Introduction

Traumatic wounds are often caused by fire, chemicals or diseases (Zhou et al. 2008). Clinically, skin substitutes are used as covers to prevent excessive fluid loss from open wounds while allowing oxygen transport, as well as creating a moist environment for promoting wound epithelialization. Such skin substitutes must also meet the requirement of biocompatibility. Nanofibers based on natural polymers have been attracting a great deal of interest as promising skin scaffolds thanks to their high porosity, large surface to volume ratio (Ding et al. 2014; Rieger et al. 2013) and low immunogenicity (Lee et al. 2009), . Electrospinning, a simple and multifarious technique, is commonly applied to fabricate nanofibers as potential skin substitutes from natural polymers including chitosan (Antunes et al. 2015), alginate (Leung et al. 2014), silk fibroin (Song et al. 2016), gelatin (Li et al. 2016a, b) and hyaluronic acid (Uppal et al. 2011). Chitosan (CS), is a derivative of chitin, which is the second most abundant polysaccharide in nature, next to cellulose (Tang et al. 2016). It is regarded as one such polymer that fulfils the requirement for a skin substitute, owing to its antimicrobial character, biocompatibility and biodegradability (Li et al. 2016a, b). In this capacity, chitosan can help wounds heal faster and prevent scarring (Anjum et al. 2016).

In the last few years, chitosan-based nanofibers have been successfully prepared by electrospinning of homogeneous chitosan (Ohkawa et al. 2004) or chitosan solutions blended with poly (ethylene oxide) (PEO) (Bhattarai et al. 2005), poly (vinyl alcohol) (PVA) (Zhou et al. 2006), or silk fibroin (Zhang et al. 2011). However, some toxic solvents such as trifluoroacetic acid, acrylic acid, 1,1,1,3,3,3-hexafluoro-2-propanol, and dimethyl sulfoxide have usually been involved in the preparation of these chitosan-based nanofibers, which limited their applications for wounded skin. To reduce the toxicity while enhancing the biocompatibility, much effort has been devoted to electrospinning water-soluble chitosan derivatives (Alipour et al. 2009; Zarandi et al. 2015; Zhou et al. 2008) into nanofibers. However, there is a major drawback with these water-soluble fibrous mats: rapid hydrolysis makes it difficult to for the mat to remain intact in the physiological environment, rendering water-soluble mats unfit for use as tissue engineering scaffolds. To prevent rapid hydrolysis, it is necessary to chemically crosslink these water-soluble fibers. Additionally, although electrospun nanofibers have fibrous porous structures which can mimic the size scale of the native extracellular matrix, one limitation for the nanofiber scaffolds is a relatively small pore size, and the resultant difficulty for cell infiltration, which retards tissue remodeling (Kurpinski et al. 2010; Sundararaghavan et al. 2010).

To address these issues, we prepared photocrosslinkable methacrylated chitosan-based nanofibrous scaffolds with micrometer size pores by PVA-aiding electrospinning, then subsequently extracted PVA from the nanofibers. Here, water was used as the solvent for MACS, which is ideal for medical applications thanks to the absence of toxic residues. Photocrosslinked MACS-based mats have micrometer size pores that can favor cell infiltration to form three-dimensional skin tissues. In addition, their nanofibrillar structures resemble the natural extracellular matrix, which favors the attachment of cells (Bhattarai et al. 2005). Finally, high specific surface area and porosity can guarantee gaseous exchange and removal of excess exudates, which also encourages wound healing (Jayakumar et al. 2011). We also characterized morphology and thermal properties. The in vitro cytotoxicity of photocrosslinked MACS-based mats iswas evaluated with L929 cells. Cellular infiltration was evaluated with DNA plasmid encoding green fluorescent protein transfected Human Embryonic Kidney 293T (HEK 293T) cells used as model cells.

