Nanoaggregates of Biodegradable Amphiphilic Random Polycations

Mar 29, 2012 - ABSTRACT: Cationic amphiphilic random copolyesters were obtained by .... changes. All tests were conducted in accordance with previousl...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/Biomac

Nanoaggregates of Biodegradable Amphiphilic Random Polycations for Delivering Water-Insoluble Drugs Benjamin Nottelet,* Manuela Patterer, Benjamin François, Marc-Alexandre Schott, Martine Domurado, Xavier Garric, Dominique Domurado, and Jean Coudane Max Mousseron Institute of Biomolecules (IBMM), Artificial Biopolymers Group, UMR CNRS 5247 University of Montpellier 1, University of Montpellier 2, Faculty of Pharmacy, 15 Av. C. Flahault, Montpellier, 34093, France S Supporting Information *

ABSTRACT: Cationic amphiphilic random copolyesters were obtained by copolymerization of 5-Z-amino-δ-valerolactone and ε-caprolactone. The amino content of the final copolymers was controlled by the polymerization feed ratio and was in the range 10 to 100%. Copolymers solubility and aggregation behavior was assessed by conductometric and zeta potential analyses. A critical aggregation concentration of ca. 0.05% (w/v) was found for all water-soluble copolymers that formed nanoaggregates. Two populations were found to be present in equilibrium with hydrodynamic diameters in the range of 30−50 and 100−250 nm. The capacity to use the amphiphilic and cationic character of the nanoaggregates to encapsulate highly hydrophobic compounds was further investigated. Finally, copolymers hemo- and cytocompatibility were evaluated by hemagglutination, hemolysis, and cells proliferation tests. The results showed that the proposed cationic amphiphilic random copolyesters are biocompatible.



INTRODUCTION Given that almost one-third of newly discovered drugs are highly insoluble in water, more compounds with poor aqueous solubility have entered the development pipeline.1,2 A number of drug delivery systems, including molecular complexation systems based on cyclodextrins, nanosuspensions, lipid formulations, prodrugs, and polymeric micellar systems, have been developed to overcome this solubility limitation.3 Of all of these different systems, polymeric micelles have attracted the greatest interest as promising novel drug delivery carriers because of their remarkable advantages that include small size and a narrow size distribution, long circulation times, good stability, and targeting ability. Polymeric micelles are macromolecular assemblies composed of an inner core and an outer shell. They can be divided in two main categories depending on the driving force responsible for their formation: hydrophobic micelles and polyion complex (PIC) micelles.4 The former, that have undergone the most development, usually consist of amphiphilic block copolymers. Typically, micellar drug carriers involve AB- or ABA-type block copolymers, which spontaneously form nanosized objects in aqueous medium as a result of the hydrophobic and hydrophilic blocks interacting differently with the solvent. Given the possibility to load lipophilic © 2012 American Chemical Society

molecules into their hydrophobic core through both chemical conjugation and physical entrapment, polymer micelles have been widely used to solubilize and deliver poorly water-soluble drugs. The various polymers proposed for the formation of block polymeric micelles include the pharmaceutically most promising poly(ethylene glycol)−poly(lactide) (PEG-b-PLA) and poly(ethylene glycol)−poly(lactide-glycolide) (PEG-bPLGA) diblock copolymers that have been extensively studied because of their biocompatibility and the biodegradability.5−8 However, some basic micelle properties can advantageously be improved by using alternative polymer blocks. For example, degradability and drug loading can be improved by using poly(malic acid) and poly(hexyl-lactic acid) blocks, and micelle stability can be increased by using graft structures instead of block copolymers.9−11 Whereas amphiphilic block copolymers are used to form hydrophobic micelles, charged polymer blocks may also be used in micelles. PIC micelles result from electrostatic interactions between ionic segments and oppositely charged species; double Received: February 16, 2012 Revised: March 28, 2012 Published: March 29, 2012 1544

