β-Cyclodextrin-conjugated amino poly(glycerol methacrylate)s for efficient insulin delivery

Abstract

A series of cationic polymers based on β-cyclodextrin (β-CD)-conjugated amino poly(glycerol methacrylate)s (PGOHMAs) was synthesized for potential insulin delivery by forming polyelectrolyte complexes (PECs). Both linear and star-shaped poly(glycidyl methacrylate)s were functionalized with mono(6-(diethylenetriamine)-6-deoxy)-β-CD and diethylenetriamine to obtain CD-DETA-PGOHMAs and DETA-PGOHMAs, respectively. The resulting polymers were characterized by ¹H NMR, FT-IR, elemental analysis, and TGA, and then used to prepare PECs with insulin. The association efficiency and loading capacity of CD-DETA-PGOHMAs were higher than those of DETA-PGOHMAs. The release of insulin depended on the introduction of β-CDs, the backbone architecture of the polymers, as well as the pH. The competitive binding release study indicted that insulin could be released rapidly when 1-aminoadamantane hydrochloride (ADA) was used as a competitive binding guest molecule. The in vitro cytotoxicity study revealed that CD-series polymers have much lower toxicity than the D-series. The CD-DETA-PGOHMA/insulin complexes, with lower cytotoxicity and proper release rate, showed great potential for insulin controlled release.

---

1. Introduction

Proteins and polypeptides are structurally and functionally complex entities, which are widely applied in chemistry and pharmaceutics owing to their specificity and high activity.¹ Insulin is a 51-amino-acid polypeptide widely used for the treatment of diabetes mellitus by injection.² However, parenteral administration brings distress and inconvenience to patients.³ As a consequence, oral insulin delivery system has received extensive attention.^4–7 Insulin is usually encapsulated by functional polymer carriers to form micro or nanoparticles to avoid proteolytic breakdown.⁸

Polyelectrolytes, i.e., polymers with ionic or ionizable groups, have constituted an important research field.⁹ Mixing of polyelectrolytes with opposite charged materials could form polyelectrolyte complexes (PECs), whose properties depend strongly on numerous factors, such as the components' molecular weights, densities of ionic groups, mixing ratio, concentrations upon mixing, the pH value and ionic strength of the surrounding medium, etc.¹⁰ Such polyelectrolytes are regarded as “smart polymers”.¹¹ Moreover, PECs were deeply and widely investigated in biopharmaceutical applications. Alessandro et al. studied the PEC formation of chitosan and heparin through electrostatic interactions.¹² Polyelectrolyte complexes of DNA and linear PEI, as well as other cationic polymers for gene therapy have also been extensively studied.¹³ Especially, PECs, composed of polyelectrolytes and insulin, polypeptides, or proteins have also been explored for the purpose of oral delivery.^14–16

β-Cyclodextrin (β-CD) is cyclic oligosaccharide consisting of seven α-D-glucopyranose units linked by α-1,4-glucosidic bonds. It contains a hydrophilic outer surface and a hydrophobic internal cavity within which can include a number of hydrophobic guests with appropriate sizes by non-covalent interactions.^17–21 CDs and their derivatives have been suggested to undergo degradation in the colon since there are vast microflora, breaking them into monosaccharides to be absorbed in the large intestine.²² Therefore, we envision biocompatible CDs and their derivatives have a bright prospect for drug delivery applications, particularly in oral administration of peptide drugs. Significantly, CD–insulin complex could stabilize insulin against aggregation, thermal denaturation and degradation, and enhance the absorption of insulin across the biological barriers by perturbing the membrane fluidity to lower the barrier function.²³ Nonetheless, the solubility of nature β-CD in water (1.8 g/100 mL at 25 °C) is usually insufficient for the stabilization of drugs at therapeutic doses.²⁴ In addition, insulin is too bulky to be wholly included into one CD's cavity, thus their hydrophobic side chains can only partially penetrate into the multiple CDs' cavities to form non-covalent inclusion complexes.²⁵ Therefore, various β-CD derivatives are expected to be developed to conquer the shortcomings.

Previously, we have successfully prepared PECs composed of amino poly(glycerol methacrylate)s (amino PGOHMAs) and insulin.²⁶ The amino PGOHMAs/insulin PECs showed great promise as carriers for insulin and potentially other therapeutic polypeptides. However, the cytotoxicity of PECs was relatively high due to the high density of amino groups in amino PGOHMAs. The association efficiency (AE) of the complexes needs to be improved as well. In addition, amino modified mono-β-CD and bis-β-CDs have been proved to interact with bovine serum albumin (BSA).²⁷ Conjugation of mono(6-(2-aminoethyl)amino-6-deoxy)-β-CD or bis-β-CDs onto polylactides could significantly improve their protein loading capacity (LC).^28,29

Herein, we designed and synthesized a series of β-CD-conjugated PGOHMAs through diethylenetriamine (DETA) linker to obtain CD-DETA-PGOHMAs. More than one CD could be introduced onto the branching backbone of poly(glycidyl methacrylate) (PGMA), providing multi-active sites to include guest molecules (Fig. 1). Consequently, CD-DETA-PGOHMA was used to form complexes with insulin utilizing host–guest inclusion complexation and electrostatic interaction between polymer and insulin. The AE, LC, and in vitro release as well as the cytotoxicity of CD-DETA-PGOHMA, as insulin carrier, were investigated by comparing with their counterpart, DETA-PGOHMA. It is expected to obtain peptide carriers with high LC and low cell toxicity bearing pH- and competitive binding activation.

