Home
Search
Collections
Journals
About
Contact us
My IOPscience
Gelatin modified ultrathin silk fibroin films for enhanced proliferation of cells
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2015 Biomed. Mater. 10 025003 (http://iopscience.iop.org/1748-605X/10/2/025003) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 169.230.243.252 This content was downloaded on 25/03/2015 at 16:41
Please note that terms and conditions apply.
Biomed. Mater. 10 (2015) 025003
doi:10.1088/1748-6041/10/2/025003
Paper
received
5 August 2014
Gelatin modified ultrathin silk fibroin films for enhanced proliferation of cells
re vised
21 January 2015 accep ted for publication
12 February 2015 published
18 March 2015
Luyuan Yang1, Mohammed Yaseen1, Xiubo Zhao1, Paul Coffey1, Fang Pan1, Yuming Wang2, Hai Xu2, John Webster3 and Jian R Lu1 1
Biological Physics Laboratory, School of Physics and Astronomy, University of Manchester, Schuster Building, Manchester M13 9PL, UK Centre for Bioengineering and Biotechnology, China University of Petroleum (East China), 66 Changjiang West Road, Qingdao 266580, People’s Republic of China 3 ISIS Neutron Facility, Rutherford Appleton Laboratory, STFC, Chilton, Didcot, Oxon OX11 0QX, UK 2
E-mail: [email protected] Keywords: polypeptides, biomedical materials, surface biocompatibility, biocompatible thin film
Abstract Silk fibroin (SF) films were modified with gelatin (G) to explore if such SF/G films could enhance the surface biocompatibility of silk as cell growth biomaterials. Ultrathin films were coated from aqueous SF solutions pre-mixed with different amounts of G. It was found that the SF/G blended films after methanol treatment were highly stable in physiological conditions. The incorporation of G smoothed the surface morphology of the SF/G films formed. Surface-exposed RGD sequences were successfully identified on the SF/G films through specific recognition of an integrin-mimicking peptide (bearing the sequence of CWDDGWLC). Cell culture experiments with 3T3 fibroblasts demonstrated that SF/G films with 1.2–20% (w/w) G gave clear improvement in promoting cell attachment and proliferation over pure SF films. Films containing 10–20% (w/w) of G showed cell attachment and growth even superior to the pure G films. The differences as observed from this study suggest that due to the lack of mechanical strength associated with its high solubility, G could not work alone as a cell growth scaffold. The enhanced cellular responses from the blended SF/G films must result from improvement in film stability arising from SF and in cytocompatibility arising from G. The results thus indicate the potential of the SF/G blends in tissue engineering and biomedical engineering where physical and biological properties could be manipulated via mixing either as bulk biomaterials or for coating purposes. S Online supplementary data available from stacks.iop.org/BMM/02/025003
1. Introduction Attachment, spreading and proliferation of anchorage-dependent cells are strongly influenced by surface physical and biological properties including surface stability, roughness, wettability, charge, chemical groups and cell-binding ligands [1–4]. In spite of extensive endeavours over the past few decades, it remains a formidable challenge to mediate the cell– surface interaction in cell culture, tissue engineering and biomedical applications. To promote initial cell growth, a good biomaterial, often a natural or synthetic protein or polypeptide is used to undertake surface modification, providing a suitable and more consistent cell growth substrate. Silk fibroin (SF) from the domesticated Bombyx mori silkworm is a competitive candidate as cell growth © 2015 IOP Publishing Ltd
substrates, not only because it is natural, biocompatible and inexpensive, but because it has many other unique properties. Earlier studies have revealed that native SF films have good mechanical strength in the wet state, and high oxygen and drug permeability [5–7]. In vivo and in vitro experiments showed that inflammatory responses to SF films were similar to or less than collagen films [8]. On the other hand, purification processes used to extract fibroin from raw silk often disrupt the native crystallized β-sheet domains, compromising their mechanical strength. However, with dehydration treatment (methanol, heat, etc), these β-sheet structures can be induced and any SF-based materials can therefore revert to being water-insoluble. This is a valuable property for a biomaterial to meet the requirements for long lasting performance in physiological conditions.
Biomed. Mater. 10 (2015) 025003
L Yang et al
SF from domesticated B. mori has been reported to support some mammalian cell growth [9–11], but unlike some wild types, it lacks specific integrin ligands or cell recognition sequences such as arginine– glycine–aspartic acid (RGD) to promote cell attachment and growth [10]. Modification of SF from B. mori is often necessary because culturing of stem cells and many primary cells require the enhancement from epitopes inherent of many natural extracellular matrix (ECM) proteins [12]. On the other hand, gelatin (G, typically type I collagen), which is hydrolyzed from collagen, shows excellent cell attachment, proliferation and differentiation properties due to its RGD and related sequences [13, 14]. It is widely used in pharmaceutical and biomedical fields since it is nontoxic, biodegradable, low cost and non-immunogenic [15]. However, because of its brittleness, high aqueous solubility, low intensity and uncontrollable degradation rates, G is rarely used alone [13, 15]. An attractive way forward would be to combine these two naturally occurring polymers together to exploit their respective strengths in enhancing stability and biocompatibility. Cell growth promoting peptide sequences such as RGD or RGDS can be introduced into polypeptidebased biomaterials in different manners, for example, through chemical grafting or insertion of moieties containing these cell recognition sequences [16, 17], through the addition of surfactants bearing such groups [18] and through recombinant expression of proteins or polypeptides that combine the features of silk, gelatin or other cell-adherent epitopes [19– 21]. These processes could also help optimize the performance of these peptide sequences, but the main drawbacks in many of these approaches stem from the cost and technical sophistications, for example, the need to validate a new synthetic route and a recombinant expression process and the effort to meet the requirements of regulatory approvals. Several previous studies have demonstrated the potential for SF/G blends as films, scaffolds and hydrogels in tissue engineering and controlled drug release [14, 15, 22–29]. So far the reported SF/G films have usually been cast by drying the mixtures of SF/G on polystyrene plates at a mild temperature overnight or for up to 3 days [15, 25–29]. This type of fabrication generates rather thick films, with thicknesses mostly around 1–10 µm, with little control of their surface morphologies and physiochemical properties. It is thus difficult to utilize these thick films to study how different film surfaces affect cell attachment and growth. However, the outer surfaces of ultrathin films with thicknesses over the nanometre range are strongly influenced by the confining effects from their substrates [30, 31]. As a result, their outer surface properties can be better controlled, enabling a better characterization of cell–biomaterial interactions. Previously, we have developed a spincoating method that can fabricate nano-scale SF films [32], with the thickness and surface morphology being precisely controlled, e.g. by varying the concentration, 2
thus providing quick and reproducible film production. In the work reported here, the same method was used to fabricate a series of SF/G blended films at the nanometre scale. In addition, most reported studies utilized chemical or enzymatic crosslinking [24, 28, 29] to stabilize SF/G blends which would affect the properties of the SF/G blends including cytocompatibility, toxicity and transparency [29]. In contrast, the films studied in this work were SF-based and blended with a small proportion of G. It is expected that no crosslinking was needed as the blended film stability could be controlled by inducing β-sheet structures among SF molecules. This study reports the assessment of film stability, secondary structures, morphology and performance in cell growth. The aim is to explore how the ratios of SF and G in blends affect film stability and cellular responses and how these physical and biological properties can be tuned to benefit cell growth.
