Protein-Engineered Large Area Adipose-derived Stem Cell Sheets for Wound Healing

Human adipose-derived stem cells (hADSCs) formed robust cell sheets by engineering the cells with soluble cell adhesive molecules (CAMs), which enabled unique approaches to harvest large area hADSC sheets. As a soluble CAM, fibronectin (FN) (100 pg/ml) enhanced the cell proliferation rate and control both cell-to-cell and cell-to-substrate interactions. Through this engineering of FN, a transferrable hADSC sheet was obtained as a free-stranding sheet (122.6 mm2) by a photothermal method. During the harvesting of hADSC sheets by the photothermal method, a collagen layer in-between cells and conductive polymer film (CP) was dissociated, to protect cells from direct exposure to a near infrared (NIR) source. The hADSC sheets were applied to chronic wound of genetically diabetic db/db mice in vivo, to accelerate 30% faster wound closure with a high closure effect (εwc) than that of control groups. These results indicated that the engineering of CAM and collagens allow hADSC sheet harvesting, which could be extended to engineer various stem cell sheets for efficient therapies.


Engineering of soluble FN for the formation of robust hADSC cell interactions and harvesting of hADSC sheets.
We have previously shown that collagen-coated PEDOT substrates are effective for the preparation of a human dermal fibroblast sheet 25 . Contrary to the human dermal fibroblasts cells, which were adhered and were homogeneously distributed on the collagens coated PEDOT surfaces, the hADSCs were not evenly adhered onto the CPP-PEDOT surfaces and left a partially aggregated mass when the hADSCs were seeded at 4 × 10 5 cells/dish. Thus, the cultured hADSCs on CPP-PEDOT for one day (Fig. 1h) afforded unstable and fragmented sheet pieces (Fig. 1i) by the photothermal detachment method. The hADSCs that were cultured for a longer period (3 days) were not detached from the CPP-PEDOT surface. This could be attributed to poor cell-to-cell (CC) and cell-to-substrate (CS) interactions, which are known to be extremely sensitive to the microenvironment of a so-called 'niche' 30,31 . To improve the C-C and C-S interactions on the CPP-PEDOT surface and to create a transferrable cell sheet, hADSCs were seeded in cell culture media containing fibronectin molecules (FNs), which are known as cell adhesion molecules (CAMs). When the hADSCs was seeded for one day at a cell number of 100,000 cells/dish, cells were well proliferated and formed a confluent layer. The cell proliferation rate (k pr ) in the presence of FNs was 184,255 cells/dish, which is higher than the rate without FNs (k pr = 99,475 cells/dish) (Fig. 1d,e).
Harvesting of hADSC sheets. To harvest the cell sheets, CPP-PEDOT substrates covered with confluent hADSCs were irradiated from the bottom face of TCPS with a coherent NIR laser (λ = 808 nm, I pw = 2.7 W cm −2 ). The temperature at the CPP-PEDOT, increased up to 43 °C which is above the temperature of dissociation of collagen 25,26 leading cell sheet detachment as shown in Fig. 1g. Since most of the light is absorbed by PEDOT (transmittance of 808 nm light <5%, Fig. 1c) and collagen layer is in-between cells and PEDOT film, cells are protected from direct exposure to a NIR source. As the collagens of triple helix structure in the film become unfolded into single strands at the light exposed area (43 °C), they are dissolved into solution media and the CS interactions are lost 25,26 . On the contrary, the CC interactions would stay intact to maintain as a sheet form during the harvesting Scientific  of sheets due to CAM (Fig. 1g). As observed before, the temperature of the cell medium at the same condition was 37~38 °C that could not damage cells within short NIR irradiation time (<10 min). For this reason, the harvested cell sheets showed a high viability after harvesting by photothermal method 25,26 . To find an optimized FN concentration for cell interactions and harvesting of a sheet, the area of the PEDOT and the cell concentration were fixed as 132.7 mm 2 and 4 × 10 5 /dish, respectively. When the FN concentration was extremely high (10 −4 to 10 −5 %; 1μg~100 ng/ml), cell sheets were not detached, even after 5 min of NIR exposure. This is probably due to the strong adhesion of the hADSCs on the CPP-PEDOT surface. With a dilute FN concentration (10 −6 to 10 −9 %; 10 ng/ml ~10 pg/ml), the hADSC sheets were detached from the border of the CPP-PEDOT and floated onto the media after 3 min of NIR exposure. The sheet was completely detached with a harvesting efficiency (ε d ) of 100% (Table 1) at a FN concentration from 10 ng/ml to 10 pg/ml: where A det and A nir are the areas of the detached cells and NIR exposure, respectively. A nir was the same as the area of CPP-PEDOT (132.7 mm 2 ) because the temperature rise mainly occurred at the PEDOT film 26 .