Fig. 1 Schematic illustration of CD-DETA-PGOHMA/insulin complex formation.

2. Experimental section

2.1. Materials

Glycidyl methacrylate (GMA), 2-bromoisobutyryl bromide, bipyridyl and CuBr were purchased from Shanghai Adamas Reagent Co. Ltd (shanghai, China). Insulin was obtained from Xuzhou Wanbangjinqiao Co. Ltd (Xuzhou, China). Other reagents were purchased from Tianjin Chemical Reagent Co. Ltd (Tianjin, China). THF was dried over sodium using benzophenone as a dryness indicator. β-CD was recrystallized twice before use. 5-arm atom transfer radical polymerization (ATRP) initiator: 1,2,3,4,6-penta-O-isobutyrylbromide-α-D-glucose was synthesized using a synthetic route developed in our previous work.³⁰ 21-arm ATRP initiator: heptakis[2,3,6-tri-O-(2-bromo-2-methylpropionyl)]-β-CD (21 Br-β-CD) was synthesized as reported.^31,32 Mono(6-(diethylenetriamine)-6-deoxy)-β-CD (DETA-CD) was synthesized according to previous reports.^33,34

2.2. Synthesis and characterization of the polymers

2.2.1. Synthesis of linear and star-shaped PGMA. Linear-PGMA (L-PGMA) and star-shaped PGMA (S5-PGMA and S21-PGMA) were synthesized by ATRP.^35–37 A three-necked flask was charged with THF ([monomer] 0.3 M) and GMA (105 equiv.). Bipyridyl (1.5 equiv.), ATRP initiator (2-bromoisobutyryl bromide, or 5- or 21-arm initiator) (1 equiv.) and CuBr (1 equiv.) were successively added into the reaction solution. The mixture was degassed under argon for 15 min at room temperature and then heated at 90 °C for 24 h. After cooling down, the mixture was passed through a silica gel column with THF as an eluent to remove copper. After evaporation of the solvent, the product was dissolved in a small volume of THF and precipitated twice in diethyl ether (Scheme 1).

Scheme 1 Synthesis of PGMA (A), DETA-PGOHMA (B) and CD-DETA-PGOHMA (C).

2.2.2. Synthesis of DETA-PGOHMA. The amino PGOHMAs were synthesized according to a previous report.^26,38 Briefly, PGMA (0.3 g) and excess diethylenetriamine were dissolved in acetonitrile (30 mL). The mixture was heated at 90 °C for 12 h. The product was purified by dialysis (cut-off molecular weight 7000 Da, Tianjin Unite Stars Biotech Co. Ltd) against distilled water for two days and lyophilized. The obtained amino PGOHMAs were named as L-D, S5-D and S21-D, respectively (L, S5, and S21 represent linear, 5-arm and 21-arm star-shaped polymer, respectively).

2.2.3. Synthesis of CD-DETA-PGOHMA. PGMA (0.1 g) and DETA-CD (1.5 g) were dissolved in DMF (30 mL). The mixture was heated at 70 °C for 12 h. Then, diethylenetriamine (10 mL) was added to the flask and reacted for another 12 h in order to complete ring-opening reaction of epoxide groups. The resulting crude CD-DETA-PGOHMA was purified by dialysis, followed by lyophilization as described above. Linear, S5- and S21-arm DETA-CD-modified PGOHMAs were abbreviated as L-CD, S5-CD and S21-CD (Scheme 1).

2.2.4. Characterization of DETA-PGOHMA and CD-DETA-PGOHMA. All ¹H NMR spectra were recorded on a Bruker AV-400 spectrometer (400 MHz, Bruker, Freemont, CA). PGMA was dissolved in deuterated chloroform while DETA-PGOHMA and CD-DETA-PGOHMA were dissolved in deuterated water. FT-IR spectra were recorded on a Bio-Rod 6000 spectrometer (Thermo Electron, USA) to identify the chemical functional groups of the samples. The substance was finely grounded and dispersed into KBr pellets using a ratio of approximately 1 mg sample/200 mg KBr. Molecular weight and polydispersity of PGMAs were detected by Gel permeation chromatography (GPC, R1-201H, SHOKO SCIENTIFIC CO. LTD, Japan). Adequate molecular weight separation was achieved using THF at a flow rate of 0.5 mL min⁻¹ at 20 °C. An elemental analysis instrument (elementar vario EL, GER) was used to determine the content of nitrogen in amino PGOHMA and CD-DETA-PGOHMA to evaluate the amino conversion. β-CD content in CD-DETA-PGOHMA was determined using the phenol–sulphuric acid assay.³⁹ Thermo-gravimetric analysis (TGA) experiments were conducted on a NETSZCH instrument (TG 209 F3 NETSZCH, Germany). The samples were heated from 40 °C to 500 °C at a heating rate of 10 °C min⁻¹ under argon protection.

