Construction of Sandwich-like cell sheets with natural electrospun gelatin/chitosan nanofibrous delivered plasmid VEGF for accelerated wound healing

Background Development of natural biodegradable electrospun nanofibrous with appropriate physical properties and biocompatibility is highly desirable to support multi-layer cell sheets construction for wound healing. Results We developed a series of electrospun gelatin/chitosan nanofibrous with different gelatin/chitosan ratios and controlled pore sizes, and impregnated plasmid VEGF into membrane, which as supporting membrane to construct sandwich-like adipose-derived stem cells (ADSCs) cell sheets with a simple and effective technique for accelerated wound healing. We found that the physical properties of the electrospun nanofibrous including water retention, stiffness, strength, elasticity and degradation could be tailored by changing the proportion of gelatin/chitosan. We further observed that the optimized electrospun nanofibrous with the optimal ratio of gelatin to chitosan (7:3) which were soft and elastic could most effectively support cell adhesion, proliferation and migration into the whole nanofibrous membranes. Nanofibrous delivered plasmid VEGF facilitating multi-layer ADSC cell sheets formation and promoting regeneration of cutaneous tissues within two weeks. Conclusions Such natural biodegradable and biocompatible electrospun gelatin/chitosan nanofibrous with plasmid have the potential to become fully cellularized and support sandwich-like ADSC cell sheets formation, which will make it suitable for widespread applications such as skin substitute or wound dressing.


Background
Regenerative medicine is promoted as a promising treatment for the difficult-to-treat diseases and physically impaired function in patients, which heralds a revolutionary shift from conventional symptomatic treatment to radical treatment [1]. Replacement of injured 3 or lost tissues with appropriate cells or tissues is one of desired mechanisms in regenerative medicine. To maximize the regenerative effects, it is indispensable to transplant amount of cells for a long time. Tissue engineering is a promising strategy to overcome inadequate cell number and cell loss after transplantation [1]. Researchers have developed biodegradable scaffolds for tissue engineering and scaffold-free cell sheetbased tissue engineering. Cell sheet engineering involves many disciplines including medicine, biology, engineering, and pharmacy [2,3], because the noninvasive cell sheet can be engrafted to the desired transplantation site without suturing the cell sheet.
Transplanted cell sheets can replace injured tissue and compensate impaired function when implanted in the ectopic region.
Although the cell sheet formation is a promising and compelling technology for tissue engineering, it is challenging to harvest cell sheet with excellent matrix stability. Single layer cell sheets are quite fragile and easily crumple when they are removed from the culture surface with forceps or other tools, so cell sheet formation and transplantation processes, such as the layering of sheets, are delicate and labor-intensive tasks [9].
Moreover, layering multiple cell sheets provides limited space for cell growth and nutrient supply, so it's hard to obtain superior cell viability and differentiation potential. A natural biodegradable and non-toxic nanofibrous membrane to support cell sheets would be of great benefit for cell-sheet engineering. The supporting membrane should have unprecedented porosity, a high surface volume ratio and three dimensional (3D) porous structures for cell growth and nutrient supply. Electrospinning is considered to be a simple and versatile method to fabricate nanofibrillar membrane, and the morphology of electrospun nanofibrous can closely mimic the structure of the native ECM and, thus, 4 facilitate cell adhesion, proliferation and differentiation [10]. Electrospun nanofibrous is well known for its native tissue mimicking properties in tissue engineering application [11]. Thus, incorporating essential components of ECM in electrospun nanofibrous to mimic native microenvironment is necessary and efficient approach to support cellular proliferation and differentiation in tissue regeneration.
As is well known, The primary constituent of the ECM is typically collagen [12], gelatin is a denatured form of collagen, it is completely resorbable in vivo, its physicochemical properties can be suitably modulated due to the existence of many functional groups, because of its biological origin, biocompatibility, biodegradability, commercial availability and has not shown any antigenicity, it has been popularly used in tissue engineered scaffolds [13]. Chitosan is also promising biopolymers for tissue engineering because of its biocompatibility, biodegradability, antibacterial and antifungal activity [14,15].
Moreover, it can be easily processed and formed into various shapes [16], and could achieve hemostasis and allow the promotion of normal tissue regeneration. Chitosan have been applied to diverse biomedical research and therapy, such as drug-delivery carriers, surgical thread and bone healing materials, especially wound dressing [15,17].
