Fabrication of Biobased Polyelectrolyte Capsules and Their

May 23, 2016 - The Key Laboratory of Food Colloids and Biotechnology Ministry of Education, School of Chemical and Material Engineering, Jiangnan Univ...
0 downloads 11 Views 9MB Size
Research Article www.acsami.org

Fabrication of Biobased Polyelectrolyte Capsules and Their Application for Glucose-Triggered Insulin Delivery Dongjian Shi,*,† Maoshuang Ran,† Li Zhang,† He Huang,† Xiaojie Li,† Mingqing Chen,*,† and Mitsuru Akashi‡ †

The Key Laboratory of Food Colloids and Biotechnology Ministry of Education, School of Chemical and Material Engineering, Jiangnan University, Wuxi 214122, P. R. China ‡ Department of Applied Chemistry, Graduate School of Engineering, Osaka University, 2-1 Yamadaoka, Suita 565-0871, Japan S Supporting Information *

ABSTRACT: To enhance the glucose sensitivity and selfregulated release of insulin, biobased capsules with glucoseresponsive and competitive properties were fabricated based on poly(γ-glutamic acid) (γ-PGA) and chitosan oligosaccharide (CS) polyelectrolytes. First, poly(γ-glutamic acid)-g-3aminophenylboronic acid) (γ-PGA-g-APBA) and galactosylated chitosan oligosaccharide (GC) were synthesized by grafting APBA and lactobionic acid (LA) to γ-PGA and CS, respectively. The (γ-PGA-g-APBA/GC)5 capsules were then prepared by layer-by-layer (LBL) assembly of γ-PGA-g-APBA and GC via electrostatic interaction. The size and morphology of the particles and capsules were investigated by DLS, SEM, and TEM. The size of the (γ-PGA-g-APBA/GC)5 capsules increased with increasing glucose concentration due to the swelling of the capsules. The capsules could be dissociated at high glucose concentration due to the breaking of the cross-linking bonds between APBA and LA by the competitive reaction of APBA with glucose. The encapsulated insulin was able to undergo selfregulated release from the capsules depending on the glucose level and APBA composition. The amount of insulin release increased with incubation in higher glucose concentration and decreased with higher APBA composition. Moreover, the on−off regulation of insulin release from the (γ-PGA-g-APBA/GC)5 capsules could be triggered with a synchronizing and variation of the external glucose concentration, whereas the capsules without the LA functional groups did not show the on−off regulated release. Furthermore, the (γ-PGA-g-APBA/GC)5 capsules are biocompatible. These (γ-PGA-g-APBA/GC)5 with good stability, glucose response, and controlled insulin delivery are expected to be used for future applications to glucose-triggered insulin delivery. KEYWORDS: insulin delivery, glucose response, biobased polymer, layer-by-layer self-assembly, diabetes mellitus (PBA).7−11 Among them, PBA and its derivatives have attracted much attention in glucose-responsive systems12 because of their higher stability than the protein-based GOx and Con A systems, easier design by incorporation of the differently functional molecules, and effective protection of the loaded insulin from damage by the ambient environment. PBA derivatives can conjugate with the 1,2-diol and 1,3-diol moieties in glucose to form a reversible ester bond for glucose-responsive materials.13−15 Moreover, these derivatives are in equilibrium between an uncharged form, which is hydrophobic, and a charged form, which is made hydrophilic by complexing with glucose.16,17 Based on these properties of the PBA derivatives, closed-loop nanoparticles, capsules, films, and gels for glucose detection and insulin release have been reported.18−21 Among them, layer-by-

1. INTRODUCTION More than 3 billion people suffered from diabetes mellitus (DM) in the world in 2015. DM is caused by dysfunctions resulting in glucose concentrations.1 A high blood glucose level can cause several complications and may even be lifethreatening.2 For maintaining the normoglycemia, a closedloop system that combines blood glucose monitoring and selfadministration of insulin with a precise dosage is required. Stimuli response is an efficient way to construct the closedloop system because stimuli-responsive polymer materials have the outstanding ability to release an effective drug dosage at a diseased region at a desired time in response to environmental changes,3,4 such as pH, temperature, and light. Thus, these intelligent polymers can be employed as a system for continuous glucose sensing and self-regulating release of insulin.5,6 Glucosemediating systems consist of glucose-responsive elements, including the glucose-binding protein concanavalin A (Con A), glucose oxidase enzyme (GOx), and phenylboronic acid © XXXX American Chemical Society

Received: February 20, 2016 Accepted: May 23, 2016

A

DOI: 10.1021/acsami.6b02121 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Scheme 1. Fabrication, Drug Loading, and Release of Polyelectrolyte Capsules Based on γ-PGA-g-APBA and GC

Scheme 1. Poly(γ-glutamic acid) (γ-PGA) and chitosan oligosaccharide (CS), which are two of the most widely investigated biomedical materials, were employed as the biological polyelectrolytes.32,33 Glucose-responsive poly(γ-glutamic acid-g-3-aminophenylboronic acid) (γ-PGA-g-APBA) polymers were synthesized by grafting functional 3-aminophenylboronic acid (APBA) to the γ-PGA chains with various APBA compositions. For precise, glucose-regulated release of drugs, galactosylated chitosan oligosaccharide (GC) was prepared using lactobionic acid (LA) to modify CS to construct a competitive system. The capsules were fabricated with multilayers by LBL self-assembly of γ-PGA-g-APBA and GC through electrostatic interaction on the silica nanoparticles and subsequent removal of the silica nanoparticles. The glucose sensitivity and insulin release were investigated in detail. The capsules swelled and dissociated at high glucose level, and the encapsulated insulin could undergo self-regulated release from the capsules, depending on the glucose concentration. Moreover, the biocompatibility of the capsules was also investigated.

layer (LBL) self-asssembled capsules have attracted much attention for the development of closed-loop systems.22,23 De Smedt and co-workers 24 fabricated PBA-based glucoseresponsive capsules by LBL self-assembly of a phenylboronic acid-modified polycation and a polyanion. The addition of glucose led to rapid dissolution of the capsules, which was induced by the change in electrostatic interactions between the polyanion and polycation. Levy et al.25 also prepared glucosesensitive films and capsules using a PBA-modified PAA and polysaccharide using the LBL technique via the ester bonds between the PBA moieties and diols. Unfortunately, these glucose responses were only observed at pH 9.0 or above. PBA-based polymer materials do not have glucose sensitivity under physiological conditions (pH 7.4) because they are weak acids with a pKa of 8.2−8.6.26 Since 1994, Kataoka et al. have enhanced the sensitivity of PBA-based polymer materials to glucose by decreasing the pKa by the substitution of electron-withdrawing groups at the para or ortho position of the boronic acid group.27,28 Shi et al. reported a series of self-assembled micelles that contained PBA, showing glucose responsiveness under physiological conditions.29,30 Zhang groups attempted to construct glucose-responsive films using PBA-bearing PAA and poly(vinyl alcohol) (PVA) through LBL assembly at neutral pH.31 Significant achievements have been made in decreasing the apparent pKa for glucose-responsive material application under physiological pH and enhancing the rate of responsiveness to glucose. However, challenges in the glucose-triggered insulin delivery system persist, especially for demonstrating a method that combines (1) high sensitivity and on−off regulated response to the blood glucose level; (2) high loading efficiency of insulin for simple long-lasting injections; and (3) biocompatibility without long-term side effects. To achieve the above objectives, we describe the construction of biobased polyelectrolyte capsules with multilayers that are designed for the combined detection of glucose and selfregulated insulin release based on the glucose level, as shown in

