\times d}$$ (3)

Where A is the size of the circle samples, which covered the beaker, d is the samples thickness, Δp is the vapor pressure difference (mmHg), W is the amount of water vapor penetrated through the samples (g/h), and the WVP rate was calculated and reported as g/h/cm³/mmHg.

The release profile of ZnO NPs from Alg-TG-ZnO, Bilayer and Tri-layer wound dressings at pH 7.4 was measured using an inductively coupled plasma-mass spectrometer (ICP-MS, Agilent 7500ce, Tokyo, Japan). For this purpose, samples of equal size were immersed in PBS solution and incubated for 1, 3, 5 and 7 d at 37 °C. At each stage, PBS was collected and fresh PBS was added, and the amount of zinc released in the collected PBS was measured. The release of IGF-1 from the Tri-layer wound dressing was measured using enzyme-linked immunosorbent assays (ELISA) kits (Pars Azmoon, Iran). In summary, 1 ml Roswell Park Memorial Institute (RPMI) culture medium without serum was added to 40 mg of each sample and incubated for 1, 3, 5 and 7 d at 37 °C, 5% CO₂ and 90% humidity. The culture medium was collected at each time point and evaluated using ELISA reader kits (Biotek-elx 808, USA).

2.5. Antibacterial activity

In order to investigate the antibacterial properties of the Tri-layer wound dressing and compare it with the Bilayer wound dressings of Alg, Alg-TG and Alg-TG-ZnO, a diffusion disc method was used. Inoculums of 1 × 10⁶ to 3 × 10⁶ CFU/ml E. coli and S. aureus species were cultured on agar medium. A piece of sterile material (10 mm × 10 mm) was then added to the agar medium and incubated at 37 °C to determine the zone of inhibition. After 24 h, the samples were removed from the incubator, pictures were taken, and the zone of inhibition was measured.

2.6. Cell viability, adhesion and migration

In vitro cell studies were performed according to ISO 10,993–5. In this regard, L929 mouse fibroblast cells were defrosted and transferred to RPMI culture medium containing 10% FBS and 1% penicillin/streptomycin mixture and incubated at 37 °C, 90% humidity and 5% CO₂. The culture medium was changed every day during the experiment. The samples (10 mm × 10 mm) were sterilized by immersion for 2 h in 70% ethanol and then exposure to UV radiation for 20 min for each side. Each sterile sample was placed in a 24-well culture plate with 1 × 10⁴ cells and 1 ml culture medium and incubated at 37 °C for 4 h to assess viability. 3 wells (without wound dressing) in each plate were considered as controls. MTT solution (500 µl, 0.5 mg/ml) was added to each well on Day 1 and 5 after removing the culture medium. Then the plate was incubated for further 4 h. Subsequently, the purple formazan crystals were dissolved in isopropanol, while the plate was placed on a shaker incubator for 15 min. The surface materials were transferred to a 96-well plate, and the concentration of the solution residue in isopropanol was measured using an ELISA reader at a wavelength of 570 nm. Considering that wells with more cells have higher absorbance compared to wells with fewer cells, the cell viability was determined using Eq. (4) [34]:

$$\text{Cell}\text{viability}(\%)=\frac{\text{Absorption}\text{of}\text{sample}}{\text{Absorption}\text{of}\text{the}\text{control}}\times 100\%$$ (4)

A 10 mm × 10 mm sterile piece of Alg, Alg-TG, Alg-TG-ZnO, Bilayer and Tri-layer wound dressing was placed in a 24-well plate to measure cell adhesion. Then, 2 × 10⁴ cells were added to the surface of each sample and were incubated for 4–5 h before being exposed to a specific volume of culture medium containing 10% FBS. After 2 d, the culture medium was removed, and the cells were fixed with 3.5% glutaraldehyde. A specific amount of glutaraldehyde was added to each sample and left at room temperature for 2 h before being removed as a fixative. Following two washes with deionized water, the samples were dehydrated with 50%, 60%, 70%, 80% and 96% alcohols and 15 min for each percentage. The samples were then dried in an oven at 40 °C for 48 h. The morphology of the wound dressings and their cell adhesion were examined through SEM imaging.

