Journal of Colloid and Interface Science Preparation of uniform
Transcription
Journal of Colloid and Interface Science Preparation of uniform
Journal of Colloid and Interface Science 323 (2008) 267–273 Contents lists available at ScienceDirect Journal of Colloid and Interface Science www.elsevier.com/locate/jcis Preparation of uniform-sized PELA microspheres with high encapsulation efficiency of antigen by premix membrane emulsification Qiang Wei a,b , Wei Wei a,b , Rui Tian a,b , Lian-yan Wang a , Zhi-Guo Su a , Guang-Hui Ma a,∗ a b State Key Lab of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing, PR China Graduate School of Chinese Academy of Sciences, Beijing, PR China a r t i c l e i n f o a b s t r a c t Article history: Received 3 February 2008 Accepted 19 April 2008 Available online 30 April 2008 Relatively uniform-sized poly(lactide-co-ethylene glycol) (PELA) microspheres with high encapsulation efficiency were prepared rapidly by a novel method combining emulsion-solvent extraction and premix membrane emulsification. Briefly, preparation of coarse double emulsions was followed by additional premix membrane emulsification, and antigen-loaded microspheres were obtained by further solidification. Under the optimum condition, the particle size was about 1 μm and the coefficient of variation (CV) value was 18.9%. Confocal laser scanning microscope and flow cytometer analysis showed that the inner droplets were small and evenly dispersed and the antigen was loaded uniformly in each microsphere when sonication technique was occupied to prepare primary emulsion. Distribution pattern of PEG segment played important role on the properties of microspheres. Compared with triblock copolymer PLA–PEG–PLA, the diblock copolymer PLA–mPEG yielded a more stable interfacial layer at the interface of oil and water phase, and thus was more suitable to stabilize primary emulsion and protect coalescence of inner droplets and external water phase, resulting in high encapsulation efficiency (90.4%). On the other hand, solidification rate determined the time for coalescence during microspheres fabrication, and thus affected encapsulation efficiency. Taken together, improving the polymer properties and solidification rate are considered as two effective strategies to yield high encapsulation. © 2008 Elsevier Inc. All rights reserved. Keywords: Premix membrane emulsification PELA microspheres Uniform-sized Coalescence Encapsulation efficiency 1. Introduction Biodegradable microspheres have received more and more attentions for therapeutic application such as controlled release and drug targeting [1–4]. Among several techniques to fabricate biodegradable microspheres [5], the water-in-oil-in-water (W/O/W) double emulsion method is generally considered as one of the most convenient ways to encapsulate water soluble protein [6–8]. Although the process is conceptually simple, many factors, either material- or process-related, are still critical to the obtained microspheres [7]. To date, polylactide (PLA) is the most commonly used polymer because of its proven biodegradable and biocompatible. However, PLA has some drawbacks, resulting from hydrophobic nature, which is related to their rapid sequestration by the mononuclear phagocyte system [9,10]. Furthermore, the affinity between hydrophobic polymer and hydrophilic protein is not satisfied, and leads to a low encapsulation efficiency of protein drug [11]. High encapsulation efficiency has been considered to be an important requirement for the successful fabrication of microspheres con- * Corresponding author. Fax: +86 10 82627072. E-mail address: ghma@home.ipe.ac.cn (G.-H. Ma). 0021-9797/$ – see front matter doi:10.1016/j.jcis.2008.04.058 © 2008 Elsevier Inc. All rights reserved. taining protein. As an alternative to PLA, copolymer poly(lactideco-ethylene glycol) (PELA) has showed much potentials because it improves the properties of the microspheres products [8,12,13]. Herein, two PELA polymers, diblock copolymers (PLA–mPEG) based on monomethoxypoly(ethylene glycol) (mPEG) and triblock copolymer (PLA–PEG–PLA) based on poly(ethylene glycol) (PEG) were synthesized and used as polymer materials, which was expected to yield high encapsulation efficiency during microspheres fabrication. Particle size is also a crucial parameter in drug release system [14–16]. However, the method for controlling particle size and size distribution is limited. In