Biomimetic synthesis of sericin and silica hybrid colloidosomes for stimuli-responsive anti-cancer drug delivery systems

Colloids and Surfaces B: Biointerfaces 151 (2017) 102–111

Contents lists available at ScienceDirect

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

Protocols

Biomimetic synthesis of sericin and silica hybrid colloidosomes for stimuli-responsive anti-cancer drug delivery systems Ying Yang a , Yurong Cai a,∗ , Ning Sun a , Ruijing Li a , Wenhua Li a , Subhas C. Kundu b,c , Xiangdong Kong d , Juming Yao a,∗ a The Key Laboratory of Advanced Textile Materials and Manufacturing Technology of Ministry of Education, National Engineering Lab for Textile Fiber Materials and Processing Technology (Zhejiang), College of Materials and Textiles, Zhejiang Sci-Tech University, Hangzhou 310018, China b Department of Biotechnology, Indian Institute of Technology (IIT), Kharagpur, West Bengal 721302, India c 3Bs Research Group, Headquarters of the European Institute of Excellence on Tissue Engineering and Regenerative Medicine, University of Minho, AvePark, 4805-017 Barco, Taipas, Guimaraes, Portugal d College of Life Sciences, Zhejiang Sci-Tech University, Hangzhou 310018, China

a r t i c l e

i n f o

Article history: Received 21 July 2016 Received in revised form 7 December 2016 Accepted 9 December 2016 Available online 12 December 2016 Keywords: Sericin Synthesis Colloidosomes Stimuli-responsive Drug delivery

a b s t r a c t Colloidosomes are becoming popular due to their significant flexibility with respect to microcapsule functionality. This study reports a facile approach for synthesizing silica colloidosomes by using sericin microcapsule as the matrix in an environment-friendly method. The silica colloid arrangement on the sericin microcapsules are orchestrated by altering the reaction parameters. Doxorubicin (DOX), used as a hydrophilic anti-cancer drug model, is encapsulated into the colloidosomes in a mild aqueous solution and becomes stimuli-responsive to different external environments, including pH values, protease, and ionic strength are also observed. Colloidosomes with sub-monolayers, close-packed monolayers, and close-packed multi-layered SiO2 colloid shells can be fabricated under the optimized reaction conditions. A flexible DOX release from colloidosomes can be obtained via modulating the SiO2 colloid layer arrangement and thickness. The close-packed and multi-layered SiO2 colloid shells can best protect the colloidosomes and delay the rapid cargo release. MG-63 cells are killed when doxorubicin is released from the microcapsules due to degradation in the microenvironment of cancer cells. The drug release period is prolonged as SiO2 shell thickness and integrity increase. This work suggests that the hybrid colloidosomes can be effective in a bioactive molecule delivery system. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Colloidosomes are hollow microcapsules with densely packed colloidal particle walls [1–4]. The functionalities and physical properties of colloidosomes, such as permeability, selectivity, and biocompatibility, can be precisely controlled using suitable colloidal particles and processing conditions [1,2,4–6]. Compared with the shortcomings of direct drug delivery systems, hollow colloidosomes possess a perfect alternative approach that is highly efficient for encapsulating drug. They control drug delivery efficiently at the target site through response to external stimuli and prevent the drug degradation in the bloodstream, prolong circulation time, and improve drug bioavailability [7–10]. These properties make colloidosomes promising vehicles in drug delivery systems

∗ Corresponding authors. E-mail addresses: [email protected] (Y. Cai), [email protected] (J. Yao). http://dx.doi.org/10.1016/j.colsurfb.2016.12.013 0927-7765/© 2016 Elsevier B.V. All rights reserved.

and seem reliable in delivering potent drugs to action sites precisely and timely, enhance accumulation in tumor sites, decrease adverse effects, and improve drug tolerance [11–14]. In recent years, many methods have been developed to synthesize colloidosomes, and they are commonly based on the self-assembly of colloidal particles at water/oil interface in immiscible liquids. Silica sol and polystyrene latex are effective building blocks for colloidosome assembly. A shell reinforcement mechanism is necessary for preparing colloidosomes to convert individual assembly units into robust microcapsules and retain their complete structure after the oil–water template removal. Moreover, stimulus-responsive colloidosomes have been developed for inverting or demulsifying the obtained colloidosomes in response to changes in pH or temperature in solutions. The convenient approach offered significant flexibility in microcapsule functionality, which are the main requirements in the fields [15,16]. Silk protein sericin, obtained from silkworm cocoons and once considered a waste product in silk production, is becoming pop-

