Amphiphilic Triblock Copolymers of Methoxy-poly(ethylene glycol)-b

Dec 9, 2008 - Australian Institute for Bioengineering and Nanotechnology, Centre for ... Australia, Institute of Health and Biomedical Innovation, Que...
0 downloads 0 Views 8MB Size
Download PDF
Biomacromolecules 2009, 10, 95–104

95

Amphiphilic Triblock Copolymers of Methoxy-poly(ethylene glycol)-b-poly(L-lactide)-b-poly(L-lysine) for Enhancement of Osteoblast Attachment and Growth Hui Peng,† Yin Xiao,*,‡ Xueli Mao,‡,§ Lan Chen,† Ross Crawford,‡ and Andrew K. Whittaker*,† Australian Institute for Bioengineering and Nanotechnology, Centre for Magnetic Resonance, University of Queensland, Brisbane QLD 4072, Australia, Institute of Health and Biomedical Innovation, Queensland University of Technology, Brisbane QLD 4059, Australia, and Guanghua College of Stomatology, Sun Yat-sen University, Guangzhou 510055, China Received August 22, 2008; Revised Manuscript Received November 2, 2008

Amphiphilic triblock copolymers of methoxy-poly(ethylene glycol)-poly(L-lactide)-poly(L-lysine) (MPEG-b-PLLAb-PLL) (Mn ) 8540-22 240) were synthesized through the ring-opening polymerization of Nε-(Z)-lysine-Ncarboxyanhydrides (Nε-(Z)-Lys-NCA) using MPEG-b-PLLA-NH2 as a macroinitiator. The triblock copolymers and diblock precursors were characterized by 1H NMR, ATR-FTIR, and GPC. The chain lengths of each block could be controlled by varying the feed ratios of the monomers. The surface properties of films of PLLA modified by blending with the triblock copolymers were investigated by XPS and AFM and demonstrated an enrichment of PLL blocks on the surface of the PLLA film. No cytotoxicity was detected on a range of modified PLLA films arising from the incorporation of the triblock copolymers. The triblock copolymers MPEG-b-PLLA-b-PLL showed better surface properties in promoting osteoblast adhesion and proliferation compared with pure PLLA and PLLA modified with MPEG-b-PLLA diblock copolymers. This study demonstrated that the triblock copolymers containing free amino groups, which self-segregate on the surface of biodegradable polyesters, have potential for applications in cell delivery and tissue engineering.

Introduction Synthetic biodegradable polymers have been increasingly utilized in pharmaceutical, medical, and biomedical engineering over the past two to three decades.1-3 Most of the common synthetic polymers can be produced under controlled conditions and therefore exhibit, in general, predictable and reproducible mechanical and physical properties. However, an important remaining problem is control over interactions between polymer and cells.4,5 Approaches to improve this aspect of the behavior of biomaterials include reduction of nonspecific protein adsorption,6,7 enhancement of adsorption of specific proteins,8,9 surface modification,6,10,11 and immobilization of cell recognition motifs to obtain controlled interactions between cells and synthetic materials.12-17 Among the large range of biodegradable polymers, aliphatic polyesters of R-hydroxyalkanoic acids, such as poly(lactic acid) (PLA), poly(glycolic acid) (PGA), and their copolymers, have attracted much attention because of their biodegradability, biocompatibility, and excellent mechanical properties.18-21 However, polyesters are not especially good at promoting interactions with cells, and because of the lack of functional groups, they cannot be easily modified with biologically active moieties. Many methods have been developed for improving the properties of hydrophobic aliphatic polyesters. Block copolymers containing polyesters, such as PEG-polyester and polyester-poly(amino acid) have been intensively investigated * Corresponding authors. (A.K.W.) Tel: +61-7-33463885. Fax: +61-733463973. E-mail: [email protected]. (Y.X.) Tel: +61-7-31386240. Fax: +61-7-31386030. E-mail: [email protected]. † Centre for Magnetic Resonance. ‡ Queensland University of Technology. § Sun Yat-sen University.

and employed in a number of medical and pharmaceutical applications.22-24 Hydrophilic poly(ethylene glycol) (PEG) blocks have often been introduced to copolymers for their outstanding biological and physicochemical properties. The PEG chains also have been regarded as the most effective structures for preventing undesirable, nonspecific adhesion of proteins and cells, and hence PEG has been proposed for many biomedical and biotechnological applications.25,26 A number of studies have therefore focused on amphiphilic block copolymers of PEG with polyesters with the expectation of achieving unique properties and corresponding applications.22,27-29 PEG polymers are known to have low surface free energy; as such, when blended with other polymers, they will be concentrated at the surface, in particular, in an aqueous environment. Various approaches have been taken to retain the PEG at the surface, such as grafting to or from a substrate. In 1992, Amiji and Park30 demonstrated that absorption of a central hydrophobic poly(propylene oxide) block of a PEG-PPO-PEG triblock copolymer onto a hydrophobic glass surface is sufficient for retaining the copolymer on the surface and in particular for preventing absorption of fibrinogen on that surface. When blended with hydrophobic polymers, block copolymers containing PEG are known to segregate at the surface, for example, as reported by Hancock and colleagues for blends of PEG-b-polysulfone blended with polysufone.31,32 More recently, Ji and coworkers6 used this approach of blending PEG-block copolymers with bulk matrix to decorate the surface of PLLA with peptide motifs attached to the end of PEG-PPOPEG triblock copolymers. In this Article, we likewise exploit the surface-segregating properties of PEG to present a triblock copolymer at the surface of PLLA.