2.3. Preparation of PECs

Insulin solution (1 mg mL⁻¹) was prepared by dissolving insulin powder in 87% (v/v) 0.01 M HCl. Subsequently, 13% (v/v) of 0.1 M Tris(hydroxymethyl)metylaminomethane solution was added. Polymer (10 mg) was dissolved in 0.25% (v/v) acetic acid solution (10 mL) under magnetic stirring, followed by adjusting the pH of solution to 5.5 with 1 M NaOH. Nanocomplexes were prepared by adding polymer solution (1 mL) to an equal volume of insulin solution in a glass vial under gentle magnetic stirring at room temperature, followed by stirring for another 10 min. The nanoparticles were separated by refrigerated ultracentrifugation at 45 000 × g at 4 °C for 30 min, washed with distilled water and lyophilized. The size and zeta potential of the complexes were evaluated with a Zetasizer Nano ZS90 (Malvern Instruments, Southborough, MA). Scattering light was detected at 90° angle through a 50 μm pin hole at 25 °C.

2.3.1. The effect of the ionic strength on the properties of the complexes. The polymer solution (1 mg mL⁻¹, pH 5.5) was mixed with insulin solution (1 mg mL⁻¹, pH 7.4) at NaCl concentrations of 0, 0.01, 0.02, 0.03, 0.04 and 0.06 mol L⁻¹, respectively, under mild magnetic stirring.

2.3.2. The effect of the pH values of insulin solutions on the properties of the complexes. Insulin solutions (1 mg mL⁻¹) of different pH values (pH = 7.0, 8.0 and 9.0) were added to equal volume (1 mL) of L-CD and L-D solutions (1 mg mL⁻¹, pH 5.5) under gentle magnetic stirring.

2.3.3. The effect of the polymer concentrations on the properties of the complexes. Insulin solution (1 mg mL⁻¹, pH 7.4) was added to equal volume (1 mL) of polymers with various concentrations (0.5, 1.0, 1.5 and 3.0 mg mL⁻¹) and different pH values (pH = 7.0, 8.0 and 9.0) respectively.

2.4. Investigations on the interactions of polymers and insulin

2.4.1. Fourier-transform infrared (FT-IR). FT-IR of insulin, L-CD, physical mixture of insulin and L-CD, and insulin–L-CD complex (ins–L-CD) were recorded on a Bio-Rod 6000 spectrometer (Thermo Electron, USA).

2.4.2. Differential scanning calorimetry (DSC). Insulin, L-CD, L-D, physical mixture of insulin and L-CD or L-D, as well as ins–L-CD, ins–L-D complexes were recorded on the NETZSCH DSC instrument (NETZSCH, F3 209, Germany). Nitrogen was used as the purging gas. Specimens of 3–5 mg sample were loaded into aluminum pans and then heated at a heating rate of 10 °C min⁻¹ to 160 °C, cooled to 20 °C, kept at 20 °C for 3 min, and heated again at a heating rate of 10 °C min⁻¹ to 160 °C. The DSC thermograms of samples were recorded during the second heating run.

Solution DSC was performed as follows: 4 mg of insulin, L-CD and L-D was placed into the crucible, respectively, and 35 μL H₂O was added. For the mixed solution DSC, 4 mg of insulin/L-CD, or insulin/L-D (w/w, 1 : 1) was loaded in the pan, and 35 μL H₂O was added as well. All the solution DSC samples were heated at a heating rate of 10 °C min⁻¹ to 80 °C, cooled to 10 °C, kept at 10 °C for 3 min, and heated again at a heating rate of 10 °C min⁻¹ to 80 °C. The solution DSC thermograms of samples were recorded during the second heating run.

2.5. Characterization of PECs

2.5.1. AE and LC of insulin. The nanoparticles were prepared by mixing equal volume of DETA-PGOHMA (or CD-DETA-PGOHMA) (1 mg mL⁻¹, pH 5.5) and insulin (1 mg mL⁻¹, pH 7.8). The nanoparticles were then refrigerated ultracentrifugated. The amount of insulin in the supernatant was determined by UV absorption band at 277 nm employing a calibration curve. The LC was related with the weight of dry nanoparticles after lyophilization. The AE and LC were calculated from the following equations:

2.5.2. Scanning electron microscope (SEM) experiments. SEM measurements of CD-series complexes and D-series complexes were carried on JSM-6700F Field Emission Scanning Electron Microscope (JEOL, Japan). All the SEM samples were prepared by dropping suspensions (0.1 mg mL⁻¹, 100 μL) of the two series nanoparticles on coverslip. The samples were dried and coated with a thin gold layer after evaporation of water.

