Injectable Small Intestinal Submucosa and Adipose-Derived Stem Cells Composite Gel Promoting The Deep Partial-Thickness Burns Repair


 Backgroud:

Burns wound treatment remains a significant clinical challenge around the world. Although stem cell-based scaffold therapies are promising strategy for burn wounds, its clinical therapeutic effect is still not satisfactory nowadays. Herein the aim of this study is to evaluate the therapeutic efficacy of injectable small intestinal submucosa (SIS) and rat adipose-derived mesenchymal stem cells (ADSCs) composite gel to repair the deep partial thickness burns in rats.
Methods

The deep partial-thickness burns model in rats were made by contacting the dorsal surface SIS memberance directly with boiled water for 10 seconds. After scalding, the wound edge and the central area were injected for phosphate-buffered saline (PBS) solution, ADSCs, injectable SIS and injectable SIS/ ADSCs composite gel, respectively. At 3, 7, 14 and 21 days post injection treatment, the burn wound closure percentages were evaluated. Moreover, micro-vascular density and epidermal thickness assessment in burn wound were performed by histopathology examination or immunofluorescence. Besides, the expression of genes related to wound angiogenesis and re-epithelialization were determined in vitro.
Results

Our data revealed that that injectable SIS gel could provide a well-grown microenvironment for ADSCs in vitro, and the ADSCs-SIS composite gel could synergistically promote the deep partial-thickness burn repair via paracrine and differentiation mechanisms.
Conclusions

Taken together, this study shows the ADSCs-SIS composite gel is a promising candidate for burn wound regeneration.

coverage includes auto-grafts, xeno-grafts, and skin substitutes [6][7][8]. However, autologous skin source is seriously inadequate, and auto-grafts exhibit insu cient angiogenesis [9]. Moreover, the use of xenografts is usually limited by antigenic rejection and the risk of infection. Consequently, the healing of the burn wounds is not promising, and hypertrophic scarring is commonly formed over time. Thus, new strategies are required to treat burn injuries.
Presently, stem cell-based scaffold therapies are research hotspots, of which mesenchymal stem cells (MSCs) are the most studied [10,11]. In particular, compared with other stem cells, adipose-derived stem cells (ADSCs) have a wide range of easily obtained donors, fast expansion, low immunogenicity, minimally invasive extraction and signi cant proliferative capacity [12]. In addition, the ADSCs can produce and secrete various growth factors, such as vascular endothelial growth factor (VEGF), hepatocyte growth factor (HGF), transforming growth factor-β (TGF-β), and broblast growth factor (FGF) [13]. Therefore, ADSCs have recently been proposed as a potential dominant cell for treating burn wounds [14][15][16]. In terms of the traditional stem cell-based therapies, the cell transplants are used either by intravenous or direct injection into the wound area, where the transplanted cells can differentiate into related epithelial cells and secrete various kinds of growth factors to promote wound healing.
Nevertheless, most studies have demonstrated that the clinical outcome of stem cell-based therapies are lower than expected due to the low survival rate and poor bio-distribution in tissues [17]. Thus, given these limitations in treatment, it is crucial to seek an effective cell micro-carrier to make transplanted cells play their maximum role.
Small intestine submucosa (SIS), derived from the submucosal layer of the porcine intestine, is an acellular, natural extracellular matrix (ECM) biomaterial [18]. SIS is rich in collagen, glycosaminoglycans, and various growth factors [19]. Besides, it has good mechanical properties, histocompatibility, and low immunogenicity [20]. Membrane SIS has been used in animal experiments and clinical practice to repair and reconstruct different types of tissue defects, especially in hernia [21], cardiovascular disease [22], urinary system diseases [23], refractory skin trauma [24], and burn injuries [25]. Currently [27]. Unlike traditional solid (including membranous) scaffold materials, injectable bio-scaffold material can be arbitrarily molded and simply operated, quickly presenting the cells to the lesion area with smaller trauma and faster recovery [28]. Although ADSCs seeding traditional solid SIS have been extensively used in soft-tissue reconstruction [29][30][31][32], little attention has been devoted to combining ADSCs with SIS gel to repair burn wounds.
Therefore, this study aimed to evaluate the therapeutic e cacy of ADSCs-SIS composite gel to treat deep partial-thickness burn wounds. The present study's speci c objectives were: 1) to evaluate whether the membrane SIS could be transformed into injectable SIS and provide a well-grown microenvironment for ADSCs in vitro, 2) to assess if the injectable SIS could promote the repair capacities of ADSCs in deep partial-thickness burns wounds, and 3) to elucidate the therapeutic mechanisms of promoting the burn wound healing.

