Temperature- and pH-responsive nanoparticles of biocompatible polyurethanes for doxorubicin delivery

International Journal of Pharmaceutics 441 (2013) 30–39

Contents lists available at SciVerse ScienceDirect

International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm

Temperature- and pH-responsive nanoparticles of biocompatible polyurethanes for doxorubicin delivery Anning Wang a , Hui Gao a,∗ , Yanfang Sun a , Yu-long Sun b , Ying-Wei Yang b , Guolin Wu c , Yinong Wang c , Yunge Fan c , Jianbiao Ma a,∗ a

School of Chemistry and Chemical Engineering, Tianjin University of Technology, Tianjin 300384, China State Key Laboratory of Supramolecular Structure and Materials, College of Chemistry, Jilin University, Changchun 130012, China c Key Laboratory of Functional Polymer Materials (Ministry of Education), Institute of Polymer Chemistry, Nankai University, Tianjin 300071, China b

a r t i c l e

i n f o

Article history: Received 18 October 2012 Received in revised form 20 November 2012 Accepted 12 December 2012 Available online 20 December 2012 Keywords: pH Temperature Responsive Nanoparticle Polyurethane Doxorubicin

a b s t r a c t A series of temperature- and pH-responsive polyurethanes based on hexamethylene diisocyanate (HDI) and 4,4 -diphenylmethane diisocyanate (MDI) were synthesized by a coupling reaction with bis-1,4(hydroxyethyl) piperazine (HEP), N-methyldiethanolamine (MDEA) and N-butyldiethanolamine (BDEA), respectively. The chemical structure, molecular weight, thermal property and crystallization properties were characterized by Fourier transform infrared (FT-IR) spectroscopy, nuclear magnetic resonance (NMR) spectroscopy, gel permeation chromatography (GPC), differential scanning calorimetry (DSC) and X-ray diffraction (XRD) spectroscopy. The resulting polyurethanes were then used to prepare nanoparticles either by direct dispersion method or dialysis method. Their pH and temperature responsibilities were evaluated by optical transmittance and size measurement in aqueous media. Interestingly, HDIbased and MDI-based polyurethanes exhibited different pH and temperature responsive properties. Nanoparticles based on HDI-HEP and HDI-MDEA were temperature-responsive, while MDI-based biomaterials were not. All of them showed pH-sensitive behavior. The possible responsive mechanism was investigated by 1 H NMR spectroscopy. The cytotoxicity of the polyurethanes was evaluated using methylthiazoletetrazolium (MTT) assay in vitro. It was shown that the HDI-based polyurethanes were non-toxic, and could be applied to doxorubicin (DOX) encapsulation. The experimental results indicated that DOX could be efficiently encapsulated into polyurethane nanoparticles and uptaken by Huh-7 cells. The loaded DOX molecules could be released from the drug-loaded polyurethane nanoparticles upon pH and temperature changes, responsively. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Intelligent polymers, whose characteristics changes in response to various external stimuli, such as light, pH, electric potential, magnetic field, temperature, etc., are spotlighted in material science, tissue engineering and drug delivery systems (Lee et al., 2007). Especially, pH and temperature changes as special triggers have drawn much attention in recent years, owing to their physiological importance in the human body and practical advantages both in vitro and in vivo (Fu et al., 2011). As a class of important smart material, pH-sensitive polymers can respond to pH changes in the surrounding medium to adjust their structures and conformations, due to the disequilibrium of ionization–deionization of polymers in aqueous solution caused by

∗ Corresponding authors. Tel.: +86 2260214259; fax: +86 2260214251. E-mail addresses: [email protected] (H. Gao), [email protected] (J. Ma). 0378-5173/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ijpharm.2012.12.021

ionizable functional groups (a weak acid or a weak base) (Filippov et al., 2008). It is well known that different parts of the body have different optimal pH levels (Bae et al., 2005a, 2005b; Zhou et al., 2012). The slightly acidic microenvironment of tumors could be employed as targeting sites for pH-sensitive drug delivery (Aryal et al., 2010; Engler et al., 2011; Haining et al., 2004; Poon et al., 2011; Wang et al., 2004). On the other hand, temperature is also one of the most commonly used stimuli due to its easy operation and many other practical advantages both in vitro and in vivo (He et al., 2008). Temperature-sensitive nanoparticles show great potential as cancer drug carriers, because the temperature in a specific pathological site is usually higher or can be externally manipulated to higher values (Chen et al., 2007). Taking advantage of this property, drug delivery can be intelligently triggered by temperature changes. Inspired by these, nanoparticles responsive to the fluctuation of both pH and temperature in physiological conditions can be more attractive as advanced drug carriers for cancer therapy due to possible synergistic advantages (Akiyoshi et al., 1997; Qiu et al., 1997).

A. Wang et al. / International Journal of Pharmaceutics 441 (2013) 30–39

Recently, researchers developed pH/temperature multiple sensitive polymers for drug delivery in the form of beads (Kim et al., 1994) or hydrogels (Dayananda et al., 2008). However, pH/temperature-responsive polymeric nanoparticles are rarely studied, to the best of our knowledge. Nanoparticles will endow the polymer specific properties that are quite different from bulk materials (Shah et al., 2012). Kang et al. (2003) synthesized a series of polymers by the copolymerization of methacryloyl poly(Nisopropylacrylamide-co-N,N-dimethylacrylamide) and methacryloyl sulfamethoxypyridazine telomers. The resulting polymers were then used to fabricate nanoparticles loaded with doxorubicin (DOX), showing a pH and temperature-sensitive release profile. However, polymers composed of acylamide were not biodegradable and their cytotoxicities were not considered in their study. Polyurethanes as important biomaterials have been widely used in stimuli-responsive drug delivery systems (DDS) for controlled drug release, tissue engineering scaffolds, artificial muscles, etc., due to their attractive physical properties and good biocompatibility (Chen et al., 2000; Gavini et al., 2009; Huynh et al., 2010; Lan et al., 1996; Loh et al., 2008; Zhang et al., 2008). Recently, pH-sensitive polyurethane micelles based on carboxylic groups (Ding et al., 2009), amino groups (Huynh et al., 2010) and acidic cleavage hydrazone linkage (Bae et al., 2003, 2005a, 2005b) have been developed for drug delivery applications. The biologically active drug is released by the disassembly of micelle in response to external pH change (Sun et al., 2011). Very recently, temperature-responsive polyurethanes were developed by our research groups (Fu et al., 2011). Herein, HDI and MDI bearing an alkyl and aromatic chain, as well as three diols, bis-1,4-(hydroxyethyl) piperazine (HEP), N-methyldiethanolamine (MDEA) and N-butyldiethanolamine (BDEA), composed of pHsensitive amino groups bearing different carbon chain length and steric structure were chosen to prepare polyurethanes. This variation in structure will influence the hydrophilic–hydrophobic balance of the polymer, and thus pH and temperature sensitivity. Nanoparticles were successfully fabricated from these polymers and their pH/temperature-responsive properties were systematically investigated. In addition, the cytotoxicity of these nanoparticles was evaluated against Human Umbilical Vein Endothelial Cells (HUVEC cells), and the potential use of these polymeric nanoparticles as drug delivery platform was also explored, employing DOX as a model anti-cancer drug (Fig. 1).