2. Materials and methods 2.1. Materials B. mori raw silk was donated by a silk reeling manufacturer in Jiangsu Province in China. Gelatin was hydrolized from porcine skin (Type I, containing RGD sequences, powder, gel strength about 300 g bloom, from Sigma-Aldrich, CAS Number 9000-70-8) and used as supplied. All the other chemicals used in this study were purchased from Fisher Scientific or SigmaAldrich if not specified. 2.2. Preparation of regenerated SF solution Raw silk was degummed in boiling aqueous solution containing 0.5% (w/v) Na2CO3 for 30 min, followed by rinsing with distilled water. The degumming process was repeated two more times, followed by drying of the degummed silk fibres in a drying cabinet at 37 °C for 2 days. The degummed silk was dissolved at 60 °C for 4 h (under mild stir by a magnetic stirrer) in 9.3 M LiBr solution to make a concentration of 16% (w/v). The resulting solution was then dialyzed against ultrahigh quality (UHQ) water for approximately 5 days, in 2 kDa molecular weight cutoff dialysis tubing at room temperature. The dialysis water was frequently changed, until the electric conductivity of the dialysis solution matched the pure UHQ water for more than 12 h. The successful removal of LiBr was confirmed by x-ray photoelectron spectroscopy experiments, in which both elements of Li and Br were not observed on the electron spectra. To facilitate longer storage, the solution after dialysis was lyophilized into powder and kept in the fridge at 4 °C. The lyophilized powder is readily dissolvable in water for future use. 2.3. Fabrication of SF/G films Both SF and G powders were dissolved in phosphate buffered saline (PBS) buffer (pH 7.4) at 4 mg ml−1 separately and sterile-filtered with 0.2 µm syringe filters as stock solutions. All mixed sample solutions
Biomed. Mater. 10 (2015) 025003
L Yang et al
were made by adding G stock solution into SF stock solution at the contents of 0.625, 1.25, 2.5, 5, 10 and 20% (w/w), with the total concentrations being kept at 4 mg ml −1. Films were coated on silica ⟨1 1 1⟩ wafers (cuts of dimensions of 1.2 cm × 1.2 cm, from Compart Technology, UK) for the stability test and on sterilized glass wafers (diameter: 13 mm) for cell culture by a specially designed spin coater (WS400-6NPP-LITE; Laurell, USA) running at a speed of 3000 rpm for 20 s. Pure and modified SF films were immersed in methanol for 10 min to stabilize the films by inducing β-sheet formation. These films were then rinsed by flowing UHQ water for 45s , dried by N2 and vacuumed overnight. Because of the high solubility of G in the solvents, some of the G films used in this work were not subject to the methanol immersion and water rinsing procedure. Instead, the spin-coated G films were directly exposed to aqueous solutions and the adsorption helped retain a much thinner G nanofilm that was used for comparative studies (see figure 2 for further details). As will be revealed from the infrared study, because the exposure of pure G films to methanol did not affect any β-sheet transformation, methanol treatment would not promote their stability. 2.4. Film thickness and stability A Jobin-Yvon UVISEL phase modulated spectroscopic ellipsometer was used to measure the thickness and stability of the pure and mixed SF/G films. Spectroscopic ellipsometry measures the changes in the polarization state of a light beam caused by the reflection on the material surface. From the changes in the polarization state, information about layer thickness and refractive index can be obtained by calculating the relative phase change, Δ and the relative amplitude change, Ψ. These parameters are determined in two components: the plane of reflection (p-plane), and that perpendicular to it (s-plane). The sample ellipticity, ρ, is defined as the ratio of the Fresnel coefficients of the p and s planes (rp and rs) and is expressed as [33]: rp ρ= = tan(Ψ ) e iΔ. (1) rs
The thickness of the film can be calculated by the software DeltaPsi II using equation (1). For films measured at the air/solid interface, the formula applied is based on the Lorentz–Lorenz relation [33], which describes the refractive index n of a mixture of substances: n2 − 1 (2) = A1N1 + A2 N2 + A3N3 + … , n2 + 2
where Ai is the molar refractivity of substance i and Ni is the number of the moles of substance i per unit volume. For a pure substance, the area density of a deposited layer can be written as: M ⎛ n2 − 1 ⎞ Γ=τ ⎜ ⎟, (3) 10A ⎝ n2 + 2 ⎠ 3
where M is the molar mass of the polymer, M/A for protein is close to 4.2 [34], layer thickness τ is in angstroms (Å), n is kept the same as the bulk value as 1.478 and the resulting area density Γ is in milligrams per square metre (mg m−2). For film stability test, coated silica wafers were immersed in PBS culture buffer (Sigma-Aldrich) for 8 days at 37 °C. Each day the wafers were taken out, dried with N2 and measured in air. When the measurements were finished, the wafers were put back into the PBS buffer. 2.5. Secondary structures The secondary structures of the pure and mixed SF/G films with or without methanol treatment were characterized by a Fourier-transform infrared (FTIR) spectroscopy attached to a special attenuated total reflectance (ATR) cell enabling direct ATR measurement at the solid (ZnSe crystal)/D2O interface (Nicolet 6700; Thermo Scientific, USA). Each protein film was fabricated by covering the ZnSe crystal surface with the relevant protein solution (mixed or pure solutions made in D2O, 10 mg ml−1, pH 7). After 20 min, the solution was removed, followed by methanol treatment on the film for 10 min. The film was then dried for about 30 min, followed by filling the cell with D2O. ATR measurements were performed in transmission mode with a nominal resolution of 4 cm−1 and 64 scans for each experiment. The recorded wavenumber ranged from 400 to 4000 cm−1. Background measurement (in the presence of D2O) was taken before the protein sample was mounted and this background was subtracted from the sample readings. The average SF and SF/G film thicknesses following such treatment were found to be around 10 nm from samples coated on an optically flat silicon wafer measured by spectroscopic ellipsometry. The thicknesses of the films coated at different ratios of SF and gelatin are shown in online supplementary figure 1 (stacks.iop.org/BMM/02/025003). As already mentioned, pure G films had high solubility and could easily become dissolved completely. Care was taken to ensure that some G films were left on the ZnSe crystal surface with and without methanol treatment. 2.6. Surface morphology Surface analysis of the SF/G films was carried out using an atomic force microscope (AFM) (Nanoscope IIIa, Digital Instruments, USA) in tapping mode with an oxide-sharpened Si3N4 tip mounted on a triangular cantilever with a spring constant of 0.58 N m−1 (specified by the manufacturer). Images were taken in air at room temperature and flattened as required. Average values of roughness were analyzed using the software Nanoscope 7.20. 2.7. RGD epitope identification The RGD motifs on the surface of SF/G films were identified using an integrin-mimicking RGD-binding peptide CWDDGWLC-biotin [35, 36] prepared by the
Biomed. Mater. 10 (2015) 025003
L Yang et al
Fmoc solid phase peptide synthesis approach at the Biocentre of China University of Petroleum (Qingdao). The binding sites were further visualized either by streptavidin-colloidal gold (10 nm diameter, BBInternational) with silver enhancer (Sigma-Aldrich) or by Alexa Fluor® 488 streptavidin-colloidal gold (10 nm, Invitrogen) at room temperature. The films were first treated in a blocking buffer (pH = 8.6) containing 0.1 mol L−1 NaHCO3, 0.02% (w/v) NaN3, and 5% (w/v) bovine serum albumin for 1 h. Each film was washed five times in wash buffer (50 mmol L−1 TrisHCl, 150 mmol l−1 NaCl, 0.1% (v/v) Tween-20, pH = 7.5). After incubation with the integrin-mimicking peptide (0.1 mg ml−1 in wash buffer, 30 min), the films were washed five times in the wash buffer to remove any unbound peptide. Part of the films were treated with streptavidin-colloidal gold (3 ng ml−1 in wash buffer, 45 min) and rinsed in wash buffer and enhanced by a silver-enhancing kit for 20 min. The other films were treated with Alexa Fluor® 488 streptavidin-colloidal gold (20 µg ml−1, 45 min). Finally, all the samples were washed in wash buffer five times, rinsed with abundant running UHQ water, N2 dried and then imaged with a light microscope or confocal microscope. Fibronectin films (20 µg ml−1, 30 min immersion, N2 dried) were used as positive control. All the samples were prepared in triplicate. 2.8. Cell culture NIH 3T3 Cells (from mouse embryonic fibroblast cells) were cultured in Dulbecco’s modified Eagle’s medium (Sigma-Aldrich) supplemented with 10% fetal bovine serum (v/v) (PAA) and 1% (v/v) penicillin/streptomycin and were incubated in a humidified atmosphere with 5% CO2 at 37 °C. Cells were passaged every 3 days when approximately 80% confluence was reached. 2.9. Cell morphology For morphological observation, 3T3 fibroblasts were seeded on coated glass wafers placed in 24-well culture plates at a density of 10 000 cells per well. The morphology of the cells was directly examined by light microscope over 4 days. For F-actin stain, Alexa Fluor® 488 phalloidin (Invitrogen) was used on the cells cultured for 3 days. Specifically, the cells were washed twice with pre-warmed PBS culture buffer and fixed in 3.7% (w/v) formaldehyde solution in PBS for 10 min at room temperature. Following the PBS wash, the cells were further treated with 0.1% (v/v) Triton X-100 in PBS for 5 min. Then the cells were incubated in phalloidin solution for 20 min. 4′,6-diamidino-2-phenylindole (DAPI) (Sigma-Aldrich) was used to stain the cell nuclei. 2.10. MTT assay MTT (3-⟨4,5-dimethylthiazol-2-yl⟩-2,5-diphe nyltetrazolium bromide, a yellow tetrazole) assay was performed on the 3T3 cells. It is an indirect method to assess cell viability and proliferation since the activity of enzymes in mitochondria reduces MTT to formazan 4
crystals. Upon dissolving in dimethyl s ulfoxide (DMSO), the solution shows a purple colour which can be quantified to equivalent cell number/viability by colorimetric assay at the wavelength of 570 nm. 2.11. Statistical analysis Single factor analysis of variance (ANOVA) and Fisher’s least significant difference (LSD) test were used to assess the statistical significance of MTT assay results between groups. Data was shown as the mean ± standard deviation (n = 3). The statistical analysis was performed with the software SPSS 16.0 at the confidence levels of 95 and 99%.