Materials And Methods
Isolation, culture, phenotype, differentiation, and labeling of ADSCs This study was approved by the China Ethics Committee and performed following the ethical standards. The inguinal fat tissues from adult male Sprague-Dawley (SD) rats (200-250g) were washed with phosphate-buffered saline (PBS) solution, followed by cutting into a paste by ophthalmic scissors. The tissues were then digested with 0.1% type I collagenase (Sigma, USA) at 37°C for 60 min in a shaking water bath. The adipose tissue suspension and cells were ltered through a 70-µm lter and added to the cell medium, vortexed at 1500 rpm for 10 min, and the fat layer and media were discarded. The cell pellet was resuspended in Dulbecco's Modi ed Eagle Medium-F12 glucose (DMEM-F12, Hyclone, USA) with 10% fetal bovine serum (FBS, Hyclone, USA) and 1% penicillin-streptomycin (Invitrogen, USA). The cells were seeded at 5×10 5 /mL in culture bottles and incubated at 37ºC in 5% carbon dioxide. The thirdgeneration cells were identi ed by ow cytometry using monoclonal antibodies speci c for CD29-APC (Biolegend), CD90-PE-Cy7 (Biolegend), CD34-PE (Biolegend), CD45-PerCP (Biolegend), and CD31-PE (Biolegend), and their isotype controls. Meanwhile, the adipogenic and osteogenic differentiation potential of the cells was identi ed using Oil red O (Sigma) and Alizarin red (Sigma). The third-passaged cells were harvested and labeled with CM-DiI (Molecular Probes, USA) for cell tracking according to the manufacturer's protocol.

Injectable SIS preparation
A standard procedure described previously [33] was used to prepare the injectable SIS. The proximal jejunum was collected from a healthy fresh pig (around 100 kg at six months) within four h of sacri ce. Then the jejunum was washed by saline solution repeatedly. The porcine small intestine was obtained by mechanically removing the tunica mucosa, serosa, and tunica muscularis, followed by careful washing with saline solution. The membrane SIS was then freeze-dried at 80ºC for 48 h using a freeze-dryer (LGJ-18C, China), followed by grinding into a powder of about 10-20 µm at -198°C using a freeze ball mill (Retsch, Germany). The obtained SIS was continuously stirred in a medium containing 3% acetic acid and 0.1% pepsin of the aqueous solution for 48 h. The pH value was adjusted to neutral by sodium bicarbonate. The neutralized solution was freeze-dried to yield the nal SIS powder, followed by sterilization with ethylene oxide gas. The nal SIS suspension was dispersed in PBS to achieve a concentration of 20% by weight of injectable SIS. Injectable SIS was evaluated by hematoxylin and eosin (H&E) staining and observed under a scanning electron microscope.

Material extract of injectable SIS preparation
According to the preparation method of GB/T16886.12-2000, the complete medium (extraction medium) was added to an ori ce plate containing the powdered SIS under sterile conditions, and the material/extraction medium was set at a ratio of 0.1 g/mL. After incubation for 24 h at 37ºC in 5% carbon dioxide, the material extract (100% w/v) in the ori ce plate yielded a concentration of 100% by volume.

Co-culture of injectable SIS and CM-DiI-labelled ADSCs
The injectable SIS/cell culture medium was mixed at a ratio of 0.1 g/mL and incubated at 37°C in 5% carbon dioxide using 6-well plates. After incubation for 24 h, CM-DiI-labelled ADSCs (1×10 6 cells) were seeded on the injectable SIS as described previously [38]. The complexes were observed by uorescence microscopy (Olympus, Japan) and scanning electron microscopy (SEM, Cambridge, England).

