- Email: [email protected]
Functionalized silk fibroin dressing with topical bioactive insulin release for accelerated chronic wound healing
Materials Science and Engineering C 72 (2017) 394–404
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Functionalized silk fibroin dressing with topical bioactive insulin release for accelerated chronic wound healing Xiufang Li a, Yan Liu b, Jian Zhang b, Renchuan You c, Jing Qu a, Mingzhong Li a,⁎ a b c
National Engineering Laboratory for Modern Silk, College of Textile and Clothing Engineering, Soochow University, No. 199 Ren'ai Road, Industrial Park, Suzhou 215123, China Department of Burn and Plastic Surgery, School of Medicine Affiliated Ruijin Hospital, Shanghai JiaoTong University, Shanghai 200025, China College of Textile Science and Engineering, Wuhan Textile University, Wuhan 430200, China
a r t i c l e
i n f o
Article history: Received 6 October 2016 Received in revised form 10 November 2016 Accepted 21 November 2016 Available online 24 November 2016 Keywords: Silk fibroin Insulin release Microparticles Dressings Wound healing
a b s t r a c t The healing of chronic wounds remains a key challenge in regenerative medicine. To promote wound healing, a bioactive wound dressing is required. In this study, a functionalized silk fibroin dressing with topical bioactive insulin release was prepared for the treatment of chronic wounds. For this purpose, insulin-encapsulated silk fibroin (SF) microparticles were prepared by coaxial electrospraying of aqueous SF solution under mild processing conditions. Insulin was successfully encapsulated in the inner layer of SF microparticles, providing a sustained insulin release for up to 28 days. It was found that the insulin released from the microparticles could maintain original molecular conformation. Moreover, the cell migration assay based on human keratinocyte and endothelial cells confirmed that the insulin released from SF microparticles retained its native bioactivity. Furthermore, the insulin-encapsulated microparticles were loaded into a SF sponge, functioned as a bioactive wound dressing, and the in vivo therapeutic effect of the sponge dressing was evaluated on dorsal full thickness wounds of diabetic Sprague-Dawley rats. The results showed that an insulin-functionalized SF dressing accelerated wound closure, collagen deposition and vascularization, thus, significantly promoting wound healing. The insulin-functionalized SF dressing provides new treatment options for chronic wounds. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Chronic cutaneous wounds are characterized by the absence of healing after 6 weeks, which is commonly seen as a consequence of diabetes mellitus and vascular compromise [1,2]. The classic treatment is the debridement of the wound followed by its compression with sterile gauze. When this method is not effective enough in some chronic wounds, the dressings based on adequate biomaterials are frequently used to promote healing [1]. However, treatment currently focuses on dressings that prevent microbial infiltration and keep a balanced moisture and gas exchange environment. To heal some chronic wounds, a bioactive dressing is frequently required. For example, in the cutaneous wound in diabetic foot ulcers, reepithelialization and angiogenesis are powerless. Therefore, functionalized wound dressings with sustained bioactive drug release are needed to offer a stimulatory function for facilitating tissue repair, thus, promoting wound healing and minimizing the recovery period [3,4]. Recently, increasing evidences demonstrated that insulin contributes to wound healing [5–7]. It was reported that insulin accelerated reepithelialization of provisional tissue by stimulating the migration and proliferation of keratinocyte [4,8,9]. In addition, insulin can ⁎ Corresponding author. E-mail address: [email protected] (M. Li).
http://dx.doi.org/10.1016/j.msec.2016.11.085 0928-4931/© 2016 Elsevier B.V. All rights reserved.
stimulate the migration and tube formation of endothelial cells which helps to improve angiogenesis during wound healing [9,10]. However, the major problem of topical administration of peptides is their short half-life and loss of bioactivity in the peptidase-rich wound environment [11]. An alternative strategy to overcome this problem is the use of biocompatible wound dressings for sustained delivery of insulin. Drug delivery using biodegradable microspheres is a promising approach for sensitive biologicals, and insulin has been encapsulated into polymer microparticles to establish a sustained delivery system [4,6, 12,13]. However, most methods for the preparation of insulin delivery microparticles, such as emulsification and the solvent extraction method [9,12], need processing in organic solvents, at extreme pH values, or mechanical stress, potentially challenging the bioactivity of insulin [14, 15]. Unlike other small-molecule drugs, insulin possessing a complex molecular conformation for its biological activity is more susceptible to salts, organic solvents and high temperature, which may cause a loss of bioactivity [16]. In this regard, the retention of bioactivity of insulin released from dressing remains the key challenge. As a natural protein, silk fibroin (SF) has been explored for various tissue engineering applications due to excellent biocompatibility and tailorable degradability [17,18]. SF materials showed excellent bioresponses in vivo with low immunogenicity for numerous clinical applications [19]. SF biomaterials have been applied as wound healing dressings in diverse structural forms such as film, sponge and
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
electrospun fibers, which supported cell growth of fibroblasts, keratinocytes and endothelial cells, and showed a positive effect for wound healing [18–24]. In addition, silk is generated in an aqueous state and is thus readily miscible with other protein solutions. Its aqueous solubility and processability under very mild conditions makes it an attractive material for the loading of sensitive biologics to avoid loss of bioactivity of the drugs to be delivered [25,26]. SF-based biomaterials are being considered to address a wide range of stabilization challenges, from labile enzymes to antibodies [27]. Glucose oxidase, lipase, and horseradish peroxidase were entrapped in SF films over 10 months and significant activity was retained, even when stored at 37 °C [27]. The SF microparticles loaded with insulin-like growth factor 1 demonstrated controlled and sustained release over 7 weeks in bioactive form [26]. Therefore, we hypothesized that insulin-encapsulated SF microparticles are able to provide a sustained release of bioactive insulin, and the microparticle-loaded SF dressing can promote the migration of skin repair cells and the reconstruction of a microvascular network for healing chronic wounds. In the present study, we prepared insulin-encapsulated SF microparticles by coaxial electrospraying and freeze-drying. This method proposed to prepare SF microparticles under very mild conditions from aqueous SF solutions, which facilitates maintaining the bioactivity of insulin. Then, the insulin-encapsulated microparticles that offered a topical and sustained release system would be incorporated into SF sponges via multilayer loading, functioning as a bioactive wound dressing for chronic wounds. The molecular conformation of insulin released from microparticles was detected by FTIR and CD, and the bioactivity of released insulin was assessed by the scratch assay of the immortal human keratinocyte (HaCaT) and the human endothelial cell (EA. hy926). Furthermore, the in vivo therapeutic effect of SF dressings, including healing rate and vascularization, was evaluated on dorsal full thickness wounds of diabetic Sprague-Dawley (SD) rats.
395
theoretical drug amount relative to entire microparticles. Eq. (1) was used to evaluate insulin loading content. Insulin loading ð%Þ Weight of the insulin in micropaticles ðmgÞ ¼ 100 Weight of the micropaticles ðmgÞ
ð1Þ
The loading ratio of insulin in the coaxial microparticles was about 16.7% according to the above equation. As a control, the pure SF and insulin microparticles were prepared, respectively, the solution delivering rate was 0.4 mL/h and the other conditions remained consistent. 2.3. Preparation of microparticle-loaded SF sponge dressing
2. Materials and methods
As shown in Fig. 1 (B), a microparticle-loaded SF sponge dressing was prepared by multilayer loading and freeze-drying method. The SF solution was diluted to 2.0 wt%, and then 2-morpholinoethanesulfonic acid (MES), N-hydroxysuccinimide (NHS) and 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) (Sigma-Aldrich) were added to the SF solution at 20%, 10% and 20% of the SF weight in solution, respectively. The mixed solution was stirred slowly and reacted in an ice bath for 1 h. 900 μL of EDC-activated SF solution was added into a round metallic box (D = 40 mm, H = 15 mm, V = 10 mL), and frozen at −80 °C for 30 min. Then 10 mg of SF microparticles were evenly laid on a frozen SF solution layer, and frozen at −80 °C for 15 min. On top of it, 700 μL of EDC-activated SF solution was poured to cover the microparticles and frozen at − 80 °C for 15 min. Finally, 10 mg of microparticles were added and frozen, and subsequently, 900 μL of EDC-activated SF solution was poured according to the above process. Then the SF mixture composite with two layers of microparticles were frozen at −80 °C for 4 h, and further lyophilized to obtain a microparticle-loaded sponge dressing. The dressings were sterilized with γ-ray irradiation and stored at 4 °C.
2.1. Preparation of SF solution
2.4. Scanning electron microscopy
Regenerated SF solution was prepared following the procedure described previously [18]. Briefly, Bombyx mori silk fibers (Huzhou, China) were degummed three times in 0.05% Na2CO3 solution at 98– 100 °C for 30 min and dried at 60 °C after thoroughly rinsing. The extracted silk fibroin was dissolved in a ternary solvent of CaCl2:CH3CH2OH:H2O (1:2:8 M ratio) at 72 ± 2 °C for 1 h. SF solution was obtained after dialysis with cellulose membranes (MWCO 9– 12 kDa) in deionized water for 4 days.
The morphology of samples was observed by a scanning electron microscopy (SEM, S-4800, Hitachi, Japan). The size of microparticles was analyzed on the basis of SEM images with the Nano Measurer analysis software (Department of Chemistry, Fudan University, China. Copyright: (C) 2008 Jie Xu). To calculate the diameter of microparticles, we measured the average diameter of a total of 100 microparticles based on using SEM images. 2.5. In vitro insulin release
2.2. Preparation of SF microparticles 100 mg insulin (27.5 IU/mg; from porcine, WangBang Bochemical Pharmaceutical Co., Ltd., Xuzhou, China) was dissolved in 10 mL of 0.01 M HCl, then the pH of the solution was adjusted to 7.0 ± 0.1 using 0.1 M NaOH. SF solution was diluted to 2.0 wt% and mixed with glycerol at 30 wt% of SF weight. The resulting insulin solution (1.0 wt%) and SF solution (2.0 wt%) were used as core and shell for coaxial electrospraying, respectively. Fig. 1 (A) shows the preparation process for insulin-encapsulated SF microparticles. The coaxial nozzle has an inner capillary of 0.6 mm (inner diameter) and an outer capillary of 1.2 mm (inner diameter). Two syringe pumps (KDS100, KD Scientific, USA) deliver core and shell layer solutions at the rates of 0.1 mL/h and 0.3 mL/h, respectively. A high-voltage power supply supplied a 13 kV high voltage between the nozzle and the collection box filled with liquid nitrogen, and the distance is 12 cm. The collected insulin-encapsulated SF microparticles were lyophilized by a Virtis Genesis 25-LE Freeze Dryer for 48 h and then balanced at 25 °C and RH 90% for 24 h to obtain the water-insoluble microparticles. Drug loading ratio was expressed as the percentage of
Insulin was labeled with fluorescein isothiocyanate (FITC) (SigmaAldrich) as previously described [28]. 300 μL of FITC solution (10 mg/mL in dimethylsulfoxide) was added to 10 mL of insulin solution (15 mg/mL in bicarbonate buffer, pH = 8.5, 0.1 M) and stirred at room temperature for 60 min. Next, 200 μL of 1 M hydroxyl ammonium chloride solution was added and stirred for 10 min at room temperature. The insulin was then purified using a 10 mm × 300 mm column with Sephadex G-50 equilibrated in 0.1 M sodium bicarbonate buffer (pH = 8.5) to remove any unreacted FITC. FITC-insulin encapsulated SF microparticles were prepared according to the condition found in Section 2.2. The fluorescent images were captured using an inverted fluorescence microscope (Olympus IX71, Japan). Physically absorbed microparticles were used as a control group. 10 mg of pure SF microparticles was directly immersed into 2 mL of FITC-insulin solution to adsorb insulin by permeating and physical absorption, and named as insulinadsorbed microparticles. The remaining amount of insulin in solution was quantified to calculate the adsorption amount using a fluorescence spectrophotometer (FM4P TCSPC, Horiba Jobin Yvon), and the loading ratio in the insulin-adsorbed microparticles was about 5.5%.