Although the ε d values were 100%, the harvested sheet area (A hcs ) was always smaller than the A det and A nir . This could be attributed to 1) the fact that the cells were not in a confluent state or had weak cellular interactions, resulting in fragmented sheets, and/or 2) the shrinkage of the cell sheet, which is often observed in cell sheet harvesting. Importantly, the harvested sheet area was strongly dependent on the FN concentration at the same cell concentration (4 × 10 5 cells/dish), as shown in Fig. 1g and Table 1. When the FN concentration decreased from 1 μg/ml to 100 pg/ml, A hcs increased and was maximized at the FN concentration of 100 pg/ml, after which A hcs decreased significantly. The maximum A hcs at 100 pg/ml reached 26.8 mm 2 at a cell concentration of 4 × 10 5 cells/dish. This result indicates that the soluble CAMs (FN) could enhance the cell proliferation rate and bring the cells in close proximity to one another and to the substrate (CPP-PEDOT) (Fig. 1g). It has been reported that the displacement of cell-cell (CC) interactions, enhancement of cell proliferation, and migration of cells could occur upon contact of the intercellular junctions with CAMs (e.g., FN, collagen, and laminin) 40 . From the DAPI (4′,6-diamidino-2-phenylindole) staining experiments, the cell proliferation number was counted at different cell concentrations under a fixed FN concentration (100 pg/ml). It was observed that the cells were concentrated and resided in proximity to one another as the cell number increased to 3.7 × 10 5 cells/dish ( Fig. 2a-c).
As shown in Fig. 1e, the CC interactions in the proliferated cells in the media containing FN went through re-formation of the CC interactions (Figs 1e and 2). E-cadherin expression was observed by fluorescence microscopy in the cell media with more than 2.7 × 10 5 cells/dish (Fig. 2b,e). As the cells were further concentrated, the cell junctions increased and were vividly observed at a cell number of 3.7 × 10 5 cells/dish (Fig. 2c,f). Thus, it was important to have a cell number larger than 3.7 × 10 5 cells/dish to reach a concentrated state with a large number of intercellular junctions.
To investigate the optimum cell concentration for the formation of a full hADSC sheet, the hADSCs were cultured on the CPP-PEDOT substrate with different cell concentrations (4, 8, 12 × 10 5 cells/dish) ( Fig. 3 and Table S1). After the cells were cultured for 1 day, the cell sheets were obtained by the photothermal method under NIR exposure (I pw : 2.7 W/cm 2 , t nir : 10 min). Interestingly, the size (A hcs ) of the harvested sheets increased as the seeded cell number increased (Fig. 3a) at the same A nir (area = 132.7 mm 2 , d = 1.3 cm). Figure 2g-i show the DAPI staining results for the harvested cell sheets. The average nucleus distance of the neighboring cells (d nn ) was identified by DAPI. The d nn value slightly decreased in the harvested cell sheet as the cell number increased; however, it was not proportional to the seeded cell number. The intercellular distances in harvested cell sheet from the DAPI data, were calculated and exhibited with different cell seeding numbers (4 × 10 5 , 8 × 10 5 , 1.2 × 10 6 cells/ dish). The cell groups are randomly selected cell groups in the same area of the cell sheet. Each cell groups are consisting of two cells to measure intercellular distance between the two cells. The average intercellular distances of cells (groups) were displayed in Fig. 2j-l. As the distance at 4 × 10 5 cells/dish was approximately 13.7 μm (Fig. 2j), an expected d nn in the cell sheet harvested from 8 × 10 5 cells/dish would be 6.9 μm assuming a proportional shrinkage of the cell sheet to cell number; however, the d nn of the 8 × 10 5 cells/dish concentration was determined as 13.1 μm by DAPI (Fig. 2k). The distance only slightly decreased with an increase in concentration, even under a high cell seeding condition such as 1.2 × 10 6 cells/dish (d nn = 12.6 μm). Instead of shrinkage, the harvested cell sheet area dramatically increased as the cell number increased. Thus, at an optimum FN concentration and cell seeding number, sheet shrinkage could be minimized, providing a large cell sheet. Finally, at the seeded cell number of 12 × 10 5 cells/dish, a full hADSC sheet was harvested by the photothermal dissociation of collagens on the CPP-PEDOT. Figure 3 shows the entire process of hADSC sheet harvesting. When the hADSCs were exposed to a NIR laser (λ: 808 nm, I pw : 2.7 W/cm 2 ) for 2 min, the cells were partially detached and floated from the CPP-PEDOT substrate (Fig. 3b). After NIR exposure for 5 min, the cell sheet was almost detached and floated from the substrate (Fig. 3c,d). The harvested sheet was free-standing and was transferrable using tweezers (Fig. 3e). The sheet was highly dense, and the size of the harvested sheet was approximately 72.9 mm 2 (Fig. 3f). E-cadherin expression (red) was observed in the cell sheet, and the cell viability was approximately 90%.