Cell migration was performed using a scratch test, as reported in previous articles [35,36]. L929 mouse fibroblast cells were seeded in a 12-well culture plate and incubated to achieve a single layer with a cell density of 90%. Then, a 10 ml pipette tip was used to scratch the monolayer cells and cell debris was collected. The scratched area was covered with a 10 mm × 10 mm sterilized sample. A non-treated group was also considered serving as the control. Cell migration and motility of L929 cells are inhibited by cell proliferation; therefore, 0.1% FBS-containing culture medium is added to solve this issue. Studies show that a culture medium containing 0.1% FBS can inhibit cell growth and guarantee that in vitro wound closure is purely due to cell migration [36]. After removing the samples and culture medium, cellular scratches were imaged using a microscope (IX53, Olympus). Each group consisted of 3 replicates.

To achieve a 3D model of cell migration stimulation by wound dressings, the Boyden chamber and Transwell migration model were utilized. After separating the upper and lower chambers with a polycarbonate filter (8 µm pore size), culture media was added to the lower chamber until the polycarbonate filter was just barely covered and in the upper chamber 1 × 10⁵ cells suspended in 50 ml media was added. The filter and the cells were removed after 24 h, and the samples were trypsinized and the cells were fixed of on the petri plates. After that, the cells were stained for 20 min with 1% crystal violet solution and observed under microscope. The density of the migrated cells was also measured and quantified by the ImageJ software from the scratch assay and crystal violet images.

2.7. In vivo wound healing model

Thirty-six male Wistar rats (4 months old with an average weight of 300 ± 84 g) were randomly divided into six groups, single-housed in cage and placed in the stress-free environment with a 12-h light/dark cycle and unlimited access to water and standard diet one week prior to surgery. The experiments were conducted under sterile and clean conditions in accordance with EU Directive 2010/63/EU ethical guidelines and under the supervision of the ethics committee of Isfahan University of Medical Sciences (ethics code #IR.MUI.AEC.1401.050). Rats were anesthetized with 10% ketamine (75 mg/kg) and xylazine (10 mg/kg), their dorsal area was cleaned with 70% alcohol, shaved, washed, and a wound with an area of 1 cm² was created on their dorsal area. Wounds were covered with Alg, Alg-TG, Alg-TG-ZnO, Bilayer and Tri-layer dressings, while a group without wound dressing was considered as the control group. The wound healing rate was monitored over time and imaged. The wound healing rate was calculated using the Eq. (5) [37]:

$$Healingrate(\%)=\frac{\text{are}{\mathrm{a}}_0-\text{are}{\mathrm{a}}_{\mathrm{n}}}{\text{are}{\mathrm{a}}_0}\times 100\%$$ (5)

where area₀ is the size of the wound on Day 0 and area_n is its size on Day n.

The ratio of re-epithelialization was measured using the equation:

$$\begin{array}{c}\hfill \mathrm{Re}-\mathrm{epithelialization}(\%)=\frac{\mathrm{B}-\mathrm{A}}{\mathrm{B}}\times 100\%\end{array}$$ (6)

In this equation, A and B are the distances between the re-epithelialization edges and the wound length, respectively.

Also, the of collagen density in the wound area was measured and compared with the normal dermis area at 100× magnification according to Eq.7 [38]:

$$\begin{array}{ccc}& & \text{Collagen}\text{density}(\%)\hfill \\ & & =\frac{\mathrm{Average}\mathrm{collagen}\mathrm{intensity}\mathrm{in}\mathrm{wounded}\mathrm{area}}{\mathrm{Average}\mathrm{collagen}\mathrm{intensity}\mathrm{of}\mathrm{normal}\mathrm{dermis}}\times 100\%\hfill \end{array}$$ (7)