Assessment of the proliferation and viability of ADSCs in the injectable SIS
To evaluate the effect of injectable SIS on the proliferation of ADSCs, a CCK-8 (Tong Ren Chemistry, Japan) assay was performed. The third-passaged ADSCs (5×10 4 cells/mL) were seeded onto 96-well plates with cell culture medium and divided into an experimental group (injectable SIS/ADSCs) and control group (only ADSCs). They were all incubated at 37°C in 5% carbon dioxide. On days 1, 3, 5, and 7 after cell incubation, the original medium was removed, and then the two groups were all added to 100 µL of serum-free medium and 10 µL of CCK-8 reagent. After incubation for two hours, the absorbance values (OD value) of the cells were measured at 450 nm.
To evaluate the effect of injectable SIS on the viability of ADSCs in injectable SIS, a live/dead assay was performed. The third-passaged ADSCs (5×10 5 cells/mL) were seeded onto 6-well plates with 1 mL of cell culture medium. After four hours, the cells were attached, and the original medium was removed. Then 100% material extract (experimental group) and fresh cell culture medium (control group) were added into the three wells. The cells were incubated at 37°C in 5% carbon dioxide. On days 1, 3, 5, and 7 after cell incubation, the original medium was removed, and the cells were washed with PBS gently. Each group was then added a freshly prepared concentration of 4 µmol/L of Eth-d1 and two µmol/L of Calcein-AM, and incubated for 30 min at 37ºC. Immediately, the luminescence of the cells was observed. Live cells exhibited green uorescence; dead cells exhibited red uorescence. Under the three elds of vision (magni cation ×100), the number of dead cells was observed randomly in the center and edge of the ori ce at each time interval, and the ratio of dead cells/total cells was calculated.

Rat skin burn model and surgical procedures
The deep partial-thickness burn model was established by applying directly boiled water (100°C) with a heated, about 2.5 cm diameter round as previously described with minor modi cation [34]. Thirty-two adult male SD rats were anesthetized with pentobarbital sodium (50 mg/kg) by intraperitoneal administration. The dorsal hair was removed with an 8% Na 2 S aqueous solution. A hollow plastic tube measuring about 2.5 cm in diameter was placed close to the rat's back, and then 10 mL of boiled water (100°C) was added into the tube by a 10-s direct contact. The deep partial-thickness burn wounds were assessed by pathologic examination.
The ADSCs were labeled by CM-DiI as described earlier in the study. The burned rats were randomly divided into four groups (n = 16): group A, control group (1 mL of PBS); group B, ADSCs only (1×10 7 cells resuspended in 1 mL of PBS); group C, injectable SIS only (1 mL of 20% injectable SIS); group D, ADSCs/SIS (1×10 7 cells mixed with 1 mL of 20% injectable SIS). After the burn, the rats were immediately injected with medication: along the wound edge at six equidistant points and the wound center, four points for multi-point intradermal injection (0.1 mL for each point) using a 1-mL syringe (25-G needle) as previously described with minor modi cation [35]. In the control group, 1 mL of PBS without ASCs or injectable SIS was injected into the intradermal layer in the same manner. Then, vaseline gauze and sterile gauze were used in turn to cover the wound in all the groups. Postoperatively, ceftazidime, and buprenex were intraperitoneally injected for seven days.
Burn wound closure measurements, calculation of capillary density and epidermal thickness The rats were observed every day, and digital photographs of the wounds were taken on days 3, 7, 14, and 21 postoperatively. The wound area was measured by tracing the wound margin and calculated using Image-Pro Plus 6.0 (IPP 6.0) software (Media Cybernetics, USA). The percentage of wound closure was calculated as follows: Wound closure rate (%) = (wound area on the postoperative day 0 -wound area on the postoperative day "X")/ (wound area on the postoperative day 0) ×100.
To assess the capillary density, H&E staining of the tissue specimens was performed in all the groups on postoperative day 7 to assess neoangiogenesis. Tissue specimens from each rat were made into a slice, and three areas with the largest number of microvessels were selected under low magni cation (×40); subsequently, ve non-repetitive elds were randomly selected under high magni cation (×100). The number of microvessels was analyzed and counted using the image analysis software IPP 6.0. Only mature vessels containing erythrocytes were counted.
To assess the epidermal thickness, H&E staining of the tissue specimens from all the groups was performed on postoperative day 21. The tissue specimens from each rat were made into a slice, and ve high-power elds per section were selected under high magni cation (×100), followed by performing ve measurements of the epidermal thickness per eld.