396
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
Fig. 1. The schematic diagram of the preparation process for: (A) SF microparticles and (B) the microparticle-loaded SF sponge dressing.
The insulin release profiles were determined by immersing 10 mg of microparticles and a sponge dressing containing 10 mg of microparticles in 10 mL of phosphate buffer solution (PBS, 10 mM, pH 7.4). The samples were incubated in a water bath at 37 °C with a 100 rpm shaking. 1 mL of medium was collected and replaced with an equal volume of fresh PBS after centrifugation at predetermined time points. The amount of released insulin was quantified using a fluorescence spectrophotometer.
2.7. Circular dichroic spectroscopy To detect the conformational stability of released insulin, the secondary structure of released insulin solutions were analyzed after 2 days release with circular dichroism (CD) spectra. CD spectra were recorded on a spectrophotometer (J-815, JASCO) equipped with a Peltier temperature controller. Spectra were obtained from 250 to 195 nm at a scanning speed of 100 nm·min−1 at 25 °C. The data was recorded as mean residue ellipticity (deg·cm2·dmol−1).
2.6. Fourier transform infrared spectroscopy 2.8. Cell scratch assay To detect whether microparticle preparation processing has an effect on the molecular conformation of insulin, the insulin particles were prepared according to the preparation methods of insulin encapsulated SF microparticles. The lyophilized insulin microparticles were cut into powder and then prepared in KBr pellets. Native insulin powder was used as a control. FTIR spectra were performed with a Nicolet 5700 spectrometer (Nicolet Company, USA). The percentage of different secondary structures in insulin microparticles and native insulin powder was analyzed by PeakFit v4.12 software (Systat Software Inc., USA) from the amide I region [29].
To evaluate the bioactivity of released insulin, a cell scratch assay was conducted following the procedure described previously [30]. The immortal human keratinocyte line HaCaT (presented by Professor S. Zhang at the Basic Medical College of Soochow University) and the human endothelial cell line EA. hy926 (ATCC, USA) were cultured in Dulbecco's modified Eagle medium (DMEM, low glucose, Gibco, USA) supplemented with 10% fetal bovine serum (FBS, Gibco, USA) and 1% streptomycin-penicillin (Gibco, USA). The medium was replaced every 3 days, and cell cultures were maintained in a humidified incubator at 37 °C and 5% CO2. After reaching 80% confluence, cells were detached
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
from the petri dish and seeded on 6 well plates at a density of 1 × 104 cells/well in 1 mL, respectively. Cells were allowed to reach confluency after 48 h, at which time a single scratch was made using the tip of a plastic disposable 10 μL pipette tip. All wells were washed twice with 1 mL of DMEM culture medium to remove cellular debris. Then, three groups of culture mediums were used for further cell culture: (a) supernatants with insulin released from microparticles within 2 days at a concentration of 10−7 M, (b) insulin solution from native insulin powder at a concentration of 10−7 M and (c) without insulin. The distance across the scratch at 0 h (d0) and 24 h (d1) was measured under a phase contrast inverted microscope (Olympus IX81, Japan). The relative migration distance (%) across the scratch was calculated as follows: Relative migration distance ð%Þ ¼
d0 −d1 100 d0
397
10 mm diameter) were created on the upper back of each rat using a pair of sharp scissors and a scalpel after being anesthetized with pentobarbital sodium (30 mg/kg). SF dressings were immersed for 24 h in sterilized PBS solution to remove crosslinker residue prior to in vivo experiments. Some insulin loss is inevitable during immersion, and the loss is about 6.4% of total insulin loading in the dressing within 24 h. The wound was covered with insulin-loaded SF sponge dressing, SF sponge dressing and gauze, respectively. The entire wound area was then wrapped with sterile gauze and fixed with an elastic bandage. After surgery, each rat was caged individually with free access to water and food, and the dressing was daily wetted by topical normal saline application. Five rats were sacrificed and the skin tissues surrounding the wound were excised to evaluate wound repair at 1, 2 and 3 weeks, respectively. The degree of wound healing was determined via measuring the area of wound closure. 2.10. Histology and immunohistochemical staining
2.9. Diabetic rat model and wound healing examination All animal experiments were conducted in accordance with the Management Ordinance of Experimental Animal of China ((2001) No. 545) and approved by the Jiangsu Province in experimental animals management rules ((2008) No. 26). Twenty SD rats (180–200 g, SPF grade, male) were used in this experiment, and type I diabetic rats were induced by a single intraperitoneal injection of 60 mg/kg streptozotocin (Sigma-Aldrich) dissolved in sodium citrate buffer solution (pH ~ 4.5) [31,32]. After 4 days, the blood glucose levels of rats were checked using a glucometer (Accu-Chek Aviva, Roche). Fifteen rats with blood glucose levels higher than 16.7 mM, which were considered as a successful diabetic model. Three full thickness wounds (approximately
Specimens were harvested and immediately fixed in 4% formaldehyde in PBS at 4 °C, and then embedded in paraffin for hematoxylineosin (HE) and Masson staining. Wounds were assessed by multiple serial tissue sections, and the stained sections were observed with a light microscope (Olympus IX51, Japan). For immunohistochemical staining, the paraffin sections (5 μm) were de-paraffinized, washed three times in PBS for 5 min, and then blocked with 5% serum for 30 min. The slides were subsequently incubated with primary antibody mouse anti-CD31 monoclonal antibody (Abcam) at 4 °C overnight. After rinsing three times with PBS, the slides were incubated with secondary antibody at 37 °C for 30 min, and further developed with 3,3′-diaminobenzidine tetrahydrochloride (DAB) solution and counterstained with hematoxylin.
Fig. 2. The SEM of microparticles and sponge dressing. (A–C) SF microparticles, (D–F) insulin-encapsulated SF microparticles, and (G, H) microparticle-loaded SF sponge dressing. (B, C, E, F, G, H) Cross-sectional views of SF microparticles, insulin-encapsulated SF microparticles and dressing, respectively. Scale bars: (A, D) 1 mm, (B, E, H) 50 μm, (C, F) 5 μm, (G) 200 μm.
398
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
Fig. 3. (A) Cumulative FITC-insulin release from microparticles and dressing: (a) insulin-adsorbed SF microparticles, (b) insulin-encapsulated SF microparticles, and (c) microparticleloaded SF dressings. (B) The fluorescent image of FITC-insulin-encapsulated SF microparticles. (C) SEM image of insulin-encapsulated SF microparticles after 28 days immersion in PBS. Scale bars: (B) 300 μm, (C) 100 μm.
2.11. Statistical analysis Data are presented as means ± SD. Statistical comparisons were performed using ANOVA (t-test), and differences at p b 0.05 were considered as statistically significant.
3. Results 3.1. Morphological observation As shown in Fig. 2 (A), the SF microparticles were close to spherical in shape, which showed an average diameter of 94.6 ± 27.7 μm by statistical calculation. The surface of microparticles was covered by a dense skin layer with a few micropores, whereas the interior consisted of a dense core layer and an outer layer with a porous structure (Fig. 2 (B)). Unlike the core layer, the porous outer layer contained a large number of nanofibers (5–150 nm diameter), nanoparticles (5–300 nm diameter) and a sheet-like structure (5–150 nm thickness), along the radial direction of the microparticles (Fig. 2 (C)). Fig. 2 (D) showed the morphology of insulin-encapsulated SF microparticles, which has a similar surface morphology and average diameter (92.9 ± 25.0 μm) with pure SF microparticles. The porous outer layer consisted of a large number of sheet-like structures (5–150 nm thickness), and some
thin fibers (5–150 nm diameter) and particles (100–1200 nm diameter) could be also observed (Fig. 2 (E)). However, numerous particles (200– 1800 nm) formed in the core layer (Fig. 2 (E, F)). These results suggest that the SF in the shell layer solution is able to assemble into nanoparticles, and further assemble into nanofibers or a sheet-like structure during the electrospraying process, while the insulin as a core layer solution is easy to assembly into micro/nano particles. Furthermore, to load insulin-encapsulated SF microparticles into an SF sponge to form a functionalized dressing, the microparticles were embedded within the SF sponge by multilayer loading and freezing. As shown in Fig. 2 (G, H), the SF microparticles were tightly embedded within the pores and evenly distributed in SF sponges. 3.2. In vitro insulin release The cumulative FITC-insulin release profiles of SF microparticles and microparticle-loaded SF dressings are shown in Fig. 3 (A). Obviously, the insulin-adsorbed SF microparticles showed a serious burst release of 98.4% within the first 2 h (Fig. 3 (A-a)). The release behavior from insulin-encapsulated SF microparticles was triphasic consisting of (i) a typical burst release within the first 12 h, followed by (ii) a period of stable and continuous release until day 14, and finally (iii) a period of low release rates from these days on. As shown in Fig. 3 (A-b), the cumulative insulin release from insulin-encapsulated SF microparticles were 62.2%,
Fig. 4. (A) FTIR spectra of: (a) insulin microparticles prepared by electrospraying and (b) native insulin powder. (B) CD spectra of: (a) released insulin from SF microparticles and (b) native insulin.
