Colloids and Surfaces B: Biointerfaces 120 (2014) 28–37

Contents lists available at ScienceDirect

Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb

Silk porous scaffolds with nanofibrous microstructures and tunable properties Guozhong Lu a , Shanshan Liu b,1 , Shasha Lin b , David L. Kaplan c , Qiang Lu b,d,∗ a

Department of Burns and Plastic Surgery, The Third Affiliated Hospital of Nantong University, Wuxi 214041, People’s Republic of China National Engineering Laboratory for Modern Silk & Collaborative Innovation Center of Suzhou Nano Science and Technology, Soochow University, Suzhou 215123, People’s Republic of China c Department of Biomedical Engineering, Tufts University, Medford, MA 02155, USA d Jiangsu Province Key Laboratory of Stem Cell Research, Medical College, Soochow University, Suzhou 215006, People’s Republic of China b

a r t i c l e

i n f o

Article history: Received 27 November 2013 Received in revised form 10 March 2014 Accepted 17 March 2014 Available online 22 May 2014 Keywords: Silk Scaffolds ECM-mimetic Tissue regeneration Biomaterials

a b s t r a c t Scaffold biomaterials derived from silk fibroin have been widely used in tissue engineering. However, mimicking the nanofibrous structures of the extracellular matrix (ECM) for achieving better biocompatibility remains a challenge. Here, we design a mild self-assembly approach to prepare nanofibrous scaffolds from silk fibroin solution. Silk nanofibers were self-assembled by slowly concentrating process in aqueous solution without any cross-linker or toxic solvent and then were further fabricated into porous scaffolds with pore size of about 200–250 ␮m through lyophilization, mimicking nano and micro structures of ECM. Gradient water/methanol annealing treatments were used to control the secondary structures, mechanical properties, and degradation behaviors of the scaffolds, which would be critical for different tissue regeneration applications. With salt-leached silk scaffold as control, the ECM-mimetic scaffolds with different secondary structures were used to culture the amniotic fluid-derived stem cells in vitro to confirm their biocompatibility. All the ECM-mimetic scaffolds with different secondary structures represented better cell growth and proliferation compared to the salt-leached scaffold, confirming the critical influence of ECM-mimetic structure on biocompatibility. Although further studies such as cell differentiation behaviours are still necessary for clarifying the influence of microstructures and secondary conformational compositions, our study provides promising scaffold candidate that is suitable for different tissue regenerations. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Biomimetic materials are booming areas in biomaterials and tissue engineering, in which mimicking microstructural features of extracellular matrix (ECM) is a feasible way to improve the biocompatibility of different materials [1–3]. In these studies, porous 3-D nanofibrous scaffold fabrication is becoming important since growing studies have confirmed that the nanofibrous structure of ECM is a critical factor in providing suitable micro-environments for cell growth and proliferation [4–8]. Because of their impressive biocompatibility, biodegradability, minimal inflammatory reactions and excellent mechanical

∗ Corresponding author at: Corresponding author at: Soochow University, National Engineering Laboratory for Modern Silk, 199 Renai Road, Suzhou, Jiangsu 215123, People’s Republic of China. Tel.: +86 512 67061649/+86 532 68061649. E-mail addresses: [email protected], [email protected] (Q. Lu). 1 The author has same contribution with the first author. http://dx.doi.org/10.1016/j.colsurfb.2014.03.027 0927-7765/© 2014 Elsevier B.V. All rights reserved.

properties, silk-based materials have been used in different tissue regenerations such as bone, cartilage, blood vessel, skin and nerve [9–16]. Although silk-based scaffolds hold promise for tissue repairs, a challenge for scaffold fabrication remains to further improve silk biocompatibility, feasibility and inductivity for different tissue regeneration needs [17–21]. The design of silk-based scaffolds with nanofibrous structures is considered as suitable choice for further improving their biocompatibility. Electrospinning is a versatile technique for producing nanofiber-based biomaterials, and has been used to prepare silk fibroin fibrous matrices with diameters from a few micrometers down to the tens of nanometers [22]. Unfortunately, it is still difficult for electrospinning to prepare complex 3-D porous structures that are suitable for tissue regenerations. Lyophilization and salt-leaching are other methods to prepare silk scaffolds, which is powerful for porous structure fabrication but weak for nanofiber assembly [23–29]. In our previous study, collagen was used to induce the self-assembly of silk fibroin to form nanofibers. The porous scaffolds were subsequently achieved through lyophilization [4].