3. Results and discussion 3.1. Secondary structures and stability of SF/G films Successful cell attachment and growth demand a stable underlying substratum to support sustainable bioactivity of ECM proteins and functional growth factors [3,37]. Failure of this property can cause problems in the initial cell attachment and further cell proliferation. SF in solution contains amorphous metastable conformations such as random coil and α-helix [38] which under certain conditions can be transformed into stable crystallized β-sheets. Such conditions include pH change, sheer force, methanol treatment (dehydrating solvent), metallic ions or heat treatment [39, 40]. In this work, methanol was used because of its efficiency and popularity in inducing stable β-sheets [41]. It was found that for such ultrathin films, 10 min of immersion in the absolute methanol gave the optimal outcome in terms of both stability and time efficiency. ATR-FTIR spectra (figure 1) show that after methanol treatment, amide I peak (C=O stretching) of the SF film shifted from 1645 to 1622 cm−1, which is a strong indication of the conformational transition from random coil to β-sheet [42]. The pure G films before (data not shown) and after methanol treatment display the amide I peak at the same wavenumber of 1653 cm−1, which indicates methanol has little or no effect on G. The blended SF/G films upon methanol treatment show signature peaks of both the methanol treated SF film (1622 cm−1) and G film (1653 cm−1), however, these peak intensities respond to the proportion of the two proteins within the blended films. There is a pronounced peak around 1460 cm−1 which is most likely to be the amide II peak (N–H bending). The position of this peak appears to be lower than reported values [15, 25, 43]. The difference could arise from the H-D exchange due to the presence of D2O in the bulk [44]. In addition, because the protein films were very thin, the material in it was strongly confined by the substrate. This could influence the peak position. We note that the films characterized in this work were adsorbed on a ZnSe surface that could impose some influence on the secondary structures of the surfaceconfined polypeptides. As a control, we undertook the same methanol treatment with the SF/G powders and
Biomed. Mater. 10 (2015) 025003
L Yang et al
Figure 1. ATR-FTIR spectra of the SF/G films with or without methanol treatment. SF/G(1.2), SF/G(5), SF/G(20): blended films containing 1.2, 5 and 20% (w/w) gelatin.
Figure 2. Stability of the spin-coated SF/G films. All films were immersed in Dulbecco’s PBS buffer at 37 °C for 8 days. SF/G(5), SF/G(20): blended films containing 5 and 20% (w/w) gelatin. Error bars indicate standard deviations from three different locations on the SiO2 wafer. All films were spin-coated at 4 mg ml−1, 3000 rpm and ambient conditions. The area densities of all films prior to methanol treatment were around 30–33 mg m−2. For methanol treated films, time zero marked the starting point immediately after methanol treatment followed by rinsing with abundant water. Methanol treatment and relatively strong water flow led to fast G film removal (yellow line), however, G film without any treatment (blue line) could retain a small amount of G due to irreversible adsorption.
observed almost identical changes, indicating little influence from the solid substrate in these cases. Thus, these studies show the clear effect of transformation to β-sheets upon methanol treatment, irrespective of the surface-confined SF/G films or bulk SF/G powders. Gelatin is a water-soluble heterogenous protein mixture derived from collagen. Following the results as shown in figure 1, it appears that by blending G with SF followed with methanol treatment, the crystallized β-sheet structures could retain some G molecules and prevent them from dissolving into the buffer. However, more studies need to be undertaken to establish 5
how these films behave upon exposure to buffer solution. Figure 2 shows the stability of the SF/G films in phosphate buffer at 37 °C mimicking the physiological conditions. Because G has a higher viscosity in solution than SF at the same concentration, spin-coated SF/G films or pure G film initially had a higher area density than the pure SF film. Methanol treatment was performed on the coated films. G film without methanol treatment was measured for comparison. The thicknesses of the spin-coated films were roughly 10–12 nm for the treated SF based films (online supplementary figure 1) (stacks.iop.org/BMM/02/025003) and 28 nm
Biomed. Mater. 10 (2015) 025003
L Yang et al
Figure 3. AFM height images of the SF/G film surfaces. (a): SF film; (b): SF/G blended film with 5% (w/w) gelatin; (c): SF/G blended film with 20% (w/w) gelatin; (d): gelatin film. All films were spin-coated at 4 mg ml−1. SF-based films were methanol treated.
Figure 4. Schematic diagram of RGD identification method using an integrin-mimicking peptide.
for the untreated G film. Noticeably, there was little film detected on the treated G film due to its high solubility and exposure to harsh water rinsing (under steady and relatively strong water flow). After the initial measurements (day 0), all the films were immersed in Dulbecco’s PBS buffer at 37 °C for 8 days which mimics cell culture conditions. The SF films and blended films showed good stability over time. Although there was some decrease in the area density for the film with the higher ratio of G content (20%, w/w) during the first day, the remaining part was stable. The untreated G film was substantially dissolved in buffer within hours as expected. However, a small amount of G remained on the substrate due to the irreversible adsorption effect. Amorphous SF/G blends appear transparent and miscible at any ratio. Upon exposure of as-cast films 6
to methanol, the amorphous part of SF transformed into stable crystallized β-sheets. It was found that the presence of G did not affect the dehydrating process induced by methanol. In fact, as suggested by Gil et al [43], the SF/G blends upon SF crystallization should yield physically cross-linked interpenetration networks in which G exhibits a triple helix structure at ambient temperature. The abundant hydrophobic crystallized SF may serve to protect G from direct exposure to the water molecules and thus lock G within the networks. A previous thermal calorimetry study [45] demonstrated that increasing crystallized SF content from 0 to 75% (w/w) in the SF/G blends increased the transition temperature of G going through the helix to coil state (i.e. gel to liquid) from 35 to 50 °C (peak value). This result can explain the mechanism of how SF entraps
Biomed. Mater. 10 (2015) 025003
L Yang et al
Figure 5. Light microscopic images identifying the existence of RGD motifs visualized as black spots by binding with CWDDGWLC–biotin–streptavidin–gold nanoparticles enhanced with silver on the surface of gelatin-modified SF films. SF/G(5), SF/G(10): blended films containing 5 and 10% (w/w) gelatin; Fn: fibronectin film as control. SF-based films were methanol treated to ensure stability prior to the assay testing. Scale bar represents 50 µm.