Molecular analysis: Real-time quantitative PCR
For in vivo RT-PCR, wound tissue specimens were harvested from all the groups on postoperative days 3, 7, 14, and 21 and stored in RNA stabilization solution (Thermo, America) at -20°C. RNA isolated from burn wounds treated with ADSCs, injectable SIS, ADSCs/SIS, or PBS alone was used to determine the relative changes in mRNA expression of rat VEGF, EGF, and bFGF. Total RNA of the burn wounds was extracted with RNeasy Mini Kit (QIAGEN, Valencia, CA) according to the manufacturer's protocol, and cDNA was synthesized using a reverse transcription kit (ReverTra Ace qPCR RT Master Mix, TOYOBO, Japan). SYBR Green Real-time PCR Master (TOYOBO, Japan) was used to amplify the target cDNA. Mix Quantitative PCR analysis was then performed on a StepOne real-time PCR system (Applied Biosystems, Alameda, CA, USA). The primers targeting VEGF, EGF, bFGF, and the control glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were designed by Sango Biotech (Shanghai, China); the primer sequences are summarized in Table 1. Cycling conditions were performed: 95ºC for 30 s, 40 cycles at 95ºC for 5 s, optimal annealing temperature for 10 s, and 72ºC for 15 s. The relative expression levels of target genes were normalized to the expression of the GAPDH gene. The expression of relative genes was determined using the 2 −△△Ct formula.

Immuno uorescence
Fresh tissue specimens on postoperative days 7 and 21 were embedded in tissue-freezing medium with optimum cutting temperature (OCT) (Leica) and immediately cut into 10-µm frozen sections each. To identify dermal angiogenesis, the CD31 antibody level was determined on postoperative days 7. Frozen sections were placed in cold acetone (4ºC) for 10 min and rinsed with PBS three times (5 min), followed by incubation with monoclonal rat-anti-CD31 primary antibody (1:100, sc-53526-FITC; Santa Cruz, CA, USA) at 37ºC for 30 min. To assay the neo-epidermis, the CK10 antibody level was determined on postoperative days 21. The frozen sections were rinsed with PBS three times (5 min) and blocked with 10% goat serum for 30 min at 37ºC, followed by incubation with primary antibody CK10 (rabbit-anti-rat; 1:100; ab76318, Abcam, Cambridge, MA, USA) at 4ºC overnight. Then, the sections were rinsed with PBS and incubated with secondary antibody (anti-rabbit IgG-FITC; 1:200; ab6717, Abcam, Cambridge, MA, USA) for one h at 37ºC. Finally, DAPI (4,6-diamidino-2-phenylindole; 2.5 µg/mL) was used to stain nuclei for 2 min. The slices were immediately observed and photographed under a uorescent microscope.

Statistical analysis
The data were presented as means ± SD. SPSS 19.0 was used to analyze the experimental data. The odds ratio between the two groups was compared using the t-test of two independent samples. A singlefactor ANOVA was used to compare the scores of multiple groups, and the difference between groups was determined by further analysis with LSD tests. The difference was statistically signi cant at P < 0.05.

Results
Characterization, labeling of ADSCs and co-culture with injectable SIS As shown in Fig. 1A, The third generation of ADSCs exhibited a long spindle shape, vortex-like growth, and robust proliferation capability. The CM-DiI-labeled ADSCs showed red uorescence and labeled only the cell membrane and cytoplasm, without labeling the nucleus (Fig. 1B).The Oil Red O and Alizarin red staining showed that these cells had the potential for adipogenic and osteogenic differentiation (Fig. 1C). Moreover, the ow cytometry data showed that the third-generation cells were positive for the phenotype of MSCs: CD29 and CD90 and negative for CD34, CD45, and CD31 (Fig. 1D). Similar to the previous study [26], SEM micrographs showed that the injectable SIS had very different sizes of three-dimensional pores, the ber was complete (Fig. 1E), and the CM-DiI-labeled ADSCs attached to the material well, exhibiting oval and cord-like growth by the SEM (Fig. 1F).