2. EXPERIMENTAL PART 2.1. Materials. N-(3-(Dimethylamino)propyl)-N′-ethylcarbodiimide (EDC), lactobionic acid (LA), 3-aminophenylboronic acid (APBA), N-hydroxysuccinimide (NHS), aminopropyltriethoxysilane, and insulin were purchased from Aladdin. Chitosan oligosaccharide (CS) with a number-average molecular weight of 10 kDa and 90% deacetylation degree was purchased from Nanjing Weikang Biotechnology Co., Ltd. and was used without further purification. Poly(γglutamic acid) (γ-PGA) with a molecular weight range of approximately 200−500 kDa was provided by Wako Pure Chemical Co., Ltd. L929 cells were purchased from Biomedical Co., Ltd. Dimethyl sulfoxide (DMSO) and 3-(4,5-dimethylthiazol-2-thiazolyl)-2,5-diphenyltetrazolium bromide (MTT) were purchased from Amersco. All the other reagents and solvents used in this research were purchased from Sinopharm Chemical Reagent Co., Ltd., Shanghai, China, and were used as received. 2.2. Characterization. Proton nuclear magnetic resonance (1H NMR) spectroscopy (400 MHz, Bruker Corp., Germany) was used to B

DOI: 10.1021/acsami.6b02121 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces Scheme 2. Synthesis of GC (a) and γ-PGA-g-APBA (b)

confirm the polymer structure using D2O as solvent at room temperature. The average particle size and size distribution of the nanoparticles and capsules were characterized by dynamic light scattering (DLS) with an ALV-5000/E DLS instrument (Hosic Limited, Germany) at a fixed scattering angle of 90°, after being filtered by 0.8 μm Milli-pore filters. Zeta potential measurement was performed on a Malvern Mastersizer 2000 (Brookhaven Instruments Corp., USA) to detect the surface charge of the functional polyelectrolytes averaged over 20 runs. The morphologies of the obtained nanoparticles and capsules were determined by scanning electron microscopy (SEM, Hitachi S-4800, Japan) at 20 kV and transmission electron microscopy (TEM) with a JEOL JEM-2100 microscope (Hitachi, Japan) under a voltage of 200 kV. The cumulative amounts of insulin loading and releasing in the supernatants were determined by a UV−vis spectrometer (Persee, China) using a standard insulin curve upon various concentrations. Circular dichroism (CD) spectra of the native and released insulin from the capsules at 200−300 nm in PBS buffer were obtained by a CD spectropolarimeter (Jasco-715, Welltech Enterprises, Inc. Maryland, United States) at room temperature. 2.3. Modification of the Biopolymers. 2.3.1. Synthesis of Galactosylated-Modified Chitosan Oligosaccharide (GC). Galactosylated-modified chitosan oligosaccharide (GC) was synthesized using EDC and NHS as coupling agents,34,35 as shown in Scheme 2a. LA (5 mmol) was dissolved in 120 mL of distilled water containing EDC (5 mmol) and NHS (15 mmol) coupling agents, to activate the carboxyl groups for 30 min. CS (5 mmol) was subsequently added to the LA solution, and the pH of the mixture was adjusted to 4.7 using 0.01 M HCl under magnetic stirring for 72 h at room temperature. The resulting product GC was purified by dialyzing against distilled water for 7 days with dialysis membranes (MWCO 5000−10 000), followed by lyophilization. 2.3.2. Synthesis of 3-Aminophenylboronic Acid-Modified Poly(γglutamic acid) (γ-PGA-g-APBA). Synthesis of 3-aminophenylboronic acid-modified poly(γ-glutamic acid) (γ-PGA-g-APBA) (Scheme 2b) was the same as GC. Briefly, EDC and NHS were added into a DMSO (150 mL) solution of γ-PGA (6 mmol) at room temperature, and the solution was kept stirring for 30 min. After APBA (4 mmol) was added to the above solution, the pH was adjusted to 4.7 to further react for 72 h at room temperature. The product was dialyzed against 0.1 M NaCl for 3 days and against distilled water for 4 days using dialysis membranes (MWCO 8000−14 000). Finally, γ-PGA-g-APBA was obtained by lyophilization. By changing the amount of APBA, a series of γ-PGA-g-APBA with various grafting degrees was prepared, and the results are shown in Table S1. 2.4. LbL Assembly of γ-PGA-g-APBA/GC Multilayers on the SiO2 Nanoparticles. SiO2 nanoparticles were prepared as previously reported.36 Amine-modified SiO2 nanoparticles (SiO2−NH2) with ζpotential at +23.8 mV were obtained using aminopropyltriethoxysilane

as a silane coupling agent and were then dispersed in water. Fabrication of the γ-PGA-g-APBA/GC multilayers on the SiO2 nanoparticles was performed by LBL self-assembly in pH 5.0 buffer solution37,38 (Scheme 1). A 20 mL volume of γ-PGA-g-APBA solution (2 mg/mL) was added dropwise into the SiO2−NH2 suspension for 30 min at ambient temperature. The excess γ-PGA-g-APBA polymer was removed by washing with ultrapure water and centrifuging three times. Then, (γPGA-g-APBA)@SiO2 (represented as shell@core) nanoparticles were redispersed in buffer solution (20 mL) at pH 5.0 and were added to the GC solution (2 mg/mL, 20 mL) under magnetic stirring for 30 min at ambient temperature. The (γ-PGA-g-APBA/GC)@SiO2 nanoparticles were obtained by washing with ultrapure water to remove excess GC and centrifuging three times. The same procedure was repeated 5 times to obtain the (γ-PGA-g-APBA/GC)5@SiO2 nanoparticles. For comparison, two types of nanoparticles, (γ-PGA/CS)5@SiO2 and (γ-PGA-g-APBA/CS)5@SiO2, were also fabricated using the unmodified γ-PGA and CS biopolymers as multilayers, using processes similar to the (γ-PGA-g-APBA/GC)5@SiO2 nanoparticles. The fabrication of insulin-loaded (γ-PGA-g-APBA/GC)5@SiO2 nanoparticles was the same as the above procedure. Prior to the γPGA-g-APBA/GC multilayer assembly on the SiO2 nanoparticles, 5 mL of insulin with a concentration of 2 mg/mL in pH 5.0 buffer solution was added to the SiO2−NH2 nanoparticle suspension (20 mL) under gentle stirring at 0 °C for 30 min. The insulin-loaded (γ-PGA-g-APBA/ GC)5@SiO2 nanoparticles were obtained by LBL assembly of the γPGA-g-APBA and GC polyelectrolytes on the silica nanoparticles by repeating the process 5 times. The entrapment efficiency (EE) and entrapment capacity (EC) of the insulin in the capsules were determined by the UV absorption band of the supernatant at 595 nm using a standard insulin curve at various concentrations (shown in Figure S7). EE and EC were calculated from the following equations