2.5.3. In vitro insulin release study. Nanoparticles (ins–L-CD, ins–S5-CD, ins–S21-CD, ins–L-D, ins–S5-D and ins–S21-D) were dispersed in phosphate buffered saline (PBS, pH 7.4, 0.136 M NaCl, 0.01 M PO₄³⁻) at concentration of 2 mg mL⁻¹, and added in a dialysis bag (cut-off molecular weight 7000 Da), which was then immersed into release medium (PBS) in a beaker. The beaker was incubated in an orbital shaker water bath (SHZ-82, Xinghang Instrument Factory of Jintan City, China) at 37 °C and 25 °C, respectively. For competitive binding release study, 100 μL of competitive binding agent 1-aminoadamantane hydrochloride (ADA) solution was added into the dialysis bag at designed time. At designed intervals, release medium (3 mL) was withdrawn and replaced with equal volume of fresh PBS. The amount of released insulin was evaluated by measuring the absorbance of release medium at 277 nm by UV-vis spectroscopy. The cumulative drug release was calculated from the following equation:

Cumulative release (%) = M_t/M_∞ × 100%

2.5.4. Circular dichroism spectra of insulin. The circular dichroism (CD) spectra of native insulin and the released insulin from ins–L-CD and ins–L-D complexes in PBS buffer were obtained by the circular dichroism spectropolarimeter (Jasco-715, Welltech Enterprises, Inc. Maryland, United States) at room temperature (20 °C). Each sample solution was scanned in the range of 200–300 nm. All the CD data was converted in molar ellipticity unit and then fitted by CDPro.

2.6. Cytotoxicity test

The in vitro cytotoxicity of the polymers was determined by the evaluation of the viability of fibroblasts L929 cells using Cell Counting Kit-8 (CCK-8) assay. CCK-8 interacts with mitochondria in cells to generate water-soluble orange formazan product. The number of living cells is in direct proportion to the amount of formazan product. L929 cells were suspended in complete medium and seeded in 96-well tissue culture plates at a concentration of 5 × 10³/100 μL per well, and cultivated in a humidified 5% CO₂ atmosphere at 37 °C for 24 h. The confluency reached 80% before treatment. Then 11 μL of various concentrations (0.2, 0.4, 0.6, 0.8, 1.0 mg mL⁻¹) of polymers was added into the wells. The final concentration of each well was 20, 40, 60, 80 and 100 μg mL⁻¹. After another 24 h of incubation, growth medium was replaced with 100 μL of fresh complete culture media. Then 10 μL of CCK-8 was added to each well. The plates were cultured in CO₂ incubator for 2 h. The optical density (OD) was measured at 450 nm using an automatic BIO-TEK (Inc EL311S, America) microplate reader, and the cell viability was calculated from following equation:

2.7. Statistical analysis

Significant differences between CD-DETA-PGOHMA and the corresponding DETA-PGOHMA were evaluated by a Student's t-test.

3. Results and discussion

3.1. Characterization of polymers

L-PGMA, S5-PGMA, and S21-PGMA were synthesized by ATRP with Mn of 9.5, 10.1, 10.6 kDa (Table S1†). ¹H NMR spectra of the obtained polymers are shown in Fig. 2. The peaks at 2.64, 2.84 and 3.26 ppm were characteristic of the epoxy groups on the side chain of S5-PGMA, and the peaks at 3.82 and 4.33 ppm were assigned to the methylene protons adjacent to the epoxy group due to the space isomerization of the epoxy group. After ring-opening reaction of epoxy groups by diethylenetriamine of PGMA, the signal for the epoxy groups completely disappeared and the peak of the methylene protons adjacent to the epoxy groups shifted to 3.84–4.05 ppm for L-D and 3.42–3.64 ppm for L-CD. The peaks corresponding to β-CD appear at 3.68–4.05 and 5.12 ppm (Fig. 2(c)). The former was corresponding to H₂–H₆ of β-CD while the latter was H₁. Compared with L-D, the proton signal of methylene from diethylenetriamine of L-CD decreased, due to the introduction of β-CD. The successful post-modification of PGMA was also evidenced by FT-IR spectra (Fig. 3). The characteristic absorption peaks of the epoxide ring at 849 and 908 cm⁻¹ disappeared after the ring–opening reaction and a broad band at 3200–3500 cm⁻¹ (b) and 3411 cm⁻¹ (c), assigned to N–H & O–H stretching vibration, appeared. A characteristic strong peak around 1039 cm⁻¹ aroused from C–O–C stretching vibration, 945 cm⁻¹ from framework vibration of α-1,4 bond, as well as 855 & 757 cm⁻¹ from glucose ring breathing vibration of β-CD units, evidencing that β-CD was conjugated onto the pending amino groups of the amino PGOHMAs.