2. Experimental 2.1. Materials Hexamethylene diisocyanate (HDI, 99%), 4,4 -diphenylmethane diisocyanate (MDI, 99%), N-methyldiethanolamine (MDEA, 99%), Nbutyldiethanolamine (BDEA, 99%) and dibutyltin dilaurate (DBTDL, 95%) were purchased from Aladdin Reagent Co., Ltd. (Shanghai, China). Bis-1,4-(hydroxyethyl) piperazine (HEP, 99%) was obtained from Nanjing Chemlin Chemical Industry Co., Ltd. (Nanjing, China). N-Dimethylformamide (DMF, 99%) was freshly distilled under reduced pressure before subjecting to any reactions. Toluene and dichloroethane were distilled over sodium or calcium hydride at ambient pressure, dried over activated 4 A˚ molecular sieves before use. Doxorubicin hydrochloride (DOX) was purchased from Beijing Huafeng United Technology Co., Ltd. All other reagents and organic solvents obtained from Tianjin Chemical Reagent Co. (Tianjin, China) were of reagent grade and used as received.

31

2.2. Synthesis of polyurethanes The two series of polyurethanes were synthesized by the condensation reaction via the coupling reaction between terminal hydroxyl groups of HEP, MDEA and BDEA and isocyanate groups of HDI and MDI, respectively. As for HDI-based polyurethanes, dibutyltin dilaurate was used as catalyst, while no catalyst was employed for MDI-based polyurethanes due to its high reaction activity. The stoichiometric ratio of OH and NCO groups was adjusted to NCO/OH = 1.0 (in the case of HDI) or 1.2 (in the case of MDI). Briefly, HEP (1.76 g, 10 mmol) was added into a dry threeneck round bottom flask equipped with a magnetic stir bar. The flask was heated at 100 ◦ C under vacuum for 10 min, followed by nitrogen replacement for three times. After cooling the flask to 80 ◦ C, dibutyltin dilaurate (0.5 wt%, with respect to the reactant, in the case of HDI-based polyurethanes) and 60 mL of anhydrous toluene/dichloroethane (50:50) as solvent were injected. Subsequently, the requisite amount of HDI (1.63 mL, 10 mmol) or MDI (3.02 g, 12 mmol) was added and the reaction mixture was allowed to react at 80 ◦ C for 4 h under nitrogen atmosphere. When MDEA (1.17 mL, 10 mmol) or BDEA (1.65 mL, 10 mmol) was used as reactant, DMF was employed as reaction solvent. Finally, the resulting products were purified twice by precipitating in a 10-fold excess of diethyl ether, before drying under vacuum to constant weight at 40 ◦ C for 48 h. The yields of all these polymers were over 70%. 2.3. Characterization of polyurethanes 2.3.1. Nuclear magnetic resonance (NMR) spectroscopy and Fourier transform infrared (FT-IR) spectroscopy The resulting polyurethanes were characterized by 1 H and 13 C NMR spectroscopy. The spectra were recorded at room temperature in chloroform-d (CDCl3 ) or (methyl sulfoxide)-d6 (DMSO-d6 ), using a bruker-400 MHz NMR spectrometer. FT-IR tests were measured by Bio-Rod 6000 (Thermo Electron, USA) spectrometer using KBr pellets. 2.3.2. Gel permeation chromatography (GPC) The molecular weights and polydispersity indexes (PDI) of synthesized polymers were determined by GPC (Waters 2414 system Milford, MA, equipped with a refractive index detector), using DMF or chloroform as eluent at a flow rate of 1.0 mL/min at 35 ◦ C. Calibration curves were obtained with nearly monodispersed polystyrene. 2.3.3. Thermal analysis The thermogravimetry analysis (TGA) of these polyurethanes was carried out under nitrogen atmosphere with a heating rate at 10 ◦ C/min using a thermogravimetric analyzer (Netzsch TG209). DSC was carried out under a nitrogen flow rate of 50 mL/min with a differential scanning calorimeter (Netzsch DSC200). Specimens (3–5 mg) in an aluminum pan was heated from room temperature to 200 ◦ C (HDI-based polyurethanes) or 160 ◦ C (MDI-based polyurethanes); cooled to −60 ◦ C rapidly and kept at 60 ◦ C for 3 min; and heated again to 200 ◦ C or 160 ◦ C with a heating rate at 10 ◦ C/min. Heating curves were collected during the second heating run. 2.3.4. X-ray diffraction (XRD) spectroscopy Crystallization properties of samples were determined by ARLX’TRA powder XRD system (Rigaku D/max 2500v/pc, Japan). Samples were freeze-dried before measurement. X-ray generator was equipped with a rotating copper anode and nickel filter. All the polyurethanes were scanned at 40 kV, 100 mA using Cu K␣1 ˚ at the range of 3–50◦ . radiation ( = 1.5406 A)

32

A. Wang et al. / International Journal of Pharmaceutics 441 (2013) 30–39

Fig. 1. Schematic illustration of DOX incorporation and release of polyurethane nanoparticles under different conditions.

2.4. Nanoparticles fabrication and evaluation of pH/temperature-responsive behavior 2.4.1. Nanoparticle fabrication The two series of polyurethanes nanoparticles were prepared using two different methods. HDI-based polyurethane nanoparticles were prepared by dispersing the polymers (10 mg) directly in hydrochloric acid (pH = 2.0, 20 mL) under magnetic stirring, while nanoparticles of MDI-based polyurethane were prepared by dialysis method. Briefly, polyurethane (5 mg) was dissolved in DMF (5 mL), and dialyzed against double distilled water.

