Dual-stimuli-responsive glycopolymer bearing a reductive and photo-cleavable unit at block junction

Abstract

New dual-stimuli-responsive and bio-recognizable glycopolymers (Glyco-ONB-s-s-PXCL) containing a photodegradable 5-hydroxy-2-nitrobenzyl alcohol (ONB), and a redox-cleavable disulfide (-s-s-) as junction point between bio-recognizable hydrophilic sugar molecule (Glyco) and hydrophobic poly(4-substituted-ε-caprolactone) (PXCL) chains were synthesized using a combination of ring-opening polymerization and nucleophilic substitution reactions. When the polymer solution was exposed to ultraviolet (UV) irradiation and/or the reducing agent DL-dithiothreitol (DTT), substantial structural and morphological changes were observed in the particles. The polymers formed micelles in the aqueous solution, had hydrodynamic sizes ≤200 nm, and were spherical. Under the combined stimulation of UV irradiation and DTT, the micellar nanoparticles dissociated and the loaded molecules could be released from the assemblies more efficiently than under the stimulation of only one of the two stimuli. Selective lectin binding experiments confirmed that glucosylated Gluco-ONB-s-s-PXCL can be used for bio-recognition. The nanoparticles exhibited nonsignificant toxicity against HeLa cells at concentrations of ≤300 μg mL⁻¹. The DOX-loaded Gluco-ONB-s-s-PMCL₂₀ micelles effectively inhibited the proliferation of HeLa cells with a half-maximal inhibitory concentration (IC₅₀) of 1.5 μg mL⁻¹.

---

Introduction

Stimuli-responsive amphiphilic polymeric micelles have been extensively studied because they are responsive to environmental changes (e.g., pH, temperature, presence or absence of chemicals and biological compounds) and the applications of external stimuli (e.g., light, magnetic, electric, ionic force).^1,2 A notable advancement is the development of polymeric micelles that can be disrupted by two or more stimuli, which increases the level of control.^3,4 Dual- or multi-stimuli-responsiveness can readily be achieved by incorporating two or more types of stimuli-reactive moieties into the polymer structure, either as pendent groups or in the chain backbone.^5–7 Zhao et al. reported the synthesis of dual- and multi-sensitive comb-like, toothbrush-like grafted copolymers.^8–11 However, making polymeric micelles undergo an efficient and controllable disruption in response to stimuli is challenging. This promising strategy regularly utilizes the position of multiple cleavable linkages on the hydrophobic polymer main chains. The cleavage of the degradable linkages in hydrophobic cores causes the micelles to destabilized or disintegrated due to hydrophobic cores loss.¹² Recently, dual-responsive degradable micelles consisting of both disulfide (s–s) and o-nitrobenzyl (ONB) linkages along hydrophobic main chains have been examined for fast and slow degradation. Disulfide can be cleaved by a reducing agent and o-nitrobenzyl ester through photolysis upon light absorption.^13–18

An additional challenge in this technology is cell-specific targeting, which enables the recognition and binding of the micelles or aggregate host to the target site.^19–22 The cell surface carbohydrates from glycoproteins and glycolipids are vital in recognizing sites between cells and between cells and microorganisms. The recognition mechanisms are based on specific interactions between the saccharide resides and protein receptors, the so-called lectins. These materials are promising carrier systems for drug delivery because of the specific cellular targeting of the drugs by membrane lectins, which often participate in the internalization of their ligands.^23,24 Therefore, synthetic polymers incorporating sugar residues are being developed using various synthetic methods. Yang et al.²⁵ demonstrated that micelles self-assembled from Gal-functionalized amphiphilic polycarbonate block copolymers exhibited efficient cellular uptake and significantly increased antitumor activity in HepG2 cells. Zhong and coworkers²⁶ designed and developed an acid copolymer for efficient hepatoma-targeting delivery of doxorubicin based on poly(ε-caprolactone)-graft-s-s-lactobionic acid.

In this study, bio-recognizable and dual-stimuli-cleavable glycol-ONB-disulfide-poly(4-substituted-ε-caprolactone) (Glyco-ONB-s-s-PXCL) polymers that carry light- and redox-cleavable units in the main chain were synthesized. Light is a particularly useful external trigger because it allows precise temporal control (when the light source is activated) and spatial (where the light is directed) control.^27,28 The ONB photolabile group was selected because of its susceptibility to degradation over a wide range of cytocompatible irradiation conditions.^6,29 The ONB group can be cleaved by UV irradiation at 365 nm, a wavelength that has been proven to be harmless to cells.^30,31 Another notable stimulus factor results from the redox potential in the cytosol and nuclei of cells, which contain 100- to 1000-fold higher concentrations of reducing GSH tripeptide (ca. 1–10 mM) than common fluids found outside cells (ca. 2–20 μM). The concentration level of cytosolic GSH in certain tumor cells has been found to be several times higher than that in normal cells.³² Therefore, disulfide-containing polymeric micelles can facilitate intracellular release of the encapsulated drugs through the cleavage of this bond. Scheme 1 shows the chemical structure of the block polymer and the schematic representation of the dual-stimuli-cleavable of the block polymer micelles. To the best of our knowledge, light and redox dual-stimuli-cleavable sugar-installed polyester has yet to be investigated. The self-assembling and redox-degradable and photo-degradable properties of the polymer as well as its recognition by using glucose-specific bacterial lectin were studied.

Scheme 1 (A) Synthesis reductive and photo-cleavable Glyco-ONB-s-s-PXCL polymers, and (B) dual reductive and photo-cleavable micelles.

Experimental

Materials

4-Methylcyclohexanone, 4-phenylcyclohexanone, ε-CL, 2,2-dithiodipyridine, 2-mercaptoethanol, potassium ethanolate, 5-hydroxy-2-nitrobenzyl alcohol, 3-mercapto propionic acid (MPA), hafnium(IV) chloride tetrahydrofuran complex (HfCl₄·2THF), indomethacin (IMC), nile red (NR), pyrene, and DL-dithiothreitol (DTT) were purchased from the Aldrich Chemical Co. (Milwaukee, WI). Additionally, m-chloroperoxybenzoic acid (m-CPBA), acetyl chloride, acetic acid, boron trifluoride diethyl etherate (BF₃OEt₂), α-D-pentaacetyl-glucopyranoside, β-D-galactose pentaacetate, and D-(+)-maltose were purchased from the Fluka Chemical Co. (Buchs SGI, Switzerland). Doxorubicin hydrochloride (DOX HCl 99%) purchased from Sigma Chemical Co. (Saint Louis, MO) was deprotonated to obtain the hydrophobic DOX in accordance with a published protocol.³³ 4-Methyl-ε-caprolactone (MCL), and 4-phenyl-ε-caprolactone (BCL) were prepared according to the methods described.³⁴ 2-(2-Pyridyldithio)-ethanol was prepared according to a reported procedure.³⁵ N,N-Dimethyl formamide (DMF) and toluene were distilled under calcium hydride. Other solvents such as tetrahydrofuran (THF), dimethylsulfoxide (DMSO), methanol, chloroform, and n-hexane of high-pressure liquid chromatography (HPLC) grade, were purchased from Merck Chemical Co. (Darmstade, Germany). Ultrapure water was obtained by purifying with a Milli-Q Plus system (Waters, Milford, MA). Dulbecco's modified Eagle's medium (DMEM), trypsin/EDTA, 100× antibiotic–antimycotic, and Hochest 33342 nuclei dye were purchased from Gibco, Invitrogen Corp. (Carlsbad, CA). Fetal bovine serum (FBS) was obtained from Biological Industry (Kibbutz Beit Haemek, Israel). A CellTiter 96® AQueuous One Solution kit was obtained from Promega (Fitchburg, Wisconsin).