Figure 6. Confocal microscopic images identifying the existence of RGD motifs visualized as fluorescent spots by binding with CWDDGWLC–biotin–streptavidin–gold nanoparticles labelled with Alexa Fluor® 488 on the surface of gelatin-modified SF films. SF/G (10): blended films containing 10% (w/w) gelatin. Films were methanol treated to ensure stability prior to the assay testing. Scale bar represents 50 µm.
G in the blended films. Namely, apart from shielding G from directly contacting water molecules, crystallized SF can maintain G in the more stable conformation of triple helix structures at elevated temperatures (e.g. 37 °C). However, a high G content in the blends can result in a much less stable film. This is why in figure 2 after the initial rinse and 1 day immersion in the buffer, more came off from the SF/G (20) films than from the SF/G (5) films. 3.2. Surface morphology of SF/G films The surface morphology of the spin-coated SF/G films was characterized by AFM in tapping mode (figure 3). The SF-based films were methanol treated, water rinsed and N2 dried before AFM characterization. An untreated SF film was characterized as control (online supplementary figure 2) (stacks.iop.org/BMM/02/025003). Methanol treatment resulted in a much rougher surface than an untreated surface and presented clustering structures which were not found in the untreated film. 7
The average roughness of the SF films with and without methanol treatment is 7.1 nm and 0.6 nm, respectively. These values were much greater than 0.2–0.3 nm measured on the bare SiO2 wafer. The substantial increase in surface roughness together with the formation of isolated small patches indicated some surface segregation upon exposure to methanol. The phase image of the methanol-treated SF film (online supplementary figure 2) (stacks.iop.org/BMM/02/025003) supported the AFM height measurements. These AFM measurements also confirmed that the SF/G films had a complete surface coverage on the substrate and that the observed formation of the morphological features formed only on the outer film surface. When G was involved, the shape of each clustering became smaller and more rounded granules were formed. For the film with 20% (w/w) G, the surface was almost completely packed with small rounded granules. The average surface roughness was reduced from 7.1 nm for the pure SF film to 5.8 nm for the 5% (w/w) SF/G film. Further
Biomed. Mater. 10 (2015) 025003
L Yang et al
Figure 7. Light microscope images of live 3T3 fibroblasts on the pure SF films, modified SF films containing 0.6–20% (w/w) gelatin and pure gelatin films from 16 h to 4 days. SF-based films were methanol treated to ensure stability prior to cell culturing. Scale bar represents 100 µm.
increasing G content to 20% (w/w) only changed the surface morphology but did not reduce the roughness. This smoothing effect and morphology change must be related to the involvement of G in the dehydrating process upon exposure to methanol. As previously shown, SF/G films were fairly stable against water. The intermixing process could promote hydrogen bonding between SF and G molecules. As expected, the surface for the spin-coated pure G film was smooth with an 8
average roughness of 0.4 nm which is close to the SiO2 surface (0.2–0.3 nm). 3.3. RGD identification Arginine–glycine–aspartic acid (RGD) is the amino acid sequence found in blood and many ECM proteins, and is recognized as an important cell-binding ligand by at least one type of integrin (receptor) on the cell surface [4, 46]. These RGD-containing proteins (e.g.
Biomed. Mater. 10 (2015) 025003
L Yang et al
Figure 8. Fluorescence micrographs of 3T3 fibroblasts growing on SF film, modified SF films containing 5, 10 and 20% (w/w) gelatin, gelatin film (G) and polystyrene surface (P) on day 3. SF-based films were methanol treated to ensure stability prior to cell culturing. F-actin cytoskeleton was stained by Alexa Fluor® 488 phalloidin (green), and the cell nuclei were stained by DAPI (blue). The scale bar represents 50 µm.
fibronectin, vitronectin, fibrinogen, laminin, collagen/ gelatin, etc) together with their integrins constitute a versatile recognition system for cell attachment, migration, differentiation and signalling [4, 35, 46, 47]. Once immobilized on the surface, such tripeptide motifs promote cell attachment, but when solubilized in solution they can inhibit cells from anchoring [47–49]. In this work, the SF films were modified by blending with a relatively small amount of G. RGD motifs can be hidden by chain entanglement and β-sheet structures, making them unavailable for cellular interaction [36]. Therefore, it is important to confirm the existence of RGD sequence on the film surface. Two methods for visualizing surface RGD motifs were performed by applying an integrin-mimicking peptide (CWDDGWLC-biotin) [35, 36]. The biotin-labelled peptide was then probed by streptavidin-colloidal gold conjugate (figure 4). With the help of silver enhancer or fluorescein label, the RGD binding sites can be visualized under microscope imaging (figures 5 and 6). No sign of gold-binding was observed for the pure SF film because SF from domesticated B. mori silkworm does not contain any RGD sequences. However, clear dark spots and fluorescent spots appeared on the SF/G blended films, demonstrating the existence of RGD motifs on the film surface. These RGD motifs can initiate increased affinity between cells and film surface. 3.4. Morphology of 3T3 fibroblast cells 3T3 fibroblast cells were cultured on the SF/G blended films in order to examine cellular responses on those films. The morphology of the cells was monitored under the light microscope over 4 days as shown in figure 7. SF-based films were methanol treated while G films were not treated to ensure a G layer was retained on the substrate. At 16 h, the cells remained small and rounded on the pure SF films, showing weak or no attachment. However, clear cell attachment was presented on the blended films, evident from the cell sizes, stretches and shape changes. Even a very small amount of G content 9
(0.6%, w/w) improved cell attachment. When the G content was increased to 1.2% (w/w) or above, most cells had already attached and spread on the surface. It is interesting to note that although many cells seemed to attach on the pure G film, the majority of cells had not yet spread at this stage. This could be due to the instability of the G film when initially immersed in the culture medium. As G was spin-coated and not physically adsorbed or chemically bound to the substrate, loosely attached molecules were substantially dissolved into the medium quickly (within the first 5–20 h) as characterized by ellipsometry. Thus, it failed to provide a stable support for cells to attach at the early stage. These results proved that stable RGD-containing G molecules on the film surface had a great positive impact on the initial cell attachment. Over time, attached cells grew and proliferated. Mature cells were spindle-shaped and exhibited a flatter configuration on the blended films than on the pure SF films. On day 4, cell clusters were found on the pure SF films and blended films containing the least amount of G (0.6%, w/w). However, on blended films with higher G contents (1.2–20%, w/w), the cells were more evenly distributed and reached above 80% confluence, demonstrating better cytocompatibility of these blended film surfaces. Dramatic differences in the actin microfilament arrangement of cells were found between the blended films and the other substrates. Figure 8 shows the F-actin cytoskeleton of 3T3 fibroblasts cultured on the SF film, blended films, G film and standard commercial polystyrene culture substratum. One difference is the larger distribution of filaments for each individual cell on the blended film, reflecting a larger cell spread area. The other difference is that cells growing on the blended film had clear filopodia which were missing or not obvious on the other substrata. It is well known that filopodia has important roles in cell migration, cell–ECM adhesion and cell–cell adhesion [50]. Integrins and cadherins, which are cell adhesion molecules, are often found in the tips of filopodia or along the
Biomed. Mater. 10 (2015) 025003
L Yang et al
2.5
Absorbance (570nm)
**
2d
2.0
* **
* **
16h
3d
*
4d
1.5
1.0
0.5
*
**
**
**
**
**
0.0
SF
0.6%
1.2%
2.5%
5%
10%
20%
G
Figure 9. MTT results showing the viability of 3T3 fibroblast cells on SF film, blended films with 0.6, 1.2, 2.5, 5, 10 and 20% (w/w) gelatin and gelatin film (G). Error bars represent means + SD for n = 3. Statistical analysis was performed for the data collected at 16 h and on day 4 only. *(p
Search
Collections
Journals
About
Contact us
My IOPscience
Gelatin modified ultrathin silk fibroin films for enhanced proliferation of cells
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2015 Biomed. Mater. 10 025003 (http://iopscience.iop.org/1748-605X/10/2/025003) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 169.230.243.252 This content was downloaded on 25/03/2015 at 16:41
Please note that terms and conditions apply.
