Cell growth and proliferation on the interface of a silk fabric tubular scaffold

Silk fibroin (SF), derived from the silkworm Bombyx mori, is a natural protein that exhibits excellent biocompatibility, biodegradability and low immunogenicity, and has been used in various tissue engineering materials including cornea, ¹ cartilage, ² ligament, ³ vessel, ⁴ and skin replacement. ⁵ The properties of the scaffold material have a significant influence on in situ tissue regeneration, and rapid endothelialization may suppress thrombosis and initial hyperplasia given adequate cell-scaffold interactions. Coating the stent surface with a cyclic-(arginine-glycine-aspartic acid-D-phenylalanine-lysine) peptide appeared to have a positive effect on cell proliferation of bovine aortic endothelial cells. ⁶ Additionally, sulfated silk fibroin tubular scaffolds (SFTSs) reduced platelet adhesion and activation, and enhanced endothelial cell adhesion, proliferation, and maintenance of cellular functions. ⁷ When cells interact with a scaffold, they sense both the material and the microarchitecture. Porosity is one of the most important factors: if pores are too small this can impede cell ingrowth, but if they are too large cells will not be able to bridge the pores and tissue generation will be difficult in the voids. The endothelial coverage of expanded polytetrafluoroethylene surfaces can be increased by changing the porosity of the graft from 30 to 60 µm. ⁸

In previous works, we developed an artificial porous vascular vessel using regenerated SF and silk fabric,^4,9,10 and explored molding approaches under different preparation conditions. These scaffolds consisted of a silk knit as the media and a polyethylene glycol diglycidyl ether (PEG-DE) cross-linked SF membrane as the intimal and adventitial layers that were integrated to form a porous tissue. Scaffolds have impressive stress-strain resistance when combined with this silk knit media. ¹⁰ Effective vascular grafts should promote intimate cell-scaffold interactions, and the size and distribution of pores on the internal surface are important factors. Scaffolds should be suitable for the growth of human vascular endothelial cells, while the middle layer should be optimized the growth of human vascular smooth muscle cells and fibroblasts, if tissue reconstruction is to be promoted. The aim of the present study was to analyze comparatively the growth and proliferation of fibroblasts and HUVECs on the interfacial layer of SFTSs. The results suggest that SFTSs with more extensive cell-scaffold interactions are superior and should be subjected to further detailed studies.

Materials and methods

Preparation of regenerated SF solution

Bombyx mori raw silk was treated three times with 0.06% Na₂CO₃ solution at 98℃–100℃ for 30 min to remove sericin. Degummed silk fiber was dissolved in the ternary solvent CaCl₂·CH₃CH₂OH·H₂O (mole ratio = 1:2:8) at 70 ± 2℃. The final SF solution was obtained by dialyzing against deionized water at 4℃ for 3 days and concentrated to 6% or 4%.

Preparation of SFTSs

SFTSs were prepared as described previously.^9,10 Briefly, silk-braided tubes were coated with solutions of SF and PEG-DE (SF:PEG-DE ratios of 1.0:0.5 and 1.0:1.0 [w/w]) from the internal to the external surface, then freeze-dried (−20℃ or −40℃) to form SFTSs with an inner diameter of 3.0 mm. SFTSs were immersed in deionized water at 4℃ for 3 days to remove unreacted residues, air-dried at room temperature, cut into small pieces, placed carefully at the bottom of 48-well plates (internal surface facing upward) and sterilized using gamma irradiation (25 KGy).

Cell culture

Cell culture was performed as described previously. ⁴ Briefly, Human umbilical vein endothelial cells (HUVEC, CRL-1730, ATCC) and fibroblasts (L929, Cell Bank of the Chinese Academy of Sciences) were used to evaluate the cytocompatibility of SFTSs. Cells were cultured in Dulbecco's Modified Eagles Medium (DMEM, Gibco, CA) with 10% fetal bovine serum (FBS, Gibco, CA), 100 U/mL penicillin and 100 µg/mL streptomycin at 37℃ and 5% CO₂ in air. While in the logarithmic growth phase, cells were trypsinized using 0.25% trypsin (Sigma) and re-suspended in fresh DMEM with FBS and antibiotics.