G. Lu et al. / Colloids and Surfaces B: Biointerfaces 120 (2014) 28–37

Then, through controlling the self-assembly process of silk to form nanofibers in aqueous solution, pure silk fibroin scaffolds composed of nanofibers were also prepared by freeze–drying the nanofiber solution [30] and showed better biocompatibility than silk scaffolds without nanofiber structure. However, further studies are still necessary to regulate the mechanical properties and degradation behaviors of the nanofiber scaffolds in order to adapt different tissue regenerations. Therefore, in our present study, the silk nanofiber scaffolds were treated by improved gradient water–methanol annealing processes to induce silk I and silk II formation. Compared to the salt-leached silk scaffolds, the nanofiber scaffolds containing different silk I and silk II contents showed tunable degradation behaviors, various mechanical properties, and better cell growth in vitro, implying better feasibility for different tissue regenerations. 2. Materials and methods 2.1. Preparation of silk solutions Bombyx mori silk fibroin solutions were prepared according to our previous published procedures [31]. Cocoons were boiled for 20 min in an aqueous solution of 0.02 M Na2 CO3 and then rinsed thoroughly with distilled water to extract the sericin proteins. After drying the extracted silk fibroin was dissolved in 9.3 M LiBr solution at 60 ◦ C for 4 h, yielding a 20% (w/v) solution. This solution was dialyzed against distilled water using dialysis tube (MWCO 3500) for 72 h to remove the salt. Then the solution was centrifuged at 9000 r min−1 for 20 min at 4 ◦ C to remove silk aggregations formed during the process. The final concentration of silk in water was about 6%, determined by weighing the remaining solid after drying. 2.2. Preparation of nanofibrous silk scaffolds Based on our silk self-assembly study [4,32], silk fibroin nanofibers were assembled by slowly concentrating the fresh solution to 25–30 wt% for 4 days at room temperature in fume hood. The nanofiber solution was diluted to 2.5 wt% with distilled water, and then poured into cylindrically-shaped container. The container was placed at −20 ◦ C for 24 h to freeze the samples and lyophilized for about 72 h to achieve silk porous scaffolds. Modulated water/methanol annealing processes were applied to induce gradual transformations from random to silk I/silk II structures. The scaffolds were placed in desiccators filled with methanol/water blend solutions with a 25 in. Hg vacuum for 6 h to produce waterinsoluble scaffolds. Methanol contents in these blend solutions were 0%, 10%, 30% and 50%, respectively. Traditional methanol annealing process also used to increase the ␤-sheet content through immersing the scaffolds in 80% (v/v) methanol solution for about 30 min [24,25,33].

29

examined using a S-4800 SEM at 3 kV. Since previous study has indicated that the nanoscale architectures of silk fibroin are usually covered by compact surface of the pore walls [4], different scaffolds were incubated at 37 ◦ C in 40 ml phosphate saline (PBS) containing 5 U ml−1 protease XIV for 6 h to degrade the compact surface. Then the nano-structural morphologies of the different scaffolds after degradation were examined using a Hitachi S-4800 SEM (Model S-4800, Hitachi, Tokyo, Japan). 2.4. Structural analysis of the scaffolds The structure of the various scaffolds was analyzed by FTIR on a NICOLET FTIR 5700 spectrometer (Thermo Scientific, FL, USA) equipped with a MIRacleTM attenuated total reflection (ATR) Ge crystal cell in reflection mode. For each measurement 64 scans were coded at a resolution of 4 cm−1 , with the wave number ranging from 400 to 4000 cm−1 [31,35]. X-ray diffraction (XRD) was also used to determine crystal structure of the scaffolds. The experiments were performed with an X-ray diffractometer (X Pert-Pro MPD, PANalytical B.V., Holland) using Cu Ka radiation at 40 kV. Irradiation conditions were at 40 mA and a scanning rate of 0.6◦ min−1 [36]. 2.5. Differential scanning calorimetry (DSC) The thermal properties of the scaffolds were measured in a TA Instrument Q100 DSC (TA Instruments, New Castle, DE) under a dry nitrogen gas flow of 50 ml min−1 . The samples were heated at 2 ◦ C min−1 from −30 ◦ C to 350 ◦ C [33]. 2.6. In vitro enzymatic degradation Different silk fibroin scaffolds were incubated at 37 ◦ C in 40 ml of phosphate-buffered saline (PBS) containing protease XIV (5 U ml−1 ). Each solution contained an approximately equivalent mass (40 ± 5 mg) of scaffolds. Solutions were replenished with enzyme and samples were collected daily. At designated time points the samples were rinsed with distilled water and prepared for mass balance assessment [37]. 2.7. Mechanical properties The compression properties of specimens (d = 10 mm, h = 11 mm) were measured with a cross head speed of 2 mm min−1 at 25 ◦ C using an Instron 3366 testing frame (Instron, Norwood, MA) with a 10 N capacity load cell. The mechanical properties of the scaffolds were determined in wet conditions. For the wet conditions, the scaffolds were first hydrated in water for 2 h and then measured at 25 ◦ C with a cross head speed of 2 mm min−1 . All samples were measured in triplicates [10].