conventional processes, the double emulsions are prepared by mechanical stirring or homogenization, and thereby the size distribution of microspheres obtained is very broad. We previously reported that preparation of uniform-sized PLA microspheres by direct SPG (Shirasu Porous Glass) membrane emulsification while methylene chloride (MC) was used as organic solvent [17,18]. Typically, the primary emulsion was permeated through the uniform pores of the membrane into the external water phase by the pressure of nitrogen gas to form uniform-sized multiple droplets. However, MC suffers from the drawback of high toxicity. In contrast, ethyl acetate (EA) has much low toxicity and could thus be good alternative [8,19]. Previous studies showed that a high interfacial tension between dispersed phase and membrane 268 Q. Wei et al. / Journal of Colloid and Interface Science 323 (2008) 267–273 surface played important role on the uniform droplet formation [20,21]. Unfortunately, the water solubility of EA (8.7%, w/v) is higher than that of MC (2.0%, w/v) [8], it cannot meet the requirement of direct SPG membrane technique for preparation of uniform emulsion with relatively hydrophilic dispersed phase. Therefore, it is still highly expected to develop a method to prepare uniform-sized microspheres with good applicability to relatively hydrophilic materials. Premix membrane emulsification would be a promising technique due to the high transmembrane flux [22]. The oil phase containing polymers does not have sufficient time to spread on the membrane surface. Thus, in present study, a novel method combining double emulsion-extraction and premix membrane emulsification was proposed. Briefly, preparation of coarse double emulsions was followed by additional premix membrane emulsification with proper pressure, and protein-loaded microspheres were obtained by further solidification. EA was employed as organic solvent and PLA was also chosen as polymer for comparison with PELA. The recombinant hepatitis B surface antigen (HBsAg) was selected as a prototype to testify the encapsulation efficiency [23]. The effect of volume ratio of oil phase and external water phase on size distribution of the microspheres was investigated. Flow cytometer analysis and confocal laser scanning microscope were occupied to compare the preparation method of primary emulsion. The effects of polymer composition and solidification technique on the encapsulation efficiency of microspheres were also investigated and explained in details. 2. Materials and methods 2.1. Materials PLA with a molecular weight (Mw ∼ 40,000 kDa) was purchased from the Institute of Medical instrument (Shandong, China). Monomethoxypoly(ethylene glycol) (mPEG 2000, Mw = 2000 Da) was obtained form Fluka (Switzerland). Poly(ethylene glycol) (PEG 2000, Mw = 2000 Da) was purchased from Bio Basic Inc. (Canada). Stannous octoate was ordered from Sigma (Germany). Poly(vinyl alcohol) (PVA-217, degree of polymerization 1700, degree of hydrolysis 88.5%) was provided by Kuraray (Japan). Hepatitis B surface antigen (Mw ∼ 24 kDa) was kindly supplied by Hualan Biological Engineering Incorporation (Xinxiang, China) and concentrated to 1.5 mg/ml. Fluorescein isothiocyanate (FITC) was purchased from Sigma–Aldrich Inc. (Germany). SPG membrane (pore size of the membrane was 5.2 μm) was provided by SPG Technology Co. Ltd. (Japan). All other reagents were of analytical grade. PELA copolymers (PLA–mPEG and PLA–PEG–PLA) were synthesized according to the procedure reported by Lucke et al. [24], in which the mPEG block and PEG block have a same molecular weight of 2000 Da. The Mw, Mn and Mw/Mn of copolymers were evaluated by gel permeation chromatography. Mean number molecular weight (Mw) of copolymers PLA–mPEG and PLA– PEG–PLA were 43,000 Da and 39,000 Da and the polydispersity (Mw/Mn) were 1.63 and 1.49, respectively. 