10.1021/bm800937g CCC: $40.75  2009 American Chemical Society Published on Web 12/09/2008

96

Biomacromolecules, Vol. 10, No. 1, 2009

Through the use of ring opening polymerization (ROP) of amino acid N-carboxyanhydrides (NCAs) initiated by primary amines, synthetic poly(R-amino acid)s can be incorporated into copolymer structures.1,23,24,33-38 The inclusion of the functional side groups of amino acids in copolymers can help to improve the affinity of the materials with proteins or cells and provide the possibility of anchoring drugs or DNA to the materials.39 In addition, the presence of peptide bonds in the polymer backbone can modify degradation patterns of the polymers, making them susceptible to degradation by peptidases.39-41 The self-aggregation behavior of the amphiphilic copolymers derived from polyesters and amino acids can provide a specific surface structure and is expected to lead to innovation in biomedical materials and cell delivery systems.24,42-45 Therefore, in this Article, amphiphilic triblock copolymers of methoxy-poly(ethylene glycol)-poly(L-lactide)-poly(L-lysine) (MPEG-b-PLLA-b-PLL) were synthesized via sequential polymerization of PLLA onto MPEG, followed by ROP of PLL on the functionalized chain end. The triblock copolymer was then blended with high-molecular-weight PLLA and cast in the form of films to modulate the initial osteoblast responses. The PLLA segments of the MPEG-b-PLLA-b-PLL copolymers improve the compatibility with the PLLA matrix to form a MPEG-b-PLLA-b-PLL layer on the PLLA surface. The hydrophilic PEG and PLL chains formed an extended surface layer in which PLL can promote cell attachment, spreading, and proliferation.46,47 PLLA surfaces modified with the amphiphilic triblock copolymer are expected to display enhanced cell adhesion and growth, serving as a potential biodegradable scaffold for cell and tissue engineering.

Experimental Section Materials. (3s)-cis-3,6-Dimethyl-1,4-dioxane-2,5-dione (L-lactide) was purchased from Sigma-Aldrich and recrystallized in ethyl acetate three times. Monomethoxy-poly(ethylene glycol) with a molecular weight of 750 (MPEG750) was obtained from Sigma-Aldrich and dried by an azeotropic distillation in toluene. The catalyst, tin(II) ethylhexanoate [Sn(Oct)2], was purchased from Sigma-Aldrich. N-tert-Butoxycarbonyl-L-phenylalanine (BOC-Phe-OH), trifluoroacetic acid (TFA), triphosgene, hydrogen bromide (33 wt % in acetic acid) (HBr/HAc), poly(L-lysine hydrobromide) (PLL-NH3+Br-) (average Mw of 15 000), and poly(L-lactide) (average Mw of 100 000-150 000) were purchased from Sigma-Aldrich and used as received. Dicyclohexylcarbodiimide (DCC), N,N-diisopropylethylamine, and N-epsilon(ε)-carbobenzyloxyε L-lysine (N -(Z)-L-Lys) were obtained from Fluka and were used without further purification. Hexane, methylene chloride, chloroform, and dimethylformamide (DMF) were refluxed over CaH2 and distilled under nitrogen. Tetrahydrofuran (THF) was dried and distilled in the presence of sodium immediately before use. Characterization. ATR-FTIR spectra were acquired on a Nicolet Nexus 870 apparatus with a Smart Endurance diamond ATR accessory. Each spectrum was collected from 32 scans with a resolution of 4 cm-1. Peak height measurements were performed using the spectral analysis software (GRAMS/32, Galactic Industrie, Salem, NH). 1 H spectra of solutions in CDCl3 or DMSO were measured on either a Bruker Avance 500 or 300 MHz spectrometer with a TXI probe at room temperature, and the residual proton signal of the deuterated solvent was used as an internal chemical shift reference. Gel permeation chromatography (GPC) measurements were performed using a Waters Alliance 2690 separations module equipped with an autosampler, column heater, differential refractive index detector, and a photodiode array connected in series. HPLC-grade THF was used as eluent at a flow rate of 1 mL · min-1. The columns consisted of three 7.8 × 300 mm Waters GPC columns connected in series, comprising two linear UltraStyragel columns and one Styragel HR3 column.