2.4.2. Acid–base titration Polymer nanoparticles (pH = 2.0, 1 mg/mL) were prepared for the titration measurement. The mixture was stirred for 24 h to insure complete disolvation of the polymer. Titrations were carried out by stepwise addition of 100 ␮L of NaOH (0.05 mol/L) with two minutes interval for equilibrium each time. The pH value was checked by a pH meter of PB-10 (Sartorius instrument, Germany). The buffer capacity and pKa values of nanoparticles could be calculated from the titration curves. The ionization degree ˛ of the amino groups in the polymer was calculated as ˛ = ([basic] − [OH− ] + [H+ ])/CNH , where [basic], [OH− ] and [H+ ] were the molarity of added NaOH for titration, free hydroxide ion and hydrogen ion, respectively, and CNH is the total molar concentration of the amino groups in molarity. The pH at ˛ = 0.5 is considered as the apparent dissociation constant pKa of polyurethanes. 2.4.3. Optical transmittance The optical transmittance tests were carried out at 410 nm using a UV-2450 instrument (Shimadzu Co., Japan), equipped with a temperature-controllable cell. HCl (0.05 M) or NaOH (0.05 M) aqueous solution was used to adjust pH values of the nanoparticle solutions. Then the temperature-responsive behavior of nanoparticles was studied at critical pH values slightly lower than pKa where nanoparticles begun to aggregate. The temperature interval was configured at 5 ◦ C ranging from 20 ◦ C to 55 ◦ C, and equilibrated for 15 min before each measurement.

2.4.4. Dynamic light scattering (DLS) The mean diameter and PDI of nanoparticles were determined on a Zetasizer Nano ZS90 instrument (Malvern Instruments, Southborough, MA). The pH and temperature of nanoparticle solutions were set corresponding to the optical transmittance assay.

2.4.5. Scanning electron microscope (SEM) Nanoparticle samples were dropped on a coverslip and dried at room temperature. The morphology of nanoparticles at different pH values or temperatures was observed on JSM-6700F type field emission SEM (JEOL, South Korea) after sputtered with gold and scanned at an accelerated voltage of 10 kV. 2.4.6. In vitro degradation tests The degradation tests of polyurethane membranes, which were prepared by solvent casting method, were evaluated by recording the weight loss over time in PBS (154 mM NaCl, pH = 7.4) at 37 ◦ C. These samples were respectively placed into small vials filled with 3 mL buffer solution containing 0.5% (w/v) sodium azide as antimicrobial agent and immersed in a 37 ◦ C water bath with cyclic shaking to simulate dynamic in vivo tissue environment. The films were washed with deionized water and dried in vacuum to a constant weight before weight loss analysis once a week. 2.4.7. Cytotoxicity analysis Human Umbilical Vein Endothelial Cells (HUVEC cells) were precultured for 18 h in Cell System-Corporation medium (Applied Cell Biology Research Institute) in a gelatin-coated 24-well plate. The cells were then exposed to a serial concentration of the polymeric nanoparticles at 37 ◦ C for 48 h. Free DOX and DOX-loaded HDI-MDEA nanoparticles (40 ␮g DOX equiv./mL) were incubated on Huh-7 hepatocarcinoma cell line. The medium was removed and 100 ␮L of MTT solution (5 mg/mL, PBS) was added to the cultures, and incubated for another 4 h. Cells incubated with media were tested for control. The growth medium was replaced with 100 ␮L of DMSO and the resulting solution was measured using a microplate reader (Model 680, BIO-RAD). The cell viability in each well was calculated from the obtained values as a percentage of control wells. The results were presented as a mean and standard deviation obtained from four samples. 2.4.8. DOX loading and in vitro release Polyurethanes (38 mg) and DOX hydrochloride (38 mg) were dissolved separately in DMF (3 mL). The DOX solution was added to the polymer solution after triethylamine (12 ␮L) was added dropwise and stirred for 2 h. The mixture was then dialyzed against deionized water at room temperature for 24 h using dialysis bag (MWCO = 7000). The mean size of DOX-loaded nanoparticles was determined by DLS. The dialysate was freeze-dried to a constant weight. The DOX-loaded particles were dissolved in DMSO, and the UV absorbance of DOX was measured at 480 nm using UV-2450 (Shimadzu Co., Japan). Final loading capacity (LC) and

A. Wang et al. / International Journal of Pharmaceutics 441 (2013) 30–39

33

encapsulation efficiency (EE) were calculated from the following equations (Eqs. (1) and (2)): LC (% w/w) = EE (%) =

mass of loaded guest × 100 mass of nano particles

final loading × 100 initial loading

(1) (2)

In vitro release tests were carried out under different pH values and temperatures. Lyophilized DOX-loaded nanoparticles (2 mg/mL) were dispersed in acetate buffer solution (154 mM NaCl, pH = 4.0) or PBS (154 mM NaCl, pH = 7.4), respectively, and dialyzed against their corresponding buffer solutions (20 mL) in capped beakers at 37 ◦ C. In vitro release at different temperatures (37 ◦ C and 50 ◦ C) was carried out at pKa of polyurethane nanoparticles. At every designated interval, buffer solutions (5 mL) in the beaker were taken out and fresh buffer solution (5 mL) was replenished to keep a constant volume. The amount of DOX released into the buffer solution was analyzed using a UV spectrophotometer at a wavelength of 480 nm. The concentration of DOX released from the nanoparticles was expressed as a percentage of the total DOX available and plotted as a function of time. The cumulative DOX release was calculated through equation below (Eq. (3)): cumulative DOX release (%) =