The whole tissue surrounding the wound was collected after 14 d and immersed in 4% paraformaldehyde solution for histomorphological analysis. Masson's trichrome and hematoxylin-eosin (H&E) staining were used for histopathological studies. IL-6, TNF-α, CD-31, CD-34 and VEGF were also monitored and the results were evaluated using the CaseViewer software. For IL-6, TNF-α, CD31, CD34 and VEGF detection, freshly collected tissues were fixed in 4% paraformaldehyde for 8 h at room temperature, rinsed with tap water for 5 min, and sequentially dehydrated in 70%, 80%, 95% and 100% ethanol. The tissue was cleared in xylene and embedded in paraffin to create blocks. Sections (4 µm thick) were cut using a microtome (Leica RM2235, Leica Biosystems), floated on water at 40 °C, mounted onto IHC slides (FLEX, Agilent), and air-dried for 8 h. Slides were de-paraffinized in xylene and rehydrated through graded alcohols (100%, 95%, 70%, 50%). Endogenous peroxidase activity was blocked by incubating the sections in 3% hydrogen peroxide in methanol for 20–30 min at room temperature. Antigen retrieval was performed using citrate buffer (10 mM, pH 6.0) at 95–98 °C for 10–20 min, followed by cooling at room temperature for 20 min. For non-specific binding prevention, sections were incubated with blocking buffer (10% FBS) for 1 h at room temperature. The sections were then incubated overnight at 4 °C with the respective primary antibodies: anti-IL-6 (1:200, ab6672, Abcam), anti-TNF-α (1:200, ab6671, Abcam), anti-CD31 (1:200, ZA-0568, Abcam), anti-CD34 (1:200, ZA-0550, Abcam), and anti-VEGF (1:200, VG1, Thermo MS-1467). After washing in PBS or Tris–HCl buffered saline (TBS), the sections were incubated with biotinylated secondary antibodies (1:300, ab64256, Abcam or DAKO) for 1 h at room temperature. Following secondary antibody incubation, sections were incubated with Streptavidin-HRP conjugate (ab171537, Abcam) for 30 min in the dark. Color development was achieved using diaminobenzidine (DAB) chromogen (Dako, Glostrup, Denmark) for 5 min. The sections were counterstained with hematoxylin, dehydrated through graded alcohols, cleared in xylene, and mounted with coverslips using mounting medium (ab64320, Abcam). Slides were visualized under a Nikon Eclipse Ts2-FL or Olympus BX41 microscope equipped with a Logene LG1000 digital camera system for detection and analysis of IL-6, TNF-α, CD31, CD34, and VEGF expression. Also, IL-6 and TNF-ɑ positive points as well as CD-31, CD-34 and VEGF positive points were analyzed and quantified using Image J software.

2.8. Statistical analysis

The acquired data was analyzed using the ANOVA and paired Student's t-test methods by the GraphPad Prism 8 Software. The average value and standard deviation of the data with three or more replications were shown (*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, and ****P ≤ 0.0001, ns: not significant) for each test.

3. Results and discussion

3.1. Design and physical characterization

Tri-layer wound dressing with antibacterial properties and angiogenic properties was designed to have significant antibacterial activity and enhanced wound healing speed. Additionally, the top hydrophobic layer of PCL nanofibers was electrospun to prevent fluid and microbial penetration into the wound area and strengthen the wound's mechanical profile. The middle layer containing ZnO NPs provided the wound dressing with antibacterial properties with gradual release of ZnO. TG with anti-inflammatory properties in the middle layer was an effective substance in wound healing. Alg in this layer was a non-toxic and biocompatible material, which was cross-linked with cationic ions, and formed coordination bonds similar to an egg box model, and provided strength and enhanced mechanical properties [39]. The bottom layer contained IGF-1 and Soluplus produced by electrospinning to increase cell biodegradability, growth, proliferation, and cell migration.

Fig. 1A depicts the SEM image of 3D printed scaffold meshes of Alg, Alg-TG and Alg-TG-ZnO. All scaffolds exhibited a filament structure with macro-pores and typical multilayer structures of 3D printed fibers [40]. It can be clearly seen that all the 3D printed scaffolds had a smooth surface. By addition of TG the fiber thickness slightly decreased and macro-pores size increased. Furthermore, the contact angle values were 40.07°, 46.72° and 49.37° for Alg, Alg-TG and Alg-TG-ZnO, respectively. Although the hydrophilicity partially decreased following by addition of TG and TG-ZnO, the value of contact angels was still in hydrophilic range. The EDS mapping confirmed uniform distribution of ZnO NPs in the middle layer (Fig. 1B & 1C). The upper layer of Bilayer dressing material, showed a uniform coverage on the surface of 3D printed layer both on the fibers and in the macropores space (Fig. 1D). Cross-sectional images indicated proper integration of nanofibers on the 3D printed layer surface (Fig. 1E). The fine structures of PCL nanofibers and the typical appearance of nanofibers prepared by electrospinning were observed in SEM images without any droplet and aggregation (Fig. 1F).

Fig. 1.

Microstructure of fabricated monolayer, Bilayer and Tri-layer dressing materials and ZnO NPs distribution in middle layer. (A) The SEM structure and contact angle value of Alg, Alg-TG and Alg-TG-ZnO, (B) EDS mapping of Zn element distribution in Alg-TG-ZnO, (C) EDS spectra of Zn element in Alg-TG-ZnO, (D) Upper layer of the fabricated Bilayer dressing material, (E) The cross-section of Bilayered dressing material, (F) Nanofiber structure of the upper layer and its contact angle value, (G) Nanofiber structure of the bottom layer of Tri-layer dressing material and contact angle value, (H) Bottom surface of Tri-layer dressing material and (I) Cross-section of the Tri-layer dressing material.