Characterization

The chemical structures of the polymers were characterized using ¹H NMR (Bruker WB/DMX-500 spectrometer, Ettlingen, Germany) operating at 500 MHz, for which the samples were dissolved in CDCl₃. Fourier transform infrared spectroscopy (FT-IR) spectra were recorded on a Bruker TENSOR 27 FT-IR spectrophotometer (Bruker, Germany) using a method wherein a sample was placed on sodium chloride plates or pressed into potassium bromide pellets. Number- and weight-average molecular weights (M_n and M_w, respectively) and molecular weight distributions (M_w/M_n) were determined through gel permeation chromatography (GPC) using Jordi Gel polydivinyl benzene columns with pore sizes of 100, 500, and 1000 Å, and an HPLC system equipped with a model PU-2031 refractive-index detector (Jasco, Tokyo, Japan). CHCl₃ was used as an eluent at a flow rate of 0.5 mL min⁻¹ at 30 °C. Poly(ethylene glycol) standards of low dispersity (Polymer Sciences) were used to generate a calibration curve.

Synthesis of Py-s-s-PXCL

Py-s-s-PMCL₂₀ was synthesized through a ring-opening reaction. 2-(2-Pyridyldithio)-ethanol (Py-ss-(CH₂)₂OH, 0.12 g, 0.64 mmol), was used as the initiator, MCL (1.65 g, 12.89 mmol) was introduced into a flask, and then dissolved in 50 mL of toluene under a dry nitrogen stream. Then, 26.6 mg of SnOct₂ [1.5 wt% based on the weight of Py-ss-(CH₂)₂OH and MCL] was added to the flask. The flask was purged with a nitrogen stream and refluxed for 24 h, and the solution was then concentrated under reduced pressure. The resulting product (Py-s-s-PMCL₂₀) was dissolved in CHCl₃, which precipitated into excess n-hexane/diethyl ether (v/v, 5 : 1) during stirring. The purified polymer was dried in vacuo at 50 °C for 24 h to obtain the polymer with 76% yield. GPC (CHCl₃) M_n = 3080 g mol⁻¹, polydispersity index (PDI) = 1.18. ¹H NMR δ (ppm): 8.49 (s, ArH), 7.71–7.60 (m, ArH), 7.11 (m, ArH), 4.35 (t, –CH₂OCO–), 4.10 (m, –CH₂O–), 3.06 (t, –S–CH₂–), 2.30 (t, –COCH₂–), 1.67 (m, –CH₂–), 1.58 (m, –CH–), 1.48 (m, –CH₂–), 0.90 (d, –CH₃).

Synthesis of (4-hydroxy-2-nitrobenzyl) 3-mercapto-propionate (HONB-SH)

The thiol-functionalized propionate (HONB-SH) was synthesized from 4-hydroxy-2-nitro-benzyl alcohol (HONB-OH) and 3-mercaptopropionic acid (MPA), as modified the report in the literature.³⁶ In a typical experiment, HONB-OH (0.80 g, 4.72 mmol) and MPA (2.5 g, 23.61 mmol) was dissolved in toluene (30 mL) along with HfCl₄·2THF (0.11 g, 0.24 mmol) were added. The mixture was refluxed at 130 °C under nitrogen for 24 h. Toluene was removed under reduced pressure. The residue was purified on a silica gel column chromatography by elution with EtOAc/n-hexane (v/v 1 : 2) to give white product. Yield: 85%. ¹H NMR δ (ppm): 8.18 (d, ArH), 7.04 (d, ArH), 6.88 (dd, ArH), 5.60 (s, benzyl CH₂–), 2.89–2.80 (m, –CH₂–CH₂–).

Synthesis of HONB-s-s-PXCL

HONB-s-s-PXCL was synthesized using a nucleophilic substitution between the thiol end group HONB-SH and the pyridyldisulfide group at the end of PXCL (Py-s-s-PXCL). A mixture of Py-s-s-PMCL₂₀ (2.28 g, 0.74 mmol, M_n = 3080 g mol⁻¹) and acetic acid (0.5 mL) was stirred in DMF (5 mL) for 30 min at ambient temperature. The HONB-SH (0.23 g, 0.89 mmol, 1.2 equiv.) solution in DMF (3 mL) was added, and the mixture was stirred at ambient temperature for 24 h. After DMF was concentrated under reduced pressure, the resulting product (HONB-s-s-PMCL₂₀) was dissolved in CHCl₃, which precipitated into excess diethyl ether during stirring. The purified polymer was dried in vacuo at 50 °C for 24 h and then analyzed. Fig. 1A and 2A depict the representative ¹H NMR, and IR spectra of HONB-s-s-PMCL₂₀.

Fig. 1 ¹H NMR spectra of (A) HONB-s-s-PMCL₂₀, and (B) Gluco-ONB-s-s-PMCL₂₀.

Fig. 2 IR spectra of (A) HONB-s-s-PMCL₂₀, and (B) Gluco-ONB-s-s-PMCL₂₀.

Synthesis of glyco-ONB-s-s-PXCL polymers

Three types of 3-bromopropyl-sugars (3-bromopropyl D-glucopyranoside, 3-bromopropyl D-maltose, and 3-bromopropyl D-galactose) were prepared according to a published protocol.³⁶ We used a typical procedure for coupling of 3-bromopropyl-sugar with HONB-s-s-PXCL. First, a mixture of HONB-s-s-PMCL₂₀ (0.5 g, 0.17 mmol) and potassium carbonate (71.6 mg, 0.52 mmol) was stirred in DMF (5 mL) for 1 h at 60 °C. 3-Bromopropyl D-glucopyranoside (62.4 mg, 0.21 mmol) in DMF (3 mL) was then added, and the mixture was stirred for 24 h at 60 °C. DMF was removed under reduced pressure. The residue was dissolved in CHCl₃, and precipitated into excess n-hexane/diethyl ether (v/v 5 : 1) while stirring. The purified polymer, Gluco-ONB-s-s-PMCL₂₀ was dried in vacuo at 60 °C for 24 h and analyzed. The yield was 84%. GPC (CHCl₃) M_n = 3800 g mol⁻¹, PDI = 1.36. Fig. 1B and 2B demonstrate representative ¹H NMR and FT-IR spectra of Gluco-ONB-s-s-PMCL₂₀.