dx.doi.org/10.1021/bm300251j | Biomacromolecules 2012, 13, 1544−1553

Biomacromolecules

Article

(BzOH), ethyl acetate (AcOEt), tetrahydrofuran (THF), toluene, dimethylformamide (DMF), diethyl ether (Et2O), dichloromethane (CH2Cl2), poly-L-lysine hydrobromide (PLL) (Mw = 4000−15 000 g/ mol and Mw = 80 000 g/mol), 17α-ethinylestradiol (EE), flufenamic acid (FA), pyrene, 1-pyrenecarboxylic acid, and 3-(4,5-dimethyl-2thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) were purchased from Sigma-Aldrich (St-Quentin Fallavier, France). Benzotriazol-1-yloxytris(dimethylamino)phosphonium hexafluorophosphate (BOP) and tris(hydroxymethyl)aminomethane (Tris) were purchased from Merck (Strasbourg, France). Modified Eagle’s medium (MEM), horse serum, penicillin, streptomycin, Glutamax, and Dulbecco’s phosphate-buffered saline (DBPS) were purchased from Invitrogen (Cergy Pontoise, France). Cellstar polystyrene tissue culture plates (TCPS) were purchased from Greiner Bio-One (Courtaboeuf, France). All chemicals and solvents were used without purification except for BzOH, ε-CL, and the polymerization solvents that were dried over calcium hydride for 24 h at room temperature and distilled under reduced pressure. Buffers. pH 7.4 phosphate-buffered and 2-amino-2(hydroxymethyl)propane-1,3-diol-buffered isoosmolar media were prepared as previously described34 and are referred to as PBS and Tris-buffer, respectively, in the following. NMR Spectroscopy. 1H NMR and 13C spectra were recorded on an AMX300 Brücker spectrometer operating at 300 and 75 MHz, respectively. Deuterated dimethyl sulfoxide or chloroform was used as solvent. Size Exclusion Chromatography. Polymer molecular weights were determined by size exclusion chromatography (SEC) on a Waters 600 controller system fitted with a Waters In-Line degasser and two PLgel 5 μm MIXED-C (300 × 7.5 mm) columns. A Waters 2410 RI detector and a Waters 2489 UV/vis detector were used. The mobile phase was DMF at 1 mL/min flow and 80 °C. Typically, the polymer (10 mg) was dissolved in DMF (2 mL), the resulting solution was filtered through a 0.45 μm Millipore filter, and 20 μL of filtrate was injected. M̅ n and M̅ w were expressed according to calibration using poly(methyl methacrylate) standards. Differential Scanning Calorimetry. Differential scanning calorimetry (DSC) measurements were made under nitrogen at a 10 °C/ min heating rate on a Perkin-Elmer DSC 6000 thermal analyzer. Dynamic Light Scattering. Polymer solutions were prepared in Milli-Q water and filtered through a 0.45 μm filter. Nanoaggregate hydrodynamic diameter (), zeta potential, and solutions conductivity were evaluated by laser doppler velocytometry (NanoZS, Malvern Instrument, U.K.). Measurements were made at 25 °C using a 633 nm laser. Zeta potential (ζ) was calculated from the electrophoretic mobility value, using Smoluchowski’s equation: ζ = (4πηu)/ε, where u is electrophoretic mobility, η is viscosity, and ε is the dielectric constant of the solvent. Microscopy. Light microscopy examinations (40× magnification) were carried out using a TE300 microscope (Nikon, Tokyo, Japan) fitted with a Canon Power Shot A620 camera. Photomicrographs were taken from two observation fields. The microscope stage was at room temperature. Transmission electron microscope (TEM) micrographs were obtained with a JEOL 1200 EXII (working voltage of 120 kV). A drop of the solution was placed onto a carbon-supported copper grid for 5 min. The excess liquid was sucked away by filter paper. Encapsulation of Hydrophobic Compounds. The encapsulation capability of the different copolymers was evaluated by adding hydrophobic compounds (HCs) to a 1% (w/v) copolymer aqueous solution. In a typical experiment, 3 mg of HC was added to 3 mL of a 1% (w/v) copolymer solution in distilled water. The mixture was stirred for 3 h before centrifugation. The supernatant was then carefully collected before dilution at 50% with acetone or ethanol (depending on the HC) to release the HC entrapped in the copolymer hydrophobic domains. The resulting solutions were analyzed on a Perkin-Elmer Lambda 35 UV/vis spectrophotometer. The total amount of dissolved HC was calculated by comparing the resulting absorbance with standard plots. The standard plot for each HC was constructed as the absorbance measured at the isosbestic point of the spectra obtained for decreasing concentrations of HC. HC chemical