The diameter of PCL nanofibers was measured to be 267.75 ± 73.88 nm and a contact angle value of 90.89° was obtained as expected. It has been widely reported that PCL fibers are hydrophobic in nature [41], and therefore, the top layer can prevent the penetration of fluids and microbes into the wound. Additionally, in previous researches, it was concluded that the nanofiber meshes of PCL on the surface of biomaterials create a non-adhesive property, eliminating the need for secondary wound dressings [35]. Cotton gauzes are commonly used as secondary wound dressings that adhere to the wound, disrupt the formation of new tissue, and delay the healing process. The bottom layer of the Tri-layer wound dressing structure showed uniform nanofibers with an approximate diameter of 94.03 ± 28.66 nm and a hydrophilic nature, which can effectively absorb excess wound exudate and body fluids (Fig. 1G). The bottom layer is also integrated with nanofibers and covered with large pores. The cross-sectional surface of the Tri-layer wound dressing showed three distinguishable layers, with the printed middle layer sandwiched with nanofiber structures (Fig. 1I).

The finger print of the single and composite materials is presented in Fig. 2. Peaks in the Alg spectrum at 1025, 1597 and 3200–3400 cm⁻¹ were ascribed, respectively, to the stretching vibration of C—O-C, COO⁻ and OH groups [35]. The symmetric stretching of CH₂ at 2935 cm⁻¹, the carbonyl stretching vibration of galacturonic acid and its ester at 1726 cm⁻¹, the carboxylate stretching vibration of d-galacturonic acid at 1600 cm⁻¹, and the antisymmetric C—O-C vibrations of glycosidic groups of polysaccharides at 1151 cm⁻¹ are the characteristic peaks of TG [41]. The FTIR spectra of the Alg-TG confirmed presence of the both materials with changes in intensity due to hydrogen bonds. Peaks around 1633 cm⁻¹ and 1726 cm⁻¹ for Sol were related to the carbonyl groups in the vinyl acetate and caprolactam segments, respectively [42]. In the spectrum of Sol-IGF1, the peak at 1685 cm⁻¹ represents amino N–H vibrations and amide C = O stretching vibrations in IGF1 [34]. In the spectrum of Alg-TG-Sol, the decrease in the intensity of the peak at 1633 cm⁻¹ can be due to the interaction of the Sol carbonyl groups with the hydroxyl groups of Alg and TG. The sharp bands of C = O, C—O and CO—O-CO were seen at 1722, 1175 cm⁻¹ and 1044 cm⁻¹, respectively, as well as the CH₂ peaks at 2945 and 2866 cm⁻¹ are an indicator for PCL. Between 1200 and 1150 cm⁻¹, the strong stretching bands were observed by the symmetric and asymmetric C—C(=O)-O vibrations [43]. The mentioned peaks of PCL available at 1722, 1175 and 1044 cm⁻¹, have been slightly shifted and their intensity has decreased in the spectrum of Alg-TG-ZnO-PCL, due to hydrogen bonds with Alg and TG. These results, in addition to the results obtained from SEM, confirm the proper entanglement of the three layers of dressing.

Fig. 2.

FTIR spectra of the Alg, TG, Alg-TG, Sol, Sol-IGF1, Alg-TG-Sol, PCL and Alg-TG-ZnO-PCL.

3.2. Swelling ratio, weight loss and WVP

The high ability to absorb wound exudate and fluids is a vital characteristic of wound dressings. If exudates are not well absorbed, not only the wound healing speed decreases, but the dressing easily separates and falls off [44]. All prepared materials, Alg, Alg-TG, Alg-TG-ZnO, Bilayer and Tri-layer dressings, had outstanding swelling characteristic. Alg, Alg-TG and Alg-TG-ZnO all had a swelling of over ∼175%. Swelling decreased with the addition of electrospun PCL and reached ∼165% (Fig. 3A). The hydrophobic nature of PCL may be the cause of this issue. Since the bottom layer is in close contact with the wound, it is crucial that it be able to absorb and sustain water. As predicted, the swelling ratio increased when the Sol/IGF-1 nanofiber was added, which exceeded 200% (Fig. 3A). However, there was no statistically significant difference in the swelling ratio of the dressings (P > 0.05).

Fig. 3.