2.3. Morphology analysis of the scaffold

2.8. Cell culture

The nanostructural transition of silk fibroin in aqueous solution was observed by AFM (Veeco, Nanoscope V, NY, America) in air. A 225 ␮m long silicon cantilever with a spring constant of 3 Nm−1 was used in tapping mode at 1.5 Hz scan rate. To prepare the samples for AFM imaging, different SF solutions were diluted to below 0.0001 wt% with deionized water to avoid masking the original morphology by multilayers of silk [34]. Once diluted, 2 ␮l of the diluted SF solution was quickly dropped onto freshly cleaved 4 × 4 mm2 mica surfaces and dried under a nitrogen gas. The morphology of the scaffolds was observed using a SEM (Hitachi S-4800, Hitachi, Tokyo, Japan). The specimens were fractured in liquid nitrogen using a razor blade and then sputter-coated with gold prior to imaging. Then the morphologies of the scaffolds were

Considering amniotic fluid-derived stem cells with a capacity to differentiate into multiple cells have been used in different tissue regeneration applications, and could avoid ethical concerns [38–41], amniotic fluid-derived stem cells were used to evaluate the cytocompatibility of the scaffolds. Samples of amniotic fluid were obtained from the First Affiliated Hospital of Soochow University (Suzhou, Jiangsu, China) following routine amniocentesis carried out on pregnant women at 15 to 35 weeks of gestation. All the procedures were performed following the guidelines established by the First Affiliated Hospital of Soochow University and the First Affiliated Hospital of Soochow University Ethics Committee; a written consent was obtained from each woman to use the amniotic fluid for research purposes. The isolation of

30

G. Lu et al. / Colloids and Surfaces B: Biointerfaces 120 (2014) 28–37

Fig. 1. AFM images of silk fibroin in fresh solution (a) and treated solution after slow concentrating process (b) on a mica surface, and then the SEM images of lyophilized silk scaffolds derived from the fresh solution (c) and the treated solution (d). The treated solution was prepared through slowly concentrating the fresh solution to 25 wt% for 3 days in fume hood an d then diluting to original concentration. For AFM investigation, the solutions were diluted to below 0.001 wt% firstly and then dried on the mica surface to form single layer of silk.

human amniotic fluid-derived stem cells has been described previously [42]. Amniotic fluid-derived stem cells were cultured in ␣-Modified Eagle Medium (␣-MEM) supplemented with 10% fetal bovine serum (FBS), 1% UI ml−1 Streptomycin-Penicillin (all from Invitrogen, Carlsbad, CA) and growth factor: Basal, C Frozen Supplement (Irvine Scientific, Santa Ana, CA). The medium was replaced every 2 d and the cultures were maintained in a humidified incubator at 37 ◦ C and 5% CO2 . The scaffolds (thickness 2 mm) were punched into small discs with a diameter of 5 mm and then sterilized by  radiation. After pre-conditioning with the culture medium overnight, the samples were transferred into a non-treated 6-well cell culture plate and seeded with the amniotic fluid-derived stem cells at the density of 1 × 105 /well. 2.9. Cell growth and morphology The cell morphology of the amniotic fluid-derived stem cells on the scaffolds was examined by confocal microscopy. Briefly, the cell-seeded scaffolds were washed three times with PBS (pH7.4). Cells were fixed for 30 min by incubating in 4% paraformaldehyde, followed by further washing. The cells were permeabilized using 0.1% Triton X-100 for 5 min and incubation with FITC-phalloidin for 20 min at room temperature, then PBS washing and finally staining with 4, 6-diamino-2-phenyl indole (DAPI) for 1 min. Fluorescence images from stained samples were obtained through a confocal laser scanning microscope (CLSM, Olympus FV10 inverted microscope, Nagano, Japan). The cell morphology on the scaffolds was confirmed by SEM. After harvest, the seeded scaffolds were washed with PBS 3 times and fixed in 4% paraformaldehyde

(Sigma-Aldrich, St. Louis, Mo) at room temperature, and then the scaffolds were washed 3 times with PBS again. Fixed samples were dehydrated through a gradient of alcohol (50%, 70%, 80%, 90%, 100% and 100%) followed by lyophilization. Specimens were examined using a Hitachi Model S-4800 SEM (Hitachi, Tokyo, Japan). 2.10. DNA content To study cell proliferation on the scaffolds, samples were harvested at the indicated time point (from day 1 to day 12), and digested with proteinase K overnight at 56 ◦ C, as described previously [43]. The DNA content was determined by using Pico Green TM DNA assay following the protocols of the manufacturer (Molecular Probes, Eugene, OR). Samples (n = 5) were measured at excitation wavelength of 480 nm and emission wavelength of 530 nm using a BioTec Synergy 4 spectrofluorometer (BioTec, Winooski, UK). The amount of DNA was calculated by interpolation from a standard curve prepared with lambda DNA in 10 × 10−3 M Tris–HCl (pH7.4), 5 × 10−3 M NaCl, 0.1 × 10−3 M EDTA over a range of concentrations. 2.11. Statistical analysis All statistical analyses were performed using SPSS v. 16.0 software. Comparison of the mean values of the data sets was performed using two-way ANOVA. Measures are presented as means ± standard deviation, unless otherwise specified, and p < 0.05 was considered significant.

G. Lu et al. / Colloids and Surfaces B: Biointerfaces 120 (2014) 28–37

31

Fig. 2. SEM images of the scaffolds derived from the silk nanofiber solution and treated with different annealing processes as follows: (a) water annealing treatment; (b) water/methanol annealing treatment with 10% methanol; (c) water/methanol annealing treatment with 30% methanol; (d) water/methanol annealing treatment with 50% methanol and (e) methanol treatment by immersing the sample in 80% methanol solution for 30 min.

