A biomimetic nanovector-mediated targeted cholesterol-conjugated siRNA delivery for tumor gene therapy

Biomaterials 33 (2012) 8893e8905

Contents lists available at SciVerse ScienceDirect

Biomaterials journal homepage: www.elsevier.com/locate/biomaterials

A biomimetic nanovector-mediated targeted cholesterol-conjugated siRNA delivery for tumor gene therapy Yang Ding a,1, Wei Wang a, *,1, Meiqing Feng b, Yu Wang c, Jianping Zhou a, *, Xuefang Ding a, Xin Zhou a, Congyan Liu a, Ruoning Wang a, Qiang Zhang d a

State Key Laboratory of Natural Medicines, Department of Pharmaceutics, China Pharmaceutical University, 24 Tongjiaxiang, Nanjing 210009, China School of Pharmacy, Fu Dan University, 826 Zhangheng Road, Shanghai 201203, China Department of Pharmacology, Nanjing Medical University, 140 Hanzhong Road, Nanjing 210029, China d State Key Laboratory of Natural and Biomimetic Drugs, School of Pharmaceutical Sciences, Peking University, 38 Xueyuan Road, Beijing 100191, China b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 9 August 2012 Accepted 23 August 2012 Available online 12 September 2012

RNA interference holds tremendous potential as a therapeutic approach of malignant tumors. However, safe and efficient nanovectors are extremely lack for systemic delivery of small interfering RNA (siRNA). The study aimed to develop a biomimetic nanovector, reconstituted high density lipoprotein (rHDL), mediating targeted cholesterol-conjugated siRNA (Chol-siRNA) delivery for Pokemon gene silencing therapy. Chol-siRNA-loaded rHDL nanoparticles (rHDL/Chol-siRNA complexes) were prepared using thinfilm dispersion method and their characteristics were investigated in detail. RHDL/Chol-siRNA complexes at the optimal volume ratio (lipid: Chol-siRNA) exhibited high Chol-siRNA-loading efficiency (w99%), desirable nanoparticle size and excellent stability in serum. In addition, by analyzing Chol-siRNA release profile, rHDL/Chol-siRNA complexes displayed sustained-release characteristic and storage stability. Observations from FACS and confocal microscopic analyses revealed that rHDL-mediated carboxyfluorescein tagged Chol-siRNA (FAM-Chol-siRNA) transfection resulted in highly efficient uptake and specific cytoplasmic delivery of FAM-Chol-siRNA into human hepatocellular carcinoma cell line HepG2 via HDL-receptor mediated mechanism. In vitro cytotoxicity, apoptosis and Western-blot analyses revealed significant cellular growth inhibition and decrease of Pokemon and Bcl-2 protein expression in HepG2 cells treated with Chol-siRNA-Pokemon-loaded rHDL nanoparticles (rHDL/Chol-siRNA-Pokemon complexes), respectively. In in vivo studies, the near-infrared (NIR) dye Cy5 labeled Chol-siRNA-loaded rHDL nanoparticles (rHDL/Cy5-Chol-siRNA complexes) obviously accumulated in tumor of nude mice after i.v. administration as compared with Cy5-Chol-siRNA-loaded lipoplexes (Lipos/Cy5-Chol-siRNA complexes). Morover, rHDL/Chol-siRNA-Pokemon complexes demonstrated great tumor growth inhibition and significant decrease of Pokemon and Bcl-2 protein expression in vivo. These results suggested that rHDL should be an ideal non-viral tumor-targeting vector for Chol-siRNA transfer, and rHDLmediated Chol-siRNA-Pokemon delivery might be a promising new strategy for gene therapy in hepatocellular carcinoma. Ó 2012 Elsevier Ltd. All rights reserved.

Keywords: Reconstituted high density lipoprotein Apolipoprotein A-I Targeted cholesterol-conjugated siRNA delivery Pokemon Tumor gene therapy

1. Introduction RNA interference (RNAi) represents a promising gene silencing technology for functional genomics and a potential gene therapeutic strategy for a variety of genetic diseases [1,2]. Since the discovery about the RNAi mechanism of gene-specific silencing, synthetic small interfering RNA (siRNA) has been applied widely in * Corresponding authors. Tel./fax: þ86 25 83271272. E-mail addresses: [email protected] (W. yahoo.com.cn (J. Zhou). 1 These authors contributed equally to this work.