EE% =

Total insulin − Free insulin × 100 Total insulin

(a)

EC% =

Total insulin − Free insulin × 100 NP weight

(b)

2.5. Fabrication of Insulin-Loaded (γ-PGA-g-APBA/GC)5 Capsules. The (γ-PGA-g-APBA/GC)5@SiO2 nanoparticles (40 mg) were redispersed in 100 mL of NH4F/HF (8 M/2 M) buffer solution at pH 5.0 for 12 h under stirring at room temperature. Then, the SiO2 template was removed. The formed (γ-PGA-g-APBA/GC)5 hollow capsules were collected after centrifuged, washed with deionized water three times to remove residual impurities, and then freeze-dried. 2.6. Glucose Response of (γ-PGA-g-APBA/GC)5 Capsules in PBS. The glucose response of the (γ-PGA-g-APBA/GC)5 capsules was studied by measuring the diameter and morphology changes of the capsules in the presence of glucose. Briefly, the capsules were dispersed in water at 2 mg/mL. A designed volume of the capsule suspension was C

DOI: 10.1021/acsami.6b02121 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces added to the glucose solutions, and the final glucose concentration was controlled to 0, 2, 6, 20, or 50 mg/mL. The diameter and morphology of the capsules before and after glucose treatment were measured for every 2 h by DLS and TEM, respectively. 2.7. In Vitro Insulin Release. The release behavior of insulin from the (γ-PGA-g-APBA/GC)5 capsules was investigated by measuring the amount of insulin released in the presence of glucose. The insulinloaded capsules were first dispersed in PBS buffer solution at 2 mg/mL. Then, the capsules were dialyzed into PBS solution containing glucose with various concentrations at 37 °C. At predetermined time points, the amount of the insulin released from the capsules was determined by the UV absorption band at 595 nm according to the standard insulin curve. The secondary structure of the released insulin was measured by a circular dichroism (CD) spectropolarimeter by scanning in the range of 200−300 nm at room temperature. “On−off” regulation of glucose was performed as follows. The capsules were first dialyzed into PBS solution (pH 7.4) with a glucose concentration of 20 mg/mL (or 6 mg/mL) for 3 h. Subsequently, the capsules were removed from the solution and immediately placed into PBS solution (pH 7.4) without glucose for another 3 h. Repeatedly changing the dialysis solutions with and without glucose, the amount of the insulin released from the capsules for each time was then determined by the UV absorption band at 595 nm. 2.8. Cell Viability. Cell viability of the capsules was evaluated using L929 cells as model cells. The cells were cultured in a 96-well plate (3 × 104 cells/well) in complete Dulbecco’s modified Eagle’s medium (DMEM, with 10% fetal bovine serum and penicillin-streptomycin supplemented) in a humidified atmosphere of 5% CO2 at 37 °C. The capsules were diluted with culture medium to achieve the predetermined concentrations. After 24 h of incubation, the growth medium was replaced with 100 mL of DMEM culture medium. Cells treated with the same amount of PBS were used as a control group and were incubated for an additional 24 h. The cell viability was assayed by adding 20 μL of PBS solution of MTT (2 mg/mL). After incubation at 37 °C for another 4 h, the formed crystals were dissolved in 150 mL of DMSO and were recorded at 595 nm with a reference wavelength of 570 nm using a 96-well plate ELLSA microplate reader (Infinite M200 Pro, Tecan, Männedorf, Switzerland). All measurements were performed in triplicate.

chemical shifts at δ = 3.0−4.0 in Figure 1a were assigned to the protons of the glucosamine units in CS. The peak at δ = 2.05 belonged to the proton of the methyl group, which was from partially acetylated chitosan. The characteristic peak of the LA group appeared in the 1H NMR spectrum of GC (Figure 1b) at δ = 4.1−4.3, suggesting the successful preparation of the GC polymer. FTIR spectrum (Figure S1) also showed the same result. The degree of substitution of LA in GC was estimated to be 15.4 mol % by calculating the peak areas between the original acetamide group of CS at δ = 2.05 and the LA group at δ = 4.16. Amide group modification of phenylboronic acid at the metaposition results in glucose response under physiological conditions. Therefore, 3-aminophenylboronic acid (APBA) was used as a functional group to modify γ-PGA for preparation of the γ-PGA-g-APBA polymer. The chemical structure of the obtained γ-PGA-g-APBA polymer was estimated by 1H NMR spectra, as shown in Figure 2. Peaks at δ = 2.2, 1.2, and 1.7 were

3. RESULTS AND DISCUSSION 3.1. Modification and Characterization of the Biobased Polymers. GC was synthesized by grafting LA to CS in the presence of EDC and NHS, as shown in Scheme 2a. 1H NMR spectra of CS and GC are shown in Figure 1. The

Figure 2. 1H NMR spectra of γ-PGA (a) and γ-PGA-g-APBA0.12 (b), γPGA-g-APBA0.29 (c), and γ-PGA-g-APBA0.40 (d).

due to the characteristic protons of the methylene groups in γPGA, and the peak at δ = 4.1 was assigned to the methylidyne protons adjacent to the carbonyl group in Figure 2a. In the 1H NMR spectrum of the typical γ-PGA-g-APBA polymer (Figure 2b), the peaks of the benzyl group at δ = 7.2−7.7 indicated that APBA was introduced into γ-PGA by amidation. Moreover, the proton of the methine groups was shifted from δ = 4.1 to δ = 4.3, due to the formation of the amide bond. These results confirmed the successful preparation of the γ-PGA-g-APBA polymer. To investigate the effect of the APBA composition in the polymer on the glucose sensitivity, three types of γ-PGA-gAPBA polymers with various APBA compositions were synthesized by changing the molar ratio of APBA and γ-PGA (as listed Table S1), and the 1H NMR spectra are shown in Figure 2b−d. With increment of the APBA molar ratio, the peak areas of the benzyl group (δ = 7.2−7.7) increased. By comparison of the peak areas between the benzyl groups and the methylene groups in γ-PGA (δ = 1.7−2.2), the composition of APBA in the polymers was 12, 29, and 40 mol % for the feed molar ratios of γ-PGA to APBA of 1:0.25, 1:0.5, and 1:0.75,