Peng et al. Polystyrene standards ranging in molecular weight from 517 to 2 × 106 g · mol-1 were used for calibration. X-ray photoelectron spectroscopy (XPS) was conducted using an Axis Ultra XPS spectrometer (Kratos Analytical). The spectra were obtained at a 90° takeoff angle. Films of virgin PLLA and PLLA modified by blending with MPEG-b-PLLA-b-PLL copolymer were examined. (See Preparation of MPEG-b-PLLA-b-PLL-Modified PLLA Films.) The survey spectra were collected using an analyzer pass energy of 160 eV, and the high-resolution spectra of individual elements were taken with a pass energy of 20 eV and a step increment of 0.1 eV. The spectrometer has a monochromatic Al KR X-ray source operating at 15 kV, 10 mA (150 W) for all data acquisitions. Binding energies were charge-corrected to 285.0 eV for aliphatic carbon. Peak fit results were imported into a graphic software package (Origin, Origin Laboratory) for final illustration. Atomic force microscopy (AFM) experiments were performed using a Multi-Mode V scanning probe microscope (VEECO Instrument) at room temperature in air; the AFM images were obtained in tapping mode. Commercially available etched Si cantilevers with a spring constant of 42 N/m and resonant frequency of 320 kHz were used, and the tip was model RTESP7. Synthesis of Triblock Copolymers of Methoxy-poly(ethylene glycol)-b-poly(L-lactide)-b-poly(L-lysine) (MPEG-b-PLLA-b-PLL). The amino-terminated methoxy-poly(ethylene glycol)-b-poly(L-lactide) (MPEG-b-PLLA-NH2) was prepared according to the procedure reported by Gotsche and coworkers.48 Two different MPEG-b-PLLANH2 polymers were synthesized by adjusting the feed ratio of L-lactide to that of MPEG (degree of polymerization (DP)MPEG ) 17, DPPLLA ) 94; DPMPEG ) 17, DPPLLA ) 56). Nε-(Z)-lysine-N-carboxyanhydride (Nε-(Z)-Lys-NCA) was synthesized and purified according to the method reported by Dorman et al.34 and stored at -20 °C under an argon atmosphere. The expected structure and purity of Nε-(Z)-Lys-NCA were confirmed by 1H NMR and melting point measurements. The triblock copolymerization of methoxy-poly(ethylene glycol)-bpoly(L-lactide)-b-poly(Nε-(Z)-lysine) (MPEG-b-PLLA-b-PZLys) was conducted through the ROP of Nε-(Z)-Lys-NCA initiated by the terminal amino group of MPEG-b-PLLA-NH2; then, the protected groups in the side chain of poly(lysine) were removed to obtain MPEG-b-PLLA-b-PLL with free amino groups. A range of MPEG-b-PLLA-b-PLL triblock copolymers were prepared by adjusting the feed ratio of Nε-(Z)-Lys-NCA to MPEG-b-PLLA diblock copolymers. The following is a typical procedure for the preparation of MPEG17-b-PLLA94-b-PLL109. Nε-(Z)-Lys-NCA (2.0 g) was dissolved in 10 mL of DMF; then, MPEG-b-PLLA-NH2 (0.44 g, DPMPEG ) 17, DPPLLA ) 94) was dissolved in 5 mL of DMF and added to the solution of Nε-(Z)-Lys-NCA. The reaction was carried out at 40 °C in the presence of dry nitrogen. After ∼72 h, the copolymer in the DMF solution was precipitated into excess diethyl ether to give a white product. The MPEGb-PLLA-b-PZLys was dried under vacuum at 40 °C for 24 h. The purified yield was 1.97 g (92.1%). DPMPEG ) 17, DPPLLA ) 94, DPPZLys ) 109. Deprotection of MPEG-b-PLLA-b-PZLys was carried out as follows: Approximately 1.5 g of the protected polymer MPEG-b-PLLA-b-PZLys was placed in a 100 mL round-bottomed flask and purged under argon for 15 min. HBr/HAc solution (10 mL) was added via syringe to the polymer under argon to form a slurry and was stirred for 90 min. The polymer was precipitated into ether, separated by vacuum filtration, and washed several times with ether until the polymer MPEG-b-PLLA-b-PLLNH3+Br- took on an off-white color. The ammonium bromine salt MPEGb-PLLA-b-PLL-NH3+Br- was then redissolved in DMF, neutralized with excess N,N-diisopropylethylamine to remove hydrogen bromide, precipitated from ether, separated by vacuum filtration, and dried under vacuum at 40 °C for at least 48 h to obtain MPEG-b-PLLA-b-PLL with free amino groups in the side chains. The purified yield was 0.75 g (81.5%). DPMPEG ) 17, DPPLLA ) 94, DPPLL ) 115. Preparation of MPEG-b-PLLA-b-PLL-Modified PLLA Films. We carried out the preparation of PLLA films by dissolving highmolecular-weight PLLA (Mw of 100 000-150 000 g/mol) and the triblock copolymers of MPEG-b-PLLA-b-PLL in chloroform (generally

