BSA的稳定性

spectroscopy for secondary structure analyses and fluorescence measurements for tertiary structure analyses. The release profile Keywords: PEG-poly(histidine) diblock copolymer; Protein stability; Poly(lactic-co-glycolic acid) [PLGA]; Acidic microenvironment; Buffering capacity of BSA from the microspheres was monitored using UV spectrophotometry. The rate of PLGA degradation was monitored by gel permeation chromatography. The pH profile within microspheres was further evaluated by confocal microscopy using a pH- sensitive dye. Approximately 19 PEG-PH molecules and one BSA molecule coalesced to form an ionic complex around a pH range of 5.0–6.0. Plain BSA/PLGA and BSA/PEG-PH/PLGA microspheres had a mean size of 27–35 Am. PLGA microspheres with a BSA loading efficiency N80% were prepared using the double emulsion method. PEG-PH significantly improved the stability of BSA both in aqueous solutions and in PLGAmicrospheres. The release profiles of BSA from different formulations of PLGA microspheres were significantly different. PEG-PH effectively buffered the local acidity inside the microspheres and improved BSA release kinetics by reducing initial burst release and extending continuous release over a period of time, when encapsulated as an ionic complex. PLGA degradation rate was found to be delayed by PEG-PH. There was clear evidence that PEG-PH played multiple roles when complexed with BSA and incorporated into PLGA microspheres. PEG-PH is an effective excipient for preserving the structural stability of BSA in aqueous solution and BSA/PLGA microspheres formulation. D 2005 Elsevier B.V. All rights reserved. Stability of bovine serum albumin complexed with PEG-poly(l-histidine) diblock copolymer in PLGA microspheres Jong-Ho Kim a,b, Ajay Taluja a, Kristine Knutson a, You Han Bae a,* a Department of Pharmaceutics and Pharmaceutical Chemistry, University of Utah, 421 Wakara Way Suite 318, Salt Lake City, UT 84108, USA b Center for Biomaterials and Biotechnology, Department of Materials and Engineering, Gwangju Institute of Science and Technology, 1 Oryong-dong, Buk-gu, Gwangju, 500-712, Korea Received 1 March 2005; accepted 15 August 2005 Available online 2 November 2005 Abstract The aim of this study was to examine the stability of bovine serum albumin (BSA) in poly(dl-lactic acid-co-glycolic acid) (PLGA) microspheres upon addition of a new excipient, poly(ethylene glycol)-poly(l-histidine) diblock copolymer (PEG-PH). Poly(l-histidine) component can form an ionic complex with BSA under acidic conditions within a narrow pH range. To optimize the ionic complexation conditions for BSA with PEG-PH, the resulting complex sizes were monitored using the Zetasizer. PLGA microspheres containing BSA as a model protein were prepared by w/o/w double emulsion method. BSA stability in aqueous solutions and after release from PLGA microspheres was determined using circular dichroism (CD) Journal of Controlled Release 109 (2005) 86–100 www.elsevier.com/locate/jconrel 0168-3659/$ - s doi:10.1016/j.jco * Correspondin E-mail addre ee front matter D 2005 Elsevier B.V. All rights reserved. nrel.2005.09.016 g author. Tel.: +1 801 585 1518; fax: +1 801 585 3614. ss: you.bae@m.cc.utah.edu (Y. Han Bae). ntroll 1. Introduction Parenteral administration of proteins and peptides has been a method of choice for systemic delivery of agents due to ease of administration, bypassing biologic barriers that prevent proteins from passing, and the ability to achieve pharmacologic levels of circulating protein over a relatively short period of time. Peptides and proteins are frequently adminis- tered using subcutaneous or intramuscular injections as well as intravenous infusions. Their short biological half-lives, usually in the range of several minutes, require frequent injection regimens and cause consid- erable discomfort to patients, especially when long- term or chronic treatment is necessary. A prominent example is the treatment of diabetes mellitus using insulin, where patient compliance and long-term complications are closely associated. There has been widespread interest in using biodegradable polymers for controlling the release of proteins [1–4]. One method available to extend the duration of protein delivery, thus overcoming repetitive administration by conventional injections, is encapsulating the protein within microspheres (MS) composed of a biodegrad- able polymer. Poly(e-caprolactone), polyanhydrides, polylactide (PLA), polyglycolide (PGA) and copoly- mer of lactic and glycolic acid (PLGA) microspheres have been studied extensively as potential long-term protein delivery systems [5–10]. Among the biode- gradable polymers most suitable for formulating injectable protein-loaded microspheres, most experi- ence has been gained with PLGA. The popularity of these biocompatible copolymers can be ascribed in part to the approval by the FDA of several products for use in humans and their success as biodegradable sutures. Microspheres have been used successfully to control the release of peptides over a period of one month [11–13]. Compared to peptides like leuprorelin [11–13], proteins have substantially higher molecular weights and are more susceptible to chemical degra- dation and physical inactivation in formulations and the body [14]. Very few proteins have shown the ability of controlled delivery from PLGA micro- spheres [15,16]. This is because proteins undergo physical denaturation and chemical degradation dur- ing the fabrication of such controlled-release formu- J.