(A) The swelling graph associated with wound dressing, (B) Weight loss of various dressings after immersion in PBS, (C) Water vapor permeability, (D) cumulative percentage in vitro release profile of Zn from Alg-TG-ZnO, Bilayer and Tri-layer dressings, (E) cumulative percentage in vitro release profile of IGF-1 from Tri-layer dressing immersed in PBS solution at 37 °C, (F) Stress-strain curve, (G) Visual image of agar plate and inhabitation zone of dressings, and (H) Quantitative analysis of inhabitation zone. Significant results were *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001 and ****P ≤ 0.0001 and ns: not significant. All tests were don in triplicates (n = 3).

In vitro biodegradability of Bilayer and Tri-layer wound dressings was obtained and compared with Alg, Alg-TG, and Alg-TG-ZnO. Alg, Alg-TG and Alg-TG-ZnO showed increasing degradation over 7 d (Fig. 3B). Addition of PCL nanofibers reduced biodegradability. However, the Tri-layer wound dressing showed a significant weight loss over 7 d with immersion in PBS. The behavior of dual-layer and triple-layer wound dressings with an additional hydrophilic layer (Sol nanofibers layer) in the triple-layer wound dressing is justified. This hydrophilic nanofiber layer with functional groups in the triple-layer wound dressing improves biodegradability (∼22.5% weight reduction in 7 d).

The WVP is a highly desirable property for wound dressings, and all prepared materials showed satisfactory amounts (Fig. 3C). The WVP had no significant (P > 0.05) difference when Alg, Alg-TG and Alg-TG-ZnO were used to cover the beaker, but indicated significant difference after addition of the upper layer in Bilayer dressing material (P ≤ 0.0001) and both bottom and upper layers in Tri-layer dressing material (P ≤ 0.0001).

3.3. Release profile of ZINC and IGF-1, and tensile results

The release of zinc from Alg-TG-ZnO, Bilayer, and triple-layer wound dressings was calculated (Fig. 3D(. The zinc release did not show a significant difference, but it was higher in Alg-TG-ZnO following by Bilayer and Tri-layer dressing materials. This order can be clarified due to trap of Zn in added layers, because Zn was released faster in Alg-TG-ZnO (monolayer), and slower in Bilayer and Tri-layer dressing respectively. The gradual release of IGF-1 in the Tri-layer wound dressing was confirmed over a period of 7 d. The results showed that over 50% of IGF-1 was released within 2 d and reached over 80% by 7d (Fig. 3E). The reason for the slow release of IGF-1 is hydrogen bonding with IGF-1 and other substances.

Mechanical properties are key characteristics that protect wounds against physical damage and provide a non-disruptive nature, essential for wound dressing. According to Table 1 and Fig. 3F, the Tri-layer wound dressing had the highest tensile strength, elongation at break, and Young's modulus of 0.44±0.07 MPa, 60% and 1.67±0.24 MPa, respectively. Tensile strength and Young's modulus were slightly reduced in the Bilayer wound dressing (0.41±0.11 MPa and 1.42±0.18 MPa). The monolayer Alg, Alg-TG and Alg-TG-ZnO had lower tensile strength and Young's modulus, confirming the effect of adding each layer on the tensile profiles.

Table 1.

Tensile profile of the as-prepared Alg, Alg-TG, Alg-TG-ZnO, Bilayer and Tri-layer dressing materials.

Sample	Tensile strength (MPa)	Youngs’ modulus (MPa)
Alg	0.23 ± 0.06	0.54 ± 0.11
Alg-TG	0.32 ± 0.09	0.96 ± 0.14
Alg-TG-ZnO	0.25 ± 0.05	1.16 ± 0.21*
Bilayer	0.41 ± 0.04*	1.42 ± 0.18**
Tri-layer	0.44 ± 0.07*	1.67 ± 0.24**

Significant results were *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, compared to Alg dressing.

3.4. Antibacterial activity

After injury, the pH of the wound approaches the pH of internal body fluids (∼7), which provides an environment conducive to microbial growth. Antibacterial activity is necessary to eliminate bacteria that are already present in the wound as well as bacteria entering from the environment [45]. The activity of antibacterial Alg, Alg-TG, Alg-TG-ZnO, Bilayer and Tri-layer wound dressings against Gram-negative (E. coli) and Gram-positive bacteria (S. aureus) was evaluated using the disc diffusion method. The image of agar culture and quality analysis are shown in Fig. 3G–3H. The printed Alg scaffold did not show any antibacterial activity, but after adding TG, the antibacterial activity against both E. coli and S. aureus increased. By adding ZnO, the antibacterial activity significantly increased compared to Alg-TG (P ≤ 0.0001). However, the antibacterial activity of Alg-TG-ZnO, Bilayer wound dressing, and Tri-layer wound dressing, with slight differences were almost equal. Several studies have shown that the antibacterial activity of ZnO NPs is effective against various species of bacteria. Therefore, ZnO can interact with bacterial surfaces and/or bacterial core and exerts its effects through specific bactericidal mechanisms. Interestingly, ZnO NPs have been shown in numerous studies to be non-toxic to human cells, making it highly efficient. Additionally, TG has been reported to be biodegradable, non-allergenic, non-toxic, and non-carcinogenic, with good antibacterial activity [46].