Y. Yang et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 102–111

ular due to its biomedical and biotechnological applications, as well as its excellent water solubility and biocompatibility. Sericin is earlier regarded as a waste product of the silk industry [17–20]. Recently, smart sericin microcapsules were fabricated through protein self-assembly in water solutions without organic solvents or surfactants in the laboratory [21,22]. Sericin microcapsules exhibit good biocompatibility and multifunctional and tunable properties. However, they show poor stability in harsh environments [21,22]. Colloidal silica microcapsules belong to a prominent class of SiO2 hollow materials. They are advantageous in drug delivery systems because of their low density, adjustable pore size, high specific surface area, mechanical stability, easy functionalization, low toxicity, and good biocompatibility [23,24]. Based on these, we utilized colloidal SiO2 nanoparticles as the protection shells for sericin microcapsules to obtain a desirable hybrid colloidosomes. The most important advantage of this approach is that drug degradation in the extreme external environment and bloodstream is prevented because the drug is encapsulated inside the core. SiO2 colloid shells protect the sericin microcapsules and keep continuous drug release. In this report, hydrophilic doxorubicin (DOX), an anti-cancer drug model, was encapsulated in the sericin microcapsules through a one-step encapsulation method under mild conditions to avoid toxic organic solvents and complicated synthesis procedures. A biomimetic approach was followed to prepare the hybrid colloidosomes by growing SiO2 colloids on DOX-encapsulated microcapsule surfaces. The SiO2 colloid shell morphology and thickness were regulated by controlling the reaction conditions. Altering some of the reaction factors, we synthesize sub-monolayer, close-packed monolayer, and close-packed multilayer shells of SiO2 nanoparticle hybrid colloidosomes and control DOX release from the core. The SiO2 colloid shell enhances the mechanical stability o and drug circulation time of the microcapsules. Moreover, polymeric core/shell colloidosome structures enable the drug carriers to counteract extreme environments, such as pH values, protease concentrations, and ionic strengths. We

103

observed MG-63 cells undergo necrosis at different stages upon culturing with different DOX-encapsulated hybrid colloidosome thicknesses. 2. Results and discussion 2.1. The morphology evolution of hybrid colloidosomes by altering the reaction conditions According to the Stober reaction, two steps are involved in the silica growth process. One is the tetraethylorthosilicate (TEOS) hydrolysis and the other is the condensation of SiO2 onto the microcapsule surfaces [25–27]. NH /Alcohol

3 Si(OC2 H5 )4 + 4H2 O−−− −−−−→Si(OH)4 + 4C2 H5 OH

NH /Alcohol

3 −−−−→SiO2 ↓ +2H2 O Si(OH)4 −−−

The ␨-potential of the microcapsules was −26.7 ± 1.83 MV. The presence of ammonia negatively charged SiO2 colloids to stabilize the surface. The electrostatic adsorption was weak because the surface charges of the microcapsules and SiO2 were negative. The sericin microcapsules had thin and smooth SiO2 surfaces upon direct TEOS hydrolysis onto the microcapsules without surface any modification (Fig. 1a). Therefore, we used poly (allyl amine hydrochloride) PAH with amine groups as scaffolding for SiO2 nucleation and growth on the substrate surface. In principle, the nucleation and growth rates of inorganic nanoparticles are strongly dependent on the environment factors, such as reaction temperature, time, surfactants, and concentration of solutions [28–30]. To improve microcapsule morphology regulation, we further investigated the influence of some reaction parameters. SEM images of resulting hybrid colloidosomes are presented in Figs. 1 and 2. We first studied the effects of reaction temperature. Submonolayer colloid shells on the microcapsules were observed at 25 ◦ C (Fig. 1b). When the temperature was elevated to 50 and 60 ◦ C,

Fig. 1. SEM images of the hybrid colloidosomes with different coating conditions: (a) without poly(allylamine hydrochloride) (PAH), (b)–(d) hydrolysis temperatures of 25, 50, and 60 ◦ C, (e)–(g) hydrolysis durations of 20, 50, and 60 min, and (h)–(j) condensation durations of 2, 8, and 20 h, respectively. The scale bar is 1 ␮m.

104

Y. Yang et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 102–111

Fig. 2. SEM and TEM images of hybrid colloidosomes obtained at different TEOS concentrations: (a) 0.5 mL; (b) 2.5 mL, and (c) 10 mL. Labels 1–3 represent SEM, TEM, and enlarged TEM images, respectively.

SiO2 rapidly grew and interacted with each other gradually, to finally cover the whole shell (Fig. 1c, d). Relatively the high temperature promoted TEOS hydrolysis and silica growth. Colloidosomes with sub-monolayer, close-packed monolayer, and close-packed multi-layer SiO2 colloid shells can be fabricated by changing the reaction temperatures. A remarkable dependence on the morphologies of composites at the time of the reaction is observed, i.e., hydrolysis and condensation durations. Hydrolysis and condensation durations in the experiment were kept 20, 50, and 60 min and 2, 8, and 20 h, respectively (Fig. 1e–j). By prolonging the reaction time, the SiO2 particles become smaller and interact until they merge into a close-packed colloidal layer. The concentrations of the silicon source (TEOS) also plays an important role on the morphology of hybrid colloidosomes. The morphology transition from monolayer to close-packed multiplelayer is observed when volume of TEOS is increased from 0.5 to 10 mL The representative images are shown in Fig. 2a1 , b1 , and c1 . This indicates that reaction parameters influence SiO2 shell formation. The relatively high temperature, long duration time, and suitable TEOS concentration create an environment, which produce enough SiO2 . They overlap and merge. All these findings illustrate the possibility to control the morphology of SiO2 colloid shells from sub-monolayer to multi-layer by regulating the reaction parameters. Therefore, we take advantage of the different shell thicknesses to control drug release for certain duration of time to meet different treatment requirements.