Fig. 2 ¹H NMR spectra of S5-PGMA (a) in CDCl₃, L-D (b) and L-CD (c) in D₂O.

Fig. 3 FT-IR spectra of (a) S5-PGMA, (b) L-D and (c) L-CD.

The amination conversion was calculated by the content of nitrogen measured by elemental analysis (Table 1). The amination conversion of L-D is 74.7%, similar to our previous report.¹⁵ Compared with L-D, the amination conversion of S5- and S21-D was a bit lower because of steric-hindrance effect caused by the branching architecture of star-shaped polymer. Analysis of β-CD content in CD-DETA-PGOHMA revealed the conjugate ratio of β-CDs to the side chain of PGMA was 10.8–15.7%, indicating that one of six to ten repeating units in CD-DETA-PGOHMAs.

Table 1 Characterization of DETA-PGOHMAs and CD-DETA-PGOHMAs^a

Polymer	N^b (%)	AC^c (%)	CD^d (wt%)	CD^e (mol%)
a N (%): elemental N, AC (%): amination conversion, CD (wt%): β-CD content, CD (mol%); conjugation ratio.b Determined by elemental analysis.c Calculated by the elemental N (%).d Determined by the phenol–sulphuric acid assay.e Calculated using phenol–sulphuric acid assay.
L-D	12.17	74.7	—	—
S5-D	11.80	72.6	—	—
S21-D	11.02	66.7	—	—
L-CD	4.51	—	16.58	10.8
S5-CD	4.48	—	17.92	13.6
21-CD	6.16	—	21.93	15.7

---

TGA was performed to evaluate the temperature–weight loss profile. All the TGA curves are exhibited in Fig. S1.† The thermal stability could be enhanced by the introduction of β-CD to the polymers. The decomposition temperature of CD-DETA-PGOHMAs was obviously higher than the corresponding DETA-PGOHMAs.

3.2. The effect of complexes forming conditions on their physicochemical properties

The size of complexes particles has been optimized to be about 200 nm by adjusting the ionic strength and the pH value in surrounding medium and the ratio of polymer/insulin (see ESI Fig. S2–S4†).

3.3. Morphology study

It was observed from SEM images (Fig. 4) that the particles were spherical in shape. No aggregation or adhesion occurred among the particles. Moreover, the particle size demonstrated in the images, was in good agreement with the result measured by DLS.

Fig. 4 SEM images of ins–S5-CD complexes (a) and ins–L-D complexes (b).

3.4. Interactions between polymers and insulin

The FT-IR spectra of insulin, L-CD, their physical mixture, and the complex are shown in Fig. S5.† The characteristic absorption peaks of insulin appear at 1656 and 1527 cm⁻¹. The spectrum of the physical mixture shows all the typical absorption peaks of insulin and L-CD, in which the characteristic absorption peaks of insulin are still present at 1656 and 1525 cm⁻¹, while that of L-CD presents at 1723 and 1155 cm⁻¹. In contrast, in the spectrum of their complex, the two characteristic absorption peaks (1656 & 1525 cm⁻¹) of insulin almost overlap with those of L-CD. The other significant difference is that the –OH characteristic absorption peak shifts from 3358 cm⁻¹ in the physical mixture to 3296 cm⁻¹ in the complex, indicated the formation of hydrogen bonds. All these observations suggested that the physical interactions between insulin and L-CD existed during the formation of the complex.

DSC is usually used for investigation of physical state of samples. Fig. 5 shows an overview of the DSC thermograms of lyophilized blank polymer and drug-loading powder as well as insulin and polymer–insulin physical mixture. The glass transition temperature (T_g) of L-CD is 75.6 °C while that of L-D is 67.4 °C. The increase in the T_g of L-CD after DETA-CD grafting suggested that the β-CD segment plasticized the adjacent PGOHMA chain. Meanwhile, there is a noticeable regulation of T_g which increases in the following order of insulin < polymer/insulin physical mixture < complex < polymer. Physical interaction existed between polymer and insulin in the complexes, resulting in a higher T_g, compared with the physical mixture of the same composition as the complexes. In addition, the insulin phase and polymer phase probably turned into one phase in the first heating run. Therefore, only one glass transition could be observed for both complexes and physical mixture.

Fig. 5 DSC thermograms of insulin (a), the physical mixture of ins/L-D (b), ins-L-D (c), L-D (d), the physical mixture of ins/L-CD (e), ins-L-CD (f), L-CD (g).

The solution DSC thermograms of insulin, L-CD and their mixture, as well as L-D, mixture of L-D and insulin was shown in Fig. S6 (ESI).† A distinct difference was observed for solution and powder DSC of L-CD/insulin. An endothermic peak appears around 70.5 °C in solution DSC of L-CD/insulin, which wasn't detected in neither powder DSC of L-CD/insulin nor solution DSC of L-D/insulin mixture, evidenced the complex formation between β-CD and insulin.