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
399
Table 1 Secondary structure contents of insulin as determined by FTIR spectra. Unordered structures include random coils and turns. Samples
Unordered structure (%)
α-Helix (%)
β-Sheet (%)
Electrosprayed insulin microparticles Native insulin powder
29.1 30.6
41.3 42.2
29.6 27.2
77.3%, 86.8% and 91.1% at 2, 5, 9 and 14 days, respectively. From 14 to 40 days, release rate continuously reduced, and about 98.2% of insulin was released from insulin-encapsulated SF microparticles. The microparticle-loaded SF dressings exhibited a similar release behavior with insulin-encapsulated SF microparticles. About 16.2%, 39.8%, 66.4% and 81.8% cumulative insulin was released from dressing at 2, 5, 9 and 14 days, respectively (Fig. 3 (A-c)), and the cumulative insulin release content was about 90.7% at 28 days. These results indicated that the microparticle-loaded SF sponge dressings provided an effective system for sustained insulin release over a month-long period. 3.3. Conformational detection Compared to native insulin powder, the FTIR spectra of insulin microparticles prepared by electrospraying showed similar absorption peaks in Amide I (1658 cm−1) and Amide II (1541 cm−1), which contributed to α-helix or random coil (Fig. 4 (A-a, b)). The α-helix content is commonly used to evaluate whether the secondary structure of insulin changed [33]. As shown in Table 1, the result of the fit covering amide I region shows the contents of different secondary structures. The αhelix content of insulin microparticles prepared by electrospraying and native insulin powder were about 41.3% and 42.2%, respectively, suggesting that microparticle fabrication process had no significant effect on the secondary structure of insulin. Furthermore, the secondary structure of released insulin was determined by CD spectra. As shown in Fig. 4 (B-b), native insulin exhibits two minima at 208 and 223 nm, indicating a dominant α-helical structure. The released insulin showed a predominant α-helical structure identical to that of native insulin (Fig. 4 (B-a)), indicating that the released insulin from SF microparticles maintained the original molecular conformation. 3.4. Bioactivity of released insulin: cell scratch assay The bioactivity of released insulin was assessed using a cell monolayer scratch with HaCaT and EA. hy926 cells. As shown in Fig. 5 (a0, b0, c0), a scratch of approximately 400 μm in width was made on a confluent
layer of HaCaT. The marginal cells migrated into the scratch to impair the gap closure after 24 h (Fig. 5 (a1, b1, c1)). The scratch without insulin was closed at about 22.4 ± 3.1% after cell migration for 24 h (Fig. 5 (c0, c1, d)), whereas the scratch with released insulin from microparticles and native insulin were closed at about 46.0 ± 2.1% and 49.7 ± 2.9% after 24 h (Fig. 5 (a0, b0, a1, b1, d)), respectively, indicating that released insulin significantly improved HaCaT cell migration. For EA. hy926 cells, the scratch without insulin was closed at about 27.2 ± 5.1% after 24 h (Fig. 6 (c0, c1, d)), whereas the scratch with released insulin and native insulin were closed at about 57.0 ± 3.0% and 60.4 ± 6.5% (Fig. 6 (a0, b0, a1, b1, d)), respectively, indicating that released insulin was able to stimulate EA. hy926 cell migration.
3.5. Macroscopic observations The SF sponge dressings were covered on the wound of streptozotocin-induced diabetic rats to investigate the effects of released insulin on chronic wound healing in vivo. Streptozotocin selectively destroys pancreatic β-cells and causes type I diabetes after 2 or 3 days injection. Blood glucose was measured after streptozotocin injection, which reached 30 mM after 3 days. The establishment of diabetes led to a drastic weight loss of rats of around 25%. The images after wound treatment at different time points are provided in Fig. 7 (A, B, C, D), and the wound images were quantified to show the closure areas of each experimental group at different time points (Fig. 7 (E)). After 1 week, local skin ulcer and effusion was observed in the gauze-covered group (Fig. 7 (B-c1)), whereas wound scab was accelerated and inflammatory response was mild in the SF dressing-covered group (Fig. 7 (Ba1, B-b1)). The wound closure rate of the microparticle-loaded dressing (about 91.3%) was significantly higher than the other two groups after 2 weeks (Fig. 7 (C-a2, C-b2, C-c2, E)) (p b 0.05). At the third week, there was no wound evident in the insulin-loaded group, while small wounds at defect sites were still observed in the other two groups (Fig. 7 (D-a3, b3, c3)). The results indicated that insulin release accelerated the wound healing process.
Fig. 5. Scratch assay for the HaCaT cell migration measurement. (a0, b0, c0) Images of the distance across the scratch at 0 h, (a1, b1, c1) images of the distance across the scratch at 24 h. Scale bars: 300 μm. (d) Relative migration distance (%) after 24 h culture (** p b 0.01).
400
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
Fig. 6. Scratch assay for the EA. hy926 cell migration measurement. (a0, b0, c0) Images of the distance across the scratch at 0 h, (a1, b1, c1) images of the distance across the scratch at 24 h. Scale bars: 300 μm. (d) Relative migration distance (%) after 24 h culture (** p b 0.01).
3.6. Histological observation and immunohistochemistry Histological analyses of dorsal skin wounds stained with HE and Masson are shown in Figs. 8 and 9, respectively. At 1 week, the newly regenerated granulation tissue appeared at the marginal region of wounds as shown in Fig. 8 (a1, b1, c1). After 2 weeks, the granulation tissue was developed to the centre of the skin defect. The tissue neogenesis in the microparticle-loaded group was more significant than the other groups at 2 weeks after wounding (Fig. 8 (a2, b2, c2)), and more dermic papilla and hair follicle were found on the wound area at 3 weeks after wounding (Fig. 8 (a3, b3, c3)), demonstrating well-defined dermal-epidermal junctions and well-differentiated epidermis. These results suggested that the insulin-loaded dressing accelerated tissue neogenesis and maturation. Fig. 9 shows the deposition of collagen by Masson's staining at different time points. At the first week, although a slight deposition of collagen was detected in all groups (Fig. 9 (a1, b1, c1)), collagen deposition was more significant in the microparticle-loaded SF dressing group than in the control groups. The deposition of collagen increased and developed to the wound central area after 1 week, and the collagen deposition was significant in the periphery of wound in the microparticle-loaded group after 2 weeks (Fig. 9 (a2)). After 3 weeks, more
collagen deposited in the microparticle-loaded group, and the collagen deposition has been observed in the central area of the wound, and the alignment of collagen was more regular (Fig. 9 (a3)), indicating that the insulin-loaded dressing accelerated extracellular matrix deposition. Furthermore, the positive expression of CD31 was detected at different time points. At the early stage (1 and 2 weeks), the CD31 expression levels in the microparticle-loaded SF dressing group were significantly higher than the other two groups (Fig. 10 (d)), indicating accelerated neovascularization. The results suggested that topical insulin release promoted the vascularization degree of regenerated tissue. 4. Discussion Chronic wounds cause significant morbidity and mortality in many diseases such as diabetes and cardiovascular diseases. The healing of cutaneous wounds involves granulation tissue formation, reepithelialization and angiogenesis. A successful dressing for wound healing not only should be able to alleviate infection and inflammation by acting as a temporary barrier to protect the underlying tissue, but it also should promote reepithelialization and angiogenesis via stimulating the activities of repair cells [1,3]. Insulin can stimulate proliferation and migration of keratinocytes and endothelial cells in the wound
Fig. 7. (A–D) Macroscopic observation of skin wounds for different times: (a) insulin-loaded dressing, (b) SF dressing, (c) gauze, (subscripts 0, 1, 2, and 3) skin wounds after covering with dressings for 0, 1, 2 and 3 weeks. (E) Quantitative evaluation of wound closure at different time points (* p b 0.05).
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
401
Fig. 8. H&E images of wound after treatment for different times. White dotted lines indicated the obvious boundary of tissue neogenesis in wound beds. Scale bars: 1 mm.
healing process, promote tube formation of endothelial cells, and thus promote wound healing [8–10]. To achieve the insulin-functionalized dressing for chronic wound healing, in this work, insulin was encapsulated into SF microparticles and then incorporated with SF sponges to assess impact on wound healing.
The SF microparticle has been utilized as a promising material for drug release due to its biocompatibility, degradability, and the ability to control drug release kinetics [15,25]. In the present study, we fabricated SF microparticles using an aqueous encapsulation process based on coaxial electrospraying. When differentiated droplets fell into liquid
Fig. 9. Masson images of wound after treatment for different times. White dotted lines indicated the obvious boundary of collagen deposition in wound beds. Scale bars: 1 mm.
402
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
nitrogen, ice nuclei formed and grew into larger ice crystals; meanwhile the SF solution was concentrated from the outer part to the inner part. Thus, the inner part of SF microparticles was denser than the outer part (Fig. 2 (B)). It has been reported that SF was able to assemble to form nanospheres or nanofibers in an aqueous environment at low concentration [34,35]. The concentration of SF in the outer part of microparticles was significantly decreased after freezing. Consequently, the outer part in microparticles exhibited a sheet-like morphology with thin fibers and some nanoparticles (Fig. 2 (C)). The insulin-encapsulated SF microparticles had a similar diameter and surface morphologies with pure SF microparticles (Fig. 2 (D)). In the inner layer, however, more
large particles (100–1200 nm) were observed in insulin-encapsulated SF microparticles than pure SF microparticles (Fig. 2 (E, F)). Different from fibrous SF macromolecules, insulin is a small-molecule globular protein with a low molecular weight of 5808 Da, which may be easier to assemble into micro/nano particles. The results suggested that insulin was mainly encapsulated in the inner layer of the microparticles in the form of nanoparticles. In practice, the yield of this technique is a key for practical clinic application. Single-nozzle electrospraying has a low-throughput with a yield of milligrams/hour, but this can be overcome with multiple electrospray sources. Particle production could be increased from milligrams to grams per hour using 19 parallel nozzles
Fig. 10. (a, b, c) Immunohistochemical staining of CD31 expression at different time points. Scale bars: 250 μm. (d) Quantitative evaluation of CD31 expression at different time points (* p b 0.05).
