- Email: [email protected]
Development and characterisation of chitosan nanoparticles for siRNA delivery
GENE DELIVERY
Journal of Controlled Release 115 (2006) 216 – 225 www.elsevier.com/locate/jconrel
Development and characterisation of chitosan nanoparticles for siRNA delivery Haliza Katas, H. Oya Alpar ⁎ School of Pharmacy, University of London, 29-39, Brunswick Square, London, WC1N 1AX, UK Received 20 April 2006; accepted 20 July 2006 Available online 25 July 2006
Abstract Gene silencing mediated by double-stranded small interfering RNA (siRNA) has been widely investigated as a potential therapeutic approach for diseases with genetic defects. The use of siRNA, however, is hampered by its rapid degradation and poor cellular uptake into cells in vitro or in vivo. Therefore, we have explored chitosan as a siRNA vector due to its advantages such as low toxicity, biodegradability and biocompatibility. Chitosan nanoparticles were prepared by two methods of ionic cross-linking, simple complexation and ionic gelation using sodium tripolyphosphate (TPP). Both methods produced nanosize particles, less than 500 nm depending on type, molecular weight as well as concentration of chitosan. In the case of ionic gelation, two further factors, namely chitosan to TPP weight ratio and pH, affected the particle size. In vitro studies in two types of cells lines, CHO K1 and HEK 293, have revealed that preparation method of siRNA association to the chitosan plays an important role on the silencing effect. Chitosan–TPP nanoparticles with entrapped siRNA are shown to be better vectors as siRNA delivery vehicles compared to chitosan–siRNA complexes possibly due to their high binding capacity and loading efficiency. Therefore, chitosan– TPP nanoparticles show much potential as viable vector candidates for safer and cost-effective siRNA delivery. © 2006 Elsevier B.V. All rights reserved. Keywords: siRNA; Chitosan; Delivery system; Simple complexation; Ionic gelation
1. Introduction Over the past few decades, antisense approaches including oligonucleotides, ribozymes and DNAzymes have been extensively investigated as tools for controlling cellular processes. Only recently, small interfering RNAs (siRNAs) have proven to be versatile agents for controlling gene expression in mammalian cells. They have also been shown more potent than conventional antisense strategies [1–3]. Overall, they appear to be a much more robust and efficient technology offering significant potential. siRNA consisting of 21–23 nucleotides can regulate gene expression in mammalian cells through RNA interfering (RNAi). As the administration of siRNA could bypass nonspecific inhibition of protein synthesis induced by long doublestranded RNA [4,5], it has therefore been employed as a novel tool in blocking the expression of genes such as those expressed ⁎ Corresponding author. Tel.: +44 207 753 5928; fax: +44 207 753 5975. E-mail address: [email protected] (H.O. Alpar). 0168-3659/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.jconrel.2006.07.021
in infectious diseases and cancers. However, similar to hydrophilic and polyanion-mediated Gene Therapy, siRNA also suffers particular problems including poor cellular uptake, rapid degradation by ubiquitous nucleases as well as limited blood stability [6,7]. As a result of these limitations, unassisted delivery of siRNA to the cells is frustrating. Although various chemical modifications of siRNA can be used to overcome these problems, these modifications posses disadvantages such as a decreased mRNA hybridization, higher cytotoxicity and increased unspecific effects [8]. Therefore, effective systems which can protect and transport siRNA to the cytoplasm of the target cells are needed to exploit the promising potential applications offered by successful delivery of siRNA. From among the gene vectors that have been studied, nonviral vectors have attracted more and more attention in comparison to viral vectors, although viral vectors have been proven to yield higher transfection efficiency in most cell lines. This is due to the advantages of non-viral vectors such as ease of synthesis, low immune response against the vector and unrestricted gene materials size in addition to potential benefits in
terms of safety [9]. In recent years, chitosan-based carriers are one of the non-viral vectors that have gained increasing interest as a safer and cost-effective delivery system for gene materials including plasmid DNA (pDNA), oligonucleotide (ODN) as well as proteins and peptides. Chitosan has beneficial qualities such as low toxicity, low immunogenicity, [10] excellent biodegradability, biocompatibility, [10,11] as well as a high positive charge that can easily form polyelectrolyte complexes with negatively charged nucleotides by electrostatic interaction. Although chitosan has been studied for more than a decade as a gene vector for pDNA and ODN [12], so far to our knowledge, there is no study that has been carried out to investigate the use of chitosan to deliver siRNA in vitro. In this report, we studied three methods of siRNA association; by simple complexation, ionic gelation (siRNA entrapment) and adsorption of siRNA onto the surface of preformed chitosan nanoparticles for preparation of chitosan-based nano-carriers. To address the lack of information regarding the behaviour of interaction between siRNA and chitosan – since siRNA's structure and size is quite different to that of pDNA, these systems were fully characterised with regards to their physical and biological features by exploiting commercial chitosan products. The ability of these nanoparticulate systems to mediate gene silencing was then assessed on CHO K1 and HEK 293 cells in vitro. 2. Materials and methods 2.1. Materials Four different types of medical grade chitosan with the degree of deacetylation (DD) of ∼ 86% were used: chitosan hydrochloride with molecular weight of 270 kDa (Cl213) and 110 kDa (Cl113) as well as chitosan glutamate with molecular weight of 470 kDa (G213) and 160 kDa (G113) (Protasan Ultrapure, Pronova Biomedical, Norway). Pentasodium tripolyphosphate (TPP) was obtained from Fluka and sodium acetate from Sigma-Aldrich. siRNA targeting against pGL3 luciferase gene (sense: 5′–CUUACGCUGAGUACUUCGATT–3′, antisense: 3′–TTGAAUGCGACUCAUGAAGCU–5′) and control siRNA (non-silencing) (sense: 3′–UUCUCCGAACGUGU CACGUTT–3′, antisense: 3′–TTAAGAGGCUUGCACA GUGCA–5′) were synthesized by Proligo (France). Reporter vectors, pGL3 control and pRL-TK which encoded firefly luciferase and renilla luciferase gene, respectively were purchased from Promega, UK. Dual-Glo Luciferase Assay system, agarose (low melting point) and Tris–borate–EDTA buffer pH 8.3 were from Promega, UK. Lipofectamine 2000 and OptiMEM I reduced serum medium were purchased from Invitrogen, UK. Other reagents were all commercially available and were of analytical grade. 2.2. Chitosan nanoparticles preparation 2.2.1. Simple complexation Four types of chitosan (chitosan hydrochloride and glutamate with two different molecular weights) were dissolved separately
217
either in distilled water or acetate buffer (0.1 M sodium acetate/ 0.1 M acetic acid, pH 4.5) to form different concentrations of chitosan solution, ranging from 25 to 300 μg/ml. ChitosansiRNA complexes were prepared by adding chitosan solution drop-wise to an equal volume of siRNA solution (20 μg/ml) and quickly mixed before incubating them at room temperature for 30 min to form chitosan-siRNA complexes with the chitosan to siRNA weight ratio of 1.25:1 to 15:1. 2.2.2. Ionic gelation Nanoparticles were produced based on modified ionic gelation of tripolyphosphate (TPP) with chitosan as described elsewhere [13]. Two different types of chitosan (chitosan hydrochloride and glutamate) and each type with two different molecular weights were investigated. Nanoparticles were spontaneously obtained upon the addition of 1.2 ml of a TPP aqueous solution (0.84 mg/ml) to 3 ml of chitosan solution (2 mg/ml, at chitosan to TPP weight ratio of 6:1) under constant magnetic stirring at room temperature. The particles were then incubated at room temperature for 30 min before use or further analysis. Nanoparticles were collected by centrifugation (IEA Micromax Ultracentrifuge) at 13,000×g for 10 min. The supernatants were discarded and nanoparticles were resuspended in filtered (Millex® GP filter unit, Millipore, Ireland; 0.25 μm) distilled water. 2.2.2.1. siRNA entrapment in chitosan–TPP nanoparticles. For the association of siRNA with the chitosan–TPP nanoparticles (chitosan–TPP–siRNA), 3 μl of siRNA (19.95 μg/μl) in double distilled water was added to the TPP solution (1.2 ml, 0.84 mg/ ml) before adding this drop-wise to the chitosan solution (3 ml, 2 mg/ml) under constant magnetic stirring at room temperature. The particles were then incubated at room temperature for 30 min before use or further analysis. 2.2.2.2. siRNA adsorption onto chitosan–TPP nanoparticles. Preprepared chitosan nanoparticles prepared by ionic gelation as described above were dispersed in distilled water to yield a chitosan concentration, ranging from 0.1 to 1 mg/ml. To adsorb siRNA onto the surface of chitosan nanoparticles, 500 μl of siRNA solution (10 μg/ml in distilled water) was added dropwise to 500 μl of chitosan suspension and quickly mixed by inverting the interaction tube up and down. Then, the particles were incubated for 2 h at room temperature before further analysis. 2.3. Characterisation of the chitosan nanoparticles or chitosan siRNA nanoparticles Mean particle diameter (Z-average) and zeta potential of the nanoparticles were determined using Submicron Particle Analyser System 4700 and Zetasizer® S (Malvern Instruments, UK), respectively. The measurements of particle size were made at 25 °C in triplicate and no further dilution was performed for these particles. For the determination of zeta potential, samples were diluted with distilled water to an appropriate concentration to yield count rate per second (KCps) in the range of 2500–3500. Each batch was analysed in triplicate.
GENE DELIVERY
H. Katas, H.O. Alpar / Journal of Controlled Release 115 (2006) 216–225
GENE DELIVERY
218
H. Katas, H.O. Alpar / Journal of Controlled Release 115 (2006) 216–225
2.4. Determination of siRNA loading efficiency The loading efficiency of siRNA (%) entrapped or adsorbed onto the chitosan nanoparticles was obtained from the determination of free siRNA concentration in the supernatant recovered after particle centrifugation (13,000×g, 15 min) by absorbance measurement (UV–Visible Spectrophotometer Cary 3E, Varian) at 260 nm. Supernatant recovered from unloaded chitosan–TPP nanoparticles (without siRNA) was used as a blank. siRNA loading efficiency (%) was the percentage of entrapped or adsorbed siRNA, (difference between the total amount of siRNA added for the nanoparticles preparation and the amount of nonentrapped or -adsorbed siRNA remaining in supernatant after centrifugation) to the total amount of siRNA added.
well plate at a density of 30,000 cells per well in Opti-MEM 1 reduced serum medium containing 5% of FBS without antibiotics, 24 h prior to transfection. On the day of transfection, pGL3-control (0.15 μg) and pRL-TK vectors (0.05 μg) were cotransfected to the cells using Lipofectamine 2000™ according to manufacturer's instructions. After 4 h of transfection, the medium was removed and the cells were washed with PBS, following this, media was replaced with 100 μl fresh medium containing serum. 50 μl of chitosan–siRNA nanoparticles, siRNA alone or Lipofectamine 2000–siRNA complexes (each well or formulation contained 4 pmol of siRNA) in the medium without serum were then added to the cells and incubated at 37 °C with a 5% CO2 atmosphere for 24 or 48 h. After 24 or 48 h, luciferase activities were determined using Dual-Glo Luciferase Assay System [14].