Cell fluorescence labeling by CM-DiI

CM-DiI (Invitrogen) labeling was performed as described previously. ⁴ After incubation in the working solution of CM-DiI for 5 min at 37℃, and 15 min at 4℃, cells were washed with phosphate buffered saline (PBS) and re-suspended at a density of 2 × 10⁵ cells/mL in fresh complete medium. Labeled cells (500 µL/well) were added onto the internal surfaces of the SFTSs and incubated at 37℃ in 5% CO₂. Fluorescently labeled cells growing on the SFTS interfaces were observed at 565 nm on an inverted fluorescence microscope (TH4-200, Olympus, Japan).

Cell fluorescence labeling by DAPI

DAPI labeling was performed as described previously. ⁴ During the logarithmic growth phase, cells were fed with fresh DMEM containing 5 µm DAPI (Beyotime, China) and incubated overnight. The medium was removed and the cells washed five times with PBS prior to trypsinization using 0.25% trypsin, and re-suspended at a density of 1.0 × 10⁵ cells/mL in fresh complete medium. DAPI-labeled cells were seeded onto internal surfaces of the SFTSs (500 µL/well) and incubated at 37℃ in 5% CO₂. Cell fluorescence intensities were measured at a wavelength of 454 nm using an inverted fluorescence microscope.

MTT assay

A total of 500 µL of cell suspension was used to seed onto the internal surface of SFTSs at a density of 5 × 10⁴ cells/mL. After different incubation periods, the SFTSs were washed three times with DMEM, and 450 µL of DMEM was added and mixed with 50 µL of 5 mg/mL 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2-H- tetrazolium bromide (MTT). After incubating for 4 h at 37℃, reagents were replaced with 500 µL of HCl-dimethylcarbinol and plates were gently agitated until the formazan precipitate completely dissolved. The absorbance at 490 nm was measured using a Synergy HT microplate reader (BIO-TEK, VT).

Pore structure analysis

The SFTSs were cut into 1 cm segments and inserted by a U type hollow fiber membrane between every two samples, and then fixed using epoxy resin adhesive. After 24 h for fixation, the samples were fully infiltrated in high pure water. Pore size including pore diameter and pore distribution was recorded using a 3H-2000PB bubble pressure method membrane pore size analyzer (Beishide, China).

Statistical analysis

Data were presented as means ± standard errors, and means were compared using one-way analysis of variance (ANOVA), followed by the independent-sample T-test using SPSS 17.0 statistical software.

Results and discussion

Pore structure of SFTS internal surface

Appropriate pore size and density are crucial to the performance of surgical vascular substitutes, affecting adhesion, growth and spreading of endothelial cells on the scaffold and accelerating the endothelialization process. ¹¹ In our previous works, we explored the control of SFTS formation and morphological structure^4,12 by silk fibroin concentration, freezing temperature and cross-linker content and were able to optimize and control the pore structures and sizes of the internal surfaces. Figure 1 shows the pore diameter of the internal surfaces of SFTSs. When SFTS was prepared with 4% SF and an SF:PEG-DE ratio of 1.0:0.5, silk fibroin molecules could not adequately cross-link with each other to cover the SFTS surface and yarns were clearly visible. The pore diameters were large (100 ∼ 300 µm, Figure 1a), and distributed unevenly. At the same SF:PEG-DE ratio of 1.0:0.5, but with an SF concentration of 6%, the pore diameters decreased (about 30 ∼ 100 µm, Figure 1b), but the porosity remained low. At a higher relative PEG-DE content (1.0:1.0), more silk fibroin molecules were cross-linked, and a layer of membrane was formed with a large number of micropores (about 25 ∼ 75 µm) distributed across the SFTS internal surface (Figure 1c). The freezing temperature also influenced pore structure; at 6% SF and the SF:PEG-DE ratio of 1.0:1.0, very small pores (5 ∼ 20 µm) were formed rarely when the freezing temperature was changed from −20℃ to −40℃ (Figure 1d). OPEN IN VIEWER

Figure 1.

SEM of SFTS internal surfaces. (a) SF:PEG-DE = 1.0:0.5, 4%, −20℃; ⁴ (b) SF:PEG-DE = 1.0:0.5, 6%, −20℃; (c) SF:PEG-DE = 1.0:1.0, 6%, −20℃; (d) SF:PEG-DE = 1.0:1.0, 6%, −40℃ (Bar of A = 1.0 mm; Bar of B, C and D = 500 µm).