hydrophilic copolymers composed of a polyionic segment and a hydrophilic segment are therefore generally used.4 Interpolyelectrolyte complexes (IPECs) are another example; they form spontaneously in solution because of strong cooperative electrostatic interactions between positively and negatively charged polyions.12 IPECs are, for example, developed for gene delivery applications, where they are known as polyplexes based on the complexation of DNA with various polycations.13−16 In addition to these purely hydrophilic compounds, amphiphilic copolyelectrolytes undergo hydrophobic associations that compete with electrostatic repulsions to form various types of micelle-like nanostructures.17−20 Although these systems have so far drawn little attention, they are of particular interest: nanosized aggregates formed by random amphiphilic polyelectrolytes can serve as carriers and are an interesting alternative to hydrophobic micelles and polyionic complexes for drug delivery.21 Morishima et al. were among the first to conduct systematic studies on random anionic copolymers with amphiphilic character and the effects of various parameters, including charge density, hydrophobicity, charge location, and molecular weight on their self-organization.22−24 Our group recently capitalized on the advantage of the hydrophobic nanodomains formed by such systems to encapsulate clofazimine in hydrophobized poly(methyl vinyl ether-altmaleic acid) (PMVEMAc). We demonstrated that this polymeric system is able to promote the solubilization of clofazimine at blood pH and that solubilization results from drug−polymer interactions that combine electrostatic and hydrophobic interactions. Synergy between these two physical phenomena resulted in the entrapment of large amounts of water-insoluble clofazimine with up to one molecule of drug for two mononeric units, corresponding to 1 g of drug for 1 g of polymer.21 We were interested in broadening this strategy to random amphiphilic polycations while at the same time adding degradability to the polymer system. Because aliphatic polyesters are well-known for their degradability, we focused our efforts on amino-functionalized lactones with the intention to generate a family of cationic copolymers. Only a few examples of such monomers can be found in the literature, mainly serine-based butyrolactones and one lysine-based lactide.25−29 More recently, Lang’s group proposed an aminated caprolactone that was polymerized with PEG to generate double hydrophilic block copolymers.30,31 Our group also reported the synthesis of 5-Z-amino-δ-valerolactone (5-NHZVL), which is easily obtained in two steps from a glutamic acid derivative and can be polymerized to yield cationic aliphatic polyesters.32 In the work reported herein, 5-NHZ-VL and εcaprolactone (ε-CL) were copolymerized to generate a family of random amphiphilic polycations with a predictable amino content. Their ability to self-assemble as nanoaggregates and generate hydrophobic domains was evaluated. Finally, poorly water-soluble model compounds with various log P and log D values were encapsulated in the copolymers to evaluate their potential for future drug delivery applications.



EXPERIMENTAL SECTION

Materials. N-Benzyloxycarbonyl-glutamic acid γ-tert-butyl ester (ZGlu(OtBu)OH, 99%), N-N′-diisopropylethylamine (DIPEA), sodium borohydride powder (NaBH4, 99%), sodium bicarbonate (NaHCO3), magnesium sulfate (MgSO4), ε-CL, tin(II) 2-ethylhexanoate (Sn(Oct)2, ∼95%), trifluoroacetic acid (TFA), solution of hydrobromic acid in acetic acid (HBr/AcOH, 33 wt %), benzyl alcohol 1545

dx.doi.org/10.1021/bm300251j | Biomacromolecules 2012, 13, 1544−1553

Biomacromolecules

Article

Table 1. Characteristics of Synthesized Poly(5-Z-amino-δ-valerolactone-co-ε-caprolactone) and Poly(5-NH3+-δ-valerolactoneco-ε-caprolactone) composition NH-VL/ε-CL (%)

thermal properties (°C)a

molecular weight (g/mol)

copolymer

f NHZVLb

FNHZVLc

FNH3+d

M̅ w before dep.

M̅ w/M̅ n

M̅ n after dep.

