- Email: [email protected]
Chitosan-coated magnetic nanoparticles as carriers of 5-Fluorouracil: Preparation, characterization and cytotoxicity studies
Colloids and Surfaces B: Biointerfaces 68 (2009) 1–6
Contents lists available at ScienceDirect
Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb
Chitosan-coated magnetic nanoparticles as carriers of 5-Fluorouracil: Preparation, characterization and cytotoxicity studies Longzhang Zhu, Jingwei Ma, Nengqin Jia, Yu Zhao, Hebai Shen ∗ Life and Environment Science College, Shanghai Normal University, Shanghai 200234, PR China
a r t i c l e
i n f o
Article history: Received 3 April 2008 Received in revised form 19 July 2008 Accepted 22 July 2008 Available online 15 August 2008 Keywords: Chitosan 5-Fluorouracil Magnetic nanoparticles Carrier Fluorescence
a b s t r a c t The chitosan-coated magnetic nanoparticles (CS MNPs) were prepared as carriers of 5-Fluorouracil (CS–5Fu MNPs) through a reverse microemulsion method. The characteristics of CS–5-Fu MNPs were determined by using transmission electron microscopy (TEM), FTIR spectroscopy and vibrating-sampling magnetometry (VSM). It was found that the synthesized CS–5-Fu MNPs were spherical in shape with an average size of 100 ± 20 nm, low aggregation and good magnetic responsivity. Meanwhile, the drug content and encapsulation rate of the nanoparticles was 16–23% and 60–92%, respectively. These CS–5-Fu MNPs also demonstrated sustained release of 5-Fu at 37 ◦ C in different buffer solutions. The cytotoxicity of CS–5Fu MNPs towards K562 cancer cells was investigated. The result showed that CS–5-Fu MNPs retained significant antitumor activities. Additionally, it was observed that the FITC-labeled CS–5-Fu MNPs could effectively enter into the SPCA-1 cancer cells and induced cell apoptosis. © 2008 Published by Elsevier B.V.
1. Introduction Nanoparticles, with highly controlled shapes, sizes and some interesting properties (such as optical and magnetic properties), have been studied extensively as drug carriers which can improve the bioavailability of drug with poor absorption characteristics [1]. In recent years, much attention has been focused on the natural and synthetic polymers such as poly -caprolactone (PCL), polylactide (PLA) and poly(lactide-co-glycolide) (PLGA) due to their good biocompatibility, biodegradability and ability to prolong drug release [2]. Among them, chitosan (CS) is a favorable type of drug deliver system. CS is a natural linear biopolyaminosaccharide derived from alkaline deacetylation of chitin, which has been found to be the second most abundant natural biopolymer in nature behind only cellulose. Because of its chemical structure, CS has received increasing attention as a renewable polymeric material and has now been widely used in many fields such as protein adsorption [3] and metal adsorption [4]. Additionally, CS and its derivative have also been investigated in the development of controlled release drug delivery systems [5–7], since CS’s mucoadhesive property can enhance drug transmucosal absorption [8] and promote sustained release of drug [9]. Nowadays, 5-Fluorouracil (5-Fu) has proved to be one of the effective chemotherapeutic anticancer drugs [10]. Generally, it is
∗ Corresponding author. Tel.: +86 21 64321800; fax: +86 21 54242640. E-mail address: [email protected] (H. Shen). 0927-7765/$ – see front matter © 2008 Published by Elsevier B.V. doi:10.1016/j.colsurfb.2008.07.020
administered either orally through feeds and drinking water or through injections for several times a day [11]. However, one of the limitations inherent in current cancer chemotherapies is the lack of specificity of anticancer drug delivery. Generally, most anticancer drugs appear to have serious cytotoxic effects on normal cells [12]. Therefore, an important approach in improving treatment effect is to better harness the potency of chemotherapeutic agents by more effectively targeting them to tumor tissues [13]. With the rapid development of nanotechnology, magnetic nanoparticles, especially superparamagnetic iron oxide nanoparticles, are currently being widely studied and have been found numerous applications in the fields of cell separation [14], cell apoptosis [15] and enzyme immobilization [16]. Magnetic targeted nanoparticles can be used as drug carriers to provide targeted delivery and sustained release of chemotherapeutic agents to improve bioavailability. Drug based on carriers of core-shell magnetic nanoparticles can be easily guided to arrive at the interest position in the body by means of physical force from magnetic field [17]. Meanwhile, the outer shell (polymer layer) of the drug can effectively slow down the rate of release. Therefore, drug delivery system using polymer-coated magnetic nanoparticles is considered as an effective strategy for passive tumor targeting, which can not only increase drug circulation but also reduce pain in patients. However, limited literatures reported the direct application of chitosan-coated magnetic nanoparticles as drug carriers. In this study, we prepared one kind of novel nano-scale carrier for 5-Fu, using ␥-Fe2 O3 nanoparticles and 5-Fu as core and CS as a polymeric shell to form drug-loaded magnetic nanoparticles.
2
L. Zhu et al. / Colloids and Surfaces B: Biointerfaces 68 (2009) 1–6
The morphology, structure and characteristic of the CS–5-Fu MNPs were studied by TEM, FTIR and VSM. The application of CS MNPs as carriers of 5-Fu was evaluated by measuring its drug content, encapsulation rate and cytotoxicity in vitro. We also examined the apoptosis of SPCA-1 cancer cells induced by FITC-labeled CS–5-Fu MNPs.
rate of CS–5-Fu MNPs were calculated by the following formulas:
2. Materials and methods
2.5. In vitro drug release experiment
2.1. Materials
Drug release behavior was studied in phosphate buffered saline solution at three different pH values (PBS, 0.2 mol/L, pH 7.4, 6.8 and 1.2, respectively). CS–5-Fu MNPs (2 mg/ml), dispersed in PBS solutions, were sealed in dialysis bag (MW cutoff: 12–16 kDa, Yuanju Bio-tech Co. Ltd., Shanghai, China). The dialysis bag was incubated in 50 ml of different incubation mediums at 37 ◦ C and gently shaken at 100 cycles/min. At predetermined time intervals, 1 ml of the incubation medium was collected for 5-Fu quantitative analysis by spectrophotometry and meanwhile, 1 ml of fresh medium was compensated. For comparison, the release of free 5-Fu particles was also performed in PBS 7.4 (pH 7.4).
5-Fluorouracil was supplied by Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Chitosan (deacetylation degree 90%, MW = 60 kDa) was purchased from Wenming Biochemistry Science and Technology Co., Ltd. (Shanghai, China). Fluorescein isothiocyanate (FITC) was obtained from Sigma. RPMI-1640 medium was purchased from Sino-American Biotechnology Company. K562 cells and SPCA-1 cells were both obtained from Cell Bank of Chinese Academy of Sciences. All other reagents were analytical grade and used directly without any purification. Double distilled water was used for all the experiments.