3.5. In vitro cell viability, adhesion and migration

The wound dressing materials were directly contacted with L929 fibroblast cells and an MTT test was performed to determine viability, cytotoxicity, and cell proliferation. The results clearly showed that none of the wound dressing materials were toxic during the evaluation period, and the viability, growth, and proliferation of all samples increased from Day 1 to 5. Based on the findings, the Tri-layer wound dressing exhibited significant cell viability compared to the control (P ≤ 0.0001). The Bilayer wound dressing and Alg-TG-ZnO did not show significant viability compared to the control sample (Fig. 4A). Additionally, cells showed higher biocompatibility on the surface of Al-TG (P ≤ 0.001) compared to Alg (P ≤ 0.001). Interestingly, this trend continued until the fifth day. Since the Tri-layer wound dressing performed better, the results confirmed the positive effect of IGF-1 on cell viability from day one to day five. In addition to the lack of IGF-1 release, the lower cell viability in Alg-TG-ZnO and Bilayer dressings compared to Tri-layer dressing can be due to the negative effect of more Zn release from these dressings, which has been mentioned in some previous studies [[47], [48], [49]].

Fig. 4.

Cytocompatibility of the Tri-layer dressing material in compare to Alg, Alg-TG, Alg-TG-ZnO, Bilayer, and control, (A) cell viability test (n = 3), (B) cell adhesion observation via SEM, (C) Cell migration assay in contact with the fabricated materials and (D) migrated cells density graph in scratch assay after 24 h (n = 3), (E)In vitro Transwell migration model stained with crystal violet, and (F) migrated cells density graph in Transwell migration model (n = 3), Significant results were *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001 and ns: not significant.

SEM micrographs show L929 mouse fibroblast cells attachment to the surface of Alg, Alg-TG, Alg-TG-ZnO, Bilayer and Tri-layer wound dressings (Fig. 4B). As observed, cells attached to the Alg surface with low density and were non-intensive, relatively same density of cells were present on the surface of Alg-TG, Alg-TG-ZnO and Bilayer, and mostly cells were adhered on the surface. Tri-layer dressing's surface was totally covered by L929 cells, confirming impressive cytocompatibility of the Tri-layer dressing in comparison to other samples.

Cell migration is a vital parameter in wound healing [35]. The scratch test was used to evaluate the effect of wound dressings on the migration of L929 cells. After 24 h, the results indicated homogeneous distribution of cells in the scratch covered with the Tri-layer wound dressing, confirming cells migration (Fig. 4C). As shown in Fig. 4D, the cell migration increased due to the presence of Alg wound dressing compared to the control (P < 0.05), also Alg-TG and Tri-layer wound dressings significantly increased cell migration in the scratch assay (P < 0.001 and P < 0.0001), while Alg-TG-ZnO and Bilayer dressings did not significantly stimulate cell migration compared to the control (P > 0.05). The in vitro results of the transwell migration model confirm the scratch test, showing the high potential of the Tri-layer wound dressing in improving cell migration (Fig. 4E). According to Fig. 4F, Alg and Alg-TG wound dressings caused a significant increase in Transwell cell migration (P < 0.01 and P < 0.0001), while the presence of ZnO NPs in Alg-TG-ZnO and Bi-layer wound dressings caused no significant change in cell migration compared to the control sample (P < 0.05). The interesting point is that the Tri-layer wound dressing with an additional nanofibrous bottom layer containing IGF-1 not only neutralized the negative effect of ZnO NPs on cell migration, but also caused a significant increase in Transwell cell migration compared to the control and other samples. In the literature, it is often mentioned that IGF-1 is a mitogen agent for fibroblast cells, which can be the main reason for the higher cell viability, proliferation, and cell adhesion to the surface of the Tri-layer wound dressing [50]. It is worth mentioning that the nanofibrous structure with a high specific surface area and high similarity to the natural ECM structure can enhance cell behavior [51].