M̅ w/M̅ n

Tm

Tg

P1 P2 P3 P4 P5

100 75 50 25 10

100 70 40 30 8

100 68 38 28 12

7000 9500 11 000 12 300 8600

1.20 1.15 1.19 1.20 1.20

5600 6200 6000 7500 5000

1.07 1.16 1.12 1.19 1.69

74 none none none

−22 −28 −28 −30

Thermal properties of poly(5-NH3+-δ-valerolactone-co-ε-caprolactone) with Br− as counterion. bCalculated using eq 1. cCalculated using 1H NMR spectra before deprotection. dCalculated using 1H NMR spectra after deprotection. a

formulas, ethanol, and water ratios used for the analyses, isosbestic points, water solubility, log P, and log D values are given in the Supporting Information. Hemocompatibility. Hemocompatibilty was evaluated by assessing whether red blood cells (RBCs) in contact with polymer solutions underwent any hemolysis, hemagglutination, or morphological changes. All tests were conducted in accordance with previously described protocols.33,34 Wistar rats were housed in the Nı̂mes University experimental research center. The drawing of blood samples was accepted by the University’s Animal Research Ethics Committee (CEEA-LR-1010). Polymer solutions were prepared in Tris-buffer at concentrations of 2 and 10 mM with regards to the amino groups. Rat venous blood was drawn onto disodium ethylenediaminetetraacetic acid salt (EDTA) for anticoagulation, and all RBCs were used within 5 h of blood collection. After centrifugation at 1980g, autologous anticoagulated plasma was collected, the buffy coat was removed, and the RBCs were washed three times with Trisbuffer. After the last wash, 0.4 mL of packed RBCs was suspended in 0.6 mL of the selected experimental medium, namely, autologous anticoagulated plasma, or Tris-buffer, to yield a 40% hematocrit. The suspensions were incubated at 37 °C for 10 min, and 0.1 mL aliquots of the different polymer solutions were then added. Controls were prepared in a similar manner with Tris-buffer alone or PLL (Mn = 80 000 g/mol, 10 mM of amino groups) as negative and positive controls, respectively. The final mixtures were further incubated for 15 min at 37 °C. Possible hemagglutination and morphological changes were evaluated by observing the RBCs under a microscope after 15 min of incubation. Possible hemolysis was evaluated by centrifuging the RBC suspensions for 15 min and separating the supernatant. We then prepared 1 % solutions of the supernatant in Tris-buffer, and the hemoglobin content was calculated from spectrophotometer measurements at λ1 = 576 nm, λ2 = 560 nm, and λ3 = 592 nm.35 The samples containing Tris-buffer as medium were measured against Tris-buffer, and the samples containing plasma as medium were measured against a 1% solution of pure plasma in Tris-buffer. The total amount of hemoglobin initially present in the RBC suspension was determined by preparing a 40% hematocrit solution in distilled water and processing this as previously described. After centrifugation, the supernatant was diluted for values to fall within the linear range of the spectrophotometer standard plot. The hemoglobin released (rHb) was defined as the percentage of hemoglobin present in the supernatant compared with the total hemoglobin initially present in the RBC suspension. Cytocompatibility. Mouse L929 fibroblasts (L929) were cultured in MEM containing 10% horse serum, penicillin (100 μg/mL), streptomycin (100 μg/mL), and Glutamax (1%). Polymers were disinfected by dissolution and incubation in ethanol for 24 h. The ethanol was then evaporated off at room temperature, and stock solutions of 1 mg/mL were prepared in culture medium. Other concentrations were obtained by diluting the stock solutions in culture medium. The in vitro cytocompatibility of the polymers was assessed by following the proliferation of L929 fibroblasts grown in 24-well TCPS in culture medium containing polymer concentrations of 1 μg/ mL to1 mg/mL. In all, 3000 cells were seeded for each test, and