3. Results and discussion 3.1. Microstructure of silk scaffolds Our recent study clarified that the self-assembly of silk in aqueous solution could be influenced by some kinetic factors such as concentration, temperature, time and charge [34]. The self-assembly process shows a transformation from nanoparticles to nanofibers coupled with the transition from random to crystal structures, in which the increase of concentration and time facilitates nanofiber conversion while the elevation of temperature induces insoluble silk II formation [32,34]. According to the mechanism, it would be possible to prepare random silk nanofibers in aqueous solution by elaborately regulating these factors. Therefore, a very slowly concentrating process at room temperature was used to facilitate soluble silk nanofiber formation

without insoluble silk II transition in our present study. As shown in Fig. 1, the nanofibers were successfully self-assembled from nanoparticles after the concentrating process. The diameter of the nanofibers ranged from 20 nm to 70 nm, providing suitable motifs for further fabrication of ECM-mimetic porous scaffolds. Since our recent study has implied that nanofibrous structures of silk could restrain separate sheet formation in the lyophilization process [4], the scaffolds were directly prepared from the nanofiber solutions through lyophilization. The lyophilized scaffolds showed excellent porous structures (Fig. 1d) rather than lamellae derived from fresh solution (Fig. 1c), confirming our previous hypothesis. Considering that the lyophilized scaffolds were soluble in water because of their random structure, different annealing processes were used to improve the stability of the scaffolds in aqueous environments, and then more importantly, to regulate their secondary conformational compositions for achieving different degradation

32

G. Lu et al. / Colloids and Surfaces B: Biointerfaces 120 (2014) 28–37

Fig. 3. SEM images of different lyophilized scaffolds after incubated in 5 U ml−1 protease XIV solution at 37 ◦ C for 6 h. The scaffolds were derived from silk nanofiber solution and treated with different annealing processes as follows: (a) water annealing treatment; (b) water/methanol annealing treatment with 10% methanol; (c) water/methanol annealing treatment with 30% methanol; (d) water/methanol annealing treatment with 50% methanol and (e) methanol treatment by immersing the sample in 80% methanol solution for 30 min.

behaviors and mechanical properties. Fig. 2 showed the influence of the different treatments on the microporous structures of the scaffolds. All the scaffolds maintained their porous structures after different treatments, indicating the slight effect of the treatments. Although the scaffolds were prepared from silk nanofiber solutions, a compact surface of the pore wall usually formed after lyophilization [4], making it impossible to investigate the changes of nanofibrous structure directly after the treatments. Therefore, the treated silk scaffolds were cultured in protease XIV solution for 6 h to degrade the compact surface and expose the nanofibers inside the scaffolds [4]. After the degradation, the nanofibers appeared in all the scaffolds treated with different processes (Fig. 3), which suggested that the different annealing processes had tiny influence on nanofiber structure. These SEM results indicated that a series of ECM-biomimetic scaffolds with nanofibrous and microporous structures was successfully achieved by the slowly concentratinglyophilizating-annealing processes.

3.2. Structural analysis FTIR and XRD were used to determine the secondary structures of the scaffolds after different annealing treatments (Fig. 4A and B). The infrared spectral region within 1700–1600 cm−1 is assigned to absorption by the peptide backbone of amide I, which has been commonly used for the analysis of different secondary structures of silk fibroin [34,35]. The peak at 1610–1630 cm−1 is characteristic of silk II secondary structure [44] while the peaks at 1648–1656 cm−1 and 1635–1645 cm−1 are indicative of silk I conformation and random coil, respectively [44]. The scaffolds before annealing treatment showed main peak at 1640 cm−1 , corresponding to random coil conformation. After water annealing treatment, the main amide I peak of the scaffolds changed to 1655 cm−1 , indicating silk I formation. Then the peak intensity at 1655 cm−1 gradually decreased while the peak intensity at 1625 cm−1 (typical silk II band) increased following the increase of methanol contents

G. Lu et al. / Colloids and Surfaces B: Biointerfaces 120 (2014) 28–37

33

peak from 12.5 to 26◦ . Then typical silk I peaks at 12.1◦ and 19.8◦ appeared in the water-annealed scaffolds while gradually transition from silk I to silk II (20.6◦ ) was found in the water/methanol annealed scaffolds when methanol contents increased from 10% to 80%. Compared to previous water-annealed silk materials [45], more silk I structure formed in our water-annealed nanofibrous silk scaffolds. Our recent study has revealed that the transition from random coil conformation to silk I or silk II structure is depended on silk molecular mobility and hydrophilic interaction. Silk fibroin molecules have to overcome the energy barrier resulting from the hydrophilic interactions to form silk II structure [34]. Therefore, stronger hydrophilic interaction could restrain silk II formation in annealing process, resulting in more silk I transformation. Considering more hydrophilic interactions formed after slowly concentrating process [34,44], silk I rather than silk II became the main crystal structure in our water-annealed scaffolds. Then, the silk II contents gradually increased following the increase of methanol in annealing processes since silk molecular became more active in the processes. Therefore, both FTIR and XRD results indicated that the secondary conformations of the ECM-biomimetic scaffolds could be easily regulated through modulated annealing processes. Since the secondary conformational composition has pivotal influence on scaffold properties including stability, degradation, mechanical properties and so on, our present study might provide a feasible way to design biomimetic scaffolds with tunable properties that are suitable for different tissue regenerations. 3.3. Scaffold properties