Mt × 100 M∞

(3)

where Mt is the amount of drug released from nanoparticles at time t and M∞ is the amount of drug released from the polymeric nanoparticles at time infinity. 2.4.9. Confocal laser scanning microscopy (CLSM) The intracellular distribution of free DOX and drug-loaded nanoparticles were followed with CLSM using Huh-7 cells (5 × 104 cells/well) cultured in a 35 mm glass base dish (Iwaki, Tokyo, Japan) for one day. Then, the complex (40 ␮g DOX equiv/mL) was added, and cells were cultured for 6, or 24 h in a humidified 5% CO2 -containing atmosphere. After washing with PBS, the cells were stained with Hoechst 33342 (Dojindo Laboratories, Kumamoto, Japan). CLSM images of cells were obtained using a Zeiss LSM510 Laser Confocal Scanning Microscopy imaging system at the excitation wavelengths of 480 nm (Ar laser) for DOX, and 710 nm (MaiTai laser, 2 photon excitation; Spectra-Physics, Mountain View, CA) for Hoechst 33342. The cells treated with free DOX (40 ␮g/mL) were used as a control. 3. Results and discussion 3.1. Synthesis and characterization of polyurethanes Two series of polyurethanes were synthesized by a condensation reaction between the isocyanate groups of HDI or MDI and hydroxyl groups of HEP, MDEA or BDEA (Scheme 1). The polyurethanes were consisted of HDI or MDI as hydrophobic segments, and HEP, MDEA or BDEA as hydrophilic segments. The chemical structures of the resulting polymers were characterized by 1 H and 13 C NMR as well as FT-IR. The molecular weights were determined by GPC. Fig. S1 shows 1 H and 13 C NMR spectra of HDIHEP. The peaks a, b and c at 3.16 ppm, 1.5 ppm and 1.34 ppm were assigned to the methylene protons of HDI. The peaks e, f and g at 4.19 ppm, 2.65 ppm and 2.59 ppm were assigned to the methylene protons of HEP. A feeble peak d appeared at 5.08 ppm was attributed to OOCNH . In 13 C NMR (Fig. S1b), the characteristic peak of isocyanate at 122.9 ppm was not observed, indicating that the HDI monomer was completely consumed during the reaction. Instead, peak d aroused at 156.5 ppm was attributed to OOCNH . The FTIR measurements further confirmed the success of condensation

Scheme 1. Synthesis of two series of polyurethanes based on HDI and MDI.

reaction. Assignments of adsorption peaks for HDI-HEP and MDIHEP are presented in Fig. S2. The peaks at 3339 cm−1 and 1532 cm−1 correspond to the N H stretching and deformation vibration band of the urethane, respectively. The C O stretching band belonging to the hard segment appears at 1713 cm−1 . In addition, the absence of peaks at around 2260 cm−1 indicates that no isocyanate group remains in the obtained polymer. These records confirmed the formation of polyurethanes. The molecular weight and PDI of polyurethanes are given in Table 1. Their molecular weights were in the range of (1.22–4.05) × 104 with PDI of 1.55–2.68, endowing the polyurethanes appropriate mechanical and processing properties for biomedical applications. As shown in Table 1, the molecular weights of HDI-based polyurethanes are similar, while MDI-based polyurethanes exhibit a decrease in molecular weight with the prolongation of alkyl side chains. The influence of steric hindrance effect resulted from the alkyl side chains of diol on MDI-bearing aromatic ring was greater than that of HDI-bearing alkyl chain. The existence of steric hindrance effect of butyl side chains hindered MDI and BDEA from polymerization, thus resulting in the lowest molecular weight for polyurethane MDI-BDEA. 3.2. Thermal analysis of polyurethanes Thermal properties of the synthesized polyurethanes were studied by DSC (Fig. S3 and Table 1). Endothermic peaks were detected for HDI-based polyurethanes, suggesting crystalline region existed in HDI-based polyurethanes, which was confirmed from the analysis of the XRD profiles (Fig. S4). The hydrogen bonding arose from HDI series may be attributed to the formation of crystalline region. Basically, these thermograms presented two characteristic transitions corresponding to hard phase transition of diisocyanate (Tg hard ) and order–disorder transition (ODT), respectively. The glass transition of diol of soft phase (Tg soft ) was probably lower than the measurement temperature. Tg hard of MDI-based polyurethanes was higher than that of HDI-based compartment, which may be attributed to the relative rigid conformation of MDI segment. For the polyurethane bearing same hard segments, Tg hard was affected by the mobility of soft segments, where BDEA < MDEA < HEP. In addition, ODT caused by micro-phase separation was observed at Table 1 Molecular weights, PDI and DSC data of polyurethanes. Polyurethanes HDI-HEP HDI-MDEA HDI-BDEA MDI-HEP MDI-MDEA MDI-BDEA

Mn 25,100 23,900 22,600 40,500 25,500 12,200

Mw 52,700 43,500 35,000 70,900 68,400 24,700

PDI

Tg hard (◦ C)

2.10 1.82 1.55 1.75 2.68 2.02

−32.9 −8.5 −6.0 3.7 4.2 4.6

ODT (◦ C)

145 139 82.5 77.1

Tm (◦ C) 30.3 56.2 65.5

34

A. Wang et al. / International Journal of Pharmaceutics 441 (2013) 30–39

Fig. 2. Acid–base titration curves of polymeric nanoparticle solutions.

higher temperature for HDI-based polyurethane, indicating the relative high stability of micro-domain of HDI-based polymers due to their intermolecular and intramolecular hydrogen-bonding interaction. 3.3. Hydrolytic degradation Degradation of the synthesized polymers was carried out by submerging the solution-cast films in 3 mL PBS (154 mM NaCl, pH = 7.4) at 37 ◦ C. The hydrolytic degradation was examined by measuring the weight loss of polyurethanes. Fig. S5 shows the weight loss of polyurethanes as a function of degradation time. The degradation rate was strongly affected by their components. HDIbased polyurethanes exhibited higher mass loss than MDI-based polyurethanes. The weight loss of HDI-MDEA could reach up to 60% at week 5, compared with the 8% of weight loss of MDI-MDEA. These results are expectable considering the lower hydrophobicity, lower initial molecular weight and more flexibility of polymer chains of the HDI-based polyurethanes. The chain mobility of MDI-based polyurethanes was less on account of the rigid MDI segments. In addition, the low initial molecular weight induced decreased interaction between polymer chains, which facilitated the attack of water molecules to the backbone. The degradation rate of polyurethanes based on the same hard segment was also different due to the different hydrophobicity of soft segment. The higher hydrophobicity of BDEA hindered water molecules from attacking the polymer backbone, resulting in the lower degradation rate of corresponding polyurethanes. 3.4. pH-responsive property of polyurethane nanoparticles In order to understand the aggregation behavior of the obtained polymers and calculate their pKa values, a stepwise titration of polymer solution (1.0 mg/mL) with NaOH (0.05 mol/L) was conducted at 37 ◦ C. The pH titration curves were acquired by recording the changes in pH values upon the addition of NaOH. The typical acid-base titration profiles are shown in Fig. 2. The buffering capacity of HDI-based polymers was confirmed, as testified by the existence of plateaus. There were two mutations in the course of titration tests. The first mutation can be defined as the process of deprotonation, which was followed by acid–base neutralization, while the second one was considered as the emergence of precipitations. The pKa values of HDI-HEP, HDI-MDEA and HDI-BDEA were extracted from the titration curves and the corresponding values were 6.40, 6.66 and 5.48, respectively, which was influenced by