proliferation was evaluated after 2, 4, and 7 days by MTT assay. Nonadhesive cells were removed from the TCPS at scheduled time points. The TCPS was washed with DPBS, and an MTT solution (250 μL, 1 mg/mL in DPBS) was added. After 3 h of MTT incubation, the reagent was removed and washed with DPBS, and isopropyl alcohol was added to solubilize the formazan. The formazan was measured at 570 nm on a spectrophotometer multiplate reader (Victor X3 multilabel plate reader, PerkinElmer). All data points and standard deviations were derived from triplicates. Synthesis of Tertiobutyl-4-(benzyloxycarbonylamino)-5-hydroxypentanoate (2). Selective reduction of the main chain carboxylic acid group is based on the formation of an activated ester using BOP as reagent, as previously described in the literature.36 Typically, DIPEA (1.17 mL, 7.1 mmol) and a BOP solution (2.87 g, 6.5 mmol) in 10 mL of THF were slowly added to a stirred suspension of Z-Glu(OtBu)OH 1 (2.02 g, 6 mmol) in 20 mL of THF at room temperature. After 10 min of stirring, NaBH4 (1.1 g, 30 mmol) was slowly added to the reaction medium, which was then stirred for 2 h at room temperature. The resulting mixture was then diluted in CH2Cl2 (150 mL) and washed with a 5% HCl solution (5 × 100 mL), followed by an NaHCO3 saturated solution (3 × 100 mL) and brine (3 × 100 mL). The organic phase was then dried over MgSO4 and concentrated under reduced pressure to yield a clear oil. Final purification of compound 2 was obtained by flash column chromatography using first-step CH2Cl2 to elute byproduct and second-step AcOEt to recover pure 2 with a 90% overall yield. 1H NMR for 2: (300 MHz; CDCl3): δ = 7.3 (m, 5H, Ph), 5.1 (m, 1H, NH), 5.0 (s, 2H, OCH2Ph), 3.8−3.7 (m, 1H, CH-NHZ), 3.6−3.5 (m, 2H, CH2−OH), 2.3−2.2 (m, 2H, CH2−CH2-CH), 1.9−1.7 (m, 2H, CH2-COOtBu), 1.4 (s, 9H, tBu). HPLC for 2: 1.52 min. Calculated monoisotopic mass for 2 (C17H25NO5): 323.17 g/mol; ES-MS (70 eV, m/z) found for 2: 324.3 [M+H]+, 346.3 [M+Na]+. Synthesis of Benzyl 6-Oxotetrahydro-2H-pyran-3-ylcarbamate (5-Z-Amino-δ-valerolactone) (3). Compound 2 was lactonized via simultaneous removal of the tertiobutyl group and cyclization of the resulting intermediate to yield 3. In a typical experiment, 2 (2 g) was dissolved in a mixture of TFA (10 mL) and CH2Cl2 (10 mL) and stirred for 3 h at room temperature. After reaction completion, cold water (80 mL) and CH2Cl2 (80 mL) were added. The organic layer was washed with water (5 × 50 mL), dried over MgSO4, then filtered and concentrated under reduced pressure. Column chromatography with AcOEt/Et2O (5:5) yielded compound 3 (0.85 g, 55%) as white crystals of at least 98% purity. mp 65 °C (from DSC). 1H NMR for 3: (300 MHz; DMSO-d6): δ = 7.6 (m, 1H, NH), 7.4−7.3 (m, 5H, Ph), 5.05 (s, 2H, OCH2Ph), 4.3−4.2 (m, 1Ha, CH2−O), 4.1−4.0 (m, 1Hb, CH2−O), 3.9−3.8 (m, 1H, CH−NHZ), 2.6−2.4 (m, 2H, CH2−CO), 2.1−2.0 (m, 1Ha, CH2−CH2−CH), 1.8− 1.7 (m, 1Hb, CH2−CH2−CH). 13C NMR for 3: (75 MHz, DMSOd6): δ = 171 (s, C(O)O), 156.2 (s, NHC(O)O), 137.5 (s, CH2CCH), 128.8−127.6 (m, CH), 70.6 (s, CH2O), 65.9 (s, CH2C(O)ONH), 44.3 (s, CHNH), 27.5 (s, CH2CHNH), 24.5 (s, CH2C(O)). HPLC for 3: 1.23 min. Calculated monoisotopic mass for 3 (C13H15NO4): 249.10 g/mol; ES-MS (70 eV, m/z) found for 3: 250.2 [M+ + H], 499.4 [2M+ + H]. 1546

dx.doi.org/10.1021/bm300251j | Biomacromolecules 2012, 13, 1544−1553

Biomacromolecules

Article

Scheme 1. Synthesis of 5-Z-Amino-δ-valerolactone, Poly(5-NH3+-δ-valerolactone), and Poly(5-NH3+-δ-valerolactone-co-εcaprolactone)a

a

(i) DIPEA, BOP, THF, 30 min, RT and NaBH4, 2 h, RT; (ii) TFA/CH2Cl2 (1:1), 3 h, RT; (iii) BzOH, Sn(Oct)2, 24h, 110°C; and (iv) CH2Cl2/ HBr/AcOH,