Biomed. Mater. 10 (2015) 025003
doi:10.1088/1748-6041/10/2/025003
Paper
received
5 August 2014
Gelatin modified ultrathin silk fibroin films for enhanced proliferation of cells
re vised
21 January 2015 accep ted for publication
12 February 2015 published
18 March 2015
Luyuan Yang1, Mohammed Yaseen1, Xiubo Zhao1, Paul Coffey1, Fang Pan1, Yuming Wang2, Hai Xu2, John Webster3 and Jian R Lu1 1
Biological Physics Laboratory, School of Physics and Astronomy, University of Manchester, Schuster Building, Manchester M13 9PL, UK Centre for Bioengineering and Biotechnology, China University of Petroleum (East China), 66 Changjiang West Road, Qingdao 266580, People’s Republic of China 3 ISIS Neutron Facility, Rutherford Appleton Laboratory, STFC, Chilton, Didcot, Oxon OX11 0QX, UK 2
E-mail: [email protected] Keywords: polypeptides, biomedical materials, surface biocompatibility, biocompatible thin film
Abstract Silk fibroin (SF) films were modified with gelatin (G) to explore if such SF/G films could enhance the surface biocompatibility of silk as cell growth biomaterials. Ultrathin films were coated from aqueous SF solutions pre-mixed with different amounts of G. It was found that the SF/G blended films after methanol treatment were highly stable in physiological conditions. The incorporation of G smoothed the surface morphology of the SF/G films formed. Surface-exposed RGD sequences were successfully identified on the SF/G films through specific recognition of an integrin-mimicking peptide (bearing the sequence of CWDDGWLC). Cell culture experiments with 3T3 fibroblasts demonstrated that SF/G films with 1.2–20% (w/w) G gave clear improvement in promoting cell attachment and proliferation over pure SF films. Films containing 10–20% (w/w) of G showed cell attachment and growth even superior to the pure G films. The differences as observed from this study suggest that due to the lack of mechanical strength associated with its high solubility, G could not work alone as a cell growth scaffold. The enhanced cellular responses from the blended SF/G films must result from improvement in film stability arising from SF and in cytocompatibility arising from G. The results thus indicate the potential of the SF/G blends in tissue engineering and biomedical engineering where physical and biological properties could be manipulated via mixing either as bulk biomaterials or for coating purposes. S Online supplementary data available from stacks.iop.org/BMM/02/025003
1. Introduction Attachment, spreading and proliferation of anchorage-dependent cells are strongly influenced by surface physical and biological properties including surface stability, roughness, wettability, charge, chemical groups and cell-binding ligands [1–4]. In spite of extensive endeavours over the past few decades, it remains a formidable challenge to mediate the cell– surface interaction in cell culture, tissue engineering and biomedical applications. To promote initial cell growth, a good biomaterial, often a natural or synthetic protein or polypeptide is used to undertake surface modification, providing a suitable and more consistent cell growth substrate. Silk fibroin (SF) from the domesticated Bombyx mori silkworm is a competitive candidate as cell growth © 2015 IOP Publishing Ltd
substrates, not only because it is natural, biocompatible and inexpensive, but because it has many other unique properties. Earlier studies have revealed that native SF films have good mechanical strength in the wet state, and high oxygen and drug permeability [5–7]. In vivo and in vitro experiments showed that inflammatory responses to SF films were similar to or less than collagen films [8]. On the other hand, purification processes used to extract fibroin from raw silk often disrupt the native crystallized β-sheet domains, compromising their mechanical strength. However, with dehydration treatment (methanol, heat, etc), these β-sheet structures can be induced and any SF-based materials can therefore revert to being water-insoluble. This is a valuable property for a biomaterial to meet the requirements for long lasting performance in physiological conditions.
Biomed. Mater. 10 (2015) 025003
L Yang et al
SF from domesticated B. mori has been reported to support some mammalian cell growth [9–11], but unlike some wild types, it lacks specific integrin ligands or cell recognition sequences such as arginine– glycine–aspartic acid (RGD) to promote cell attachment and growth [10]. Modification of SF from B. mori is often necessary because culturing of stem cells and many primary cells require the enhancement from epitopes inherent of many natural extracellular matrix (ECM) proteins [12]. On the other hand, gelatin (G, typically type I collagen), which is hydrolyzed from collagen, shows excellent cell attachment, proliferation and differentiation properties due to its RGD and related sequences [13, 14]. It is widely used in pharmaceutical and biomedical fields since it is nontoxic, biodegradable, low cost and non-immunogenic [15]. However, because of its brittleness, high aqueous solubility, low intensity and uncontrollable degradation rates, G is rarely used alone [13, 15]. An attractive way forward would be to combine these two naturally occurring polymers together to exploit their respective strengths in enhancing stability and biocompatibility. Cell growth promoting peptide sequences such as RGD or RGDS can be introduced into polypeptidebased biomaterials in different manners, for example, through chemical grafting or insertion of moieties containing these cell recognition sequences [16, 17], through the addition of surfactants bearing such groups [18] and through recombinant expression of proteins or polypeptides that combine the features of silk, gelatin or other cell-adherent epitopes [19– 21]. These processes could also help optimize the performance of these peptide sequences, but the main drawbacks in many of these approaches stem from the cost and technical sophistications, for example, the need to validate a new synthetic route and a recombinant expression process and the effort to meet the requirements of regulatory approvals. Several previous studies have demonstrated the potential for SF/G blends as films, scaffolds and hydrogels in tissue engineering and controlled drug release [14, 15, 22–29]. So far the reported SF/G films have usually been cast by drying the mixtures of SF/G on polystyrene plates at a mild temperature overnight or for up to 3 days [15, 25–29]. This type of fabrication generates rather thick films, with thicknesses mostly around 1–10 µm, with little control of their surface morphologies and physiochemical properties. It is thus difficult to utilize these thick films to study how different film surfaces affect cell attachment and growth. However, the outer surfaces of ultrathin films with thicknesses over the nanometre range are strongly influenced by the confining effects from their substrates [30, 31]. As a result, their outer surface properties can be better controlled, enabling a better characterization of cell–biomaterial interactions. Previously, we have developed a spincoating method that can fabricate nano-scale SF films [32], with the thickness and surface morphology being precisely controlled, e.g. by varying the concentration, 2
thus providing quick and reproducible film production. In the work reported here, the same method was used to fabricate a series of SF/G blended films at the nanometre scale. In addition, most reported studies utilized chemical or enzymatic crosslinking [24, 28, 29] to stabilize SF/G blends which would affect the properties of the SF/G blends including cytocompatibility, toxicity and transparency [29]. In contrast, the films studied in this work were SF-based and blended with a small proportion of G. It is expected that no crosslinking was needed as the blended film stability could be controlled by inducing β-sheet structures among SF molecules. This study reports the assessment of film stability, secondary structures, morphology and performance in cell growth. The aim is to explore how the ratios of SF and G in blends affect film stability and cellular responses and how these physical and biological properties can be tuned to benefit cell growth.