Experimental

Materials

Chitosan (Mw = 1.0 × 105 Da) was purchased from Zhejiang Golden-Shell Biochemical Co., Ltd., China. The deacetylation degree of chitosan was determined to be 67.2% by 1H NMR (Hirai et al. 1991). Poly (vinyl alcohol) (1.7 × 105 Da, 88% hydrolyzed) was obtained from Kuraray Co, Ltd., Japan. 2-Hydroxyethyl methacrylate (HEMA) was obtained from Tianjin Institute of Chemical Reagents. Acryloyl chloride was obtained from Shanghai Chemical Reagents Company, China. Darocur 2959 (D-2959, 2-hydroxy-1-[4-(hydroxyethoxy) phenyl]-2-methyl-1-propanone) was donated from IGM Resins B.V. (Netherlands). Mouse fibroblasts (L929) were purchased from Wuhan Beinglay Biological Technology Co., Ltd., China. DNA plasmid encoding green fluorescent protein transfected Human Embryonic Kidney 293T (HEK 293T) cells were kindly provided by the Institute of Hydrobiology, Chinese Academy of Sciences.

Synthesis of methacrylated chitosan (MACS)

Water-soluble methacrylated chitosan was prepared as described in our previous report (Gao et al. 2010). Briefly, 83.81 g HEMA, 105 mL triethylamine and 9.79 g acryloyl chloride were mixed in 300 mL toluene and reacted under a temperature of 0–5 °C for 12 h. Then, the mixture was filtered and extracted with water, 1 mol L−1 HCl and 1 mol L−1 NaHCO3 solution, respectively. After that, ethylene glycol acrylate methacrylate (EGAMA, yellow liquid) was obtained by removing the toluene.

Then, 2.0 g chitosan, 0.69 mL acetic acid and 4.0 g EGAMA were mixed in water/ethanol (5:4) solution (90 mL) and reacted for 2 days at 60 °C. After that, NaHCO3 was added to the mixture solution, which then was dialyzed against water and lyophilized to get pure MACS. 1H NMR spectroscopy was recorded in CD3COOD/D2O as solvents on a Bruker AV 400 NMR instrument.

Preparation of MACS/PVA solutions

We prepared 15% (w/v) MACS solution and 10% (w/v) PVA solution. The MACS solution was combined with the PVA solution at MACS/PVA mass ratios of 100/0, 75/25, 50/50, 25/75, and 0/100.

The conductivity of the MACS/PVA solution was evaluated by electrical conductivity meter (DDB-6200, Shanghai Rex Xinjing Instrument Co. Ltd., China), and the viscosity of the MACS/PVA solution was examined as shear rate of ~400 s−1 on an AR 2000ex rheometer (TA Instrument, USA) by using a steel parallel plate geometry with a diameter of 40 mm and a 1 mm gap at 25 °C.

Preparation of electrospun membranes

The electrospinning set-up was composed of a 20 mL plastic syringe with a stainless needle (0.57 mm inner diameter), grounded aluminum foil as the collector and a high voltage–power supply (EST705, Beijing Huajinghui Technology Co., Ltd., China). A voltage of 25 kV dc and 12 cm distance was applied between the needle and the plate; the solution flow rate was 1.0 mL/h.

Photocrosslinking of the nanofibers

For further photopolymerization, the MACS/PVA solution at MACS/PVA mass ratio of 50/50 was prepared containing cytocompatible UV photoinitiator D-2959 (0.1 w/v %) (Williams et al. 2005). The nanofiber mats were obtained by the aforementioned electrospinning process and irradiated directly under a 250 W Hg lamp with a light intensity of 30 mW/cm2 for 30 min. Upon UV irradiation, the photoinitiator in the mats initiated further crosslinking of MACS nanofiber mats. The photocrosslinked nanofibrous mats were collected and then dried at 25 °C in z vacuum for 12 h. After the photocrosslinking process, the electrospun nanofibrous scaffold was soaked in purified water for 48 h at 37 °C, refreshing water daily to remove PVA. The mats were lyophilized and the morphology was observed by SEM. These lyophilized mats (photocrosslinked MACS-based nanofibers) were used for subsequent MTT evaluation and cell culture study.