Wang),

zhoujpcpu@

0142-9612/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.biomaterials.2012.08.057

the research field of gene therapy [3,4], as it selectively and persistently suppresses the expression of pathogenic proteins through the sequence-specific degradation of mRNA, with almost no interferon response [5]. However, due to the relatively large molecular weight, polyanionic nature and susceptibility to enzymatic degradation of siRNA duplexes, the effective cellular uptake and intracellular delivery of siRNA remains a major challenge for widespread use of RNAi as a therapeutic modality or even as an investigational tool in vivo [6,7]. A critical requirement for achieving systemic RNAi in vivo is the introduction of “drug-like” properties, such as stability, cellular delivery and tissue bioavailability, into synthetic siRNA. In

8894

Y. Ding et al. / Biomaterials 33 (2012) 8893e8905

exploring the potential of synthetic siRNAs to silence endogenous target genes, chemically modified siRNAs, including cholesterolconjugated, 20 -O-methyl sugar modified and antibody-linked siRNAs, have been found to markedly improve pharmacological activities in vitro and in vivo [8e10]. Covalent conjugation of cholesterol to siRNA increases the nucleaseeresistant activity of free siRNA, facilitates cellular import and elicits RNAi, which results in silencing of endogenous genes in vivo [9,11]. New formulation strategies have been reported to form siRNA complexes with several cationic vectors such as polymers, liposomes or peptides [6,12,13], but the safety profiles of them remain an obstacle for clinical application. To overcome the cytotoxicity of siRNA vectors caused by excess of positive charge, non-specific interaction and aggregation with serum proteins in the blood, more specific methods are needed for systemic delivery of siRNA. The most challenging issue in successful application of the RNAi strategy for gene therapy to human neoplasms is the choice of an effective and non-cytotoxic vector for delivering siRNA into tumor cells. High density lipoprotein (HDL) is one of the essential components of the lipid transport system, and plays a critical role in the reverse cholesterol transport from peripheral tissues to liver [14]. Mature HDL consists of apolipoproteins, phospholipid monolayer, cholesterol and hydrophobic core, into which highly hydrophobic drugs can be incorporated [15]. The major protein component (w70%) in HDL is apolipoprotein A-I (apoA-I), a highly a-helical polypeptide (28 kDa), which promotes cholesterol efflux to cells and is important in maintaining cellular cholesterol homeostasis following interaction of HDL with scavenger receptor class B type I (SR-BI) that is primarily expressed in the liver [16,17] and most malignant cells [18]. Endogenous HDL particles are of complete biodegradation and non-immunogenicity. Moreover, HDL particles could escape from the reticuloendothelial system thereby exhibiting longer residence time in the circulation [19] than other lipoproteins or most drug formulations. The core contents of the circulating HDL particles are selectively taken up into cells through a SR-BI receptor-mediated mechanism [18]. On the basis of attractive features mentioned above, HDL could be an ideal non-viral siRNA vector, which has been confirmed in Christian et al.’s report [5]. Reconstituted HDL (rHDL) is the synthetic form of the circulating human HDL, and they have similar physicochemical properties. The composition of rHDL includes phospholipids (PC), apoAI, cholesterol and cholesteryl esters [20]. In the past decades, rHDL has been successfully developed as a drug carrier. RHDL/aclacinomycin (ACM) complexes prepared by sonication of PC, ACM and apoA-I from human serum, kept the basic physical and biological binding properties of native HDL and showed a preferential cytotoxicity for human hepatoma SMMC-7721 cells to normal L02 hepatocytes [21]. RHDL/paclitaxel (PTX) complexes prepared with substantially higher PTX content than that in earlier reports, had superior cytotoxicity against several cancer cell lines that overexpressed SR-BI receptors and the half maximal inhibitory concentration (IC50) had been found to be 5e20 times lower than that of the free drug [22]. Therefore, rHDL was considered attractive as a drug carrier for selective delivery of therapeutic payloads. Another important consideration about the RNAi strategy for tumor treatment involves the gene, itself, targeted for silencing. Thus far, many gene candidates have been studied. Pokemon, a proto-oncogene, has been recently identified to have crucial but versatile functions in embryonic development, cell differentiation, proliferation and tumorigenesis [23,24]. Moreover, aberrant overexpression of Pokemon has been observed in many human cancers [25e29], and its expression levels are effective for the prediction of tumorigenic behavior and clinical outcome [23,24,29]. The underlying molecular mechanism of carcinogenesis mediated