2. Materials and methods 2.1. Materials B. mori raw silk was donated by a silk reeling manufacturer in Jiangsu Province in China. Gelatin was hydrolized from porcine skin (Type I, containing RGD sequences, powder, gel strength about 300 g bloom, from Sigma-Aldrich, CAS Number 9000-70-8) and used as supplied. All the other chemicals used in this study were purchased from Fisher Scientific or SigmaAldrich if not specified. 2.2. Preparation of regenerated SF solution Raw silk was degummed in boiling aqueous solution containing 0.5% (w/v) Na2CO3 for 30 min, followed by rinsing with distilled water. The degumming process was repeated two more times, followed by drying of the degummed silk fibres in a drying cabinet at 37 °C for 2 days. The degummed silk was dissolved at 60 °C for 4 h (under mild stir by a magnetic stirrer) in 9.3 M LiBr solution to make a concentration of 16% (w/v). The resulting solution was then dialyzed against ultrahigh quality (UHQ) water for approximately 5 days, in 2 kDa molecular weight cutoff dialysis tubing at room temperature. The dialysis water was frequently changed, until the electric conductivity of the dialysis solution matched the pure UHQ water for more than 12 h. The successful removal of LiBr was confirmed by x-ray photoelectron spectroscopy experiments, in which both elements of Li and Br were not observed on the electron spectra. To facilitate longer storage, the solution after dialysis was lyophilized into powder and kept in the fridge at 4 °C. The lyophilized powder is readily dissolvable in water for future use. 2.3. Fabrication of SF/G films Both SF and G powders were dissolved in phosphate buffered saline (PBS) buffer (pH 7.4) at 4 mg ml−1 separately and sterile-filtered with 0.2 µm syringe filters as stock solutions. All mixed sample solutions
Biomed. Mater. 10 (2015) 025003
L Yang et al
were made by adding G stock solution into SF stock solution at the contents of 0.625, 1.25, 2.5, 5, 10 and 20% (w/w), with the total concentrations being kept at 4 mg ml −1. Films were coated on silica ⟨1 1 1⟩ wafers (cuts of dimensions of 1.2 cm × 1.2 cm, from Compart Technology, UK) for the stability test and on sterilized glass wafers (diameter: 13 mm) for cell culture by a specially designed spin coater (WS400-6NPP-LITE; Laurell, USA) running at a speed of 3000 rpm for 20 s. Pure and modified SF films were immersed in methanol for 10 min to stabilize the films by inducing β-sheet formation. These films were then rinsed by flowing UHQ water for 45s , dried by N2 and vacuumed overnight. Because of the high solubility of G in the solvents, some of the G films used in this work were not subject to the methanol immersion and water rinsing procedure. Instead, the spin-coated G films were directly exposed to aqueous solutions and the adsorption helped retain a much thinner G nanofilm that was used for comparative studies (see figure 2 for further details). As will be revealed from the infrared study, because the exposure of pure G films to methanol did not affect any β-sheet transformation, methanol treatment would not promote their stability. 2.4. Film thickness and stability A Jobin-Yvon UVISEL phase modulated spectroscopic ellipsometer was used to measure the thickness and stability of the pure and mixed SF/G films. Spectroscopic ellipsometry measures the changes in the polarization state of a light beam caused by the reflection on the material surface. From the changes in the polarization state, information about layer thickness and refractive index can be obtained by calculating the relative phase change, Δ and the relative amplitude change, Ψ. These parameters are determined in two components: the plane of reflection (p-plane), and that perpendicular to it (s-plane). The sample ellipticity, ρ, is defined as the ratio of the Fresnel coefficients of the p and s planes (rp and rs) and is expressed as [33]: rp ρ= = tan(Ψ ) e iΔ. (1) rs
The thickness of the film can be calculated by the software DeltaPsi II using equation (1). For films measured at the air/solid interface, the formula applied is based on the Lorentz–Lorenz relation [33], which describes the refractive index n of a mixture of substances: n2 − 1 (2) = A1N1 + A2 N2 + A3N3 + … , n2 + 2
where Ai is the molar refractivity of substance i and Ni is the number of the moles of substance i per unit volume. For a pure substance, the area density of a deposited layer can be written as: M ⎛ n2 − 1 ⎞ Γ=τ ⎜ ⎟, (3) 10A ⎝ n2 + 2 ⎠ 3
where M is the molar mass of the polymer, M/A for protein is close to 4.2 [34], layer thickness τ is in angstroms (Å), n is kept the same as the bulk value as 1.478 and the resulting area density Γ is in milligrams per square metre (mg m−2). For film stability test, coated silica wafers were immersed in PBS culture buffer (Sigma-Aldrich) for 8 days at 37 °C. Each day the wafers were taken out, dried with N2 and measured in air. When the measurements were finished, the wafers were put back into the PBS buffer. 2.5. Secondary structures The secondary structures of the pure and mixed SF/G films with or without methanol treatment were characterized by a Fourier-transform infrared (FTIR) spectroscopy attached to a special attenuated total reflectance (ATR) cell enabling direct ATR measurement at the solid (ZnSe crystal)/D2O interface (Nicolet 6700; Thermo Scientific, USA). Each protein film was fabricated by covering the ZnSe crystal surface with the relevant protein solution (mixed or pure solutions made in D2O, 10 mg ml−1, pH 7). After 20 min, the solution was removed, followed by methanol treatment on the film for 10 min. The film was then dried for about 30 min, followed by filling the cell with D2O. ATR measurements were performed in transmission mode with a nominal resolution of 4 cm−1 and 64 scans for each experiment. The recorded wavenumber ranged from 400 to 4000 cm−1. Background measurement (in the presence of D2O) was taken before the protein sample was mounted and this background was subtracted from the sample readings. The average SF and SF/G film thicknesses following such treatment were found to be around 10 nm from samples coated on an optically flat silicon wafer measured by spectroscopic ellipsometry. The thicknesses of the films coated at different ratios of SF and gelatin are shown in online supplementary figure 1 (stacks.iop.org/BMM/02/025003). As already mentioned, pure G films had high solubility and could easily become dissolved completely. Care was taken to ensure that some G films were left on the ZnSe crystal surface with and without methanol treatment. 2.6. Surface morphology Surface analysis of the SF/G films was carried out using an atomic force microscope (AFM) (Nanoscope IIIa, Digital Instruments, USA) in tapping mode with an oxide-sharpened Si3N4 tip mounted on a triangular cantilever with a spring constant of 0.58 N m−1 (specified by the manufacturer). Images were taken in air at room temperature and flattened as required. Average values of roughness were analyzed using the software Nanoscope 7.20. 2.7. RGD epitope identification The RGD motifs on the surface of SF/G films were identified using an integrin-mimicking RGD-binding peptide CWDDGWLC-biotin [35, 36] prepared by the
Biomed. Mater. 10 (2015) 025003
L Yang et al
Fmoc solid phase peptide synthesis approach at the Biocentre of China University of Petroleum (Qingdao). The binding sites were further visualized either by streptavidin-colloidal gold (10 nm diameter, BBInternational) with silver enhancer (Sigma-Aldrich) or by Alexa Fluor® 488 streptavidin-colloidal gold (10 nm, Invitrogen) at room temperature. The films were first treated in a blocking buffer (pH = 8.6) containing 0.1 mol L−1 NaHCO3, 0.02% (w/v) NaN3, and 5% (w/v) bovine serum albumin for 1 h. Each film was washed five times in wash buffer (50 mmol L−1 TrisHCl, 150 mmol l−1 NaCl, 0.1% (v/v) Tween-20, pH = 7.5). After incubation with the integrin-mimicking peptide (0.1 mg ml−1 in wash buffer, 30 min), the films were washed five times in the wash buffer to remove any unbound peptide. Part of the films were treated with streptavidin-colloidal gold (3 ng ml−1 in wash buffer, 45 min) and rinsed in wash buffer and enhanced by a silver-enhancing kit for 20 min. The other films were treated with Alexa Fluor® 488 streptavidin-colloidal gold (20 µg ml−1, 45 min). Finally, all the samples were washed in wash buffer five times, rinsed with abundant running UHQ water, N2 dried and then imaged with a light microscope or confocal microscope. Fibronectin films (20 µg ml−1, 30 min immersion, N2 dried) were used as positive control. All the samples were prepared in triplicate. 2.8. Cell culture NIH 3T3 Cells (from mouse embryonic fibroblast cells) were cultured in Dulbecco’s modified Eagle’s medium (Sigma-Aldrich) supplemented with 10% fetal bovine serum (v/v) (PAA) and 1% (v/v) penicillin/streptomycin and were incubated in a humidified atmosphere with 5% CO2 at 37 °C. Cells were passaged every 3 days when approximately 80% confluence was reached. 2.9. Cell morphology For morphological observation, 3T3 fibroblasts were seeded on coated glass wafers placed in 24-well culture plates at a density of 10 000 cells per well. The morphology of the cells was directly examined by light microscope over 4 days. For F-actin stain, Alexa Fluor® 488 phalloidin (Invitrogen) was used on the cells cultured for 3 days. Specifically, the cells were washed twice with pre-warmed PBS culture buffer and fixed in 3.7% (w/v) formaldehyde solution in PBS for 10 min at room temperature. Following the PBS wash, the cells were further treated with 0.1% (v/v) Triton X-100 in PBS for 5 min. Then the cells were incubated in phalloidin solution for 20 min. 4′,6-diamidino-2-phenylindole (DAPI) (Sigma-Aldrich) was used to stain the cell nuclei. 2.10. MTT assay MTT (3-⟨4,5-dimethylthiazol-2-yl⟩-2,5-diphe nyltetrazolium bromide, a yellow tetrazole) assay was performed on the 3T3 cells. It is an indirect method to assess cell viability and proliferation since the activity of enzymes in mitochondria reduces MTT to formazan 4
crystals. Upon dissolving in dimethyl s ulfoxide (DMSO), the solution shows a purple colour which can be quantified to equivalent cell number/viability by colorimetric assay at the wavelength of 570 nm. 2.11. Statistical analysis Single factor analysis of variance (ANOVA) and Fisher’s least significant difference (LSD) test were used to assess the statistical significance of MTT assay results between groups. Data was shown as the mean ± standard deviation (n = 3). The statistical analysis was performed with the software SPSS 16.0 at the confidence levels of 95 and 99%.