The pore diameter of whole SFTSs, including the surface pores and internal pores, was measured using capillary flow porometry. The integral flow pore size distribution of all samples is shown in Table 1. The average pore diameter of sample A was 198 µm, but the most probable pore diameter was 65 µm. The result showed that the SFTS prepared with 4% SF and SF:PEG-DE ratio of 1.0:0.5 could not form well, which also can be seen from Figure 1a. Uniform pores were distributed in sample B (about 50 µm) and sample C (about 30 µm), respectively. The pore diameter of sample D was very small (average pore diameter <10 µm) and very little that would not benefit the cells adhesion and cells ingrowth. Results by capillary flow porometry indicated that the pores in SFTSs were interconnected.

OPEN IN VIEWER

Table 1.

Pore diameter of SFTSs

Sample*	A	B	C	D
Pore size distribution (µm)	60 ∼ 200	25 ∼ 75	10 ∼ 50	5 ∼ 20
Average pore diameter (µm)	198	43	30	9
Most probable pore diameter (µm)	65	50	28	9

Note: *Samples A, B, C and D were the same as A, B, C and D of Figure 1.

Growth of fibroblasts

The MTT assay is a quantitative method for evaluating cell viability, and was applied to fibroblasts on the internal surface and in the scaffold pores (Figure 2). On day 1 and day 3, the A₄₉₀ values of fibroblasts were similar, but after day 3 the A₄₉₀ values of samples C and D continued to increase markedly, while there were no obvious changes with samples A and B. It indicated that the living cell numbers in samples C and D were more than that in samples A and B on day 5. When the SF:PEG-DE ratio was1.0:0.5, there were more macropores or exposed silk braids distributed on the internal surfaces. Compared with the silk braid, porous 3D interface was better for growth and proliferation of fibroblasts. The pores of sample B were slightly large and cells could not bridge the macropores. OPEN IN VIEWER

Figure 2.

MTT analysis of fibroblasts. (A) SF:PEG-DE = 1.0:0.5, 4%, −20℃; (B) SF:PEG-DE = 1.0:0.5, 6%, −20℃; (C) SF:PEG-DE = 1.0:1.0, 6%, −20℃; (D) SF:PEG-DE = 1.0:1.0, 6%, −40℃.

In order to intuitively evaluate cells viability with the different scaffolds, fluorescence labeling was employed to follow cell proliferation using the Cell Tracker CM-DiI lipophilic fluorochrome dye. When combined with cell membrane, the fluorescence intensity of CM-DiI increases greatly, which makes it particularly suitable for long-term labeling and tracking of cells.^4,13 DAPI can penetrate the cell membrane to stain living cell nuclei, leading to emission of blue fluorescence, and may be used to detect living cells. ⁴

As shown in Figures 3 and 4, fibroblasts showed high proliferation capability from days 1 to 5 on all scaffold interfaces. On day 3, the number of cells in samples B, C and D were similar, and comparable to the tissue culture plate, but slightly greater than sample A. On day 5, fibroblasts continued to proliferate, but the cell number was significantly higher in sample C, and the fluorescence intensity was also higher. Because the pore diameters of samples A and B were too large for fibroblasts, cells could not bridge the pores and therefore could not effectively grow into the scaffolds. The pore structure of sample C, with a higher pore density, was the most suitable for growth of fibroblasts, and facilitated ingrowth. The pore density of sample D was very low, and this prevented cells from growing into the scaffold. Cell proliferation of sample C was also significantly higher than the tissue culture plate on day 5. OPEN IN VIEWER

Figure 3.

Fluorescent cells observation labeled by CM-DiI of fibroblasts (Bar = 100 µm). (a) SF:PEG-DE = 1.0:0.5, 4%, −20℃; (b) SF:PEG-DE = 1.0:0.5, 6%, −20℃; (c) SF:PEG-DE = 1.0:1.0, 6%, −20℃; (d) SF:PEG-DE = 1.0:1.0, 6%, −40℃; (e)Tissue culture plate.

OPEN IN VIEWER

Figure 4.

Fluorescent cells observation labeled by DAPI of fibroblasts (Bar = 100 µm). (a) SF:PEG-DE = 1.0:0.5, 4%, −20℃; (b) SF:PEG-DE = 1.0:0.5, 6%, −20℃; (c) SF:PEG-DE = 1.0:1.0, 6%, −20℃; (d) SF:PEG-DE = 1.0:1.0, 6%, −40℃; (e)Tissue culture plate.