3.6. In vivo wound healing and animal studies

Wounds in the back area of the rats were covered with Alg, Alg-TG, Alg-TG-ZnO, Bilayer and Tri-layer dressings, while uncovered wounds were considered as controls. The healing process was monitored for 14 d, with wound dressings being changed every other day. The wound healing rate of the mouse was investigated. The wound healing rate in the control group was approximately 44% over a period of 14 d (Fig. 5A and 5B). The group treated with the tri-layer wound dressing healed faster than the group treated with the Bilayer wound dressing (P ≤ 0.001), reaching over 80% improvement after one week and complete healing after 14 d, while the healing rate of the group treated with the Bilayer wound dressing was around 80% after 14 d. This comparison demonstrates the positive effect of IGF-1 on facilitating wound healing.

Fig. 5.

In vivo wound healing evaluation on rat model, (A) Visual image of the wound treated with different material and un-treated wound as control on Day 1, 7 and 14, (B) Quantitative analysis of the wound closure rate (%) on Day 7 and 14 (n = 6). Significant results were *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001 and ****P ≤ 0.0001, and ns: not significant.

3.7. Histomorphological and immunohistochemical evaluations

Histopathological sections related to the control group and treated group were examined after 14 d) Fig. 6. The presence of epidermis (Ep), dermis (D), hypodermis (HD), keratin layer (Kr), hair follicles (H.F), sweat glands (Sc.G), absence of inflammatory cells, and areas of neovascularization (N.V) (vessels with open lumen and clusters of endothelial cells) indicate the completion of the wound healing process. The regeneration of the epidermal layer was almost similar in the groups treated with Alg-TG, Alg-TG-ZnO, and the Bilayer group. These groups showed better healing compared to the groups treated with Alg and the control groups after 14 d. In addition to the improvement of the Ep layer in the Tri-layer group, H.Fs also grew towards the Ep layer. Other skin appendages such as Sc.G were also distinguishable in the group treated with the Tri-layer wound dressing. The results also indicate that inflammatory cells were more abundant in the wound treated with Alg, which decreased after the addition of TG and further decreased after mixing with Sol-IGF-1 (Fig. 6A). The re-epithelialization in different groups was also quantitatively compared (Fig. 6B), and it was determined that with the addition of TG, re-epithelialization significantly increased compared to the Alg and control groups (P < 0.001). The highest rate of re-epithelialization after 14 d was related to the group treated with Tri-layer dressing.

Fig. 6.

Histomorphological evaluation of wound treated with Alg, Alg-TG, Alg-TG-ZnO, Bilayer and Tri-layer dressing materials after 14 d, (A) H&E staining images, (B) ratio of re-epithelialization (n = 6), (C) Masson's trichrome staining images, (D) collagen density in comparison with normal dermis (n = 6). Significant results were **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001, and ns: not significant.

According to Fig. 6C, The Masson's trichrome staining of tissues on Day 14showed the presence of dense and thick collagen fibers in the wound treated with the Tri-layer wound dressing. Unlike the Tri-layer group, the presence of N.V areas in Alg-TG, Alg-TG-ZnO, and Bilayer wound dressing was similar to or even greater than the control group. Additionally, the presence of skin appendages such as Sc.G and H.F similar to Kr was observed in the wounds treated with the Tri-layer wound dressing, and H&E staining also yielded similar results. Quantitative amounts of collagen density in comparison with normal dermis were measured and the results are presented in Fig. 6D. As observed in the case of re-epithelialization, the collagen density in the groups treated with dressings containing TG also increased significantly compared to the Alg and control groups (P < 0.01) and the highest collagen density was observed in the group treated with Tri-layer dressing.

Immunohistochemical examination results after 14 d of treatment can be seen in Fig. 7. IL-6 and TNF-α markers were used to investigate inflammation, and CD-31, CD-34 and VEGF markers were used to investigate angiogenesis. As a result of damage to the skin, tissue macrophages and damaged keratinocytes secrete TNF-α and IL-6. Therefore, the presence of TNF-α and IL-6 indicate inflammation and these are among the most important pro-inflammatory cytokines [52]. On the other hand, CD31 (control of vessel density) [[53], [54], [55]], CD-34 (hematopoietic marker) [56], and VEGF (important signaling molecule in angiogenesis) [57,58] are among the most important markers of angiogenesis in the healing process. In essence, venules and capillaries, and small skin arteries are CD-31 and CD-34 positive points [49,59,60].