Thus in this study, by combining chitosan and gelatin, with impregnation of plasmid DNA pCMV6-AC-VEGF, we synthesized a crosslinkable electrospun gelatin/chitosan nanofibrous with plasmid. Modification of gelatin and chitosan proportion further introduced glutaraldehyde crosslinking, endowing the natural electrospun gelatin/chitosan nanofibrous with controllable porosity, hydrophilicity, degradation and mechanical properties, as well as good biocompatibility with ADSC, and nanofibrous with plasmid VEGF also have good biocompatibility with ADSC, which all required for aiding wound healing. By mimicking the structure of the native skin, we constructed ADSCsgelatin/chitosan nanofibrous with plasmid-ADSCs sandwich-like cell sheets, and proved 5 that the sandwich-like cell sheets with plasmid VEGF can effectively aid wound healing by guiding cellular processes, making them ideal candidates for skin regeneration.

Results
The schematic protocol of electrospun gelatin/chitosan nanofibrous development and sandwich-like ADSC cell sheets construction for wound repairing was summarized in Fig. 1.
Electrospun gelatin/chitosan nanofibrous with plasmid pCMV6-AC-VEGF were developed by electrospinning. The cross-linked electrospun gelatin/chitosan (7:3) nanofibrous had appropriate degradation behavior, water retention, mechanical property, and the best biocompatibility with ADSCs compare to the other membranes. Gelatin/chitosan (7:3) nanofibrous was selected for plasmid loading and cells-membrane-cells sandwich-like cell sheets construction, ADSCs were cultured on temperature responsive surface for cell sheet formation. After transplantation, compare to no transplantation and transplantation of nanofibrous with or without plasmid, sandwich-like cell sheets can promote the wound healing significantly, and the structure of repaired skin healing by cell sheets with plasmid is close to normal skin.
Physical characteristics of the electrospun nanofibrous SEM was used to characterize morphologies of different electrospun nanofibrous and to image the morphological change after cross-linking. Figure 2A shows SEM images of electrospun nanofibrous (non-cross-linked and cross-linked) with different proportion. A well interconnected and randomly oriented electrospun nanofibrous formed a highly pore network structure. SEM images revealed that the electrospun nanofibrous were fine, smooth and overlapped, forming thick network of electrospun nanofibrous, but after crosslinking the uniformity and smoothness of nanofibrous were lost. When the concentration of chitosan was increased, both the average nanofibrous diameter and the pore size of non-6 cross-linked electrospun nanofibrous decreased. After cross-linking, both the average nanofibrous diameter and the pore size almost remained unchanged (no significance)  Figure 2E shows the residual mass as a percentage of the original sample mass at day 0. The largest loss of mass was observed in the first 10 days, and electrospun gelatin/chitosan nanofibrous 4:6 have the largest mass loss, all the membranes degraded slowly.
The cross-linked electrospun nanofibrous were further characterized to determine their fluid absorption ability after submersion in the PBS solution. The degree of swelling of the gelatin/chitosan nanofibrous 9:1 and 8:2 was 450%, while that of 4:6 electrospun nanofibrous were 230%. The membrane with more chitosan showed less absorption ( Fig. 2G). It supported our results in SEM which showed that nanofibrous pore size was decreased with the increase of chitosan showed less percent swelling than all. 7 ADSC adhesion, spreading and proliferation on electrospun nanofibrous Figure 3A showed that the ADSCs were spread on the electrospun gelatin/chitosan nanofibrous after seeded for 1 day, well grown and had spread extensively, adhered on the surfaces of the nanofibrous, and more gelatin seems help more cells adhesion and spreading. After 3 days proliferation, under SEM, we observed that the shape of ADSCs changed from round to elongated and almost covered the electrospun nanofibrous 7:3 ( Fig. 3B). The cells almost covered the whole membrane after 1 week proliferation on electrospun gelatin/chitosan nanofibrous 7:3 and 6:4 ( Fig. 3C), the shape of ADSCs changed from round to elongate. Nanofibrous with more chitosan (5:5 and 4:6) seems not helpful for ADSCs proliferation. The above results demonstrated that the degradation behavior, water retention, and mechanical strength of g electrospun elatin/chitosan nanofibrous could be changed upon variation of the ratio of gelatin and chitosan. To increase the amount of chitosan in the cross-linked electrospun gelatin/chitosan nanofibrous resulted in decreased water retention, strength and elasticity but increased degradation, and the cross-linked electrospun gelatin/chitosan (7:3) nanofibrous had the best biocompatibility with ADSCs compare to the other nanofibrous. Gelatin/chitosan (7:3) nanofibrous was selected for plasmid loading and in vivo experiments because it was more comparable with natural skin in terms of the capability to retain water, elasticity, strength and good biocompatibility with ADSCs.