Fig. 4. FTIR spectra (A), X-ray diffraction spectra (B) and DSC curves (C) of different nanofibrous scaffolds before and after various annealing processes: (a) untreated scaffolds; (b) water annealing treatment; (c) water/methanol annealing treatment with 10% methanol; (d) water/methanol annealing treatment with 30% methanol; (e) water/methanol annealing treatment with 50% methanol and (f) methanol treatment by immersing the sample in 80% methanol solution for 30 min.

in the water/methanol annealing processes. These results implied that silk I and silk II contents in the insoluble scaffolds could be regulated by adjusting the ratios of water and methanol in the annealing processes, which would further alter the characters of the scaffolds. The structural changes of the scaffolds were confirmed by XRD curves. Similar to FTIR results, the untreated scaffolds exhibited typical amorphous state of silk fibroin, having a broad

The thermal stability of the scaffolds after different annealing treatments were investigated with standard DSC curves (Fig. 4C). The scaffolds before annealing treatments showed an endothermic water peak between 50 and 100 ◦ C, a minor non-isothermal crystallization peak at 220 ◦ C and a degradation peak at around 270 ◦ C (Fig. S1). The appearance of crystallization peak indicated that the scaffolds contained unstable non-crystal structures that were transformed to ␤-sheet at the crystallization peak. Then the scaffolds started to degrade, with an endothermal peak at around 270 ◦ C. When treated with water/methanol annealing processes, the crystallization peak disappeared because of the formation of stable silk I and silk II crystals. Interestingly, the water-annealed scaffold showed two degradation peaks that related to silk I and silk II structures respectively in the region 250–280 ◦ C, a main peak at 257 ◦ C and a minor peak at 272 ◦ C. The result further confirmed that the water-annealed scaffolds were mainly composed of silk I structure. When the scaffolds were treated with 10% methanol annealing process, the intensity of peak at 257 ◦ C decreased while the peak at 272 ◦ C became the main peak. The results indicated that more silk II crystals formed after the annealing process although the changes seemed insignificant in FTIR and XRD results. Following the further increase of methanol in the water/methanol annealing processes, the intensity of the peak at 272 ◦ C gradually increased and changed to 281 ◦ C, accompanying with the decrease and finally disappearance of the peak at 257 ◦ C, which indicated tunable crystal composition of the scaffolds and gradual improvement of thermal stability. These results suggested that the thermal stability of the insoluble silk scaffolds could be effectively regulated by changing the crystal compositions of the scaffolds in annealing processes. Scaffolds used in tissue repairs should have corresponding degradation behaviors that are suitable for specific tissue regeneration. To functionally assess the differences of the annealed scaffolds in vitro, degradation of the water-insoluble silk scaffolds treated by different annealing processes was assessed. After incubation in protease XIV solution at 37 ◦ C for 24 h, the scaffolds showed different degradation behaviors (Fig. 5A). Similar to the relation between the secondary structures and thermal stability of the scaffolds, the

34

G. Lu et al. / Colloids and Surfaces B: Biointerfaces 120 (2014) 28–37

Fig. 6. Amniotic fluid-derived stem cell growth on different silk scaffolds: S–S, silk scaffold prepared with salt-leaching process; WA-S, nanofibrous scaffold treated through water annealing process; M30-S, nanofibrous scaffold treated through water/methanol annealing process with 30% methanol; M80-S, nanofibrous scaffold treated through immersing in 80% methanol solution for 30 min. Error bars represent mean ± standard deviation with N = 5 (* p < 0.05).

Fig. 5. Enzymatic degradation behaviors (A) and compressive modulus (B) of the nanofibrous scaffolds after various annealing processes: (a) water annealing treatment, WA-S; (b) water/methanol annealing treatment with 10% methanol, M10-S; (c) water/methanol annealing treatment with 30% methanol, M30-S; (d) water/methanol annealing treatment with 50% methanol, M50-S and (e) methanol treatment by immersing the sample in 80% methanol solution for 30 min, M80-S.