2.2. Characterizing hybrid colloidosomes The thickness of the SiO2 colloid shells was verified by transmission electron microscopy (TEM) (Fig. 2a2 , b2 , c2 ). The variation of shell thickness are visualized by following the magnified images (Fig. 2a3 , b3 , c3 ). The coverings of the microcapsule shell surface compact and thicker. The changes are observed on the thin submonolayer to close-packed multi-layers. It is observed when TEOS concentration was increased, and is consistent with the SEM observations (Fig. 2a1 , b1 , c1 ). From the TEM images, controlling the surface layers by adjusting the reaction conditions was feasible. The composition of as-prepared hybrid colloidosomes was detected using XRD and FTIR analyses. The XRD showed that the main composition of the hybrid colloidosomes is SiO2 with poor crystallinity (Figs. S1, ESI†). The FTIR transmittance spectra of hybrid colloidosomes are shown in Fig. 3a. The position and the shape of the main Si O vibrational band at 1090 cm−1 prove a stoichiometric silicon dioxide structure [26]. A small peak at 954 cm−1 belongs to the Si OH stretching and that at 796 cm−1 is attributed to the Si O bending [31]. Moreover, the peak at 1648 cm−1 in the spectral is attributed to vibrations of amide bond coming from the sericin part of the hybrid colloidosomes [22,32,33]. The results show that SiO2 colloids were successfully incorporated into microcapsule surfaces. Scanning force microscopy (SFM) was used to detect local elastic properties of colloidosomes through the interaction between the surface and AFM tip [34–36]. The samples were dispersed in distilled water, dropped onto the silicon wafer, and dried at

Y. Yang et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 102–111

105

Fig. 3. (a) The FTIR spectra of sericin microcapsules and hybrid colloidosomes. (b) Force-distance curves of various samples based on AFM measurements. The line denotes the curve during retraction from the sample surface. Data are presented as mean ± SD, n = 4.

37 ◦ C. Surface deformation, the maximum adhesive force between tip and sample (Fadh ), dissipation energy, and elastic modulus are acquired from the resulting force curve. The mechanical properties are alsoextracted from the force curve from which the elastic modulus of the samples can be obtained by fitting the DerjaguinMuller-Toporov (DMT) model [34] to the part of the force curve where the sample and tip should be contact, as given by the equation F = 4/3E ∗ (Rd3 )1/2 + Fadh Here, F is the force, E* is the effective elastic modulus, R is the tip radius, d is the deformation value at a given force, and Fadh is the maximum adhesion force. The slope of the force–distance curve (F–D curve) indicates the sample surface stiffness against the applied force. Based on the curves, the elastic properties were deduced, e.g., Young’s modulus of sample surface. Representative F–D curves for naked microcapsules and hybrid colloidosomes with different shell thicknesses are presented in Fig. 3b. Compared with other samples, multilayer hybrid colloidosomes have steep slopes in the curve that indicates high stiffness. Naked microcapsules have slopes that are milder than other, implying their good elasticity. The monolayer hybrid colloidosome slope is between the multi-layer hybrid colloidosomes and naked microcapsules, proving that microcapsule stiffness increases due to the presence of SiO2 colloid shell in a shell thickness-dependent manner. Increasing the stiffness can resist the shear stresses in the drug injection. Moreover, the microcapsules may not disintegrade, which is good for drug delivery and treatment system. Young’s modulus of the various samples are approximately 620, 2430, and 5250 MPa, which correspond the naked microcapsules, monolayer, and multi-layer hybrid colloidosomes. This result indicates that the reproducing this action is possible. 2.3. Characterization of DOX-encapsulated microcapsules Microcapsule morphology was observed under SEM to investigate the influence of encapsulating DOX into the microcapsules (Fig. 4a). The morphologies of DOX-encapsulated microcapsules were the same with the origin microcapsules. Adding DOX during microcapsule preparation does not make any difference in terms of morphology. To determine whether DOX had been successfully encapsulated into the microcapsules or not, we used confocal scanning micro-

scope (CLSM) for observation (Fig. 4b) because DOX emits red light in excited state. Subsequently, red fluorescence was observed inside the microcapsules, indicating that the cores of the microcapsules are filled with DOX. Almost all microcapsules had DOX inside the microcapsules. It is a good indication for sustain release of the drug. The preparation process is unstable being water soluble sericin that needs some further improvement to adjust reaction condition in the future. The chemical composition of the DOX-encapsulated microcapsules was investigated using FTIR (Fig. 4c). The peak at 806 cm−1 is due to the stretching bands of C O CH3 [37,38]. The characteristic bands at 1207 and 966 cm−1 are related to the bending vibration of CO H and C C O coming from DOX [39]. These prove that DOX was successfully loaded into the microcapsules and corroborated the findings under CLSM. The actual DOX loading percentage in the microcapsules was approximately 12–23 wt% of the gross, which has been measured by UV–vis. 2.4. In vitro analysis of drug loading and release The stimuli-responsive offers an opportunity for programmable drug delivery systems to optimize cancer therapy [40]. In the stimuli-responsive delivery system it is expected that the anticancer agent may be released by an appropriate stimulus (for example, pH, protease, and ionic strength). We evaluated the release behavior of DOX-loaded microcapsules against different pH values (2.8, 5, and 7.4), protease concentrations (0, 4, 8, and 12 U), and ion concentrations (0, 0.1, 0.4, and 0.7 M NaCl). The effects of these three factors on the drug release were followed. Moreover, we also check the drug release of the different SiO2 thicknesses (sub-monolayer, monolayer, and multi-layer) of the DOX-loaded hybrid colloidosomes against extreme pH value (pH 2.8), protease concentration (12 U), and ion concentration (0.7 M NaCl). We investigate further the influence of the thickness of SiO2 colloid shells on the drug release profiles from hybrid colloidosomes. The loss rate of DOX was approximately 8.8% of the load amount during the process of coating SiO2 nanoparticles. 2.4.1. Effect of pH on release of DOX Acidic pH is considered an ideal trigger as an internal stimulus for the selective release of anticancer drugs. For example, taking advantage of slightly acidic environments in cancerous tissues (pH 6.5–7.2), endosomes (pH 5.0–6.5), and lysosomes (pH 4.5–5.0) compared with physiological pH 7.4 in the blood and normal tissues.