Enhancement of Osteoblast Attachment and Growth

Biomacromolecules, Vol. 10, No. 1, 2009

97

Scheme 1. Synthesis Route and Structure of MPEG-b-PLLA-b-PLL

5 wt % of MPEG-b-PLLA-b-PLL compared with PLLA), casting 2 mL (50 mg/mL) of the solution obtained on glass dishes (35 mm in diameter), and allowing the solvent to evaporate in air. To remove the last traces of chloroform, the dishes were kept under vacuum for 48 h and exposed to UV light for 30 min for sterilization. Films containing the diblock copolymers MPEG-b-PLLA, commercial PLL-NH3+Br-, triblock block copolymer MPEG-b-PLLA-b-PLL-NH3+Br-, and pure PLLA films were prepared in the same way. Cell Testing. Cytotoxicity. The cytotoxicity assay kit of extracellular lactate dehydrogenaze (LDH) (Sigma-Aldrich) was utilized to measure the membrane integrity of cells on a range of modified PLLA films. The round polymer films were removed from the glass dishes, cut into small pieces, and placed in each well of 96-well tissue culture plates; then, 50 µL of culture medium was added to settle the polymer films. The polymer films completely covered the bottom surface of the wells before seeding the cells. Six tests were performed for each polymer film. Human osteoblasts were isolated from alveolar bone, as previously described.49 Briefly, normal human alveolar bone specimens, obtained from consenting healthy young orthodontic patients (13-19 years old) with institutional ethics committee approval, were used as explants for the establishment of cell cultures. Cells (1 × 104) were seeded on each membrane in a 96-well plate and were cultured for 1 day in Dulbecco’s modified Eagle’s medium (DMEM, Invitrogen, Melbourne, Australia) supplemented with 10% fetal calf serum (FCS, HyClone, Logan UT) and 1% penicillin/streptomycin (GIBCO, Invitrogen, Melbourne, Australia) in a standard humidified incubator at 37 °C containing 5:95% CO2/atmospheric air. The amount of LDH leakage in the medium was measured according to the protocols provided by the company. Briefly, 200 µL of cell supernatant was transferred to a clean flat-bottomed plate, and 100 µL of LDH assay mix was added. The plate was covered with aluminum foil and incubated for 20-30 min. The reaction was stopped by the addition of 30 µL of 1 N HCl, and the absorbance was measured at a wavelength of 490 nm. To view the long-term cytotoxic effects of various polymer films, the osteoblasts were cultured for 7 days and visualized by staining with 10% crystal violet for 10 min,

followed by washing in PBS or water until the color ran clear. Photographs were taken under a reverse microscope using a digital camera (Nikon Coolpix 4500; Maxwell Optical, Lidcombe, NSW, Australia). Cell Adhesion, Spreading, and Proliferation. For the measurement of cell adhesion and spreading, round pieces of polymer films were cut and placed in each well of 96-well tissue culture plates, completely covering the bottom surface of the well. Approximately 7000 cells were seeded to each well with 10% FCS-supplemented or FCS-free medium. For the initial cell attachment, cells were cultured at 37 °C in a humidified 5% CO2 incubator for 4 h; then, unattached cells were removed by rinsing with PBS three times. The total number of cells attached on the films was measured using the CyQuant NF cell proliferation assay kit (Invitrogen, Australia). Briefly, cells attached on the polymer films were measured by quantification of the total amount of DNA with CyQuant NF 1× dye binding solution. The DNA samples were quantified using a fluorescence microplate reader (Polarstar optima, Germany) with the excitation wavelength at 485 nm and emission detection at 585 nm. The attached cells were also visualized by staining with 10% crystal violet for 10 min, followed by washing in PBS or water until the color ran clear. Photographs were taken under a reverse microscope using a digital camera (Nikon Coolpix 4500). We measured cell proliferation on polymer films by examining the total cell numbers after 3 days culture following initial attachment. Five polymer films from each sample were used. Osteoblasts at passage seven were seeded on the films, washed after 1 h and cultured at 37 °C for 3 days. The cell number was quantified by using the CyQuant NF cell proliferation assay kit (Invitrogen). To investigate cell spreading on different polymer surfaces further, we seeded osteoblasts at passage seven on the films and incubated them for 7 days; they were then washed in PBS three times and fixed in 3% glutaraldehyde. The cells were then stained with 1% osmium tetroxide, washed with distilled water, and dehydrated with a graded series of ethanol solutions (50, 70, 90, and 100%). After critical point drying, the samples were

98

Biomacromolecules, Vol. 10, No. 1, 2009

Peng et al.

Figure 1. 1H NMR (in DMSO) spectra of (A) MPEG-b-PLLA-b-PZLys, (B) MPEG-b-PLLA-b-PLL-NH3+Br-, and (C) MPEG-b-PLLA-b-PLL (DPMPEG ) 17, DPPLLA ) 94, DPPLL ) 109).

mounted on aluminum foil and coated with gold, and the morphology was viewed by environmental scanning electron microscopy (SEM) (Quanta 200 FEI) at an accelerating voltage of 10 kV. Statistical Analysis. Statistical evaluation of the data was carried out by one-way analysis of variance (ANOVA) and by Dunnett’s test. A level of significance was set at P < 0.05. Original data were collected and normalized according to the cell numbers, cell percentage, and OD value. Statistical analysis was performed to compare the variations within multiple groups. Tissue culture polystyrene (TCPS) plates were used as the control for all test groups.

Results and Discussion Synthesis of Triblock Copolymers of Methoxy-poly(ethylene glycol)-b-poly(L-lactide)-b-poly(L-lysine) (MPEG-bPLLA-b-PLL). As described above, the amino-terminated diblock copolymers of MPEG-b-PLLA-NH2 were prepared according to the procedure reported by Gotsche and coworkers.48 We synthesized two different MPEG-b-PLLA-NH2 polymers by adjusting the feed ratio of L-lactide to MPEG (DPMPEG ) 17, DPPLLA ) 94; DPMPEG ) 17, DPPLLA ) 56). Primary amines can be used as initiators for the ROP of NCA to prepare poly(R-amino acid)s through a nucleophilic addition to the C-5 carbonyl group of the NCA.34 Therefore, in this study, triblock copolymers of MPEG-b-PLLA-b-PZLys were synthesized via the ROP of Nε-(Z)-Lys-NCA using MPEG-b-PLLANH2 as a macroinitiator. The MPEG-b-PLLA-b-PZLys was then treated with HBr/HAc solution and neutralized with excess N,Ndiisopropylethylamine to remove the protected groups in the side chain and yield MPEG-b-PLLA-b-PLL with free amino groups. The amino groups in the side chain of MPEG-b-PLLAb-PLL can help to improve the affinity of the polymer for proteins and cells or to combine the polymer with drugs, antibodies, or DNA covalently or ionically. The synthetic route is outlined in Scheme 1. A more detailed discussion of the synthetic steps is given below. The 1H spectra of the triblock copolymers before and after deprotection are shown in Figure 1. In the spectrum of MPEGb-PLLA-b-PZLys (Figure 1A), in addition to the signals from the protons of the MPEG and PLLA segments (signals a-f), peaks at 7.90, 4.98, 3.81, 2.92, and 1.36 ppm (signals g-o) assigned to protons of the PZLys block were all identified. The intense peak at 7.30 ppm is assigned to the benzene protons from the side Z group of the protected lysine. The ammonium bromine salt MPEG-b-PLLA-b-PLLNH3+Br- was obtained after the protected copolymer was treated with HBr/HAc solution. As seen in the NMR spectrum,