-H. Kim et al. / Journal of Co lations [17–23]. In addition, such formulations often cause unpredictable release profiles, characterized by a burst effect and incomplete release [24–26]. Issues such as protein instability at the water/organic solvent interface in double emulsion w/o/w method, nonspe- cific interactions with the matrix-forming polymers [27], and local acidity development within the degrading polymer matrix need to be thoroughly addressed [28–30]. The successful application of PLGA microspheres for controlled delivery of pro- teins is therefore contingent on maintaining protein stability during formulation, storage and release and also achieving a well-defined release pattern. One approach to tackle protein instability during formula- tion is to add excipients to the inner aqueous core that compete with the proteins at the water-organic solvent interface and can prevent emulsification-induced denaturation and aggregation. There are reported studies on protein stability during the primary w/o emulsion step and the effect of stabilizing excipients (e.g. cyclodextrins, polyols, and PEGs) [31–34]. Stabilization using surfactants at interfaces and surfaces has been shown to prevent denaturation. However, this does not solve other problems associ- ated with local acidity and undesirable release patterns. One solution to improve protein release is to neutralize the acidic microenvironment within the PLGA microspheres. PLGA degrades through an acid-catalyzed hydrolysis and produces carboxylic acids. Various research groups have investigated methods for stabilizing proteins against the acidic microenvironment [35–37]. Schwendeman group has investigated magnesium hydroxide (Mg(OH)2), to neutralize the acidic microenvironment within the degrading PLGA microspheres system [37]. This caused an improved release kinetics and stability of encapsulated proteins. Since no single method can address all of the problems associated with protein stability and delivery, an optimal solution will require a combination approach. Proteins interact strongly with both synthetic and natural polyelectrolytes. Complexation is dependent upon the ionic strength of the solution as well as the charge properties of the protein and the polyelectro- lyte. Protein complexation with polymers in aqueous solution may be driven by hydrogen bonding, hydrophobic interactions, and electrostatic interac- tions. Hydrogen bonding can occur only in special ed Release 109 (2005) 86–100 87 circumstances under stringent conditions since it requires not only donor and acceptor atoms, but also overcoming aggregation and the acidic microenviron- ment. The imidazole ring in l-histidine (His) has an ntroll electron pair on the unsaturated nitrogen that endows polyHis with an amphoteric nature by rapid and reversible protonation–deprotonation [41–43]. The pKa value of PEG-polyHis(or PEG-PH) lies between 6.5 and 7.0 [42,43]. Therefore, PEG-polyHis may form an ionic complex with a negatively charged protein such as bovine serum albumin (BSA), within a pH range of 4.9 to 6.5. The present study is aimed at investigating the effect of poly(ethylene glycol)-b- poly(l-histidine) (PEG-PH) diblock copolymer on the stability of a model protein, BSA. In this report, we describe the stabilization and release profile of BSA from PLGA microspheres. Additionally, poly(l-histi- dine) is a weakly basic polymeric component which can neutralize the acidic microenvironment within PLGA microspheres created by acidic degradation products such as lactic and glycolic acids and acidic polymer products. The hydrophilic PEG is expected to prevent nonspecific interactions between BSA and PLGA matrix polymer. When the BSA/PEG-PH complex is exposed to physiological pH, it is expected to dissociate, and expose the native protein. 2. Materials and methods 2.1. Materials Poly(dl-lactide-co-glycolide) (PLGA, nominal LA:GA ratio: 50 :50; MW 40,000–75,000), bovine serum albumin (BSA, fraction V), and poly(vinyl alcohol) (PVA, MW 85,000–146,000) were purchased from Aldrich Chemical company. Methylene chloride (MC) from Fisher Scientific Co. was used without particular spatial alignment of the donor-acceptor pairs. Consequently, electrostatic and hydrophobic interactions are more common factors in protein- polyelectrolyte complexation [38–40]. These forces have been widely utilized in a variety of biological processes, especially for protein separation and purification. The concept proposed here aims to utilize smart macromolecules as a novel dtemporal and reversible molecular shieldT to achieve improved release kinetics and to retain protein stability while J.