by Pokemon involves the specific repression of two key tumor suppressive pathways, the alterative reading frame (ARF)-multiple murine double minute gene 2 (HDM2)-p53 pathway [30] and the retinoblastoma (Rb)-early-region-2 transcription factor (E2F) pathway [28]. These findings suggest that Pokemon plays a critical role in cellular transformation and that it may be a potent target of the RNAi strategy for cancer treatment. In this study, rHDL-based system (Scheme 1) was introduced for targeted Chol-siRNA delivery and gene therapy in the cancer cells that over-expressed SR-BI receptors. The complex-formation and Chol-siRNA-loading efficiency studies were performed to optimize the volume ratio (lipid: Chol-siRNA) in the formulation of rHDL/ Chol-siRNA complexes. The incubation of Chol-siRNA-loaded lipoplexes (Lipos/Chol-siRNA complexes) with apoA-I was evaluated by the observations based on dynamic light scattering (DLS) measurement and atomic force microscopy (AFM). The serum stability and Chol-siRNA release behavior of rHDL/Chol-siRNA complexes were investigated in detail. FAM-Chol-siRNA was complexed with rHDL to explore the possibility of transfection in human hepatocellular carcinoma HepG2 cells in vitro. The effects of gene therapy were evaluated by cytotoxicity, apoptosis and Western-blot analyses after transfection of rHDL/Chol-siRNAPokemon complexes in HepG2 cells in vitro. In vivo biodistribution and targeting efficiency of Cy5 labeled Chol-siRNAloaded rHDL nanoparticles (rHDL/Cy5-Chol-siRNA complexes) in tumor-bearing nude mice was investigated using a non-invasive NIR optical imaging technique. Finally, the tumor growth inhibition effect and protein expression levels in vivo were evaluated after i.v. administration of rHDL/Chol-siRNA-Pokemon complexes in a xenograft nude mouse model. 2. Materials and methods 2.1. Chemicals Unmodified siRNA (Nontargeted control siRNA, 50 -UUC UCC GAA CGU GUC ACG UTT-30 ), Chol-siRNA (Nontargeted control cholesterol-conjugated siRNA, 50 -Chol-UUC UCC GAA CGU GUC ACG UTT-30 ), Chol-siRNA-Pokemon (Chol-siRNA targeting the Pokemon gene, 50 -Chol-GCA CTT TAA GGA CGA GGA CTT-30 ), and FAM or Cy5 labeled Chol-siRNA were synthesized and purified with HPLC by Shanghai GenePharma Co., Ltd. (Shanghai, China). Apolipoprotein A-I (apoA-I) was a generous gift from Dr. Meiqing Feng (Pharmacy of Fu Dan University). Soybean phospholipids (PC, purity 90%) were obtained from Evonik Degussa China Co., Ltd. (Shanghai, China). Cholesterol and cholesteryl esters were purchased from Pharmacia Biotech (NJ, USA). DEPCtreated water was obtained from Sunshine Biotechnology Co., Ltd. (Nanjing, China). LipofectamineÔ 2000 was purchased from Invitrogen Corporation (CA, USA). Hoechst 33258 and branched PEI (25 kDa) were purchased from SigmaeAldrich (MO, USA).