Preparation of dye-loaded micelles and characterization

NR-loaded micelles were prepared in the following method. Gluco-ONB-s-s-PMCL₂₀ (2 mg) and NR (0.67 mg) were initially dissolved in THF (1 mL). The solution was then quickly added to pure water (10 mL) under ultrasonic agitation using an ultrasonic cleaner (42 KHz). The entire solution was stirred at room temperature overnight to completely remove THF through evaporation. The obtained aqueous solution was filtered (0.45 μm filter) to remove NR not solubilized by the micelles. The final NR-loaded micellar solution was adjusted to reach a polymer concentration of 0.2 mg mL⁻¹ and stored in the dark before use. Blank micelles (without loaded NR) were prepared using the same procedure.

Fluorescence emission spectra were recorded on a Hitachi F-4500 fluorescence spectrophotometer. DLS measurements were performed with a Malvern Zetasizer Nano ZS dynamic light scattering particle sizer with a helium–neon laser (λ = 632.8 nm). All measurements were performed at a scattering angle of 90°. The morphologies of the polymer micelles were examined using a JEM 1200-EXII transmission electron microscopy (TEM) operating at 100 kV.

In vitro drug loading and release

The IMC-loaded micelles were prepared as follows: a polymer (25 mg) was dissolved in 6 mL of CH₂Cl₂, following which 25 mg of IMC was added to serve as a model drug. Subsequently, 15 mL of Milli-Q water was added dropwise to the solution under vigorous stirring at an ambient temperature overnight to evaporate the CH₂Cl₂. The unloaded IMC residue was removed using a Teflon filter (Whatman); average pore size 0.45 μm. The micelles were obtained through vacuum drying. A weighted among of micelles was then disrupted by adding a 10-fold excess volume of DMF. Drug content was assayed spectrophotometrically at 320 nm using a diode array UV-Vis spectrophotometer. The following equations were used to calculate the drug-loading content and drug entrapment efficiency:

Drug-loading content (%) = (weight of drug loaded in micelles/weight of micelles) × 100

Drug entrapment efficiency (%) = (weight of drug loaded in micelles/weight of drug in feed) × 100

The experiments for the in vitro release of IMC were performed at 37 °C in PBS (pH 7.4, in the presence or absence of 10 mM DTT and UV irradiation) with constant shaking. Fifty milligrams of IMC-loaded micelles were suspended in 10 mL of PBS (0.1 M, pH 7.4) in the presence or absence of 10 mM DTT, and then placed in a dialysis bag (MWCO 3500), which in turn was immersed in 20 mL of buffer solution incubated at 37 °C with continuous shaking or exposure to UV light. At predetermined intervals, 3 mL aliquots of the aqueous solution were withdrawn from the release medium, and an identical volume of fresh buffer solution was added. The rate of controlled drug release was measured according to the cumulatively released weight of IMC by using the calibration curve for IMC.

Carbohydrate–lectin binding recognition

The lectin recognition activity of the glycopolymer solutions was analyzed by applying a change in turbidity at 600 nm at room temperature. A 2 mg mL⁻¹ sample of Concanavalin A (Con A) lectin was first prepared in 0.01 M PBS at pH 7.4. Then, 600 μL of the lectin solution was transferred into a cuvette, and a baseline measurement was obtained. A solution of 60 μL glycopolymer with different concentrations (0.2–2.0 mg mL⁻¹) in PBS was added to the cuvette containing the lectin solution. The solution in the cuvette was gently mixed using a pipette, and the absorbance at 600 nm was recorded every 200 s.

In vitro cytotoxicity tests

The cytotoxicity of free DOX, blank micelles, and DOX-loaded micelles were investigated using an MTS assay with a Promega CellTiter 96® AQueuous One Solution kit. The assay was performed according to the manufacturer's instructions with minor modifications. HeLa cells were seeded in a 24-well plate (3 × 10⁴ cells per well) overnight and treated with various concentrations of free DOX, polymers, and freeze-dried DOX-loaded micelles or a DMSO vehicles added to DMEM/F12 1 : 1 medium with 1% FBS in a humidified 37 °C incubator supplied with 5% CO₂. After 48 h, the medium in each well was removed and replaced with 350 μL of warm PBS and 35 μL of CellTiter 96® AQueuous One Solution. The mixture was incubated at 37 °C for 4 h. After incubation, 110 μL of supernatant from each well was moved to a 96-well plate and absorbance was measured at 485 nm by using an ELISA reader (Hidex, Finland).

Results and discussion

Synthesis and characterization of Glyco-ONB-s-s-PXCL

Scheme 1 presents a strategy for synthesizing a dual light- and redox-cleavable Glyco-ONB-s-s-PXCL polymer containing a photo-cleavable o-nitrobenzyl (ONB) moiety and a redox-cleavable disulfide linkage at the junction between the hydrophilic sugar and the hydrophobic block. First, Py-s-s-PXCL was prepared through ROP of ε-XCL (X = H, CH₃, Ph), initiated by 2-(2-pyridyldithio)-ethanol and catalyzed by SnOct₂ (1.5 wt%) to yield the Py-s-s-PXCL pre-polymer. Different compositions of Py-s-s-PXCLs were prepared by changing the feed ratio of the monomers ε-XCL to the initiator. Then, the coupling reaction of HONB-SH and Py-s-s-PXCL was carefully performed at room temperature under an N₂ atmosphere to minimize air oxidation and homodimerization. The ¹H NMR of HONB-s-s-PMCL₂₀ in Fig. 1A exhibits resonances at δ 8.18, 7.05, and 6.89 ppm, which assigned to protons of the aromatic ONB group. The resonance peaks of the PMCL block were at δ 4.12, 2.31, 1.78–1.39, and 0.90 ppm. The pyridyl group resonances at δ 8.50, 7.72–7.61, and 7.13 completely disappeared, which indicated the successful conversion of Py-s-s-PMCL to HONB-s-s-PMCL. Finally, nucleophilic substitution etherified the phenolic hydroxyl group of the resulting HONB-s-s-PXCL with Glyco-O(CH₂)₃Br in DMF at 60 °C to yield a Glyco-ONB-s-s-PXCL polymer.³⁷ Various polymers with different compositions were prepared by altering glycose (e.g. glucose, maltose, and galactose) and the length of PXCL segments. Table 1 lists the coupling results. The isolated yields were moderate to high. The theoretical number-average molecular weight (M_n,th), number-average molecular weight determined by ¹H NMR (M_n,NMR), and number-average molecular weight determined by GPC (M_n,GPC) showed good agreement, and the polydispersity index (M_w/M_n) ranged from 1.31 to 1.50. ¹H NMR and FT-IR results confirmed the effective coupling of Glyco-O(CH₂)₃Br with HONB-s-s-PXCL to yield Glyco-ONB-s-s-PXCL (Fig. 1B and 2B). The representative ¹H NMR spectrum revealed the presence of signals from the Glyco- and -ONB-PXCL segments. The -ONB-PMCL₂₀ blocks demonstrated resonance peaks at δ = 8.40 (H_c, aromatic –CH), 7.18 (H_a, aromatic –CH), 6.98 (H_b, aromatic –CH), 5.82 (H_d, benzylic –CH₂O–), 4.12 (H_m, –CH₂O–), 2.32 (H_i, –CH₂–), 1.85–1.38 (H_j+k+l, –CH₂–) and 0.95 (H_n, –CH₃) ppm. Glucose resonance peaks presented at δ = 6.12 (H_o, methine CH), 3.91–3.49 (H_p, –CH–) ppm. However, because of the low ratio of glycose end-group and ONB to the main-chain PMCL repeat units, the glycose and ONB proton resonance were notably weak. The IR spectrum of Gluco-ONB-s-s-PMCL₂₀ indicated a typical O–H absorptions of glucose at 3500 cm⁻¹, C=O absorptions of an ester at 1720 cm⁻¹, NO₂ absorption at 1560 and 1480 cm⁻¹, and C–O absorption at 1150 cm⁻¹ (Fig. 2B). Compared with the IR spectrum of HONB-s-s-PMCL₂₀, strong O–H and C–O absorptions observed in the spectrum of Gluco-ONB-s-s-PMCL₂₀, confirmed the effective coupling of Gluco-O(CH₂)₃Br with HONB-s-s-PXCL to yield Gluco-ONB-s-s-PXCL. Fig. 3 shows the GPC curves of Gluco-ONB-s-s-PMCL₂₀. The GPC curves of Gluco-ONB-s-s-PMCL₂₀ revealed unimodal distribution.