Drug content =
W1 × 100%, W2
Encapsulation rate =
W1 × 100% W3
where W1 is the weight of 5-Fu encapsulated in the nanoparticles, W2 is the gross weight of the nanoparticles, and W3 is the total weight of the feeding 5-Fu.
2.6. Assay of in vitro cytotoxicity 2.2. Preparation of CS–5-Fu MNPs The ␥-Fe2 O3 magnetic nanoparticles were prepared by our reported method [18]. CS–5-Fu MNPs were prepared by a reverse microemulsion method. Briefly, CS powders (20 mg) were dissolved in 1 ml of 1% (w/v) acetic acid solution containing magnetite suspension and then a predetermined amount of 5-Fu (5–10 mg) was added to the CS solution. The solution was then dropwised into 50 ml of fluid wax containing emulsifier (Span-80) in a 100ml three-necked flask at 55 ◦ C. A water-in-oil microemulsion was formed by continuous stirring at 800 rmp/min for 12 h in a water bath. Then, 1 ml of 20% (w/w) sodium citrate solution was gradually added to the flask and the reaction was still kept the same condition for 1 h. After reaction, CS–5-Fu MNPs were collected with a permanent magnet and rinsed consecutively with light petroleum and isopropanol for three times. Finally, the obtained nanoparticles were dried overnight at 65 ◦ C. The preparation of CS MNPs (not adding 5-Fu) was carried in a similar way mentioned above.
2.3. Characterizations of CS–5-Fu MNPs The morphology and size of the CS–5-Fu MNPs were observed and estimated by TEM (H-600, HITACHI, Japan). The sample for TEM analysis was obtained by placing a drop of the CS–5-Fu MNPs dispersed aqueous solution onto a copper micro-grid and evaporated in air at room temperature. FTIR spectroscopy of nanoparticles was obtained by using a FTIR spectrophotometer (AVATAR 370, Nicolet, USA). The magnetization of the prepared nanoparticles was measured on a VSM (HH-15, Nanjing University Instrument Plant, China) at room temperature.
2.4. Determination of 5-Fu content in the nanoparticles CS–5-Fu MNPs (10 mg) were distributed in 50 ml of 1 mol/L HCL by sonication. After 5 h, the supernatant was collected with centrifugation and magnetic separation. The concentration of 5-Fu in the supernatant was assayed by UV–vis spectrophotometer (UV–vis 8500, Techcomp, China) at 265 nm and the supernatant from CS MNPs was used as a contrast. The drug content and encapsulation
The MTT [3-(4,5-dimethylthazol-2-yl)-2,5-diphenyltetrazolium bromide blue-indicator dye]-based assay is a simple nonradioactive colorimetric assay to measure cell cytotoxicity, proliferation or viability [19]. K562 cells were used for the analysis of cytotoxicity in vitro. The cells (1 × 105 cells/ml) were placed into 96-well tissue-culture plates and incubated at 37 ◦ C. After 24 h, cells were treated with three different concentrations of CS–5-Fu MNPs (5 g/l, 2.5 g/l, 1.25 g/l, respectively) and CS MNPs (5 g/l), respectively. Untreated cells were used as controls. Plates were incubated in a humidified 5% CO2 balanced-air incubator at 37 ◦ C for 24 h, 48 h, 72 h and 96 h, respectively. Then 30 l of 5 mg/ml MTT solution was added to each well and the plates were incubated for another 4 h, then to each well was added 30 l DMSO (dimethyl sulfoxide). The results were read on ELISA plate reader (A5002, Tecan, Austria), using a wavelength of 570 nm and the cell apoptosis rate was calculated by the following equation: cell apoptosis = 1 −
T × 100% C
where C is the number of viable cells after 24 h, 48 h, 72 h, 96 h of incubation without nanoparticles and T is the number of viable cells after 24 h, 48 h, 72 h, 96 h of incubation with nanoparticles. 2.7. Fluorescence observation on cells uptake of FITC-labeled CS–5-Fu MNPs The fluorescent labeling is one of the most important methods of modem analysis as a nonradioactive labeling technique, especially fluorescently labeled nanoparticles that can provide a rapid, simple and sensitive means to quantify cell-associated nanoparticles by fluorometry [20]. Fluorescein isothiocyanate is often used for microscopy on biological samples. Many studies have investigated the spectral properties of this dye [21,22]. FITC-labeled CS–5-Fu MNP were prepared for the confocal laser scanning microscopy study. Synthesis of the FITC-labeled CS–5-Fu MNP was based on the reaction between the isothiocyanate group of FITC and the primary amino group of CS as reported in the literature [23]. Briefly, 3 mg of FITC in 3 ml of DMSO was added to 10 ml of PBS 7.4 containing 10 mg of CS–5-Fu MNPs. After 12 h reaction in the dark at ambient conditions, the FITC-labeled CS–5-Fu MNPs were collected with a permanent magnet to remove the unconjugated FITC and washed
L. Zhu et al. / Colloids and Surfaces B: Biointerfaces 68 (2009) 1–6
3
Fig. 3. Magnetization vs. magnetic field for the CS–5-Fu MNPs. Fig. 1. TEM image of CS–5-Fu MNPs.
several times with PBS 7.4. Finally, the obtained nanoparticles were dispersed in fresh RPMI-1640. The FITC-labeled CS–5-Fu MNPs were filtered through a 0.22 m membrane for sterilization. Subsequently, 1 ml of nanoparticles (1 mg/ml) was added to SPCA-1 cancer cells (1 × 105 cells/ml), which were pre-cultured on a 35-mm glass dish for 24 h at 37 ◦ C. After incubation for specific time intervals, the fluorescence images were obtained using an confocal laser scanning microscope at ex = 488 nm and emission collected at a range of 510–540 nm. 3. Results and discussion 3.1. Characteristics of CS–5-Fu MNPs A typical TEM micrograph of CS–5-Fu MNPs was shown in Fig. 1. Result suggested that CS–5-Fu MNPs were fairly smooth and spherical in shape, and the average size was 100 ± 20 nm with a narrow distribution (measured by TEM over 500 particles). The FTIR spectra of pure CS, 5-Fu, ␥-Fe2 O3 and CS–5-Fu MNPs were presented in Fig. 2. The spectrum of CS–5-Fu MNPs showed C–N stretching peak in the amide groups at 1395 cm−1 (at 1398 cm−1 in the spectrum
of CS), C–O in the ether groups at 1086 cm−1 (at 1094 cm−1 in the spectrum of CS). Comparing the spectrum of CS–5-Fu MNPs with that of 5-Fu and ␥-Fe2 O3 , it can be seen that the peak at 1603 cm−1 appeared significantly in the 5-Fu encapsulated CS MNPs. This peak was attributed to C O group and can be seen in the pure 5-Fu spectrum (at 1673 cm−1 ). And the peak at 632 cm−1 was assigned to the Fe–O bond vibration of ␥-Fe2 O3 in CS–5-Fu MNPs (at 591 cm−1 in the spectrum of ␥-Fe2 O3 ). Besides, the peak at 3430 cm−1 corresponded to O–H bond vibration in the spectra of CS–5-Fu MNPs. These results indicated that 5-Fu and ␥-Fe2 O3 were both encapsulated in the CS MNPs. The magnetic property of CS–5-Fu MNPs was measured by VSM (result shown in Fig. 3). The saturated magnetization of the CS–5-Fu MNPs was about 3.4 emu/g, which was lower than those of free ␥Fe2 O3 [24]. Although the magnetism has decreased, nanoparticles can still be adsorbed quickly and firmly by the magnet. On the other hand, it is well known that magnetic particles less than 30 nm will demonstrate the characteristic of superparamagnetism [24], which can be verified by the magnetization curve [25]. The remanence (Mr ) and coercivity (Hc ) for CS–5-Fu MNPs in the figure were close to zero, exhibiting the characteristic of superparamagnetism. 3.2. Drug content and encapsulation rate of CS–5-Fu MNPs The drug content and encapsulation rate were determined by varying the feed weight ratio of CS and 5-Fu particles. The loading characteristics of CS–5-Fu MNPs were summarized in Table 1. When the feed weight ratio was 4:1 (sample I), the drug content was above 20%. However, if the feed ratio was 2:1 (sample II), the drug content slightly decreased to 16–18% and meanwhile the excess of 5-Fu could not totally be loaded into CS (encapsulation rate decreased to about 64%). The reason could be, under the condition of maintaining constant amount of drug, the shielding effect of CS skeleton on drug became stronger with the increasing amount of CS. Moreover, it was favorable to coat drug and to prevent drug from escaping into aqueous solution. That is also the reason that encapsulation rate decreased from over 85% to only about 64%. Table 1 The loading characterizations of CS–5-Fu MNPs
Fig. 2. FTIR spectra of (a) CS, (b) 5-Fu, (c) ␥-Fe2 O3 , and (d) CS–5-Fu MNPs.
Sample
Feed weight ratio (CS:5-Fu)
Drug content (%)
Encapsulation rate (%)
I II
4:1 2:1
20–23 16–18
85–92 60–68
4
L. Zhu et al. / Colloids and Surfaces B: Biointerfaces 68 (2009) 1–6
Fig. 5. In vitro cytotoxicity of CS–5-Fu MNPs and CS MNPs in K562 cells after 24 h, 48 h, 72 h, and 96 h.
was possibly because increasing the polymer concentration could form the thicker polymer wall, which effectively prevented 5-Fu from releasing in buffer. It could also be seen from Fig. 4(a) that no matter which kinds of feed ratio of nanoparticles could continuously release the drug. These results indicated that CS–5-Fu MNPs could be a good candidate for drug carriers. Compared to the release of drug in PBS 6.8 (pH 6.8, simulated intestinal fluid) and PBS 1.2 (pH 1.2, simulated gastric fluid), CS–5Fu MNPs (feed ratio of 4:1) released only 5% of the total loaded 5-Fu in PBS 7.4 in the first 1 h in Fig. 4(b). The release behavior of these nanoparticles, however, remained relatively less sustainable than those in two other buffers when time was elongated. Considering that 5-Fu belongs to acidic drug, it can be ionized to anions under the alkaline condition and thus its solubility will possibly be significantly improved. Therefore, at the end of 82 h, there was approximate 82% amount of loaded 5-Fu released in PBS 7.4 in comparison with 57% and 67% in other two buffers, respectively. Fig. 4. In vitro release of 5-Fu from CS–5-Fu MNPs (a) with different feed weight ratios in PBS 7.4, and (b) with feed weight ratio of 4:1 in different buffers.
3.3. In vitro 5-Fu release experiment In order to evaluate the potential of employing CS MNPs as carriers of 5-Fu, the release behaviors of 5-Fu from CS MNPs were assessed at 37 ◦ C in PBS at pH 7.4, 6.8 and 1.2, respectively. In these experiments, the conditions were maintained by regularly replacing the incubation medium. The in vitro release profile of CS–5-Fu MNPs was illustrated in Fig. 4. Fig. 4(a) displayed the release behaviors of nanoparticles with feed weight ratios of 2:1 and 4:1 in PBS 7.4. It showed that 5-Fu was released to the extent of 99% within 2 h, but CS–5-Fu MNPs can be sustainably released for 80 h. The result that more than 80% of 5-Fu was released for 80 h suggested the potential of the nanoparticles as a sustained drug delivery system. A small (about 23% of the total loaded drug) burst release occurred in the first 2 h for the nanoparticles with feed ratio of 2:1, and then followed by a very slow drug release. Zhou et al. suggested that the drug release involved two different mechanisms of drug molecules diffusion and polymer matrix degradation [26]. The burst release was probably related with the excess drug particles (close to the CS surface) dispersing rapidly from matrix into buffer in the first few hours. Because of drug concentration decreasing, most drug in the nanoparticles with feed ratio of 4:1 was relatively far from the CS surface. This might be one of the reasons for no obvious burst release of drug for such nanoparticles. In addition, it
3.4. In vitro cytotoxicity K562 cells were incubated with CS–5-Fu MNPs and CS MNPs, respectively, to test that the cytotoxicity was caused by 5-Fu and not by the CS MNPs. As shown in Fig. 5, drug continually released from the matrix accelerated cell apoptosis and the apoptosis ten-
Fig. 6. Comparison of fluorescence emission spectra between FITC-labeled CS–5-Fu MNPs (a) and free FITC (b).
L. Zhu et al. / Colloids and Surfaces B: Biointerfaces 68 (2009) 1–6
5
Fig. 7. Confocal microscopy images of SPCA-1 cells incubated with FITC-labeled CS–5-Fu MNPs after (a) 2 h, (b) 24 h, and (c) 48 h.
dency was more obvious when the incubation time continued. This phenomenon indirectly suggested that the nanoparticles can effectively prolong drug release. Cell apoptosis rate was increased from 31% after 24 h to 87% at the end of 96 h induced by CS–5Fu MNPs with the same concentration at 5 g/l. It was also observed that cell apoptosis rates were decreased at lower concentrations of nanoparticles. That is due to lower concentrations of drug loaded into nanoparticles. Surprisingly, when incubated with control nanoparticles, the cell apoptosis rate also reached about 20%. This may be because free CS itself can induce cancer cell apoptosis in some extent [27]. As expected, the cytotoxicity was caused mainly by 5-Fu itself and CS MNPs only played a limited role in cell apoptosis. 3.5. Cellular uptake of FITC-labeled CS–5-Fu MNPs The fluorescence peak of FITC-labeled CS–5-Fu MNPs was measured by fluorescence spectrophotometer (Cary Eclipse, VARIAN,
U.S.A.) to test the linkage between FITC and CS–5-Fu MNPs. Fluorescence emission spectra of free FITC in DMSO and FITC-labeled CS–5-Fu MNPs in PBS 7.4 was presented in Fig. 6. In contrast with free FITC, FITC-labeled CS–5-Fu MNPs showed a large blue shift of 16 nm (max ) changed from 532 nm to 516 nm and the decrease in the emission intensity and full width of half-maximum (FWHM), these could be probably attributed to change of solvent and electronic cloud distribution after FITC conjugated to the surface of CS, which was in agreement with previous study [28,29]. It should be noted that FITC was successfully labeled on the CS–5-Fu MNPs. Confocal laser scanning microscope was further used to observe the impact of 5-Fu released from nanoparticles on the cancer cell apoptosis. The confocal images of SPCA-1 cells after incubation with the FITC-labeled CS–5-Fu MNPs at 37 ◦ C for distinct durations were illustrated in Fig. 7. As shown, the fluorescence (green) areas were observed in SPCA-1 cells after incubation with the luminescent nanoparticles. The 5-Fu loaded into CS MNPs was gradually released
6
L. Zhu et al. / Colloids and Surfaces B: Biointerfaces 68 (2009) 1–6
into cancer cells. Thus, some cells, having phagocytized nanoparticles, have started apoptosis in 24 h. Further, when incubated with FITC-labeled CS–5-Fu MNPs for 48 h, the cells displayed a characteristic morphology change comprising of cell transfiguration, and cell apoptosis tended to be more obvious. In accordance with the results of MTT assay, these results indicated that the luminescent nanoparticles had great effect on cancer cell apoptosis via 5-Fu release. 4. Conclusion 5-Fu was loaded into CS MNPs using a reverse microemulsion method. The synthesized CS–5-Fu MNPs showed small size, narrow distribution and relatively good magnetic responsivity. In vitro studies showed that 5-Fu was released slowly from CS MNPs in different buffer solutions. In addition, the CS MNPs exhibited low cytotoxicity, but the 5-Fu released from CS MNPs had obvious effect on cancer cell apoptosis. Confocal laser scanning microscopy results were very similar to that of MTT assay. Thus, based on our obtained results, it can be concluded that CS MNPs will have excellent potential as novel carriers of 5-Fu for cancer chemotherapy. Acknowledgements This work has been supported by National Basic Research Program of China(2008CB617504), Shanghai Unilever & Research Development Fund and the Nanotechnology special project of Shanghai (Nos. 0752nm028 and 06DZ05037). References [1] A.T. Florence, A.M. Hillery, N. Hussain, P.U. Jani, J. Control. Release 36 (1995) 39–46.
[2] F. Qian, F. Cui, J. Ding, C. Tang, C. Yin, Biomacromolecules 7 (2006) 2722– 2727. [3] C. Mao, J.J. Zhu, Y.F. Hu, Q.Q. Ma, Y.Z. Qiu, A.P. Zhu, W.B. Zhao, J. Shen, Colloids Surf. B 38 (2004) 47–53. [4] P. Chassary, T. Vincent, E. Guibal, React. Funct. Polym. 60 (2004) 137–149. [5] A. Zhao, P. Yao, C. Kang, X. Yuan, J. Chang, P. Pu, J. Magn. Magn. Mater. 295 (2005) 37–43. [6] J.H. Kim, Y.S. Kim, S. Kim, J.H. Park, K. Kim, K. Choi, H. Chung, S.Y. Jeong, R.W. Park, I.S. Kim, I.C. Kwon, J. Control. Release 111 (2006) 228–234. [7] F.Q. Hu, G.F. Ren, H. Yuan, Y.Z. Du, S. Zeng, Colloids Surf. B 50 (2006) 97–103. [8] E. Reverchon, A. Antonacci, Ind. Eng. Chem. Res. 45 (2006) 5722–5728. [9] J. Akbuga, Int. J. Pharm. 100 (1993) 257–261. [10] Y.J. Fu, S.S. Shyu, F.H. Su, P.C. Yu, Colloids Surf. B 25 (2002) 269–279. [11] A.Z. Chen, X.M. Pu, Y.Q. Kang, L. Liao, Y.D. Yao, G.F. Yin, Macromol. Rapid. Commun. 27 (2006) 1254. [12] Y.W. Cho, J.R. Lee, S.C. Song, Bioconjug. Chem. 16 (2005) 1529–1535. [13] C.J. Sunderland, M. Steiert, J.E. Talmadge, A.M. Derfus, S.E. Barry, Drug Dev. Res. 67 (2006) 70–93. [14] W. Chen, H. Shen, X. Li, N. Jia, J. Xu, Appl. Surf. Sci. 253 (2006) 1762–1769. [15] H.B. Shen, D.H. Long, L.Z. Zhu, X.Y. Li, Y.M. Dong, N.Q. Jia, H.Q. Zhou, X. Xin, Y. Sun, Biophys. Chem. 122 (2006) 1–4. [16] X. Liu, J. Xing, Y. Guan, G. Shan, H. Liu, Colloids Surf. A 238 (2004) 127–131. [17] T.J. Yoon, J.S. Kim, B.G. Kim, K.N. Yu, M.H. Cho, J.K. Lee, Angew. Chem. Int. Ed. 44 (2005) 1068–1071. [18] H.B. Shen, Y.B. Wang, H.F. Yang, J.S. Jiang, Chin. Sci. Bull. 48 (2003) 2252–2256. [19] D.H. Kim, S.H. Lee, K.H. Im, K.N. Kim, K.M. Kim, I.B. Shim, M.H. Lee, Y.K. Lee, Curr. Appl. Phys. 6s1 (2006) e242–e246. [20] F. Delie, Adv. Drug Deliv. Rev. 34 (1998) 221–233. [21] R.M. Young, J.K. Arnette, D.A. Roess, B.G. Barisas, Biophys. J. 67 (1994) 881–888. [22] A. Imhof, M. Megens, J.J. Engelberts, D.T.N. Lang, R. Sprik, W.L. Vos, J. Phys. Chem. B 103 (1999) 1408–1415. [23] O. Hiraku, M. Yoshiharu, Biomaterials 20 (1999) 175–182. [24] J. Zhi, Y. Wang, Y. Lu, J. Ma, G. Luo, React. Funct. Polym. 66 (2006) 1552– 1558. [25] C.L. Chiang, C.S. Sung, T.F. Wu, C.Y. Chen, C.Y. Hsu, J. Chromatogr. B 822 (2005) 54–60. [26] S. Zhou, X. Deng, X. Li, J. Control. Release 75 (2001) 27–36. [27] H. Masumi, Y. Keiko, I. Seigo, H. Midori, Jpn. J. Cancer Res. 92 (2001) 459– 466. [28] D. Wang, J. He, N. Rosenzwelg, Z. Rosenzweig, Nano Lett. 4 (2004) 409–413. [29] O.V. Makarova, A.E. Ostafin, H. Miyoshi, J.R. Norris, J. Phys. Chem. B 103 (1999) 9080–9084.
Contents lists available at ScienceDirect
Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb
Chitosan-coated magnetic nanoparticles as carriers of 5-Fluorouracil: Preparation, characterization and cytotoxicity studies Longzhang Zhu, Jingwei Ma, Nengqin Jia, Yu Zhao, Hebai Shen ∗ Life and Environment Science College, Shanghai Normal University, Shanghai 200234, PR China
a r t i c l e
i n f o
Article history: Received 3 April 2008 Received in revised form 19 July 2008 Accepted 22 July 2008 Available online 15 August 2008 Keywords: Chitosan 5-Fluorouracil Magnetic nanoparticles Carrier Fluorescence
a b s t r a c t The chitosan-coated magnetic nanoparticles (CS MNPs) were prepared as carriers of 5-Fluorouracil (CS–5Fu MNPs) through a reverse microemulsion method. The characteristics of CS–5-Fu MNPs were determined by using transmission electron microscopy (TEM), FTIR spectroscopy and vibrating-sampling magnetometry (VSM). It was found that the synthesized CS–5-Fu MNPs were spherical in shape with an average size of 100 ± 20 nm, low aggregation and good magnetic responsivity. Meanwhile, the drug content and encapsulation rate of the nanoparticles was 16–23% and 60–92%, respectively. These CS–5-Fu MNPs also demonstrated sustained release of 5-Fu at 37 ◦ C in different buffer solutions. The cytotoxicity of CS–5Fu MNPs towards K562 cancer cells was investigated. The result showed that CS–5-Fu MNPs retained significant antitumor activities. Additionally, it was observed that the FITC-labeled CS–5-Fu MNPs could effectively enter into the SPCA-1 cancer cells and induced cell apoptosis. © 2008 Published by Elsevier B.V.
1. Introduction Nanoparticles, with highly controlled shapes, sizes and some interesting properties (such as optical and magnetic properties), have been studied extensively as drug carriers which can improve the bioavailability of drug with poor absorption characteristics [1]. In recent years, much attention has been focused on the natural and synthetic polymers such as poly -caprolactone (PCL), polylactide (PLA) and poly(lactide-co-glycolide) (PLGA) due to their good biocompatibility, biodegradability and ability to prolong drug release [2]. Among them, chitosan (CS) is a favorable type of drug deliver system. CS is a natural linear biopolyaminosaccharide derived from alkaline deacetylation of chitin, which has been found to be the second most abundant natural biopolymer in nature behind only cellulose. Because of its chemical structure, CS has received increasing attention as a renewable polymeric material and has now been widely used in many fields such as protein adsorption [3] and metal adsorption [4]. Additionally, CS and its derivative have also been investigated in the development of controlled release drug delivery systems [5–7], since CS’s mucoadhesive property can enhance drug transmucosal absorption [8] and promote sustained release of drug [9]. Nowadays, 5-Fluorouracil (5-Fu) has proved to be one of the effective chemotherapeutic anticancer drugs [10]. Generally, it is
∗ Corresponding author. Tel.: +86 21 64321800; fax: +86 21 54242640. E-mail address: [email protected] (H. Shen). 0927-7765/$ – see front matter © 2008 Published by Elsevier B.V. doi:10.1016/j.colsurfb.2008.07.020
administered either orally through feeds and drinking water or through injections for several times a day [11]. However, one of the limitations inherent in current cancer chemotherapies is the lack of specificity of anticancer drug delivery. Generally, most anticancer drugs appear to have serious cytotoxic effects on normal cells [12]. Therefore, an important approach in improving treatment effect is to better harness the potency of chemotherapeutic agents by more effectively targeting them to tumor tissues [13]. With the rapid development of nanotechnology, magnetic nanoparticles, especially superparamagnetic iron oxide nanoparticles, are currently being widely studied and have been found numerous applications in the fields of cell separation [14], cell apoptosis [15] and enzyme immobilization [16]. Magnetic targeted nanoparticles can be used as drug carriers to provide targeted delivery and sustained release of chemotherapeutic agents to improve bioavailability. Drug based on carriers of core-shell magnetic nanoparticles can be easily guided to arrive at the interest position in the body by means of physical force from magnetic field [17]. Meanwhile, the outer shell (polymer layer) of the drug can effectively slow down the rate of release. Therefore, drug delivery system using polymer-coated magnetic nanoparticles is considered as an effective strategy for passive tumor targeting, which can not only increase drug circulation but also reduce pain in patients. However, limited literatures reported the direct application of chitosan-coated magnetic nanoparticles as drug carriers. In this study, we prepared one kind of novel nano-scale carrier for 5-Fu, using ␥-Fe2 O3 nanoparticles and 5-Fu as core and CS as a polymeric shell to form drug-loaded magnetic nanoparticles.
2
L. Zhu et al. / Colloids and Surfaces B: Biointerfaces 68 (2009) 1–6
The morphology, structure and characteristic of the CS–5-Fu MNPs were studied by TEM, FTIR and VSM. The application of CS MNPs as carriers of 5-Fu was evaluated by measuring its drug content, encapsulation rate and cytotoxicity in vitro. We also examined the apoptosis of SPCA-1 cancer cells induced by FITC-labeled CS–5-Fu MNPs.
rate of CS–5-Fu MNPs were calculated by the following formulas:
2. Materials and methods
2.5. In vitro drug release experiment
2.1. Materials
Drug release behavior was studied in phosphate buffered saline solution at three different pH values (PBS, 0.2 mol/L, pH 7.4, 6.8 and 1.2, respectively). CS–5-Fu MNPs (2 mg/ml), dispersed in PBS solutions, were sealed in dialysis bag (MW cutoff: 12–16 kDa, Yuanju Bio-tech Co. Ltd., Shanghai, China). The dialysis bag was incubated in 50 ml of different incubation mediums at 37 ◦ C and gently shaken at 100 cycles/min. At predetermined time intervals, 1 ml of the incubation medium was collected for 5-Fu quantitative analysis by spectrophotometry and meanwhile, 1 ml of fresh medium was compensated. For comparison, the release of free 5-Fu particles was also performed in PBS 7.4 (pH 7.4).
5-Fluorouracil was supplied by Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Chitosan (deacetylation degree 90%, MW = 60 kDa) was purchased from Wenming Biochemistry Science and Technology Co., Ltd. (Shanghai, China). Fluorescein isothiocyanate (FITC) was obtained from Sigma. RPMI-1640 medium was purchased from Sino-American Biotechnology Company. K562 cells and SPCA-1 cells were both obtained from Cell Bank of Chinese Academy of Sciences. All other reagents were analytical grade and used directly without any purification. Double distilled water was used for all the experiments.
Drug content =
W1 × 100%, W2
Encapsulation rate =
W1 × 100% W3
where W1 is the weight of 5-Fu encapsulated in the nanoparticles, W2 is the gross weight of the nanoparticles, and W3 is the total weight of the feeding 5-Fu.
2.6. Assay of in vitro cytotoxicity 2.2. Preparation of CS–5-Fu MNPs The ␥-Fe2 O3 magnetic nanoparticles were prepared by our reported method [18]. CS–5-Fu MNPs were prepared by a reverse microemulsion method. Briefly, CS powders (20 mg) were dissolved in 1 ml of 1% (w/v) acetic acid solution containing magnetite suspension and then a predetermined amount of 5-Fu (5–10 mg) was added to the CS solution. The solution was then dropwised into 50 ml of fluid wax containing emulsifier (Span-80) in a 100ml three-necked flask at 55 ◦ C. A water-in-oil microemulsion was formed by continuous stirring at 800 rmp/min for 12 h in a water bath. Then, 1 ml of 20% (w/w) sodium citrate solution was gradually added to the flask and the reaction was still kept the same condition for 1 h. After reaction, CS–5-Fu MNPs were collected with a permanent magnet and rinsed consecutively with light petroleum and isopropanol for three times. Finally, the obtained nanoparticles were dried overnight at 65 ◦ C. The preparation of CS MNPs (not adding 5-Fu) was carried in a similar way mentioned above.
2.3. Characterizations of CS–5-Fu MNPs The morphology and size of the CS–5-Fu MNPs were observed and estimated by TEM (H-600, HITACHI, Japan). The sample for TEM analysis was obtained by placing a drop of the CS–5-Fu MNPs dispersed aqueous solution onto a copper micro-grid and evaporated in air at room temperature. FTIR spectroscopy of nanoparticles was obtained by using a FTIR spectrophotometer (AVATAR 370, Nicolet, USA). The magnetization of the prepared nanoparticles was measured on a VSM (HH-15, Nanjing University Instrument Plant, China) at room temperature.
2.4. Determination of 5-Fu content in the nanoparticles CS–5-Fu MNPs (10 mg) were distributed in 50 ml of 1 mol/L HCL by sonication. After 5 h, the supernatant was collected with centrifugation and magnetic separation. The concentration of 5-Fu in the supernatant was assayed by UV–vis spectrophotometer (UV–vis 8500, Techcomp, China) at 265 nm and the supernatant from CS MNPs was used as a contrast. The drug content and encapsulation
The MTT [3-(4,5-dimethylthazol-2-yl)-2,5-diphenyltetrazolium bromide blue-indicator dye]-based assay is a simple nonradioactive colorimetric assay to measure cell cytotoxicity, proliferation or viability [19]. K562 cells were used for the analysis of cytotoxicity in vitro. The cells (1 × 105 cells/ml) were placed into 96-well tissue-culture plates and incubated at 37 ◦ C. After 24 h, cells were treated with three different concentrations of CS–5-Fu MNPs (5 g/l, 2.5 g/l, 1.25 g/l, respectively) and CS MNPs (5 g/l), respectively. Untreated cells were used as controls. Plates were incubated in a humidified 5% CO2 balanced-air incubator at 37 ◦ C for 24 h, 48 h, 72 h and 96 h, respectively. Then 30 l of 5 mg/ml MTT solution was added to each well and the plates were incubated for another 4 h, then to each well was added 30 l DMSO (dimethyl sulfoxide). The results were read on ELISA plate reader (A5002, Tecan, Austria), using a wavelength of 570 nm and the cell apoptosis rate was calculated by the following equation: cell apoptosis = 1 −
T × 100% C
where C is the number of viable cells after 24 h, 48 h, 72 h, 96 h of incubation without nanoparticles and T is the number of viable cells after 24 h, 48 h, 72 h, 96 h of incubation with nanoparticles. 2.7. Fluorescence observation on cells uptake of FITC-labeled CS–5-Fu MNPs The fluorescent labeling is one of the most important methods of modem analysis as a nonradioactive labeling technique, especially fluorescently labeled nanoparticles that can provide a rapid, simple and sensitive means to quantify cell-associated nanoparticles by fluorometry [20]. Fluorescein isothiocyanate is often used for microscopy on biological samples. Many studies have investigated the spectral properties of this dye [21,22]. FITC-labeled CS–5-Fu MNP were prepared for the confocal laser scanning microscopy study. Synthesis of the FITC-labeled CS–5-Fu MNP was based on the reaction between the isothiocyanate group of FITC and the primary amino group of CS as reported in the literature [23]. Briefly, 3 mg of FITC in 3 ml of DMSO was added to 10 ml of PBS 7.4 containing 10 mg of CS–5-Fu MNPs. After 12 h reaction in the dark at ambient conditions, the FITC-labeled CS–5-Fu MNPs were collected with a permanent magnet to remove the unconjugated FITC and washed
L. Zhu et al. / Colloids and Surfaces B: Biointerfaces 68 (2009) 1–6
3
Fig. 3. Magnetization vs. magnetic field for the CS–5-Fu MNPs. Fig. 1. TEM image of CS–5-Fu MNPs.
several times with PBS 7.4. Finally, the obtained nanoparticles were dispersed in fresh RPMI-1640. The FITC-labeled CS–5-Fu MNPs were filtered through a 0.22 m membrane for sterilization. Subsequently, 1 ml of nanoparticles (1 mg/ml) was added to SPCA-1 cancer cells (1 × 105 cells/ml), which were pre-cultured on a 35-mm glass dish for 24 h at 37 ◦ C. After incubation for specific time intervals, the fluorescence images were obtained using an confocal laser scanning microscope at ex = 488 nm and emission collected at a range of 510–540 nm. 3. Results and discussion 3.1. Characteristics of CS–5-Fu MNPs A typical TEM micrograph of CS–5-Fu MNPs was shown in Fig. 1. Result suggested that CS–5-Fu MNPs were fairly smooth and spherical in shape, and the average size was 100 ± 20 nm with a narrow distribution (measured by TEM over 500 particles). The FTIR spectra of pure CS, 5-Fu, ␥-Fe2 O3 and CS–5-Fu MNPs were presented in Fig. 2. The spectrum of CS–5-Fu MNPs showed C–N stretching peak in the amide groups at 1395 cm−1 (at 1398 cm−1 in the spectrum
of CS), C–O in the ether groups at 1086 cm−1 (at 1094 cm−1 in the spectrum of CS). Comparing the spectrum of CS–5-Fu MNPs with that of 5-Fu and ␥-Fe2 O3 , it can be seen that the peak at 1603 cm−1 appeared significantly in the 5-Fu encapsulated CS MNPs. This peak was attributed to C O group and can be seen in the pure 5-Fu spectrum (at 1673 cm−1 ). And the peak at 632 cm−1 was assigned to the Fe–O bond vibration of ␥-Fe2 O3 in CS–5-Fu MNPs (at 591 cm−1 in the spectrum of ␥-Fe2 O3 ). Besides, the peak at 3430 cm−1 corresponded to O–H bond vibration in the spectra of CS–5-Fu MNPs. These results indicated that 5-Fu and ␥-Fe2 O3 were both encapsulated in the CS MNPs. The magnetic property of CS–5-Fu MNPs was measured by VSM (result shown in Fig. 3). The saturated magnetization of the CS–5-Fu MNPs was about 3.4 emu/g, which was lower than those of free ␥Fe2 O3 [24]. Although the magnetism has decreased, nanoparticles can still be adsorbed quickly and firmly by the magnet. On the other hand, it is well known that magnetic particles less than 30 nm will demonstrate the characteristic of superparamagnetism [24], which can be verified by the magnetization curve [25]. The remanence (Mr ) and coercivity (Hc ) for CS–5-Fu MNPs in the figure were close to zero, exhibiting the characteristic of superparamagnetism. 3.2. Drug content and encapsulation rate of CS–5-Fu MNPs The drug content and encapsulation rate were determined by varying the feed weight ratio of CS and 5-Fu particles. The loading characteristics of CS–5-Fu MNPs were summarized in Table 1. When the feed weight ratio was 4:1 (sample I), the drug content was above 20%. However, if the feed ratio was 2:1 (sample II), the drug content slightly decreased to 16–18% and meanwhile the excess of 5-Fu could not totally be loaded into CS (encapsulation rate decreased to about 64%). The reason could be, under the condition of maintaining constant amount of drug, the shielding effect of CS skeleton on drug became stronger with the increasing amount of CS. Moreover, it was favorable to coat drug and to prevent drug from escaping into aqueous solution. That is also the reason that encapsulation rate decreased from over 85% to only about 64%. Table 1 The loading characterizations of CS–5-Fu MNPs
Fig. 2. FTIR spectra of (a) CS, (b) 5-Fu, (c) ␥-Fe2 O3 , and (d) CS–5-Fu MNPs.
Sample
Feed weight ratio (CS:5-Fu)
Drug content (%)
Encapsulation rate (%)
I II
4:1 2:1
20–23 16–18
85–92 60–68
4
L. Zhu et al. / Colloids and Surfaces B: Biointerfaces 68 (2009) 1–6
Fig. 5. In vitro cytotoxicity of CS–5-Fu MNPs and CS MNPs in K562 cells after 24 h, 48 h, 72 h, and 96 h.
was possibly because increasing the polymer concentration could form the thicker polymer wall, which effectively prevented 5-Fu from releasing in buffer. It could also be seen from Fig. 4(a) that no matter which kinds of feed ratio of nanoparticles could continuously release the drug. These results indicated that CS–5-Fu MNPs could be a good candidate for drug carriers. Compared to the release of drug in PBS 6.8 (pH 6.8, simulated intestinal fluid) and PBS 1.2 (pH 1.2, simulated gastric fluid), CS–5Fu MNPs (feed ratio of 4:1) released only 5% of the total loaded 5-Fu in PBS 7.4 in the first 1 h in Fig. 4(b). The release behavior of these nanoparticles, however, remained relatively less sustainable than those in two other buffers when time was elongated. Considering that 5-Fu belongs to acidic drug, it can be ionized to anions under the alkaline condition and thus its solubility will possibly be significantly improved. Therefore, at the end of 82 h, there was approximate 82% amount of loaded 5-Fu released in PBS 7.4 in comparison with 57% and 67% in other two buffers, respectively. Fig. 4. In vitro release of 5-Fu from CS–5-Fu MNPs (a) with different feed weight ratios in PBS 7.4, and (b) with feed weight ratio of 4:1 in different buffers.
3.3. In vitro 5-Fu release experiment In order to evaluate the potential of employing CS MNPs as carriers of 5-Fu, the release behaviors of 5-Fu from CS MNPs were assessed at 37 ◦ C in PBS at pH 7.4, 6.8 and 1.2, respectively. In these experiments, the conditions were maintained by regularly replacing the incubation medium. The in vitro release profile of CS–5-Fu MNPs was illustrated in Fig. 4. Fig. 4(a) displayed the release behaviors of nanoparticles with feed weight ratios of 2:1 and 4:1 in PBS 7.4. It showed that 5-Fu was released to the extent of 99% within 2 h, but CS–5-Fu MNPs can be sustainably released for 80 h. The result that more than 80% of 5-Fu was released for 80 h suggested the potential of the nanoparticles as a sustained drug delivery system. A small (about 23% of the total loaded drug) burst release occurred in the first 2 h for the nanoparticles with feed ratio of 2:1, and then followed by a very slow drug release. Zhou et al. suggested that the drug release involved two different mechanisms of drug molecules diffusion and polymer matrix degradation [26]. The burst release was probably related with the excess drug particles (close to the CS surface) dispersing rapidly from matrix into buffer in the first few hours. Because of drug concentration decreasing, most drug in the nanoparticles with feed ratio of 4:1 was relatively far from the CS surface. This might be one of the reasons for no obvious burst release of drug for such nanoparticles. In addition, it
3.4. In vitro cytotoxicity K562 cells were incubated with CS–5-Fu MNPs and CS MNPs, respectively, to test that the cytotoxicity was caused by 5-Fu and not by the CS MNPs. As shown in Fig. 5, drug continually released from the matrix accelerated cell apoptosis and the apoptosis ten-
Fig. 6. Comparison of fluorescence emission spectra between FITC-labeled CS–5-Fu MNPs (a) and free FITC (b).
L. Zhu et al. / Colloids and Surfaces B: Biointerfaces 68 (2009) 1–6
5
Fig. 7. Confocal microscopy images of SPCA-1 cells incubated with FITC-labeled CS–5-Fu MNPs after (a) 2 h, (b) 24 h, and (c) 48 h.