Figure 1. 1H NMR spectra of CS (a) and GC (b). D

DOI: 10.1021/acsami.6b02121 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces respectively. For clarity, these polymers were abbreviated as γPGA-g-APBA0.12, γ-PGA-g-APBA0.29, and γ-PGA-g-APBA0.40. 3.2. Fabrication of (γ-PGA-g-APBA/GC)5@SiO2 Nanoparticles and Capsules. Nanoparticles based on the γ-PGA-gAPBA and GC polyelectrolytes were fabricated by LBL assembly using SiO2 nanoparticles as a template. Prior to the LBL assembly, amino functional SiO2 nanoparticles (SiO2−NH2) were prepared by aminopropyltriethoxysilane modification of the SiO2 nanoparticles, and the zeta potential of +23.8 mV was measured by Zeta PALS (Figure 3). Then, the polyanion, γ-

diameter of approximately 195 nm (Figure 4a). After successively coating the polyelectrolytes on the silica nanoparticles at pH 5.0, the uniform (γ-PGA-g-APBA0.12/GC)5@ SiO2 nanoparticles with core−shell structures were observed to have a rough surface and a diameter of approximately 235 nm (Figure 4b), which was increased by approximately 40 nm compared with the silica nanoparticles, indicating the successful assembly of the polyelectrolytes on the nanoparticles. The DLS results also indicated the increased size of the multilayer-coated nanoparticles with a narrow particle size distribution, corresponding to the results from the TEM images, as show in Figure S3. By removing the SiO2 template in NH4F/HF (8 M/2 M) buffer solution, (γ-PGA-g-APBA/GC)5 capsules with hollow structure were obtained (Figure 4c). The diameter of the (γPGA-g-APBA/GC)5 capsules was 240 nm without the templates, which was similar to that of the (γ-PGA-gAPBA0.12/GC)5)@SiO2 nanoparticles. The size distribution of the capsules characterized by DLS measurements was narrow (Figure S3). After removing the template, the (γ-PGA-g-APBA/ GC)5 capsule was dispersed in PBS buffer (pH 7.4) and then dried by lyophilization. The obtained capsules maintained the unique hollow structures, as observed in the TEM and SEM images (Figure S4a and S4b). DLS measurement showed that the diameter of the capsules after freeze-drying was approximately 165 nm (Figure S4c). When the (γ-PGA-g-APBA/GC)5 capsules were dispersed in PBS solution (pH 7.4), boronate esters could be formed between the boronic acid groups in APBA and the 1,2-diol groups in LA, which cross-linked the multilayers with each other. These cross-linking bonds can cause close packing of the cross-linked polymer chains, resulting in the size reduction of the capsules after freeze-drying (from 240 nm to about 165 nm). Moreover, the cross-linking bonds kept the capsules more stable. 3.3. Glucose Sensitivity of the Polyelectrolyte Capsules. The glucose sensitivity of phenylboronic acid at physiological pH was achieved in the presence of sugars. Accordingly, APBA-modified capsules can demonstrate glucose sensitivity. Thus, it is necessary to investigate the changes of the capsule size and morphology during glucose response. The effect of the glucose concentration on the diameter of the (γPGA-g-APBA/GC)5 capsules was determined, and the result is shown in Figure 5. Without glucose treatment, the capsules ((γPGA-g-APBA0.12/GC)5 as an example, in Figure 5a) were relatively stable, and the size increased slowly with time in the buffer solution at pH 7.4. After the capsules were placed into the glucose solution, vigorous size change was observed, and the increment was dependent on the glucose concentration. For the low glucose level of 2 mg/mL, the diameter of the capsules increased from 150 to 208 nm after 4 h and to 315 nm after 36 h. The TEM images of the capsule morphologies with time after glucose treatment are shown in Figure 6a−d. With increasing time, the size of the capsules gradually increased from about 135 to 260 nm. Moreover, the capsules maintained the spherical morphology without dissociation of the multilayers, indicating that the capsules swelled in glucose solution but maintained a relatively stable structure. After changing the glucose level to 6 mg/mL (the critical blood sugar concentration for diabetes mellitus), the diameter of the (γ-PGA-g-APBA 0.12 /GC) 5 capsules increased from 150 to 350 nm after 36 h (Figure 5a). When incubating the capsules in higher glucose concentrations of 10 mg/mL, 20 mg/mL, and even 50 mg/mL, the (γ-PGA-gAPBA0.12/GC)5 capsules were significantly swollen, and the

Figure 3. Zeta potential values of (γ-PGA-g-APBA0.12/GC)5@SiO2 nanoparticles by LBL assembly.

PGA-g-APBA0.12 (as an example), could be assembled onto the SiO2−NH2 nanoparticles at pH 5.0. The zeta potential of the (γPGA-g-APBA0.12)@SiO2 nanoparticles changed to −27.86 mV, indicating the successful assembly of the γ-PGA-g-APBA polyanion. The GC polycation was subsequently assembled onto the (γ-PGA-g-APBA0.12)@SiO2 nanoparticles by the driving force of the electrostatic interaction. The zeta potential of the obtained (γ-PGA-g-APBA0.12/GC)@SiO2 nanoparticles completely reversed polarity to +21.4 mV due to the GC polycation coating on the particle surface. While further alternating the γ-PGA-g-APBA0.12 and GC polyelectrolytes to continue of the LBL assembly process, the zeta potential values relatively alternated between negative and positive, as shown in Figure 3. This LBL assembly process could be repeated at least 11 times in our system. These results suggest that the γ-PGA-gAPBA and GC polyelectrolytes could form multilayers on the silica nanoparticles. The size and morphology of the nanoparticles coated with multilayers were investigated by TEM, as shown in Figure 4. The silica nanoparticles were spherical and smooth with a

Figure 4. TEM images of the SiO2 nanospheres (a), (γ-PGA-g-APBA/ GC)5@SiO2 nanoparticles (b), and (γ-PGA-g-APBA/GC)5 capsules (c). E

DOI: 10.1021/acsami.6b02121 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 6. TEM images of the (γ-PGA-g-APBA0.12/GC)5 capsules with time at glucose concentration of 2 mg/mL (a−d) and 20 mg/mL (e− h). Figure 5. Glucose sensitivity of the (γ-PGA-g-APBA0.12/GC)5 (a), (γPGA-g-APBA0.29/GC)5 (b), and (γ-PGA-g-APBA0.40/GC)5 (c) capsules in PBS solution (pH 7.4) at different glucose levels at 37 °C.