2.2. Preparation of microspheres Microspheres loaded with antigen were prepared by two-step procedure. Briefly, the coarse double emulsions were first prepared, 0.4 ml HBsAg aqueous solution was mixed with 4 ml ethyl acetate containing polymers (200 mg) by sonication (Branson Sonicator) for 15 s in an ice bath to form primary emulsion, and homogenization with 6000 rpm for 15 s (IKA homogenizer) was used as comparison. The W/O was further emulsified into external aqueous phase containing 1% (w/v) PVA and 1% (w/v) NaCl by magnetic stirring for 60 s at 300 rpm to prepare coarse double emulsions. The Fig. 1. Schematic diagram of a miniature kit for premix membrane emulsification. coarse double emulsions were then poured into the premix reservoir. As schematized in Fig. 1, double emulsions with smaller and relatively uniform size were achieved by extruding the coarse double emulsions through the SPG membrane with a high pressure of 300 kPa. The microspheres were obtained by different solidification technique. In direct solidification technique, the obtained double emulsion droplets were poured quickly into 800 ml solution containing 1% (w/v) NaCl (solidification solution) under magnetic stirring for 5 min to solidify the microspheres. In step-solidification technique, the obtained double emulsion droplets were preliminarily solidified by a little amount volume (20 or 40 ml) of presolidification solution containing 1% (w/v) PVA and 1% (w/v) NaCl for 1 min, and then were completely solidified by 800 ml solution containing 1% (w/v) NaCl (solidification solution) under magnetic stirring for 5 min. The obtained microspheres were collected by centrifugation and washed with distilled water for three times and then lyophilized for 2 days. Unless specified, the formulation conditions were as follows: polymer was PLA–mPEG, primary emulsion was prepared by sonication technique and microspheres were solidified by direct solidification technique. 2.3. Characterization of microspheres The surface morphology of microspheres was observed by a JSM-6700F (JEOL, Japan) scanning electron microscope (SEM). The volume-mean diameter of microspheres was measured by laser diffraction using Mastersizer 2000 (Malvern, UK). The particle size distribution was expressed as a coefficient of variation (CV) value, which is defined as: CV = n (di − d̄)2 i =1 N 1/ 2 d̄, where di is the diameter of the ith particle, d̄ is the numberaverage diameter and N is the total number of particles counted. 2.4. Flow cytometer analysis of antigen distribution in microspheres FITC was covalently linked to HBsAg in a way as described previously [25]. Briefly, a solution of FITC (5 mg/ml) was slowly added into HBsAg solution. After an incubation period of 12 h at 4 ◦ C, unbound FITC was separated by ultrafiltration. The FITCHBsAg loaded microspheres and blank microspheres were added into phosphate buffered saline (PBS) to form a suspension with a concentration of 2.0 × 106 microspheres/ml. The samples were Q. Wei et al. / Journal of Colloid and Interface Science 323 (2008) 267–273 Fig. 2. Effects of volume ratio of oil phase to external water phase on size distribution. analyzed on FL1-H channel by a LSR flow cytometer (Becton Dickinson). 2.5. CLSM imaging of microspheres The microspheres samples prepared for flow cytometer analysis were also observed by CLSM (Confocal Laser Scanning Microscope). Suspensions containing microspheres in Petri dish were observed by a TCS SP2 CLSM (Leica). The transmitted image was taken. 2.6. Turbidity determination of primary emulsion The stability of primary emulsion was indicated by its turbidity. The turbidity of primary emulsion was determined according to a method previously described [26]. Briefly, the primary emulsions were prepared by different polymers and added into a quartz cell and absorbance was then measured at 620 nm with a spectrophotometer as a function of time. The turbidity was calculated from standard absorbance–turbidity curve with formazin suspension as standards. Formazin was prepared by reacting hydrazine sulfate with hexamethylenetetrammonium, and standards of formazin turbidity units (FTU) was prepared by appropriate dilution. 2.7. Measurement of protein encapsulation efficiency The total antigen loaded efficiency in PELA microspheres was determined by dissolving twenty milligrams of the freeze-dried microspheres in 1 ml of 5% (w/v) sodium dodecyl sulfate (SDS) containing 0.1 M sodium hydroxide solution. The amount of antigen was determined by micro bicinchoninic acid (micro-BCA) assay (Pierce, USA). Blank microspheres were used as control. The encapsulation efficiency of antigen was calculated by the following formula: EE = total amount of antigen loaded total amount of antigen added × 100%. 