The injectable SIS gel had good biocompatibility and improved the proliferation of ADSCs
As shown in Fig. 2A, a live/dead assay showed the number of cells in the experimental and control groups increased over time. The ratio of dead cells/total cells was not different in the experimental group on days 1 (1.8%), 3 (2.1%), 5 (2.5%), and 7 (3.9%) compared with the control group (1.5%, 1.6%, 2.4% and 3.6%, all P>0.05, respectively) (Fig. 2B). Besides, the OD value of the cells was also calculated by a CCK-8 assay to compare the proliferation rate of ADSCs seeded in the injectable SIS with the traditional 2D culture (Fig. 2C). Not surprisingly, the OD value of the experimental and control groups increased gradually on days 1, 3, and 5 after culturing, and on day 7, the OD value of the two groups decreased. Speci cally, there was an increase in OD value on day 1 (0.5004±0.1073, *P < 0.05) and day 5 (1.1032±0.0984, *P < 0.05) in the experimental group compared with the control group (0.3568±0.0577 and 0.9302±0.0864, respectively), suggesting that injectable SIS could improve ADSCs proliferation and had good biocompatibility.

Enhanced effects of injectable SIS gel on potential of ADSCs in rat burn wound healing
As shown in Fig. 3A, four groups of wounds decreased in size over time. Moreover, the wounds in the D group (ADSCs/injectable SIS gel) healed faster than that in other groups. Form the results of Fig. 3B, it was shown that compared with the B group (injectable SIS) and C group (ADSCs), the D group (ADSCs/injectable SIS gel) leaded to a higher would closure rate at every time intervals.
ADSCs/injectable SIS gel had an enhanced effect on angiogenesis in burn wound repair HE staining revealed greater penetration of blood vessels in the wound bed of the four groups. Most of the capillaries in D group showed clear blood vessel contours and red blood cells within the vessels (Fig. 4D). However, only a small part of the capillaries was found in groups B and C; moreover, some of the blood vessels were not clear, and the erythrocytes in the blood vessels were scattered (Fig. 4B, C). Additionally, the capillaries in group A were not clear, and only scattered red blood cells were visible (Fig. 4A). The capillary density in group D (39.2 ± 7.3 vessels per eld) was higher than groups A (11.5 ± 4.6 vessels per eld), B (24.6 ± 6.3 vessels per eld), and C (20.1 ± 9.7 vessels per eld) (*P < 0.05) (Fig. 4E). The capillary densities in groups B and C were both signi cantly higher than that in group A (*P < 0.05), while no signi cant difference was observed between groups B and C (P > 0.05) (Fig. 4E).

ADSCs/ injectable SIS gel had enhanced effects on reepithelialization in burn wound repair
Re-epithelialization of the wounds was observed in groups B, C, and D, while there was no apparent epidermogenesis in group A (Fig. 5A, B, C, D). In addition, the Groups B, C, and D showed a signi cantly thicker (#P < 0.01) epidermis than group A (Fig. 5E). Moreover, group D exhibited a signi cantly thicker epidermis (*P < 0.05) than groups B and C, while no signi cant difference was found between groups B and C (P > 0.05) (Fig. 5E).

ADSCs/ injectable SIS gel improved the wound healing by paracrine the angiogenic and epidermal growth factors
To elucidate the mechanisms of promoting the burn wound healing, the Real-time polymerase chain reaction analysis showed that the gene expressions of VEGF, bFGF, and EGF were signi cantly higher in group D than groups A, B, and C at the four time intervals (P < 0.01 or P < 0.05 ) (Fig. 6). Speci cally, the expression levels of VEGF and bFGF genes were much higher in groups B and C than group A at some time intervals (P < 0.01 or P < 0.05); in addition, VEGF and bFGF exhibited higher expression in group D than groups B and C at some time intervals (P < 0.01 or P < 0.05), while VEGF gene expression on days 3 and 7 and bFGF gene expression at the four time intervals groups B and C were not signi cantly different. The EGF level was signi cantly higher in group D than the other three groups on days 14 and 21 (P < 0.01). Moreover, statistically signi cant enhancement in EGF gene expression was detected in groups B and C than group A on day 21 (P < 0.01 or P < 0.05), with no signi cant difference between all the study groups on days 3 and 7. Interestingly, the EGF level was signi cantly higher in group D than in groups B and C (P < 0.05). These ndings indicated that ADSCs/SIS could signi cantly improve growth factor secretion from ADSCs or SIS and might be an enhanced strategy to promote ADSCs or SIS-assisted wound repair.