2.5. Gel retardation assay 2.8. Cytotoxicity assay The binding of siRNA with chitosan was determined by 4% agarose (low melting point) gel electrophoresis. A series of different chitosan to siRNA weight ratios was loaded (20 μl of the sample containing 0.2 μg of siRNA). A 1:6 dilution of loading dye was added to each well and electrophoresis was carried out at a constant voltage of 55 V for 2 h in TBE buffer (4.45 mM Tris–base, 1 mM sodium EDTA, 4.45 mM boric acid, pH 8.3) containing 0.5 μg/ml ethidium bromide. The siRNA bands were then visualised under a UV transilluminator at a wavelength of 365 nm. 2.6. Assay for serum stability Chitosan–siRNA nanoparticles (200 μl, equivalent to 5 μg of siRNA) were incubated at 37 °C with equal volume of Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 5% and 50% final concentration of foetal bovine serum (FBS). At each predetermined time interval (0, 30 min, 2, 4, 7, 24 48 and 72 h), 30 μl of the mixture was removed and stored at − 20 °C until gel electrophoresis was performed. To terminate serum activity, samples were incubated in bath incubator at 80 °C for 5 min and 5 μl heparin (1000 U/ml) was added for displacing the siRNA from the chitosan nanoparticles. The integrity of the siRNA was then analysed by 15% polyacrylamide gel containing 7 M urea and TBE buffer (0.089 M Tris base, 0.089 M boric acid, and 2 mM sodium EDTA, pH 8.3). Polyacrylamide–Urea gel (15%) was used due to its high efficiency in separating small fragments of possibly degraded siRNA. Electrophoresis was then carried out with 1× TBE buffer at a constant voltage of 200 V for 1 h. siRNA bands were visualised under a UV transilluminator after staining for 40 min with a 1:10,000 dilution of SYBR-Green II RNA gel stain (Molecular Probes) prepared in DEPC treated water. SYBRGreen for RNA was chosen as it was more sensitive than ethidium bromide in staining small RNA. 2.7. Biological activity of chitosan–siRNA nanoparticles In vitro transfection studies were performed in CHO K1 (Chinese Hamster Ovary) cells. The cells were seeded in a 96-
The effect of chitosan on cytotoxicity was measured by determining cell viability of chitosan–siRNA nanoparticles, calculated as a percentage of the cell viability of untreated/nontransfected cell samples. CHO K1 cells were seeded in a 96-well plate at a density of 30,000 cells per well in Opti-MEM 1 reduced serum medium containing 5% of FBS and grown overnight. After 24 and 48 h post-incubation of chitosan–siRNA nanoparticles or complexes at 37 °C, 20 μl of MTT (3-(4,5-dimethyl-thiazol-2-yl)2,5-diphenyl-tetrazolium bromide, 5 mg/ml) in sterile filtered PBS was added to each well and then incubated for 4 h to allow formation of formazan crystals. After 4 h, the unreduced MTT and medium was removed and the cells were washed with PBS. 200 μl of DMSO was then added to each well to dissolve the MTT formazan crystals and the plate was incubated at 37 °C for 5 min. The absorbance of formazan products was measured at 540 nm using a microplate reader (Wallac Victor2 1420 Multilabel Counter). 2.9. Statistical analysis Data are presented as the means ± standard deviations. The statistical significance was determined by using Independent Sample t-test or one-way analysis of variance (ANOVA). P values of b 0.05 were considered significant. The statistical analyses were carried out using SPSS Base 12.0 for Windows (SPSS). 3. Results 3.1. Particle size Mean particle size of chitosan–siRNA complexes prepared by simple complexation for both chitosan hydrochloride and glutamate were increased when the concentration of chitosan was increased from 25 to 300 μg/ml in distilled water (Fig. 1). However, a smaller mean particle size of chitosan nanoparticles was obtained when the lower molecular weight of chitosan was used compared to the higher molecular weight for the individual chitosan derivatives. In addition, chitosan glutamate (G213, G113) which had a higher molecular weight than chitosan
hydrochloride (Cl213, Cl113) produced smaller complexes with siRNA than chitosan hydrochloride. On the other hand, the formation of chitosan–TPP nanoparticles occurred spontaneously upon the addition of the TPP ions into the chitosan solution. The results showed that the appearance of the solution changed when a certain amount of TPP ions was added to the chitosan solution, from a clear to opalescent solution that indicated a change of the physical states of the chitosan to form nanoparticles, then microparticles and eventually aggregates [15]. For all types of chitosan tested in this study, chitosan and TPP concentration were adjusted to a weight ratio of 6:1 to obtain nanoparticles and the mean particle size of chitosan–TPP nanoparticles was 510 ± 22.9, 276 ± 17.9, 709 ± 50.3 and 415 ± 44.6 nm for G213, G113, Cl213 and Cl113, respectively. The chitosan nanoparticles entrapped with siRNA, however showed no significant difference in particle size compared to the plain particles. The weight ratio of chitosan to TPP however was changed from 6:1 to 5:1 as the pH of chitosan solution used in the ionic gelation was changed from pH 6 (distilled water) to pH 4.5 (acetate buffer, 0.1 M). 3.2. Surface charge The comparative positive value of surface charge (zeta potential) of the chitosan–siRNA complexes increased with the increasing concentration of chitosan at a constant siRNA concentration (Fig. 2). The increment was due to the increase in the number of positive charges which counteracts with negatively charged siRNA as the amount of siRNA was fixed. The net positive charge of the particles was desirable to prevent particle aggregation and promote electrostatic interaction with the overall negative charge of the cell membrane [16]. In addition, body distribution of nanoparticles after i.v. injection is highly influenced by their interaction with the biological environment which also dependent on their physicochemical properties like surface charge of nanoparticles [17]. The values of particle surface charge for each of the tested chitosans are presented in the Fig. 2. The surface charge of chitosan–TPP nanoparticles was ranged from approximately from + 40 to + 60 mV by changing the weight ratio of chitosan to TPP from 4:1 to 6:1. The addition of siRNA (10 μg) at the chitosan to TPP
Fig. 1. The effect of chitosan molecular weight and concentration (or chitosan to siRNA weight ratio) on mean particle diameter of chitosan–siRNA complex (n = 3).
219
Fig. 2. Effect of chitosan concentration on particle surface charge of chitosan– siRNA complexes (n = 3).
weight ratio of 6 reduced the surface charge of nanoparticles to approximately + 30 mV. 3.3. Effect of pH To investigate the effect of pH on the interaction between TPP ions and the chitosan derivatives that were used here, chitosan was dissolved in acetate buffer (0.1 M, pH ranging from 4 to 6) at a fixed concentration of 2 mg/ml and a chitosan to TPP mass ratio of 5:1. Interestingly, the smallest particles were formed at pH 4.5 (Fig. 3) for all the tested chitosans except for Cl213 (pH 6). In addition, the surface charge of the yielded chitosan–TPP nanoparticles was increased approximately twofold to + 40 mV when the pH of the chitosan solution was changed from 6 to 4. In contrast, investigation into the influence of pH on the particle size revealed that no significant difference in particle size was observed when complexing chitosan with siRNA in acetate buffer at pH 4.5 (0.1 M) compared to in distilled water for all the tested chitosans. 3.4. Interaction of siRNA with chitosan For siRNA adsorbed onto the chitosan nanoparticles, complete binding of siRNA with the chitosan (without the presence of a trailing band) could only be observed when the chitosan nanoparticles to siRNA weight ratio was approaching 100:1 except for lower molecular weight chitosan hydrochloride, Cl113 (Fig. 4B) whereas all the tested chitosans showed complete
Fig. 3. Effect of pH on mean particle size of chitosan–TPP nanoparticles (n = 6).
GENE DELIVERY
H. Katas, H.O. Alpar / Journal of Controlled Release 115 (2006) 216–225
GENE DELIVERY
220
H. Katas, H.O. Alpar / Journal of Controlled Release 115 (2006) 216–225
Fig. 4. Binding efficiency of siRNA to the chitosan or chitosan associated with TPP nanoparticles: (A) chitosan-siRNA complexes, (B) chitosan–TPP with adsorbed siRNA and (C) chitosan–TPP–siRNA nanoparticles (entrapped siRNA).
binding of siRNA for chitosan–TPP nanoparticles with entrapped siRNA (Fig. 4C). On the other hand, the migration of siRNA from the chitosan–siRNA complex was observed even at higher concentrations of chitosan (1 mg/ml) or higher chitosan to siRNA weight ratio (100:1) as shown on Fig. 4A. Similar observations were also observed for the chitosan–siRNA complexes prepared in acetate buffer at pH 4.5 as determined by gel retardation assay.
chitosan–TPP with adsorbed siRNA nanoparticles at a weight ratio of 100:1 were varied and depended on the type of chitosan used. Higher siRNA loading efficiency was observed for siRNA adsorbed onto chitosan glutamate (83% ± 0.9% for G213 and 90% ± 0.3% for G113) compared to chitosan hydrochloride (72% ± 1.1% for Cl213 and 59% ± 0.8% for Cl113). 3.6. siRNA stability in serum
3.5. siRNA loading efficiency 100% siRNA loading efficiency was obtained as measured by spectrophotometry for all the entrapped siRNA chitosan– TPP nanoparticles. However, the loading efficiencies of
siRNA must be stable to digestion by nuclease for maximal activity in cells [18]. Therefore, to address the question of chitosan–siRNA nanoparticles stability and protection from nuclease degradation, they were incubated in 5% and 50% of
FBS at 37 °C. Contrasting results were obtained in a previous report by Braasch et al. [18] who reported that naked siRNA was stable after incubation in 5% serum up to 72 h. However, in this study siRNA was intact only up to 30 min and it was fully degraded after 48 h. On the other hand, siRNA recovered from chitosan–TPP nanoparticles started to degrade after 24 h incubation and full degradation was only observed after 72 h incubation (Fig. 5A). This experiment was therefore repeated by incubating siRNA as well as chitosan–siRNA nanoparticles in a higher serum concentration. The chitosan–siRNA nanoparticles significantly protected siRNA from nuclease activity. Complete degradation of siRNA was observed as early as time-point 0 with degradation occurring during the mixing of siRNA with serum and freezing steps. In contrast with unformulated siRNA, the siRNA recovered from chitosan–siRNA nanoparticles was intact up to 7 h and fully degraded after 48 h incubation in 50% serum (Fig. 5B).