106

Y. Yang et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 102–111

Fig. 4. SEM image (a), CLSM image (b), and FTIR spectra (c) of DOX-loaded microcapsules. The image on the right-top corner of (b) is the magnified fluorescent image.

Fig. 5. DOX release behavior from the sericin microcapsules (a1 , b1 , c1 ) at pH 2.8, 5.0, and 7.4, protease concentrations (4, 8, and 12 U), and salt concentrations at 0.1, 0.4, and 0.7 M, respectively. The behaviours of DOX release from hybrid colloidosomes at pH 2.8 (a2 ), protease concentration at 12 U (b2 ), and salt concentration at 0.7 M (c2 ). The number in the below Images 1–3 represent the sub-monolayer, monolayer, and multi-layer of the silica shell, respectively. Temperature was maintained at 37 ◦ C for the release study. Data are presented as mean ± SD, n = 4.

Y. Yang et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 102–111

The pH-sensitive nanoparticles were designed and developed to release drugs in the tumor site and/or endo-/lysosomal compartments [41,42]. The release profiles of DOX are shown in Fig. 5a1 and a2 . The drug release at pH 7.4 was considerably slow, with an initial burst of approximately 6.7%, and only 26% of the drug was released after 25 h. By contrast, the drug release was faster at pH 5.0 and 2.8, with approximately 64% and 86%, respectively, of the drug released within 25 h. It is observed that DOX-loaded microcapsules are well dispersed in the buffer at pH 7.4. Moreover, they aggregated and settled at the bottom of the tubes at pH 5.0 and 2.8. pH changes from 7.4 to 5.0 or 2.8 led to the deformation and precipitation of the microcapsules. This may cause the release of the enclosed drug. These release profiles suggest that the product will be stable in blood during in vivo circulation. The dox is probably released due to gradual degradation of microcapsules in the presence of cancer cell-ECM-enzymatic environment. This may act as sustain delivery for cancerous niche due tumor microenvironment. These biomimetic colloidosomes may serve effectively to release the drug. Due to hydrolytic/enzymatic degradation of microcapsules drug will be released. A different types of interaction is playing to release the drug due to their protective environment with different layers of SiO2 -shells. Stimuli–responsive factors-cancer cell-ECM interaction as well enzymatic activities are required to study in future. The micro-environment of cancer cells are different and are based on many factors (pH, ECM, hypoxia and others) [43]. The DOX release rate reached approximately 74%, 65%, and 56% within 150 h, when the thickness of hybrid colloidosomes changed from sub-monolayer to close-packed monolayer and close-packed multi-layer (Fig. 5a2 ). We observed the upward tendency of dotted and dashed lines. The decrease release rate may have been due to the reduction of permeability of the microcapsules after coating the SiO2 -shell. The shell might have inhibited rapid DOX release from sericin microcapsules and maintained a constant and longer release period of the cargo. The colloid shells provided good protection for microcapsules and drugs. However, the protective degrees may be somewhat different. Delivery time may increase in case of thicker SiO2 . The SiO2 shells are thinner so the drug release from the hybrids is easier. This drug delivery system is different than other drug delivery system as microcapsules do not enter the cells.

107

This helps the sustained DOX release in low concentrations in the long run. 2.4.3. Effect of ionic strength on release of DOX Ionic concentration strength can affect the rate of drug release from hydrophilic-extended release matrices [48]. The normal range of serum sodium in human blood is 0.135–0.145 M, and we chose 0–0.7 M sodium chloride within a reasonable range. The release rate increased from 18% to 87% as the concentration of NaCl gradually increased from 0 M to 0.7 M (Fig. 5c1 ), indicating that the salt ions entered the sericin microcapsules via osmosis, as well as a high concentration of salt shrunk the microcapsules, thereby accelerating DOX release. The release rate decreased from 36% to 11% as the thickness of SiO2 shell increased from sub-monolayer to closepacked multi-layers (Fig. 5c2 ). The drug had been released slowly owing to the protection of SiO2 shell. After a certain period of time, the rate became faster. The ions may have shielded the polyelectrolyte charges, which resulted in destabilization of the complex. Therefore, some channels were created and drug release was accelerated through them [49]. We deduced that hybrid colloidosomes may have weak stimuli-responsive to ionic strength. In general, the release behavior of the drug from all samples expressed a similar trend: an initial burst release and a long plateau. Minor differences among the cumulative drug release contents can be observed among different samples. This may be due to the influence of the SiO2 shell. The whole profile of this experimentation is schematically illustrated in Fig. 6. Sericin microcapsules were formed by self-assembly with the assistance of calcium ions and stirring. DOX was dispersed in the solution by stirring and finally encapsulated into the sericin microcapsules during microcapsule formation via one-step method to obtain the drug-loaded sericin microcapsules. Moreover, the morphology of the SiO2 shell changed from sub-monolayer, mono-layer, to close-packed multi-layer by altering the reaction conditions. The naked sericin microcapsules and hybrid colloidosomes were broken and shrunk, thereby contributing to the stimuli-responsive DOX release under different external environments. 2.5. Cytocompatibility and function investigations