Fig. 7.

Immunohistochemical evaluation of wounds treated with Alg, Alg-TG, Alg-TG-ZnO, Bilayer and Tri-layer dressing materials after 14 d: (A) IL-6 stained tissue sections, (B) IL-6 positive points density graph (n = 6), (C) TNF-α stained tissue sections, (D) TNF-a positive points density graph (n = 6), (E) CD-31 stained tissue sections, (F) CD-31 positive points density graph (n = 6), (G) CD-34 stained tissue sections, (H) CD-34 positive points density graph (n = 6), (I) VEGF stained tissue sections and (J) VEGF positive points density graph (Some positive expression points of IL-6, TNF-a, CD-31, CD-34 and VEGF are indicated by yellow arrows on each image) (n = 6). Significant results were *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001, and ns: not significant.

According to Fig. 7A–7D, control group experienced a strong inflammatory reaction, which was somewhat reduced in the treated groups. It is worth mentioning that the addition of TG and ZnO NPs to the Alg wound dressing significantly reduced the expression of IL-6 and TNF-α. Also, the presence of a hydrophobic PCL nanofibrous layer on the Bilayer wound dressing, which prevented wound dehydration, caused a significant reduction in inflammation compared to Alg, Alg-TG and Alg-TG-ZnO wound dressings. Trilayer and Bilayer groups were significantly different only in IL-6 expression (P < 0.05), and TNF-ɑ expression was not significantly different for both groups (P > 0.05). These quantitative data were obtained based on image analysis using Image J Software, and the images show the statistical difference between the two groups in IL-6 staining can be detected (only one star*).

The closeness of the anti-inflammatory activity of the two groups treated with Trilayer and Bilayer wound dressings was mostly related to the presence of ZnO NPs and the presence of a hydrophobic PCL nanofibrous top layer, which prevented wound dehydration in both dressings.

The superiority of Trilayer wound dressing due to IGF1 release over Bilayer wound dressing was clear in promoting angiogenesis, accelerating epithelialization and ultimately the formation of skin appendages. Otherwise, both groups exhibited almost similar anti-inflammatory activity, with the slight superiority of Trilayer wound dressing.

Qualitative and quantitative results related to the evaluation of N.V areas in tissue sections are given in Fig. 7E–7J. N.V areas are stained as brown granules in the images, and these granules represent the positive CD31, CD-34, and VEGF points. The reduction of these points is evident in the treatment groups compared to the control group. The existence of N.V areas with high density indicates the initial stages of healing. The significant reduction of N.V areas in the tissue slices related to the Tri-layer group confirms the completion of the healing period after 14 d.

The tissue staining results revealed the positive effect of anti-inflammatory agents such as TG and ZnO NPs, and it was especially found that the release of IGF-1 as a mitogen growth factor from ECM-like nanofibrous bottom layer of Tri-layer dressing can remarkably accelerate the wound healing process. Also, the presence of a PCL hydrophobic nanofibrous layer on the Bilayer sample made this dressing provide a higher quality and healing rate than the Alg-TG-ZnO dressing, due to the prevention of dehydration. Based on in vivo studies, Tri-layer dressing is introduced as a perfect dressing for full-thickness wound repairing.

4. Conclusion

The Tri-layer wound dressing was prepared using hydrophobic PCL nanofibers in the top layer, Alg-TG and ZnO NPs fibers obtained through 3D printing in the middle layer, and nanofibers prepared from Sol and IGF-1 in the bottom layer. Its performance was evaluated in vitro and in vivo, comparing the rate of wound healing with Bilayer wound dressings and single layers of Alg, Alg-TG and Alg-TG-ZnO. The results showed that the hydrophobic layer of PCL nanofibers provided a barrier to avoid fluid infiltration into the wound. IGF-1 loaded in the wound dressing resulted in faster wound healing compared to the Bilayer wound dressing. Additionally, compared to other evaluated wound dressings, the Tri-layer wound dressing demonstrated the best performance in reduction of inflammatory cytokines, enhanced N.V, and wound closure in vivo within 14 d. Despite the promising results obtained in this study, animal studies on infected wounds and additional pre-clinical studies are necessary to ensure the proper performance of the prepared wound dressing.

Conflict of interest

Authors declare that they have no conflict of interest. This article contains animal subjects and the experiments were conducted in accordance with EU Directive 2010/63/EU ethical guidelines for animal experiments and under the supervision of the Ethics Committee of Isfahan University of Medical Sciences (ethical code number # IR.MUI.AEC.1401.050).