To obtain large-area cell sheets for in vivo chronic wound-healing experiments, the area of the CPP-PEDOT substrate was increased to 471.5 mm 2 and hADSCs (1.4 × 10 6 cells/dish) were seeded on the large CPP-PEDOT substrate with an optimized concentration of FN (100 pg/ml). After culturing the cells for 1 day, a large cell sheet was detached (Fig. 3h) and floated on the surface of the media (Fig. 3i) from the photothermal method using a NIR laser (I pw : 2l7 W/cm 2 , t nir : 2 min and 5 min, respectively). The ε eff.d of the detached cell sheets in each condition was also 100%, and the detached area of the hADSC sheet was 122.6 mm 2 , which was a suitable area for chronic wound-healing applications (Table 1). cell seeding numbers were (a,d) 1.8 × 10 5 cells/dish, (b,e) 2.7 × 10 5 cells/dish, and (c,f) 3.7 × 10 5 cells/dish. DAPI-stained images of the harvested sheets with different cell seeding numbers: 400,000 cells/dish (g), 800,000 cells/dish (h), and 1,200,000 cells/dish (i). Graphs of the calculated intercellular distance from the DAPI data at 400,000 cells/dish (j), 800,000 cells/dish (k), and 1,200,000 cells/dish (l). The cell groups are randomly selected cell groups in the same area of the cell sheet. Each cell groups in x-axis are consisting of two cells to measure intercellular distance between the two cells(y-axis).
Chemical analysis of the harvested cell sheet and media. Before the wound-healing application, the viability of the harvested cell sheet was further examined to identify any remaining toxic impurities, including (1) collagens from the CPP-PEDOT and (2) chemicals, such as iron and the monomers used for the preparation of PEDOT. To identify the remaining collagen in the harvested hADSC sheet, a sheet was harvested from the fluorescein isothiocyanate (FITC)-stained collagen layer that was coated on the PEDOT surface. Before NIR exposure, the FITC-stained collagen was detected with green fluorescence (Fig. 3k,l). Upon exposure to the NIR light source, the fluorescence intensities between the cell sheet and PP-PEDOT decreased within 2 min (Fig. 3l). This result is attributed to the photothermal dissolution of the collagen layer, in which collagens of insloluble triple helix structure were unfolded into soluble single strands upon phothermal heating, then dissolved out into ECM media. Before NIR irradiation, the collagen layer was not dissolved into the culture medium and the cell sheet was not floated from the CPP-PEDOT, as shown in Fig. 3k,l. After NIR irradiation, the collagen dissociation was started by the photothermally generated heat from the PP-PEDOT face. As NIR irradiation time goes by the distance between cell sheet to PP-PEDOT increased to 5.7 μm (60 sec), 9.8 μm (90 sec), and 14.7 μm (120 sec) (Figs 3k,l and S3). Finally, the fluorescence from FITC was almost undetectable in the harvested cell sheet (Figs 3k,l and Movie S1), indicating that the sheet is unlikely to transfer collagen from CPP-PEDOT.