dency was more obvious when the incubation time continued. This phenomenon indirectly suggested that the nanoparticles can effectively prolong drug release. Cell apoptosis rate was increased from 31% after 24 h to 87% at the end of 96 h induced by CS–5Fu MNPs with the same concentration at 5 g/l. It was also observed that cell apoptosis rates were decreased at lower concentrations of nanoparticles. That is due to lower concentrations of drug loaded into nanoparticles. Surprisingly, when incubated with control nanoparticles, the cell apoptosis rate also reached about 20%. This may be because free CS itself can induce cancer cell apoptosis in some extent [27]. As expected, the cytotoxicity was caused mainly by 5-Fu itself and CS MNPs only played a limited role in cell apoptosis. 3.5. Cellular uptake of FITC-labeled CS–5-Fu MNPs The fluorescence peak of FITC-labeled CS–5-Fu MNPs was measured by fluorescence spectrophotometer (Cary Eclipse, VARIAN,
U.S.A.) to test the linkage between FITC and CS–5-Fu MNPs. Fluorescence emission spectra of free FITC in DMSO and FITC-labeled CS–5-Fu MNPs in PBS 7.4 was presented in Fig. 6. In contrast with free FITC, FITC-labeled CS–5-Fu MNPs showed a large blue shift of 16 nm (max ) changed from 532 nm to 516 nm and the decrease in the emission intensity and full width of half-maximum (FWHM), these could be probably attributed to change of solvent and electronic cloud distribution after FITC conjugated to the surface of CS, which was in agreement with previous study [28,29]. It should be noted that FITC was successfully labeled on the CS–5-Fu MNPs. Confocal laser scanning microscope was further used to observe the impact of 5-Fu released from nanoparticles on the cancer cell apoptosis. The confocal images of SPCA-1 cells after incubation with the FITC-labeled CS–5-Fu MNPs at 37 ◦ C for distinct durations were illustrated in Fig. 7. As shown, the fluorescence (green) areas were observed in SPCA-1 cells after incubation with the luminescent nanoparticles. The 5-Fu loaded into CS MNPs was gradually released
6
L. Zhu et al. / Colloids and Surfaces B: Biointerfaces 68 (2009) 1–6
into cancer cells. Thus, some cells, having phagocytized nanoparticles, have started apoptosis in 24 h. Further, when incubated with FITC-labeled CS–5-Fu MNPs for 48 h, the cells displayed a characteristic morphology change comprising of cell transfiguration, and cell apoptosis tended to be more obvious. In accordance with the results of MTT assay, these results indicated that the luminescent nanoparticles had great effect on cancer cell apoptosis via 5-Fu release. 4. Conclusion 5-Fu was loaded into CS MNPs using a reverse microemulsion method. The synthesized CS–5-Fu MNPs showed small size, narrow distribution and relatively good magnetic responsivity. In vitro studies showed that 5-Fu was released slowly from CS MNPs in different buffer solutions. In addition, the CS MNPs exhibited low cytotoxicity, but the 5-Fu released from CS MNPs had obvious effect on cancer cell apoptosis. Confocal laser scanning microscopy results were very similar to that of MTT assay. Thus, based on our obtained results, it can be concluded that CS MNPs will have excellent potential as novel carriers of 5-Fu for cancer chemotherapy. Acknowledgements This work has been supported by National Basic Research Program of China(2008CB617504), Shanghai Unilever & Research Development Fund and the Nanotechnology special project of Shanghai (Nos. 0752nm028 and 06DZ05037). References [1] A.T. Florence, A.M. Hillery, N. Hussain, P.U. Jani, J. Control. Release 36 (1995) 39–46.
[2] F. Qian, F. Cui, J. Ding, C. Tang, C. Yin, Biomacromolecules 7 (2006) 2722– 2727. [3] C. Mao, J.J. Zhu, Y.F. Hu, Q.Q. Ma, Y.Z. Qiu, A.P. Zhu, W.B. Zhao, J. Shen, Colloids Surf. B 38 (2004) 47–53. [4] P. Chassary, T. Vincent, E. Guibal, React. Funct. Polym. 60 (2004) 137–149. [5] A. Zhao, P. Yao, C. Kang, X. Yuan, J. Chang, P. Pu, J. Magn. Magn. Mater. 295 (2005) 37–43. [6] J.H. Kim, Y.S. Kim, S. Kim, J.H. Park, K. Kim, K. Choi, H. Chung, S.Y. Jeong, R.W. Park, I.S. Kim, I.C. Kwon, J. Control. Release 111 (2006) 228–234. [7] F.Q. Hu, G.F. Ren, H. Yuan, Y.Z. Du, S. Zeng, Colloids Surf. B 50 (2006) 97–103. [8] E. Reverchon, A. Antonacci, Ind. Eng. Chem. Res. 45 (2006) 5722–5728. [9] J. Akbuga, Int. J. Pharm. 100 (1993) 257–261. [10] Y.J. Fu, S.S. Shyu, F.H. Su, P.C. Yu, Colloids Surf. B 25 (2002) 269–279. [11] A.Z. Chen, X.M. Pu, Y.Q. Kang, L. Liao, Y.D. Yao, G.F. Yin, Macromol. Rapid. Commun. 27 (2006) 1254. [12] Y.W. Cho, J.R. Lee, S.C. Song, Bioconjug. Chem. 16 (2005) 1529–1535. [13] C.J. Sunderland, M. Steiert, J.E. Talmadge, A.M. Derfus, S.E. Barry, Drug Dev. Res. 67 (2006) 70–93. [14] W. Chen, H. Shen, X. Li, N. Jia, J. Xu, Appl. Surf. Sci. 253 (2006) 1762–1769. [15] H.B. Shen, D.H. Long, L.Z. Zhu, X.Y. Li, Y.M. Dong, N.Q. Jia, H.Q. Zhou, X. Xin, Y. Sun, Biophys. Chem. 122 (2006) 1–4. [16] X. Liu, J. Xing, Y. Guan, G. Shan, H. Liu, Colloids Surf. A 238 (2004) 127–131. [17] T.J. Yoon, J.S. Kim, B.G. Kim, K.N. Yu, M.H. Cho, J.K. Lee, Angew. Chem. Int. Ed. 44 (2005) 1068–1071. [18] H.B. Shen, Y.B. Wang, H.F. Yang, J.S. Jiang, Chin. Sci. Bull. 48 (2003) 2252–2256. [19] D.H. Kim, S.H. Lee, K.H. Im, K.N. Kim, K.M. Kim, I.B. Shim, M.H. Lee, Y.K. Lee, Curr. Appl. Phys. 6s1 (2006) e242–e246. [20] F. Delie, Adv. Drug Deliv. Rev. 34 (1998) 221–233. [21] R.M. Young, J.K. Arnette, D.A. Roess, B.G. Barisas, Biophys. J. 67 (1994) 881–888. [22] A. Imhof, M. Megens, J.J. Engelberts, D.T.N. Lang, R. Sprik, W.L. Vos, J. Phys. Chem. B 103 (1999) 1408–1415. [23] O. Hiraku, M. Yoshiharu, Biomaterials 20 (1999) 175–182. [24] J. Zhi, Y. Wang, Y. Lu, J. Ma, G. Luo, React. Funct. Polym. 66 (2006) 1552– 1558. [25] C.L. Chiang, C.S. Sung, T.F. Wu, C.Y. Chen, C.Y. Hsu, J. Chromatogr. B 822 (2005) 54–60. [26] S. Zhou, X. Deng, X. Li, J. Control. Release 75 (2001) 27–36. [27] H. Masumi, Y. Keiko, I. Seigo, H. Midori, Jpn. J. Cancer Res. 92 (2001) 459– 466. [28] D. Wang, J. He, N. Rosenzwelg, Z. Rosenzweig, Nano Lett. 4 (2004) 409–413. [29] O.V. Makarova, A.E. Ostafin, H. Miyoshi, J.R. Norris, J. Phys. Chem. B 103 (1999) 9080–9084.