2.4.2. Effect of protease on release of DOX A variety of enzymes exist in the tissues of organisms, such as hydrolases, transferases, and ligases and others. Protease is a hydrolase. The biochemical signatures of proteases can act as a trigger when spatially oriented drug release is required by introducing specific enzyme substrate sequences either into the nanocarrier scaffold or in the linker segment through which the drug is anchored to the nanocarrier [44]. We chose the Streptomyces protease, which is a protease available to us. Enzyme-responsive drug carriers mainly rely on cleavage peptide sequences by esterases or proteases [45]. We first investigated whether this protease could trigger DOX release from the sericin microcapsules and hybrid colloidosomes or not. Furthermore, we investigated the effects of the protease concentrations (Fig. 5b1 , b2 ). Initially, the protease-trigger release of DOX was faster, which might have been due to protease reacting directly with the sericin microcapsules [46,47]. Finally, DOX release was concentration dependent and was released approximately 52%, 66%, and 75% with 4, 8, and 12 U protease in the solutions respectively within 24 h. The results show that the sericin microcapsules were responsive to Streptomyces protease. The release rates of the drug were 48%, 57%, and 66% within 150 h in 12 U of protease when microcapsules were coated with different thicknesses of SiO2 colloids, indicating that SiO2 shells somehow prevented the reaction between protease and sericin microcapsules. Therefore, delivery time is longer with thicker SiO2 shells.

The MG-63 cell morphology was examined by fluorescence microscopy after 1, 4, and 7 culturing days (Fig. 7a). The images show that MG-63 cells maintain a fusiform shape on sericin microcapsules and hybrid colloidosomes (without DOX loading) after culturing for 1 day. Therefore, both types of materials (sericin microcapsules and hybrid colloidosomes) can support cell adhesion and spreading. More cells were observed on the hybrid colloidosomes after culturing for 7 days than the control group. Moreover, we analyzed the cytotoxic response of MG-63 cells due to the presence of microcapsules and hybrid colloidosomes using a cell counting assay kit-8 (CCK-8). Cell viability increased with culturing time (Fig. 7b). The cells cultured with sericin microcapsules and hybrid colloidosomes showed a significantly higher viability (P < 0.05) than the control samples. Moreover, results indicated that sericin microcapsules and hybrid colloidosomes can promote MG-63 cell viability and proliferation [22,50–52]. Furthermore, the hybrid colloidosomes may be cytocompatible biomaterials. To evaluate material function of DOX-loaded microcapsules and hybrid colloidosomes with different thicknesses, MG-63 cells were co-cultured with these microcapsules in 96-well plates (Fig. 8). All the MG-63 cells adhered at the bottom of the well plates within 2 h. We observe small fusiform shapes. MG-63 cells, grown on DOXloaded microcapsules, underwent necrosis after 6 h. After 17 h, red fluorescent from the DOX-microcapsules could be observed. This

108

Y. Yang et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 102–111

Fig. 6. Schematic of the preparations of colloidosomes, drug loading, and release from the sericin microcapsules and hybrid colloidosomes at different pH values, protease concentrations, and ionic strengths.

Fig. 7. (a) Morphologies of DiO-stained MG-63 cells after co-culturing with sericin microcapsules and hybrid colloidosomes (without DOX loading) for 1, 4, and 7 days. A: control samples; B: sericin microcapsules; C: hybrid colloidosomes. The subscripts 1, 4, and 7 represent the respective culture days. The scale bar is 20 ␮m. (b) MG-63 cell proliferation after co-culturing with sericin microcapsules and hybrid colloidosomes (without DOX loading) for 1, 4, and 7 days, respectively. *P < 0.05, n = 3 at each time point (one way ANOVA followed).

indicates that DOX releases from the microcapsules due to degradation in the presence of microenvironment of cancer cells. The drug probably enters through endocytosis and/or phagocytosis and kills almost all MG-63 cells. Similarly, the treated cells underwent some metabolic changes upon being cultured with hybrid colloidosomes. MG-63 cells began to transform after 12, 17, and 48 h, according to the sub-monolayer, close-packed monolayer, and close-packed multi-layer SiO2 shells, respectively. Finally, all MG-63 cells were dead after co-culturing within 48, 72, and 140 h. Throughout the process, MG-63 cells become round fusiform shape in all groups.The color in the fluorescence images changed from green to light greenish red. This result explains the manner by which the released DOX killed the MG-63 cells and demonstrates the influence of the SiO2 shells on the DOX release. As the thickness increased, DOX was released from hybrid colloidosomes for

a longer time. Moreover, SiO2 colloid shells provided good protection from the sericin microcapsules. This type of fabrication of the colloidosomes helps to control the burst release or slow release for the long-term to meet diverse treatment requirements by changing the thickness of the protection shells made of natural and watersoluble sericin microcapsules. 3. Experimental 3.1. Materials Natural mulberry silk sericin (8 kD, Xintiansi Bio-Tech Huzhou, China), sodium chloride, calcium chloride, ammonia, ethanol, TEOS, PAH, 17.5 kD, ≥95%), and DOX hydrochloride (DOX·HCl, 98%) were purchased from Sigma-Aldrich. Dimethyl sulfoxide

Y. Yang et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 102–111

109

Fig. 8. Morphologies of DiO-stained MG-63 cells after co-culturing with DOX-loaded sericin microcapsules and DOX-loaded colloidosomes with sub-monolayer, monolayer, and multi-layers of SiO2 colloid shells over different durations, respectively. The scale bar is 20 ␮m.