Table 1 Result of the coupling of HONB-ONB-s-s-PXCL with Glyco(CH₂)₃Br

Polymer^a	Isolated yield (%)	Mn,th^b	Mn,NMR^c	Mn,GPC^d	Mw/Mn^d (PDI)
a Abbreviations: Gluco = glucose; Malto = maltose; Galacto = galactose; ONB = 4-hydroxy-2-nitrobenzyl alcohol; PCL = poly(ε-caprolactone); PMCL = poly(4-methyl-ε-caprolactone); PBCL = poly(4-phenyl-ε-caprolactone).b Mn,th = Mn,Glyco(CH2)3 + Mn,HONB-s-s-PXCLn.c Determined by ^1H NMR.d Determined by GPC.
Gluco-ONB-s-s-PCL20	85	2830	2770	2350	1.31
Gluco-ONB-s-s-PMCL20	84	3110	3560	3800	1.36
Gluco-ONB-s-s-PBCL18	75	3970	3650	3250	1.40
Gluco-ONB-s-s-PMCL34	78	4902	4980	5370	1.50
Malto-ONB-s-s-PCL20	73	2993	2820	2500	1.30
Galacto-ONB-s-s-PCL20	82	3110	3220	3360	1.34

---

Fig. 3 GPC curve of Gluco-ONB-s-s-PMCL₂₀.

Photo- and redox-cleavage of Glyco-ONB-s-s-PXCL

UV-Vis spectroscopy was used to investigate the photo- and reductive-triggered degradation of the polymer by observing the spectra changes of a polymer solution treatment for 5 h under UV irradiation, with the reducing agent DTT, or using a combination of the two stimuli conditions. The obtained Glyco-ONB-s-s-PXCL polymer was expected to undergo two controlled block junction cleavages under the effect of UV light and a reducing agent. Accordingly, 1.0 mg Gluco-ONB-s-s-PMCL₂₀ was dissolved in 1 mL of inhibitor-free THF and the solution was irradiated with a UV lamp (352 nm, 64 W, 58.8 mW cm⁻²) or treated with 10 mM DTT (intracellular-mimicking reducing environment) for 5 h. The experiment was performed at ambient temperature. Fig. 4A displays the UV-Vis spectra change of Gluco-ONB-s-s-PMCL₂₀ in THF (1 mg mL⁻¹) with treatment under UV irradiation, with the reducing agent DTT, and with a combination of the two stimuli conditions. The absorption spectra of Gluco-ONB-s-s-PMCL₂₀ is presented in curve a. Initially, as the polymer solution was treated with DTT at the same concentration, the intensity of the band located at 300 nm, associated with the nitro-aromatic moiety, decreased significantly (curve b). Similarly, the intensity of the band located at 300 nm significantly decreased when the treatment was exposed to UV light (curve c). When the two stimuli were combined, the intensity of the band located at 300 nm decreased by approximately 50% (curve d). In addition, the intensity at λ_max decreased and a blue shifted from 263 nm to 251 nm when treated with both UV irradiation and the reducing agent DTT. The intensity of absorption decreased faster under UV light exposure than in the presence of the reducing agent because the reduction reaction of the disulfide group was slower than the photolysis of ONB. Additionally, the polymer degradation was observed in the ¹H NMR and FT-IR spectra after treatment with UV and DTT. ¹H NMR indicated significant changes in chemical shifts (Fig. 4B). Regarding UV degradation, the resonance peaks of the ONB at δ 8.40, 7.18, and 6.98 ppm disappeared, indicating the photolysis, resulting in the ester linkage between ONB and PMCL being converted to an aldehyde functionality (at approximately 9.8 ppm). FT-IR (Fig. 4C) further confirmed the photo-degradation and redox cleavage; the absorption of nitrogen dioxide at 1560 and 1460 cm⁻¹ almost disappeared, and demonstrated new absorptions of aromatic aldehyde C=O at 1660 and nitroso N–O at 1070 cm⁻¹. These results indicate that Glyco-ONB-s-s-PXCL can be degraded by either exposure to UV light or the presence of a reducing agent in the solution.

Fig. 4 (A) UV-Vis spectra change of Gluco-ONB-s-s-PMCL₂₀ in THF (1 mg mL⁻¹) recorded with treatment in different stimuli, (B) ¹H NMR spectra of Gluco-ONB-s-s-PMCL₂₀ after treatment with UV irradiation (352 nm, 64 W, 58.8 mW cm⁻²) and DTT (10 mM), (C) IR spectra of Gluco-ONB-s-s-PMCL₂₀: (a) original, (b) after treatment with UV irradiation and DTT (10 mM) for 5 h.