the chemical structure of the polymers and the density of the amino groups in the polymers. the buffering property of MDI-based Nonetheless, polyurethanes was feeble because of the high ratio of hydrophobic part to hydrophilic part. For the relatively hydrophobic polyurethane backbone, changing the ionization degree seriously impacts the hydrophobic–hydrophilic balance (Filippov et al., 2008). The precipitation occurred at the same time as deprotonation took place. Therefore, there was only one mutation appeared in the titration curve. Two methods were employed for nanoparticle preparation due to their different solubility. HDI-based polyurethanes were not soluble, or self-assembled into nanoparticles at neutral pH. HCl solution was used to solubilize HDI-based polyurethane and form nanoparticles. MDI-based polyurethanes could not be solubilized in HCl solution, but soluble in DMF. Therefore, dissolving the MDI-based polyurethane followed by dialysis was used to form MDI-based polymers. Among HDI-based polymers, HDI-HEP was not soluble in DMF. All three HDI-based polymers could well be dispersed in HCl solution. In order to use the same nanoparticle preparation method, HCl solution was used for preparation of HDI-based nanoparticles. The pH-dependent transitions of polyurethane nanoparticles in aqueous solution were investigated by transmittance measurements at 410 nm. As shown in Fig. 3a, at the beginning of measurement, the solutions of HDIbased polymers were all transparent because the ionized groups prevented hydrophobic segment from aggregating. Specifically, all the nanoparticles became opaque when the pH values were higher than their pKa values, resulting in a significant decrease in transmittance rate. In addition, the transition pH value was strongly affected by the polymer composition. HDI-HEP and HDIMDEA exhibited similar transition pH value regardless of their different hydrophilic segment in the architecture of their backbones, indicating that HEP and MDEA have similar influence on ionization/deionization and hydrophobic/hydrophilic balance of polyurethanes. The existence of butyl side chains in HDI-BDEA increased the hydrophobic/hydrophilic ratio of polyurethanes, which made the nanoparticles readily aggregate. Thus, the transition pH value was lower than those of HDI-HEP and HDI-MDEA. The pH-responsive properties were also checked for MDI-based polyurethane nanoparticles. The transmittance did not change versus pH due to the strong hydrophobic instinct of MDI segments (data not shown). The mean z-average diameter of the nanoparticles was determined by DLS. In terms of HDI-based polyurethanes, there was no significant increase of diameter at pH value below pKa , as shown in

A. Wang et al. / International Journal of Pharmaceutics 441 (2013) 30–39

35

Fig. 3. UV transmittance of HDI-based polyurethanes as a function of pH (a) and temperature (b).

Fig. 4a. As pH value was higher than pKa , a sharp increase in particle size occurred. The z-average diameter enlarged 3–4 times as compared with those at lower pH. The ratio of the hydrophobic part to the hydrophilic part in polyurethane backbone was larger due to the decrease in ionization degree (Filippov et al., 2008). Further increased pH will induce precipitation, and the quantity of the suspensions became poor according to DLS analysis. Meanwhile, size curves clearly showed the transition pH values in good agreement with UV/vis measurements. As for the MDI-based polyurethanes, the influence of pH values upon average nanoparticle diameter was tremendous (Fig. 4b). When the medium pH was higher than the transition pH value, the nanoparticles exhibited an intense aggregation, resulting in the dramatically enlargement of particle size to 2.5–5.5 ␮m. The ratio of hydrophobic/hydrophilic part in MDIbased polyurethanes was higher than HDI-based ones, so the aggregation extent of MDI-based polymer nanoparticles was much greater than the latter. All these MDI-based polyurethanes had an approximate transition pH value of 4.5. We can infer that the influence of hydrophilic segments on hydrophobic/hydrophilic ratio was neglectable due to the relative high hydrophobicity of the aromatic ring and highly symmetrical chemical structure

in hard segment. Further increased pH will induce precipitation. 3.5. Temperature-responsive property of polyurethanes We investigated the temperature-responsibility of different polymers at different pH and found that only at critical pH values slightly lower than pKa , they showed temperature sensitivity. Due to the different pKa values of polymers, the temperature sensitivity was performed at pH 6.4, 6.6 and 5.4 for HDI-HEP, HDI-MDEA and HDI-BDEA. Fig. 3b shows that temperature-responsive aggregation of HDI-based polyurethane was strongly affected by chemical composition. The transmittance of HDI-HEP and HDI-MDEA polymer suspensions decreased during heating process while the transition of HDI-BDEA was feeble. The existence of butyl side chains in HDI-BDEA probably impeded the rearrangement of molecular packing upon heating. The aggregation of polyurethanes is aroused by the hydrophobic interactions between hard segments. When the hydration layer on hydrophobic surface was destroyed, the van der Waals’ force between exposed hydrophobic segments increased and played a dominant role in the aggregation of nanoparticles (Chuang et al., 2009). Fig. 5 shows the size of nanoparticles at

Fig. 4. Mean diameter of HDI/MDI-based polymers at different pH values as measured by DLS.

36

A. Wang et al. / International Journal of Pharmaceutics 441 (2013) 30–39

Fig. 5. Mean diameter and PDI of HDI-based polymers at different temperatures measured by DLS.

different temperatures. The average diameter of HDI-HEP and HDIMDEA nano-assemblies changed from 230–281 nm to 859–995 nm with temperature increased from 20 ◦ C to 55 ◦ C, while the size of HDI-BDEA nanoparticles was observed to be stable in accordance with the results of UV/vis measurements. The PDI of HDI-based polyurethanes were 0.12–0.22 at room temperature. When the temperature reached 55 ◦ C, the PDI of both HDI-HEP and HDIMDEA increased significantly to 1.0, and the PDI of HDI-BDEA changed slightly. However, MDI-based polyurethane nanoparticles did not exhibit thermo-responsive behavior (data not shown), because of their high hydrophobic/hydrophilic balance in the backbone. 3.6. 1 H NMR characterization of polyurethanes under different conditions In order to understand the environmental-responsive mechanism of HDI-based polyurethane, 1 H NMR of HDI-HEP was recorded at different pH values and temperatures. The influence of pH on the chemical structure and self-assembly of HDI-HEP was shown

Fig. 6.