(DMSO) (Shanghai Reagent Chemical Co., China); fetal bovine serum (FBS); 3,30-dioctadecyloxacarbocyanine perchlorate (Dio), pancreatin, L-glutamine, Cell Counting Kit-8 (CCK-8) (Beyotime Co. Ltd, China); streptomyces protease; Dulbecco’s Modified Eagle’s Medium (DMEM) (Gibco company, USA); and MG-63 cells, a human osteosarcoma cell (Shanghai Institute of Biochemistry and Cell Biology, China), were purchased for the experimentations. The prepared solutions were filtered through a Millipore filter (0.22 ␮m) prior to use. 3.2. Instruments The freshly prepared microcapsules were observed using field emission scanning electron microscopy (FESEM, Zeiss Ultra 55, Germany) and TEM (JEM-2100, Japan). The samples were freeze dried at −40 ± 5 ◦ C in a lyophilizer. The surface charges of microcapsules were tested using Zeta potential (Malvern 3000 Zetasizer, UK). The X-ray diffractograms were measured using an X-ray diffractometer system (XRD, ARL XTRA, Switzerland). FTIR spectra were acquired using a Nicolet 5700 (USA) spectrometer with a resolution of 4 cm−1 and a spectral range of 3500–500 cm−1 . Surface elasticity force was performed using the “force-distance analysis” (F/D analysis) mode on an atomic force microscope instrument (MultiMode 8 AFM, Bruker, Germany). The images were captured using a confocal laser scanning microscopy (CLSM, Nikon C2, Japan). 3.3. Sericin microcapsule preparation The sericin microcapsules were fabricated following our methods [21,22]. Commercially available sericin, with a molecular

weight of 8 kD, was used for all the experiments in this report. Briefly, the solutions of sericin (0.2%) and CaCl2 (60 mM) were mixed, and the pH value was adjusted to 7 ± 0.05 by adding HCl and NaOH (0.1 M). The mixture was kept at 37 ◦ C for 24 h and centrifuged at 8000 rpm for 10 min to obtain the microcapsules. The microcapsules were stored at 4 ◦ C for further use. 3.4. DOX encapsulation in the microcapsules DOX·HCl (1 mg/mL) was added into the mixed solution of sericin (0.2%) and CaCl2 (180 mM). The pH of the mixture was adjusted to 7 ± 0.05 by adding HCl (aq, 0.1 M) and NaOH (aq, 0.1 M). The mixture was kept at 37 ◦ C for 24 h. The mixture was centrifuged at 8000 rpm for 10 min to remove any residual DOX. The precipitate was collected. DOX encapsulation efficiency was determined by measuring the UV–vis absorbance at 480 nm. 3.5. Growth of silica DOX and sericin microcapsules were added into the polycyclic aromatic hydrocarbon (PAH solution (1 mg/mL) under continuous stirring for 2 h and then centrifuged and washed with ethanol several times. The precipitate was dispersed in 6 mL ethanol. As the solution was magnetically stirred, certain amounts of NH4 OHH2 O (1:5 in volume) and TEOS-ethanol (1:2 in volumes) solutions were introduced into the suspension. The nucleation and growth of the silica nanoparticles were first accelerated at a relatively high temperature (T = 25 ◦ C–60 ◦ C) at different durations (20–60 min). A further growth of silica was allowed to ripen the composite shells at a lower temperature (25 ◦ C) for 0–20 h under continuous stirring

110

Y. Yang et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 102–111

[25]. Finally, the suspensions were again centrifuged and washed with ethanol and distilled water several times. The loss rate of DOX after being coated with SiO2 colloids was calculated by measuring the UV–vis absorbance at 480 nm.

The loss rate of DOX overtime was measured. The cells were cultured on the tissue culture in 96-well plates (1 × 105 cells/well) using DMEM supplemented with 10% FBS and 1% L-glutamine. The cells were incubated with the materials at 37 ◦ C in a humidified 5% CO2 -containing atmosphere.