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
[36]. For wound treatments, however, microparticles are difficult to distribute directly in a homogeneous way and retain over a wound surface. Therefore, the microparticles were tightly embedded within the SF sponge to form an even distribution in the SF sponge dressing after multilayer loading (Fig. 2 (G, H)). The drug release mechanism from SF materials is complex, and diffusion control or biodegradation alone is also unable to fully explain the release mechanism, thus, further studies are needed to fully understand the underlying release mechanism from SF materials [26]. Although the biodegradation of SF materials is slow, the biodegradability of SF may help to sustain the release of insulin from the microparticles [15,37]. In this work, the inner structure of insulin-encapsulated SF microparticles is a porous morphology and covered with a thin skin layer. Although the SF microparticles maintained its main body morphology, the thin surface layer could be degraded to form a porous morphology after 28 days of immersion (Fig. 3 (C)), which favors the release of insulin from the microparticles' interior. Moreover, the insulin layer of SF microparticles consisted of numerous nanoparticles, which significantly increased the specific surface area of the microparticles. The increased specific surface area is able to promote the efficiency of drug release. In vitro release results showed that 96.6% of the insulin was released from SF microparticles after 28 days. In contrast, insulin was mainly adsorbed on the microparticle surface by physisorption for insulinadsorbed SF microparticles, resulting in an initial burst phenomenon. Furthermore, the insulin release of the SF sponge dressings was monitored for 40 days; the accumulated release amount of insulin was up to 90.7% after 28 days. In the present study, the insulin-encapsulated SF microparticles were fabricated by coaxial electrospraying in liquid nitrogen bath. This method possesses mild process conditions, using aqueous SF solutions and low processing temperature, which is an appealing feature for the retention of insulin bioactivity. The FTIR and CD results demonstrated that released insulin could maintain original molecular conformation (Fig. 4). Insulin can stimulate the migration of keratinocyte and endothelial cells [8–10]. The strong stimulatory effect on the migration of HaCaT cells and EA. hy926 cells confirmed the bioactivity of insulin from SF microparticles (Figs. 5 and 6). The cellular and intracellular signaling pathways of insulin stimulation on wound healing are involved in the proliferation and migration of cells. In our previous studies, it was found that topical insulin applications accelerated re-epithelialization and “maturation” of the healing tissue through stimulating the migration and proliferation of keratinocyte, and these effects are dependent on the insulin receptor and PI3K-Akt-Rac1 signaling pathways, but are independent of EGF/EGF-R [8]. On the other hand, insulin could promote angiogenesis by stimulating endothelial cell migration and tube formation. The studies involving molecular mechanisms revealed that these effects occur independently of VEGF/VEGFR signaling, but are dependent upon the insulin receptor itself, and that downstream signaling pathways involve PI3K, Akt, sterol regulatory element-binding protein 1 (SREBP-1) and Rac1 [9,10]. The released insulin from SF microparticles showed a strong stimulatory effect on the migration of keratinocyte and endothelial cells (Fig. 6), which would be positive to reepithelialization and vascularization in the wound healing process. We observed that insulin in the SF dressing improved overall wound healing in streptozotocin-induced diabetic SD rats (Fig. 7). H&E and Masson staining results showed increased wound closure rate and collagen deposition in the insulin-loaded group (Figs. 8 and 9), suggesting that local insulin release stimulated the viability of tissue repair cells, such as keratinocyte and fibroblast. The immunohistochemical determination showed a significantly increased CD31 expression in the insulin-loaded group (Fig. 10), suggesting that the functionalized SF dressing with topical insulin release promoted vascularization via stimulating endothelial cell viability. Altogether, our results demonstrate that the functionalized SF dressing, mainly through the benefits of insulin loading and sustained stimulation ability enhanced the wound healing process in diabetic rats. Therefore,
403
this insulin-functionalized SF dressing offers promising options for promoting wound healing of chronic wounds. 5. Conclusion In this study, the insulin-encapsulated SF microparticles were prepared by coaxial electrospraying and freeze-drying. This method not only allowed for a sustained release of insulin over a prolonged period, it also helped to preserve the bioactivity of the encapsulated insulin. Insulin was successfully encapsulated in the inner layer of SF microparticles, providing a sustained release of bioactive insulin. Then, we developed a functionalized dressing for chronic wound healing by loading the insulin-encapsulated SF microparticles into SF sponges. The results from the in vivo test using diabetic rat models indicated that the dressing accelerated chronic wound closure rate, and enhanced collagen deposition and vascularization, which provides great potential for the treatment of chronic wounds. Acknowledgements This work was supported by the National Nature Science Foundation of China (31370968, 81270909), the Natural Science Foundation of Jiangsu Province (BK20131177), the College Natural Science Research Project of Jiangsu Province (12KJA430003), and the Jiangsu Province Ordinary Universities and Colleges Graduate Scientific and Innovation Plan (KYLX15_1247). We also appreciate the Systat Software Inc. for providing the PeakFit v4.12 demonstration software in its web site, and the authors used the software to analyze their experimental data during the free-trial period. References [1] C. Helary, A. Abed, G. Mosser, L. Louedec, D. Letourneur, T. Coradin, M.M. GiraudGuille, A. Meddahi-Pellé, Evaluation of dense collagen matrices as medicated wound dressing for the treatment of cutaneous chronic wounds, Biomater. Sci. 3 (2015) 373–382. [2] P. Inpanya, A. Faikrua, A. Ounaroon, A. Sittichokechaiwut, J. Viyoch, Effects of the blended fibroin/aloe gel film on wound healing in streptozotocin-induced diabetic rats, Biomed. Mater. 7 (2012) 035008. [3] M.B. Dreifke, A.A. Jayasuriya, A.C. Jayasuriya, Current wound healing procedures and potential care, Mater. Sci. Eng. C 48 (2015) 651–662. [4] M. Hrynyk, M. Martins-Green, A.E. Barron, R.J. Neufeld, Alginate-PEG sponge architecture and role in the design of insulin release dressings, Biomacromolecules 13 (2012) 1478–1485. [5] X. Chen, Y. Liu, X. Zhang, Topical insulin application improves healing by regulating the wound inflammatory response, Wound Repair Regen. 20 (2012) 425–434. [6] E.J. Pierre, R.E. Barrow, H.K. Hawkins, T.T. Nguyen, Y. Sakurai, M. Desai, R.R. Wolfe, D.N. Herndon, Effects of insulin on wound healing, J. Trauma Acute Care Surg. 44 (1998) 342–345. [7] M. Hrynyk, R.J. Neufeld, Insulin and wound healing, Burns 8 (2014) 1433–1446. [8] Y. Liu, M. Petreaca, M. Yao, M. Martins-Green, Cell and molecular mechanisms of keratinocyte function stimulated by insulin during wound healing, BMC Cell Biol. 10 (2009) 1. [9] M. Hrynyk, M. Martins-Green, A.E. Barron, R.J. Neufeld, Sustained prolonged topical delivery of bioactive human insulin for potential treatment of cutaneous wounds, Int. J. Pharm. 398 (2010) 146–154. [10] Y. Liu, M. Petreaca, M. Martins-Green, Cell and molecular mechanisms of insulin-induced angiogenesis, J. Cell. Mol. Med. 13 (2009) 4492–4504. [11] L.I.F. Moura, A.M.A. Dias, E.C. Leal, L. Carvalho, H. Sousa, E. Carvalho, Chitosan-based dressings loaded with neurotensin-an efficient strategy to improve early diabetic wound healing, Acta Biomater. 10 (2014) 843–857. [12] Y. Zhang, W. Wei, P. Lv, L. Wang, G. Ma, Preparation and evaluation of alginate-chitosan microspheres for oral delivery of insulin, Eur. J. Pharm. Biopharm. 77 (2011) 11–19. [13] A. Elsayed, M. Al-Remawi, I. Maghrabi, M. Hamaidi, N. Jaber, Development of insulin loaded mesoporous silica injectable particles layered by chitosan as a controlled release delivery system, Int. J. Pharm. 461 (2014) 448–458. [14] S. Mao, C. Guo, Y. Shi, L.C. Li, Recent advances in polymeric microspheres for parenteral drug delivery-part 1, Expert Opin. Drug Deliv. 9 (2012) 1161–1176. [15] O. Germershaus, V. Werner, M. Kutscher, L. Meine, Deciphering the mechanism of protein interaction with silk fibroin for drug delivery systems, Biomaterials 35 (2014) 3427–3434. [16] R. Mo, T. Jiang, J. Di, W. Tai, Z. Gu, Emerging micro- and nanotechnology based synthetic approaches for insulin delivery, Chem. Soc. Rev. 43 (2014) 3595–3629. [17] B. Kundu, R. Rajkhowa, S.C. Kundu, X. Wang, Silk fibroin biomaterials for tissue regenerations, Adv. Drug Deliv. Rev. 65 (2013) 457–470.
404
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
[18] X. Li, R. You, Z. Luo, G. Chen, M. Li, Silk fibroin scaffolds with a micro-/nano-fibrous architecture for dermal regeneration, J. Mater. Chem. B 4 (2016) 2903–2912. [19] A.E. Thurber, F.G. Omenetto, D.L. Kaplan, In vivo bioresponses to silk proteins, Biomaterials 71 (2015) 145–157. [20] S. Çalamak, C. Erdoğdu, M. Özalp, K. Ulubayram, Silk fibroin based antibacterial bionanotextiles as wound dressing materials, Mater. Sci. Eng. C 43 (2014) 11–20. [21] S.E. Wharram, X. Zhang, D.L. Kaplan, S.P. McCarthy, Electrospun silk material systems for wound healing, Macromol. Biosci. 10 (2010) 246–257. [22] E.S. Gil, B. Panilaitis, E. Bellas, D.L. Kaplan, Functionalized silk biomaterials for wound healing, Adv. Healthcare Mater. 2 (2013) 206–217. [23] S. Min, X. Gao, C. Han, Y. Chen, M. Yang, L. Zhu, H. Zhang, L. Liu, J. Yao, Preparation of a silk fibroin spongy wound dressing and its therapeutic efficiency in skin defects, J. Biomater. Sci. Polym. Ed. 23 (2012) 97–110. [24] B.M. Min, G. Lee, S.H. Kim, Y.S. Nam, T.S. Lee, W.H. Park, Electrospinning of silk fibroin nanofibers and its effect on the adhesion and spreading of normal human keratinocytes and fibroblasts in vitro, Biomaterials 25 (2004) 1289–1297. [25] K. Numata, D.L. Kaplan, Silk-based delivery systems of bioactive molecules, Adv. Drug Deliv. Rev. 62 (2010) 1497–1508. [26] E. Wenk, A.J. Wandrey, H.P. Merkle, L. Meinel, Silk fibroin spheres as a platform for controlled drug delivery, J. Control. Release 132 (2008) 26–34. [27] J.A. Kluge, A.B. Li, B.T. Kahn, F.G. Omenettoa, D.L. Kaplan, Silk-based blood stabilization for diagnostics, Proc. Natl. Acad. Sci. U. S. A. 113 (2016) 5892–5897. [28] Y. Zhao, X. Li, C. Lu, Y. Xu, H. Lv, D. Dai, L. Zhang, C. Sun, W. Yang, X. Li, Y. Zhao, H. Fu, L. Cai, M. Lin, L. Chen, M. Zhang, Experiment on the feasibility of using modified gelatin nanoparticles as insulin pulmonary administration system for diabetes therapy, Acta Diabetol. 49 (2012) 315–325.
[29] X. Li, C. Zhang, L. Wang, C. Ma, W. Yang, M. Li, Acylation modification of Antheraea pernyi silk fibroin using succinic anhydride and its effects on enzymatic degradation behavior, J. Chem. (2013). [30] C.C. Liang, A.Y. Park, J.L. Guan, In vitro scratch assay: a convenient and inexpensive method for analysis of cell migration in vitro, Nat. Protoc. 2 (2007) 329–333. [31] Y. Yang, T. Xia, W. Zhi, L. Wei, J. Weng, C. Zhang, X. Li, Promotion of skin regeneration in diabetic rats by electrospun core-sheath fibers loaded with basic fibroblast growth factor, Biomaterials 32 (2011) 4243–4254. [32] Y.H. Lee, J.J. Chang, M.C. Yang, C.T. Chien, W.F. Lai, Acceleration of wound healing in diabetic rats by layered hydrogel dressing, Carbohydr. Polym. 88 (2012) 809–819. [33] C. Manoharan, J. Singh, Evaluation of poly (1,6-bis-(p-carboxyphenoxy)) hexaneco-sebacic acid microspheres for controlled basal insulin delivery, Pharm. Res. 30 (2013) 627–640. [34] M. Zhu, K. Wang, J. Mei, C. Li, J. Zhang, W. Zheng, D. An, N. Xiao, Q. Zhao, D. Kong, L. Wang, Fabrication of highly interconnected porous silk fibroin scaffolds for potential use as vascular grafts, Acta Biomater. 10 (2014) 2014–2023. [35] Z. Cao, X. Chen, J. Yao, L. Huang, Z. Shao, The preparation of regenerated silk fibroin microspheres, Soft Matter 3 (2007) 910–915. [36] N. Bock, T.R. Dargaville, M.A. Woodruff, Electrospraying of polymers with therapeutic molecules: state of the art, Prog. Polym. Sci. 37 (2012) 1510–1551. [37] R. You, Y. Xu, Y. Liu, X. Li, M. Li, Comparison of the in vitro and in vivo degradations of silk fibroin scaffolds from mulberry and nonmulberry silkworms, Biomed. Mater. 10 (2015) 015003.