3. Results and discussion 3.1. Secondary structures and stability of SF/G films Successful cell attachment and growth demand a stable underlying substratum to support sustainable bioactivity of ECM proteins and functional growth factors [3,37]. Failure of this property can cause problems in the initial cell attachment and further cell proliferation. SF in solution contains amorphous metastable conformations such as random coil and α-helix [38] which under certain conditions can be transformed into stable crystallized β-sheets. Such conditions include pH change, sheer force, methanol treatment (dehydrating solvent), metallic ions or heat treatment [39, 40]. In this work, methanol was used because of its efficiency and popularity in inducing stable β-sheets [41]. It was found that for such ultrathin films, 10 min of immersion in the absolute methanol gave the optimal outcome in terms of both stability and time efficiency. ATR-FTIR spectra (figure 1) show that after methanol treatment, amide I peak (C=O stretching) of the SF film shifted from 1645 to 1622 cm−1, which is a strong indication of the conformational transition from random coil to β-sheet [42]. The pure G films before (data not shown) and after methanol treatment display the amide I peak at the same wavenumber of 1653 cm−1, which indicates methanol has little or no effect on G. The blended SF/G films upon methanol treatment show signature peaks of both the methanol treated SF film (1622 cm−1) and G film (1653 cm−1), however, these peak intensities respond to the proportion of the two proteins within the blended films. There is a pronounced peak around 1460 cm−1 which is most likely to be the amide II peak (N–H bending). The position of this peak appears to be lower than reported values [15, 25, 43]. The difference could arise from the H-D exchange due to the presence of D2O in the bulk [44]. In addition, because the protein films were very thin, the material in it was strongly confined by the substrate. This could influence the peak position. We note that the films characterized in this work were adsorbed on a ZnSe surface that could impose some influence on the secondary structures of the surfaceconfined polypeptides. As a control, we undertook the same methanol treatment with the SF/G powders and
Biomed. Mater. 10 (2015) 025003
L Yang et al
Figure 1. ATR-FTIR spectra of the SF/G films with or without methanol treatment. SF/G(1.2), SF/G(5), SF/G(20): blended films containing 1.2, 5 and 20% (w/w) gelatin.
Figure 2. Stability of the spin-coated SF/G films. All films were immersed in Dulbecco’s PBS buffer at 37 °C for 8 days. SF/G(5), SF/G(20): blended films containing 5 and 20% (w/w) gelatin. Error bars indicate standard deviations from three different locations on the SiO2 wafer. All films were spin-coated at 4 mg ml−1, 3000 rpm and ambient conditions. The area densities of all films prior to methanol treatment were around 30–33 mg m−2. For methanol treated films, time zero marked the starting point immediately after methanol treatment followed by rinsing with abundant water. Methanol treatment and relatively strong water flow led to fast G film removal (yellow line), however, G film without any treatment (blue line) could retain a small amount of G due to irreversible adsorption.
observed almost identical changes, indicating little influence from the solid substrate in these cases. Thus, these studies show the clear effect of transformation to β-sheets upon methanol treatment, irrespective of the surface-confined SF/G films or bulk SF/G powders. Gelatin is a water-soluble heterogenous protein mixture derived from collagen. Following the results as shown in figure 1, it appears that by blending G with SF followed with methanol treatment, the crystallized β-sheet structures could retain some G molecules and prevent them from dissolving into the buffer. However, more studies need to be undertaken to establish 5
how these films behave upon exposure to buffer solution. Figure 2 shows the stability of the SF/G films in phosphate buffer at 37 °C mimicking the physiological conditions. Because G has a higher viscosity in solution than SF at the same concentration, spin-coated SF/G films or pure G film initially had a higher area density than the pure SF film. Methanol treatment was performed on the coated films. G film without methanol treatment was measured for comparison. The thicknesses of the spin-coated films were roughly 10–12 nm for the treated SF based films (online supplementary figure 1) (stacks.iop.org/BMM/02/025003) and 28 nm
Biomed. Mater. 10 (2015) 025003
L Yang et al
Figure 3. AFM height images of the SF/G film surfaces. (a): SF film; (b): SF/G blended film with 5% (w/w) gelatin; (c): SF/G blended film with 20% (w/w) gelatin; (d): gelatin film. All films were spin-coated at 4 mg ml−1. SF-based films were methanol treated.
Figure 4. Schematic diagram of RGD identification method using an integrin-mimicking peptide.
for the untreated G film. Noticeably, there was little film detected on the treated G film due to its high solubility and exposure to harsh water rinsing (under steady and relatively strong water flow). After the initial measurements (day 0), all the films were immersed in Dulbecco’s PBS buffer at 37 °C for 8 days which mimics cell culture conditions. The SF films and blended films showed good stability over time. Although there was some decrease in the area density for the film with the higher ratio of G content (20%, w/w) during the first day, the remaining part was stable. The untreated G film was substantially dissolved in buffer within hours as expected. However, a small amount of G remained on the substrate due to the irreversible adsorption effect. Amorphous SF/G blends appear transparent and miscible at any ratio. Upon exposure of as-cast films 6
to methanol, the amorphous part of SF transformed into stable crystallized β-sheets. It was found that the presence of G did not affect the dehydrating process induced by methanol. In fact, as suggested by Gil et al [43], the SF/G blends upon SF crystallization should yield physically cross-linked interpenetration networks in which G exhibits a triple helix structure at ambient temperature. The abundant hydrophobic crystallized SF may serve to protect G from direct exposure to the water molecules and thus lock G within the networks. A previous thermal calorimetry study [45] demonstrated that increasing crystallized SF content from 0 to 75% (w/w) in the SF/G blends increased the transition temperature of G going through the helix to coil state (i.e. gel to liquid) from 35 to 50 °C (peak value). This result can explain the mechanism of how SF entraps
Biomed. Mater. 10 (2015) 025003
L Yang et al
Figure 5. Light microscopic images identifying the existence of RGD motifs visualized as black spots by binding with CWDDGWLC–biotin–streptavidin–gold nanoparticles enhanced with silver on the surface of gelatin-modified SF films. SF/G(5), SF/G(10): blended films containing 5 and 10% (w/w) gelatin; Fn: fibronectin film as control. SF-based films were methanol treated to ensure stability prior to the assay testing. Scale bar represents 50 µm.