221
negligible silencing effect was observed for naked siRNA as well as Lipofectamine 2000–mismatch siRNA (non-silencing) complexes. In this study, chitosan glutamate, G213 (470 kDa) showed the highest gene silencing effect at 24 h post-transfection either by simple complexation (51% of gene knockdown) or ionic gelation (82% and 63% of gene knockdown for siRNA entrapment and siRNA adsorption, respectively) compared to its lower molecular weight or chitosan hydrochloride. siRNA entrapped in chitosan–TPP nanoparticles prepared from chitosan glutamate, G213 even showed a slightly higher efficiency than Lipofectamine 2000 (Fig. 6B). However, transfection studies performed in HEK 293 human kidney cells showed lower gene silencing activities for these chitosan nanoparticles either at 24 or 48 h post-transfection compared to CHO K1 cells. Chitosan G213 for example was 22%, 14% and 64% less efficient than Lipofectamine 2000 for chitosan–TPP with adsorbed siRNA, chitosan–TPP with entrapped siRNA and chitosan–siRNA complex, respectively (Fig 6B).
3.7. Biological activity of chitosan–siRNA nanoparticles 3.8. Cytotoxicity study Transfections performed with nanoparticles prepared with different types of chitosan and methods of siRNA association to the chitosan revealed that the silencing effect of certain chitosan–siRNA nanoparticles or complexes had a comparable effect compared to Lipofectamine 2000 (Fig. 6B). In contrast, a
To investigate the potential cytotoxicity of chitosan–siRNA nanoparticulate systems, the cell viability was determined by MTT assays. Over 90% average cell viability was observed for chitosan–siRNA complex and naked siRNA in comparison to
Fig. 5. Electrophoretic mobility of chitosan–TPP–siRNA nanoparticles following incubation with FBS: (A) 5% FBS and (B) 50% FBS.
GENE DELIVERY
H. Katas, H.O. Alpar / Journal of Controlled Release 115 (2006) 216–225
GENE DELIVERY
222
H. Katas, H.O. Alpar / Journal of Controlled Release 115 (2006) 216–225
Fig. 6. Effect of chitosan types and method of siRNA association to the chitosan nanoparticles on gene silencing: (A) percentage of gene knockdown of pGL3 luciferase in CHO K1 cells at 24 and 48 h post-transfection and (B) relative response ratio of chitosan nanoparticles prepared from chitosan glutamate, G213 in CHO K1 and HEK 293 cells at 24 h post-transfection. Lipo – Lipofectamine 2000; siRNA M – siRNA mismatch.
Fig. 7. Effect of chitosan types and method of siRNA association to the chitosan nanoparticles on percentage of cell viability of CHO K1 (n = 3).
untreated cells. However, 18–40% loss of cell viability was observed for siRNA associated with chitosan–TPP nanoparticles although individual TPP and chitosan solution did not show any loss of cell viability (data not shown). The cell viabilities observed with siRNA associated with chitosan–TPP nanoparticles were however not significantly different between the different chitosan derivatives. Nevertheless, the cell viability for certain siRNA associated with chitosan–TPP formulations was increased at 48 h post-incubation which indicates the recovery of cells was observed within this period (Independent Sample ttest, p b 0.05). From the studied formulations, chitosan G113 showed significant recovery of cells either by adsorbing or incorporating siRNA with chitosan–TPP nanoparticles during the preparation process (chitosan–TPP–siRNA). However, for Cl213, significant recovery was observed when adsorbing
siRNA onto the chitosan–TPP nanoparticles. In contrast, Cl113 showed high recovery of the cells when siRNA was incorporated to the chitosan during the ionic gelation process (Fig. 7). Although only a small improvement of cell viability was observed for other chitosan–TPP formulations, this finding still gave an indication of a transient physiological stress of the cells by these formulations. 4. Discussion In this study, the simple complexation and ionic gelation methods were chosen to prepare chitosan–siRNA complexes or nanoparticles since both processes are simple and mild [11]. In addition, the use of TPP in ionic gelation as a polyanion to cross-link with the cationic chitosan through electrostatic interaction could avoid possible toxicity of reagents used in chemical cross-linking (e.g. glutaraldehyde). Modulation of size and surface charge of the particles could also be easily done by using both methods [19], by adjusting certain process parameters such as chitosan concentration and stirring rate. Adsorption of siRNA onto the surface of preformed chitosan nanoparticles is also another simple approach to associate siRNA to the chitosan nanoparticles. Therefore, it was investigated in this study as another option for associating siRNA to the chitosan. Particle size and shape play an important role in transferring genes to the cells and they also greatly influence particle distribution in the body [20]. It has been reported that particles in the nanometer size have a relatively higher intracellular uptake compared to microparticles [21–23]. This characteristic is very important in gene transfer because cellular uptake of the chitosan/DNA complex as well as its subsequent release from the endo-lysosome pathway are two of the rate limiting steps in this process [24,25]. Similar to pDNA and ODN, siRNA are likely taken up by cells through endocytosis and if siRNA is not effectively delivered to the cytoplasmic compartment, it will not induce RNAi [26]. This is because the uptake of siRNA by fluid phase endocytosis does not result in the release of siRNA into the cytoplasm and mammalian cells appear to lack of the effective double stranded RNA-uptake machinery that is found in other species such as Caenorhabditis elegans [26]. Therefore, siRNA and its vector should be able to be taken up by the cells and escape the endosomal vesicle to avoid lysosomal degradation, thus, allow the RNAi to occur. In this study, we tried to obtain nanoparticles of less than 500 nm to facilitate the uptake of the particles. This can be achieved by manipulating certain process parameters that are involved in nanoparticle preparation. Generally, particle size of chitosan–siRNA complexes or nanoparticles was highly dependent on the molecular weight, type and concentration of chitosan used. However, in the case of ionic gelation, other factors like weight ratio of chitosan to TPP and pH of the chitosan solution were also shown to affect the particle size. In this study, smaller particle size was obtained with the lower molecular weight or concentration for both types of chitosan. This was expected due to the decreased viscosity of the lower concentration or molecular weight of chitosan which
223
resulted in better solubility and generated its molecular character as a polyelectrolyte material that allows more efficient interaction between negatively charged siRNA and the cationic chitosan. In addition, a smaller particle size was obtained with the chitosan glutamate compared to chitosan hydrochloride although it had a higher molecular weight than chitosan hydrochloride. Similar results were also observed for ionic gelation as smaller particle size was produced from a lower molecular weight of chitosan and glutamate chitosan apparently yielded smaller particles than chitosan hydrochloride. The reason chitosan glutamate produced a smaller particle size than chitosan hydrochloride is currently still unclear. However, it has been reported that formation of chitosan nanoparticles may vary significantly depending on the purity, acid salt, and molecular weight of chitosan employed [27]. Chitosan is a weak base polysaccharide that contains an average amino group density of 0.837 per disaccharide unit and the amine groups will be protonated in acidic medium, resulting in a high positive charge. As the charge number of TPP also decreases in acidic conditions, the formation of chitosan–TPP nanoparticles by ionic gelation is highly pH-dependent which explained the changing of chitosan to TPP weight ratio from 6:1 to 5:1 when the pH of chitosan solution was changed from 6 to 4.5. For TPP ions, the decrease of solution pH to acidic conditions resulted in a reduction in the charge number of TPP accordingly which subsequently leads to the need for more TPP ions to cross-link chitosan by electrostatic forces [11]. Nevertheless, a reduction in particle size was observed when lowering the pH to 4.5 as at this point, the number of protonated amine groups was increased and could efficiently interact with siRNA and TPP ions to produce smaller and fine particles. By using agarose gel electrophoresis, the ability of chitosan to interact with siRNA was investigated. In this study, the association of siRNA either by simple mixing or by ionic gelation (siRNA entrapment or adsorption) showed retardation or immobilisation of the siRNA migration that demonstrated the strength of interaction between chitosan and siRNA. Fig. 4 demonstrates that the interaction between chitosan and siRNA by simple mixing is apparently weak as chitosan and siRNA are easily dissociated from each other compared to siRNA associated with chitosan–TPP nanoparticles at the same chitosan to siRNA weight ratio. These results were concomitant with previous study which reported that the association of ODNs was found to be more efficient when nanoparticles were formed by ionic gelation with TPP in comparison to simple complexation [27,28]. In addition, this was not observed when chitosan was complexed with pDNA in the same manner because pDNA was completely condensed to the chitosan even at a very low concentration of chitosan, (as low as 25 μg/ml or 1.25:1 chitosan to DNA ratio – data not shown). This suggests that siRNA binds to the chitosan in a different manner from that observed with pDNA. It has been reported that the volume occupied by a condensed DNA nanoparticle is about 10,000 times smaller than its uncondensed counterpart [29,30] and a minimal length of 800 bp is required for the DNA condensation process to occur [29,31–33]. Contrary to pDNA, the shorter length of siRNA (21 bp) as well as its linearity can be expected
GENE DELIVERY
H. Katas, H.O. Alpar / Journal of Controlled Release 115 (2006) 216–225
GENE DELIVERY
224
H. Katas, H.O. Alpar / Journal of Controlled Release 115 (2006) 216–225
to contribute to the weak interaction with the chitosan as it retains the initial volume after complexing and several copies of siRNA could be complexed with chitosan instead of one molecule per cationic entity. This hypothesis was supported by the weak binding between chitosan and siRNA (for chitosan– siRNA complexes) as determined by the gel retardation assay and the formation of larger complexes of chitosan–siRNA in comparison with chitosan–pDNA complexes (data not shown). Determination of gene silencing activity of these chitosan nanoparticulate systems was first performed in CHO K1 cells and chitosan to siRNA ratios of 100:1 and 10:1 were used for siRNA associated with chitosan–TPP nanoparticles and chitosan–siRNA complexes, respectively. The efficiency of gene silencing was measured as a normalised ratio of the silencing of firefly luciferase expression from pGL3 to renilla luciferase from pRL-TK which was used as an internal standard. In consideration that luciferase expression could be reduced not only by the silencing effect of siRNA but also due to cellular death, a nontargeted gene (renilla luciferase form pRL-TK) was measured to normalise the firefly luciferase expression from pGL3. The observed expression was then also represented as a relative response ratio (RRR) in which Lipofectamine 2000– siRNA complexes were used as a positive control and cells without siRNA silencing treatment (co-transfected with pGL3 and pRL-TK) as a negative control. The higher ratio of firefly to renilla luciferase expression (low percentage of gene knockdown) for naked siRNA and the cells treated with Lipofectamine 2000–mismatch siRNA complexes indicated that naked siRNA might be degraded before reaching the cytoplasm and the specificity of siRNA silencing effect was important as would be expected. The ability of these chitosan nanoparticles in delivering siRNA to the cells was apparently independent of the type or molecular weight of chitosan as no obvious correlation could be seen between these parameters with the gene silencing effect. However, the method of siRNA association to the chitosan apparently had an effect on siRNA silencing. The addition of siRNA together with the TPP ions during the ionic gelation process (siRNA entrapment) showed better gene silencing activities compared to other methods for all types of chitosan. In contrast to that, a low inhibition of gene expression was observed for particles prepared by simple complexation and adsorption of siRNA onto the preformed chitosan–TPP nanoparticles. This finding might be attributed to lack of protection against degradation due to the weak binding between the chitosan and siRNA complex (determined by gel retardation assay) or exposure of adsorbed siRNA to the nuclease activity. Further study also showed that no significant increase in the gene silencing effect was detected for the higher weight ratio of chitosan–siRNA complexes (100:1) which indicates that the lower gene silencing effect of chitosan–siRNA complexes was not due to the 10-fold lower amount of chitosan compared to siRNA associated with chitosan–TPP nanoparticles. Moreover, stability studies revealed that these chitosan nanoparticles were stable in medium containing up to 10% serum for chitosan–siRNA complexes and chitosan–TPP–siRNA nanoparticles, which indicates that poor results or gene silencing effect for those particles were unlikely to be due to aggregation as only 5%
serum was used during the transfection (data not shown). When comparing the gene silencing activity at 24 and 48 h posttransfection, higher activity was obtained at 24 than 48 h posttransfection for the cells treated with siRNA associated with chitosan nanoparticles or complexes, suggesting the release of siRNA might occur within the first 24 h for most of the tested chitosans. Although chitosan–TPP nanoparticles associated with siRNA showed higher loss of cell viability than Lipofectamine 2000 and chitosan–siRNA complexes, this effect may be due to the 26.5and 10-fold higher concentration of chitosan added to the cells (5.3 μg) compared to only 0.2 μg of Lipofectamine 2000 and 0.53 μg of chitosan–siRNA complexes, respectively. Moreover, the gene silencing effect of siRNA delivered by chitosan nanoparticles not thought to be facilitated by damaged cells or compromise the cellular membrane as no reduction in cell viability was observed after transfecting pGL3 and pRL-TK vectors into the cells by Lipofectamine 2000 (before transfecting siRNA into the cells) as determined by MTT assay after 4 h of post-incubation. 5. Conclusions In conclusion, we have demonstrated that chitosan could be used as a delivery system for siRNA and from among four types of chitosan studied, chitosan glutamate with higher molecular weight, G213 is the best candidate as a vector for siRNA. Furthermore, in vitro study has revealed the transfection efficiency of siRNA depends on the method of siRNA association to the chitosan and entrapping siRNA using ionic gelation has shown to yield a better biological effect than simple complexation or siRNA adsorption onto the chitosan nanoparticles. Acknowledgements The authors would like to thank Ministry of Science, Technology and Environment of Malaysia for the funding of this work and also Dr. S. Somavarapu as well as Dr. Xiong Wei Li for useful discussions and technical assistance. References [1] A. Khan, M. Benboubetra, P.Z. Sayyed, K.W. Ng, S. Fox, G. Beck, I.F. Benter, S. Akhtar, Sustained polymeric delivery of gene silencing antisense ODNs, siRNA, DNAzymes and ribozymes: in vitro and in vivo studies, Journal of Drug Targeting 12 (2004) 393–404. [2] G. Beale, A.J. Hollins, M. Benboubetra, M. Sohail, S.P. Fox, I. Benter, S. Akhtar, Gene silencing nucleic acids designed by scanning arrays: antiEGFR activity of siRNA, ribozyme and DNA enzymes targeting a single hybridization-accessible region using the same delivery system, Journal of Drug Targeting 11 (2003) 449–456. [3] M. Miyagishi, H. Sumimoto, H. Miyoshi, Y. Kawakami, K. Taira, Optimization of an siRNA-expression system with an improved hairpin and its significant suppressive effects in mammalian cells, The Journal of Gene Medicine 6 (2004) 715–723. [4] M. Sioud, D.R. Sorensen, Cationic liposome-mediated delivery of siRNAs in adult mice, Biochemical and Biophysical Research Communications 312 (2003) 1220–1225. [5] D.R. Sorensen, M. Leirdal, M. Sioud, Gene silencing by systemic delivery of synthetic siRNAs in adult mice, Journal of Molecular Biology 327 (2003) 761–766.
[6] C.A. Stein, Phosphorothioate antisense oligodeoxynucleotides: questions of specificity, Trends in Biotechnology 14 (1996) 147–149. [7] B. Urban-Klein, S. Werth, S. Abuharbeid, F. Czubayko, A. Aigner, RNAimediated gene-targeting through systemic application of polyethylenimine (PEI)-complexed siRNA in vivo, Gene Therapy 1–6 (2004). [8] J. Weyermann, D. Lochmann, C. Georgens, A. Zimmer, Albumin– protamine–oligonucleotide–nanoparticles as a new antisense delivery system: Part 2. Cellular uptake and effect, European Journal of Pharmaceutics and Biopharmaceutics 59 (2005) 431–438. [9] T.H. Kim, S.I. Kim, T. Akaike, C.S. Cho, Synergistic effect of poly (ethylenimine) on the transfection efficiency of galactosylated chitosan/ DNA complexes, Journal of Controlled Release 105 (2005) 354–366. [10] M.K. Lee, S.K. Chun, W.J. Choi, J.K. Kim, S.H. Choi, A. Kim, K. Oungbho, J.S. Park, W.S. Ahn, C.K. Kim, The use of chitosan as a condensing agent to enhance emulsion-mediated gene transfer, Biomaterials 26 (2005) 2147–2156. [11] X.Z. Shu, K.J. Zhu, The influence of multivalent phosphate structure on the properties of ionically cross-linked chitosan films for controlled drug release, European Journal of Pharmaceutics and Biopharmaceutics 54 (2002) 235–243. [12] S. Gao, J. Chen, L. Dong, Z. Ding, Y.h. Yang, J. Zhang, Targeting delivery of oligonucleotide and plasmid DNA to hepatocyte via galactosylated chitosan vector, European Journal of Pharmaceutics and Biopharmaceutics 60 (2005) 327–334. [13] T. Lopez-Leon, E.L.S. Carvalho, B. Seijo, J.L. Ortega-Vinuesa, D. BastosGonzalez, Physicochemical characterization of chitosan nanoparticles: electrokinetic and stability behavior, Journal of Colloid and Interface Science 283 (2005) 344–351. [14] Y. Kakizawa, S. Furukawa, K. Kataoka, Block copolymer-coated calcium phosphate nanoparticles sensing intracellular environment for oligodeoxynucleotide and siRNA delivery, Journal of Controlled Release 97 (2004) 345–356. [15] X.W. Li, D.K.L. Lee, A.S.C. Chan, H.O. Alpar, Sustained expression in mammalian cells with DNA complexed with chitosan nanoparticles, Biochimica et Biophysica Acta (BBA) – Gene Structure and Expression 1630 (2003) 7–18. [16] R.M. Schiffelers, M.C. Woodle, P. Scaria, Pharmaceutical prospects for RNA interference, Pharmaceutical Research 21 (2004) 1–7. [17] H.J. Jeon, Y.I. Jeong, M.K. Jang, Y.H. Park, J.W. Nah, Effect of solvent on the preparation of surfactant-free poly(-lactide-co-glycolide) nanoparticles and norfloxacin release characteristics, International Journal of Pharmaceutics 207 (2000) 99–108. [18] D.A. Braasch, S. Jensen, Y. Liu, K. Kaur, K. Arar, M.A. White, D.R. Corey, RNA interference in mammalian cells by chemically-modified RNA, Biochemistry 42 (2003) 7967–7975.
225
[19] Q. Gan, T. Wang, C. Cochrane, P. McCarron, Modulation of surface charge, particle size and morphological properties of chitosan–TPP nanoparticles intended for gene delivery, Colloids and Surfaces. B, Biointerfaces 44 (2005) 65–73. [20] R. Gref, A. Domb, P. Quellec, T. Blunk, R.H. Muller, J.M. Verbavatz, R. Langer, The controlled intravenous delivery of drugs using PEG-coated sterically stabilized nanospheres, Advanced Drug Delivery Reviews 16 (1995) 215–233. [21] M. Bivas-Benita, S. Romeijn, H.E. Junginger, G. Borchard, PLGA–PEI nanoparticles for gene delivery to pulmonary epithelium, European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 1–6. [22] J. Panyam, V. Labhasetwar, Biodegradable nanoparticles for drug and gene delivery to cells and tissue, Advanced Drug Delivery Reviews 55 (2003) 329–347. [23] W. Zauner, N.A. Farrow, A.M.R. Haines, In vitro uptake of polystyrene microspheres: effect of particle size, cell line and cell density, Journal of Controlled Release 71 (2001) 39–51. [24] M. Huang, C.W. Fong, E. Khor, L.Y. Lim, Transfection efficiency of chitosan vectors: effect of polymer molecular weight and degree of deacetylation, Journal of Controlled Release 106 (2005) 391–406. [25] V. Chan, H.Q. Mao, K.W. Leong, Effect of Chitosan Molecular Weight in Gene Delivery Process, American Chemical Society, 2001 221zt-COLL260. [26] D.B. Rozema, D.L. Lewis, siRNA delivery technologies for mammalian systems, Targets 2 (2003) 253–260. [27] K.A. Janes, P. Calvo, M.J. Alonso, Polysaccharide colloidal particles as delivery systems for macromolecules, Advanced Drug Delivery Reviews 47 (2001) 83–97. [28] P. Calvo, A. Bougaba, M. Appel, E. Fattal, M.J. Alonso, P. Couvreur, Oligonucleotide–chitosan nanoparticles as new gene therapy vector, Proc. World Meet. APGI/APV, vol. 2, 1998, pp. 1111–1112. [29] M. Keller, Lipidic carriers of RNA/DNA oligonucleotides and polynucleotides: what a difference a formulation makes! Journal of Controlled Release 103 (2005) 537–540. [30] P. Tam, M. Monck, D. Lee, O. Ludkovski, E.C. Leng, K. Clow, H. Stark, P. Scherrer, R.W. Graham, P.R. Cullis, Stabilized plasmid–lipid particles for systemic gene therapy, Gene Therapy 7 (2000) 1867–1874. [31] V.A. Bloomfield, S. He, A.Z. Li, P.B. Arscott, Light scattering studies on DNA condensation, Biochemical Society Transactions 19 (1991) 496. [32] V.A. Bloomfield, Condensation of DNA by multivalent cations: considerations on mechanism, Biopolymers 31 (1991) 1471–1481. [33] V.A. Bloomfield, DNA condensation, Current Opinion in Structural Biology 6 (1996) 334–341.
GENE DELIVERY
H. Katas, H.O. Alpar / Journal of Controlled Release 115 (2006) 216–225
Journal of Controlled Release 115 (2006) 216 – 225 www.elsevier.com/locate/jconrel
Development and characterisation of chitosan nanoparticles for siRNA delivery Haliza Katas, H. Oya Alpar ⁎ School of Pharmacy, University of London, 29-39, Brunswick Square, London, WC1N 1AX, UK Received 20 April 2006; accepted 20 July 2006 Available online 25 July 2006
Abstract Gene silencing mediated by double-stranded small interfering RNA (siRNA) has been widely investigated as a potential therapeutic approach for diseases with genetic defects. The use of siRNA, however, is hampered by its rapid degradation and poor cellular uptake into cells in vitro or in vivo. Therefore, we have explored chitosan as a siRNA vector due to its advantages such as low toxicity, biodegradability and biocompatibility. Chitosan nanoparticles were prepared by two methods of ionic cross-linking, simple complexation and ionic gelation using sodium tripolyphosphate (TPP). Both methods produced nanosize particles, less than 500 nm depending on type, molecular weight as well as concentration of chitosan. In the case of ionic gelation, two further factors, namely chitosan to TPP weight ratio and pH, affected the particle size. In vitro studies in two types of cells lines, CHO K1 and HEK 293, have revealed that preparation method of siRNA association to the chitosan plays an important role on the silencing effect. Chitosan–TPP nanoparticles with entrapped siRNA are shown to be better vectors as siRNA delivery vehicles compared to chitosan–siRNA complexes possibly due to their high binding capacity and loading efficiency. Therefore, chitosan– TPP nanoparticles show much potential as viable vector candidates for safer and cost-effective siRNA delivery. © 2006 Elsevier B.V. All rights reserved. Keywords: siRNA; Chitosan; Delivery system; Simple complexation; Ionic gelation
1. Introduction Over the past few decades, antisense approaches including oligonucleotides, ribozymes and DNAzymes have been extensively investigated as tools for controlling cellular processes. Only recently, small interfering RNAs (siRNAs) have proven to be versatile agents for controlling gene expression in mammalian cells. They have also been shown more potent than conventional antisense strategies [1–3]. Overall, they appear to be a much more robust and efficient technology offering significant potential. siRNA consisting of 21–23 nucleotides can regulate gene expression in mammalian cells through RNA interfering (RNAi). As the administration of siRNA could bypass nonspecific inhibition of protein synthesis induced by long doublestranded RNA [4,5], it has therefore been employed as a novel tool in blocking the expression of genes such as those expressed ⁎ Corresponding author. Tel.: +44 207 753 5928; fax: +44 207 753 5975. E-mail address: [email protected] (H.O. Alpar). 0168-3659/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.jconrel.2006.07.021
in infectious diseases and cancers. However, similar to hydrophilic and polyanion-mediated Gene Therapy, siRNA also suffers particular problems including poor cellular uptake, rapid degradation by ubiquitous nucleases as well as limited blood stability [6,7]. As a result of these limitations, unassisted delivery of siRNA to the cells is frustrating. Although various chemical modifications of siRNA can be used to overcome these problems, these modifications posses disadvantages such as a decreased mRNA hybridization, higher cytotoxicity and increased unspecific effects [8]. Therefore, effective systems which can protect and transport siRNA to the cytoplasm of the target cells are needed to exploit the promising potential applications offered by successful delivery of siRNA. From among the gene vectors that have been studied, nonviral vectors have attracted more and more attention in comparison to viral vectors, although viral vectors have been proven to yield higher transfection efficiency in most cell lines. This is due to the advantages of non-viral vectors such as ease of synthesis, low immune response against the vector and unrestricted gene materials size in addition to potential benefits in
terms of safety [9]. In recent years, chitosan-based carriers are one of the non-viral vectors that have gained increasing interest as a safer and cost-effective delivery system for gene materials including plasmid DNA (pDNA), oligonucleotide (ODN) as well as proteins and peptides. Chitosan has beneficial qualities such as low toxicity, low immunogenicity, [10] excellent biodegradability, biocompatibility, [10,11] as well as a high positive charge that can easily form polyelectrolyte complexes with negatively charged nucleotides by electrostatic interaction. Although chitosan has been studied for more than a decade as a gene vector for pDNA and ODN [12], so far to our knowledge, there is no study that has been carried out to investigate the use of chitosan to deliver siRNA in vitro. In this report, we studied three methods of siRNA association; by simple complexation, ionic gelation (siRNA entrapment) and adsorption of siRNA onto the surface of preformed chitosan nanoparticles for preparation of chitosan-based nano-carriers. To address the lack of information regarding the behaviour of interaction between siRNA and chitosan – since siRNA's structure and size is quite different to that of pDNA, these systems were fully characterised with regards to their physical and biological features by exploiting commercial chitosan products. The ability of these nanoparticulate systems to mediate gene silencing was then assessed on CHO K1 and HEK 293 cells in vitro. 2. Materials and methods 2.1. Materials Four different types of medical grade chitosan with the degree of deacetylation (DD) of ∼ 86% were used: chitosan hydrochloride with molecular weight of 270 kDa (Cl213) and 110 kDa (Cl113) as well as chitosan glutamate with molecular weight of 470 kDa (G213) and 160 kDa (G113) (Protasan Ultrapure, Pronova Biomedical, Norway). Pentasodium tripolyphosphate (TPP) was obtained from Fluka and sodium acetate from Sigma-Aldrich. siRNA targeting against pGL3 luciferase gene (sense: 5′–CUUACGCUGAGUACUUCGATT–3′, antisense: 3′–TTGAAUGCGACUCAUGAAGCU–5′) and control siRNA (non-silencing) (sense: 3′–UUCUCCGAACGUGU CACGUTT–3′, antisense: 3′–TTAAGAGGCUUGCACA GUGCA–5′) were synthesized by Proligo (France). Reporter vectors, pGL3 control and pRL-TK which encoded firefly luciferase and renilla luciferase gene, respectively were purchased from Promega, UK. Dual-Glo Luciferase Assay system, agarose (low melting point) and Tris–borate–EDTA buffer pH 8.3 were from Promega, UK. Lipofectamine 2000 and OptiMEM I reduced serum medium were purchased from Invitrogen, UK. Other reagents were all commercially available and were of analytical grade. 2.2. Chitosan nanoparticles preparation 2.2.1. Simple complexation Four types of chitosan (chitosan hydrochloride and glutamate with two different molecular weights) were dissolved separately
217
either in distilled water or acetate buffer (0.1 M sodium acetate/ 0.1 M acetic acid, pH 4.5) to form different concentrations of chitosan solution, ranging from 25 to 300 μg/ml. ChitosansiRNA complexes were prepared by adding chitosan solution drop-wise to an equal volume of siRNA solution (20 μg/ml) and quickly mixed before incubating them at room temperature for 30 min to form chitosan-siRNA complexes with the chitosan to siRNA weight ratio of 1.25:1 to 15:1. 2.2.2. Ionic gelation Nanoparticles were produced based on modified ionic gelation of tripolyphosphate (TPP) with chitosan as described elsewhere [13]. Two different types of chitosan (chitosan hydrochloride and glutamate) and each type with two different molecular weights were investigated. Nanoparticles were spontaneously obtained upon the addition of 1.2 ml of a TPP aqueous solution (0.84 mg/ml) to 3 ml of chitosan solution (2 mg/ml, at chitosan to TPP weight ratio of 6:1) under constant magnetic stirring at room temperature. The particles were then incubated at room temperature for 30 min before use or further analysis. Nanoparticles were collected by centrifugation (IEA Micromax Ultracentrifuge) at 13,000×g for 10 min. The supernatants were discarded and nanoparticles were resuspended in filtered (Millex® GP filter unit, Millipore, Ireland; 0.25 μm) distilled water. 2.2.2.1. siRNA entrapment in chitosan–TPP nanoparticles. For the association of siRNA with the chitosan–TPP nanoparticles (chitosan–TPP–siRNA), 3 μl of siRNA (19.95 μg/μl) in double distilled water was added to the TPP solution (1.2 ml, 0.84 mg/ ml) before adding this drop-wise to the chitosan solution (3 ml, 2 mg/ml) under constant magnetic stirring at room temperature. The particles were then incubated at room temperature for 30 min before use or further analysis. 2.2.2.2. siRNA adsorption onto chitosan–TPP nanoparticles. Preprepared chitosan nanoparticles prepared by ionic gelation as described above were dispersed in distilled water to yield a chitosan concentration, ranging from 0.1 to 1 mg/ml. To adsorb siRNA onto the surface of chitosan nanoparticles, 500 μl of siRNA solution (10 μg/ml in distilled water) was added dropwise to 500 μl of chitosan suspension and quickly mixed by inverting the interaction tube up and down. Then, the particles were incubated for 2 h at room temperature before further analysis. 2.3. Characterisation of the chitosan nanoparticles or chitosan siRNA nanoparticles Mean particle diameter (Z-average) and zeta potential of the nanoparticles were determined using Submicron Particle Analyser System 4700 and Zetasizer® S (Malvern Instruments, UK), respectively. The measurements of particle size were made at 25 °C in triplicate and no further dilution was performed for these particles. For the determination of zeta potential, samples were diluted with distilled water to an appropriate concentration to yield count rate per second (KCps) in the range of 2500–3500. Each batch was analysed in triplicate.
GENE DELIVERY
H. Katas, H.O. Alpar / Journal of Controlled Release 115 (2006) 216–225
GENE DELIVERY
218
H. Katas, H.O. Alpar / Journal of Controlled Release 115 (2006) 216–225
2.4. Determination of siRNA loading efficiency The loading efficiency of siRNA (%) entrapped or adsorbed onto the chitosan nanoparticles was obtained from the determination of free siRNA concentration in the supernatant recovered after particle centrifugation (13,000×g, 15 min) by absorbance measurement (UV–Visible Spectrophotometer Cary 3E, Varian) at 260 nm. Supernatant recovered from unloaded chitosan–TPP nanoparticles (without siRNA) was used as a blank. siRNA loading efficiency (%) was the percentage of entrapped or adsorbed siRNA, (difference between the total amount of siRNA added for the nanoparticles preparation and the amount of nonentrapped or -adsorbed siRNA remaining in supernatant after centrifugation) to the total amount of siRNA added.
well plate at a density of 30,000 cells per well in Opti-MEM 1 reduced serum medium containing 5% of FBS without antibiotics, 24 h prior to transfection. On the day of transfection, pGL3-control (0.15 μg) and pRL-TK vectors (0.05 μg) were cotransfected to the cells using Lipofectamine 2000™ according to manufacturer's instructions. After 4 h of transfection, the medium was removed and the cells were washed with PBS, following this, media was replaced with 100 μl fresh medium containing serum. 50 μl of chitosan–siRNA nanoparticles, siRNA alone or Lipofectamine 2000–siRNA complexes (each well or formulation contained 4 pmol of siRNA) in the medium without serum were then added to the cells and incubated at 37 °C with a 5% CO2 atmosphere for 24 or 48 h. After 24 or 48 h, luciferase activities were determined using Dual-Glo Luciferase Assay System [14].
2.5. Gel retardation assay 2.8. Cytotoxicity assay The binding of siRNA with chitosan was determined by 4% agarose (low melting point) gel electrophoresis. A series of different chitosan to siRNA weight ratios was loaded (20 μl of the sample containing 0.2 μg of siRNA). A 1:6 dilution of loading dye was added to each well and electrophoresis was carried out at a constant voltage of 55 V for 2 h in TBE buffer (4.45 mM Tris–base, 1 mM sodium EDTA, 4.45 mM boric acid, pH 8.3) containing 0.5 μg/ml ethidium bromide. The siRNA bands were then visualised under a UV transilluminator at a wavelength of 365 nm. 2.6. Assay for serum stability Chitosan–siRNA nanoparticles (200 μl, equivalent to 5 μg of siRNA) were incubated at 37 °C with equal volume of Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 5% and 50% final concentration of foetal bovine serum (FBS). At each predetermined time interval (0, 30 min, 2, 4, 7, 24 48 and 72 h), 30 μl of the mixture was removed and stored at − 20 °C until gel electrophoresis was performed. To terminate serum activity, samples were incubated in bath incubator at 80 °C for 5 min and 5 μl heparin (1000 U/ml) was added for displacing the siRNA from the chitosan nanoparticles. The integrity of the siRNA was then analysed by 15% polyacrylamide gel containing 7 M urea and TBE buffer (0.089 M Tris base, 0.089 M boric acid, and 2 mM sodium EDTA, pH 8.3). Polyacrylamide–Urea gel (15%) was used due to its high efficiency in separating small fragments of possibly degraded siRNA. Electrophoresis was then carried out with 1× TBE buffer at a constant voltage of 200 V for 1 h. siRNA bands were visualised under a UV transilluminator after staining for 40 min with a 1:10,000 dilution of SYBR-Green II RNA gel stain (Molecular Probes) prepared in DEPC treated water. SYBRGreen for RNA was chosen as it was more sensitive than ethidium bromide in staining small RNA. 2.7. Biological activity of chitosan–siRNA nanoparticles In vitro transfection studies were performed in CHO K1 (Chinese Hamster Ovary) cells. The cells were seeded in a 96-
The effect of chitosan on cytotoxicity was measured by determining cell viability of chitosan–siRNA nanoparticles, calculated as a percentage of the cell viability of untreated/nontransfected cell samples. CHO K1 cells were seeded in a 96-well plate at a density of 30,000 cells per well in Opti-MEM 1 reduced serum medium containing 5% of FBS and grown overnight. After 24 and 48 h post-incubation of chitosan–siRNA nanoparticles or complexes at 37 °C, 20 μl of MTT (3-(4,5-dimethyl-thiazol-2-yl)2,5-diphenyl-tetrazolium bromide, 5 mg/ml) in sterile filtered PBS was added to each well and then incubated for 4 h to allow formation of formazan crystals. After 4 h, the unreduced MTT and medium was removed and the cells were washed with PBS. 200 μl of DMSO was then added to each well to dissolve the MTT formazan crystals and the plate was incubated at 37 °C for 5 min. The absorbance of formazan products was measured at 540 nm using a microplate reader (Wallac Victor2 1420 Multilabel Counter). 2.9. Statistical analysis Data are presented as the means ± standard deviations. The statistical significance was determined by using Independent Sample t-test or one-way analysis of variance (ANOVA). P values of b 0.05 were considered significant. The statistical analyses were carried out using SPSS Base 12.0 for Windows (SPSS). 3. Results 3.1. Particle size Mean particle size of chitosan–siRNA complexes prepared by simple complexation for both chitosan hydrochloride and glutamate were increased when the concentration of chitosan was increased from 25 to 300 μg/ml in distilled water (Fig. 1). However, a smaller mean particle size of chitosan nanoparticles was obtained when the lower molecular weight of chitosan was used compared to the higher molecular weight for the individual chitosan derivatives. In addition, chitosan glutamate (G213, G113) which had a higher molecular weight than chitosan
hydrochloride (Cl213, Cl113) produced smaller complexes with siRNA than chitosan hydrochloride. On the other hand, the formation of chitosan–TPP nanoparticles occurred spontaneously upon the addition of the TPP ions into the chitosan solution. The results showed that the appearance of the solution changed when a certain amount of TPP ions was added to the chitosan solution, from a clear to opalescent solution that indicated a change of the physical states of the chitosan to form nanoparticles, then microparticles and eventually aggregates [15]. For all types of chitosan tested in this study, chitosan and TPP concentration were adjusted to a weight ratio of 6:1 to obtain nanoparticles and the mean particle size of chitosan–TPP nanoparticles was 510 ± 22.9, 276 ± 17.9, 709 ± 50.3 and 415 ± 44.6 nm for G213, G113, Cl213 and Cl113, respectively. The chitosan nanoparticles entrapped with siRNA, however showed no significant difference in particle size compared to the plain particles. The weight ratio of chitosan to TPP however was changed from 6:1 to 5:1 as the pH of chitosan solution used in the ionic gelation was changed from pH 6 (distilled water) to pH 4.5 (acetate buffer, 0.1 M). 3.2. Surface charge The comparative positive value of surface charge (zeta potential) of the chitosan–siRNA complexes increased with the increasing concentration of chitosan at a constant siRNA concentration (Fig. 2). The increment was due to the increase in the number of positive charges which counteracts with negatively charged siRNA as the amount of siRNA was fixed. The net positive charge of the particles was desirable to prevent particle aggregation and promote electrostatic interaction with the overall negative charge of the cell membrane [16]. In addition, body distribution of nanoparticles after i.v. injection is highly influenced by their interaction with the biological environment which also dependent on their physicochemical properties like surface charge of nanoparticles [17]. The values of particle surface charge for each of the tested chitosans are presented in the Fig. 2. The surface charge of chitosan–TPP nanoparticles was ranged from approximately from + 40 to + 60 mV by changing the weight ratio of chitosan to TPP from 4:1 to 6:1. The addition of siRNA (10 μg) at the chitosan to TPP
Fig. 1. The effect of chitosan molecular weight and concentration (or chitosan to siRNA weight ratio) on mean particle diameter of chitosan–siRNA complex (n = 3).
219
Fig. 2. Effect of chitosan concentration on particle surface charge of chitosan– siRNA complexes (n = 3).
weight ratio of 6 reduced the surface charge of nanoparticles to approximately + 30 mV. 3.3. Effect of pH To investigate the effect of pH on the interaction between TPP ions and the chitosan derivatives that were used here, chitosan was dissolved in acetate buffer (0.1 M, pH ranging from 4 to 6) at a fixed concentration of 2 mg/ml and a chitosan to TPP mass ratio of 5:1. Interestingly, the smallest particles were formed at pH 4.5 (Fig. 3) for all the tested chitosans except for Cl213 (pH 6). In addition, the surface charge of the yielded chitosan–TPP nanoparticles was increased approximately twofold to + 40 mV when the pH of the chitosan solution was changed from 6 to 4. In contrast, investigation into the influence of pH on the particle size revealed that no significant difference in particle size was observed when complexing chitosan with siRNA in acetate buffer at pH 4.5 (0.1 M) compared to in distilled water for all the tested chitosans. 3.4. Interaction of siRNA with chitosan For siRNA adsorbed onto the chitosan nanoparticles, complete binding of siRNA with the chitosan (without the presence of a trailing band) could only be observed when the chitosan nanoparticles to siRNA weight ratio was approaching 100:1 except for lower molecular weight chitosan hydrochloride, Cl113 (Fig. 4B) whereas all the tested chitosans showed complete
Fig. 3. Effect of pH on mean particle size of chitosan–TPP nanoparticles (n = 6).
GENE DELIVERY
H. Katas, H.O. Alpar / Journal of Controlled Release 115 (2006) 216–225
GENE DELIVERY
220
H. Katas, H.O. Alpar / Journal of Controlled Release 115 (2006) 216–225
Fig. 4. Binding efficiency of siRNA to the chitosan or chitosan associated with TPP nanoparticles: (A) chitosan-siRNA complexes, (B) chitosan–TPP with adsorbed siRNA and (C) chitosan–TPP–siRNA nanoparticles (entrapped siRNA).
binding of siRNA for chitosan–TPP nanoparticles with entrapped siRNA (Fig. 4C). On the other hand, the migration of siRNA from the chitosan–siRNA complex was observed even at higher concentrations of chitosan (1 mg/ml) or higher chitosan to siRNA weight ratio (100:1) as shown on Fig. 4A. Similar observations were also observed for the chitosan–siRNA complexes prepared in acetate buffer at pH 4.5 as determined by gel retardation assay.
chitosan–TPP with adsorbed siRNA nanoparticles at a weight ratio of 100:1 were varied and depended on the type of chitosan used. Higher siRNA loading efficiency was observed for siRNA adsorbed onto chitosan glutamate (83% ± 0.9% for G213 and 90% ± 0.3% for G113) compared to chitosan hydrochloride (72% ± 1.1% for Cl213 and 59% ± 0.8% for Cl113). 3.6. siRNA stability in serum
3.5. siRNA loading efficiency 100% siRNA loading efficiency was obtained as measured by spectrophotometry for all the entrapped siRNA chitosan– TPP nanoparticles. However, the loading efficiencies of
siRNA must be stable to digestion by nuclease for maximal activity in cells [18]. Therefore, to address the question of chitosan–siRNA nanoparticles stability and protection from nuclease degradation, they were incubated in 5% and 50% of
FBS at 37 °C. Contrasting results were obtained in a previous report by Braasch et al. [18] who reported that naked siRNA was stable after incubation in 5% serum up to 72 h. However, in this study siRNA was intact only up to 30 min and it was fully degraded after 48 h. On the other hand, siRNA recovered from chitosan–TPP nanoparticles started to degrade after 24 h incubation and full degradation was only observed after 72 h incubation (Fig. 5A). This experiment was therefore repeated by incubating siRNA as well as chitosan–siRNA nanoparticles in a higher serum concentration. The chitosan–siRNA nanoparticles significantly protected siRNA from nuclease activity. Complete degradation of siRNA was observed as early as time-point 0 with degradation occurring during the mixing of siRNA with serum and freezing steps. In contrast with unformulated siRNA, the siRNA recovered from chitosan–siRNA nanoparticles was intact up to 7 h and fully degraded after 48 h incubation in 50% serum (Fig. 5B).
221
negligible silencing effect was observed for naked siRNA as well as Lipofectamine 2000–mismatch siRNA (non-silencing) complexes. In this study, chitosan glutamate, G213 (470 kDa) showed the highest gene silencing effect at 24 h post-transfection either by simple complexation (51% of gene knockdown) or ionic gelation (82% and 63% of gene knockdown for siRNA entrapment and siRNA adsorption, respectively) compared to its lower molecular weight or chitosan hydrochloride. siRNA entrapped in chitosan–TPP nanoparticles prepared from chitosan glutamate, G213 even showed a slightly higher efficiency than Lipofectamine 2000 (Fig. 6B). However, transfection studies performed in HEK 293 human kidney cells showed lower gene silencing activities for these chitosan nanoparticles either at 24 or 48 h post-transfection compared to CHO K1 cells. Chitosan G213 for example was 22%, 14% and 64% less efficient than Lipofectamine 2000 for chitosan–TPP with adsorbed siRNA, chitosan–TPP with entrapped siRNA and chitosan–siRNA complex, respectively (Fig 6B).
3.7. Biological activity of chitosan–siRNA nanoparticles 3.8. Cytotoxicity study Transfections performed with nanoparticles prepared with different types of chitosan and methods of siRNA association to the chitosan revealed that the silencing effect of certain chitosan–siRNA nanoparticles or complexes had a comparable effect compared to Lipofectamine 2000 (Fig. 6B). In contrast, a
To investigate the potential cytotoxicity of chitosan–siRNA nanoparticulate systems, the cell viability was determined by MTT assays. Over 90% average cell viability was observed for chitosan–siRNA complex and naked siRNA in comparison to
Fig. 5. Electrophoretic mobility of chitosan–TPP–siRNA nanoparticles following incubation with FBS: (A) 5% FBS and (B) 50% FBS.
GENE DELIVERY
H. Katas, H.O. Alpar / Journal of Controlled Release 115 (2006) 216–225
GENE DELIVERY
222
H. Katas, H.O. Alpar / Journal of Controlled Release 115 (2006) 216–225
Fig. 6. Effect of chitosan types and method of siRNA association to the chitosan nanoparticles on gene silencing: (A) percentage of gene knockdown of pGL3 luciferase in CHO K1 cells at 24 and 48 h post-transfection and (B) relative response ratio of chitosan nanoparticles prepared from chitosan glutamate, G213 in CHO K1 and HEK 293 cells at 24 h post-transfection. Lipo – Lipofectamine 2000; siRNA M – siRNA mismatch.
Fig. 7. Effect of chitosan types and method of siRNA association to the chitosan nanoparticles on percentage of cell viability of CHO K1 (n = 3).
untreated cells. However, 18–40% loss of cell viability was observed for siRNA associated with chitosan–TPP nanoparticles although individual TPP and chitosan solution did not show any loss of cell viability (data not shown). The cell viabilities observed with siRNA associated with chitosan–TPP nanoparticles were however not significantly different between the different chitosan derivatives. Nevertheless, the cell viability for certain siRNA associated with chitosan–TPP formulations was increased at 48 h post-incubation which indicates the recovery of cells was observed within this period (Independent Sample ttest, p b 0.05). From the studied formulations, chitosan G113 showed significant recovery of cells either by adsorbing or incorporating siRNA with chitosan–TPP nanoparticles during the preparation process (chitosan–TPP–siRNA). However, for Cl213, significant recovery was observed when adsorbing
siRNA onto the chitosan–TPP nanoparticles. In contrast, Cl113 showed high recovery of the cells when siRNA was incorporated to the chitosan during the ionic gelation process (Fig. 7). Although only a small improvement of cell viability was observed for other chitosan–TPP formulations, this finding still gave an indication of a transient physiological stress of the cells by these formulations. 4. Discussion In this study, the simple complexation and ionic gelation methods were chosen to prepare chitosan–siRNA complexes or nanoparticles since both processes are simple and mild [11]. In addition, the use of TPP in ionic gelation as a polyanion to cross-link with the cationic chitosan through electrostatic interaction could avoid possible toxicity of reagents used in chemical cross-linking (e.g. glutaraldehyde). Modulation of size and surface charge of the particles could also be easily done by using both methods [19], by adjusting certain process parameters such as chitosan concentration and stirring rate. Adsorption of siRNA onto the surface of preformed chitosan nanoparticles is also another simple approach to associate siRNA to the chitosan nanoparticles. Therefore, it was investigated in this study as another option for associating siRNA to the chitosan. Particle size and shape play an important role in transferring genes to the cells and they also greatly influence particle distribution in the body [20]. It has been reported that particles in the nanometer size have a relatively higher intracellular uptake compared to microparticles [21–23]. This characteristic is very important in gene transfer because cellular uptake of the chitosan/DNA complex as well as its subsequent release from the endo-lysosome pathway are two of the rate limiting steps in this process [24,25]. Similar to pDNA and ODN, siRNA are likely taken up by cells through endocytosis and if siRNA is not effectively delivered to the cytoplasmic compartment, it will not induce RNAi [26]. This is because the uptake of siRNA by fluid phase endocytosis does not result in the release of siRNA into the cytoplasm and mammalian cells appear to lack of the effective double stranded RNA-uptake machinery that is found in other species such as Caenorhabditis elegans [26]. Therefore, siRNA and its vector should be able to be taken up by the cells and escape the endosomal vesicle to avoid lysosomal degradation, thus, allow the RNAi to occur. In this study, we tried to obtain nanoparticles of less than 500 nm to facilitate the uptake of the particles. This can be achieved by manipulating certain process parameters that are involved in nanoparticle preparation. Generally, particle size of chitosan–siRNA complexes or nanoparticles was highly dependent on the molecular weight, type and concentration of chitosan used. However, in the case of ionic gelation, other factors like weight ratio of chitosan to TPP and pH of the chitosan solution were also shown to affect the particle size. In this study, smaller particle size was obtained with the lower molecular weight or concentration for both types of chitosan. This was expected due to the decreased viscosity of the lower concentration or molecular weight of chitosan which
223
resulted in better solubility and generated its molecular character as a polyelectrolyte material that allows more efficient interaction between negatively charged siRNA and the cationic chitosan. In addition, a smaller particle size was obtained with the chitosan glutamate compared to chitosan hydrochloride although it had a higher molecular weight than chitosan hydrochloride. Similar results were also observed for ionic gelation as smaller particle size was produced from a lower molecular weight of chitosan and glutamate chitosan apparently yielded smaller particles than chitosan hydrochloride. The reason chitosan glutamate produced a smaller particle size than chitosan hydrochloride is currently still unclear. However, it has been reported that formation of chitosan nanoparticles may vary significantly depending on the purity, acid salt, and molecular weight of chitosan employed [27]. Chitosan is a weak base polysaccharide that contains an average amino group density of 0.837 per disaccharide unit and the amine groups will be protonated in acidic medium, resulting in a high positive charge. As the charge number of TPP also decreases in acidic conditions, the formation of chitosan–TPP nanoparticles by ionic gelation is highly pH-dependent which explained the changing of chitosan to TPP weight ratio from 6:1 to 5:1 when the pH of chitosan solution was changed from 6 to 4.5. For TPP ions, the decrease of solution pH to acidic conditions resulted in a reduction in the charge number of TPP accordingly which subsequently leads to the need for more TPP ions to cross-link chitosan by electrostatic forces [11]. Nevertheless, a reduction in particle size was observed when lowering the pH to 4.5 as at this point, the number of protonated amine groups was increased and could efficiently interact with siRNA and TPP ions to produce smaller and fine particles. By using agarose gel electrophoresis, the ability of chitosan to interact with siRNA was investigated. In this study, the association of siRNA either by simple mixing or by ionic gelation (siRNA entrapment or adsorption) showed retardation or immobilisation of the siRNA migration that demonstrated the strength of interaction between chitosan and siRNA. Fig. 4 demonstrates that the interaction between chitosan and siRNA by simple mixing is apparently weak as chitosan and siRNA are easily dissociated from each other compared to siRNA associated with chitosan–TPP nanoparticles at the same chitosan to siRNA weight ratio. These results were concomitant with previous study which reported that the association of ODNs was found to be more efficient when nanoparticles were formed by ionic gelation with TPP in comparison to simple complexation [27,28]. In addition, this was not observed when chitosan was complexed with pDNA in the same manner because pDNA was completely condensed to the chitosan even at a very low concentration of chitosan, (as low as 25 μg/ml or 1.25:1 chitosan to DNA ratio – data not shown). This suggests that siRNA binds to the chitosan in a different manner from that observed with pDNA. It has been reported that the volume occupied by a condensed DNA nanoparticle is about 10,000 times smaller than its uncondensed counterpart [29,30] and a minimal length of 800 bp is required for the DNA condensation process to occur [29,31–33]. Contrary to pDNA, the shorter length of siRNA (21 bp) as well as its linearity can be expected
GENE DELIVERY
H. Katas, H.O. Alpar / Journal of Controlled Release 115 (2006) 216–225
GENE DELIVERY
224
H. Katas, H.O. Alpar / Journal of Controlled Release 115 (2006) 216–225
to contribute to the weak interaction with the chitosan as it retains the initial volume after complexing and several copies of siRNA could be complexed with chitosan instead of one molecule per cationic entity. This hypothesis was supported by the weak binding between chitosan and siRNA (for chitosan– siRNA complexes) as determined by the gel retardation assay and the formation of larger complexes of chitosan–siRNA in comparison with chitosan–pDNA complexes (data not shown). Determination of gene silencing activity of these chitosan nanoparticulate systems was first performed in CHO K1 cells and chitosan to siRNA ratios of 100:1 and 10:1 were used for siRNA associated with chitosan–TPP nanoparticles and chitosan–siRNA complexes, respectively. The efficiency of gene silencing was measured as a normalised ratio of the silencing of firefly luciferase expression from pGL3 to renilla luciferase from pRL-TK which was used as an internal standard. In consideration that luciferase expression could be reduced not only by the silencing effect of siRNA but also due to cellular death, a nontargeted gene (renilla luciferase form pRL-TK) was measured to normalise the firefly luciferase expression from pGL3. The observed expression was then also represented as a relative response ratio (RRR) in which Lipofectamine 2000– siRNA complexes were used as a positive control and cells without siRNA silencing treatment (co-transfected with pGL3 and pRL-TK) as a negative control. The higher ratio of firefly to renilla luciferase expression (low percentage of gene knockdown) for naked siRNA and the cells treated with Lipofectamine 2000–mismatch siRNA complexes indicated that naked siRNA might be degraded before reaching the cytoplasm and the specificity of siRNA silencing effect was important as would be expected. The ability of these chitosan nanoparticles in delivering siRNA to the cells was apparently independent of the type or molecular weight of chitosan as no obvious correlation could be seen between these parameters with the gene silencing effect. However, the method of siRNA association to the chitosan apparently had an effect on siRNA silencing. The addition of siRNA together with the TPP ions during the ionic gelation process (siRNA entrapment) showed better gene silencing activities compared to other methods for all types of chitosan. In contrast to that, a low inhibition of gene expression was observed for particles prepared by simple complexation and adsorption of siRNA onto the preformed chitosan–TPP nanoparticles. This finding might be attributed to lack of protection against degradation due to the weak binding between the chitosan and siRNA complex (determined by gel retardation assay) or exposure of adsorbed siRNA to the nuclease activity. Further study also showed that no significant increase in the gene silencing effect was detected for the higher weight ratio of chitosan–siRNA complexes (100:1) which indicates that the lower gene silencing effect of chitosan–siRNA complexes was not due to the 10-fold lower amount of chitosan compared to siRNA associated with chitosan–TPP nanoparticles. Moreover, stability studies revealed that these chitosan nanoparticles were stable in medium containing up to 10% serum for chitosan–siRNA complexes and chitosan–TPP–siRNA nanoparticles, which indicates that poor results or gene silencing effect for those particles were unlikely to be due to aggregation as only 5%
serum was used during the transfection (data not shown). When comparing the gene silencing activity at 24 and 48 h posttransfection, higher activity was obtained at 24 than 48 h posttransfection for the cells treated with siRNA associated with chitosan nanoparticles or complexes, suggesting the release of siRNA might occur within the first 24 h for most of the tested chitosans. Although chitosan–TPP nanoparticles associated with siRNA showed higher loss of cell viability than Lipofectamine 2000 and chitosan–siRNA complexes, this effect may be due to the 26.5and 10-fold higher concentration of chitosan added to the cells (5.3 μg) compared to only 0.2 μg of Lipofectamine 2000 and 0.53 μg of chitosan–siRNA complexes, respectively. Moreover, the gene silencing effect of siRNA delivered by chitosan nanoparticles not thought to be facilitated by damaged cells or compromise the cellular membrane as no reduction in cell viability was observed after transfecting pGL3 and pRL-TK vectors into the cells by Lipofectamine 2000 (before transfecting siRNA into the cells) as determined by MTT assay after 4 h of post-incubation. 5. Conclusions In conclusion, we have demonstrated that chitosan could be used as a delivery system for siRNA and from among four types of chitosan studied, chitosan glutamate with higher molecular weight, G213 is the best candidate as a vector for siRNA. Furthermore, in vitro study has revealed the transfection efficiency of siRNA depends on the method of siRNA association to the chitosan and entrapping siRNA using ionic gelation has shown to yield a better biological effect than simple complexation or siRNA adsorption onto the chitosan nanoparticles. Acknowledgements The authors would like to thank Ministry of Science, Technology and Environment of Malaysia for the funding of this work and also Dr. S. Somavarapu as well as Dr. Xiong Wei Li for useful discussions and technical assistance. References [1] A. Khan, M. Benboubetra, P.Z. Sayyed, K.W. Ng, S. Fox, G. Beck, I.F. Benter, S. Akhtar, Sustained polymeric delivery of gene silencing antisense ODNs, siRNA, DNAzymes and ribozymes: in vitro and in vivo studies, Journal of Drug Targeting 12 (2004) 393–404. [2] G. Beale, A.J. Hollins, M. Benboubetra, M. Sohail, S.P. Fox, I. Benter, S. Akhtar, Gene silencing nucleic acids designed by scanning arrays: antiEGFR activity of siRNA, ribozyme and DNA enzymes targeting a single hybridization-accessible region using the same delivery system, Journal of Drug Targeting 11 (2003) 449–456. [3] M. Miyagishi, H. Sumimoto, H. Miyoshi, Y. Kawakami, K. Taira, Optimization of an siRNA-expression system with an improved hairpin and its significant suppressive effects in mammalian cells, The Journal of Gene Medicine 6 (2004) 715–723. [4] M. Sioud, D.R. Sorensen, Cationic liposome-mediated delivery of siRNAs in adult mice, Biochemical and Biophysical Research Communications 312 (2003) 1220–1225. [5] D.R. Sorensen, M. Leirdal, M. Sioud, Gene silencing by systemic delivery of synthetic siRNAs in adult mice, Journal of Molecular Biology 327 (2003) 761–766.
[6] C.A. Stein, Phosphorothioate antisense oligodeoxynucleotides: questions of specificity, Trends in Biotechnology 14 (1996) 147–149. [7] B. Urban-Klein, S. Werth, S. Abuharbeid, F. Czubayko, A. Aigner, RNAimediated gene-targeting through systemic application of polyethylenimine (PEI)-complexed siRNA in vivo, Gene Therapy 1–6 (2004). [8] J. Weyermann, D. Lochmann, C. Georgens, A. Zimmer, Albumin– protamine–oligonucleotide–nanoparticles as a new antisense delivery system: Part 2. Cellular uptake and effect, European Journal of Pharmaceutics and Biopharmaceutics 59 (2005) 431–438. [9] T.H. Kim, S.I. Kim, T. Akaike, C.S. Cho, Synergistic effect of poly (ethylenimine) on the transfection efficiency of galactosylated chitosan/ DNA complexes, Journal of Controlled Release 105 (2005) 354–366. [10] M.K. Lee, S.K. Chun, W.J. Choi, J.K. Kim, S.H. Choi, A. Kim, K. Oungbho, J.S. Park, W.S. Ahn, C.K. Kim, The use of chitosan as a condensing agent to enhance emulsion-mediated gene transfer, Biomaterials 26 (2005) 2147–2156. [11] X.Z. Shu, K.J. Zhu, The influence of multivalent phosphate structure on the properties of ionically cross-linked chitosan films for controlled drug release, European Journal of Pharmaceutics and Biopharmaceutics 54 (2002) 235–243. [12] S. Gao, J. Chen, L. Dong, Z. Ding, Y.h. Yang, J. Zhang, Targeting delivery of oligonucleotide and plasmid DNA to hepatocyte via galactosylated chitosan vector, European Journal of Pharmaceutics and Biopharmaceutics 60 (2005) 327–334. [13] T. Lopez-Leon, E.L.S. Carvalho, B. Seijo, J.L. Ortega-Vinuesa, D. BastosGonzalez, Physicochemical characterization of chitosan nanoparticles: electrokinetic and stability behavior, Journal of Colloid and Interface Science 283 (2005) 344–351. [14] Y. Kakizawa, S. Furukawa, K. Kataoka, Block copolymer-coated calcium phosphate nanoparticles sensing intracellular environment for oligodeoxynucleotide and siRNA delivery, Journal of Controlled Release 97 (2004) 345–356. [15] X.W. Li, D.K.L. Lee, A.S.C. Chan, H.O. Alpar, Sustained expression in mammalian cells with DNA complexed with chitosan nanoparticles, Biochimica et Biophysica Acta (BBA) – Gene Structure and Expression 1630 (2003) 7–18. [16] R.M. Schiffelers, M.C. Woodle, P. Scaria, Pharmaceutical prospects for RNA interference, Pharmaceutical Research 21 (2004) 1–7. [17] H.J. Jeon, Y.I. Jeong, M.K. Jang, Y.H. Park, J.W. Nah, Effect of solvent on the preparation of surfactant-free poly(-lactide-co-glycolide) nanoparticles and norfloxacin release characteristics, International Journal of Pharmaceutics 207 (2000) 99–108. [18] D.A. Braasch, S. Jensen, Y. Liu, K. Kaur, K. Arar, M.A. White, D.R. Corey, RNA interference in mammalian cells by chemically-modified RNA, Biochemistry 42 (2003) 7967–7975.
225
[19] Q. Gan, T. Wang, C. Cochrane, P. McCarron, Modulation of surface charge, particle size and morphological properties of chitosan–TPP nanoparticles intended for gene delivery, Colloids and Surfaces. B, Biointerfaces 44 (2005) 65–73. [20] R. Gref, A. Domb, P. Quellec, T. Blunk, R.H. Muller, J.M. Verbavatz, R. Langer, The controlled intravenous delivery of drugs using PEG-coated sterically stabilized nanospheres, Advanced Drug Delivery Reviews 16 (1995) 215–233. [21] M. Bivas-Benita, S. Romeijn, H.E. Junginger, G. Borchard, PLGA–PEI nanoparticles for gene delivery to pulmonary epithelium, European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 1–6. [22] J. Panyam, V. Labhasetwar, Biodegradable nanoparticles for drug and gene delivery to cells and tissue, Advanced Drug Delivery Reviews 55 (2003) 329–347. [23] W. Zauner, N.A. Farrow, A.M.R. Haines, In vitro uptake of polystyrene microspheres: effect of particle size, cell line and cell density, Journal of Controlled Release 71 (2001) 39–51. [24] M. Huang, C.W. Fong, E. Khor, L.Y. Lim, Transfection efficiency of chitosan vectors: effect of polymer molecular weight and degree of deacetylation, Journal of Controlled Release 106 (2005) 391–406. [25] V. Chan, H.Q. Mao, K.W. Leong, Effect of Chitosan Molecular Weight in Gene Delivery Process, American Chemical Society, 2001 221zt-COLL260. [26] D.B. Rozema, D.L. Lewis, siRNA delivery technologies for mammalian systems, Targets 2 (2003) 253–260. [27] K.A. Janes, P. Calvo, M.J. Alonso, Polysaccharide colloidal particles as delivery systems for macromolecules, Advanced Drug Delivery Reviews 47 (2001) 83–97. [28] P. Calvo, A. Bougaba, M. Appel, E. Fattal, M.J. Alonso, P. Couvreur, Oligonucleotide–chitosan nanoparticles as new gene therapy vector, Proc. World Meet. APGI/APV, vol. 2, 1998, pp. 1111–1112. [29] M. Keller, Lipidic carriers of RNA/DNA oligonucleotides and polynucleotides: what a difference a formulation makes! Journal of Controlled Release 103 (2005) 537–540. [30] P. Tam, M. Monck, D. Lee, O. Ludkovski, E.C. Leng, K. Clow, H. Stark, P. Scherrer, R.W. Graham, P.R. Cullis, Stabilized plasmid–lipid particles for systemic gene therapy, Gene Therapy 7 (2000) 1867–1874. [31] V.A. Bloomfield, S. He, A.Z. Li, P.B. Arscott, Light scattering studies on DNA condensation, Biochemical Society Transactions 19 (1991) 496. [32] V.A. Bloomfield, Condensation of DNA by multivalent cations: considerations on mechanism, Biopolymers 31 (1991) 1471–1481. [33] V.A. Bloomfield, DNA condensation, Current Opinion in Structural Biology 6 (1996) 334–341.
GENE DELIVERY
H. Katas, H.O. Alpar / Journal of Controlled Release 115 (2006) 216–225