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Functionalized silk fibroin dressing with topical bioactive insulin release for accelerated chronic wound healing Xiufang Li a, Yan Liu b, Jian Zhang b, Renchuan You c, Jing Qu a, Mingzhong Li a,⁎ a b c
National Engineering Laboratory for Modern Silk, College of Textile and Clothing Engineering, Soochow University, No. 199 Ren'ai Road, Industrial Park, Suzhou 215123, China Department of Burn and Plastic Surgery, School of Medicine Affiliated Ruijin Hospital, Shanghai JiaoTong University, Shanghai 200025, China College of Textile Science and Engineering, Wuhan Textile University, Wuhan 430200, China
a r t i c l e
i n f o
Article history: Received 6 October 2016 Received in revised form 10 November 2016 Accepted 21 November 2016 Available online 24 November 2016 Keywords: Silk fibroin Insulin release Microparticles Dressings Wound healing
a b s t r a c t The healing of chronic wounds remains a key challenge in regenerative medicine. To promote wound healing, a bioactive wound dressing is required. In this study, a functionalized silk fibroin dressing with topical bioactive insulin release was prepared for the treatment of chronic wounds. For this purpose, insulin-encapsulated silk fibroin (SF) microparticles were prepared by coaxial electrospraying of aqueous SF solution under mild processing conditions. Insulin was successfully encapsulated in the inner layer of SF microparticles, providing a sustained insulin release for up to 28 days. It was found that the insulin released from the microparticles could maintain original molecular conformation. Moreover, the cell migration assay based on human keratinocyte and endothelial cells confirmed that the insulin released from SF microparticles retained its native bioactivity. Furthermore, the insulin-encapsulated microparticles were loaded into a SF sponge, functioned as a bioactive wound dressing, and the in vivo therapeutic effect of the sponge dressing was evaluated on dorsal full thickness wounds of diabetic Sprague-Dawley rats. The results showed that an insulin-functionalized SF dressing accelerated wound closure, collagen deposition and vascularization, thus, significantly promoting wound healing. The insulin-functionalized SF dressing provides new treatment options for chronic wounds. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Chronic cutaneous wounds are characterized by the absence of healing after 6 weeks, which is commonly seen as a consequence of diabetes mellitus and vascular compromise [1,2]. The classic treatment is the debridement of the wound followed by its compression with sterile gauze. When this method is not effective enough in some chronic wounds, the dressings based on adequate biomaterials are frequently used to promote healing [1]. However, treatment currently focuses on dressings that prevent microbial infiltration and keep a balanced moisture and gas exchange environment. To heal some chronic wounds, a bioactive dressing is frequently required. For example, in the cutaneous wound in diabetic foot ulcers, reepithelialization and angiogenesis are powerless. Therefore, functionalized wound dressings with sustained bioactive drug release are needed to offer a stimulatory function for facilitating tissue repair, thus, promoting wound healing and minimizing the recovery period [3,4]. Recently, increasing evidences demonstrated that insulin contributes to wound healing [5–7]. It was reported that insulin accelerated reepithelialization of provisional tissue by stimulating the migration and proliferation of keratinocyte [4,8,9]. In addition, insulin can ⁎ Corresponding author. E-mail address: [email protected] (M. Li).
http://dx.doi.org/10.1016/j.msec.2016.11.085 0928-4931/© 2016 Elsevier B.V. All rights reserved.
stimulate the migration and tube formation of endothelial cells which helps to improve angiogenesis during wound healing [9,10]. However, the major problem of topical administration of peptides is their short half-life and loss of bioactivity in the peptidase-rich wound environment [11]. An alternative strategy to overcome this problem is the use of biocompatible wound dressings for sustained delivery of insulin. Drug delivery using biodegradable microspheres is a promising approach for sensitive biologicals, and insulin has been encapsulated into polymer microparticles to establish a sustained delivery system [4,6, 12,13]. However, most methods for the preparation of insulin delivery microparticles, such as emulsification and the solvent extraction method [9,12], need processing in organic solvents, at extreme pH values, or mechanical stress, potentially challenging the bioactivity of insulin [14, 15]. Unlike other small-molecule drugs, insulin possessing a complex molecular conformation for its biological activity is more susceptible to salts, organic solvents and high temperature, which may cause a loss of bioactivity [16]. In this regard, the retention of bioactivity of insulin released from dressing remains the key challenge. As a natural protein, silk fibroin (SF) has been explored for various tissue engineering applications due to excellent biocompatibility and tailorable degradability [17,18]. SF materials showed excellent bioresponses in vivo with low immunogenicity for numerous clinical applications [19]. SF biomaterials have been applied as wound healing dressings in diverse structural forms such as film, sponge and
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
electrospun fibers, which supported cell growth of fibroblasts, keratinocytes and endothelial cells, and showed a positive effect for wound healing [18–24]. In addition, silk is generated in an aqueous state and is thus readily miscible with other protein solutions. Its aqueous solubility and processability under very mild conditions makes it an attractive material for the loading of sensitive biologics to avoid loss of bioactivity of the drugs to be delivered [25,26]. SF-based biomaterials are being considered to address a wide range of stabilization challenges, from labile enzymes to antibodies [27]. Glucose oxidase, lipase, and horseradish peroxidase were entrapped in SF films over 10 months and significant activity was retained, even when stored at 37 °C [27]. The SF microparticles loaded with insulin-like growth factor 1 demonstrated controlled and sustained release over 7 weeks in bioactive form [26]. Therefore, we hypothesized that insulin-encapsulated SF microparticles are able to provide a sustained release of bioactive insulin, and the microparticle-loaded SF dressing can promote the migration of skin repair cells and the reconstruction of a microvascular network for healing chronic wounds. In the present study, we prepared insulin-encapsulated SF microparticles by coaxial electrospraying and freeze-drying. This method proposed to prepare SF microparticles under very mild conditions from aqueous SF solutions, which facilitates maintaining the bioactivity of insulin. Then, the insulin-encapsulated microparticles that offered a topical and sustained release system would be incorporated into SF sponges via multilayer loading, functioning as a bioactive wound dressing for chronic wounds. The molecular conformation of insulin released from microparticles was detected by FTIR and CD, and the bioactivity of released insulin was assessed by the scratch assay of the immortal human keratinocyte (HaCaT) and the human endothelial cell (EA. hy926). Furthermore, the in vivo therapeutic effect of SF dressings, including healing rate and vascularization, was evaluated on dorsal full thickness wounds of diabetic Sprague-Dawley (SD) rats.
395
theoretical drug amount relative to entire microparticles. Eq. (1) was used to evaluate insulin loading content. Insulin loading ð%Þ Weight of the insulin in micropaticles ðmgÞ ¼ 100 Weight of the micropaticles ðmgÞ
ð1Þ
The loading ratio of insulin in the coaxial microparticles was about 16.7% according to the above equation. As a control, the pure SF and insulin microparticles were prepared, respectively, the solution delivering rate was 0.4 mL/h and the other conditions remained consistent. 2.3. Preparation of microparticle-loaded SF sponge dressing
2. Materials and methods
As shown in Fig. 1 (B), a microparticle-loaded SF sponge dressing was prepared by multilayer loading and freeze-drying method. The SF solution was diluted to 2.0 wt%, and then 2-morpholinoethanesulfonic acid (MES), N-hydroxysuccinimide (NHS) and 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) (Sigma-Aldrich) were added to the SF solution at 20%, 10% and 20% of the SF weight in solution, respectively. The mixed solution was stirred slowly and reacted in an ice bath for 1 h. 900 μL of EDC-activated SF solution was added into a round metallic box (D = 40 mm, H = 15 mm, V = 10 mL), and frozen at −80 °C for 30 min. Then 10 mg of SF microparticles were evenly laid on a frozen SF solution layer, and frozen at −80 °C for 15 min. On top of it, 700 μL of EDC-activated SF solution was poured to cover the microparticles and frozen at − 80 °C for 15 min. Finally, 10 mg of microparticles were added and frozen, and subsequently, 900 μL of EDC-activated SF solution was poured according to the above process. Then the SF mixture composite with two layers of microparticles were frozen at −80 °C for 4 h, and further lyophilized to obtain a microparticle-loaded sponge dressing. The dressings were sterilized with γ-ray irradiation and stored at 4 °C.
2.1. Preparation of SF solution
2.4. Scanning electron microscopy
Regenerated SF solution was prepared following the procedure described previously [18]. Briefly, Bombyx mori silk fibers (Huzhou, China) were degummed three times in 0.05% Na2CO3 solution at 98– 100 °C for 30 min and dried at 60 °C after thoroughly rinsing. The extracted silk fibroin was dissolved in a ternary solvent of CaCl2:CH3CH2OH:H2O (1:2:8 M ratio) at 72 ± 2 °C for 1 h. SF solution was obtained after dialysis with cellulose membranes (MWCO 9– 12 kDa) in deionized water for 4 days.
The morphology of samples was observed by a scanning electron microscopy (SEM, S-4800, Hitachi, Japan). The size of microparticles was analyzed on the basis of SEM images with the Nano Measurer analysis software (Department of Chemistry, Fudan University, China. Copyright: (C) 2008 Jie Xu). To calculate the diameter of microparticles, we measured the average diameter of a total of 100 microparticles based on using SEM images. 2.5. In vitro insulin release
2.2. Preparation of SF microparticles 100 mg insulin (27.5 IU/mg; from porcine, WangBang Bochemical Pharmaceutical Co., Ltd., Xuzhou, China) was dissolved in 10 mL of 0.01 M HCl, then the pH of the solution was adjusted to 7.0 ± 0.1 using 0.1 M NaOH. SF solution was diluted to 2.0 wt% and mixed with glycerol at 30 wt% of SF weight. The resulting insulin solution (1.0 wt%) and SF solution (2.0 wt%) were used as core and shell for coaxial electrospraying, respectively. Fig. 1 (A) shows the preparation process for insulin-encapsulated SF microparticles. The coaxial nozzle has an inner capillary of 0.6 mm (inner diameter) and an outer capillary of 1.2 mm (inner diameter). Two syringe pumps (KDS100, KD Scientific, USA) deliver core and shell layer solutions at the rates of 0.1 mL/h and 0.3 mL/h, respectively. A high-voltage power supply supplied a 13 kV high voltage between the nozzle and the collection box filled with liquid nitrogen, and the distance is 12 cm. The collected insulin-encapsulated SF microparticles were lyophilized by a Virtis Genesis 25-LE Freeze Dryer for 48 h and then balanced at 25 °C and RH 90% for 24 h to obtain the water-insoluble microparticles. Drug loading ratio was expressed as the percentage of
Insulin was labeled with fluorescein isothiocyanate (FITC) (SigmaAldrich) as previously described [28]. 300 μL of FITC solution (10 mg/mL in dimethylsulfoxide) was added to 10 mL of insulin solution (15 mg/mL in bicarbonate buffer, pH = 8.5, 0.1 M) and stirred at room temperature for 60 min. Next, 200 μL of 1 M hydroxyl ammonium chloride solution was added and stirred for 10 min at room temperature. The insulin was then purified using a 10 mm × 300 mm column with Sephadex G-50 equilibrated in 0.1 M sodium bicarbonate buffer (pH = 8.5) to remove any unreacted FITC. FITC-insulin encapsulated SF microparticles were prepared according to the condition found in Section 2.2. The fluorescent images were captured using an inverted fluorescence microscope (Olympus IX71, Japan). Physically absorbed microparticles were used as a control group. 10 mg of pure SF microparticles was directly immersed into 2 mL of FITC-insulin solution to adsorb insulin by permeating and physical absorption, and named as insulinadsorbed microparticles. The remaining amount of insulin in solution was quantified to calculate the adsorption amount using a fluorescence spectrophotometer (FM4P TCSPC, Horiba Jobin Yvon), and the loading ratio in the insulin-adsorbed microparticles was about 5.5%.