Immuno uorescent analysis
To elucidate the differentiation mechanisms of promoting the burn wound healing, the immune-staining for CD31, CK10 endothelial protein were consequently performed. The staining showed CD31 vascular endothelial cell marker in green uorescence (Fig. 7A). The cluster-like CM-DiI-labeled ADSCs were scattered and clustered at the wound bed and exhibited red uorescence (Fig. 7B), parts of which were colocalized with the vascular endothelial cell marker CD31 (Fig. 7C), with a dendritic vascular structure, indicating that ADSCs can be spontaneously differentiated neovascularization in the wound bed. Likewise, Immuno uorescence staining of DAPI revealed the cell nuclei in blue (Fig. 7D); the red uorescent signal of CM-DiI-labeled ADSCs was distributed in the wound bed (Fig. 7E); the green uorescent signal of the clear epidermal structure was seen after CK10 immuno uorescence staining (Fig. 7F). The red uorescence of ADSCs, cell nucleus stained by DAPI, and neoepidermis stained by CK10 were well fused (Fig. 10G) [44], indicating that the ADSCs/SIS can be spontaneously differentiated into neoepidermis.

Discussion
In this study, we evaluated the therapeutic e cacy of ADSCs-SIS composite gel in treating deep partialthickness burn wounds and whether SIS gel could enhance the capacities of ADSCs in burn wound repair synergistically. The results showed that injectable SIS gel could provide a well-grown microenvironment for ADSCs in vitro, and the ADSCs-SIS composite gel could synergistically promote the deep partialthickness burn repair via paracrine and differentiation mechanisms, suggesting their use as a promising candidate for burn wound regeneration.
Although the MSC-based therapies have been reported to enhance wound healing[36], the poor biodistribution and low in vivo survival of the transplanted cells limit their applications, which is attributed to the hypoxic environment in defect tissue [17]. Thus, optimizing an appropriate stem cell vehicle that promotes stem cell survival and retention is an essential component in wound healing. Lots of natural and synthetic injectable scaffolds have been proposed to support the attachment, proliferation, and differentiation of ADSCs to provide an ideal environment for cell survival, including hyaluronic acid, ECMbased natural materials, synthetic peptides, and polymer-based materials [27,28,[37][38][39]. The injectable hydrogels derived from natural ECM have been reported to have better biocompatibility and lower immunogenicity than synthetic polymer-based hydrogels [40]. It is noteworthy to mention that SIS is among the most commonly used non-immunogenic xenogenic ECM scaffolds that retain various growth factors, cytokines, or other functional proteins and can be easily converted into injectable hydrogels [41,42]. Liu et al demonstrated that ADSC-seeded scaffolds enhanced the proliferation, anti-apoptosis, and angiogenesis compared with the non-seeded scaffolds in a murine skin injury model. Moreover, the SIS and acellular dermal matrix promoted the vascularization capacity of ADSC than that of Co-CS-HA [43]. In similar studies, Jeong-Seok Choi recently reported that the injectable SIS hydrogels could promote AdMSCs survival and enhance the remodeling e cacy of adipose-derived mesenchymal stem cells in a radiation-damaged salivary gland model [44]. Consistent with these studies, we also successfully isolated and cultured ADSCs from inguinal fat tissues in vitro and combined it and injectable SIS gel in the present study. From the SEM, CCK8, and uorescence microscopy analysis results, the injectable SIS was found to be an appropriate carrier with good biocompatibility to promote the adherence and proliferation of the transplanted ADSCs.
This study also investigated the e cacy of ADSCs-SIS composite gel in treating burn wounds in a rat model. It was demonstrated that the injectable SIS and ADSCs composite gel could provide the cues for better burn wound healing. The wound closure was obviously accelerated in the ADSCs, SIS, and ADSCs/SIS gel groups compared to the control group. Additionally, the wound closure rate in the ADSCs/SIS gel group was the highest. Besides, ADSC/SIS gel groups exhibited a signi cantly thicker epidermis than ADSCs, SIS groups, respectively. Likewise, several studies have reported similar capacity of stem cells or SIS/stem cells in wound repair. IZhang et al reported that the nano-silver-modi ed porcine small intestinal submucosa could improve the healing of infected partial-thickness burn wounds [25]. In the pig animal model, scars injected with ADSCs exhibited a reduction in surface area and improvements in color and pliability compared with the control group [45]. Thus, it is not surprising to observe that the injectable SIS scaffold could promote wound healing, and ASCs/SIS gel could enhance this effect in this study.
Besides, the therapeutic mechanisms of promoting the healing of burn wounds by the ADSC/SIS gel were also explored in this work. The results showed that after the injection of ADSC or SIS gel into the burn wounds, increased capillary density and promoted epidermis were observed by the HE assay compared to the control group. Furthermore, among the four groups, the capillary density in the ADSC/SIS gel group was the highest (39.2 ± 7.3 vessels per eld). The results indicated that SIS gel could enhance the angiogenic capacities of ADSC, consistent with the previous studies [34]. The critical factors are related to the SIS containing massive ECM, which is very important for cells to respond to signals in the cellular microenvironment and provides a supporting medium to form blood vessels [46]. Another vital reason may lie in that the SIS can secret numerous growth factors, which improved angiogenic factor secretion ability of the ADSCs. Interestingly, we found gene expressions of VEGF and bFGF were the highest in the ADSCs/SIS gel group in this study. It was previously have been demonstrated that the VEGF, EGF, and bFGF stimulated ADSCs migration, enhancing their residence seeding time onto SIS, with a co-stimulatory effect on ADSCs endotheliogenesis [31]. In other words, it could be a mutual promotion in regulating neovascularization after ADSCs were seeded on the injectable SIS. Similarly, injectable SIS and ADSCs composite gel made the ADSCs and SIS synergistic in promoting re-epithelization by the paracrine mechanism.
Furthermore, Immuno-uorescent histology experiments demonstrated that CM-DiI-labeled ADSCs were co-localized with staining for CD31 and CK10 in this study. We propose the vital role of ADSCs differentiation in wound repair, including endothelial and epithelial lineages. However, this might still be the result of cell fusion between ADSCs, vascular endothelial host cells, and epidermal host cells in rats, with the cell fusion being considered a mechanism of newly formed functional cells [47]. Recently, other mechanisms of promoting the wound burn were explored for the MSCs/SIS gel. Zhang et al. reported that they created urine-derived stem cells/SIS gel composite and demonstrated that hypoxic preconditioning would improve its wound healing potential [48]. Some other studies also demonstrated that exosomes could effectively improve skin cell functions in vitro, enhancing wound healing via Wnt4/β-catenin pathways [49]. It can also modulate the biological cell-cell interactions or endocrine mechanisms by in uencing distant cells via PI3K/Akt in a paracrine manner [50]. Meanwhile, the exosomes can also enhance ATP levels and decrease oxidative stress levels [51]. Recently, ESC-derived miRNAs, such as miR-291a-3p, exhibited an anti-aging action in human dermal broblasts via the TGFβ receptor-2 pathway and improved wound healing [52]. There are also some limitations in this research: 1) No immune-uorescent quantitative analysis was performed to evaluate the differentiation of ADSC/SIS composite. 2) The level of immune and in ammatory factors in burn tissues were not evaluated, which were important for burn wound repair. 3) The gene-related mechanism of the ADSC/SIS composite in wound burn repair was not explored, necessitating further investigations.

Conclusion
In this study, we successfully transformed the traditional membranous SIS into injectable SIS and demonstrated that it could provide a well-grown microenvironment for ADSCs in vitro. Furthermore, our in vivo data revealed that locally administered injectable SIS and ADSCs composite gel was superior to the un-seeded SIS or local injection of ADSCs in accelerating wound closure. Besides, injectable SIS and ADSCs composite gel secreted more VEGF, bFGF, and EGF in comparison with only injectable SIS or ADSCs, which were synergistic in promoting neovascularization and re-epithelialization. These nal ndings suggest that injectable SIS and ADSCs composite gel promotes the deep partial-thickness burns repair mainly via paracrine and differentiation mechanisms, suggesting their use as a promising candidate for burn wound regeneration. The data that support the ndings of this study are available from the corresponding author upon reasonable request. All the data are presented in this study.

Funding
This work was nancially supported by the National Science Foundation of China (grant no. 31271049). None of the authors had professional or nancial a liations that could potentially bias this study.
Ethics approval and consent to participate All animal experiments were approved by the Animal Ethics Committee of the A liated Hospital of Southwest Medical University Hospital and were conducted following the Principles of Laboratory Animal Care of Southwest Medical University. No patients were participated in this study.

Consent for publication
Not applicable.

Disclosure of potential con icts of interest
The authors declare no con icts of interest. Table 1 is not available with   The expressions of EGF, VEGF and bFGF mRNA were assayed by qRT-PCR on postoperative days 3, 7, 14, and 21. *p < 0.05