Figure 6. Confocal microscopic images identifying the existence of RGD motifs visualized as fluorescent spots by binding with CWDDGWLC–biotin–streptavidin–gold nanoparticles labelled with Alexa Fluor® 488 on the surface of gelatin-modified SF films. SF/G (10): blended films containing 10% (w/w) gelatin. Films were methanol treated to ensure stability prior to the assay testing. Scale bar represents 50 µm.
G in the blended films. Namely, apart from shielding G from directly contacting water molecules, crystallized SF can maintain G in the more stable conformation of triple helix structures at elevated temperatures (e.g. 37 °C). However, a high G content in the blends can result in a much less stable film. This is why in figure 2 after the initial rinse and 1 day immersion in the buffer, more came off from the SF/G (20) films than from the SF/G (5) films. 3.2. Surface morphology of SF/G films The surface morphology of the spin-coated SF/G films was characterized by AFM in tapping mode (figure 3). The SF-based films were methanol treated, water rinsed and N2 dried before AFM characterization. An untreated SF film was characterized as control (online supplementary figure 2) (stacks.iop.org/BMM/02/025003). Methanol treatment resulted in a much rougher surface than an untreated surface and presented clustering structures which were not found in the untreated film. 7
The average roughness of the SF films with and without methanol treatment is 7.1 nm and 0.6 nm, respectively. These values were much greater than 0.2–0.3 nm measured on the bare SiO2 wafer. The substantial increase in surface roughness together with the formation of isolated small patches indicated some surface segregation upon exposure to methanol. The phase image of the methanol-treated SF film (online supplementary figure 2) (stacks.iop.org/BMM/02/025003) supported the AFM height measurements. These AFM measurements also confirmed that the SF/G films had a complete surface coverage on the substrate and that the observed formation of the morphological features formed only on the outer film surface. When G was involved, the shape of each clustering became smaller and more rounded granules were formed. For the film with 20% (w/w) G, the surface was almost completely packed with small rounded granules. The average surface roughness was reduced from 7.1 nm for the pure SF film to 5.8 nm for the 5% (w/w) SF/G film. Further
Biomed. Mater. 10 (2015) 025003
L Yang et al
Figure 7. Light microscope images of live 3T3 fibroblasts on the pure SF films, modified SF films containing 0.6–20% (w/w) gelatin and pure gelatin films from 16 h to 4 days. SF-based films were methanol treated to ensure stability prior to cell culturing. Scale bar represents 100 µm.
increasing G content to 20% (w/w) only changed the surface morphology but did not reduce the roughness. This smoothing effect and morphology change must be related to the involvement of G in the dehydrating process upon exposure to methanol. As previously shown, SF/G films were fairly stable against water. The intermixing process could promote hydrogen bonding between SF and G molecules. As expected, the surface for the spin-coated pure G film was smooth with an 8
average roughness of 0.4 nm which is close to the SiO2 surface (0.2–0.3 nm). 3.3. RGD identification Arginine–glycine–aspartic acid (RGD) is the amino acid sequence found in blood and many ECM proteins, and is recognized as an important cell-binding ligand by at least one type of integrin (receptor) on the cell surface [4, 46]. These RGD-containing proteins (e.g.
Biomed. Mater. 10 (2015) 025003
L Yang et al
Figure 8. Fluorescence micrographs of 3T3 fibroblasts growing on SF film, modified SF films containing 5, 10 and 20% (w/w) gelatin, gelatin film (G) and polystyrene surface (P) on day 3. SF-based films were methanol treated to ensure stability prior to cell culturing. F-actin cytoskeleton was stained by Alexa Fluor® 488 phalloidin (green), and the cell nuclei were stained by DAPI (blue). The scale bar represents 50 µm.
fibronectin, vitronectin, fibrinogen, laminin, collagen/ gelatin, etc) together with their integrins constitute a versatile recognition system for cell attachment, migration, differentiation and signalling [4, 35, 46, 47]. Once immobilized on the surface, such tripeptide motifs promote cell attachment, but when solubilized in solution they can inhibit cells from anchoring [47–49]. In this work, the SF films were modified by blending with a relatively small amount of G. RGD motifs can be hidden by chain entanglement and β-sheet structures, making them unavailable for cellular interaction [36]. Therefore, it is important to confirm the existence of RGD sequence on the film surface. Two methods for visualizing surface RGD motifs were performed by applying an integrin-mimicking peptide (CWDDGWLC-biotin) [35, 36]. The biotin-labelled peptide was then probed by streptavidin-colloidal gold conjugate (figure 4). With the help of silver enhancer or fluorescein label, the RGD binding sites can be visualized under microscope imaging (figures 5 and 6). No sign of gold-binding was observed for the pure SF film because SF from domesticated B. mori silkworm does not contain any RGD sequences. However, clear dark spots and fluorescent spots appeared on the SF/G blended films, demonstrating the existence of RGD motifs on the film surface. These RGD motifs can initiate increased affinity between cells and film surface. 3.4. Morphology of 3T3 fibroblast cells 3T3 fibroblast cells were cultured on the SF/G blended films in order to examine cellular responses on those films. The morphology of the cells was monitored under the light microscope over 4 days as shown in figure 7. SF-based films were methanol treated while G films were not treated to ensure a G layer was retained on the substrate. At 16 h, the cells remained small and rounded on the pure SF films, showing weak or no attachment. However, clear cell attachment was presented on the blended films, evident from the cell sizes, stretches and shape changes. Even a very small amount of G content 9
(0.6%, w/w) improved cell attachment. When the G content was increased to 1.2% (w/w) or above, most cells had already attached and spread on the surface. It is interesting to note that although many cells seemed to attach on the pure G film, the majority of cells had not yet spread at this stage. This could be due to the instability of the G film when initially immersed in the culture medium. As G was spin-coated and not physically adsorbed or chemically bound to the substrate, loosely attached molecules were substantially dissolved into the medium quickly (within the first 5–20 h) as characterized by ellipsometry. Thus, it failed to provide a stable support for cells to attach at the early stage. These results proved that stable RGD-containing G molecules on the film surface had a great positive impact on the initial cell attachment. Over time, attached cells grew and proliferated. Mature cells were spindle-shaped and exhibited a flatter configuration on the blended films than on the pure SF films. On day 4, cell clusters were found on the pure SF films and blended films containing the least amount of G (0.6%, w/w). However, on blended films with higher G contents (1.2–20%, w/w), the cells were more evenly distributed and reached above 80% confluence, demonstrating better cytocompatibility of these blended film surfaces. Dramatic differences in the actin microfilament arrangement of cells were found between the blended films and the other substrates. Figure 8 shows the F-actin cytoskeleton of 3T3 fibroblasts cultured on the SF film, blended films, G film and standard commercial polystyrene culture substratum. One difference is the larger distribution of filaments for each individual cell on the blended film, reflecting a larger cell spread area. The other difference is that cells growing on the blended film had clear filopodia which were missing or not obvious on the other substrata. It is well known that filopodia has important roles in cell migration, cell–ECM adhesion and cell–cell adhesion [50]. Integrins and cadherins, which are cell adhesion molecules, are often found in the tips of filopodia or along the
Biomed. Mater. 10 (2015) 025003
L Yang et al
2.5
Absorbance (570nm)
**
2d
2.0
* **
* **
16h
3d
*
4d
1.5
1.0
0.5
*
**
**
**
**
**
0.0
SF
0.6%
1.2%
2.5%
5%
10%
20%
G
Figure 9. MTT results showing the viability of 3T3 fibroblast cells on SF film, blended films with 0.6, 1.2, 2.5, 5, 10 and 20% (w/w) gelatin and gelatin film (G). Error bars represent means + SD for n = 3. Statistical analysis was performed for the data collected at 16 h and on day 4 only. *(p