Characterizations

Morphology of the nanofibers

A scanning electron microscope (JSM-6510, JEOL Ltd., Japan) was used to evaluate the basic characteristics of the nanofibers, such as fiber morphology and diameter. The accelerated voltage was 10 kV. The diameters of nanofibers in SEM Images were measured using an image analyzer (Image J, version 1.37 v, National Institutes of Health, USA).

Thermal analysis of the nanofibers

Differential Scanning Calorimetry (DSC822e, Mettler Toledo, Switzerland) was used to investigate thermal characterization of the electrospun fibers. Samples were heated from −50 to 250 °C at a 10 °C/min heating rate under nitrogen flow of 50 mL/min.

[3-(4,5-dimethylthiazol-2-yl)-2,5-d diphenyltetrazolium bromide] (MTT) assays

To measure the cytotoxicity of nanofibers toward mouse fibroblasts, the sterilized photocrosslinked MACS-based nanofibers were incubated in culture medium at an extraction ratio of 6.0 cm2/mL for 24 h at 37 °C, after which the nanofibers were removed and the extraction medium was obtained.

200 μL of L929 cell suspension was seeded in 96–well plate at a concentration of 105 cells/well. After incubation at 37 °C for 1 day, the culture medium was used instead of the extraction medium. After 1 day, the extract was removed and 10 μL of MTT solution was added to each well. L929 cells were allowed to incubate at 37 °C (5% CO2) for 4 h, and then the formazan reaction products were dissolved in dimethyl sulfoxide and the plates were shaken for 10 min. The optical density of the formazan solution was read on an ELISA reader at 490 nm.

Dulbecco’s modified eagle medium (DMEM) containing 10% fetal bovine serum (FBS) was used as negative control for toxicity.

The resulting values were expressed as cell viability (mean value ± standard deviation). The comparison between the values was made through ANOVA one-way analysis, with P < 0.05 considered significant. Analysis was performed using Origin software (7.0 SRO, Northampton, MA, USA).

Cell cultures

HEK 293T cells were cultured in growth medium containing DMEM, 10% fetal bovine serum, 100 U/mL of penicillin, 100 mg/mL of streptomycin and 2 mM/L l-glutamine. HEK 293T cells were incubated at 37 °C in a humidified atmosphere in the presence of 5% CO2. Photocrosslinked MACS-based mats (a thickness of ~0.2 mm) were fixed on cover slips, sterilized and then extensively washed with sterile PBS prior to transfer to 24-well plates. HEK 293T cells suspension with 4.0 × 104 cells/mL were seeded on the mats. The cells were then incubated for 48 h, and investigation of the cells was conducted using a fluorescence microscope. Cells on tissue culture plastic were used as a control group.

Results and discussion

In this research, unsaturated methacrylate groups were introduced onto the chitosan backbones as side groups to prepared nanofiber precursors for the subsequent UV crosslinking process. The synthesis of MACS was achieved by the Michael addition reaction between –NH2 of CS and the acrylate group of EGAMA. 1HNMR of MACS and CS are illustrated in Fig. 1. Methacrylation by the Michael addition is confirmed by 1H NMR, as indicated by the appearance of proton signals of the double bond of the methacrylate group at δ 5.7 ppm and δ 6.1 ppm, as well as methyl groups of methacrylate at δ 1.9 ppm. In addition, N-methacrylation can be identified as a reduction in peak area at δ 3.1 ppm (H-2 of GlcN) in the 1H NMR spectrum. Based on the assignment of the 1H NMR spectrum, the degree of substitution of MACS is calculated as DS = 0.328 × I1.9/I2.0, where I1.9 is the integral of the methyl protons at δ 1.9 ppm, and I2.0 is the integral of the methyl protons of –NHCOCH3 at δ 2.0 ppm. The calculated DS is 0.11, which means that 16.4% of the –NH2 group is substituted by methacrylate group. Other peaks were ascribed as follows. 4.38 (e, d, –COO–CH2–CH2OCO–), 3.7–4.0 (H-3,4,5,6 of GlcN and GlcNAc), 3.4 (g, N–CH2– of N-alkyl group), 3.2 (H-2 of GlcNAc), 2.9 (g, H-2 of alkylated GlcN), 2.6 (f, –CH2COO–), 2.0 (h, –NHCOCH3).

Fig. 1
figure 1

1HNMR spectra of MACS in D2O and CS in CD3COOD/D2O

Full size image

Electrospinning from pure MACS aqueous solution is highly desirable because it generates pure MACS naonofibers in an environmentally friendly manner. However, only micro- and nanoparticles instead of ultrafine fibers were observed (Fig. 2a), when pure MACS aqueous solution was subjected to electric field force regardless of viscosity or conductivity of the MACS solution. This can be ascribed to two aspects of the difficulty of electrospinning chitosan. First, chitosan chains are closely overlapped and cannot form effective chain entanglements, due to its polycationic nature and semi-rigid chain structure (Desai et al. 2008; Pakravan et al. 2011). Second, chitosan solution displays high electrical conductivities, unacceptably large for electrospinning, which requires jets to experience deep atomization to break into polydisperse eletrosprays (Subramanian et al. 2005). So, an effort was made to enhance the chain entanglements in the chitosan solution and to decrease the conductivity of the chitosan solution. A suitable candidate as a non-ionogenic partner for electrospinning of MACS is PVA, which has the ability to produce nanofibers because of its viscoelastic properties (Rathna et al. 2011). Figure 2 shows SEM images of MACS/PVA nanofibrous mats with different mass ratios. It is clear that the morphology depends on the content of PVA. As depicted in Fig. 2b, a mixture of thin fibers and beads were observed when the MACS/PVA weight ratio was 75/25. However, when MACS content was less than 50%, the MACS/PVA nanofibers exhibited a more regular morphology and a smooth surface, which made the electrospinning relatively easy.

Fig. 2
figure 2

Effect of the composition of spinning solution on the morphology of MACS/PVA nanofibers. MACS/PVA weight ratio: a 100/0; b 75/25; c 50/50; d 25/75; e 0/100; f 50/50, after UV irradiation

Full size image

In many applications, the effect of the average electrospun fiber diameter and its uniformity on the nanofibers’ performance is important. Figure 3 presents the diameter distribution of MACS/PVA nanofibers. By statistical calculation, the average diameter of MACS/PVA nanofiber progressively increased from 162 to 476 nm when MACS content gradually fell from 75/25 to 25/75, and fiber diameter distribution became increasingly wider.

Fig. 3
figure 3

Diameter distribution of nanofibers for different compositions. MACS/PVA weight ratio: a 75/25; b 50/50; c 25/75; d 0/100; e 50/50, after UV irradiation

Full size image

It is well documented that the structure and morphology of electrospun fibers are influenced by solution properties such as polymer content, viscosity and conductivity (Bhattarai et al. 2005). The viscosity as a function of the mass ratio of MACS/PVA in the mixed solutions is shown in Fig. 4. As the figure shows, the viscosity of the MACS/PVA solution obviously increased from 0.60 to 1.14 Pa S with the increase of PVA content; the increased viscosity can be ascribed to the increased interactions and entanglements between polymer chains as a result of increasing concentration of PVA. Correspondingly, as shown in Fig. 2, a transformation from beaded fibers to uniform fibers took place as the PVA increased. When the MACS/PVA weight ratio was high and the viscosity of the blend solution was thus lowered below the limit, the polymer chain entanglements were not enough to form the fiber and beaded fibers were often produced. However when the viscosity of the blend solutions increased due to an increase of PVA content, gradual increase of intra/inter molecular chain entanglement in MACS/PVA solution favored the electrospinnability of the MACS. Similar results have also been reported by others (Li and Hsieh 2006). At the same time, improved of viscosity of the MACS/PVA solution may also contribute to an increase of fiber diameter. The conductivity of MACS/PVA solutions is also presented in Fig. 4. The conductivity of the MACS/PVA blend solutions rose from 1.13 to 2.91 mS/cm as MACS content increased. Because MACS is a polycation, while PVA is a non-ionogenic polymer, increased MACS content could result in an increase of solution conductivity. As mentioned above, high electrical conductivity was unacceptable for electrospinning of the chitosan solution, so when PVA content increased, the conductivity of the MACS/PVA solution was gradually reduced to favor the electrospinning of MACS. On the other hand, viscosity of MACS/PVA solution was improved to enhance chain entanglements that occurred between the MACS and PVA. In this way, fine fibers were obtained. The studies based on MACS/PVA blends indicated that it is essential for polymer solutions to have the desired viscosity and conductivity to form smooth nanofibers. The increase in average diameter and broader distribution with decreased MACS content may be due to the increase of viscosity and lower conductivity of the MACS/PVA blend solution. Both effects resulted in lower stretching rate and subsequently thicker fibers.

Fig. 4
figure 4

Conductivity and viscosity of the MACS/PVA blend aqueous solutions

Full size image

The SEM micrographs of fibrous mats after UV irradiation are also illustrated in Fig. 2. As expected, distinct morphological changes were not observed in the electrospun mats after irradiation.

DSC results of MACS, PVA and spun fibers can be found in Fig. 5. In DSC measurements, the second heating run was used. As observed, glass transition temperature (T g) of the MACS was not detected, possibly owing to different characteristics of natural polysaccharides, such as crystallinity, molecular weight and deacetylation degree, from different sources or extraction methods (Neto et al. 2005). MACS/PVA nanofibers displayed Tg around 70 °C and the T g slightly increased from 71 to 77 °CC as the MACS/PVA weight ratio increased from 0/100 to 50/50. This was possibly due to intermolecular interactions via hydrogen bonds between MACS and PVA limiting movement of the molecular chain segment. In Fig. 5, the PVA presents a melting peak (T m) at 150 °C. However, for MACS/PVA blends, the endothermic curve broadened and shifted to the lower temperature as MACS content increased. The X-ray diffraction results (see the X-ray diffraction analysis section in the supporting information file) indicated that the introduction of MACS decreased the crystallinity of MACS/PVA blends, which can be attributed to the fact that ordered chain associations of PVA were hindered by the hydrogen bonding interaction between MACS and PVA during electrospinning process (Cay et al. 2014). However, after photopolymerization, MACS/PVA (50/50) nanofibers showed weak Tm peaks around 140 °C, indicating that photocrosslinking could efficiently increase the crystallinity of PVA electrospun nanofibers by increasing hydrogen bonding interactions between the chains and forming a more stable organizing structure.

Fig. 5
figure 5

DSC curves of MACS, PVA, and the nanofibrous membranes with different MACS/PVA weight ratios. a 100/0; b 50/50; c 25/75; d 0/100; e 50/50 after UV irradiation

Full size image

Pore size and porosity of electrospun fibrous structures play an important role in cell attachment, proliferation and migration for tissue engineering. Although electrospun mats have nanofibrous structures similar to the natural extracellular matrix and can support cell attachment and growth, poor cellular infiltration has been an ongoing problem owing to the mats having a pore sizesmaller than typical cellular diameters, limiting cell migration within the nanofibrous scaffold. This can lead to a situation in which the scaffold surface is always covered by a membrane of cells while cells are seldom observed within the scaffold, a circumstance which is not suitable for tissue engineering of three-dimensional tissues (Tuzlakoglu et al. 2011). To obtain micro-pore mats, PVA was chosen as a template polymer to aid electrospinning of MACS to get nanofibers due to its excellent electrospinnability and biocompatibility (Jannesari et al. 2011; Millon et al. 2008). The PVA could be easily removed by immersion in water. Here, photocrosslinked MACS/PVA nanofibers were prepared and then incubated in distilled water at 37 °C for 48 h to remove the PVA.