3.5. AE and LC of polymers

The AE and LC of CD-series were higher than that of D-series. The AE and LC of S21-CD could reach as high as 87.5% and 43.8% respectively, evidencing that the host–guest interaction contributed to enhanced drug loading efficiency of polymers. Additionally, the degree of branching and amino density also had an effect upon AE and LC. For D-series, the LC increased from 31.2% to 42.3% with increased degree of branching. The AE of S5-D was lower than L-D because the amino density of S5-D was lower than that of L-D (Table 2).

Table 2 Insulin AE and LC

Polymer	AE (%)	LC (%)
L-D	61.5	31.2
S5-D	49.4	37.9
S21-D	72.7	42.3
L-CD	73.8	36.8
S5-CD	85.7	42.4
S21-CD	87.5	43.8

---

3.6. Melting temperature in tris buffer

Melting temperature is an important parameter for proteins and peptides. Many different techniques have been used to determine T_m, such as DSC, IR spectroscopy and capillary electrophoresis.^40,41 DLS measurement was employed in this study. As shown in Fig. S7.† the scattering intensity kept constant at temperatures less than 60 °C. At 65 °C, the scattering intensity increase sharply, indicating the presence of denatured aggregates. The T_m determined by DLS is in accord with the previous report.⁴² The T_m of insulin obtained in this study was about 64 °C.

3.7. In vitro release study of insulin

In vitro release study of insulin was performed in PBS (pH 7.4, 0.136 M NaCl, 0.01 M PO₄³⁻). As can be seen in Fig. 6, the cumulative release of insulin from D-series complexes was lower than that of CD-series. The electrostatic attraction is strong between amino PGOHMAs with high charge density and insulin, which made the PECs compact so firm that insulin couldn't be released easily.⁴³ The inclusion interaction between CD-DETA-PGOHMAs and insulin is relatively weaker than electrostatic interaction. Therefore, CD-series exhibited an enhanced release rate. Interestingly, the star-shaped polymers released insulin slower than their linear counterpart, which probably due to the steric effect. The release was also performed at 25 °C, the cumulative release of insulin from polymer/insulin complexes was a little less than that at 37 °C (Fig. S8†). The molecule movement and diffusion may become slower at lower temperature, resulting in the decreased release rate.

Fig. 6 In vitro release of insulin from nanoparticles formed by DETA-PGOHMAs and CD-DETA-PGOHMAs with insulin.

A competitive binding release study was performed against ADA (Fig. S9† and 7). The addition of ADA into the release solution didn't affect the release of insulin from its L-D complexes. While the encapsulated drug released from its L-CD complexes more rapidly compared with the control without ADA in PBS. The binding capacity with β-CD of ADA (association constant, 9.4 × 10³ M⁻¹, 25 °C)⁴⁴ is greater than that of insulin. The insulin captured by CD was dislodged from its hydrophobic cavity. Therefore the insulin released rapidly. As shown in Fig. 7, after ADA was added at designated time, 30 min, 60 min and 75 min, the released insulin was more than that of released in adjacent interval. The release became less with extended time. The residue CD groups became less as time pass by. The above release study showed that insulin could be released promptly when competitive binding agent ADA was added for the strong binding capacity between β-CD and ADA. Over the competitive binding stimulus, the insulin could be precisely and fleetly controlled released.

Fig. 7 In vitro release of insulin from nanoparticles formed by L-CD and L-D with insulin when ADA was added at 30, 60 and 75 min.

Circular dichroism (CD) spectroscopy is a common method to analyze the secondary structure of a protein with high reliability. In the CD spectrum of the native insulin in PBS (pH 7.4), there are two extremum at 209 nm and 222 nm.⁴⁵ The CD spectra of the released insulin from the release test after 6 hours were shown in Fig. 8. The far-UV-CD band at 209 nm primarily ascribes to the α-helix structure, while that at 222 nm was for the β-structure.⁴⁶ The ratio between both bands ([Φ]₂₀₉/[Φ]₂₂₂])could be used to generate a qualitative measure of the overall conformational structure of insulin. In the present study, the [Φ]₂₀₉/[Φ]₂₂₂ ratio for standard insulin and released insulin from ins–L-D and ins–L-CD were 1.21, 1.25 and 1.12, respectively. The secondary structure of the insulin released from nanoparticles kept the same as the original insulin, as analyzed by decoupling the composition of the secondary structures with CDPro (ESI Table S2 and Fig. S10†).

Fig. 8 Circular dichroism spectra of the insulin solutions. Native insulin (a), insulin released after 6 hours from ins–L-D complexes (b), and from ins–S5-CD complexes (c).