3.6. In vitro DOX release at different pH values 3.11. Statistical analysis DOX encapsulated in the sericin microcapsules and hybrid colloidosomes were used in the release experiment. The sericin microcapsules and hybrid colloidosomes (different thicknesses) containing DOX were dispersed in deionized water (3 mL), respectively. The dispersed solution was dispensed into three equal aliquots in three different centrifuge tubes containing phosphate buffer solutions (0.1 M, 9 mL) with different pH values of 2.8, 5, and 7. The centrifuge tubes were kept in an incubator maintained at 37 ◦ C. After a certain period of time, the solution was centrifuged and 1 mL of supernatant was withdrawn. To maintain a constant volume, 1 mL of fresh buffer solution was added to the solution after each sampling. The amounts of released DOX were analyzed using UV–vis spectrophotometry at 480 nm. All DOX drug release data were averaged based on three measurements. 3.7. In vitro DOX release at different ionic strengths Sodium chloride solutions with different concentrations (0.1, 0.4, and 0.7 M) were prepared for different ionic strengths. The steps were similar to that for maintaining the different pH values of the solutions. The amount of DOX released from the microcapsules was determined using a UV–vis spectrophotomer at 480 nm. The drug release data were averaged from three measurements. 3.8. In vitro DOX release at different percentage enzyme Streptomyces protease (1 mg/mL) was added into the solutions with different concentrations (0, 4, 8, and 12 U). The steps were similar to those mentioned above. 3.9. CCK-8 assays The MG-63 cells were cultured in DMEM with 10% FBS and 1% penicillin-streptomycin and maintained at 37 ◦ C in a humidified atmosphere with 5% CO2 . The medium was changed every 2 days. The MG-63 cells were dyed with Dio. A total of 20 ␮L Dio was added to 4 mL DMEM to prepare the medium. The cells were cultured in this medium for 30 min and washed with PBS thrice. The dyed cells were detached with pancreatin. The dyed MG-63 cells were seeded on the tissue culture in the 96-well plates (control) with sericin microcapsules and hybrid colloidosomes. They were cultured in the same conditions as mentioned above. The cell morphologies were observed under a fluorescence microscope after 1, 4, and 7 culturing days. The cell viability on the sericin microcapsules and hybrid colloidosomes (without DOX loading) was determined using CCK-8. After 1, 4, and 7 culturing days, the cells were incubated in 10% CCK-8 solution in a 5% CO2 incubator at 37 ◦ C for 2 h. The intense orange-cultured formazan derivative that was formed by cell metabolism was soluble in the culture medium. The absorbance of the culture medium was recorded at 450 nm using a microplate reader (Bio-Rad 680, USA). The viability of cells was measured based on optical density. 3.10. Material function study DOX-loaded microcapsules and hybrid colloidosomes were imaged with a fluorescence microscope using MG-63 cells, which were dyed with Dio. The cells were treated with encapsulated-DOX in the sericin microcapsules and different hybrid colloidosomes.

All data were expressed as mean ± SD of samples in triplicate. Statistical data evaluation from different compositions was done using ANOVA. *P < 0.05 between the groups was considered statistically significant. 4. Conclusions Colloidosomes were synthesized facilely using sericin microcapsule as the matrix in a simple and environmentally friendly method using SiO2 colloids as building blocks of protective shells by hydrolyzing the tetraethylorthosilicate solution. The SiO2 colloid arrangement was controlled by modulating the reaction parameters. A series of colloidosomes with sub-monolayer, close-packed monolayer, and close-packed multi-layer SiO2 colloid shells was fabricated. The hybrid colloidosomes showed improved stability, shell strength, and flexible responsiveness toward external stimulations via pH values, protease, and ionic strength secondary to the combination of the intrinsic physical/chemical properties of the organic protein core and inorganic SiO2 colloid shells. The release of DOX was controlled by regulating the SiO2 shell structure under different environmental conditions. Moreover, the hybrid colloidosomes showed good cytocompatibility. DOX-encapsulated colloidosomes are released in the microenvironment of the cancer cells and shows destructive effect on MG-63 cells. The hybrid colloidosomes, due to its facile synthesis, efficient loading, and controllable release, may act as a potential drug carrier in bioactive molecule delivery system for regenerative medicine. Acknowledgements The work is financially supported by the Program for National Natural Science Foundation of China (51202219 and 51372226), Zhejiang Provincial Natural Science Foundation of China (LY16E020013), and Zhejiang Top Priority Discipline of Textile Science and Engineering. SCK is grateful to Zhejiang Sci-Tech University, Hangzhou for providing excellent working facilities during his short stay in the laboratory. SC Kundu holds ERA Chair Full Professor of European Commission Project (FoReCaST) at 3Bs Research Group, University of Minho, Portugal. References [1] A. Böker, J. He, T. Emrick, T.P. Russell, Soft Matter 3 (2007) 1231. [2] A.D. Dinsmore, M.F. Hsu, M.G. Nikolaides, M. Marquez, A.R. Bausch, D.A. Weitz, Science 298 (2002) 1006. [3] D. Lee, D.A. Weitz, Adv. Mater. 20 (2008) 3498. [4] R.K. Shah, J.W. Kim, D.A. Weitz, Langmuir 26 (2009) 1561. [5] H. Skaff, Y. Lin, R. Tangirala, K. Breitenkamp, A. Böker, T.P. Russell, T. Emrick, Adv. Mater. 2005 (17) (2005) 2082. [6] B.P. Binks, Adv. Mater. 14 (2002) 1824. [7] J. Wu, L. Liu, Y.R. Cai, J.M. Yao, Mini-Rev. Med. Chem. 1501 (2013) 1507. [8] H. Koide, A. Okamoto, H. Tsuchida, H. Ando, S. Ariizumi, C. Kiyokawa, N. Oku, J. Control. Release 228 (2016) 1. [9] A.J. Morse, E.C. Giakoumatos, S.Y. Tan, G.B. Webber, S.P. Armes, S. Ata, E.J. Wanless, Soft Matter 12 (2016) 1477. [10] L. Dang, H. Ma, J. Xu, Y. Jin, J. Wang, Q. Lu, F. Gao, CrystEngComm 18 (2016) 544. [11] A. Jana, L. Bai, X. Li, H. Ågren, Y. Zhao, ACS Appl. Mater. Interfaces 8 (2016) 2336. [12] J. Tan, L. Fu, X. Zhang, Y. Bai, L. Zhang, J. Mater. Sci. 51 (2016) 9455. [13] W. Chen, K. Achazi, B. Schade, R. Haag, J. Control. Release 205 (2015) 15. [14] R. Haag, F. Kratz, Angew. Chem. Int. Ed. 45 (2006) 1198.