Micelles of Glyco-ONB-s-s-PXCL

The amphiphilic nature of Glyco-ONB-s-s-PXCL polymers, which consist of a hydrophilic glycose molecule and a hydrophobic ONB-s-s-PXCL segment, enables the formation of micelles in water. In this study, the characteristics of the Glyco-ONB-s-s-PXCL micelles in the aqueous phase were investigated using fluorescence techniques. The CMC of Glyco-ONB-s-s-PXCL in the aqueous phase was determined using pyrene as a probe molecule. The fluorescence intensity of the excitation spectrum of pyrene increased with the Gluco-ONB-s-s-PMCL₂₀ polymer concentration (Fig. 5A). In the pyrene excitation spectrum, a red-shift of the (0,0) band of pyrene from 334 to 337 nm was observed during partitioning into the micellar hydrophobic core. Fig. 5B displays the intensity ratios I₃₃₇/I₃₃₄ of pyrene excitation spectra versus the logarithm of Gluco-ONB-s-s-PMCL₂₀ polymers concentrations. The CMC was determined on the basis of the interaction of straight line segments drawn through the points of the lowest polymer concentrations which lie on a nearly horizontal line and the points of the rapidly rising region of the plot. Table 2 lists the CMC values of the various Glyco-ONB-s-s-PXCL polymers. The Glyco-ONB-s-s-PXCL polymers formed micelles in the aqueous phase, with their CMC values ranging from 1.2 to 5.2 mg L⁻¹. The Glyco-ONB-s-s-PXCL polymers exhibited lower CMC values than the surfactant (e.g. 2.3 g L⁻¹ for sodium dodecyl sulfate in water), indicating thermodynamically favorable self-association for Glyco-ONB-s-s-PXCL. Declining CMC values were observed with increasing hydrophilicity of the hydrophilic segment or increasing hydrophobicity of the hydrophobic segment. At a fixed hydrophilic sugar (glucose) level, the CMC values of the Gluco-ONB-s-s-PMCL series of polymers decreased from 2.1 to 1.4 mg L⁻¹ when the hydrophobic PMCL chain length increased from PMCL₂₀ to PMCL₃₄. The steric hindrance of the hydrophobic segment increased and the CMC values decreased. The mean hydrodynamic diameters of micelles incorporating IMC and the blank micelles, ranged from 122.6 to 144.0 nm and from 81.7 to 137.0 nm, respectively. Therefore, the micelles incorporating IMC were slightly larger than the blank micelles because of the incorporation of the hydrophobic drug; however, the micelle size remained smaller than 200 nm for all formations. This micelle size is suitable for intracellular drug delivery because an appropriate diameter (less than 200 nm) of aggregations is favorable to maintain the lowered level of reticuloendothelial system (RES) uptake, minimal renal excretion, and efficient EPR effect for passive tumor targeting.³⁸ At fixed glucose levels, micelles size increased with the length of the hydrophobic segment (PMCL). These responses indicated that micelle size is dependent on polymer composition (i.e., the length of the hydrophobic segment or the hydrophilicity of the hydrophilic moiety in the chain). The particle size distribution histogram for Glyco-ONB-s-s-PXCL micelles measured by DLS shows a monomodal peak, with a relatively narrow distribution (PD ≤ 0.18). Fig. 6A displays the spherical morphology of the Gluco-ONB-s-s-PMCL₂₀ micelles. When the drug was incorporated, micelle size marginally increased (Fig. 6C). The average micelle diameter was smaller than that recorded by the DLS (Fig. 6B and D). This is because TEM reveals the actual core dimensions of the micelles in a dry state, whereas DLS reveals the average dimensions of the micelles in aqueous solution. The stability of blank and IMC-loaded Gluco-ONB-s-s-PMCL₂₀ micelles was detected using DLS to monitor the diversification of the size and particle size distribution (PD). As shown in Fig. 6E, at pH 7.4 under UV irradiation and DTT dual stimuli conditions, the size and PD change from 140.6 (PD = 0.08) to 130 nm (PD = 0.07) for blank micelle, and from 126.3 (PD = 0.07) to 122.3 nm (PD = 0.09) for IMC-loaded micelle over the entire testing time, showing the stability of the micelles at pH 7.4, which was advantageous for circulation.

Fig. 5 (A) Excitation spectra of pyrene loaded Gluco-ONB-s-s-PMCL₂₀ micelles monitored at λ_em = 390 nm with different concentrations, (B) plot of the I₃₃₇/I₃₃₄ intensity ratio (from pyrene excitation spectra: pyrene concentration = 6.1 × 10⁻⁷ M) versus the logarithm of the concentration (log C) of Gluco-ONB-s-s-PBCL₁₈ (■); Gluco-ONB-s-s-PMCL₂₀ (▼); Gluco-ONB-s-s-PCL₂₀ (▲); Galacto-ONB-s-s-PMCL₂₀ (●); Malto-ONB-PMCL₂₀ (◆).

Table 2 Properties of Glycol-ONB-s-s-PXCL polymeric micelles

Polymer	CMC (mg L^−1)	Drug loading content^a (%)	Drug entrapment efficiency^a (%)	Micelle size^b (nm)
				Blank	PD^b	Zeta potential (mV)	With IMC	PD	Zeta potential (mV)
a Feed weight ratio IMC/polymer = 1/1.b Micelle size and particle size distribution (PD) determined by DLS.
Gluco-ONB-s-s-PCL20	3.5	34.4 ± 1.8	68.9 ± 3.5	81.7 ± 23.8	0.18	−32.6
Gluco-ONB-s-s-PMCL20	2.1	44.0 ± 0.3	88.0 ± 0.6	119.4 ± 51.0	0.11	−41.6	122.6 ± 48.5	0.11	−32.2
Gluco-ONB-s-s-PBCL18	1.2	35.6 ± 6.3	71.3 ± 12.7	120.6 ± 49.4	0.10	−43.4	126.9 ± 44.5	0.11	−34.2
Gluco-ONB-s-s-PMCL34	1.4	35.9 ± 2.5	71.8 ± 6.0	133.8 ± 34.3	0.07	−34.7	140.3 ± 39.9	0.10	−35.7
Malto-ONB-s-s-PCL20	5.2	29.7 ± 2.2	59.4 ± 4.5	102.1 ± 38.9	0.12	−38.5
Galacto-ONB-s-s-PCL20	3.7	16.6 ± 1.2	33.2 ± 2.3	137.0 ± 48.4	0.11	−56.3	144.0 ± 49.4	0.11	−35.4

---

Fig. 6 TEM and size distribution of the Gluco-ONB-s-s-PMCL₂₀ micelles: blank (A) (B), IMC-loaded (C) (D); (E) time-dependent size and PD of Gluco-ONB-s-s-PMCL₂₀ micelles at pH 7.4 under UV and DTT stimuli conditions: blank (▼, ▽), IMC-loaded (▲, △); TEM image change of the Gluco-ONB-s-s-PMCL₂₀ micelle treat with different stimuli for 5 h: (F) only with 10 mM DTT, (G) only with UV irradiation, (H) combined UV irradiation and 10 mM DTT (scale bar 0.2 μm).

Zeta potential (i.e., surface charge) greatly influences particle stability in suspension through electrostatic repulsion between particles. The dividing line between stable and unstable suspensions is generally taken at either +30 mV or −30 mV.²⁵ Micelle surfaces are negatively charged with comparable zeta potentials in the range of −32.2 mV to −56.3 mV demonstrating the resulting micelles are stable. A less negatively charged surface is favorable to prevent uptake by the phagocytic cells, which results in faster clearance from the blood.