To trace the iron ion (Fe 3+ ) from the oxidant, dipped CPP-PEDOT was analyzed by an inductive coupled plasma mass spectrometer (ICP-MS). The PEDOT-coated substrate, which had gone through the washing step 4 times (dipped in fresh ethanol for 2 h and washed out), showed only a trace amount of Fe components. This metal content of Fe quantity (18 ng mL −1 ) from the 4-times repeated washing step was much lower than the content in the cell medium (284 ng mL −1 ) (Table S3). On the other hand, the Fe 3+ quantity in the solution of the PEDOT that was dipped for 1 and 2 h followed by washing with ethanol was higher (1 h: 4837 ng mL −1 , 2 h: 6328 ng mL −1 ) compared with that of the cell medium. This result confirmed that the PP-PEDOT substrate purified by the washing step may be suitable as a cell culture media and does not have the problem of residual Fe 3+ ions. Figure S4 shows Fourier transform infrared spectrometer (FT-IR) of the fully washed PP-PEDOT (blue line). The peak at 3115 cm −1 for EDOT monomer (Fig. S4, black) was due to the 2,5-hydrogen atoms on the thiophene ring 42 . The peak at 3115 cm −1 disappeared in the fully washed PP-PEDOT (blue line). Furthermore, the peaks from oxidants due to S-O stretching peaks (810 and 1006 cm −1 ) were not observed on the spectrum of PP-PEDOT after it was fully washed 43 . Therefore, for the fully washed PEDOT substrate, the FT-IR peaks of the monomer and oxidant were not observed, confirming that the ethanol washing process was highly suitable for PP-PEDOT without toxicity to cells.
Furthermore, to examine the stability of PP-PEDOT under NIR exposure during cell sheet harvesting, UV/ Vis/NIR spectroscopy was used to determine any PEDOT residue in the harvested cell media. As shown in Fig. S5, there is no absorption at NIR range (700~900 nm) in the spectrum of the harvested cells in PBS obtained from the traditional trypsinization method (Fig. S5, PBS + cell (TCPS), blue line), because it was prepared without PP-PEDOT. Interestingly, the spectrum of the photothermally harvested cell sheet using CPP-PEDOT (PBS + cell sheet (PP-PEDOT), magenta) showed almost the same absorption spectrum as the sample obtained from the traditional trypsinization method. In addition, NIR absorption was not observed in the spectrum, even though it was prepared by PP-PEDOT. This result was strong evidence for using PP-PEDOT, as it was highly stable with NIR laser exposure for photothermal cell sheet harvesting. PP-PEDOT was not dissolved or decomposed in the medium during or after NIR exposure. Furthermore, PP-PEDOT did not remain in the harvested cell sheet. As a result, CPP-PEDOT and the photothermal cell sheet harvesting method are non-toxic, highly stable, and suitable for wound-healing applications.
hADSC sheets for chronic wound healing. The potential use of the hADSC sheet for chronic wound healing was examined in vivo using genetically diabetic db/db mice. Firstly, we carried out the wound healing effect by the cells suspension injection group (cells alone group) and cell sheet treated group for normal mice prior to the db/db mice. The cell sheets treated group showed higher wound healing effect than the cells suspension injection group (Supplementary information, Fig. S6). This is in good agreement with the previous reports that cell sheets are quite effective for the chronic wound healing [44][45][46] .
A photothermally harvested hADSC sheet (50.2 mm 2 ) containing 1 × 10 6 cells was transferred onto the wound (area = 50.2 mm 2 ). The free-standing hADSC sheet was transferred and attached onto the wound by virtue of the wet surface nature of the wound. In the control group (untreated), each wound was wrapped with PBS wet gauze for a while, and then the gauze was removed to reach a hydrated environment that was similar to the hADSC sheet-treated group. Figure 4 shows the progression of the wound-healing process over time. Treatment with a hADSC sheet on the skin wound resulted in more effective wound closure compared with the control (Fig. 4a). On the 5 th postoperative day, wound closure in the hADSC sheet-treated groups was similar to that of the control. However, at day 7, the hADSC sheet-treated group led to significantly faster wound closure with a wound closure effect (ε wc ) of 55% compared with that of the control (ε wc = 41%). The wound closure effect (ε wc ) was determined as follows: where A wc and A iwa are the wound closure area after treatment and the initial wound area (50.2 mm 2 ), respectively. Furthermore, at day 9, wound closure was much greater in the hADSC sheet-treated group (ε wc = 73%) than in the control (ε wc = 55%) (Fig. 4b). These results indicated the superior performance of the harvested stem cell sheets over the control in healing the chronic wound. As the stem cell sheet could be harvested from the non-invasive photothermal method within a very short time (5~10 min), the wound-healing properties of hADSC sheet could be designed to help pave the way for various stem cell therapies and organ regeneration.