Scheme 3. Schematic Representation of the Competitive Reaction in the Capsules with and without the Addition of Glucose

sizes increased to 400 nm, 450 nm, and 480 nm, respectively, in 36 h. The morphology of the (γ-PGA-g-APBA0.12/GC)5 capsules in the high-concentration glucose solution (20 mg/mL as an example) is shown in Figure 6e−h based on the TEM observation. From the TEM image (Figure 6e), the capsules significantly swelled with a hollow structure in the first 4 h of treatment. The capsules were partially destroyed at 8 h (Figure 6f) and completely disappeared after more than 24 h (Figure 6g,h), indicating the assembled multilayers dissociated at high glucose levels. Two actions occurred in the interior of the (γ-PGA-g-APBA/ GC)5 capsules when incubated at physiological conditions in the presence of glucose. The first was that glucose gradually diffused into the interlayers of the capsule and conjugated with APBA to form a more hydrophilic polymer layer, resulting in incompact multilayers and swelling of the capsules. Second, the interaction between APBA and LA decreased or disappeared due to the competitive interaction of APBA and glucose, breaking the cross-linked structure among the multilayers, as shown in Scheme 3. In the case of a low glucose concentration, glucose diffused into the polyelectrolyte multilayers and then reacted with part of the APBA groups to make the layer more

hydrophilic, loosening the polyelectrolyte layers. However, the unbroken cross-linking bonds between APBA and LA maintained the capsule stability. Accordingly, the size of the (γ-PGA-g-APBA/GC)5 capsules increased, but the morphology of the capsules remained almost constant. When the medium was changed to the high glucose concentration, the glucose diffused into the multilayers to react with most of the APBA groups, and most of the interactions between the APBA and LA groups were broken, resulting in significant swelling and even destruction of the capsules. To further investigate the influence of the APBA composition on the glucose sensitivity, two more capsules with various APBA F

DOI: 10.1021/acsami.6b02121 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 7. Cumulative release of insulin from the (γ-PGA-g-APBA0.12/GC)5 (a), (γ-PGA-g-APBA0.29/GC)5 (b), and (γ-PGA-g-APBA0.40/GC)5 (c) capsules after exposure to glucose solutions with various levels over time and (d) cumulative release of insulin after 12 h at 37 °C.

groups. The diameter of the (γ-PGA-g-APBA0.29/CS)5 capsules increased with increasing time and glucose concentration (Figure S6b), similarly to the (γ-PGA-g-APBA0.29 /GC) 5 capsules. Although the (γ-PGA-g-APBA0.29/CS)5 capsules also showed glucose sensitivity, their size change was larger than that of the (γ-PGA-g-APBA0.29/GC)5 capsules. There was no competitive reaction between APBA with LA and APBA with sugar because of the (γ-PGA-g-APBA0.29/CS)5 capsules without the LA groups. Thus, the APBA moiety reacted with glucose immediately, and the size of the capsules swelled quickly due to the lack of cross-linking bonds formed between the APBA and LA groups. Accordingly, the (γ-PGA-g-APBA/GC)5 capsules were more stable and may have glucose-triggered properties for potential application in drug-controlled release systems. 3.4. In Vitro Insulin Release. To demonstrate the selfregulated drug release from the polyelectrolyte capsules under physiological conditions, insulin as a model drug was used to monitor the glucose-triggered release. Insulin is a proteincontaining hydrophilic and hydrophobic amino acid residue that can be loaded onto the SiO2 nanoparticles and subsequently encapsulated into the polymer capsules by electrostatic and hydrogen bonding interactions at pH 5.0, as reported by previous papers.13,18,19,39 The encapsulation efficiency into the capsules was 63%, 60%, and 59% for (γ-PGA-g-APBA0.12/GC)5, (γ-PGA-g-APBA0.29/GC)5, and (γ-PGA-g-APBA0.40/GC)5, respectively, based on the standard curve of insulin (Figure S7). The encapsulation efficiency in this study was higher than the reported micelles due to the hollow structure of the (γ-PGA-gAPBA/GC)5 capsules. The in vitro release profile of insulin from the (γ-PGA-g-APBA/GC)5 capsules in response to different glucose levels was evaluated by dispersing the insulin-loaded capsules into the glucose solutions, as shown in Figure 7. Less than 17% of the insulin was released from the (γPGA-g-APBA0.12/GC)5 capsules after incubation for 12 h in buffer solution without glucose (Figure 7a), which was induced

grafting degrees, (γ-PGA-g-APBA0.29/GC)5 and (γ-PGA-gAPBA0.40/GC)5, were manufactured to study the size change with time at different glucose levels (Figure 5b and 5c). The size changes of the (γ-PGA-g-APBA0.29/GC)5 and (γ-PGA-gAPBA0.40/GC)5 capsules were the same as those of the (γPGA-g-APBA0.12/GC)5 in the system without glucose. After the addition of glucose, the diameter of the capsules increased with increasing glucose concentration, which was similar to the previous results. However, the size changes decreased at higher APBA composition. The diameters were 474, 375, and 320 nm for the (γ-PGA-g-APBA0.12/GC)5, (γ-PGA-g-APBA0.29/GC)5, and (γ-PGA-g-APBA0.40/GC)5 capsules, respectively, when stimulated by a glucose solution at 50 mg/mL. The morphologies of the (γ-PGA-g-APBA0.29/GC)5 and (γ-PGA-gAPBA0.40/GC)5 capsules were observed with the addition of the glucose at 20 mg/mL by TEM, as shown in Figure S5. For the (γ-PGA-g-APBA0.29/GC)5 capsules, the capsule structure was partially destroyed with time (Figure S5a−c), whereas the (γPGA-g-APBA0.40/GC)5 capsules maintained the structure with a slight increase of size (Figure S5d−f). These results indicated that the capsules with high APBA composition were more stable. At the same glucose concentration, a similar amount of APBA groups in the three capsules can react with glucose. Excess interactions between the LA and APBA groups occurred to form the cross-linked bonds at the higher APBA composition, which was the driving force to keep the capsules without morphology destruction. Therefore, the capsules were more stable at higher APBA composition. For a better understanding of the advantage of the (γ-PGA-gAPBA/GC)5 capsules with respect to glucose response, the (γPGA/CS)5 and (γ-PGA-g-APBA0.29/CS)5 capsules were fabricated to study the glucose sensitivity based on the unmodified γ-PGA and CS biopolymers. For the (γ-PGA/CS)5 capsules, the size increased with incubation time independent of the glucose concentration (Figure S6a), due to a lack of APBA functional G

DOI: 10.1021/acsami.6b02121 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 8. Cumulative release of insulin from the (γ-PGA-g-APBA0.29/GC)5 capsules in PBS triggered by alternating glucose concentrations of 20−0 (a) and 6−0 (b) at 37 °C.