3. Results and discussion In premix membrane emulsification, fine double emulsions could be obtained by disrupting the coarse double emulsions with SPG membrane. The effects of volume ratio of oil phase and external water phase on particle size and size distribution were investigated in present study. As shown in Fig. 2, with the increase of external water phase volume, the obtained microspheres showed a bigger size and broader size distribution (SEM was not shown). The viscosity of the coarse double emulsions could be a good explanation for these results. When external water phase volume 269 Fig. 3. Comparison of the preparation method of primary emulsion: (a) sonication technique, (b) homogenization technique and (c) corresponding flow cytometer analysis. increased, a larger amount of EA diffused from the oil phase into the external water due to its good solubility in water (8.7%, w/v). Consequently, the coarse double emulsions became more viscous and difficult to be disrupted into fine double emulsions, which affected the uniformity of the obtained double emulsions. The optimum volume ratio of oil phase and external water phase (1:6) was performed in the following study. The corresponding particle size of the obtained microspheres was about 1 μm and CV value was 18.9%. When the size is in particular ranges, the microspheres possess passive targeting property. What’s more, microspheres with a diameter under 5 μm would be ideal for passive targeting of antigen presenting cells [14–16], and induce high humoral immunization response. However, preparation of microspheres with such small particle requires an efficient method to prepare stable primary emulsion. It is well known that reducing droplets size of the primary emulsion is known as one promising approach to improve the stability of double emulsion [27]. Herein, homogenization and sonication were occupied to testify their effects on the primary emulsion. The antigen was labeled with FITC and then encapsulated into the microspheres. During the solvent-extraction process, the microspheres were solidified so rapidly that inner droplets did not have sufficient time to coalesce. Thus, the interior structure of microspheres almost could represent the original distribution of inner droplets within oil droplets. As shown in Figs. 3a and 3b, the inner droplets of primary emulsion prepared by sonication were small and evenly dispersed in the microsphere, whereas the size of inner droplets was very large and could not be effectively encapsulated while using homogenization technique. Correspondingly, the signal of FITC labeled antigen showed a sharp peak for the microspheres fabricated by sonication during flow cytometer analysis, whereas a broad distribution for the microspheres prepared by homogenization and the obtained signal was partly overlapped with the signal of the blank microspheres (Fig. 3c), indicating that some of the microspheres prepared by homogenization did not encapsulate any FITC labeled antigen and that the antigen was loaded uniformly in each microsphere under the sonication process. From these two aspects, sonication was an adequate technique and adopted in following study. 3.1. Effects of polymer composition The morphologies of microspheres fabricated by premix membrane emulsification were visualized by SEM and shown in Fig. 4. Relatively uniform-sized microspheres were successfully prepared 270 Q. Wei et al. / Journal of Colloid and Interface Science 323 (2008) 267–273 Fig. 5. The stability of primary emulsion of different polymers. Fig. 4. SEM photographs of microspheres prepared by different polymers: (a) PLA, (b) PLA–PEG–PLA and (c) PLA–mPEG. Table 1 Preparative parameters and properties of corresponding microspheres Batch Polymer Pre-solidification solution volume (ml) Particle size (μm) Encapsulation efficiency (%) A B C D E PLA PLA–PEG–PLA PLA–mPEG PLA–mPEG PLA–mPEG 0 0 0 40 20 1.39 ± 0.05 1.21 ± 0.07 1.13 ± 0.03 1.22 ± 0.05 1.33 ± 0.07 61.3 ± 3.7 75.1 ± 5.1 90.4 ± 4.3 83.9 ± 3.9 77.3 ± 4.1 by different polymers. The CV values of microspheres were 21.5, 24.1 and 18.9% for PLA, PLA–PEG–PLA and PLA–mPEG, respectively. Compared with smooth surface of PLA microspheres, the PLA–PEG– PLA microspheres exhibited pitted and porous surface, whereas the PLA–mPEG yielded microspheres with rough appearances and numerous saliences. Irregular surface structures were observed similarly in the case of microspheres prepared with another amphiphilic copolymer, such as PLA–mPEO [28]. The heterogeneous surface of microspheres might be formed due to different distribution pattern of PEG segment in copolymers [7], which could be seen from