Photo- and redox-cleavable behaviors of micelles

The photo- and redox-degradation of the micelles were evaluated at various treatment and irradiation intervals by monitoring changes in NR fluorescence, according to the reduction in the NR fluorescence intensity at 605 nm.³⁹ After NR and Gluco-ONB-s-s-PMCL₂₀ were dissolved in THF (1 mg mL⁻¹, NR to polymer ratio = 1 : 3), water was added to induce micelle formation and concomitant NR encapsulation by the micelle core. Subsequently, THF was removed through evaporation, and the nonsolubilized NR was filtered through microfiltration (0.2 μm pore filter). The final micelle concentration was adjusted to 0.2 mg mL⁻¹. Fig. 7 displays the fluorescence emission spectra of NR loaded in the Gluco-ONB-s-s-PMCL₂₀ micelle in PBS (pH 7.4, 37 °C) before and after treatment with DTT (10 mM) and/or UV irradiation (352 nm, 64 W, 58.8 mW cm⁻²) for 5 h. In the absence of DTT and UV, the intensity of the fluorescence emission spectra of NR remained generally constant over time, indicating micellar stability in the solution. By contrast, under UV irradiation, the intensity of the fluorescence emission spectra of NR decreased as the exposed time increased (Fig. 7A). Similarly, in the presence of DTT, the intensity of the fluorescence emission spectra of NR decreased with treatment time (Fig. 7B). The combined stimuli of UV irradiation and DTT rapidly decreased the intensity of the fluorescence emission spectra of NR within the treatment time (Fig. 7C). Fig. 7D displays normalized fluorescence plotted against time under different conditions: without DTT and UV irradiation, only with DTT, only with UV irradiation, and with a combination of both stimuli. In the absence of DTT and UV light, the emission intensity remained constant over time, indicating the stable encapsulation of the dye in a hydrophobic environment. By contrast, when expose to UV for 1 h, the emission intensity decreased to 63%. Over the same period (1 h), DTT treatment decreased emission intensity to 82%. However, under the condition of UV irradiation and DTT, the emission intensity rapidly decreased to 11% over 1 h. These results indicate that micelles disintegrate more rapidly under UV light than when exposed to DTT during the first 1 h period. This is caused by two different reaction mechanisms. Regarding the structure of micelle, the ONB moieties composed the micelle corona, the disulfide group located in the micelle core. In the presence of DTT, the reduction reaction of the disulfide group takes place at the micelle corona–core interface because water-soluble DTT molecules cannot diffuse into the rigid micelles core. The reaction is similar to a slow erosion of the micelles. By contrast, under UV light exposure, the micelle corona can absorb photons quickly, and the photolysis of o-nitrobenzyl methyl esters can proceed quickly and result in rapid micelle disintegration. The micelles exhibited redox and photo-labile properties in response to the reducing agent and light activation. The TEM images in Fig. 6F–H provided more insight into the effect of photocleavage or redox-cleavage both alone and combined on the micelles of NR-loaded Gluco-ONB-s-s-PMCL₂₀. Fig. 6F depicts the morphology of the micelles after treatment with DTT for 5 h. The TEM micrographs indicate distinct morphologies with some aggregation, confirming the disintegration of the micelles in aqueous solution. Before irradiation, micelles size was uniform. However, after UV irradiation, disintegrated micelles were observed, indicating that irradiation for a specific duration induced changes in the assembly state (Fig. 6G). Under the combined stimulation of UV irradiation and DTT, the spherical micelle disintegrated, and smaller irregularly shaped nanoaggregates were formed (Fig. 6H). Therefore, the TEM micelle characterization results revealed that altering the external environment with a reducing agent and light engenders the dissociation of the micelles. The sizes and distributions change of the NR-loaded micelles in response to 10 mM DTT and/or UV irradiation for 5 h were recorded by DLS (Fig. 7E). When expose to UV light or the presence of DTT, the micelle size of NR-loaded Gluco-ONB-s-s-PMCL₂₀ increased from 211.7 (PD = 0.23) to 267.2 nm (PD = 0.23), and from 211.7 (PD = 0.23) to 274.4 nm (PD = 0.25), with a slightly broader size distribution and larger aggregates with an average diameter greater than 1000 nm (curves c and d). Notably, under static experimental conditions, dissociated micelle fragments may randomly aggregate to form particles of larger sizes. However, in the presence of the combined stimuli, smaller sizes and distributions of micelles were detected (119.3 nm, PD = 0.10), indicating that ONB photo-cleavage and disulfide bond cleavage resulted in the dissolution of polymers. By contrast, no obvious size change was detected for Gluco-b-PMCL₂₀ micelles without disulfide linkage after 5 h of incubation under the same conditions (in the presence of 10 mM DTT).

Fig. 7 Fluorescence spectra of Nile red-loaded Gluco-ONB-s-s-PMCL₂₀ micelle in PBS (0.1 M, pH 7.4) at 25 °C for 0–5 h: (A) only with UV light (352 nm), (B) only with DTT (10 mM), (C) combined with UV light and DTT (10 mM), and (D) normalized fluorescence emission intensity vs. time of the NR-loaded micelle solution, (E) size distribution change of NR-loaded Gluco-ONB-s-s-PMCL₂₀ micelles treat with different stimuli for 5 h.

Carbohydrate–lectin binding recognition

The ability of the synthesized Glyco-ONB-s-s-PXCLs glycopolymers to interact with a biological system was assessed. Carbohydrates play a major role in biological recognition events mediated by specific carbohydrate–lectin interactions. The in vitro evaluation of this specific binding event thus provides an initial test of the ability of a synthetic Glyco-ONB-s-s-PXCL to interact with biological systems (e.g., drug delivery, tissue engineering, or other biomedical materials). These tests are typically conducted by mixing the Glyco-ONB-s-s-PXCL with a lectin that is selective for the sugar conjugated to the polymer.⁴⁰ A positive result is obtained by the appearance of a precipitation resulting from the aggregation of lectins, and measured as a reduced transparency of the solution. Con A is a specific lectin for the selective binding of glucosyl residues. Therefore, we investigated the change in absorbance of solutions of Gluco-ONB-s-s-PXCL with Con A at 600 nm. Fig. 8A indicates that the absorbance (i.e., turbidity) was greater for higher concentrations of Gluco-ONB-s-s-PMCL₂₀ because of the formation of larger aggregates. As a control experiment, a PBS buffer solution without glycopolymer was added to Con A. Only a slight absorbance (<0.006) was detected,⁴¹ which was substantially lower than that of the sample containing glucopolymers. These experiments confirmed that Gluco-ONB-s-s-PMCL₂₀ synthesized through the nucleophilic coupling of glucose to HONB-s-s-PMCL entails active bio-recognition. The lectin binding was selective, and depending on the concentration, instantaneous. The binding efficiency of various kinds of PXCL conjugated with glucose, glucopolymers with diverse PXCL compositions were also investigated. The influence of the PXCL on the rate of multivalent ligand–lectin clustering was assessed using Con A as the model lectin. The results were shown in Fig. 8B. In the four kinds of Gluco-ONB-s-s-PPXCL: Gluco-ONB-s-s-PBCL₁₈, Gluco-ONB-s-s-PCL₂₀, Gluco-ONB-s-s-PMCL₃₄, and Gluco-ONB-s-s-PMCL₂₀, the Gluco-ONB-s-s-PMCL₂₀ owned the highest binding efficiency with Con A. When the hydrophobicity or crystallization of the hydrophobic PXCL segment increased, the binding efficiency with Con A decreased, presumably because the hydrophobicity caused by the rigid backbone of PBCL and PCL may preclude the further binding of the glucose residues to the Con A tetramer. However, the absorbance of the binding glucopolymer was only slightly higher than the unbinding control due to the weak interaction between one sugar molecule and lectin.