Conclusion
Transferrable and free-standing hADSC sheets were obtained via the photothermal method within few minutes by controlling the cell-cell (CC) and cell-substrate (CS) interactions. A soluble protein solution based on fibronectin (FN) was employed to enhance the CC interactions and the cell proliferation rate on the CPP-PEDOT. In the presence of FN molecules, cells were in contact with neighboring cells to promote CC and CS interactions, leading to full sheet formation. The cell proliferation rate and the harvested cell sheet area could be optimized by the combination of FN concentration (10 −8 %, 100 pg/mL) and cell seeding number (1.2 × 10 6 cells) to produce a large area, transferrable, free-standing hADSC sheet (122.6 mm 2 ) by the photothermal method. A hADSC sheet was transferred to a wound on diabetic db/db mice to observe the superior performance of the harvested stem cell sheets over the control. The wound closure with hADSC was faster than that of the control (30%↑). This result indicates that the hADSC sheet harvesting system can be applied to various therapies, such as skin reproduction and organ regeneration, with immense promise as an emerging strategy in tissue engineering. Real Light) was modulated by changing the applied current using a power supply, and a thermoelectric cooling module was used to cool the hot diode laser at high power modulated by the power supply. The focusing of the laser exposed area and power density of the laser were controlled by the distance between the end of the fiber and the PS substrate and collimator. The laser exposed area was determined by NIR fluorescent detection card (Edmond Optics). The thermal image and temperature distribution data were taken by a thermal image camera (875i, Testo). A confocal microscope (Axio imager Z2, LSM 700, Carl Zeiss) was used to obtain images of the side view. Fluorescein isothiocyanate (FITC)-stained collagen was used to trace collagen dissociation. A laser was used as the excitation source for FITC (λ exi = 495 nm, λ emi = 519 nm). Preparation of the CPP-PEDOT substrate for cell culture. The PP-PEDOT substrate was prepared with pyridine (13.54 mg) and PEPG triblock co-polymer (200 mg). They were added into 1 g of the oxidative solution containing 40 wt% of iron(III) tris-p-toluenesulfonate (Fe-Tos) in n-butanol. After stirring for 6 h, the solution was sonicated for 10 min to make a homogeneous solution. The coating solution consisted of 52.8 wt% of the mixture solution of pyridine, PEPG, Fe-Tos, monomer, and 47.2 wt% of n-butanol. The molar ratio of pyridine: iron(III) tosylate: monomer was fixed at 0.55: 2.25: 1. Then, it was cooled before adding the EDOT monomers. The oxidative solution containing EDOT was spin-coated onto the polystyrene petri dish substrates at 2000 rpm for 30 sec. All the samples were polymerized at 90 °C for 1 h. After cooling to room temperature, the conductive polymer-coated substrate was washed several times with 70 wt% ethanol to remove residual oxidant, low molecular weight oligomers, and impurities. The film was then dried under nitrogen flow. The patterned PP-PEDOT surface was obtained by the polymerization of a patterned monomer layer (including oxidant and pyridine), which was prepared by the removal of an unwanted area of the pattern using a cotton pencil that was wet with solvent. The patterned monomer layer (including oxidant and pyridine) was then subjected to polymerization at 60 °C. The prepared patterned PP-PEDOT substrate was washed several times with 70 wt% ethanol and was dried before collagen adsorption for the cell-detachment experiments. Non-adherent cells were removed after 24 h by exchanging the culture medium with repeated washing with PBS. The medium was changed every 3 days. Upon reaching 90% confluence, the hADSCs were recovered using 0.05% Trypsin/EDTA and were re-plated onto a CPP-PEDOT substrate. The CPP-PEDOT substrate was washed with 1× PBS buffer and was placed in a 35-mm culture dish (Corning Inc., NY). Next, the CPP-PEDOT substrate was coated with an aqueous solution of collagen (0.3 wt%) by drop-casting, washed out with distilled water, and sterilized with a UV treatment for 5 min. Cell counts were determined by counting the number of cells with a hemacytometer after Trypan blue staining in a counting chamber. To control the shape of the cell sheet, a circle-patterned PP-PEDOT substrate (diameter = 13 mm) was used. After culturing for 1 day, CPP-PEDOT substrates with confluent cells were irradiated for 5 min with an NIR coherent diode laser (808 nm, 2.7 W cm −2 ). The shape and size of the detached cell sheets were observed using an Olympus inverted research microscope (Model IX71). Harvesting efficiency (ε d ) of the detached sheet was defined as the ratio of the detached sheet size (A hcs ) divided by the sheet-detached area (A det ).