Sustained release of plasmid VEGF by electrospun gelatin/chitosan nanofibrous The structure of nanofibrous-plasmid complex was observed under SEM. The electrospun nanofibrous of gelatin/chitosan were fine and smooth, however, there were some beads, which are complexes of gelatin, chitosan and plasmid DNA on the nanofibrous, and the nanofibrous with beads overlapped layer by layer, and the diameter of nanofibrous seems thinner than the nanofibrous without plasmid (Fig. 4A). Interestingly, the elastic strength of both nanofibrous has no significant difference (Fig. 4C). The in vitro release profile of plasmid DNA from nanofibrous showed that plasmid DNA was released from nanofibrous quickly in the first 8 days, nearly 75% plasmid was released. After two weeks, plasmid began to release slowly, and nearly 90% plasmid was released from electrospun nanofibrous at that time (Fig. 4B). ADSCs can spread and grow very well on both electrospun gelatin/chitosan nanofibrous and nanofibrous with plasmid, and VEGF proteins (green fluorescence) in ADSCs were shown on nanofibrous with plasmid ( Fig. 4D), after 3 days proliferation, it seems that plasmid pCMV6-AC-VEGF has no obvious effect on ADSC growing on nanofibrous (Fig. 4E, 4F).
Wound closure was affected by cell sheets constructed by nanofibrous with plasmid Wound closure was assessed by macroscopic observation along the time of implantation and by planimetric analysis of the wound area. There was no significant difference in wound closures between the nanofibrous with or without plasmid groups and the control group (Fig. 5). However, a significant difference in wound area (p 0.05) was observed from day 3 post-transplantation in both cell sheets groups, compared to the control and nanofibrous groups (Fig. 6A). It seemed that the presence of the electrospun nanofibrous did not affect wound closure significantly, and the plasmid in the nanofibrous seems has little help to wound healing compare to control group, while the nanofibrous with or without plasmid together with ADSCs synergized to promote wound closure. At day 13, wound transplanted cell sheets with or without plasmid were both healed, while after transplanted 7:3 gelatin/chitosan nanofibrous or nanofibrous with plasmid, the wound was closed at 17 days and 16 days separately, and the wound without transplantation was closed till day 19.
The mechanical property of neo-skin in the cell sheets and cell sheets with plasmid group were similar to that of normal skin, while that of the neo-skin in the control group was significantly lower compared to normal skin (Fig. 6B). To examine the wound healing effect, the thickness of neo-skins at the day of healing was determined considering both the epidermal and dermal layers in H&E histological sections (Fig. 6C, 6D). Masson's trichrome staining carried out at the same time shows the nascent collagen (blue staining) in the regenerating skin. Quantitative analysis demonstrated that the overall thickness of neo-skin in the cell sheets and cell sheets with plasmid group was significantly thicker than that in the other groups ( Fig., 6C), and the matrix density has increased significantly with amount of collagen (Fig. 6E). Interestingly, the presence of nanofibrous with or without plasmid slightly increased the thickness and collagen deposition of the neo-skin, compared to the control, and the plasmid in the nanofibrous alone or in the cell sheets can increase the skin thickness and collagen deposition, then improve the wound healing.

Discussion
In this study, we demonstrated a simple and effective technique to construct multi-layer cell sheets using a natural electrospun gelatin/chitosan nanofibrous and ADSCs for accelerating wound healing. The architecture of electrospun nanofibrous made from natural biodegradable and biocompatible material (gelatin and chitosan) was similar to real tissue, which can satisfy physiological requirements. However, the strength properties of electro spun chitosan fibers are weak [19], and gelatin is sensitive to degradation [20,21]. In this work, the blending of chitosan with gelatin can improve the mechanical properties of chitosan fibers, what's more, chitosan is a hard electrospinning biomaterial, and gelatin can increase the spinnability of chitosan. The blend electro spun method also creates a new component without the need for synthesizing a new co-polymer. Therefore, we combined gelatin with different ratios of chitosan to fabricate electrospun gelatin/chitosan nanofibrous.