396
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
Fig. 1. The schematic diagram of the preparation process for: (A) SF microparticles and (B) the microparticle-loaded SF sponge dressing.
The insulin release profiles were determined by immersing 10 mg of microparticles and a sponge dressing containing 10 mg of microparticles in 10 mL of phosphate buffer solution (PBS, 10 mM, pH 7.4). The samples were incubated in a water bath at 37 °C with a 100 rpm shaking. 1 mL of medium was collected and replaced with an equal volume of fresh PBS after centrifugation at predetermined time points. The amount of released insulin was quantified using a fluorescence spectrophotometer.
2.7. Circular dichroic spectroscopy To detect the conformational stability of released insulin, the secondary structure of released insulin solutions were analyzed after 2 days release with circular dichroism (CD) spectra. CD spectra were recorded on a spectrophotometer (J-815, JASCO) equipped with a Peltier temperature controller. Spectra were obtained from 250 to 195 nm at a scanning speed of 100 nm·min−1 at 25 °C. The data was recorded as mean residue ellipticity (deg·cm2·dmol−1).
2.6. Fourier transform infrared spectroscopy 2.8. Cell scratch assay To detect whether microparticle preparation processing has an effect on the molecular conformation of insulin, the insulin particles were prepared according to the preparation methods of insulin encapsulated SF microparticles. The lyophilized insulin microparticles were cut into powder and then prepared in KBr pellets. Native insulin powder was used as a control. FTIR spectra were performed with a Nicolet 5700 spectrometer (Nicolet Company, USA). The percentage of different secondary structures in insulin microparticles and native insulin powder was analyzed by PeakFit v4.12 software (Systat Software Inc., USA) from the amide I region [29].
To evaluate the bioactivity of released insulin, a cell scratch assay was conducted following the procedure described previously [30]. The immortal human keratinocyte line HaCaT (presented by Professor S. Zhang at the Basic Medical College of Soochow University) and the human endothelial cell line EA. hy926 (ATCC, USA) were cultured in Dulbecco's modified Eagle medium (DMEM, low glucose, Gibco, USA) supplemented with 10% fetal bovine serum (FBS, Gibco, USA) and 1% streptomycin-penicillin (Gibco, USA). The medium was replaced every 3 days, and cell cultures were maintained in a humidified incubator at 37 °C and 5% CO2. After reaching 80% confluence, cells were detached
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
from the petri dish and seeded on 6 well plates at a density of 1 × 104 cells/well in 1 mL, respectively. Cells were allowed to reach confluency after 48 h, at which time a single scratch was made using the tip of a plastic disposable 10 μL pipette tip. All wells were washed twice with 1 mL of DMEM culture medium to remove cellular debris. Then, three groups of culture mediums were used for further cell culture: (a) supernatants with insulin released from microparticles within 2 days at a concentration of 10−7 M, (b) insulin solution from native insulin powder at a concentration of 10−7 M and (c) without insulin. The distance across the scratch at 0 h (d0) and 24 h (d1) was measured under a phase contrast inverted microscope (Olympus IX81, Japan). The relative migration distance (%) across the scratch was calculated as follows: Relative migration distance ð%Þ ¼
d0 −d1 100 d0
397
10 mm diameter) were created on the upper back of each rat using a pair of sharp scissors and a scalpel after being anesthetized with pentobarbital sodium (30 mg/kg). SF dressings were immersed for 24 h in sterilized PBS solution to remove crosslinker residue prior to in vivo experiments. Some insulin loss is inevitable during immersion, and the loss is about 6.4% of total insulin loading in the dressing within 24 h. The wound was covered with insulin-loaded SF sponge dressing, SF sponge dressing and gauze, respectively. The entire wound area was then wrapped with sterile gauze and fixed with an elastic bandage. After surgery, each rat was caged individually with free access to water and food, and the dressing was daily wetted by topical normal saline application. Five rats were sacrificed and the skin tissues surrounding the wound were excised to evaluate wound repair at 1, 2 and 3 weeks, respectively. The degree of wound healing was determined via measuring the area of wound closure. 2.10. Histology and immunohistochemical staining
2.9. Diabetic rat model and wound healing examination All animal experiments were conducted in accordance with the Management Ordinance of Experimental Animal of China ((2001) No. 545) and approved by the Jiangsu Province in experimental animals management rules ((2008) No. 26). Twenty SD rats (180–200 g, SPF grade, male) were used in this experiment, and type I diabetic rats were induced by a single intraperitoneal injection of 60 mg/kg streptozotocin (Sigma-Aldrich) dissolved in sodium citrate buffer solution (pH ~ 4.5) [31,32]. After 4 days, the blood glucose levels of rats were checked using a glucometer (Accu-Chek Aviva, Roche). Fifteen rats with blood glucose levels higher than 16.7 mM, which were considered as a successful diabetic model. Three full thickness wounds (approximately
Specimens were harvested and immediately fixed in 4% formaldehyde in PBS at 4 °C, and then embedded in paraffin for hematoxylineosin (HE) and Masson staining. Wounds were assessed by multiple serial tissue sections, and the stained sections were observed with a light microscope (Olympus IX51, Japan). For immunohistochemical staining, the paraffin sections (5 μm) were de-paraffinized, washed three times in PBS for 5 min, and then blocked with 5% serum for 30 min. The slides were subsequently incubated with primary antibody mouse anti-CD31 monoclonal antibody (Abcam) at 4 °C overnight. After rinsing three times with PBS, the slides were incubated with secondary antibody at 37 °C for 30 min, and further developed with 3,3′-diaminobenzidine tetrahydrochloride (DAB) solution and counterstained with hematoxylin.
Fig. 2. The SEM of microparticles and sponge dressing. (A–C) SF microparticles, (D–F) insulin-encapsulated SF microparticles, and (G, H) microparticle-loaded SF sponge dressing. (B, C, E, F, G, H) Cross-sectional views of SF microparticles, insulin-encapsulated SF microparticles and dressing, respectively. Scale bars: (A, D) 1 mm, (B, E, H) 50 μm, (C, F) 5 μm, (G) 200 μm.
398
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
Fig. 3. (A) Cumulative FITC-insulin release from microparticles and dressing: (a) insulin-adsorbed SF microparticles, (b) insulin-encapsulated SF microparticles, and (c) microparticleloaded SF dressings. (B) The fluorescent image of FITC-insulin-encapsulated SF microparticles. (C) SEM image of insulin-encapsulated SF microparticles after 28 days immersion in PBS. Scale bars: (B) 300 μm, (C) 100 μm.
2.11. Statistical analysis Data are presented as means ± SD. Statistical comparisons were performed using ANOVA (t-test), and differences at p b 0.05 were considered as statistically significant.
3. Results 3.1. Morphological observation As shown in Fig. 2 (A), the SF microparticles were close to spherical in shape, which showed an average diameter of 94.6 ± 27.7 μm by statistical calculation. The surface of microparticles was covered by a dense skin layer with a few micropores, whereas the interior consisted of a dense core layer and an outer layer with a porous structure (Fig. 2 (B)). Unlike the core layer, the porous outer layer contained a large number of nanofibers (5–150 nm diameter), nanoparticles (5–300 nm diameter) and a sheet-like structure (5–150 nm thickness), along the radial direction of the microparticles (Fig. 2 (C)). Fig. 2 (D) showed the morphology of insulin-encapsulated SF microparticles, which has a similar surface morphology and average diameter (92.9 ± 25.0 μm) with pure SF microparticles. The porous outer layer consisted of a large number of sheet-like structures (5–150 nm thickness), and some
thin fibers (5–150 nm diameter) and particles (100–1200 nm diameter) could be also observed (Fig. 2 (E)). However, numerous particles (200– 1800 nm) formed in the core layer (Fig. 2 (E, F)). These results suggest that the SF in the shell layer solution is able to assemble into nanoparticles, and further assemble into nanofibers or a sheet-like structure during the electrospraying process, while the insulin as a core layer solution is easy to assembly into micro/nano particles. Furthermore, to load insulin-encapsulated SF microparticles into an SF sponge to form a functionalized dressing, the microparticles were embedded within the SF sponge by multilayer loading and freezing. As shown in Fig. 2 (G, H), the SF microparticles were tightly embedded within the pores and evenly distributed in SF sponges. 3.2. In vitro insulin release The cumulative FITC-insulin release profiles of SF microparticles and microparticle-loaded SF dressings are shown in Fig. 3 (A). Obviously, the insulin-adsorbed SF microparticles showed a serious burst release of 98.4% within the first 2 h (Fig. 3 (A-a)). The release behavior from insulin-encapsulated SF microparticles was triphasic consisting of (i) a typical burst release within the first 12 h, followed by (ii) a period of stable and continuous release until day 14, and finally (iii) a period of low release rates from these days on. As shown in Fig. 3 (A-b), the cumulative insulin release from insulin-encapsulated SF microparticles were 62.2%,
Fig. 4. (A) FTIR spectra of: (a) insulin microparticles prepared by electrospraying and (b) native insulin powder. (B) CD spectra of: (a) released insulin from SF microparticles and (b) native insulin.