-H. Kim et al. / Journal of Co88 further purification. SNARF-1k dextran conjugate was from Molecular Probes, Oregon. Poly(ethylene glycol)-b-poly(l-histidine) (PEG-PH) diblock copoly- mers were synthesized and characterized as previously reported [42,43]. The reproducibility of the polymer synthesis was confirmed. The molecular weight of PEG-polyHis used in this preliminary study was 7 kDa consisting of two blocks; poly(ethylene glycol) PEG 2K and polyHis (PH) 5K as determined by 1H NMR and a polydispersity index of 1.2 as determined by MALDI-TOF analysis. 2.2. Determination of particle size and stability of BSA The positively charged diblock copolymer should form an ionic complex with a negatively charged protein such as BSA. The isoelectric point (pI) of BSA is 4.9. The particle size studies helped to determine the working pH range for complexation of BSA and PEG-PH. The aim was to determine an optimal pH that resulted in a stable ionic complex, so that it could be incorporated in the PLGA matrix. The particle size of the BSA and PEG-PH ionic complexes in an aqueous medium was measured by using Zetasizer (Brookhaven Instruments Corp. ZetaPALS) in pH range, from 4.5 to 7.5 at 25 8C. PEG-PH concentration was fixed at 0.5 mg/mL. BSA concen- tration was varied from 0.17 to 1.5 mg/mL. Therefore, the weight ratio of PEG-PH to BSA ranged from 0.33 to 3. The size of uncomplexed PEG-PH and BSAwas measured within the experimental pH range as controls. 2.3. Conformational analyses of BSA and BSA/PEG- PH complexes at different pH 2.3.1. Tertiary structure To confirm the retention of the tertiary structure of a protein containing tryptophan residues, researchers have typically used fluorescence techniques [44,45]. The working pH range for complex formation as determined from particle size was employed; pH 4.5 and 7.5 were selected as non-complexing pH and pH 5.5 was selected as the optimal pH for ionic complexation based upon ionic interactions between BSA and PEG-PH. Fluorescence measurements were performed using a fluorescence spectrophotometer ed Release 109 (2005) 86–100 (PerkinElmer LS55). All samples were incubated at 25 8C. To estimate the degree of retention of BSA ntroll folded structure, the emission peak of tryptophan (Trp) in BSA was measured at 25 8C. Trp in a hydrophobic environment (protein core) presents a peak at 335 nm and upon exposure to a hydrophilic environment (unfolding) the peak is shifted to a wavelength in the range 345 to 380 nm (red shift) when the excitation wavelength is set at 290 nm [46]. As an indication of BSA’s folded structure, the ratio of the peaks of BSA (Pt) to native BSA (Pn) at 335 nm was presented as (Pt /Pn)�100 (%), where Pt is the peak intensity of released or treated BSA at time t, and Pn is the peak intensity of native BSA at 335 nm at time zero. The spectra of PEG-PH and BSA at pH 7.4 were run as controls. 2.3.2. Secondary structure Retention of secondary structure was monitored using circular dichroism (CD). The CD spectra were recorded on a Jasco J-720 spectropolarimeter. A quartz cuvette with 0.1 cm path length was used. The spectra were scanned between 190 and 260 nm with 0.5 nm resolution; 16 scans were taken with a scan rate of 100 nm min-1 and a time constant of 0.125 s. The spectral analysis was performed by deconvolution of CD spectra by CDNN v2.1 software (Gerald Bohm of University of Halle (Germany)). 2.4. Preparation of PLGA microspheres PLGA microspheres were prepared using a con- ventional w/o/w double emulsion and solvent evapo- ration method. BSA and PEG-PH (total 20 mg) were dissolved in 2 mL of a buffer solution with different pH. The ionic strength of each buffer was adjusted to 0.15 M with sodium chloride. The pH 4.5 and 5.0 buffer solutions are based on potassium hydrogen phthalate as per the pharmacopoeia specifications. The protein solution served as the internal aqueous phase for the primary emulsion. They were emulsified with 10 mL methylene chloride containing 500 mg PLGA for 1 min using a homogenizer (20,000 rpm). This w/o emulsion was poured into 200 mL of 0.5% (w/v) poly(vinyl alcohol) (PVA) and 0.5% (w/v) sodium chloride (NaCl) for secondary emulsion formation. Emulsification was continued using a mechanical stirrer at 2000 rpm for 1 min. This J.