The advantages of the electrospinning technique include the production of very thin nanofibrous membranes with large specific growth surface, which allows for cell proliferation and functionalization. It is reported that both fiber diameter and pore size of electrospun nanofibrous membrane affect cell proliferation and differentiation [22]. The morphology of gelatin/chitosan nanofibrous clearly demonstrates that gelatin 11 concentration was very important in obtaining fine and large gelatin/chitosan fibers, more gelatin in membrane exhibited much higher mechanical strength than the composite containing more chitosan. The mechanical property of skin is important to its structure, appearance and functionality [23], so the mechanical strength of electrospun nanofibrous is also important for the cell sheets construction and wound healing. However, electrospun nanofibrous of natural biomaterials usually possess inferior mechanical properties and fast degradation rates, crosslinking is a common approach for improving the mechanical and degradation properties of natural biomaterials for tissue engineering applications [24]. A crosslinking treatment of gelatin/chitosan nanofibrous with glutaraldehyde effectively optimizes mechanical properties and modifies biodegradation rate. Our crosslinked gelatin/chitosan (7:3) electrospun nanofibrous showed the highest mechanical strength because this fiber was uniform and had medium-size in diameter, both of which are important factors dictating the mechanical properties of electro spinning [25].
The property of nanofibrous biodegradation is beneficial for skin regeneration as it can support and regulate skin regeneration [20]. Gelatin absorbed an amount of water and made the nanofibrous more hydrophilic and sensitive to degradation, while more chitosan exhibited slower degradation, which may inhibit wound healing [26]. So the crosslinked electrospun gelatin/chitosan (7:3) nanofibrous with appropriate biodegradation may be more suitable for skin regeneration. Additionally, the degradation products of gelatin and chitosan are relatively non-toxic small molecules, which can easily be excreted directly or after entry and exit from various metabolic pathways [20,27].
Natural biomaterials usually possess better biocompatibility and biofunctionality, we indicated that electrospun gelatin/chitosan nanofibrous had good biocompatibility with adipose-derived stem cells (ADSCs). ADSC is a promising adult stem cell for clinical therapy, because it can be easily isolated from adipose tissue, and has strong proliferation and differentiation ability [28]. The electrospun gelatin/chitosan nanofibrous (7:3) with appropriate mechanical strength and biodegradation can accelerate more ADSC adhesion and proliferation. Studies indicated that culture surface with higher stiffness favored cell adhesion and spreading [29,30], and hydrophilic surfaces favored BMSC spreading [31], which are consistent with our observation. The electrospun gelatin/chitosan nanofibrous was shown to not be toxic at all compositions.
The electrospun gelatin/chitosan nanofibrous are positively charged, which can be a promising nonviral vector for gene transfection. Negatively charged molecules can be ionically bound to positively charged free amino groups of chitosan, which can protect molecules from the destruction caused by enzymes or other harmful factors [18]. Some researchers combined cationized gelatin with plasmid DNA and impregnated the complex into a collagen/polyglycolic acid scaffold to enhance the formation of engineered bone tissue [32]. Endogenous angiogenic factors, vascular endothelial growth factor (VEGF), are naturally produced in response to tissue hypoxia and during wound healing [33]. Because of inadequate secretion of VEGF, this restorative process is insufficient to prevent tissue ischemia and necrosis, exogenous delivery of VEGF is necessary for wound healing. After impregnation of plasmid DNA VEGF into electrospun gelatin/chitosan nanofibrous (7:3), the positively charged gelatin and chitosan firmly immobilize negatively charged plasmid DNA, the release of plasmid DNA was unlikely to be simple diffusion through the surface of nanofibrous, it will be completely released only when the nanofibrous is degraded to generate water-soluble fragments. The electrostatic interaction between the negatively charged cell membrane and the positively charged polymer attaches the gelatin/chitosanplasmid complex to ADSCs, which promote the long-term expression of VEGF.