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
399
Table 1 Secondary structure contents of insulin as determined by FTIR spectra. Unordered structures include random coils and turns. Samples
Unordered structure (%)
α-Helix (%)
β-Sheet (%)
Electrosprayed insulin microparticles Native insulin powder
29.1 30.6
41.3 42.2
29.6 27.2
77.3%, 86.8% and 91.1% at 2, 5, 9 and 14 days, respectively. From 14 to 40 days, release rate continuously reduced, and about 98.2% of insulin was released from insulin-encapsulated SF microparticles. The microparticle-loaded SF dressings exhibited a similar release behavior with insulin-encapsulated SF microparticles. About 16.2%, 39.8%, 66.4% and 81.8% cumulative insulin was released from dressing at 2, 5, 9 and 14 days, respectively (Fig. 3 (A-c)), and the cumulative insulin release content was about 90.7% at 28 days. These results indicated that the microparticle-loaded SF sponge dressings provided an effective system for sustained insulin release over a month-long period. 3.3. Conformational detection Compared to native insulin powder, the FTIR spectra of insulin microparticles prepared by electrospraying showed similar absorption peaks in Amide I (1658 cm−1) and Amide II (1541 cm−1), which contributed to α-helix or random coil (Fig. 4 (A-a, b)). The α-helix content is commonly used to evaluate whether the secondary structure of insulin changed [33]. As shown in Table 1, the result of the fit covering amide I region shows the contents of different secondary structures. The αhelix content of insulin microparticles prepared by electrospraying and native insulin powder were about 41.3% and 42.2%, respectively, suggesting that microparticle fabrication process had no significant effect on the secondary structure of insulin. Furthermore, the secondary structure of released insulin was determined by CD spectra. As shown in Fig. 4 (B-b), native insulin exhibits two minima at 208 and 223 nm, indicating a dominant α-helical structure. The released insulin showed a predominant α-helical structure identical to that of native insulin (Fig. 4 (B-a)), indicating that the released insulin from SF microparticles maintained the original molecular conformation. 3.4. Bioactivity of released insulin: cell scratch assay The bioactivity of released insulin was assessed using a cell monolayer scratch with HaCaT and EA. hy926 cells. As shown in Fig. 5 (a0, b0, c0), a scratch of approximately 400 μm in width was made on a confluent
layer of HaCaT. The marginal cells migrated into the scratch to impair the gap closure after 24 h (Fig. 5 (a1, b1, c1)). The scratch without insulin was closed at about 22.4 ± 3.1% after cell migration for 24 h (Fig. 5 (c0, c1, d)), whereas the scratch with released insulin from microparticles and native insulin were closed at about 46.0 ± 2.1% and 49.7 ± 2.9% after 24 h (Fig. 5 (a0, b0, a1, b1, d)), respectively, indicating that released insulin significantly improved HaCaT cell migration. For EA. hy926 cells, the scratch without insulin was closed at about 27.2 ± 5.1% after 24 h (Fig. 6 (c0, c1, d)), whereas the scratch with released insulin and native insulin were closed at about 57.0 ± 3.0% and 60.4 ± 6.5% (Fig. 6 (a0, b0, a1, b1, d)), respectively, indicating that released insulin was able to stimulate EA. hy926 cell migration.
3.5. Macroscopic observations The SF sponge dressings were covered on the wound of streptozotocin-induced diabetic rats to investigate the effects of released insulin on chronic wound healing in vivo. Streptozotocin selectively destroys pancreatic β-cells and causes type I diabetes after 2 or 3 days injection. Blood glucose was measured after streptozotocin injection, which reached 30 mM after 3 days. The establishment of diabetes led to a drastic weight loss of rats of around 25%. The images after wound treatment at different time points are provided in Fig. 7 (A, B, C, D), and the wound images were quantified to show the closure areas of each experimental group at different time points (Fig. 7 (E)). After 1 week, local skin ulcer and effusion was observed in the gauze-covered group (Fig. 7 (B-c1)), whereas wound scab was accelerated and inflammatory response was mild in the SF dressing-covered group (Fig. 7 (Ba1, B-b1)). The wound closure rate of the microparticle-loaded dressing (about 91.3%) was significantly higher than the other two groups after 2 weeks (Fig. 7 (C-a2, C-b2, C-c2, E)) (p b 0.05). At the third week, there was no wound evident in the insulin-loaded group, while small wounds at defect sites were still observed in the other two groups (Fig. 7 (D-a3, b3, c3)). The results indicated that insulin release accelerated the wound healing process.
Fig. 5. Scratch assay for the HaCaT cell migration measurement. (a0, b0, c0) Images of the distance across the scratch at 0 h, (a1, b1, c1) images of the distance across the scratch at 24 h. Scale bars: 300 μm. (d) Relative migration distance (%) after 24 h culture (** p b 0.01).
400
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
Fig. 6. Scratch assay for the EA. hy926 cell migration measurement. (a0, b0, c0) Images of the distance across the scratch at 0 h, (a1, b1, c1) images of the distance across the scratch at 24 h. Scale bars: 300 μm. (d) Relative migration distance (%) after 24 h culture (** p b 0.01).
3.6. Histological observation and immunohistochemistry Histological analyses of dorsal skin wounds stained with HE and Masson are shown in Figs. 8 and 9, respectively. At 1 week, the newly regenerated granulation tissue appeared at the marginal region of wounds as shown in Fig. 8 (a1, b1, c1). After 2 weeks, the granulation tissue was developed to the centre of the skin defect. The tissue neogenesis in the microparticle-loaded group was more significant than the other groups at 2 weeks after wounding (Fig. 8 (a2, b2, c2)), and more dermic papilla and hair follicle were found on the wound area at 3 weeks after wounding (Fig. 8 (a3, b3, c3)), demonstrating well-defined dermal-epidermal junctions and well-differentiated epidermis. These results suggested that the insulin-loaded dressing accelerated tissue neogenesis and maturation. Fig. 9 shows the deposition of collagen by Masson's staining at different time points. At the first week, although a slight deposition of collagen was detected in all groups (Fig. 9 (a1, b1, c1)), collagen deposition was more significant in the microparticle-loaded SF dressing group than in the control groups. The deposition of collagen increased and developed to the wound central area after 1 week, and the collagen deposition was significant in the periphery of wound in the microparticle-loaded group after 2 weeks (Fig. 9 (a2)). After 3 weeks, more
collagen deposited in the microparticle-loaded group, and the collagen deposition has been observed in the central area of the wound, and the alignment of collagen was more regular (Fig. 9 (a3)), indicating that the insulin-loaded dressing accelerated extracellular matrix deposition. Furthermore, the positive expression of CD31 was detected at different time points. At the early stage (1 and 2 weeks), the CD31 expression levels in the microparticle-loaded SF dressing group were significantly higher than the other two groups (Fig. 10 (d)), indicating accelerated neovascularization. The results suggested that topical insulin release promoted the vascularization degree of regenerated tissue. 4. Discussion Chronic wounds cause significant morbidity and mortality in many diseases such as diabetes and cardiovascular diseases. The healing of cutaneous wounds involves granulation tissue formation, reepithelialization and angiogenesis. A successful dressing for wound healing not only should be able to alleviate infection and inflammation by acting as a temporary barrier to protect the underlying tissue, but it also should promote reepithelialization and angiogenesis via stimulating the activities of repair cells [1,3]. Insulin can stimulate proliferation and migration of keratinocytes and endothelial cells in the wound
Fig. 7. (A–D) Macroscopic observation of skin wounds for different times: (a) insulin-loaded dressing, (b) SF dressing, (c) gauze, (subscripts 0, 1, 2, and 3) skin wounds after covering with dressings for 0, 1, 2 and 3 weeks. (E) Quantitative evaluation of wound closure at different time points (* p b 0.05).
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
401
Fig. 8. H&E images of wound after treatment for different times. White dotted lines indicated the obvious boundary of tissue neogenesis in wound beds. Scale bars: 1 mm.
healing process, promote tube formation of endothelial cells, and thus promote wound healing [8–10]. To achieve the insulin-functionalized dressing for chronic wound healing, in this work, insulin was encapsulated into SF microparticles and then incorporated with SF sponges to assess impact on wound healing.
The SF microparticle has been utilized as a promising material for drug release due to its biocompatibility, degradability, and the ability to control drug release kinetics [15,25]. In the present study, we fabricated SF microparticles using an aqueous encapsulation process based on coaxial electrospraying. When differentiated droplets fell into liquid
Fig. 9. Masson images of wound after treatment for different times. White dotted lines indicated the obvious boundary of collagen deposition in wound beds. Scale bars: 1 mm.
402
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
nitrogen, ice nuclei formed and grew into larger ice crystals; meanwhile the SF solution was concentrated from the outer part to the inner part. Thus, the inner part of SF microparticles was denser than the outer part (Fig. 2 (B)). It has been reported that SF was able to assemble to form nanospheres or nanofibers in an aqueous environment at low concentration [34,35]. The concentration of SF in the outer part of microparticles was significantly decreased after freezing. Consequently, the outer part in microparticles exhibited a sheet-like morphology with thin fibers and some nanoparticles (Fig. 2 (C)). The insulin-encapsulated SF microparticles had a similar diameter and surface morphologies with pure SF microparticles (Fig. 2 (D)). In the inner layer, however, more
large particles (100–1200 nm) were observed in insulin-encapsulated SF microparticles than pure SF microparticles (Fig. 2 (E, F)). Different from fibrous SF macromolecules, insulin is a small-molecule globular protein with a low molecular weight of 5808 Da, which may be easier to assemble into micro/nano particles. The results suggested that insulin was mainly encapsulated in the inner layer of the microparticles in the form of nanoparticles. In practice, the yield of this technique is a key for practical clinic application. Single-nozzle electrospraying has a low-throughput with a yield of milligrams/hour, but this can be overcome with multiple electrospray sources. Particle production could be increased from milligrams to grams per hour using 19 parallel nozzles
Fig. 10. (a, b, c) Immunohistochemical staining of CD31 expression at different time points. Scale bars: 250 μm. (d) Quantitative evaluation of CD31 expression at different time points (* p b 0.05).