1

in Fig. 6a. The proton signals of HDI-HEP are different in CDCl3 and D2 O due to the salvation effect. At pH 4.0, the chemical shifts of protons in the hydrophilic chain (peaks d, e and f) of HDI-HEP in D2 O moved to lower field compared with those in D2 O at pH 6.4, which probably due to the electron receptor effect of the protonated amino groups in D2 O. The resolution of proton signal (peaks e and f) reduced, indicating that the appearance of phase separation around pH 6.4 resulted in the suppressing movements of polymer chain (Han et al., 2003; Gao et al., 2005). The temperature responsibility of HDI-HEP was also investigated (Fig. 6b). All the peaks were visible at temperatures from 25 ◦ C to 55 ◦ C, and moved to lower field with an increased temperature, indicating that the gradual breakdown of polymer–water hydrogen bonding upon increasing temperature induced thermally independent water solubility of the polymers. 3.7. SEM analysis The morphology of polyurethane HDI-MDEA under different conditions was investigated by SEM (Fig. 7). All the

H NMR spectra of HDI-HEP polyurethane (a) at 20 ◦ C and different pH values and (b) at pH 6.4 and different temperatures.

A. Wang et al. / International Journal of Pharmaceutics 441 (2013) 30–39

37

Fig. 7. SEM images of HDI-MDEA at (a) 25 ◦ C, pH 6.6, (b) 25 ◦ C, pH 7.4 and (c) 45 ◦ C, pH 6.6.

3.8. In vitro cytotoxic activity Cytotoxicity is a major barrier in vitro and/or in vivo application. Ideal biomaterials should not release toxic substances or produce adverse reactions in vivo (Nan et al., 2011). The cytotoxicity of two series of polyurethanes was evaluated by MTT assay against HUVEC cells. Dose–response curves for the preparations of biomaterials are shown in Fig. 8, presenting that HDI-based polyurethanes exhibited lower toxicity than MDI-based polyurethanes. About 100% of cell viability was observed for HDI-MDEA even at a higher dose (250 ␮g/mL), indicating that it can be regarded as safe drug delivery carrier and is promising for controlled drug delivery. On the contrary, MDI-MDEA exhibited higher toxicity against HUVEC cells, maybe due to the presence of toxic aromatic ring ingredients. Therefore, aromatic polyurethane compounds can be potentially useful for the fabrication of extracorporeal medical devices instead of intracorporeal applications.

3.9. DOX incorporation and release of polyurethane nanoparticles Fig. 8. The cytotoxicity assays of polyurethane against HUVEC cells.

nano-assemblies show near-spherical shape. For the samples prepared at room temperature, the mean diameter of HDI-MDEA was about 260 nm at pH 6.6, and 500 nm at pH 7.4. While for the samples prepared at 45 ◦ C and pH 6.6, the mean diameter was about 750 nm. The SEM observation was in good agreement with DLS results.

Doxorubicin was used as a model anti-cancer drug to examine the release behavior of polymeric nanoparticles at different pH values and temperatures. HDI-MDEA and HDI-BDEA were chosen as drug carriers, due to their high biocompatibility and good solubility in DMF. DOX was successfully encapsulated into HDI-MDEA and HDI-BDEA. The characteristics of drug-loaded/blank nanoparticles of the polymers are listed in Table 2. The hydrophobic interaction of BDEA and DOX attributed to the enhancement of EE and LC. Therefore, the EE and LC of HDI-BDEA were higher than that of HDI-MDEA. In

Fig. 9. Drug (DOX) release from polyurethane nanoparticles.

38

A. Wang et al. / International Journal of Pharmaceutics 441 (2013) 30–39

Table 2 Characteristics of polyurethanes nanoparticles. Polymers

Diameter/nm

HDI-MDEAa HDI-BDEAa HDI-MDEAb HDI-BDEAb

195 172 481 390

a b

LC (%)

39.8% 43%

EE (%)

79.6% 86%

Before DOX encapsulation. After DOX encapsulation.

addition, there is a significant increase in diameter after drug loading. Fig. 9 shows the cumulative release profiles of DOX from drug-loaded nanoparticles under different conditions. Basically, HDI-BDEA-based biomaterial released drug slowly, compared with HDI-MDEA under the same situation because the butyl side chain increased the stability of polymer aggregation. The in vitro release rates of DOX were evaluated at different pH values (pH 4.0 and 7.4) and temperatures (37 ◦ C and 50 ◦ C). As can be seen from Fig. 9a, the drug release rate of all the nanoparticles was obviously pHdependent. The cumulative release of DOX from HDI-MDEA and HDI-BDEA could reach up to 75% and 39% after 60 h at pH 4.0, higher than that at pH 7.4, which was 9% and 7%, respectively. The differences in drug release behaviors under different pH were mainly attributed to the change of the protonation extent of amino groups. Apparently, nanoparticle formulation exhibited rapid release at the initial 15 h at pH 4.0, the repulsion between ionized amino groups at lower pH led to looser nanoparticles, making the drug diffuse out of the complexes much easier. Whereas at pH 7.4, amino groups were deprotonated and the complexes tended to be more compact which is not favorable for drug diffusion. Therefore, DOX-loaded HDIbased nanoparticles exhibited a pH sensitive drug release behavior. It is well known that the extracellular pH of tumors is slightly lower than that of blood and normal tissue (Ma et al., 2012). Therefore, the accelerated drug release at lower pH can be potentially useful for tumor-targeted drug delivery. Fig. 9b shows the DOX release profiles from polymeric nanoparticles under different temperatures. The temperature did not affect the release profile of DOX from HDI-BDEA because HDI-BDEA nanoparticles were not temperature-responsive (Fig. 3b). For HDIMDEA, DOX release was selectively accelerated upon heating; the cumulative release could reach up to 30% at 50 ◦ C compared to 16.5% at 37 ◦ C. This was probably because the interspace between the polymer chains increased after polymer shrinkage due to temperature increase, so that the nanoparticles were easily attacked and the drug release rate increased (Shen et al., 2008). The intracellular distribution of DOX in Huh-7 cells was followed by CLSM. Small amount of DOX fluorescence was observed in Huh7 cells following 6 h incubation (Fig. 10a), while significant DOX fluorescence was observed with the longer incubation time of 24 h that was distributed to the whole cells of Huh-7 cells (Fig. 10b). It should be noted that HDI-MDEA-DOX complexes nanoparticles displayed a faster approach to the cells than free DOX under similar conditions (Fig. 10c) due to the electrostatic interaction between the carrier and the negatively charged cell membrane (Gao et al., 2012). DOX, used widely in the treatment of different types of tumors, is known to exert drug effects via intercalation with DNA and inhibition of macromolecular biosynthesis (Zhan et al., 2011). However, the cytotoxicity of DOX cannot be ignored. As can be seen in Fig. 10b, only 43% of cell remained alive for free DOX. After encapsulated into polyurethane nanoparticles, even though the antitumor activity was reduced, the cell viability was greatly improved.