degradation rate decreased following the increase of silk II structure, in which the water-annealed scaffolds totally degraded within 10 h while the weight loss from the methanol-annealed scaffolds was about 25% after 24 h enzyme exposure. Although further investigation is necessary to clarify the degradation behaviors of the scaffolds in vivo, our present study implies an effective way of regulating the degradation behaviors of ECM-biomimetic silk scaffolds to satisfy the requirements of different tissue repairs. The mechanical properties of the scaffolds also have significant influence on cell behaviors and tissue regenerations. The mechanical properties of the annealed scaffolds were measured in the wet state (Fig. 5B). Following the increase of silk II content in these nanofibrous scaffolds, the modulus of the scaffolds gradually increased from 8 kPa to 25 kPa. The results indicated that tunable mechanical properties could also be achieved through different annealing processes, which would satisfy the specific requirements of different tissue regenerations. 3.4. Cell culture and evaluation The cell culture in vitro was used to assess the influence of nanofibrous structures and secondary conformational compositions of the scaffolds on cell behaviors. After annealed by water, 30% methanol solution and 80% methanol solution, the scaffolds

achieved various silk I and silk II contents and were used to compare the influence of secondary structures on cell behaviors. Because lyophilized silk scaffolds derived from silk nanoparticle solutions generally formed separate sheets rather than porous structures, it is difficult to prepare silk porous scaffolds without nanofibers prepared through lyophilization process. Therefore, salt-leached silk porous scaffolds without nanofibers (Fig. S2) were used as control to assess the influence of nanofibrous structure on the biocompatibility of the scaffolds. On the other hand, since amniotic fluid stem cells show immunosuppressive activity and multilineage differentiation potential into various tissue types such as skin, cartilage, cardiac tissue, nerves, muscle and bone, amniotic fluid stem cells were chosen to analyze the biocompatibility of the scaffolds [46]. Mammalian cells response to lots of growth factors that regulate their function including proliferation, differentiation, and apoptosis. Both autocrine growth mechanisms (in which the individual cell synthesizes and/or secretes a substance that stimulates that same cell type to undergo a growth response) and paracrine growth mechanisms (in which the individual cells responding to the growth factor synthesize and/or secrete a substance that stimulates neighboring cells of another cell type) are important in cell growth in vivo and in vitro. The low cell seeding density generally results in the decrease of concentration of substance secreted through autocrine and/or paracrine, which could inhibit the cell growth at the initial period after seeding. Therefore, different cell growth behaviors on scaffolds could be investigated at different cell seeding densities. Considering that the excellent cell-compatibility of silk-based scaffolds generally results in over-rapid growth of cells, lower initial cell seeding density compared to previous study was used to reduce the cell growth on the scaffolds [47]. DNA content results showed different cell proliferation behaviors on the different scaffolds (Fig. 6). We noted that a few cells were observed on the scaffolds within 6 days, which was due to the low initial cell seeding density. More importantly, significant higher cell numbers were found on the nanofibrous scaffolds treated with different processes after day 6. Considering the numbers of the cell on all the scaffolds at day 1 were almost same, the higher cell numbers on the nanofibrous scaffolds indicated that better cell compatibility was achieved following improved nanofibrous architecture. Confocal microscopy and SEM results confirmed the better cytocompatibility of the nanofibrous scaffolds (Figs. 7 and 8). All the nanofibrous silk scaffolds maintained whole architecture after the

G. Lu et al. / Colloids and Surfaces B: Biointerfaces 120 (2014) 28–37

35

Fig. 7. Confocal microscopy of amniotic fluid-derived stem cells cultivated on different scaffolds at day 1 and 12: (a) silk scaffold prepared with salt-leaching process, day 1; (b) silk scaffold prepared with salt-leaching process, day 12; (c) nanofibrous scaffold treated through water annealing process, day 1; (d) nanofibrous scaffold treated through water annealing process, day 12; (e) nanofibrous scaffold treated through water/methanol annealing process with 30% methanol, day 1; (f) nanofibrous scaffold treated through water/methanol annealing process with 30% methanol, day 12; (g) nanofibrous scaffold treated through immersing in 80% methanol solution for 30 min, day 1; (h) nanofibrous scaffold treated through immersing in 80% methanol solution for 30 min, day 12.

36

G. Lu et al. / Colloids and Surfaces B: Biointerfaces 120 (2014) 28–37

Fig. 8. SEM images of amniotic fluid-derived stem cells cultivated on different scaffolds at day 1 and 12: (a) silk scaffold prepared with salt-leaching process, day 1; (b) silk scaffold prepared with salt-leaching process, day 12; (c) nanofibrous scaffold treated through water annealing process, day 1; (d) nanofibrous scaffold treated through water annealing process, day 12; (e) nanofibrous scaffold treated through water/methanol annealing process with 30% methanol, day 1; (f) nanofibrous scaffold treated through water/methanol annealing process with 30% methanol, day 12; (g) nanofibrous scaffold treated through immersing in 80% methanol solution for 30 min, day 1; (h) nanofibrous scaffold treated through immersing in 80% methanol solution for 30 min, day 12.

in vitro cell culture process. At day 1, only a few cells were observed on the nanofibrous scaffolds. After day 12, cells proliferated significantly and interacted to form aggregates on the nanofibrous scaffolds while maintained separate states on the salt-leached scaffolds. More cell-ECM monolayer structures formed on the surface of the nanofibrous scaffolds, implying better ECM formation on the scaffolds. The cell culture in vitro validated that ECM-biomimetic structure fabrication provided better microenvironment for cells,

which would further improve the biocompatibility of silk-based scaffolds. For the nanofibrous silk scaffolds with different secondary structures, the cell growth slightly decreased firstly and then increased following the addition of silk II structure, resulting in similar cell growth behaviors on the water-annealed scaffolds and methanol-annealed scaffolds. Similar to previous studies, the results indicated that the secondary structural changes could also affect the cytocompatibility of silk biomaterials [45]. However,