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
[36]. For wound treatments, however, microparticles are difficult to distribute directly in a homogeneous way and retain over a wound surface. Therefore, the microparticles were tightly embedded within the SF sponge to form an even distribution in the SF sponge dressing after multilayer loading (Fig. 2 (G, H)). The drug release mechanism from SF materials is complex, and diffusion control or biodegradation alone is also unable to fully explain the release mechanism, thus, further studies are needed to fully understand the underlying release mechanism from SF materials [26]. Although the biodegradation of SF materials is slow, the biodegradability of SF may help to sustain the release of insulin from the microparticles [15,37]. In this work, the inner structure of insulin-encapsulated SF microparticles is a porous morphology and covered with a thin skin layer. Although the SF microparticles maintained its main body morphology, the thin surface layer could be degraded to form a porous morphology after 28 days of immersion (Fig. 3 (C)), which favors the release of insulin from the microparticles' interior. Moreover, the insulin layer of SF microparticles consisted of numerous nanoparticles, which significantly increased the specific surface area of the microparticles. The increased specific surface area is able to promote the efficiency of drug release. In vitro release results showed that 96.6% of the insulin was released from SF microparticles after 28 days. In contrast, insulin was mainly adsorbed on the microparticle surface by physisorption for insulinadsorbed SF microparticles, resulting in an initial burst phenomenon. Furthermore, the insulin release of the SF sponge dressings was monitored for 40 days; the accumulated release amount of insulin was up to 90.7% after 28 days. In the present study, the insulin-encapsulated SF microparticles were fabricated by coaxial electrospraying in liquid nitrogen bath. This method possesses mild process conditions, using aqueous SF solutions and low processing temperature, which is an appealing feature for the retention of insulin bioactivity. The FTIR and CD results demonstrated that released insulin could maintain original molecular conformation (Fig. 4). Insulin can stimulate the migration of keratinocyte and endothelial cells [8–10]. The strong stimulatory effect on the migration of HaCaT cells and EA. hy926 cells confirmed the bioactivity of insulin from SF microparticles (Figs. 5 and 6). The cellular and intracellular signaling pathways of insulin stimulation on wound healing are involved in the proliferation and migration of cells. In our previous studies, it was found that topical insulin applications accelerated re-epithelialization and “maturation” of the healing tissue through stimulating the migration and proliferation of keratinocyte, and these effects are dependent on the insulin receptor and PI3K-Akt-Rac1 signaling pathways, but are independent of EGF/EGF-R [8]. On the other hand, insulin could promote angiogenesis by stimulating endothelial cell migration and tube formation. The studies involving molecular mechanisms revealed that these effects occur independently of VEGF/VEGFR signaling, but are dependent upon the insulin receptor itself, and that downstream signaling pathways involve PI3K, Akt, sterol regulatory element-binding protein 1 (SREBP-1) and Rac1 [9,10]. The released insulin from SF microparticles showed a strong stimulatory effect on the migration of keratinocyte and endothelial cells (Fig. 6), which would be positive to reepithelialization and vascularization in the wound healing process. We observed that insulin in the SF dressing improved overall wound healing in streptozotocin-induced diabetic SD rats (Fig. 7). H&E and Masson staining results showed increased wound closure rate and collagen deposition in the insulin-loaded group (Figs. 8 and 9), suggesting that local insulin release stimulated the viability of tissue repair cells, such as keratinocyte and fibroblast. The immunohistochemical determination showed a significantly increased CD31 expression in the insulin-loaded group (Fig. 10), suggesting that the functionalized SF dressing with topical insulin release promoted vascularization via stimulating endothelial cell viability. Altogether, our results demonstrate that the functionalized SF dressing, mainly through the benefits of insulin loading and sustained stimulation ability enhanced the wound healing process in diabetic rats. Therefore,
403
this insulin-functionalized SF dressing offers promising options for promoting wound healing of chronic wounds. 5. Conclusion In this study, the insulin-encapsulated SF microparticles were prepared by coaxial electrospraying and freeze-drying. This method not only allowed for a sustained release of insulin over a prolonged period, it also helped to preserve the bioactivity of the encapsulated insulin. Insulin was successfully encapsulated in the inner layer of SF microparticles, providing a sustained release of bioactive insulin. Then, we developed a functionalized dressing for chronic wound healing by loading the insulin-encapsulated SF microparticles into SF sponges. The results from the in vivo test using diabetic rat models indicated that the dressing accelerated chronic wound closure rate, and enhanced collagen deposition and vascularization, which provides great potential for the treatment of chronic wounds. Acknowledgements This work was supported by the National Nature Science Foundation of China (31370968, 81270909), the Natural Science Foundation of Jiangsu Province (BK20131177), the College Natural Science Research Project of Jiangsu Province (12KJA430003), and the Jiangsu Province Ordinary Universities and Colleges Graduate Scientific and Innovation Plan (KYLX15_1247). We also appreciate the Systat Software Inc. for providing the PeakFit v4.12 demonstration software in its web site, and the authors used the software to analyze their experimental data during the free-trial period. References [1] C. Helary, A. Abed, G. Mosser, L. Louedec, D. Letourneur, T. Coradin, M.M. GiraudGuille, A. Meddahi-Pellé, Evaluation of dense collagen matrices as medicated wound dressing for the treatment of cutaneous chronic wounds, Biomater. Sci. 3 (2015) 373–382. [2] P. Inpanya, A. Faikrua, A. Ounaroon, A. Sittichokechaiwut, J. Viyoch, Effects of the blended fibroin/aloe gel film on wound healing in streptozotocin-induced diabetic rats, Biomed. Mater. 7 (2012) 035008. [3] M.B. Dreifke, A.A. Jayasuriya, A.C. Jayasuriya, Current wound healing procedures and potential care, Mater. Sci. Eng. C 48 (2015) 651–662. [4] M. Hrynyk, M. Martins-Green, A.E. Barron, R.J. Neufeld, Alginate-PEG sponge architecture and role in the design of insulin release dressings, Biomacromolecules 13 (2012) 1478–1485. [5] X. Chen, Y. Liu, X. Zhang, Topical insulin application improves healing by regulating the wound inflammatory response, Wound Repair Regen. 20 (2012) 425–434. [6] E.J. Pierre, R.E. Barrow, H.K. Hawkins, T.T. Nguyen, Y. Sakurai, M. Desai, R.R. Wolfe, D.N. Herndon, Effects of insulin on wound healing, J. Trauma Acute Care Surg. 44 (1998) 342–345. [7] M. Hrynyk, R.J. Neufeld, Insulin and wound healing, Burns 8 (2014) 1433–1446. [8] Y. Liu, M. Petreaca, M. Yao, M. Martins-Green, Cell and molecular mechanisms of keratinocyte function stimulated by insulin during wound healing, BMC Cell Biol. 10 (2009) 1. [9] M. Hrynyk, M. Martins-Green, A.E. Barron, R.J. Neufeld, Sustained prolonged topical delivery of bioactive human insulin for potential treatment of cutaneous wounds, Int. J. Pharm. 398 (2010) 146–154. [10] Y. Liu, M. Petreaca, M. Martins-Green, Cell and molecular mechanisms of insulin-induced angiogenesis, J. Cell. Mol. Med. 13 (2009) 4492–4504. [11] L.I.F. Moura, A.M.A. Dias, E.C. Leal, L. Carvalho, H. Sousa, E. Carvalho, Chitosan-based dressings loaded with neurotensin-an efficient strategy to improve early diabetic wound healing, Acta Biomater. 10 (2014) 843–857. [12] Y. Zhang, W. Wei, P. Lv, L. Wang, G. Ma, Preparation and evaluation of alginate-chitosan microspheres for oral delivery of insulin, Eur. J. Pharm. Biopharm. 77 (2011) 11–19. [13] A. Elsayed, M. Al-Remawi, I. Maghrabi, M. Hamaidi, N. Jaber, Development of insulin loaded mesoporous silica injectable particles layered by chitosan as a controlled release delivery system, Int. J. Pharm. 461 (2014) 448–458. [14] S. Mao, C. Guo, Y. Shi, L.C. Li, Recent advances in polymeric microspheres for parenteral drug delivery-part 1, Expert Opin. Drug Deliv. 9 (2012) 1161–1176. [15] O. Germershaus, V. Werner, M. Kutscher, L. Meine, Deciphering the mechanism of protein interaction with silk fibroin for drug delivery systems, Biomaterials 35 (2014) 3427–3434. [16] R. Mo, T. Jiang, J. Di, W. Tai, Z. Gu, Emerging micro- and nanotechnology based synthetic approaches for insulin delivery, Chem. Soc. Rev. 43 (2014) 3595–3629. [17] B. Kundu, R. Rajkhowa, S.C. Kundu, X. Wang, Silk fibroin biomaterials for tissue regenerations, Adv. Drug Deliv. Rev. 65 (2013) 457–470.
404
X. Li et al. / Materials Science and Engineering C 72 (2017) 394–404
[18] X. Li, R. You, Z. Luo, G. Chen, M. Li, Silk fibroin scaffolds with a micro-/nano-fibrous architecture for dermal regeneration, J. Mater. Chem. B 4 (2016) 2903–2912. [19] A.E. Thurber, F.G. Omenetto, D.L. Kaplan, In vivo bioresponses to silk proteins, Biomaterials 71 (2015) 145–157. [20] S. Çalamak, C. Erdoğdu, M. Özalp, K. Ulubayram, Silk fibroin based antibacterial bionanotextiles as wound dressing materials, Mater. Sci. Eng. C 43 (2014) 11–20. [21] S.E. Wharram, X. Zhang, D.L. Kaplan, S.P. McCarthy, Electrospun silk material systems for wound healing, Macromol. Biosci. 10 (2010) 246–257. [22] E.S. Gil, B. Panilaitis, E. Bellas, D.L. Kaplan, Functionalized silk biomaterials for wound healing, Adv. Healthcare Mater. 2 (2013) 206–217. [23] S. Min, X. Gao, C. Han, Y. Chen, M. Yang, L. Zhu, H. Zhang, L. Liu, J. Yao, Preparation of a silk fibroin spongy wound dressing and its therapeutic efficiency in skin defects, J. Biomater. Sci. Polym. Ed. 23 (2012) 97–110. [24] B.M. Min, G. Lee, S.H. Kim, Y.S. Nam, T.S. Lee, W.H. Park, Electrospinning of silk fibroin nanofibers and its effect on the adhesion and spreading of normal human keratinocytes and fibroblasts in vitro, Biomaterials 25 (2004) 1289–1297. [25] K. Numata, D.L. Kaplan, Silk-based delivery systems of bioactive molecules, Adv. Drug Deliv. Rev. 62 (2010) 1497–1508. [26] E. Wenk, A.J. Wandrey, H.P. Merkle, L. Meinel, Silk fibroin spheres as a platform for controlled drug delivery, J. Control. Release 132 (2008) 26–34. [27] J.A. Kluge, A.B. Li, B.T. Kahn, F.G. Omenettoa, D.L. Kaplan, Silk-based blood stabilization for diagnostics, Proc. Natl. Acad. Sci. U. S. A. 113 (2016) 5892–5897. [28] Y. Zhao, X. Li, C. Lu, Y. Xu, H. Lv, D. Dai, L. Zhang, C. Sun, W. Yang, X. Li, Y. Zhao, H. Fu, L. Cai, M. Lin, L. Chen, M. Zhang, Experiment on the feasibility of using modified gelatin nanoparticles as insulin pulmonary administration system for diabetes therapy, Acta Diabetol. 49 (2012) 315–325.
[29] X. Li, C. Zhang, L. Wang, C. Ma, W. Yang, M. Li, Acylation modification of Antheraea pernyi silk fibroin using succinic anhydride and its effects on enzymatic degradation behavior, J. Chem. (2013). [30] C.C. Liang, A.Y. Park, J.L. Guan, In vitro scratch assay: a convenient and inexpensive method for analysis of cell migration in vitro, Nat. Protoc. 2 (2007) 329–333. [31] Y. Yang, T. Xia, W. Zhi, L. Wei, J. Weng, C. Zhang, X. Li, Promotion of skin regeneration in diabetic rats by electrospun core-sheath fibers loaded with basic fibroblast growth factor, Biomaterials 32 (2011) 4243–4254. [32] Y.H. Lee, J.J. Chang, M.C. Yang, C.T. Chien, W.F. Lai, Acceleration of wound healing in diabetic rats by layered hydrogel dressing, Carbohydr. Polym. 88 (2012) 809–819. [33] C. Manoharan, J. Singh, Evaluation of poly (1,6-bis-(p-carboxyphenoxy)) hexaneco-sebacic acid microspheres for controlled basal insulin delivery, Pharm. Res. 30 (2013) 627–640. [34] M. Zhu, K. Wang, J. Mei, C. Li, J. Zhang, W. Zheng, D. An, N. Xiao, Q. Zhao, D. Kong, L. Wang, Fabrication of highly interconnected porous silk fibroin scaffolds for potential use as vascular grafts, Acta Biomater. 10 (2014) 2014–2023. [35] Z. Cao, X. Chen, J. Yao, L. Huang, Z. Shao, The preparation of regenerated silk fibroin microspheres, Soft Matter 3 (2007) 910–915. [36] N. Bock, T.R. Dargaville, M.A. Woodruff, Electrospraying of polymers with therapeutic molecules: state of the art, Prog. Polym. Sci. 37 (2012) 1510–1551. [37] R. You, Y. Xu, Y. Liu, X. Li, M. Li, Comparison of the in vitro and in vivo degradations of silk fibroin scaffolds from mulberry and nonmulberry silkworms, Biomed. Mater. 10 (2015) 015003.