Fig. 10. CLSM images of Huh-7 cells incubated with HDI-MDEA-DOX (DOX concentration: 40 ␮g/mL) for (a) 6 h, (b) 24 h, and (c) free DOX, 24 h incubation, for each panel, images from left to right show DOX fluorescence in cells (red), cell nuclei stained by Hoechst 33342 (blue), and overlays of two images, the scale bars correspond to 20 ␮m in all images, and (d) the cytotoxicity of DOX with/without HDI-MDEA. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

4. Conclusions In the present work, research efforts have been focused on the preparation and material property studies of polyurethanes and their nanoparticles. Polyurethane nanoparticles were found to be environmentally responsive, in which their particle sizes and transmittance varied upon changing the pH value of surrounding media. Nanoparticles based on HDI-HEP and HDI-MDEA showed temperature-responsive behavior, while MDI-based biomaterials were not temperature-sensitive, due to their high hydrophobic/hydrophilic ratio in the polymer backbone. Cytotoxicity analysis indicated that HDI-based polymers possessed

A. Wang et al. / International Journal of Pharmaceutics 441 (2013) 30–39

negligible toxicity while the cytotoxicity of MDI-based polymers was high. High LC and EE of DOX were obtained from HDI-BDEA and HDI-MDEA nanoparticles prepared by dialyzing method. Both polyurethanes exhibited faster release profile at low pH values. In addition, HDI-MDEA showed an accelerated release at high temperature. Encapsulation of DOX into HDI-MDEA could improve the cell viability and cellular uptake. The pH/temperature responsive nanoparticles fabricated from biocompatible polyurethanes provide an opportunity for the controlled release of DOX. Acknowledgments Financial support from NSFC (21074092, 21272093, 21244004), Program for New Century Excellent Talents in University (NCET-11-1063), Tianjin Municipal Natural Science Foundation (10JCYBJC26800), Foundation of Tianjin Educational Committee (20090505), Studying Abroad Program of Tianjin Municipal Education Commission for Prominent Young College Teachers, and Open Project of State Key Laboratory of Supramolecular Structure and Materials (SKLSSM201209) is highly acknowledged. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ijpharm. 2012.12.021. References Akiyoshi, K., Deguchi, S., Tajima, H., Nishikawa, T., Sunamoto, J., 1997. Microscopic structure and thermoresponsiveness of a hydrogel nanoparticle by self-assembly of a hydrophobized polysaccharide. Macromolecules 30, 857–861. Aryal, S., Hu, C.M.J., Zhang, L.F., 2010. Polymer-cisplatin conjugate nanoparticles for acid-responsive drug delivery. ACS Nano 4, 251–258. Bae, Y., Fukushima, S., Harada, A., Kataoka, K., 2003. Design of environment-sensitive supramolecular assemblies for intracellular drug delivery: polymeric micelles that are responsive to intracellular pH change. Angew. Chem. Int. Ed. 42, 4640–4643. Bae, Y., Jang, W.D., Nishiyama, N., Fukushima, S., Kataoka, K., 2005a. Multifunctional polymeric micelles with folate-mediated cancer cell targeting and pH-triggered drug releasing properties for active intracellular drug delivery. Mol. Biosyst. 1, 242–250. Bae, Y., Nishiyama, N., Fukushima, S., Koyama, H., Yasuhiro, M., Kataoka, K., 2005b. Preparation and biological characterization of polymeric micelle drug carriers with intracellular pH-triggered drug release property: tumor permeability, controlled subcellular drug distribution, and enhanced in vivo antitumor efficacy. Bioconjug. Chem. 16, 122–130. Chen, H.Y., Gu, Y.Q., Hu, Y.Z., Qian, Z.Y., 2007. Characterization of pH- and temperature-sensitive hydrogel nanoparticles for controlled drug release, PDA. J. Pharm. Sci. Technol. 61, 303–313. Chen, K.Y., Kuo, J.F., Chen, C.Y., 2000. Synthesis, characterization and platelet adhesion studies of novel ion-containing aliphatic polyurethanes. Biomaterials 21, 161–171. Chuang, C.Y., Don, T.M., Chiu, W.Y., 2009. Synthesis and properties of chitosan-based thermo- and pH-responsive nanoparticles and application in drug release. J. Polym. Sci. Part A: Polym. Chem. 47, 2798–2810. Dayananda, K., He, C., Lee, D.S., 2008. In situ gelling aqueous solutions of pH- and temperature-sensitive poly(ester amino urethane)s. Polymer 49, 4620–4625. Ding, M.M., Li, J.H., Fu, X.T., Zhou, J., Tan, H., Gu, Q., Fu, Q., 2009. Synthesis, degradation, and cytotoxicity of multiblock poly(epsilon-caprolactone urethane)s containing gemini quaternary ammonium cationic groups. Biomacromolecules 10, 2857–2865. Engler, A.C., Bonner, D.K., Buss, H.G., Cheung, E.Y., Hammond, P.T., 2011. The synthetic tuning of clickable pH responsive cationic polypeptides and block copolypeptides. Soft Matter 7, 5627–5637.