Scheme 1. Schematic cross-section of rHDL/Chol-siRNA complexes. rHDL/Chol-siRNA complexes consist of an apolar core containing cholesteryl esters, that is surrounded by a shell composed of soybean phospholipids, cholesterol, apoA-I and Chol-siRNA. The chemical modification of siRNA with a lipophilic anchor yields the amphiphilic CholsiRNA that can be readily incorporated into rHDL.

Y. Ding et al. / Biomaterials 33 (2012) 8893e8905 3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyl tetrazolium bromide (MTT) was purchased from Fluka (MO, USA). Annexin V-FITC apoptosis detection kit was purchased from BioVision (CA, USA). Primary antibodies were obtained from Santa Cruz Biotechnology Inc. (CA, USA), and IRDyeÔ 800-conjugated second antibody was obtained from Rockland Inc. (PA, USA). All other chemicals were from various suppliers and were reagent grade or better. 2.2. Cell culture Human hepatocellular carcinoma cell line HepG2 was purchased from Cell Bank of Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences (Shanghai, China). Cells were grown in 90% DMEM (Gibco, USA) supplemented with 10% FBS (Sijiqing, Hangzhou, China), 100 U/mL penicillin, and 100 mg/mL streptomycin. Exponentially growing cultures were maintained in a humidified atmosphere of 5% CO2 at 37  C. 2.3. Preparation of Lipos/Chol-siRNA and rHDL/Chol-siRNA complexes The preparation process consisted of the construction of Lipos/Chol-siRNA complexes and subsequent formation of rHDL/Chol-siRNA complexes by incubation of Lipos/Chol-siRNA complexes with apoA-I. Thin-film dispersion method [31] was alternatively employed to prepare Lipos/Chol-siRNA complexes with soybean phospholipids (PC), cholesterol (Chol) and cholesteryl esters (CE). Briefly, 30 mg of PC, 3 mg of Chol and 6 mg of CE dissolved in 500 mL of organic solvent (chloroform: methanol ¼ 1: 1, v/v), and the solvent of lipid solutions (20 mL, 40 mL, 60 mL, 80 mL, 100 mL and 120 mL) was evaporated with a rotary evaporator at 30  C until a thin film was formed. The trace solvent residue was finally removed with a stream of nitrogen gas. 4 mL of Chol-siRNA solution (1 O.D. Chol-siRNA dissolved in 150 mL of DEPC-treated water), appropriate volume of sodium cholate solution (30 mg/mL in PBS buffer) and Tris buffer (0.1 M KCl, 10 mM Tris, 1 mM EDTA, pH 8.0) were added to dissolve the thin film, then the mixture was sonicated using a probe-type ultrasonicator (JY 92-2D, Ningbo Scientz Biotechnology Co., Ltd, Nanjing, China) at 100 W in an ice bath until a clear suspension was obtained. The dispersion was then filtered through a 0.22 mm pore-sized microporous membrane and the filtrate was dialyzed to remove the free sodium cholate using dialysis bags (MW cut off 14, 000 Da). Finally, the prepared Lipos/Chol-siRNA complexes were collected and stored at 4  C until further use. Lipos/Chol-siRNA complexes were incubated with 16 mL of apoA-I solution (30 mg/mL in PBS buffer) to form rHDL/Chol-siRNA complexes under 600 rpm stirring at 25  C for 8 h. The prepared rHDL/Chol-siRNA complexes were obtained and stored at 4  C to ensure their intact structures until further use. Complex formation was confirmed by electrophoresis on a 2.0% agarose gel at 100 mV for 30 min. Then, Chol-siRNA was visualized by 5 mL/100 mL Goldview (Amresco, USA) staining using Gel-Pro analyzer (Genegenius, Syngene, UK). 2.4. Determination of Chol-siRNA-loading efficiency The loading efficiency of Chol-siRNA in rHDL at various volume ratios (lipid: Chol-siRNA) was determined from free Chol-siRNA concentration in the supernatant recovered after complexes centrifugation (20, 000 rpm, 15 min, 4  C) by absorbance measurement at 260 nm. For comparison, rHDL/unmodified siRNA complexes were prepared according to Section 2.3. The supernatant recovered from unloaded rHDL (without Chol-siRNA or unmodified siRNA) was used as a blank. The loading efficiency (LE) was calculated from the total concentration of the added amount of CholsiRNA (or unmodified siRNA) present in the system ([siRNA]T) and the concentration of that in the supernatant ([siRNA]S) using the following equation:

 LE % ¼ ½siRNAT  ½siRNAS ½siRNAT  100

2.5. Measurement of particle size, zeta potential and morphology RHDL/Chol-siRNA complexes used in this and the following experiments were prepared at the optimal volume ratio (lipid: Chol-siRNA) of 20: 1. The particle size and zeta potential of the complexes were measured by dynamic light scattering (DLS) using a Malvern Zetasizer (Nano-ZS90, Malvern instruments, UK). All of the DLS measurements were performed at 25  C and at a scattering angle 90 . The size distribution and morphology of complexes were observed using atomic force microscopy (AFM, Veeco diNanoScope V, USA). An aqueous solution of complexes was placed on a clean mica surface, washed with distilled water and purged with nitrogen. Subsequently, the surface containing complexes was imaged by AFM in a scanned area of 2.0  2.0 mm, and the particle size was obtained using Nano Rule software. The height differences on the surface were indicated by the color code (lighter regions indicated higher heights). 2.6. Serum stability assay Serum stability assay was routinely performed to evaluate the effect of rHDL on protecting Chol-siRNA from serum degradation [32]. The free Chol-siRNA (0.5 mg)

8895

and rHDL/Chol-siRNA (0.5 mg) complexes were separately mixed with FBS (Sijiqing, Hangzhou, China) in a volume ratio of 1: 1 and incubated at 37  C for the indicated durations, then the results were analyzed by 2.0% agarose gel eletrophoresis. 2.7. Release of Chol-siRNA in vitro RHDL/Chol-siRNA complexes (100 mL) were prepared at the optimal volume ratio of 20:1 (lipid: Chol-siRNA) in 0.1 M PBS buffer (pH 7.4), and gently shaken at 37  C in the water bath at 100 rpm. For comparison, Chol-siRNA plus rHDL solution was a mixture of the same amount of Chol-siRNA and rHDL with those of rHDL/CholsiRNA complexes. The supernatant and complexes were periodically collected by centrifugation at 20,000 rpm for 15 min. The amount of residual Chol-siRNA in complexes was determined using the same analytical method of Chol-siRNA-loading efficiency. The amount of released siRNA was also determined by absorbance measurement at 260 nm. The remaining Chol-siRNA (%) in complexes was calculated as a ratio (%) of the remaining, to the total amount of Chol-siRNA. The stability of rHDL/Chol-siRNA complexes during the release test was evaluated as follows. The collected complexes resolved in 100 mL Tris buffer (0.1 M, pH 8.0), and then 20 mL of complexes was mixed with 4 mL of loading buffer. Samples were analyzed by 2.0% agarose gel electrophoresis (100 mV, 30 min). 2.8. Chol-siRNA cellular uptake by flow cytometry HepG2 cells were seeded into 6-well plates. After incubation overnight (to reach 80e90% confluence at the time of transfection), the culture media were replaced, respectively, with the media containing carboxyfluorescein tagged Chol-siRNA (FAM-Chol-siRNA, 100 nM) alone, Lipos/FAM-Chol-siRNA (100 nM) and rHDL/ FAM-Chol-siRNA (100 nM) complexes. Transfection with LipofectamineÔ 2000/ FAM-Chol-siRNA (100 nM) complexes was performed as a positive control according to the manufacture’s protocol, and untreated cells were used as a control. After incubation for 6 h at 37  C, cells were washed with PBS for three times, harvested by trypsinization and collected in PBS to measure the fluorescence intensity. The fluorescence intensity at 490/530 nm of FAM per cell was determined by flow cytometer (BD FACS Calibur, USA). 2.9. Chol-siRNA cellular distribution by confocal microscopy Intracellular location of Chol-siRNA was investigated using confocal laser scanning microscopy (CLSM). HepG2 cells were cultured in 35 mm glass bottom culture dishes for 24 h. The media were then replaced with 1 mL of the media containing FAM-Chol-siRNA (100 nM) alone and FAM-Chol-siRNA (100 nM)-loaded complexes. After incubation at 37  C for 6 h, the culture media were subsequently removed and the cells were rinsed with PBS for three times to remove complexes which were not ingested by the cells. Then cells were fixed with 4% paraformaldehyde, and Hoechst 33258 (final working concentration 10 mg/mL) was added to stain the cell nuclei for 10 min before the fluorescent imaging. Intracellular location of Chol-siRNA was observed using a confocal microscope (Leica TCS SP5, Heidelberg, Germany) excited at 346 nm and 494 nm, and emitted at 460 nm and 522 nm for Hoechst 33258 and FAM, respectively. The images were analyzed using Leica CLSM software. 2.10. MTT assay In vitro cytotoxicity of rHDL and rHDL/Chol-siRNA-Pokemon complexes was evaluated with HepG2 cells by 3-(4,5-dimethylthiazol-2-yl)-2, 5-diphenyl tetrazolium bromide (MTT) assay. HepG2 cells at the density of 5  103 cells/well were seeded into 96-well plates. After 24 h, the culture media were replaced, respectively, with 200 mL of the media containing PEI (0.05e1000 mg/mL), rHDL (0.05e1000 mg/ mL), rHDL/Chol-siRNA (100 nM) complexes, Chol-siRNA-Pokemon (100 nM) alone and rHDL/Chol-siRNA-Pokemon (100 nM) complexes and additionally incubated for 6 h. Then the media were exchanged for fresh media, and allowed to continue incubating. After 24 h, 48 h and 72 h incubation, 5 mg/mL MTT solution (20 mL/well) was added and cultured for 4 h, then the supernatant was discarded and DMSO was added (100 mL/well), respectively. Untreated cells were taken as a control with 100% viability and cells without addition of MTT were used as blank to calibrate the spectrophotometer to zero absorbance. The suspension was placed on microvibrator for 5 min and the optical density (O.D.) was measured at 570 nm by the Universal Microplate Reader (EL800, BIO-TEK Instruments Inc., USA). 2.11. Cell morphological assessment and apoptosis study HepG2 cells were seeded in a 6-well plate at a density of 2  105 cells per well and grown for 24 h. The culture media were removed and the cells were treated with Chol-siRNA-Pokemon (100 nM) alone, rHDL/Chol-siRNA-Pokemon (100 nM) complexes or rHDL with equivalent concentration of lipids to rHDL/Chol-siRNAPokemon complexes, respectively, and untreated cells were used as a control. After 48 h post-transfection, the morphology of cells was monitored under an inverted light microscope (OlympusIX51, Japan). Then all attached and floating cells were collected and washed twice with PBS. An annexin V-FITC apoptosis detection