Y. Yang et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 102–111 [15] W. Tong, X. Song, C. Gao, Chem. Soc. Rev. 41 (2012) 6103. [16] J. Shi, Y. Jiang, X. Wang, H. Wu, D. Yang, F. Pan, Z. Jiang, Chem. Soc. Rev. 43 (2014) 5192. [17] B. Zhu, H. Wang, W.R. Leow, Y. Cai, X.J. Loh, M.Y. Han, X. Chen, Adv. Mater. 28 (2015) 4250. [18] S.C. Kundu, B.C. Dash, R. Dash, D.L. Kaplan, Prog. Polym. Sci. 33 (2008) 998. [19] D. Aytemiz, W. Sakiyama, Y. Suzuki, N. Nakaizumi, R. Tanaka, Y. Ogawa, T. Asakura, Adv. Healthc. Mater. 2 (2013) 361. [20] V. Werner, L. Meinel, Eur. J. Pharm. Biopharm. 97 (2015) 392. [21] Z. Liu, Y. Cai, Y. Jia, L. Liu, X. Kong, S.C. Kundu, J. Yao, Macromol. Rapid Commun. 35 (2014) 1668. [22] W. Li, Y. Cai, Q. Zhong, Y. Yang, S.C. Kundu, J. Yao, J. Mater. Chem. B 4 (2016) 340. [23] X. Du, J. He, Nanoscale 3 (2011) 3984. [24] S. Zhou, A. Sugawara-Narutaki, S. Tsuboike, J. Wang, A. Shimojima, T. Okubo, Langmuir 31 (2015) 13214. [25] H. Gao, D. Wen, G.B. Sukhorukov, J. Mater. Chem. B 3 (2015) 1888. [26] B. Sun, S.A. Mutch, R.M. Lorenz, D.T. Chiu, Langmuir 21 (2005) 10763. [27] S.C. Warren, M.R. Perkins, A.M. Adams, M. Kamperman, A.A. Burns, H. Arora, J. Werner, Nat. Mater. 11 (2015) 460. [28] G.L. Davies, A. Barry, Y.K. Gun’ko, Chem. Phys. Lett. 468 (2015) 239. [29] B. Li, H. Gao, Y. Guo, S. Hou, CrystEngComm 14 (2012) 4168. [30] A.S. Myerson, B.L. Trout, Science 341 (2013) 855. ´ N. Filipovic-Vincekovi ´ ´ L. Sekovanic, ´ J. Chem. Eng. 28 (2011) 89. [31] S. Music, c, [32] J.P. Bange, L.S. Patil, D.K. Gautam, Prog. Electromagn. Res. 3 (2008) 165. ´ [33] A. Slósarczyk, Z. Paszkiewicz, C. Paluszkiewicz, J. Mol. Struct. 744 (2005) 657.

111

[34] B. Zuo, F.F. Zheng, Y.R. Zhao, T. Chen, Z.H. Yan, H. Ni, X. Wang, Langmuir 28 (2012) 4283. [35] Y.F. Dufrêne, D. Martínez-Martín, I. Medalsy, D. Alsteens, D.J. Müller, Nat. Methods 10 (2013) 847. [36] A. Boularas, F. Baudoin, C. Villeneuve-Faure, S. Clain, G. Teyssedre, J. Appl. Phys. 116 (2014) 084106. [37] Q. Yuan, S. Hein, R.D.K. Misra, Acta Biomater. 6 (2010) 2732. [38] H. Huang, Q. Yuan, J.S. Shah, R.D.K. Misra, Adv. Drug Deliv. Rev. 63 (2011) 1332. [39] G. Das, A. Nicastri, M.L. Coluccio, F. Gentile, P. Candeloro, G. Cojoc, E. Di Fabrizio, Microsc. Res. Technique 73 (2011) 991. [40] B.Y. Kim, J.W. Bae, K.D. Park, J. Control. Release 172 (2013) 535. [41] R. Chen, F. Meng, C. Deng, H. Klok, Z. Zhong, Biomaterials 3647 (2013) 3657. [42] J.Z. Du, Y.Q. Tang, A.L. Lewis, S.P. Armes, J. Am. Chem. Soc. 127 (2005) 17982. [43] O. Trédan, C.M. Galmarini, K. Patel, I.F. Tannock, J. Natl. Cancer Inst. 99 (2007) 1441. [44] E. Fleige, M.A. Quadir, R. Haag, Adv. Drug Deliv. Rev. 866 (2012) 884. [45] E. Fleige, M.A. Quadir, R. Haag, Adv. Drug Deliv. Rev. 64 (2012) 866. [46] Y. Cao, B. Wang, Int. J. Mol. Sci. 10 (2009) 1514. ¨ [47] S. Muller-Herrmann, T. Scheibel, ACS Biomater. Sci. Eng. 1 (2015) 247. [48] K. Asare-Addo, M. Levina, A.R. Rajabi-Siahboomi, A. Nokhodchi, Carbohydr. Polym. 85 (2011) 93. [49] H. Cai, C. Ni, L. Zhang, Eur. J. Pharm. Sci. 45 (2012) 43. [50] M. Yang, Y. Shuai, C. Zhang, Y. Chen, L. Zhu, C. Mao, H. OuYang, Biomacromolecules 15 (2014) 1185. [51] X. Feng, S. Zhang, X. Lou, Colloid Surf. B 107 (2013) 220. [52] W. Xu, L. Wang, Y. Ling, K. Wei, S. Zhong, RSC Adv. 4 (2014) 13495.