Figure 2. The ATR-FTIR spectra of (A) MPEG-b-PLLA-b-PZLys, (B) MPEG-b-PLLA-b-PLL-NH3+Br-, and (C) MPEG-b-PLLA-b-PLL (DPMPEG ) 17, DPPLLA ) 94, DPPLL ) 109).

peaks due to benzylic methylene and aromatic protons at 4.93 and 7.30 ppm, respectively (signals n and o, Figure 1A), were completely removed, and an ammonium group was detected in the side chain. Furthermore, after neutralization of the ammonium bromide salt with excess N,N-diisopropylethylamine, a single amino signal was left in the 1H spectrum of MPEGb-PLLA-b-PLL (Figure 1C). Minor peaks at 1.2, 3.1, and 3.6 ppm are due to residual salt of N,N-diisopropylethylamine. The results clearly confirmed that the benzyl groups were completely removed and that the deprotected MPEG-b-PLLA-b-PLL with free amino groups was successfully synthesized. The structure of the triblock copolymers was also confirmed by ATR-FTIR spectroscopy (Figure 2). The spectrum of the MPEG-b-PLLA-b-PZLys (Figure 2A) shows absorption peaks at 3292 cm-1 assigned to the NH stretching vibration, at 1640 and 1690 cm-1 assigned to the two amide I (CO) bands, and at 1540 cm-1 assigned to the amide II (CO-NH) band. These bands from the main-chain and side-chain amide groups indicated the formation of the polypeptide block. The CH vibrations of the benzyl groups from the protected side group of the lysine at 750 and 700 cm-1 were also observed. The peak at 1751 cm-1 is assigned to the carbonyl groups belonging to the PLLA blocks, and the peak at 1190 cm-1 is ascribed to the MPEG blocks. After deprotection, the CH vibration of the benzyl group at 750 and 700 cm-1 and the amide I (CO) vibration in the side-chain amide groups at 1690 cm-1 disappeared (Figure 2B,C), which clearly indicated that the deprotected MPEG-bPLLA-b-PLL was successfully synthesized.

Enhancement of Osteoblast Attachment and Growth

Biomacromolecules, Vol. 10, No. 1, 2009

99

Table 1. Feed Composition and Composition of MPEG-b-PLLA-b-PLL polymer MPEG-b-PLLA-b-PZLys

MPEG-b-PLLA-b-PLL-NH3+Br–

MPEG-b-PLLA -b-PLL

a

Mn (theory)

Mn (1H NMR)a

composition of feed (EO/LLA/LL)

composition of polymer (EO/LLA/LL)a

10 100 16 120 26 930 17 250 20 920 35 330 8780 13 580 23 550 15 080 18 010 29 500 6880 9820 15 840 12 770 14 500 21 600

12 300 17 540 28 410 19 830 21 140 36 080 10 780 18 270 30 060 16 510 20 060 31 550 8540 13 150 19 810 13 660 15 200 22 240

17:40:24 17:40:47 17:40:94 17:80:41 17:80:55 17:80:110 17:40:24 17:40:47 17:40:94 17:80:41 17:80:55 17:80:110 17:40:24 17:40:47 17:40:94 17:80:41 17:80:55 17:80:110

17:56:28 17:56:48 17:56:100 17:94:47 17:94:57 17:94:109 17:56:28 17:56:64 17:56:125 17:94:43 17:94:60 17:94:115 17:56:28 17:56:64 17:56:125 17:94:48 17:94:60 17:94:115

Calculated from results of 1H spectroscopy.