These cross-linking bonds kept the capsules more stable with less swelling. Moreover, glucose and LA competitively reacted with APBA, which retarded the swelling and dissociation of the capsules (Scheme 3), resulting in the slow release of insulin from the capsules, as mentioned above. On the contrary, in the absence of LA groups in the capsules, there were no crosslinking bonds and competition in the glucose solution, leading to rapid release of insulin from the (γ-PGA-g-APBA0.29/CS)5 capsules (Figure S8 at 20 mg/mL of glucose concentration). Therefore, the polyelectrolyte capsules overcome the defect of the lacking glucose response and sustained release by introducing the functional LA and APBA groups into the γPGA and GC polymers. Generally, the glucose level fluctuates from hunger to satiation. Thus, the continuous glucose change with regulated insulin release was designed to simulate a person who suffers from diabetes. The release of insulin with “on−off” regulation of glucose was performed by alternately immersing the capsules into PBS solutions with and without glucose, and the result is shown in Figure 8. Vigorous increase of the release of insulin from the (γ-PGA-g-APBA0.29/GC)5 capsules with 20 mg/mL of glucose was observed from 0 to 15% for 3 h. After changing the glucose level to 0, the release speed quickly decreased, and the release behavior was almost shut off (Figure 8a). After returning the glucose level to 20 mg/mL, the release switched to the rapid delivery of insulin. For a glucose concentration of 6 mg/mL, the release behavior of insulin can be regulated by the “on−off” glucose level (Figure 8b), but the amount of insulin released during the glucose “on” was lower than in the case of 20 mg/mL of glucose. The on−off regulation of the insulin release could be repeated for at least two cycles with a synchronizing manner and a variation of the external glucose concentration. These results were consistent with the glucose-triggered insulin release profiles. However, in the absence of LA groups in the capsules, insulin release from the (γ-PGA-g-APBA0.29/CS)5 capsules could not be achieved with “on−off” regulation by glucose (Figure S8). On the contrary, insulin was continuously and quickly released because of the lack of competition between LA and glucose for APBA. Therefore, the (γ-PGA-g-APBA/GC)5 capsules have the advantage of controlled release and can be used in the diagnosis and treatment of diabetes. To evaluate the changes of the activity and structure of insulin released from the capsules, circular dichroism (CD) spectroscopy was performed as a common method to analyze the

by an osmotic-release phase. This small cumulative release was attributed to the capsules having a multilayer structure, which could improve upon the defect of high permeability for a single polyelectrolyte layer. Sustained release of insulin was observed in the (γ-PGA-g-APBA0.12/GC)5 capsules at 2 mg/mL of glucose. The release of insulin was gradual to 30% in 12 h. This basal release rate of insulin at the normoglycemic level is desirable in insulin-dependent therapy to manage blood-glucose fluctuations.40 The amount of insulin released increased with increasing glucose concentration. At the critical blood sugar concentration (6 mg/mL), the rapid release of insulin increased from 12% to 60% for 4 to 12 h. When the glucose concentration in the media was at 20 mg/mL, the phenomenon of a fast release occurred from 1 to 12 h, and the amount of insulin released was from 4% to 75%. However, the insulin release increased gradually from 12 to 36 h, independent of the glucose concentration. These results demonstrated that the insulin release from the (γ-PGA-g-APBA/GC)5 capsules was triggered by the glucose concentration. Because there is a competitive reaction between APBA with LA and APBA with glucose, glucose diffused into the multilayers of the capsules to react with APBA, and the cross-linking bonds of LA and APBA were broken, resulting in swelling and even dissociation of the capsules with time. Thus, the breaking of the cross-linking bonds, the swelling of the capsules, and the enhanced permeability were synergetic effects, which resulted in the gradual release of insulin. However, the dissociation of the capsules at the high glucose level was the primary driving force of insulin release, leading the rapid release of insulin. The release profile of insulin was also detected after changing the APBA composition, as shown in Figure 7b and 7c. The (γPGA-g-APBA0.29/GC)5 capsules gradually released insulin at the low glucose level and quickly released insulin at the high level (Figure 7b), similarly to the (γ-PGA-g-APBA0.12/GC)5 capsules. Interestingly, the (γ-PGA-g-APBA0.40/GC)5 capsules showed delayed release, even at the high glucose concentration at 20 mg/mL (Figure 7c). Moreover, the amount of insulin released decreased with increasing APBA concentration (Figure 7d). The amount of insulin released was 30%, 18%, and 17% at 2 mg/mL of glucose and increased to 75%, 62%, and 40% at 20 mg/mL within 12 h for (γ-PGA-g-APBA0.12/GC)5, (γ-PGA-g-APBA0.29/ GC)5, and (γ-PGA-g-APBA0.40/GC)5, respectively. As the APBA composition in (γ-PGA-g-APBA/GC)5 capsules increased, more LA-mediated association occurred to form cross-linking bonds. H

DOI: 10.1021/acsami.6b02121 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces secondary structure of a protein with high reliability.41 In the CD spectra of the native insulin in PBS (pH 7.4) and at pH 5.5, there were two extrema at 210 and 220 nm (Figure S9a and S9b), due to the α-helix structure and β-structure, respectively. The CD spectrum of the insulin after releasing for 12 h was similar to that of the native insulin with two extrema, as shown in Figure S9c. The ratios between bands ([φ]205/[φ]222) for the native and released insulin were 1.26 and 1.22, respectively. Therefore, the secondary structure of the insulin released from the capsules was similar to the original insulin. Accordingly, the released insulin maintained its structure and properties. 3.5. Biocompatibility. In order to verify the noncytotoxicity of the (γ-PGA-g-APBA/GC)5 capsules, the biocompatibility of the capsules was estimated using L929 cells as model cells in the MTT assay method. Figure S10 shows inverted microscope images of the cells after incubation of the (γ-PGA-g-APBA/ GC)5 capsules for 24 h at 37 °C. The cells exhibited good growth with a fibrous structure. The cell viabilities were approximately 105%, 100%, and 96% for the (γ-PGA-gAPBA0.12/GC)5, (γ-PGA-g-APBA0.29/GC)5, and (γ-PGA-gAPBA0.40/GC)5 capsules, respectively, at a 0.1 mg/mL capsule concentration, which decreased with increasing amount of APBA in the capsules (Figure 9). Moreover, after increasing the

level. The cumulative release of insulin was approximately 80% at 20 mg/mL glucose within 12 h and increased with increasing glucose concentration, while it decreased with increasing APBA composition. Moreover, the on−off regulation of insulin release from the (γ-PGA-g-APBA/GC)5 capsules could be triggered by a synchronized variation of the external glucose concentration, whereas the capsules without the LA functional groups did not demonstrate the on−off regulated release. The cell viabilities of the capsules exceeded 88% for L929 cells, which indicated that the capsules had low cell toxicity and had good biocompatibility. These (γ-PGA-g-APBA/GC)5 capsules with good stability, glucose response, and controlled insulin delivery can be used in controlled release drug delivery systems.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.6b02121. (1) Synthesis and characterization of γ-PGA-g-APBA by FTIR spectra (Figures S1−S2, Table S1); (2) DLS results, TEM and SEM images of the particles and capsules (Figures S3−S4); (3) TEM images of the capsules with addition of glucose (Figure S5); (4) Glucose sensitivity of the capsules without functional group modification (Figure S6); (5) Insulin of the calibration curve and release from the (γ-PGA-gAPBA0.29/CS)5 capsules (Figure S7−S8); (6) CD spectra of the released insulin (Figure S9); (7) Inverted microscope images of cell adhesion on the capsules (Figure S10) (PDF)



AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Tel.: +86-510-85917019. Fax: +81-510-85917763. Notes

The authors declare no competing financial interest.