Fig. 1 in Supporting materials. As shown in Table 1, the PELA microspheres possessed higher encapsulation efficiency as compared to PLA counterpart. The result can be ascribed to the improvement of affinity between polymer and antigen. Interestingly, there is still an evident difference between the PELA copolymers. Although PLA–mPEG and PLA–PEG– PLA possessed the same content of PEG segment, the PLA–mPEG microspheres yielded higher encapsulation efficiency (90.4%) than that of PLA–PEG–PLA microspheres (75.1%). To our knowledge, limited study has performed on detailed mechanism about the results. We supposed that these might be affected by the different distribution pattern of PEG segments in polymers and these would be discussed in present study. It was known that the encapsulation efficiency of microspheres was strongly affected by the stability of the primary emulsion during the double emulsion process [29,30]. Coalescence between the inner droplets within primary emulsion leads to separation of oil phase and water phase. The stability of the primary emulsion was measured by the time required for phase separation [31]. However, the time was not easy to be determined. In present study, we proposed a more quantitative method to evaluate the stability by measuring turbidity of the primary emulsion. The turbidity decreased with the growth of inner droplets by coalescence and was easily monitored by spectrophotometer. The turbidity changes of primary emulsion fabricated by different polymers as a function of time were investigated and are shown in Fig. 5. All primary emulsions had a same initial turbidity, but the turbidity of the primary emulsion prepared by PLA decreased dramatically from 250 to 81 FTU in the initial 10 h, whereas the turbidity of the PLA– PEG–PLA copolymer decreased slightly to 219 FTU in the initial 10 h and the turbidity of the PLA–mPEG copolymer was kept constant during 24 h, indicating that PELA could improve the primary emulsion stability. These results were in good agreements with encapsulation efficiency of different polymers. When PELA was employed in double emulsion process, the hydrophilic PEG segments would extend into the water phase and hydrophobic PLA segments would spread into the oil phase. As illustrated in Fig. 6, in the present case of PELA, the polymer itself would form an interfacial layer and act as surfactant because of its amphiphilic character, thereby stabilizing the primary emulsion. Surfactants could improve the stability of the primary emulsion [30,32], but should be limited to the minimum level to avoid possible toxic [1,33]. On the other hand, distribution pattern of PEG segment would also play important role on the properties of the interface between oil phase and water phase. As shown in Fig. 6, the PLA–mPEG would form a more stable interfacial layer than that of PLA–PEG–PLA probably due to spatial hindrance of triblock copolymer [29], thus led to a more stable primary emulsion. Q. Wei et al. / Journal of Colloid and Interface Science 323 (2008) 267–273 271 Fig. 6. Schematic illustration of polymer distribution on the interface between internal phase (W 1 ) and oil phase. Fig. 7. Schematic illustration of the coalescence process between inner droplets (W 1 ) and external aqueous phase (W 2 ) (oil droplet was dispersed in external aqueous phase W 2 ). The encapsulation efficiency of microspheres was determined by the protein lose in microspheres fabrication. How the protein lost during double emulsion-solvent extraction method has not been studied thoroughly. A reasonable explanation was that the loss of encapsulated protein during the fabrication process essentially from the fact that the protein in the inner droplets tend to merge with external water phase [7]. Herein, we speculated that an oil film prevented the coalescence of the inner droplets and external aqueous phase. As illustrated in Fig. 7, inner droplets moved to the interface of oil phase and external water phase and separated from external water phase by a thin oil film. Once the oil film was evacuated, the inner droplets would merge into the external water phase rapidly and disappear, resulting in the loss of protein. It would be an effective strategy to yield high encapsulation by restraining the evacuation of the oil film separating the inner droplets and external water phase. When PELA was used, the PLA segment was anchored at the interface due to interfacial layer (Fig. 6), thus the evacuation of oil film therein would be restrained, and higher encapsulation efficiency of antigen is an expected consequence. As mentioned above, the interfacial layer of PLA–mPEG was more stable than that of PLA–PEG–PLA. The properties of oil film within PLA–mPEG copolymer was more suitable for protecting the coalescence of inner droplets and external water phase, led to a higher encapsulation efficiency of antigen. 