The molecular weights of the protected and deprotected copolymers were calculated from the results of 1H spectroscopy. The DP and Mn of the block copolymers were listed in Table 1. The value of DPPZLys was very close to the calculated values according to the feed ratio of NCA monomer to the macroinitiator. This suggests that the DP of the PZLys segment could be controlled by adjusting the ratio of NCA monomer to initiator. The molecular weights did not change significantly before and after deprotection, which suggested that polymer main-chain cleavage did not occur during the deprotection reaction. Surface Analysis. The triblock copolymers were blended in high-molecular-weight PLLA and cast into films, as described in the Experimental Section, to examine the effect on cell adhesion. In the following sections, the surface composition and cell assays are described. X-ray Photoelectron Spectroscopy Analysis. Figure 3 shows the XPS spectra of the virgin PLLA film (A) and PLLA modified with the MPEG-b-PLLA-b-PLL triblock copolymer (11.21 wt % MPEG-b-PLLA-b-PLL in PLLA) (B) at 90° takeoff angle. The XPS survey spectrum of the unmodified PLLA (Figure 3A) film shows characteristic peaks at 284 and 531 eV attributed to C 1s and O 1s, respectively. However for the PLLA film modified with MPEG-b-PLLA-b-PLL, a small peak corresponding to N 1s was observed at 400 eV (Figure 3B) that was due solely to PLL on the surface of the PLLA film. As expected, the unmodified PLLA surface generates three peaks in the C 1s high-resolution spectrum (Figure 3A), indicating that this surface contains three carbon types. The binding energies of the peaks due to C-C, C-O, and CdO were 284.3, 286.6, and 288.8 eV, respectively. After modification with MPEG-b-PLLA-b-PLL, there was a significant change in the surface chemistry of the PLLA. Changes in the C 1s spectrum were apparent, and the fitting of the spectrum allowed the identification of a new peak at 285.7 eV, which was assigned to C-N carbons. This was assigned to the PLL component of the copolymer. The bulk and surface C, O, and N contents and atomic ratio of N/C of the PLLA film modified with MPEG-b-PLLA-b-PLL are listed in Table 2. The bulk values were calculated from the results of elemental analysis. The C, O, and N contents on the surface were different from the bulk, and the atomic ratio of N/C on the surface was much higher than that in the bulk, which was consistent with the enrichment of MPEG-b-PLLA-b-PLL on the PLLA surface.

Figure 3. XPS survey and C1s spectra of (A) virgin PLLA film and (B) film modified with MPEG17-b-PLLA56-b-PLL28 triblock copolymer at 90° takeoff angle. Table 2. Surface Composition of the PLLA Films Prior to and after Modification by Blending mass concentration surface composition bulka surfaceb a

O%

C%

N%

N/C

45.1 ( 0.05 36.4 ( 0.05

54.0 ( 0.05 59.7 ( 0.05

0.8 ( 0.05 3.9 ( 0.05

0.015 0.066

From elemental analysis of whole film.

b

XPS analysis.

By using the XPS data, it is possible to propose possible structures of the surface of the films. It is clear from the higherthan-expected nitrogen content for bulk mixing of the two components that the triblock copolymer is partitioning on the surface of the film. However, a close examination of the

100

Biomacromolecules, Vol. 10, No. 1, 2009

Scheme 2. Model of MPEG-b-PLLA-b-PLL Triblock Copolymer on the Surface of PLLA Matrix

observed C, O, and N contents and the comparison with the C, O, and N contents of the component polymers suggest that the triblock copolymer does not form a continuous layer on the surface but that there is a contribution to the XPS results from the bulk PLLA homopolymer. Therefore a reasonable model of the morphology of the surface of the blends, shown in Scheme 2, has MPEG and PLL blocks of the copolymer presented to the surface, the PLLA segments of the triblock providing miscibility with the PLLA matrix, and a proportion of the surface being pure PLLA. Ji and coworkers6 have reported a similar surface structure in films of end-functionalized triblock copolymers of PEO and PPO blended with PLLA. These authors also reported enhanced cell attachment and growth in copolymers functionalized with monopeptides and with the RGD sequence. Atomic Force Microscopy Analysis. The surface structure of these films was also studied by AFM. As shown in Figure 4A, the virgin PLLA film presented a smooth surface. However, we can clearly see some phase segregation on the surface of the PLLA film modified with MPEG-b-PLLA-b-PLL. Furthermore, when the DP of the PLL block increased from 28, to 48, to 100, the size of the segregated phases became larger (∼145, 280, and 420 nm, respectively), as shown in Figure 4B-D. These images, in addition to the XPS analysis, confirm the selfsegregation behavior of MPEG-b-PLLA-b-PLL on the surface of the PLLA films. The proposed model depicted in Scheme 2 is therefore supported by the AFM images. The hydrophobic PLLA segments of the MPEG-b-PLLA-b-PLL triblock copolymer are miscible with the PLLA matrix, and the significantly more hydrophilic PEG and PLL chains will spread on the surface of the PLLA. A similar surface phase structure was reported by Ji et al.6 in blends of triblock copolymers with PLLA. The observed self-segregation of MPEG-b-PLLA-b-PLL triblock copolymer may be broadly useful in facilitating the presentation of functional groups in tissue scaffolds. Cell Behavior. Human osteoblasts derived from healthy alveolar bone chips were used to test the cellular response on the polymer surfaces. Cell adhesion and proliferation on various films were evaluated by culturing the osteoblasts in DMEM medium containing 10% FBS. The test samples were virgin PLLA films and PLLA films modified with 5 wt % of either MPEG-b-PLLA diblock copolymer or MPEG-b-PLLA-b-PLL triblock copolymer. Cytotoxicity. We examined the in vitro cytotoxicity by measuring cell membrane integrity using the extracellular LDH leakage assay after 24 h of incubation at 37 °C. To assess the long-term cytotoxic effects of various polymer films, osteoblasts were cultured for 7 days and stained by crystal violet to monitor morphological changes. TCPS plates were used as a control. The results are presented in Figure 5. The copolymer films were labeled according to the ratios of DP in each block; that is, the ratio 17:94 represents the diblock copolymer with 17 units of PEG joined to 94 units of PLLA. The optical density (OD) at 490 nm is a measure of the rate of oxidation of nicotinamide adenine dinucleotide released from cytoplasmic LDH and is hence a measure of the cell membrane integrity. This is a direct measurement of the cytotoxicity of the polymer films. It was noted that no significant change in extracellular LDH was