39

ˇ Macková, H., Spírková, ˇ ˇ epánek, P., 2008. Novel ´ M., Konák, ˇ Filippov, S., Hruby, C., M., Stˇ pH-responsive nanoparticles. Langmuir 24, 9295–9301. Fu, H.G., Gao, H., Wu, G.L., Wang, Y.N., Fan, Y.G., Ma, J.B., 2011. Preparation and tunable temperature sensitivity of biodegradable polyurethane nanoassemblies from diisocyanate and poly(ethylene glycol). Soft Matter 7, 3546–3552. Gao, H., Wang, Y.N., Fan, Y.G., Ma, J.B., 2005. Synthesis of a biodegradable tadpole-shaped polymer via the coupling reaction of polylactide onto mono(6(2-aminoethyl)amino-6-deoxy)-ˇ-cyclodextrin and its properties as the new carrier of protein delivery system. J. Control. Release 107, 158–173. Gao, Y.H., Yang, C.H., Liu, X., Ma, R.J., Kong, D.L., Shi, L.Q., 2012. A multifunctional nanocarrier based on nanogated mesoporous silica for enhanced tumor-specific uptake and intracellular delivery. Macromol. Biosci. 12, 251–259. Gavini, E., Mariani, A., Rassu, G., Bidali, S., Spada, G., Bonferoni, M.C., Giunchedi, P., 2009. Frontal polymerization as a new method for developing drug controlled release systems (DCRS) based on polyacrylamide. Eur. Polym. J. 45, 690–699. Haining, W.N., Anderson, D.G., Little, S.R., von Bergwelt-Baildon, M.S., Cardoso, A.A., Alves, P., Kosmatopoulos, K., Nadler, L.M., Langer, R., Kohane, D.S., 2004. pHtriggered microparticles for peptide vaccination. J. Immunol. 173, 2578–2585. Han, S., Hagiwara, M., Ishizone, T., 2003. Synthesis of thermally sensitive watersoluble polymethacrylates by living anionic polymerizations of oligo(ethylene glycol) methyl ether methacrylates. Macromolecules 36, 8312–8319. He, C.L., Kim, S.W., Lee, D.S., 2008. In situ gelling stimuli-sensitive block copolymer hydrogels for drug delivery. J. Control. Release 127, 189–207. Huynh, T.T.N., Padois, K., Sonvico, F., Rossi, A., Zani, F., Pirot, F., Doury, J., Falson, F., 2010. Characterization of a polyurethane-based controlled release system for local delivery of chlorhexidine diacetate. Eur. J. Pharm. Biopharm. 74, 255–264. Kang, S.I., Na, K., Bae, Y.H., 2003. Physicochemical characteristics and doxorubicinrelease behaviors of pH/temperature-sensitive polymeric nanoparticles. Colloid Surf. A 231, 103–112. Kim, Y.H., Bae, Y.H., Kim, S.W., 1994. pH/temperature-sensitive polymers for macromolecular drug loading and release. J. Control. Release 28, 143–152. Lan, P.N., Corneillie, S., Schacht, E., Davies, M., Shard, A., 1996. Synthesis and characterization of segmented polyurethanes based on amphiphilic polyether diols. Biomaterials 17, 2273–2280. Lee, Y., Fukushima, S., Bae, Y., Hiki, S., Ishii, T., Kataoka, K., 2007. A protein nanocarrier from charge-conversion polymer in response to endosomal pH. J. Am. Chem. Soc. 129, 5362–5363. Loh, X.J., Sng, K.B.C., Li, J., 2008. Synthesis and water-swelling of thermo-responsive poly(ester urethane)s containing poly(ε-caprolactone), poly(ethylene glycol) and poly(propylene glycol). Biomaterials 29, 3185–3194. Ma, Y.N., Gao, H., Gu, W.X., Yang, Y.W., Wang, Y.N., Fan, Y.G., Wu, G.L., Ma, J.B., 2012. Carboxylated poly(glycerol methacrylate)s for doxorubicin delivery. Eur. J. Pharm. Sci. 45, 65–72. Nan, Y.F., Luo, Y.L., Xu, F., Chen, Y.S., Di, H.W., 2011. Synthesis and characterization of novel h-HTBN/PEG PU copolymers: effect of surface properties on hemocompatibility. Polym. Adv. Technol. 22, 802–810. Poon, Z., Chang, D., Zhao, X., Hammond, P.T., 2011. Layer-by-layer nanoparticles with a pH-sheddable layer for in vivo targeting of tumor hypoxia. ACS Nano 33, 4284–4292. Qiu, X.P., Kwan, C.M.S., Wu, C., 1997. Laser light scattering study of the formation and structure of poly(N-isopropylacrylamide-co-acrylic acid) nanoparticles. Macromolecules 30, 6090–6094. Shah, M., Ullah, N., Choi, M.H., Kim, M.O., Yoon, S.C., 2012. Amorphous amphiphilic P(3HV-co-4HB)-b-mPEG block copolymer synthesized from bacterial copolyester via melt transesterification: nanoparticle preparation, cisplatin-loading for cancer therapy and in vitro evaluation. Eur. J. Pharm. Biopharm. 80, 518–527. Shen, Z.Y., Ma, G.H., Dobashi, T., Marki, Y., Su, Z.G., 2008. Preparation and characterization of thermo-responsive albumin nanospheres. Int. J. Pharm. 346, 133–142. Sun, X.K., Gao, H., Wu, G.L., Wang, Y.N., Fan, Y.G., Ma, J.B., 2011. Biodegradable and temperature-responsive polyurethanes for adriamycin delivery. Int. J. Pharm. 412, 52–58. Wang, C., Ge, Q., Ting, D., Nguyen, D., Shen, H.R., Chen, J.Z., Eisen, H.N., Heller, J., Langer, R., Putnam, D., 2004. Molecularly engineered poly(ortho ester) microspheres for enhanced delivery of DNA vaccines. Nat. Mater. 3, 190–196. Zhan, F.X., Chen, W., Wang, Z.J., Lu, W.T., Cheng, R., Deng, C., Meng, F.H., Liu, H.Y., Zhong, Z.Y., 2011. Acid-activatable prodrug nanogels for efficient intracellular doxorubicin release. Biomacromolecules 12, 3612–3620. Zhang, C.H., Zhao, K.J., Hu, T.Y., Cui, X.F., Brown, N., Boland, T., 2008. Loading dependent swelling and release properties of novel biodegradable, elastic and environmental stimuli-sensitive polyurethanes. J. Control. Release 131, 128–136. Zhou, L.J., Liang, D., He, X.L., Li, J.H., Tan, H., Li, J.S., Fu, Q., Gu, Q., 2012. The degradation and biocompatibility of pH-sensitive biodegradable polyurethanes for intracellular multifunctional antitumor drug delivery. Biomaterials 33, 2734–2745.