3.2. Effects of solidification technique In order to control the solidification rate, the process of solidification was modified with a procedure of additional presolidification. The double emulsions obtained by premix membrane emulsification were preliminarily partially-solidified by a lit- Fig. 8. SEM photographs of microspheres prepared by different solidification techniques: (a) batch C, (b) batch D and (c) batch E. tle amount aqueous solution and then completely solidified. PLA– mPEG was selected as polymer, and the detailed preparative parameters and properties of the microspheres are listed in Table 1. The solidification rate of microspheres fabricated by direct solidification technique is faster than those of step-solidification techniques. Among the step-solidification techniques, the solidification rate of microspheres would increase with the increase of pre-solidification aqueous solution volume. Thus, batch C > batch D > batch E in terms of solidification rate of obtained microspheres. The surface morphologies of microspheres were observed by SEM and shown in Fig. 8. All the microspheres revealed rough appearances, which was more developed with the increase of solidification rate. As shown in Table 1, the particle size of microspheres fabricated by direct solidification technique had a lightly smaller size compared with those of step-solidification techniques. Solidification rate is known to directly influence the internal structure of microspheres [34]. The coalescence of inner droplets within oil droplets might form interconnecting channels inside of the obtained microspheres [8]. In present study, rapid solidification would reduce the coalescence time, and thus led to a more compact structure. To represent their effects on internal structure of 272 Q. Wei et al. / Journal of Colloid and Interface Science 323 (2008) 267–273 4. Summary Relatively uniform-sized amphiphilic PELA microspheres were prepared rapidly and conveniently by a novel method combining emulsion-solvent extraction and premix membrane emulsification while EA was used as organic solvent. Under the optimum condition, the particle size was about 1 μm and the CV value was 18.9%. Confocal laser scanning and flow cytometer analysis showed that the inner droplets were small and evenly dispersed and the antigen was loaded uniformly in each microsphere when sonication technique was occupied to prepare primary emulsion. Effects of polymer composition and solidification technique on encapsulation efficiency were also investigated in this study. Improving the properties of polymers and shorting the time required for solidification of microspheres were found as two effective strategies to yield high encapsulation of antigen. Distribution pattern of PEG segment of polymers played important role on the properties of microspheres. Compared with triblock copolymer PLA–PEG–PLA, the diblock copolymer PLA–mPEG yielded a more stable interfacial layer at the interface of oil and water phase, and thus was more suitable to stabilize primary emulsion and protect coalescence of inner droplets and external water phase, resulting in high encapsulation efficiency (90.4%). On the other hand, solidification technique determined solidification time during microsphere fabrication, and thus affected coalescence time. The encapsulation efficiency increased with the increase of solidification rate of microspheres. These uniform-sized PELA microspheres prepared by this novel method might have great potential as protein drug carrier. Acknowledgments We thank the financial support of the National Nature Science Foundation of China (20536050, 20221603 and 20636010) and Chinese Academy of Sciences (KJCX2.YW.M02). Supplementary material Supplementary material for this article may be found on ScienceDirect, in the online version. Please visit DOI: 10.1016/j.jcis.2008.04.058. Fig. 9. SEM photographs of microspheres after incubation in PBS for 70 days: (a) batch C, (b) batch D and (c) batch