3.8. Cytotoxicity test

The cytotoxicity against fibroblast L929 test of the D-series and CD-series was then evaluated by CCK-8 assay. As shown in Fig. 9, the cytotoxicity of CD-series was much lower than that of D-series. The cell viability of CD-series was nearly 100%, even at a high concentration of 0.1 mg mL⁻¹. The cytotoxicity mainly relies on the functional groups of the polymers. The introduction of CD reduced the toxicity of amino PGOHMA by decreasing the density of amino groups.⁴⁷ Meanwhile, the large CD molecules can sterically reduce nonspecific binding affinities,⁴⁸ which resulted in the good biocompatibility of CD-DETA-PGOHMA.

Fig. 9 Cell viability of each polymer at different concentrations against L929 cell line by CCK-8 assay for the incubation time of 24 h. Asterisk (*) indicates significant differences (p < 0.05) between every two groups. Results are shown as mean ± standard deviation (n = 3).

4. Conclusions

In this study, a series of DETA-PGOHMAs and CD-DETA-PGOHMAs with linear/star-shaped backbone were synthesized. The formation of polymer/insulin complexes were influenced by a variety of parameters, such as the ionic strength and the pH value surrounding medium, as well as mass ratio of polymer/insulin. The AE and LC of CD-DETA-PGOHMAs were higher than that of CD-DETA-PGOHMAs. DSC and FT-IR indicated the existence of physical interaction between the insulin and polymer. Meanwhile, the degree of branching also had an effect on AE and LC. The AE and LC of S21-CD can reach as high as 87.5% and 43.8%, respectively. The release of insulin relied on degree of branching, introduction of β-CD, and the pH of the medium as well. Competitive binding release study manifested that insulin could be precisely and fleetly controlled released. The results of cytotoxicity test revealed that the introduction of multi-β-CD into cationic PGOHMA reduced the density of amino groups, thus significantly enhanced the cell viability. These multi-β-CD contained polyelectrolytes show great promise as carriers for insulin and potentially other therapeutic polypeptides.

Acknowledgements

Financial support from NSFC (21074092, 21244004, 21374079), Program for New Century Excellent Talents in University (NCET-11-1063), Program of for Prominent Young and Middle-aged College Teachers of Tianjin Educational Committee is highly acknowledged.