8896

Y. Ding et al. / Biomaterials 33 (2012) 8893e8905

kit was used to analyze cell apoptosis. The cells were resuspended in binding buffer (500 mL) and then stained with fluorescein isothiocyanate (FITC)-conjugated annexin V (5 mL) and propidium iodide (PI, 10 mL) according to the manufacturer’s instructions. After double supravital staining, the samples were immediately analyzed by flow cytometry mentioned above (excitation 488 nm for FITC and 540 nm for PI). Cells in the early apoptotic process should be stained by the annexin V-FITC alone, necrotic cells should be stained by both annexin V-FITC and PI, and live cells should not be stained by either annexin V-FITC or PI. 2.12. Western-blot analysis HepG2 cells were seeded in a 6-well plate (2  105 cells per well) and grown for 24 h. The cells were then transfected with Chol-siRNA-Pokemon (100 nM) alone or rHDL/Chol-siRNA-Pokemon (100 nM) complexes for 6 h, respectively, and then incubated with fresh culture media for 48 h. Proteins of cells were isolated by lysis buffer (100 mM Tris-Cl, pH 6.8, 4% (m/v) SDS, 20% (v/v) glycerol, 200 mM b-mercaptoethanol, 1 mM PMSF, and 1 g/mL aprotinin) and measured using the BCA protein assay method with Varioskan spectrofluorometer and spectrophotometer (Thermo) at 562 nm. Untreated cells were used as a control. Protein samples were separated with 15% SDS-polyacrylamide gel (SDS-PAGE) and transferred onto the PVDF membranes (Millipore). Immune complexes were formed by incubation of the proteins with primary antibodies (rabbit anti-Pokemon and rabbit anti-Bcl-2) overnight at 4  C. Blots were washed and incubated for 1 h with IRDyeÔ 800 conjugated anti-rabbit second antibody. Immunoreactive protein bands were detected with an Odyssey Scanning System (LI-COR inc., USA). 2.13. In vivo anti-tumor studies of rHDL/Chol-siRNA-Pokemon complexes 2.13.1. Animal BALB/c nude mice (six-week-old, 18e22 g) were obtained from Shanghai Laboratory Animal Center (SLAC, China) and maintained at 22  2  C on a 12 h lighte

Fig. 1. The determination of the optimal volume ratio (lipid: Chol-siRNA) to prepare rHDL/Chol-siRNA complexes. (A) Formation analysis of rHDL/Chol-siRNA complexes at various volume ratios (lipid: Chol-siRNA) by agarose gel electrophoresis. Lane 1: Naked Chol-siRNA (0.5 mg); Lane 2-7: Chol-siRNA (0.5 mg) with progressively increasing volumes of lipid solution. (B) The loading efficiency for Chol-siRNA or unmodified siRNA into rHDL at various volume ratios (lipid: Chol-siRNA or lipid: unmodified siRNA) ranging from 5 to 30. The error bars in the graph represent standard deviations (n ¼ 3).

dark cycle with access to food and water ad libitum. All care and handling of animals were approved by the University Ethics Committee for the use of experimental animals and carried out in strict accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals. 2.13.2. In vivo imaging analysis Approximately 1  106 HepG2 cells were inoculated subcutaneously in the armpit region of BALB/c nude mice. After 21st days of tumor inoculation, the NIR dye Cy5 labeled Chol-siRNA-loaded Lipos and rHDL nanoparticles (Lipos/Cy5-CholsiRNA complexes and rHDL/Cy5-Chol-siRNA complexes) were injected into the tail vein of tumor-bearing nude mice at a dose of 50 mg of Cy5-Chol-siRNA per mouse to investigate their biodistribution and tumor-targeting efficacy in the living animals (n ¼ 6 for each group). The NIR fluorescence imaging experiments were performed at 6 and 24 h post-injection using an in vivo imaging system (FX PRO, Kodak, USA) equipped with an excitation bandpass filter at 720 nm and an emission at 790 nm. Images were analyzed using the Kodak Molecular Imaging Software 5.X. After living imaging, mice were sacrificed and fresh tumor tissues were excised for ex vivo imaging using the same imaging system. 2.13.3. Treatment of tumor xenografts in nude mice Freshly harvested HepG2 cells were rinsed and resuspended in PBS at 1  107 cells/mL. BALB/c nude mice were injected with HepG2 cells (106 cells) subcutaneously in the armpit region. When the tumor size reached approximately 50 mm3, the animals were randomly divided into three groups with six mice per group, and treated with normal saline (control), Chol-siRNA-Pokemon or rHDL/Chol-siRNAPokemon complexes, respectively, by i.v. administration at a dose of 25 mg of Chol-siRNA-Pokemon per mouse. The initial day of administration was defined as day 0. Administration was then repeated seven times at 2-day interval such that day 14 was the final day of administration. Tumor sizes were measured every other day and tumor volumes were calculated by using the formula V ¼ a  b2/2, where a is the largest diameter and b is the perpendicular diameter and V is given in mm3. Furthermore, Pokemon and Bcl-2 protein expression in vivo was measured by Western-blot analysis carried out as described in Section 2.12. Briefly, at day 14 after

Fig. 2. Particle size and its changes of rHDL/Chol-siRNA complexes at the optimal volume ratio (lipid: Chol-siRNA) of 20:1 measured in Tris buffer (pH 8.0) by DLS. (A) Particle size distribution profiles of representative Lipos/Chol-siRNA and rHDL/CholsiRNA complexes. (B) Particle size changes of rHDL/Chol-siRNA complexes for 24 h. The error bars in the graph represent standard deviations (n ¼ 3).