E. References microspheres, the morphologies of microspheres fabricated by different solidification techniques after incubation in PBS buffer solution at 37 ◦ C for 60 days were observed and shown in Fig. 9. All the microspheres were in the erosion phase, microspheres fabricated by direct solidification technique (batch C) showed incomplete erosion owing to denser structure of microspheres, whereas microsphere prepared with slow solidification technique (batch E) exhibited extensive erosion due to comparatively looser structure of microspheres. The encapsulation efficiencies of batch C, batch D and batch E, were 90.4, 83.9 and 77.3%, respectively. Obviously, the encapsulation efficiency was increased with the increase of solidification rate of microspheres. Based on the explanation for encapsulation efficiency during the microspheres fabrication shown in Fig. 7, it is easily to understand the difference among three solidification techniques. The evacuation of oil film separating inner droplets and external aqueous phase would be restrained if the solidification rate was faster. Thus, the batch C yielded a highest encapsulation efficiency of antigen among the three batches due to its fastest solidification rate. [1] H. Tamber, P. Johansen, H.P. Merkle, B. Gander, Adv. Drug Delivery Rev. 57 (2005) 357. [2] L.Y. Wang, G.H. Ma, Z.G. Su, J. Controlled Release 106 (2005) 62. [3] J.J. Wang, K.M. Chua, C.H. Wang, J. Colloid Interface Sci. 271 (2004) 92. [4] P. Couvreur, M.J. Blanco-Prieto, F. Puisieux, B. Roques, E. Fattal, Adv. Drug Delivery Rev. 28 (1997) 85. [5] R.A. Jain, Biomaterials 21 (2000) 2475. [6] Y. Ogawa, M. Yamamoto, S. Takada, H. Okada, T. Shimamoto, Chem. Pharm. Bull. 36 (1988) 1502. [7] G. Ruan, S.S. Feng, Q.T. Li, J. Controlled Release 84 (2002) 151. [8] F.T. Meng, G.H. Ma, W. Qiu, Z.G. Su, J. Controlled Release 91 (2003) 407. [9] D. Chognot, J.L. Six, M. Leonard, F. Bonneaux, C. Vigneron, E. Dellacherie, J. Colloid Interface Sci. 268 (2003) 441. [10] D. Chognot, M. Léonard, J. Six, E. Dellacherie, Colloids Surf. B 51 (2006) 86. [11] X.M. Deng, X.H. Li, M.L. Yuan, C.D. Xiong, Z.T. Huang, W.X. Jia, Y.H. Zhang, J. Controlled Release 58 (1999) 123. [12] S.B. Zhou, X.Y. Liao, X.H. Li, X.M. Deng, H.M. Li, J. Controlled Release 86 (2003) 195. [13] J.J. Cheng, B.A. Teply, I. Sherifi, J. Sung, G. Luther, F.X. Gu, E. Levy-Nissenbaum, A.F. Radovic-Moreno, R. Langer, O.C. Farokhzad, Biomaterials 28 (2007) 869. [14] S. Akhtar, K.J. Lewis, Int. J. Pharm. 151 (1997) 57. [15] C. Evora, I. Soriano, R.A. Rogers, K.M. Shakesheff, J. Hanes, R. Langer, J. Controlled Release 51 (1998) 143. [16] S.R. Little, D.M. Lynn, Q. Ge, D.G. Anderson, S.V. Puram, J. Chen, H.N. Eisen, R. Langer, Proc. Natl. Acad. Sci. USA 101 (2004) 9534. [17] R. Liu, G.H. Ma, F.T. Meng, Z.G. Su, J. Controlled Release 103 (2005) 31. [18] R. Liu, G.H. Ma, Y.H. Wan, Z.G. Su, Colloids Surf. B 45 (2005) 144. Q. Wei et al. / Journal of Colloid and Interface Science 323 (2008) 267–273 [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] C. Sturesson, J. Carlfors, J. Controlled Release 67 (2000) 171. Q.Z. Zhou, L.Y. Wang, G.H. Ma, Z.G. Su, J. Colloid Interface Sci. 311 (2007) 118. J. Wu, W. Wei, L.Y. Wang, Z.G. Su, G.H. Ma, Colloids Surf. B, in press. G.T. Vladisavljević, M. Shimizu, T. Nakashima, J. Membr. Sci. 244 (2004) 97. L. Feng, X.R. Qi, X.J. Zhou, Y. Maitani, S.C. Wang, Y. Jiang, T. Nagai, J. Controlled Release 112 (2006) 35. A. Lucke, J. Teßmar, E. Schnell, G. Schmeer, A. Göpferich, Biomaterials 21 (2000) 2361. A.H. Krauland, D. Guggi, A. Bernkop-Schnürch, Int. J. Pharm. 307 (2006) 270. J. Wu, Z.G. Su, G.H. Ma, Int. J. Pharm. 315 (2006) 1. N. Garti, A. Aserin, Adv. Colloid Interface Sci. 65 (1996) 37. P. Bouillot, N. Ubrich, F. Sommer, T.M. Duc, J. Loeffler, E. Dellacherie, Int. J. Pharm. 181 (1999) 159. 273 [29] Y.Y. Yang, J.P. Wan, T.S. Chung, P.K. Pallathadka, S. Ng, J. Heller, J. Controlled Release 75 (2001) 115. [30] G. De Rosa, R. Iommelli, M.I. La Rotonda, A. Miro, F. Quaglia, J. Controlled Release 69 (2000) 283. [31] N. Nihant, C. Schugens, C. Grandfils, R. Jérôme, P. Teyssié, Pharm. Res. 11 (1994) 1479. [32] K. Pays, J. Giermanska-Kahn, B. Pouligny, J. Bibette, F. Leal-Calderon, Langmuir 17 (2001) 7758. [33] O. Lambert, O. Nagele, V. Loux, J. Bonny, L. Marchal-Heussler, J. Controlled Release 67 (2000) 89. [34] R. Jeyanthi, B.C. Thanoo, R.C. Metha, P.P. DeLuca, J. Controlled Release 38 (1996) 235.