The results indicated that pores promoted cell ingrowth by providing greater surface interfaces for fibroblasts proliferation. We predict that a suitable pore structure (diameter and density) would also benefit the ingrowth of human vascular fibroblasts and vascular adventitia regeneration.

Growth of HUVECs

HUVECs also exhibited high cell viability and at 3 days after seeding cells proliferation was significant, but slower than that of fibroblasts (Figure 5); after 4 days, there remained spare interface area for cell proliferation and growth. On day 7, the A₄₉₀ value of sample A was slightly higher than that of the other samples (Figure 5A). HUVECs covered the inner surface of samples B and C, but the pore diameter of sample A was larger, and some cells leaked into the scaffold. We hypothesized that there was more spare interface for cell proliferation on the inner surface for cells to enter and proliferate within the scaffold. However there were no significant differences in the A₄₉₀ values of the samples. Cells contact interface was required differently for different cells. We concluded that pore size and distribution on the internal surface of SFTS are important for adhesion and spreading of HUVECs, as is avoiding cells leakage. OPEN IN VIEWER

Figure 5.

MTT analysis of HUVECs. (A) SF:PEG-DE = 1.0:0.5, 4%, −20℃; (B) SF:PEG-DE = 1.0:0.5, 6%, −20℃; (C) SF:PEG-DE = 1.0:1.0, 6%, −20℃; (D) SF:PEG-DE = 1.0:1.0, 6%, −40℃.

Because the pore diameter of sample A was larger (Figure 1a), on day 1 the cell number was lower than the other samples, as shown in Figures 6 and 7. On day 4, there were no obvious differences in cell number between samples B, C and D. The pore diameter of sample B (30 ∼ 100 µm, Figure 1b) was larger than sample C (25 ∼ 75 µm, Figure 1c), and the larger pore (> 50 µm) of sample B was more than that of sample C, in which most of the pores were about 30 ∼ 50 µm, but the cell number was comparable and cell ingrowth of HUVECs lower than that observed with fibroblasts. At 7 days after seeding, cell numbers on the internal surface of samples B and C were significantly higher than that of sample D, in which the pores were much smaller (5 ∼ 20 µm, Figure 1d). Sample D also contained fewer pores, which likely also contributes to poor cell spreading. The morphological structure of the interface of sample B and C was therefore suitable for growth and spreading of HUVECs from day 1 to 7. OPEN IN VIEWER

Figure 6.

Fluorescent cells observation labeled by CM-DiI of HUVECs (Bar = 100 µm). (a) SF:PEG-DE = 1.0:0.5, 4%, −20℃; (b) SF:PEG-DE = 1.0:0.5, 6%, −20℃; (c) SF:PEG-DE = 1.0:1.0, 6%, −20℃; (d) SF:PEG-DE = 1.0:1.0, 6%, −40℃; (e)Tissue culture plate.

OPEN IN VIEWER

Figure 7.

Fluorescent cells observation labeled by DAPI of HUVECs (Bar = 100 µm). (a) SF:PEG-DE = 1.0:0.5, 4%, −20℃; (b) SF:PEG-DE = 1.0:0.5, 6%, −20℃; (c) SF:PEG-DE = 1.0:1.0, 6%, −20℃; (d) SF:PEG-DE = 1.0:1.0, 6%, −40℃; (e) Tissue culture plate.

PEG-DE is a cross-linker comprising flexible macromolecular chains. SFTS cross-linked by PEG-DE possesses a higher elastic recovery rate, ¹⁰ which is essential for artificial blood vessel. Our results showed that PEG-DE has no cytotoxicity. However, a high SF:PEG-DE ratio of 1.0:1.0 would make SFTSs stiff compared with the SF:PEG-DE ratio of 1.0:0.5. We are currently exploring a more suitable SF:PEG-DE ratio in the range of 1.0:0.5 to 1.0:1.0 in order to further optimize pore size and distribution for vascular cells growth.

Conclusions

Silk fibroin tubular scaffolds with different pore size and pore density were prepared by controlling the silk fibroin concentration, freezing temperature and cross-linker content. Fibroblasts and HUVECs displayed different cellular responses when in contact with the porous interface. For fibroblasts, a pore diameter of 50 ∼ 75 µm and a high pore density provided superior cell adhesive, proliferation and ingrowth, and maintained a high cell motility and fast dispersion in the scaffold, which is conducive to connective tissue regeneration. For HUVECs, 30 ∼ 50 µm micropores were optimal for cell adhesion, spreading and proliferation.