G. Lu et al. / Colloids and Surfaces B: Biointerfaces 120 (2014) 28–37

with the changes of the secondary structures, some different key influencing factors for cytocompatibility of biomaterials including mechanical properties, degradation behaviors or hydrophilic properties generally alter, making it difficult to elucidate the relation between the secondary structures and cytocompatibility of silk scaffolds. More importantly, the cytocompatibility had no significant difference between the nanofibrous scaffolds with different secondary structures, which confirmed that the formation of nanofibrous structures was the main factor in improving the cell growth on silk scaffolds. Although further study is still necessary to clarify the influence of secondary structures on biocompatibility, our present results provide a feasible way to design silk scaffolds with tunable properties and improved biocompatibility, which would facilitate the applications of silk in different tissue regenerations. 4. Conclusions The ECM-biomimetic porous scaffolds composed of silk nanofibers were prepared by slowly concentrating-lyophilizing processes. Then the secondary conformational compositions of the scaffolds were regulated through modulated water/methanol annealing methods to achieve water-stability, tunable degradation behaviors and mechanical properties. Following the formation of nanofibrous structure, the scaffolds showed better biocompatibility than the salt-leached silk scaffolds. Considering the improvement of biocompatibility as well as the control for different properties of the scaffolds, our present study provides a promising way to design silk-based scaffolds that are more suitable for different tissue regenerations. Acknowledgments We thank National Basic Research Program of China (973 Program 2013CB934400), and NSFC (21174097, 81272106) for support of this work. We also thank the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD), the Excellent Youth Foundation of Jiangsu Province (BK2012009), the NIH (EB002520), and the Key Natural Science Foundation of the Jiangsu Higher Education Institutions of China (11KGA430002) for support of this work. Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at http://dx.doi.org/10.1016/j.colsurfb. 2014.03.027. References [1] Y. Huang, X.F. Duan, Q.Q. Wei, C.M. Lieber, Science 291 (2001) 630–633. [2] A. Vasconcelos, A.C. Gomes, A. Cavaco-Paulo, Acta Biomater. 8 (2012) 3049–3060. [3] H. Shin, S. Jo, A.G. Mikos, Biomaterials 24 (2003) 4353–4364. [4] Q. Lu, X.L. Wang, S.Z. Lu, M.Z. Li, D.L. Kaplan, H.S. Zhu, Biomaterials 32 (2011) 1059–1067. [5] X.H. Zhang, M.R. Reagan, D.L. Kaplan, Adv. Drug Delivery Rev. 61 (2009) 988–1006.