Peng et al.

detected in the cultures on all polymer films compared with the culture in TCPS (p > 0.05) (Figure 5A), indicating that no significant cytotoxicity was detected for the PLLA, MPEG-bPLLA diblock copolymers, and MPEG-b-PLLA-b-PLL triblock copolymers. This confirms the biocompatibility of the synthetic copolymers. In a study similar to ours, Deng and coworkers50 reported the synthesis of poly(ethylene glycol)-b-poly(L-lactide)-b-poly(Llysine) triblock copolymers and the conjugation of RGD sequences to the lysine units. They reported enhanced cell adhesion and spreading on the PEG-b-PLLA-b-PLL/RGD film compared with pure PLLA films; however, after 20 h of incubation, the number and size of the cells on the PEG-bPLLA-b-PLL copolymer (without RGD) decreased. The authors suggested that PLL was harmful to the cells. In the preparation of the triblock copolymer, Deng et al.50 used a solution of HBr/ acetic acid to affect the deprotection of PEG-b-PLLA-b-PZLL. Using this procedure, it is likely that the final product is not the amino form of lysine but rather the ammonium bromide salt. Fan et al.42 reported the deprotection PLLA-b-PZLL by using the same procedure as Deng, and the amine region of the NMR spectrum of the copolymer in the form of lysine bromide was essentially identical to that reported by Deng. Fan et al. assigned this to the hydrogen bromide salt. Previously, others reported the neutralization of the bromide salt of lysine-containing peptides by reaction with amines. For example, Hrkach et al.1 reported the deprotection of PLLA-b-PZLL by HBr/HAc solution followed by neutralization to the free amino side chains with excess N,N-diisopropylethylamine. In our study, the amine region of the NMR spectrum of the MPEG-b-PLLA-b-PLL-NH3+Br- after deprotection of MPEGb-PLLA-b-PZLys with HBr/acetic acid solution was the same as the spectrum reported by Deng et al. (Figure 1B). However, after the sample was neutralized with N,N-diisopropylethylamine, only a single peak was observed in the appropriate region, indicating the formation of MPEG-b-PLLA-b-PLL with free amino groups (Figure 1C). To confirm the assignments to the NMR spectra of MPEG-b-PLLA-b-PLL and MPEG-bPLLA-b-PLL-NH3+Br-, we added one drop of D2O or 33 wt % HBr solution to solutions of MPEG-b-PLLA-b-PLL separately. As shown in Figure 6, the deuterium (D) replaced the protons of the NH2 groups when D2O was added, so the single peak derived from amine in Figure 6A disappeared in Figure 6B. When the HBr solution was added to MPEG-b-PLLA-bPLL, the amine groups reacted with HBr to form the ammonium bromide salt, and the amine region in Figure 6C was that in the spectrum of the MPEG-b-PLLA-b-PLL-NH3+Br- (Figure 1B). From these simple experiments, we have confirmed the assignments to the 1H spectra of MPEG-b-PLLA-b-PLL and MPEGb-PLLA-b-PLL-NH3+Br-. To study the effect of hydrogen bromide on the cell cytotoxicity, we also tested the cytotoxicity of PLLA films modified with 5 wt % MPEG-b-PLLA-b-PLL-NH3+Br- and commercial PLL-NH3+Br- (used without dialysis). The results of these experiments were compared with the cytotoxicities of our final films of PLLA modified with MPEG-b-PLLA-b-PLL triblock copolymer. As shown in Figure 5B, significantly higher OD values were noted in the cultures on PLLA films modified with MPEG-b-PLLA-b-PLL-NH3+Br- as well as PLLNH3+Br- films (p < 0.05) compared with the TCPS, PLLA, and MPEG-b-PLLA-b-PLL triblock copolymer modified films, indicating the damage of the cellular membranes after 24 h of incubation in these two types of films. These results were further

Enhancement of Osteoblast Attachment and Growth

Biomacromolecules, Vol. 10, No. 1, 2009

101

Figure 4. AFM phase images of (A) virgin PLLA film and (B-D) PLLA film modified with 5 wt % MPEG-b-PLLA-b-PLL (B: MPEG17-b-PLLA56b-PLL28, C: MPEG17-b-PLLA56-b-PLL48, and D: MPEG17-b-PLLA56-b-PLL100).