Figure 9. Cell viability of (γ-PGA-g-APBA/GC)5 capsule-treated L929 cells in the MTT assay after incubation for 24 h at 37 °C.

ACKNOWLEDGMENTS This study was supported by the National Nature Science Foundation of China (No. 51173072), the Fundamental Research Funds for the Central Universities (JUSRP51408B), the Key Laboratory of Food Colloids and Biotechnology, Ministry of Education, Jiangnan University (JDSJ2015-06), and MOE & SAFEA for the 111 Project (B13025).

concentration of the (γ-PGA-g-APBA/GC)5 capsules to 1 mg/ mL, the cell viabilities slightly decreased to 96%, 89%, and 88% for the (γ-PGA-g-APBA0.12/GC)5, (γ-PGA-g-APBA0.29/GC)5, and (γ-PGA-g-APBA0.40/GC)5 capsules, respectively. These results indicated that the boronated moieties had slight cytotoxic activity; however, the (γ-PGA-g-APBA/GC)5 capsules had good biocompatibility, due to excellent biocompatibility of the γ-PGA and CS biopolymers. These capsules have potential application to insulin delivery for the treatment of diabetes.



REFERENCES

(1) Ziegler, A. G.; Rewers, M.; Simell, O.; Simell, T.; Lempainen, J.; Steck, A.; Eisenbarth, G. S.; et al. Seroconversion to Multiple Islet Autoantibodies and Risk of Progression to Diabetes in Children. JAMA, J. Am. Med. Assoc. 2013, 309, 2473−2479. (2) Iyer, H.; Khedkar, A.; Verma, M. Oral Insulin−A Review of Current Status. Diabetes, Obes. Metab. 2010, 12, 179−185. (3) Thabit, H.; Hovorka, R. Bringing Closed-Loop Home: Recent Advances in Closed-Loop Insulin Delivery. Curr. Opin. Endocrinol., Diabetes Obes. 2014, 21, 95−101. (4) Shi, D. J.; Matsusaki, M.; Akashi, M. Preparation of Degradable and Photo-Responsive Crosslinked Nanoparticle Based on Cinnamic Acid Derivative as Protein Carriers. J. Controlled Release 2011, 149, 182−189. (5) Nan, F.; Wu, J.; Qi, F.; Fan, Q.; Ma, G.; Ngai, T. Preparation of Uniform-Sized Colloidosomes Based on Chitosan-Coated Alginate

4. CONCLUSIONS In conclusion, we designed and prepared two biopolymers γPGA-g-APBA and GC by modification with functional APBA and LA groups to enhance the glucose sensitivity for the regulated release of insulin. The (γ-PGA-g-APBA/GC)5 capsules were fabricated by LBL self-assembly of GC and γ-PGA-g-APBA with different APBA compositions using SiO2 nanoparticles as the template. The capsules swelled in glucose solution at a low level and dissociated at high glucose concentration due to the competitive reaction. The encapsulated insulin underwent selfregulated release from the capsules depending on the glucose I

DOI: 10.1021/acsami.6b02121 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces Particles and Its Application for Oral Insulin Delivery. J. Mater. Chem. B 2014, 2, 7403−7409. (6) Yogendraji, K. A.; Lokwani, P.; Singh, N. Newer Strategies for Insulin Delivery. Int. J. Res. Ayurveda Pharm. 2011, 2, 1717−1721. (7) Ye, T.; Yan, S.; Hu, Y.; Ding, L.; Wu, W. Synthesis and Volume Phase Transition of Concanavalin A-Based Glucose-Responsive Nanogels. Polym. Chem. 2014, 5, 186−194. (8) Alonso, B.; Armada, P. G.; Losada, J.; Cuadrado, I.; González, B.; Casado, C. M. Amperometric Enzyme Electrodes for Aerobic and Anaerobic Glucose Monitoring Prepared by Glucose Oxidase Immobilized in Mixed Ferrocene-Cobaltocenium Dendrimers. Biosens. Bioelectron. 2004, 19, 1617−1625. (9) Cella, L. N.; Chen, W.; Myung, N. V.; Mulchandani, A. SingleWalled Carbon Nanotube-Based Chemiresistive Affinity Biosensors for Small Molecules: Ultrasensitive Glucose Detection. J. Am. Chem. Soc. 2010, 132, 5024−5026. (10) Caves, M. S.; Derham, B. K.; Jezek, J.; Freedman, R. B. The Mechanism of Inactivation of Glucose Oxidase from Penicillium Amagasakiense under Ambient Storage Conditions. Enzyme Microb. Technol. 2011, 49, 79−87. (11) Wang, X. L.; Jiang, Z. Y.; Shi, J. F.; Liang, Y. P.; Zhang, C. H.; Wu, H. Metal-Organic Coordination-Enabled Layer-by-Layer Self-Assembly to Prepare Hybrid Microcapsules for Efficient Enzyme Immobilization. ACS Appl. Mater. Interfaces 2012, 4, 3476−3483. (12) Ma, R.; Shi, L. Q. Phenylboronic Acid-Based Glucose-Responsive Polymeric Nanoparticles: Synthesis and Applications in Drug Delivery. Polym. Chem. 2014, 5, 1503−1518. (13) Guo, Q. Q.; Zhang, T. Q.; An, J. X.; Wu, Z. M.; Zhao, Y.; Dai, X. M.; Zhang, X. G.; Li, C. X. Block versus Random Amphiphilic Glycopolymer Nanopaticles as Glucose-Responsive Vehicles. Biomacromolecules 2015, 16, 3345−3356. (14) Tan, L.; Yang, M. Y.; Wu, H. X.; Tang, Z. W.; Xiao, J. Y.; Liu, C. J.; Zhuo, R. X. Glucose- and pH-Responsive Nanogated Ensemble Based on Polymeric Network Capped Mesoporous Silica. ACS Appl. Mater. Interfaces 2015, 7, 6310−6316. (15) Wu, Z.; Zhang, X.; Guo, H.; Li, C.; Yu, D. An Injectable and Glucose-Sensitive Nanogel for Controlled Insulin Release. J. Mater. Chem. 2012, 22, 22788−22796. (16) Ye, T.; Jiang, X.; Xu, W.; Zhou, M.; Hu, Y.; Wu, W. Tailoring The Glucose-Responsive Volume Phase Transition Behaviour of Ag@Poly (Phenylboronic Acid) Hybrid Microgels: from Monotonous Swelling to Monotonous Shrinking upon Adding Glucose at Physiological pH. Polym. Chem. 2014, 5, 2352−2362. (17) Yang, T.; Ji, R.; Deng, X. X.; Du, F. S.; Li, Z. C. GlucoseResponsive Hydrogels Based on Dynamic Covalent Chemistry and Inclusion Complexation. Soft Matter 2014, 10, 2671−2678. (18) Gu, Z.; Dang, T. T.; Ma, M. L.; Tang, B. C.; Cheng, H.; Jiang, S.; Dong, Y. Z.; Zhang, Y. L.; Anderson, D. G. Glucose-Responsive Microgels Integrated with Enzyme Nanocapsules for Closed-Loop Insulin Delivery. ACS Nano 2013, 7, 6758−6766. (19) Ma, R. J.; Yang, H.; Li, Z.; Liu, G.; Sun, X. C.; Liu, X. J.; An, Y. L.; Shi, L. Q. Phenylboronic Acid-Based Complex Micelles with Enhanced Glucose- Responsiveness at Physiological pH by Complexation with Glycopolymer. Biomacromolecules 2012, 13, 3409−3417. (20) Zheng, C.; Guo, Q. Q.; Wu, Z. M.; Sun, L.; Zhang, Z. P.; Li, C. X.; Zhang, X. G. Amphiphilic Glycopolymer Nanoparticles as Vehicles for Nasal Delivery of Peptides and Proteins. Eur. J. Pharm. Sci. 2013, 49, 474−482. (21) Ding, Z. B.; Guan, Y.; Zhang, Y. J.; Zhu, X. X. Layer-by-Layer Multilayer Films Linked with Reversible Boronate Esterbonds with Glucose-Sensitivity under Physiological Conditions. Soft Matter 2009, 5, 2302−2309. (22) Sato, K.; Yoshida, K.; Takahashi, S.; Anzai, J. pH- and SugarSensitive Layer-by-Layer Films and Microcapsules for Drug Delivery. Adv. Drug Delivery Rev. 2011, 63, 809−821. (23) Cuomo, F.; Lopez, F.; Ceglie, A. Templated Globules− Applications and Perspectives. Adv. Colloid Interface Sci. 2014, 205, 124−133.