Fig. 8 (A) Absorbance (600 nm) of the Gluco-ONB-s-s-PMCL₂₀ solutions in the presence of different concentrations upon reaction with lectin Con A (2 mg mL⁻¹) in PBS buffer: (■) without glycopolymer, (◆) with glycopolymer 0.2 mg mL⁻¹, (▼) with glycopolymer 0.5 mg mL⁻¹, (▲) with glycopolymer 1.0 mg mL⁻¹, and (●) with glycopolymer 2.0 mg mL⁻¹; (B) absorbance (600 nm) of various kinds of glycopolymers solution (2 mg mL⁻¹) upon reaction with lectin Con A (2 mg mL⁻¹) in PBS buffer: (■) without glycopolymer, (▼) with Gluco-ONB-s-s-PBCL₁₈, (●) with Gluco-ONB-s-s-PCL₂₀, (▲) with Gluco-ONB-s-s-PMCL₃₄, and (◆) with Gluco-ONB-s-s-PMCL₂₀.

Evaluation of drug-loading content, drug entrapment efficiency and stimuli-responsive release of IMC

The drug-loading content and drug entrapment efficiency of the polymeric micelles were determined using UV-Vis absorption spectroscopy. IMC, a common hydrophobic, non-steroidal, anti-inflammatory drug was selected as a model drug to investigate drug-loading in the hydrophobic core. The maximum absorption peak of IMC was proportional to its concentration at 320 nm. After IMC was released and the polymer precipitate was removed, the amount of loaded IMC was determined according to an absorbance at 320 nm. Table 2 lists the calculated drug-loading content and drug entrapment efficiency values. At a constant feed weight ratio (1 : 1), the IMC-loaded Gluco-ONB-s-s-PMCL₂₀ micelles maximized their drug-loading content (44%) and drug entrapment efficiency (88.0%). Compare with Dextran₁₀-ONB-PMCL₂₆, the DLC and DEE values of the Gluco-ONB-s-s-PMCL₂₀ micelles were high.⁴²

Dialysis was used to investigate the in vitro release of IMC from the dual-stimuli-cleavage IMC-loaded Glyco-ONB-s-s-PXCL micelles in the PBS (pH 7.4) at 37 °C under various conditions: (i) in the absence of UV and DTT; (ii) only with DTT (10 mM); (iii) only with UV light; and (iv) in the presence of UV and DTT. The cumulative release profiles of IMC from the IMC-loaded Gluco-ONB-s-s-PMCL₂₀ micelles are shown in Fig. 9. Environmental factors such as reducing agents and light significantly affected the release of IMC from the micelles. Without stimulation (in the absence of UV irradiation and DTT), only 43% of IMC was released after 24 h of incubation at pH 7.4 PBS. However, the release rate of IMC was considerably faster in the presence of DTT (10 mM), with approximately 70% of the encapsulated IMC being released sustainably during a 24 h period. Similarly, the release of IMC was accelerated under UV irradiation at pH 7.4, approximately 78% of the IMC was released in 24 h. The combined stimuli (UV and DTT), IMC release rate was promoted to 83% in 24 h. These results confirmed our hypothesis that release is achieved by rapid disintegration of the Gluco-ONB-s-s-PML₂₀ micelle corona in the presence of UV and DTT. This finding implies that micelle disintegration might be faster in response to light stimulus than to a reducing agent. The reduction and UV dual-responsive of Gluco-ONB-s-s-PMCL₂₀ micelles may have applications for the selective release of drugs in intracellular environments and with externally controlled stimuli.

Fig. 9 IMC release from the micelle of Gluco-ONB-s-s-PMCL₂₀ treated with different stimuli in 0.1 M PBS (pH 7.4, 0.1 M) at 37 °C: (▼) in the absence of DTT and UV light; (■) only with DTT (10 mM); (▲) only with UV light (352 nm, 64 w, 58.8 mW cm⁻²); and (●) in the presence of UV light and DTT (10 mM). Data are shown as mean ± S.E. (n = 3).

In vitro cytotoxicities of the polymer and DOX-loaded micelles

Cytotoxicity is crucial consideration in the design of drug carriers. An MTS assay was used to evaluate the in vitro cytotoxicities of the Gluco-ONB-s-s-PMCL₂₀ polymer, Gluco-b-PMCL₂₀ without stimuli (ONB, -s-s-) linkage polymer, DOX-loaded micelles, and free DOX by incubating HeLa cells with various polymer concentrations or DOX dosages. Fig. 10A illustrates the relative cell viability of cells treated with various concentrations of Gluco-ONB-s-s-PMCL₂₀ and Gluco-b-PMCL₂₀ for 48 h. Compared with the control of DMSO, the measured cell viability was ≥90% at Gluco-ONB-s-s-PMCL₂₀ concentration with a range from 1 to 300 μg mL⁻¹. The nanoparticles exhibited nonsignificant toxicity against HeLa cells. However, the cytotoxicity of Gluco-b-PMCL₂₀ showed higher than Gluco-ONB-s-s-PMCL₂₀. In addition, Fig. 10B presents the in vitro cytotoxicities of the DOX-loaded micelles and free DOX at various DOX dosages (0.125–2.0 μg mL⁻¹). The DOX-loaded Gluco-ONB-s-s-PMCL₂₀ micelles effectively inhibited the proliferation of HeLa cells with a half-maximal inhibitory concentration (IC₅₀) of 1.5 μg mL⁻¹. The DOX-loaded micelles possess slightly higher IC₅₀ values than the free DOX (1.25 μg mL⁻¹), which is observed in many polymeric system.⁴³ This is likely caused by the longer time required for DOX to be released from micelles into to tumor cells. Moreover, the dual-responsive DOX-loaded Gluco-ONB-s-s-PMCL₂₀ micelles showed a stronger inhibition of cell proliferation than the DOX-loaded Gluco-b-PMCL₂₀ micelles. The IC₅₀ values of DOX-loaded Gluco-ONB-s-s-PMCL₂₀ and Gluco-b-PMCL₂₀ micelles are 1.5 and 1.75 μg mL⁻¹, respectively. These results indicate that the DOX release more rapidly in the Gluco-ONB-s-s-PMCL₂₀ micelle than the Gluco-b-PMCL₂₀ micelle under UV and DTT condition.

Fig. 10 Cell viabilities of HeLa cells treated: (A) with various concentration of Gluco-ONB-s-s-PMCL₂₀ (black) and Gluco-b-PMCL₂₀ lack of stimuli linkage (gray), (B) with DOX-loaded Gluco-ONB-s-s-PMCL₂₀ micelles (●), DOX-loaded Gluco-b-PMCL₂₀ micelles (▲), and free DOX (■) for 48 h. Data are shown as mean ± S.E. (n = 3).