demonstrated in the subsequent observation of cell morphological changes after 7 days of incubation (Figure 5C). Only cell debris were detected on PLL-NH3+Br- (Figure 5C, i) and MPEG-b-PLLA-b-PLL-NH3+Br- (Figure 5C, ii) membranes, whereas a normal spindle cell morphology was observed on the films modified with triblock copolymers of MPEG17-bPLLA94-b-PLL57 (Figure 5C, iii) and MPEG17-b-PLLA56-bPLL28 (Figure 5C, iv). The osteoblasts grown on the films modified with MPEG17-b-PLLA94-b-PLL57 (Figure 5C, iii) and MPEG17-b-PLLA56-b-PLL28 (Figure 5C, iv) triblock copolymers showed cell morphology that was similar to that of the osteoblasts on TCPS (Figure 5C, v). We therefore suggest that the poor cell growth reported for MPEG-b-PLLA-b-PLL copolymer by Deng et al. may be due to incomplete conversion of the hydrogen bromide salt of the lysine groups. It has been reported that polypeptides containing lysine hydrogen bromide prevent cell growth and may lead to cell death in some circumstances.51 Cell Adhesion and Proliferation. The cell adhesion and proliferation were evaluated from the relative cell responses on different surfaces compared with pure PLLA films, and the cell number was quantified by the CyQuant NF cell proliferation assay. TCPS was used as a positive control. After incubation at 37 °C for 4 h, as shown in Figure 7, the number of cells on the surface of films modified with the triblock copolymer was

greater than that of diblock-modified and pure PLLA films (Figure 7A). This indicates that cells adhere and spread more extensively on the triblock-modified surface than on the other surfaces. It was noted that there was a correlation between the number of cells attached and the molecular weight of PLL block. As the chain length of PLL blocks in the triblock copolymers having compositions of either MPEG17-b-PLLA56-b-PLLx or MPEG17-b-PLLA94-b-PLLy increased, the cell numbers increased (Figure 7A). This demonstrates, as is expected, that PLL can enhance the cell attachment and spreading in triblockmodified PLLA films. The cells attached to polymer films were also visualized by staining with crystal violet (Figure 7B, i-iii). After 4 h, osteoblasts showed a spindle morphology on these films. No differences in morphology were noted on the polymer films tested. The cell proliferation on the PLLA films modified by the incorporation of the diblock and triblock copolymers is shown in Figure 8. Enhanced cell proliferation was achieved on PLLA films modified with MPEG17-b-PLLA56-b-PLLx and MPEG17b-PLLA94-b-PLLy compared with that on the relevant diblock copolymer-modified films and virgin PLLA films (Figure 8A). The PLLA films modified with triblock copolymers exhibited the best results with significant increases in cell numbers, which indicates that the triblock copolymers are more effective in

102

Biomacromolecules, Vol. 10, No. 1, 2009

Peng et al.

Figure 5. (A) Membrane integrity values of OBM cells seeded on TCPS, virgin PLLA films, and PLLA films modified with 5 wt % MPEG-b-PLLA or 5 wt % MPEG-b-PLLA-b-PLL evaluated by extracellular LDH assay incubation at 37 °C for 1 day. Columns and error bars represent the means ( SD of six measurements. Significance was calculated by Dennett’s test: *P < 0.05. (B) Membrane integrity values of OBM cells seeded on TCPS, virgin PLLA film, PLLA film modified with 5 wt % PLL-NH3+Br-, 5 wt % MPEG-b-PLLA-b-PLL-NH3+Br-, and 5 wt % MPEGb-PLLA-b-PLL evaluated by extracellular LDH assay incubation at 37 °C for 1 day. Columns and error bars represent the means ( SD of three measurements. Significance was calculated by Dennett’s test: *P < 0.05. (C) Microscopic images of OBM cells seeded on PLLA film modified with (i) 5 wt % PLL-NH3+Br-, (ii) 5 wt % MPEG-b-PLLA-b-PLL-NH3+Br-, (iii) 5 wt % MPEG17-b-PLLA94-b-PLL57, (iv) 5 wt % MPEG17-b-PLLA56b-PLL28, and (v) and TCPS for 7 days. Scale bar is 50 µm.

Figure 6. 1H NMR (in DMSO) spectra of (A) MPEG-b-PLLA-b-PLL, (B) MPEG17-b-PLLA94-b-PLL 109 + one drop D2O, and (C) MPEG17b-PLLA94-b-PLL109 + one drop 33 wt % HBr solution.

promoting cell proliferation than are the diblock copolymers and pure PLLA. After 7 days of incubation, a notable increase in cell proliferation was noted on the triblock copolymer-modified PLLA surface when it was investigated under SEM, with large numbers of spreading cells attached to the film surface. Higher cell densities and multilayered spindle cells demonstrated enhanced spreading and faster proliferation on the surface modified with triblock copolymer (Figure 8B, iii) compared with that on the diblock polymer (Figure 8B, ii) and PLLA (Figure 8B, i) films. These results also indicate that the triblock copolymers are significantly more effective at promoting cell adhesion and proliferation than are the diblock copolymers and PLLA only. As mentioned above, Ji and colleagues6 have studied the effects of the incorporation of functionalized PEO-PPO-PEO triblock copolymers on chondrocyte attachment to PLLA. These

authors found that conjugation of a single peptide residue to the terminal units of the triblock copolymers resulted in a material that when blended with PLLA enhanced the extent of cell attachment. When the monopeptide was replaced by the RGD tripeptide, which is well known for promoting cell adhesion, results comparable to those of the positive control were obtained. In our study however, we have demonstrated that comparable results can be achieved by the use of a polypeptide (in this case PLL) without the need to incorporate the adhesion motif RGD.52-54 In 1993, Cook and coworkers55 were the first to report the modification of the copolymer of L-lactide and lysine with RGDpeptide, and they later39 demonstrated the attachment of enhanced bovine aortic endothelial (BAE) cells in blends with PLLA. However, in their early work, the material prepared was a random copolymer with a lysine-like monomer, and the resultant polymer contained