References

1. G. Fuhrmann, A. Grotzky, R. Lukić, S. Matoori, P. Luciani, H. Yu, B. Zhang, P. Walde, A. D. Schlüter, M. A. Gauthier and J.-C. Leroux, Nat. Chem., 2013, 5, 582.
2. C. L. Li and Y. J. Deng, J. Pharm. Pharmacol., 2004, 56, 1101.
3. A. Gholamipour-Shirazi, Expert Opin. Drug Delivery, 2008, 5, 1335.
4. P. Mukhopadhyay, K. Sarkar, S. S. Soam and P. P. Kundu, J. Appl. Polym. Sci., 2013, 129, 835.
5. D. Tahtat, M. Mahlous, S. Benamer, A. N. Khodja, H. Oussedik-Oumehdi and F. Laraba-Djebari, Int. J. Biol. Macromol., 2013, 58, 160.
6. Y. L. Zhang, W. Wei, P. P. Lv, L. Y. Wang and G. H. Ma, Eur. J. Pharm. Biopharm., 2011, 77, 11.
7. P. F. Minimol, W. Paul and C. P. Sharma, Carbohydr. Polym., 2013, 95, 1.
8. J. A. Fix, J. Pharm. Sci., 1996, 85, 1282.
9. F. Carnal and S. Stoll, J. Phys. Chem. B, 2011, 115, 12007.
10. H. Z. Wang, C. Qian and M. Roman, Biomacromolecules, 2011, 12, 3708.
11. M. R. Aguilar, C. Elvira, A. Gallardo, B. Vázquez and J. S. Román, Tissue Eng., 2007, 3, 1.
12. A. F. Martins, J. F. Piai, I. T. A. Schuquel, A. F. Rubira and E. C. Muniz, Colloid Polym. Sci., 2011, 289, 1133.
13. I. Y. Perevyazko, M. Bauer, C. M. Pavlov, S. Hoeppener, S. Schubert, D. Fisher and U. S. Schubert, Langmuir, 2012, 28, 16167.
14. T. V. Burova, N. V. Grinberg, D. R. Tur, V. S. Papkov, A. S. Dubovik, E. D. Shibanova, D. I. Bairamashvili, V. Y. Grinberg and A. R. Khokhlov, Langmuir, 2013, 29, 2273.
15. C. Thompson, W. P. Cheng, P. Gadad, K. Skene, M. Smith, G. Smith, A. McKinnon and R. Knott, Pharm. Res., 2011, 28, 886.
16. S. J. Shu, L. Sun, X. G. Zhang, Z. M. Wu, Z. Wang and C. X. Li, J. Nanopart. Res., 2011, 13, 3657.
17. S. Shang, W. Zeng and X. M. Tao, RSC Adv., 2012, 2, 4657.
18. Y. Liu, Y. W. Yang, L. Li and Y. Chen, Org. Biomol. Chem., 2004, 2, 1542.
19. J. L. Manasco, C. D. Saquing, C. Tang and S. A. Khan, RSC Adv., 2012, 2, 3778.
20. J. Szejtli, Chem. Rev., 1998, 98(5), 1743.
21. M. L. Denadai, D. Ianzer, A. F. C. Alcântara, M. M. Santoro, C. F. F. Santos, I. S. Lula, A. C. M. Camargo, A. Faljoni-Alario, R. A. S. Santos and R. D. Sinisterra, Int. J. Pharm., 2007, 336, 90.
22. Z. Z. Shao, Y. P. Li, T. Chermak and A. K. Mitra, Pharm. Res., 1994, 11, 1174.
23. K. Uekama, Chem. Pharm. Bull., 2004, 52(8), 900.
24. U. Gupta, H. B. Agashe, A. Asthana and N. K. Jain, Biomacromolecules, 2006, 7, 649.
25. T. Irie and K. Uekama, Adv. Drug Delivery Rev., 1999, 36, 101.
26. X. Y. Lu, H. Gao, C. Li, Y. W. Yang, Y. N. Wang, Y. G. Fan, G. L. Wu and J. B. Ma, Int. J. Pharm., 2012, 423, 195.
27. H. Gao, Y. N. Wang, Y. G. Fan and J. B. Ma, Bioorg. Med. Chem., 2006, 14, 131.
28. H. Gao, Y. N. Wang, Y. G. Fan and J. B. Ma, J. Controlled Release, 2005, 107, 158.
29. H. Gao, Y.-W. Yang, Y. G. Fan and J. B. Ma, J. Controlled Release, 2006, 112, 301.
30. H. Gao, M.-C. Jones, P. Tewari, M. Ranger and J.-C. Leroux, J. Polym. Sci., Part A: Polym. Chem., 2007, 45, 2425.
31. J. S. Li and H. N. Xiao, Tetrahedron Lett., 2005, 46, 2227.
32. J. S. Li, Z. Z. Guo, J. Y. Xin, G. L. Zhao and H. N. Xiao, Carbohydr. Polym., 2010, 79, 277.
33. Y. Y. Liu, X. D. Fan and L. Gao, Macromol. Biosci., 2003, 3, 715.
34. Y. Liu, Y. W. Yang, Y. Song, H. Y. Zhang, F. Ding, T. Wada and Y. Inoue, ChemBioChem, 2004, 5, 868.
35. H. Gao, M. Elsabahy, E. V. Giger, D. K. Li, R. E. Prud'homme and J.-C. Leroux, Biomacromolecules, 2010, 11, 889.
36. M.-C. Jones, P. Tewari, C. Blei, K. Hales, D. J. Pochan and J.-C. Leroux, J. Am. Chem. Soc., 2006, 128, 14599.
37. M.-C. Jones, H. Gao and J.-C. Leroux, J. Controlled Release, 2008, 132, 208.
38. H. Gao, X. Y. Lu, Y. N. Ma, Y. W. Yang, J. F. Li, G. L. Wu, Y. N. Wang, Y. G. Fan and J. B. Ma, Soft Matter, 2011, 7, 9239.
39. G. L. Koh and I. G. Tucker, Int. J. Pharm., 1986, 34, 183.
40. K. McIntosh, W. Charman and S. Charman, J. Pharm. Biomed. Anal., 1998, 16, 1097.
41. R. Chehin, M. Thorolfsson and P. Knappskog, FEBS Lett., 1998, 422, 225.
42. A. Elsayed, M. Remawi, N. Qinna, A. Farouk and A. Badwan, Eur. J. Pharm. Biopharm., 2009, 73, 269.
43. N. Zhang, J. H. Li, W. F. Jiang, C. H. Ren, J. S. Li, J. Y. Xin and K. Li, Int. J. Pharm., 2010, 393, 212.
44. H. Goto, Y. Furusho and E. Yashima, J. Am. Chem. Soc., 2007, 129, 109.
45. B. Sarmento, D. C. Ferreira, L. Jorgensen and M. van de Weert, Eur. J. Pharm. Biopharm., 2007, 65, 10.
46. L. Sun, X. G. Zhang, C. Zheng, Z. M. Wu and C. X. Li, J. Phys. Chem. B, 2013, 117, 3852.
47. X. J. Zhang, X. G. Zhang, Z. M. Wu, X. J. Gao, S. J. Shu, Z. Wang and C. X. Li, Carbohydr. Polym., 2011, 84, 1419.
48. S. H. Pun, N. C. Bellocq, A. J. Liu, G. Jensen, T. Machemer, E. Quijano, T. Schluep, S. F. Wen, H. Engler, J. Heidel and M. E. Davis, Bioconjugate Chem., 2004, 15, 831.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c3ra47150k

---