The electrospun gelatin/chitosan nanofibrous (7:3) with plasmid VEGF can promote ADSC proliferation and VEGF expression, so it could be a good supporting nanofibrous 13 membrane for sandwich-like cell sheets construction. Cell sheet was established with a confluence layer of cell populations on a temperature responsive surface [10,34]. Our previous study suggested that at least 5 days of culture was necessary for sufficient matrix deposition, and then the thick cell sheet was easily to be peeled off from culture surface with PVDF membrane [35]. Even so, it was hard to transplant intact cell sheets without PVDF membrane because they were too thin and fragile. Using biodegradable electrospun gelatin/chitosan nanofibrous as supporting membrane to replace of PVDF membrane, it was much easier to operate and form ADSC-nanofibrous -ADSC sandwichlike cell sheets, and transplant for wound healing.
A mouse dorsal wound healing model was used to investigate the skin regeneration potential of sandwich-like cell sheets. The whole wound healing process included cell adhesion, proliferation as well as migration, and eventually leads to skin regeneration [36]. Our in vivo skin defect models demonstrated that complete wound closure and reepithelialization was only observed after implanting sandwich-like cell sheets, and the time of wound healing was faster than the case of no transplantation. In contrast, the treatment with electrospun gelatin/chitosan nanofibrous only had no obvious beneficial effect compared to the no-transplantation group with respect to wound closure, and both of them lack of a complete epithelial layer. It may be reasoned that the ADSC adding to electrospun gelatin/chitosan nanofibrous promoted its biomechanical features, and provided a better wound healing environment. So both ADSCs and nanofibrous were the key elements for wound healing and skin regeneration. Although wound closed almost at the same time after transplantation of cell sheets and cell sheets with plasmid VEGF, the wound skin transplanting of cell sheets with plasmid had thicker and complete epithelial layer than that of cell sheets only, which indicate that VEGF in the nanofibrous can promote wound healing and skin regeneration. Of course, it will have better repairing 14 effects if we differentiated ADSCs into fibroblasts and epidermal cells before seeding on the nanofibrous, it will be our next work.

Conclusions
In this study, we successfully fabricated electrospun gelatin/chitosan nanofibrous with different proportions. The gelatin/chitosan (7:3)  Electrospinning was done at a voltage of 8 kV to create micro/nanofibers with a needle having an inner diameter of 20 gauge and a 0.6 mL/h feeding rate of solution using a syringe pump. The cellophane collector plate was placed 12 cm away from the tip of the needle. Upon the completion of the electrospinning, the electrospun nanofibrous were removed from the collector and cross-linked with glutaraldehyde for 1 hour.
For the impregnation of plasmid DNA into electrospun nanofibrous, 2 mg plasmid DNA pCMV6-AC-VEGF (500 µg/µL, vector pCMV6-AC-GFP was bought from ORIGENE, Beijing) were gently mixed with gelatin/chitosan electrospun solution, electrospinning was done as previously described. The charge ratio (N:P) that is indicated as the molar ratio of the free amino groups of electrospun gelatin/chitosan nanofibrous to the phosphate molecules of plasmid DNA is 5.0, according to our previous results, to obtain the highest transfection efficiency [18]. The electrospun nanofibrous with plasmid also need to be cross-linked with glutaraldehyde to avoid fast-degrading.

Characterization of electrospun nanofibrous
The morphology of the electrospun nanofibrous was studied by a scanning electron Where M is the weight of each sample after immersion in the PBS for 24 h, Md is the weight of the sample in its dry state. Only cross-linked nanofibrous was examined.
The nanofibrous were collected using cellophane, and then the thickness of each sample was measured by SEM. These samples were measured, cut carefully and glued on the frame, and then the cellophane was removed. The paper frame was cut before initiating the tensile strength measurements. The tensile properties of these samples were Ethical and Welfare Committee of Shenzhen University (SYXK: 2018 − 0140). All efforts were made to minimize suffering and numbers of mice used. Specific pathogen free kunming mice, were purchased from Guangdong Medical Laboratory Animal Center, weighting 32 ± 2 g, were anesthetized in a chamber of 2% isoflurane (Jinan Shengqi pharmaceutical Co, China) with 40% oxygen and 60% nitrous oxide. After shaving, one excision of 2 cm diameter round shape was performed in the dorsum of each animal.

Statistical analysis
The results are shown as mean ± SEM. All experiments were independently repeated for at least 3 times unless otherwise stated. ANOVA with a post hoc Dunn or Bonferroni test was 20 used to analyze the data. p < 0.05 was considered to be significant unless otherwise

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request..