Conclusions

Reductive and light dual-responsive glycopolymers (Glyco-ONB-s-s-PXCL) with dual light- and redox-cleavable junction between the hydrophilic sugar molecules and hydrophobic blocks were successfully synthesized using ROP and nucleophilic substitution reactions. The polymers formed micelles in an aqueous solution, had hydrodynamic sizes ≤200 nm, and were spherical. Under the combined stimulation of UV irradiation and a reducing agent DTT, the micellar nanoparticles dissociated; therefore the loaded molecules could be released from the assemblies more efficiently than under exposure only one of the two stimuli. Cell viabilities were evaluated in response to the polymer concentration (1–1000 μg mL⁻¹), and the nanoparticles exhibited nonsignificant toxicity against HeLa cells at concentrations ≤300 μg mL⁻¹. Rapid and selective biorecognition was demonstrated through lectin clustering experiments. For the HeLa cells, the effectively inhibited proliferation of the DOX-loaded Gluco-ONB-s-s-PMCL₂₀ micelles is similar to the free DOX. Therefore, the study results indicate that redox- and light-dual-cleavage Gluco-ONB-s-s-PMCL₂₀ can be used for targeted drug delivery.

Acknowledgements

This research was supported by grants from the Ministry of Science and Technology, Taiwan, (MOST 104-2221-E-182-061) and Chang Gung Memorial Hospital (CNRPD5D0012).

Notes and references

1. F. Meng, Z. Zhong and J. Feijen, Biomacromolecules, 2009, 10, 197–209.
2. E. Fleige, M. A. Quadir and R. Haag, Adv. Drug Delivery Rev., 2012, 64, 866–884.
3. A. Klaikherd, C. Nagamani and S. Thayumanavan, J. Am. Chem. Soc., 2009, 131, 4830–4838.
4. F. Meng, W. E. Hennink and Z. Zhong, Biomaterials, 2009, 30, 2180–2198.
5. R. Cheng, F. Meng, C. Deng, H. A. Klok and Z. Zhong, Biomaterials, 2013, 34, 3647–3657.
6. S. M. A. Soliman, L. Colombeau, C. Nouvel, J. Babin and J. L. Six, Carbohydr. Polym., 2016, 136, 598–608.
7. H. Mori, I. Kato and T. Endo, Macromolecules, 2009, 42, 4958–4992.
8. W. Shao, K. Miao, H. Liu, C. Ye, J. Du and Y. Zhao, Polym. Chem., 2013, 4, 3398–3410.
9. K. Miao, H. Liu and Y. Zhao, Polym. Chem., 2014, 5, 3335–3345.
10. H. Liu, C. Li, D. Tang, X. An, Y. Guo and Y. Zhao, J. Mater. Chem. B, 2015, 3, 3959–3971.
11. M. Tong, X. An, W. Pan, H. Liu and Y. Zhao, Polym. Chem., 2016, 7, 2209–2221.
12. Q. Zhang, N. R. Ko and J. K. Oh, Chem. Commun., 2012, 48, 7542–7552.
13. J. Liu, Y. Pang, W. Huang, X. Huang, L. Meng, X. Zhu, Y. Zhou and D. Yan, Biomacromolecules, 2011, 12, 1567–1577.
14. D. Han, X. Tong and Y. Zhao, Langmuir, 2012, 28, 2327–2331.
15. J. Xuan, D. Han, H. Xia and Y. Zhao, Langmuir, 2014, 30, 410–417.
16. D. Li, Y. Bu, L. Zhang, X. Wang, Y. Yang, Y. Zhuang, F. Yang, H. Shen and D. Wu, Biomacromolecules, 2016, 17, 291–300.
17. P. Zhang, H. Zhang, W. He, D. Zhao, A. Song and Y. Luan, Biomacromolecules, 2016, 17, 1621–1632.
18. P. M. Kharkar, A. M. Kloxin and K. L. Klick, J. Mater. Chem. B, 2014, 2, 5511–5521.
19. J. D. Byrne, T. Betancourt and L. Brannon-Peppas, Adv. Drug Delivery Rev., 2008, 60, 1615–1626.
20. S. G. Spain and N. R. Cameron, Polym. Chem., 2011, 2, 60–68.
21. Y. H. Bae and K. Park, J. Controlled Release, 2011, 153, 198–205.
22. R. Sunasee and R. Narain, Macromol. Biosci., 2013, 13, 9–27.
23. J. Rieger, F. Stoffelbach, D. Cui, A. Imberty, E. Lameignere, J. L. Putaux, R. Jérôme, C. Jérôme and R. Auzély-Velty, Biomacromolecules, 2007, 8, 2717–2725.
24. L. M. Amanda, L. Bo and R. G. Elizabeth, J. Am. Chem. Soc., 2009, 131, 734–741.
25. Y. Zhang, J. He, D. Cao, M. Zhang and P. Ni, Polym. Chem., 2014, 5, 5124–5138.
26. W. Chen, Y. Zou, F. Meng, R. Cheng, C. Deng, J. Feijen and Z. Zhong, Biomacromolecules, 2014, 15, 900–907.
27. S. Kumar, J. F. Allard, D. Morris, Y. L. Dory, M. Lepage and Y. Zhao, J. Mater. Chem., 2012, 22, 7252–7253.
28. G. Liu and C. M. Dong, Biomacromolecules, 2012, 13, 1573–1583.
29. Y. Wu, D. Zhou, Y. Qi, Z. Xie, X. Chen, X. Jing and Y. Huang, RSC Adv., 2015, 5, 31972–31983.
30. K. Usui, T. KiKuchi, K. Y. Tomizaki, T. Kakiyama and H. Mihara, Chem. Commun., 2013, 49, 6394–6396.
31. Y. Zhang, Z. Dong, M. Nomura, S. Zhong, N. Chen, A. M. Bode and Z. Dong, J. Biol. Chem., 2001, 276, 20913–20923.
32. J. Ding, J. Chen, D. Li, C. Xiao, J. Zhang, C. He, X. Zhuang and X. Chen, J. Mater. Chem. B, 2013, 1, 69–81.
33. H. Huang, S. Wu, Z. Xie, F. Meng, X. Jing and Y. Huang, Biomacromolecules, 2012, 13, 3004–3012.
34. K. Y. Peng, S. W. Wang and R. S. Lee, J. Polym. Sci., Part A: Polym. Chem., 2013, 51, 2769–2781.
35. N. Murthy, J. Campbell, N. Fauston, A. S. Hoffman and P. S. Stayton, Bioconjugate Chem., 2003, 14, 412–419.
36. L. Jia, D. Cui, J. Bignon, A. D. Cicco, J. Wdzieczak-Bakala, J. Liu and M. H. Li, Biomacromolecules, 2014, 15, 2206–2217.
37. H. Cai, G. Jiang, Z. Shen and X. Fan, Macromolecules, 2012, 45, 6176–6184.
38. K. Kataoka, A. Harada and Y. Nagasaki, Adv. Drug Delivery Rev., 2001, 47, 113–131.
39. C. Lv, Z. Wang, P. Wang and X. Tang, Langmuir, 2012, 28, 9387–9394.
40. M. Ambrosi, N. R. Cameron and B. G. Davis, Org. Biomol. Chem., 2005, 3, 1593–1608.
41. J. Huang, G. Habraken, F. Audouin and A. Heise, Macromolecules, 2010, 43, 6050–6057.
42. R. S. Lee, Y. C. Li and S. W. Wang, Carbohydr. Polym., 2015, 117, 201–210.
43. Y. Wu, D. Zhou, Y. Qi, Z. Xie, X. Chen, X. Jing and Y. Huang, RSC Adv., 2015, 5, 31972–31983.

---