(24) De Geest, B. G.; Jonas, A. M.; Demeester, J.; De Smedt, S. C. Glucose-Responsive Polyelectrolyte Capsules. Langmuir 2006, 22, 5070−5074. (25) Levy, T.; Déjugnat, C.; Sukhorukov, G. B. Polymer Microcapsules with Carbohydrate-Sensitive Properties. Adv. Funct. Mater. 2008, 18, 1586−1594. (26) Matsumoto, A.; Ikeda, S.; Harada, A.; Kataoka, K. GlucoseResponsive Polymer Bearing A Novel Phenylborate Derivative as A Glucose-Sensing Moiety Operating at Physiological pH Conditions. Biomacromolecules 2003, 4, 1410−1416. (27) Shiino, D.; Murata, Y.; Kataoka, K.; Koyama, Y.; Yokoyama, M.; Okano, T.; Sakurai, Y. Preparation and Characterization of A GlucoseResponsive Insulin-Releasing Polymer Device. Biomaterials 1994, 15, 121−128. (28) Shiino, D.; Murata, Y.; Kubo, A.; Kim, Y. J.; Kataoka, K.; Koyama, Y.; Okano, T.; et al. Amine Containing Phenylboronic Acid Gel for Glucose-Responsive Insulin Release under Physiological pH. J. Controlled Release 1995, 37, 269−276. (29) Yang, H.; Sun, X.; Liu, G.; Ma, R.; Li, Z.; An, Y.; Shi, L. GlucoseResponsive Complex Micelles for Self-Regulated Release of Insulin under Physiological Conditions. Soft Matter 2013, 9, 8589−8599. (30) Wang, B.; Ma, R.; Liu, G.; Li, Y.; Liu, X.; An, Y.; Shi, L. GlucoseResponsive Micelles from Self-Assembly of Poly(Ethylene Glycol)-bPoly(Acrylic Acid- co-Acrylamidophenylboronic Acid) and the Controlled Release of Insulin. Langmuir 2009, 25, 12522−12528. (31) Zhang, X.; Guan, Y.; Zhang, Y. Dynamically bonded layer-bylayer films for self-regulated insulin release. J. Mater. Chem. 2012, 22, 16299−16305. (32) Seth, A.; Heo, M. B.; Lim, Y. T. Poly(γ-Glutamic Acid) Based Combination of Water-Insoluble Paclitaxel and TLR7 Agonist for Chemo-Immunotherapy. Biomaterials 2014, 35, 7992−8001. (33) Yang, Q.; He, C.; Xu, Y.; Liu, B.; Shao, Z.; Zhu, Z.; Shen, Y. M.; et al. Chitosan Oligosaccharide Copolymer Micelles with Double Disulphide Linkage in The Backbone Associated by H-Bonding Duplexes for Targeted Intracellular Drug Delivery. Polym. Chem. 2015, 6, 1454−1464. (34) Park, I. K.; Yang, J.; Jeong, H. J.; Bom, H. S.; Harada, I.; Akaike, T.; Cho, C. S. Galactosylated Chitosan as A Synthetic Extracellular Matrix for Hepatocytes Attachment. Biomaterials 2003, 24, 2331− 2337. (35) Zhou, N.; Zan, X.; Wang, Z.; Wu, H.; Yin, D.; Liao, C.; Wan, Y. Galactosylated Chitosan−Polycaprolactone Nanoparticles for Hepatocyte-Targeted Delivery of Curcumin. Carbohydr. Polym. 2013, 94, 420− 429. (36) Stö ber, W.; Fink, A.; Bohn, E. Controlled Growth of Monodisperse Silica Spheres in The Micron Size Range. J. Colloid Interface Sci. 1968, 26, 62−69. (37) Yoshida, K.; Hasebe, Y.; Takahashi, S.; Sato, K.; Anzai, J. I. Layerby-Layer Deposited Nano- and Micro-Assemblies for Insulin Delivery: A Review. Mater. Sci. Eng., C 2014, 34, 384−392. (38) Ran, M. S.; Shi, D. J.; Dong, H. X.; Chen, M. Q.; Zhao, Z. L. Glucose Responsive Bio-based Polyelectrolyte Capsules by Layer-byLayer Assembly: Synthesis and Properties. Acta Chim. Sin. 2015, 73, 1047−1054. (39) Ma, Z.; Yeoh, H. H.; Lim, L. Y. Formulation pH Modulates the Interaction of Insulin with Chitosan Nanoparticles. J. Pharm. Sci. 2002, 91, 1396−1404. (40) Tai, W.; Mo, R.; Di, J.; Subramanian, V.; Gu, X.; Buse, J. B.; Gu, Z. Bio-Inspired Synthetic Nanovesicles for Glucose-Responsive Release of Insulin. Biomacromolecules 2014, 15, 3495−3502. (41) Sun, L.; Zhang, X.; Zheng, C.; Wu, Z. M.; Li, C. X. A pH Gated, Glucose-Sensitive Nanoparticle Based on Worm-Like Mesoporous Silica for Controlled Insulin Release. J. Phys. Chem. B 2013, 117, 3852− 3860.

J

DOI: 10.1021/acsami.6b02121 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX