Bombyx mori silk was purchased from Jiangsu and Zhejiang (China). Dulbecco’s modified Eagle’s medium and fetal bovine serum were obtained from Gibco (United States). Forskolin, heregulin, cytosine arabinoside, and 488-labeled goat anti-mouse IgG were obtained from Sigma-Aldrich (United States). Rabbit anti-S100 beta monoclonal antibody was obtained from Abcam. CCK-8 kit was purchased from Ribobio (Guangzhou, China).

Fresh silkworm raw silk from different sources (Jiangsu and Zhejiang) was weighed and put into a sodium carbonate solution with different concentrations (0.05 and 0.5 wt%) in proportion. The silk was boiled three times in Na₂CO₃ solution (half an hour each time). The degummed silk was thoroughly rinsed with Millipore water and air-dried on a super clean platform to obtain refined silk, that is, silk fibroin sample for standby. There are 4 samples in total: the raw silk from Jiangsu degummed with 0.05 and 0.5% Na₂CO₃ are sample 1 and sample 2, respectively, while the raw silk from Zhejiang degummed with 0.05 and 0.5% Na₂CO₃ are sample 3 and sample 4, respectively. The tussah silk degummed with 0.5% Na₂CO₃ is labeled as sample 5.

The refined silk of silkworm was dissolved in a ternary solvent system of CaCl₂/H₂O/C₂H₅OH (mole ratio, 1:8:2) at 75 ± 2°C and then dialyzed against Millipore water in a cellulose tube (molecular cutoff = 12,000–14,000) at room temperature for 3 days.

Picric acid–carmine staining was used to evaluate the degumming degree of silk degummed with two kinds of Na₂CO₃ solution (Zhang et al., 2014). First, in preparing the staining solution, carmine was dissolved in 25% ammonia, followed by adding saturated picric acid aqueous solution and adjusting the pH to 8.0–9.0. Then, the refined bombyx silk samples were immersed in the staining solution in test tubes, and the tubes were heated in boiling water bath for 5 min. Lastly, the samples were thoroughly rinsed with ddH₂O and air-dried. Raw silk was taken as the control.

The refined samples were dissolved in a tertiary solvent system of CaCl₂/H₂O/C₂H₅OH (mole ratio, 1:8:2) and then kept in water bath at 20°C for 2 h. Apparent viscosity was then measured with NDJ-7 rotary viscometer.

Samples under different degumming treatments were fixed in pre-cooled 2.5% glutaraldehyde, post-fixed in 1% osmium acid, dehydrated with gradient ethanol, embedded in EPON 812 epoxy resin, and cut into slices of 70-nm thickness. The ultra-thin sections were stained with lead citrate and uranium acetate, followed by observation under a transmission electron microscope (JEOL Ltd., Tokyo, Japan). In Photoshop 7.0, the fiber diameter was measured by measuring tools. The diameter of at least 60 fibers was measured from 10 photos taken in different fields randomly. The average value and standard deviation of the fiber diameter were calculated.

In this study, all experimental procedures involving animals were conducted as per the institutional animal care guidelines and approved ethically by the administration committee of experimental animals, Jiangsu Province, China.

Schwann cells (SCs) were harvested as described previously (Chen et al., 2017). Briefly, the sciatic nerves and dorsal root ganglia were isolated from neonatal Sprague–Dawley rats (1 to 2 days) to get primary rat Schwann cells. The obtained tissues were triturated and enzymatically digested with 0.25% trypsin at 37°C for 30 min. Then, the mixture was centrifuged and re-suspended in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), followed by plating on poly-l-lysine pre-coated dishes. After incubation for 24 h, cytosine arabinoside was added to allow cell incubation for another 24 h to remove fibroblasts. Next, the cells were cultured in DMEM supplemented with 10% FBS, 2 mM forskolin, and 2 ng/ml heregulin to stimulate cell proliferation. When cells covered 90% of the dish surface, they were further purified with anti-Thy1 antibody (1:1,000, AbD Serotec, Raleigh, NC, United States) and complement (Jackson Immuno, West Grove, PA). Lastly, the purified SCs were cultured in DMEM with FBS and growth factor until the cells were sufficient to seed on the SF membrane.

The prepared SF membranes were sterilized with 75% alcohol for 30 min and rinsed extensively with sterilized phosphate-buffered saline (PBS). Then, all samples were put in 24-well culture plates, and cell suspension was added in. The seeding cell density was 1 × 10⁵ cells/well. At different times of culture, the morphological changes of Schwann cells on these different SF membranes were observed under an inverted light microscope.

CCK-8 kit was used to evaluate the viability and proliferation of SCs on different samples after culturing for 1 and 3 days, respectively. At different times, fresh DMEM medium with CCK-8 reagent (V _medium / V _CCK-8 = 10:1) was added to replace the cell medium to allow cell incubation at 37°C for 4 h. Then, 150 μl of suspension was transferred to a 96-well plate. The absorbance was measured at 450 nm by an ElX-800 micro-ELISA reader (Bio-Tek Inc., United States).

The morphology of Schwann cells on different samples was examined by the immunostaining method. In brief, after 2 days of culture, cells were rinsed with PBS for three times thoroughly; then, they were fixed in 4% paraformaldehyde at 4°C for 4 h and stained with S100 (1:400) at 4°C for 24 h, followed by further reaction with IgG-488 (1:400) at 37°C for 2 h. Subsequently, the cells were incubated with Hoechst (final concentration: 5 μg/ml) at room temperature for 15 min. Finally, cell samples were observed under an immuno-fluorescence microscope (Leica, Germany).

The viability of Schwann cells was evaluated with Calcein-AM/PI Double Staining Kit (Invitrogen, L3224). Schwann cells were seeded on four different silk fibroin membranes and control plates at a density of 1 × 10⁵ cells/well. After the SCs were cultured for 1 and 3 days, DMEM with 10% FBS was discarded, and the samples were rinsed with PBS. Then, calcein-AM and propidium iodide (PI) were added, and the cells were incubated at 37°C for 30 min. Images were captured by an immuno-fluorescence microscope.

To find out the different proteins between samples and select the appropriate silk materials, we used isobaric tags for relative and absolute quantification (iTRAQ) to analyze proteins in different samples quantitatively (Zou et al., 2019). The phenol extraction method was used for protein extraction, which can effectively remove the small molecular interference in the sample. At the same time, filter-aided proteome preparation enzymolysis strategy was used for protein digestion. Then, the peptides of each sample were labeled with iTRAQ, and the labeled samples were graded by high-pH reversed-phase classification strategy. Finally, data collection of samples obtained by classification was carried out by the ultra-high-resolution mass spectrometer Q-Exactive.

The statistical significance was analyzed by GraphPad Prism 7.0 (GraphPad Software, Inc.). One-way ANOVA followed by Tukey’s post-hoc test was used to compare individuals among different groups of the same time. Two-way analysis of variance (two-way ANOVA), followed by Sidak’s multiple-comparisons test, was employed when comparing −ES and +ES groups of all groups. Data were presented as mean ± SD. p < 0.05 was considered statistically significant.

Two kinds of Na₂CO₃ solution (0.05 and 0.5 wt%) were chosen to degum silk from different sources. All silk from different sources in the two solutions were boiled for 0.5 h, and the test was repeated three times. The degumming degree of two kinds of Na₂CO₃ solutions on the raw silk from different sources was determined by the picric acid–carmine staining method. SF and sericin have different absorbance capacity on picric acid and carmine. Silk fibroin turns yellow in alkaline solution due to its selective adsorption of picric acid. However, sericin has a strong capacity to absorb both picric acid and carmine; red covers yellow, so it appears red. Therefore, after dyeing and washing, the yellow surface of raw silk indicates that sericin has been completely removed; otherwise, it indicates that sericin has not been completely removed. It can be seen from Figure 1 that, after degumming with the Na₂CO₃ solution, samples 1–4 are yellow (Figures 1A,B,D,E), indicating that sericin has been removed, while the control group without degumming can be clearly observed to be red (Figures 1C,F). With the increase of Na₂CO₃ solution concentration, the samples appear pure yellow (Figures 1B,E), showing that the degumming degree is higher.

The SF obtained by different degumming methods was dissolved in ternary solution in the same proportion, and its apparent viscosity was measured (Figure 1G). The viscosity of Zhejiang silk decreased from 630 to 78 CP, while that of Jiangsu silk changed from 450 to 75 CP as the concentration of Na₂CO₃ changed from 0.05 to 0.5%, indicating that the high concentration of Na₂CO₃ can catalyze the degradation or hydrolysis of SF. Under the catalysis of the high concentration of Na₂CO₃, the molecular chain of SF becomes shorter, the number of entangled nodes in the solution decreases, and the friction resistance between molecules decreases, which leads to the decrease of viscosity. Therefore, according to the viscosity change, the concentration of Na₂CO₃ should be reduced as much as possible to reduce the damage to the viscosity of SF. At the same time, we also found that the viscosity of the SF solution of Zhejiang silkworm was higher than that of Jiangsu silkworm under two kinds of Na₂CO₃ concentration.

Silk from different sources was degummed with different concentrations of Na₂CO₃ solution. After degumming with 0.05% Na₂CO₃, the cross-section of the SF sample demonstrates a homogeneous oval with complete structure, smooth edge, and no attachment on the surface, indicating that 0.05% Na₂CO₃ can not only remove sericin on the surface but also maintain the structure of SF. While treated with 0.5% Na₂CO₃, the section of degummed SF presents oval of different sizes and a long fusiform shape. It appears wrinkled with incomplete edge, and the diameter gets smaller, demonstrating that the high concentration of Na₂CO₃ makes sericin completely removed, but it also destroys the structure of SF to some extent. Comparing the effect of source on the structure of SF, despite the concentration of the Na₂CO₃ solution, the cross-section diameter of SF obtained from Zhejiang silkworms was larger than that from Jiangsu silkworms, indicating that the structure of Zhejiang silkworms was relatively full (Figure 2).

Schwann cells play an important role in the formation of myelin sheath during peripheral nerve regeneration. Therefore, in this study, the adhesion, survival, and growth of Schwann cells on different silk fibroin membranes were studied to choose a suitable SF to prepare peripheral nerve grafts which can better promote peripheral nerve regeneration.

Figure 3 shows the morphology observation of SCs on different samples for 24 and 48 h, respectively. After 1 day of culture, there were certain amounts of cell on all samples, but the number of cells on JS-0.5 and ZJ-0.5 was more than that on JS-0.05 and ZJ-0.05, and some cells on JS-0.05 and ZJ-0.05 shrank into a circle, indicating that cells could grow on all samples, but samples JS-0.5 and ZJ-0.5 were more conducive to cell adhesion. It may be that a small amount of residual sericin in JS-0.05 and ZJ-0.05 affected the interaction between cells and samples, resulting in inferior number and the morphology of cells on JS-0.05 and ZJ-0.05. Two days later, the cell density on sample JS-0.5 and ZJ-0.5 increased, indicating good cell growth and proliferation, and the number of cells on samples JS-0.05 and ZJ-0.05 increased, too, but not as much as that of JS-0.5 and ZJ-0.5, indicating that JS-0.05 and ZJ-0.05 can also support cell growth and promote cell proliferation, but the cell growth was slower than that of JS-0.5 and ZJ-0.5. The results of immunofluorescence staining tells us that, on samples JS-0.5 and ZJ-0.5, SC cells exhibited better adhesion and spreading and displayed a spindle-like shape with filopodia at both ends, while on sample JS-0.05 and ZJ-0.05 the cell bodies were spindle or round, which is consistent with the brightfield images. Moreover, the morphology and quantity of the cells on ZJ-SF were better than those on JS-SF.

The viability and proliferation of Schwann cells on all samples were evaluated by calcein-AM/PI double staining and CCK-8 test, respectively. The results are shown in Figure 4. It was found that Schwann cells can grow on all samples, and almost all cells are green (Figure 4A), indicating that all samples can support cell survival and growth. Both staining test and CCK-8 test (Figure 4B) showed that the number of cells increased over time, which further proved that all the samples exhibited no cytotoxicity and supported cell proliferation. However, the number of cells on the sample made of silk fibroin degummed with 0.05% Na₂CO₃ was relatively larger, as the thoroughly degummed silk fibroin membrane was more conducive to cell adhesion. In addition, under the same degumming conditions, compared with JS-SF, the number of cells on the ZJ-SF samples is relatively more in 1 or 3 days, and the morphology of the cells was better.

In order to find out the reasons for the differences of cell adhesion and growth on different products, we analyzed the protein of samples and found that there were 55 kinds of proteins which were classified according to their function. Figure 5 shows the correlation analysis of each sample. It can be seen that the correlation between sample JS-0.05 and sample ZJ-0.05 is very high, as well as that between sample JS-0.5 and ZJ-0.5, indicating that the silk fibroin composition of different sources treated with the same degumming method is similar. The sample Antheraea pernyi silk is different from the other samples and is closest to sample ZJ-0.5, which tells us that the protein composition of Zhejiang silk treated with 0.5% is the most similar to that of A. pernyi silk.

Both A. pernyi silk fibroin and mulberry silk fibroin are fibrous proteins, and the category of amino acids is almost identical. However, there are obvious differences in quantity (Kim et al., 2012). The amino acid residues in the silk fibroin of A. pernyi are large, and the side chain contains rich active groups, including aspartic acid (ASP) and arginine (Arg), which can form a special RGD tripeptide sequence with glycine. Thus, A. pernyi silk is favorable to cell adhesion (Kopp et al., 2020), and its elongation and elasticity are higher than that of silkworm silk. Therefore, in our quantitative analysis, tussah silk was selected as the control. Through iTRAQ analysis, the silk with similar properties as tussah silk was determined so as to find the suitable silk source and degumming treatment method. According to the above-mentioned results, sample ZJ-0.5 is similar to sample A. pernyi silk, that is, the silk sample from Zhejiang Province being degummed in 0.5% Na₂CO₃. Previous cell experiments show that sample ZJ-0.5 is the best for cell adhesion, viability, or proliferation, which is well explained by the sample protein clustering here.

The 55 proteins identified in the samples could be classified into several categories based on their annotated molecular function: extracellular (16), binding (13), antimicrobial (1), fibroin (8), cell adhesion (2), integral of membrane (11), sericin (1), cyclosketen (2), and so on (Figure 6). It can be found that, except silk fibroin and sericin, other proteins are closely related to cell growth, which is one of the reasons that silk fibroin has good biological activity and is widely used in the biomedical field.

To compare protein abundances across different samples, we used scale function to realize data standardization. Figure 7 shows that the content of silk fibroin in samples JS-0.5 and ZJ-0.5 is higher, while the sericin content is lower, which is consistent with the degumming treatment method. Samples JS-0.5 and ZJ-0.5 were treated with 0.5% Na₂CO₃, and sericin is removed more thoroughly. Cell adhesion and DNA polymerase are also higher in samples JS-0.5 and ZJ-0.5, and there is more toxin-activity protein in samples 1 and 3, explaining why the cells on samples JS-0.5 and ZJ-0.5 survived more and grew better. However, the contents of cytoskeleton, binding, and extracellular proteins in samples JS-0.05 and ZJ-0.05 are higher than those in samples JS-0.5 and ZJ-0.5, indicating that a high concentration of Na₂CO₃ can effectively remove sericin but also make silk fibroin lose some useful proteins. Therefore, the content of each protein in every sample is different. They play various roles in the process of cell adhesion and proliferation. We need to consider their effects on cell growth comprehensively, retaining as much natural matrix components as possible on the premise of thoroughly removing sericin.