According to previous work (Liu et al. 2009), the as-spun pure PVA nanofibers were dissolved immediately when immersed in an aqueous environment, owing to the water-solubility of the PVA nanofibers. Here, the vinyl groups of the MACS molecular chain were crosslinked by UV irradiation to form water-insoluble MACS networks, while PVA without crosslinking could dissolve in water to be removed from fibers. ATR-FTIR results showed that nearly all PVA was removed after the soaking process (shown in the supporting information). Figure 6 presents the SEM images of the nanofibers with and without UV polymerization after water immersion at 37 °C for 48 h. It can be seen that uncrosslinked MACS/PVA nanofiber (50/50) dissolved in water and lost its nanofibrous structure (Fig. 6a). However, photocrosslinked MACS-based mats kept their micro-pore structure with many fibrils (Fig. 6b-c), and the micrometer-size pores in the electrospun mats were clearly visible.

Fig. 6
figure 6

SEM micrographs of the nanofibers (MACS/PVA = 50/50) with/without UV irradiation after water immersion at 37 °C. a Without UV irradiation; b with UV irradiation, ×2000 magnification; c with UV irradiation, ×15,000 magnification

Full size image

The photocrosslinked MACS-based mats had a tensile strength of 2.36 ± 0.11 MPa and water-uptake of 6.41 ± 0.65 g/g (shown in the supporting information).

In vitro cytotoxic tests were carried out as an initial verification of the biocompatibility for biomedical materials, with an eye to future applications of biomaterials in tissue engineering. Here, in vitro cytotoxicity of photocrosslinked MACS-based nanofibers was evaluated with L929 cells, as shown in Fig. 7. There was no statistically significant difference (p < 0.05) between photopolymerized MACS-based mats and the negative control group when diluted extract concentration at 50% or below 50% was adopted. Although statistical analysis indicated a significant difference (p < 0.05) between cytotoxicity of photopolymerized MACS-based mats and the negative control group when 100% extract was used, cell viability could still be more than 90% of control group. According to International Standard ISO 10993-5 (2009), reduction of cell viability by more than 30% is considered a cytotoxic effect. So, the photopolymerized MACS-based mats were not cytotoxic to L929 cells, which suggests they were biocompatible and suitable for biomedical applications.

Fig. 7
figure 7

Plots of cell viability obtained from photocrosslinked MACS-based nanofibers and negative control, respectively

Full size image

Our previous work revealed good adherence and proliferation of the L929 cells on the surface of mats (Zhou et al. 2008). However, cellular infiltration into the nanofibers was constrained due to small pore sizes, which led to a non-uniform cellular distribution among the nanofibers, limiting vascularization and tissue growth (Khorshidi et al. 2016). Figure 8 shows fluorescent images of HEK 293T cells cultured under different conditions. The images show HEK 293T cells at various penetration levels with in-focus cells and out-of-focus cells, indicating various depths throughout the mats (Fig. 8b). The results indicated that the highly micro-porous structure of MACS-based mats allowed for cellular migration into the 3D matrix, which could play a key role in scaffolds for tissue engineering and 3D cell culture (Wingate et al. 2012).

Fig. 8
figure 8

Fluorescent images of HEK 293T cells cultured under different conditions. a Negative control (without nanofibers); b photocrosslinked MACS-based mats

Full size image

Conclusions

In the present study, photocrosslinkable nanofibrous scaffolds based on MACS were successfully prepared by photocrosslinking electrospun MACS/PVA mats, then subsequently removing PVA from the nanofibers. SEM images showed that MACS/PVA nanofibrous mats had smaller diameters and narrower diameter distributions as the MACS/PVA weight ratio decreased. DSC results demonstrated the hydrogen bonding interaction between MACS and PVA molecular chains. The photopolymerized MACS-based nanofibrous scaffolds had micro-porous structures with many fibrils after removing PVA with water. The results of in vitro cytotoxicity assessment suggested photocrosslinked MACS-based mats were biocompatible. More interesting, cell culture results indicated that the highly micro-porous structure of MACS-based mats allowed for cellular migration into the 3D matrix. The photocrosslinked MACS-based mats thus have potential applications as skin replacement materials.