Materials and
For the shrinkage constant (sh) of the cell sheets, the cell-cell distance was calculated with DAPI staining before and after cell detachment. The distance was measured between center of cells with staining DAPI. Cells were randomly selected within the area of 200 × 200 μm 2 . It was measured 20 times at the nearest cell and the average value was calculated. The cells were analyzed via a live and dead assay to determine cell viability. The viability of the detached cell sheet by NIR exposure was analyzed using the Live/Dead viability kit (Molecular probes, cat no. L3224) according to the instructions of the manufacturer. Briefly, harvested cells were stained with phosphate-buffered saline (PBS) containing 2 μM of a calcein AM solution and 4 μM of an ethidium homodimer-1 solution for 30 min at room temperature in the dark. After washing with PBS, images were acquired using an Olympus inverted research microscope (Model IX71).
Effect of sheet formation with soluble fibronectin molecules. For sheet formation, soluble fibronectin molecules were used as a cell adhesion molecule. Cells were seeded onto the CPP-PEDOT substrates at 4 × 10 5 cells/dish with different soluble factor concentrations (fibronectin: 1 μg/ml to 1 pg/ml). The size of the detached cell sheets was calculated by a scale bar in the optical microscope, and the sheets were compared with those at other concentrations of the soluble fibronectin molecule. To study the effect of cell junction generation with soluble fibronectin, E-cadherin expression (Santa Cruz Biotechnology, cat no. sc-7870), which represents the protein of the cell junction, was observed by fluorescence microscopy at different cell concentrations.
Characterization. An inductive coupled plasma mass spectrometer (ICP-MS, PerkinElmer: MexION300) was used to detect the dissolved Fe ions in the wash solution after washing the conductive polymer by analyzing the change of metal components according to time (1, 2 h) and number (4 times) to confirm the removal of residual metal ions. A Fourier transform infrared spectrometer (FT-IR, Bruker: Vertex70) was used to detect the unreacted monomer and oxidant after washing the PEDOT. From the measurement, the detected residual unreacted impurities were used to confirm the toxicity of PP-PEDOT. UV/Vis/NIR spectroscopy was measured to confirm the conductive polymer that was dissolved in the cell medium. After cell and cell sheet detachment, the solution samples were prepared with cell, cell sheets and medium in a quartz cell with an air reference.
Anesthesia and creation of excisional wound. All experimental procedures performed in this study followed ethical guidelines for animal studies and were approved by the Institutional Animal Care and Use Committee of Yonsei University College of Medicine (IACUC No. 2016-0095). All experiments were performed in accordance with relevant guidelines and regulations. Seven-week-old male genetically diabetic db/db mice were selected for a chronic wound-healing model. Twelve mice were randomly divided into control and experimental groups. We applied hair removal cream (Veet ® , Cat. No. 20070747) on the dorsal skin of anesthetized mice and completely removed the cream and hair after five minutes. A 50.2 mm 2 size skin wound was made on the dorsal side using an 8 mm-biopsy punch. The skin flap was removed from the wound bed.

Stem cell transplantation.
For the hADSC sheet-treated group (n = 10), a 50.2 mm 2 hADSC sheet containing 1.2 × 10 6 cells was transferred on the wound by using a polyester container that was poured into the open weave of the polyester (Urgotul ® , Laboratoires URGO, Dijon, France) with the medium. After the medium has drained through the mesh, the sheet was carefully transferred to the wound by removing only the polyester mesh after 1 minute. In the control group (n = 10), Urgotul ® that was buried in PBS was put on the wound and removed as the sheet transfer process to make a hydrated environment similar to that of the hADSC sheet-treated group. Subsequently, all wounds were covered with Tegaderm 3M sterile transparent dressing.
Wound analysis. We uncovered the wound dressings and took photographs of individual wounds every third day from the 5th day. A 16 mm-diameter circular template was used as a reference to analyze the size of the remaining wound. Using the ImageJ program, the size of the template and the remaining wound size were measured. We used Statistical analysis system (SAS) (version 9.2, SAS Inc., Cary, NC, USA) along with the LMM (linear mixed model) method for measuring and calculating the wound size for confound symmetry covariance within mouse P-value with bonferroni correction. Setting the first wound size as 100%, we calculated the time for the wound to reduce to 70%, 50%, 30%, and 10%.