-H. Kim et al. / Journal of Co dispersion was stirred for 4 h at 35 8C for solvent evaporation. The microspheres were collected by centrifugation at 3000 rpm for 10 min. The obtained microspheres were washed with water and freeze dried for 24 h. The shape and size of microspheres were visualized with scanning electron microscopy (SEM, Hitachi S-3000N). The actual BSA loading efficiency in the microspheres was determined by extraction of BSA from the MS according to a technique patented by Tice et al. [47]. The micro- spheres were first dissolved in methylene chloride under magnetic stirring. Then 2 mL of a phosphate buffer solution (PBS, pH 7.4) was added and the contents were agitated for 1 min. After centrifugation for 10 min at 3000 rpm twice, the aqueous phase was transferred into a vial. The aqueous fractions were collected in the vial and the final volume was made to 5 mL with PBS. After extraction, the encapsulated amount of protein was determined using UV spec- troscopy (CARY 3E UV–Visible Spectrophotometer). The loading efficiency was calculated using the following equation: Loading efficiency of BSA %ð Þ ¼ La=Ltð Þ � 100% where La is the amount of BSA extracted from PLGA microspheres and Lt is the theoretical loading of BSA in the PLGA microspheres calculated from the feeding amount during the preparation. 2.5. Characterization of microspheres Plain PLGA microspheres containing only BSA (BSA/PLGA) were smooth and spherical with a mean size of 23.4F2.3 Am. The loading efficiency of BSA was 81.3F7.4% as determined by UV spectropho- tometer. The characterization of microspheres pre- pared under other conditions was similar to those of plain PLGA microspheres. Characterization of all microspheres (including mean size and loading efficiency) is summarized in Table 1. 2.6. In vitro degradation of PLGA microspheres and BSA release kinetics The release of BSA from the microspheres containing complexed diblock copolymer/BSA was performed in a simulated physiological fluid; the microspheres were immersed in a vial containing ed Release 109 (2005) 86–100 89 PBS, pH 7.4, 0.15 M at 37 8C. The inner water phase pH in one w/o/w emulsion procedure was 7.4. aq. p ntroll This condition was adopted to examine the effect of the presence of PEG-PH without ionic association with BSA during microspheres fabrication on BSA release profiles. Microspheres containing BSA com- plexed with PEG-PH at pH 5.5 and BSA only were also subjected to release kinetics studies. These control experiments were designed to assess the role played by PEG-PH in modulating the release characteristics. The vials were horizontally shaken at 50 rpm. Aliquots of supernatant were withdrawn periodically. An equal amount of fresh PBS was added to the vials containing the microspheres at each sample withdrawal. Protein content in the release medium was measured at 280 nm by using UV (CARY 3E UV–Visible Spectrophotometer). The structural stability of released BSA was ana- lyzed as described in the previous section. After the estimation of the protein content, samples were dried in vacuo for 24 h to stop further degradation of PLGA. The samples were kept in a desiccator until further analysis. The change in molecular weight of PLGA was monitored by GPC using N,N-dimethyl- Table 1 The characteristics of PLGA microspheres Microsphere Ratio (w/w) Inner BSAa PEG-PH BSA/PLGA 1 0 7.4 BSA/PH05/PLGA 1 0.5 7.4 BSA/PH10/PLGA 1 1 7.4 BSA/PH20/PLGA 1 2 7.4 BSA/PH20-5/PLGA 1 2 5.5 BSA/PH20-4/PLGA 1 2 4.5 a BSA concentration was fixed at 0.5 mg/mL. J.-H. Kim et al. / Journal of Co90 formamide (DMF) as an eluent. The change was determined by comparing the molecular weight at a given time (Mnt) with the initial molecular weight (Mn0). The change of molecular weight was defined as: Mn % ¼ Mnt=Mn0ð Þ � 100%: The surface morphology and internal structure of microspheres were investigated by SEM. To assess surface morphology, the microspheres were mounted onto metal stubs using double-sided adhesive tape, vacuum-coated with a gold film and directly observed by SEM. For internal structure analysis, the micro- spheres were embedded in gelatin and cross-sectioned using an ultra-microtome, coated with gold and viewed by SEM. These two studies provided a relative rate of PLGA degradation in MS and the buffering effect provided by the diblock copolymer. 2.7. pH measurement by confocal microscopy Further evidence of PEG-PH buffering effect and the neutralization of acidic microenvironment in PLGA microspheres were obtained using the combi- nation of two confocal microscopes, Nikon Diaphot 200 and Yokagowa spinning dish confocal. The pH within microspheres was determined by SNARF-1R dextran conjugate, which is a dye that exhibits a significantly different emission ratio 580 /640 nm when excited at 488 nm [28]. Emission images at different wavelengths were collected in photon- counting mode using 63� objective and intensified video camera. The ratio was calculated using Openlab 3.09 (Improvision). The calibration bar and pH calibration bar were constituted using dye solutions hase pH Size (Am) meanFSD Loading effic