37

[6] H. Perry, A. Gopinath, D.L. Kaplan, L.D. Negro, F.G. Omenetto, Adv. Mater. 20 (2008) 3070–3072. [7] T.P. Knowles, A.W. Fitzpatrick, S. Meehan, H.R. Mott, M. Vendruscolo, C.M. Dobson, M.E. Welland, Science 318 (2007) 1900–1903. [8] X.H. Liu, X.B. Jin, P.X. Ma, Nat. Mater. 10 (2011) 398–406. [9] B. Kundu, R. Rajkhowa, S. Kundu, C. Subhas, X. Wang, Adv. Drug Delivey Rev. 65 (2013) 457–470. [10] D.Y. Yao, S. Dong, Q. Lu, X. Hu, D.L. Kaplan, B.B. Zhang, H.S. Zhu, Biomacromolecules 13 (2012) 3723–3729. [11] C. Correia, S. Bhumiratana, L.P. Yan, A.L. Oliveira, J.M. Gimble, D. Rockwood, D.L. Kaplan, R.A. Sousa, R.L. Reis, G. Vunjak-Novakovic, Acta Biomater. 8 (2012) 2483–2492. [12] N. Kasoju, U. Bora, Adv. Healthcare Mater. 1 (2012) 393–412. [13] G.H. Altman, F. Diaz, C. Jakuba, T. Calabro, R.L. Horan, J.S. Chen, H. Lu, J. Richmond, D.L. Kaplan, Biomaterials 24 (2003) 401–416. [14] Y.F. Zhang, W. Fan, Z.C. Ma, C.T. Wu, W. Fang, G. Liu, Y. Xiao, Acta Biomater. 6 (2010) 3021–3028. [15] Y. Wang, E. Bella, S.D. Lee-Christopher, C. Migliaresi, L. Pelcastre, Z. Schwartz, D.B. Barbara, M. Antonella, Biomaterials 31 (2010) 4672–4681. [16] S. Huang, X.B. Fu, J. Controlled Release 142 (2010) 149–159. [17] D.G. Harkin, K.A. George, P.W. Madden, I.R. Schwab, D.W. Hutmacher, T.V. Chirila, Biomaterials 32 (2011) 2445–2458. [18] R. Rajkhowa, E.S. Gil, J. Kluge, K. Numata, L.J. Wang, X.G. Wang, D.L. Kaplan, Macromol. Biosci. 10 (2010) 599–611. [19] S. Das, F. Pati, S. Chameettachal, S. Pahwa, A.R. Ray, S. Dhara, S. Ghosh, Biomacromolecules 14 (2013) 311–321. [20] L.S. Wray, J. Rnjak-kovacina, B.B. Mandal, D.F. Schmidt, E.S. Gil, D.L. Kaplan, Biomaterials 33 (2012) 9214–9224. [21] D.M. Lin, H. Tao, J. Trevino, J.P. Mondia, D.L. Kaplan, F.G. Omenetto, L. Dal-Negro, Adv. Mater. 24 (2012) 6088–6093. [22] J.X. Zhu, Y.P. Zhang, H.L. Shao, X.C. Hu, Polymer 49 (2008) 2880–2885. [23] Y.H. Ruan, H. Lin, J.R. Yao, Z.G. Chen, Z.Z. Shao, Macromol. Biosci. 11 (2011) 419–426. [24] R. Nazarov, H.J. Jin, D.L. Kaplan, Biomacromolecules 5 (2004) 718–726. [25] U.J. Kim, J. Park, H.J. Kim, M. Wada, D.L. Kaplan, Biomaterials 26 (2005) 2775–2785. [26] Q. Lv, Q.L. Feng, J. Mater. Sci. Mater. Med. 17 (2006) 1349–1356. [27] Y. Tamada, Biomacrommolecules 6 (2005) 3100–3106. [28] X.H. Zhang, C.B. Baughman, D.L. Kaplan, Biomaterials 29 (2008) 2217–2227. [29] H. Cao, X. Chen, J.R. Yao, Z.Z. Shao, J. Mater. Sci. 48 (2013) 150–155. [30] B.B. Mandal, S.C. Kundu, Biomaterials 30 (2009) 2956–2965. [31] Q. Lu, X.H. Zhang, X. Hu, D.L. Kaplan, Macromol. Biosci. 10 (2010) 289–298. [32] S.M. Bai, S.S. Liu, C.C. Zhang, W.A. Xu, Q. Lu, H.Y. Han, D.L. Kaplan, H.S. Zhu, Acta Biomater. 9 (2013) 7806–7813. [33] Q. Lu, Y.L. Huang, M.Z. Li, B.Q. Zuo, S.Z. Lu, J.N. Wang, H.S. Zhu, D.L. Kaplan, Acta Biomater. 7 (2011) 2394–2400. [34] Q. Lu, H.S. Zhu, C. Zhang, F. Zhang, B. Zhang, D.L. Kaplan, Biomacromolecules 13 (2012) 826–832. [35] T. Asakura, A. Kuzuhara, R. Tabeta, H. Saito, Macromolecules 18 (1985) 1841–1845. [36] S. Sohn, H.H. Strey, S.P. Gido, Biomacromolecules 5 (2004) 751–757. [37] K. Numata, P. Cebe, D.L. Kaplan, Biomaterials 31 (2010) 2926–2933. [38] T. Xu, W.X. Zhao, J.M. Zhu, M.Z. Albanna, J.J. Yoo, A. Atala, Biomaterials 34 (2013) 130–139. [39] S. Bollini, M. Pozzobon, M. Nobles, J. Riegler, X.B. Dong, M. Piccoli, K.K. Cheung, A. Cabrelle, P.R. O’Mahoney, E. Cozzi, S. Sartore, A. Tinker, M.F. Lythgoe, P. DeCoppi, Stem Cell Rev. Rep. 7 (2011) 364–380. [40] I. Antonucci, A. Pantalone, S. Tete, V. Salini, C.V. Borlongan, D. Hess, L. Stuppia, Curr. Pharm. Des. 18 (2012) 1846–1863. [41] P. De Coppi, G. Bartsch, M.M. Siddiqui, T. Xu, C.C. Santos, L. Perin, G. Mostoslavsky, A.C. Serre, E.Y. Snyder, J.J. Yoo, M.E. Furth, S. Soker, A. Atala, Nat. Biotechnol. 25 (2007) 100–106. [42] S. Dutta, B. Talukdar, R. Bharal, R. Rajkhowa, D. Dev, Biopolymers 99 (2013) 326–333. [43] P.H. Chao, S. Yodmuang, X. Wang, L. Sun, D.L. Kaplan, G. Vunjak-Novakovic, J. Biomed. Mater. Res. B: Appl. Biomater. 95 (2010) 84–90. [44] Q. Lu, X. Hu, X.Q. Wang, J.A. Kluge, S.Z. Lu, P. Cebe, D.L. Kaplan, Acta Biomater. 6 (2010) 1380–1387. [45] H.J. Jin, J. Park, V. Karageorgiou, U.J. Kim, R. Valluzzi, D.L. Kaplan, Adv. Funct. Mater. 15 (2005) 1241–1247. [46] N.H. Kang, K.A. Hwang, S.U. Kim, Y.B. Kim, S.H. Hyun, E.B. Jeung, K.C. Choi, Cancer Gene Ther. 19 (2012) 517–522. [47] S. Hofmann, K.S. Stok, T. Kohler, A.J. Meinel, R. Muller, Acta Biomater. 10 (2014) 308–317.