Y. Ding et al. / Biomaterials 33 (2012) 8893e8905

8897

Fig. 3. AFM images of Lipos/Chol-siRNA and rHDL/Chol-siRNA complexes at the optimal volume ratio (lipid: Chol-siRNA) of 20:1. (A) AFM 3D image of rHDL/Chol-siRNA complexes. (B) AFM height images of Lipos/Chol-siRNA (a) and rHDL/Chol-siRNA complexes (b).

8898

Y. Ding et al. / Biomaterials 33 (2012) 8893e8905

i.v. administration of Chol-siRNA-Pokemon alone or rHDL/Chol-siRNA-Pokemon complexes, mice were sacrificed and fresh tumor tissues were collected. Proteins of tissue were isolated by lysis buffer and measured using the BCA protein assay method with Varioskan spectrofluorometer and spectrophotometer (Thermo) at 562 nm. Normal saline group were used as a control. Protein samples were separated with 15% SDS-PAGE and transferred onto the PVDF membranes (Millipore). Immune complexes were formed by incubation of the proteins with primary antibodies (rabbit anti-Pokemon and rabbit anti-Bcl-2) and subsequently with IRDyeÔ 800 conjugated anti-rabbit second antibody. Immunoreactive protein bands were detected with an Odyssey Scanning System (LI-COR inc., USA). 2.14. Statistical analysis Data were expressed as means  S.D. from triplicate experiments performed in a parallel manner unless otherwise indicated. The results were analyzed using an unpaired, two-tailed Student’s t-test. All comparisons were made relative to corresponding controls and significance of difference was indicated as *P < 0.05, **P < 0.01 or ***P < 0.001.

3. Results and discussion 3.1. Formation and characterization of rHDL/Chol-siRNA complexes The main obstacle to achieving in vivo gene silencing by RNAi technologies is safe delivery [1,33]. To date, multiple delivery systems, e.g., cationic lipids [12], cationic polymers [34] and cell penetrating peptides [13] were investigated in vitro and in vivo. However, the positively charged surface characteristics and nonspecificity of cationic gene vectors would cause severe cytotoxicity, wide body distribution and aggregation with serum proteins in the body. To alleviate these problems, selective and noncytotoxic delivery systems are needed to be exploited. Herein, we prepared rHDL, a biomimetic nanovector, for targeted Chol-siRNA delivery. The rHDL was generated by the interaction of phospholipids and apoA-I, an amphipathic a-helix structure, and derivatization of siRNA with a lipophilic anchor (Chol) yielded the amphiphilic Chol-siRNA that could be readily incorporated into rHDL. Fig. 1A showed the influence of volume ratios (lipid: CholsiRNA) on complex formation from the gel retardation assay. The results indicated that the electrophoretic mobility of Chol-siRNA was retarded with the increasing volume of lipid solution and Chol-siRNA even remained at the top of the gel at volume ratios of 20e30, suggesting that rHDL formed complexes with Chol-siRNA successfully. The intensity of free Chol-siRNA band was weaker than that of the bands of rHDL/Chol-siRNA complexes at the same amount of Chol-siRNA, indicating that rHDL/Chol-siRNA complexes showed significantly enhanced resistance toward nuclease degradation occurred in the experiment. As expected, rHDL was capable of effectively encapsulating Chol-siRNA as a non-viral vector.

Lipos/Chol-siRNA complexes (data not shown). Meanwhile, taking the results from complex-formation analysis into consideration, we chose the optimal volume ratio (lipid: Chol-siRNA) of 20: 1 to prepare rHDL/Chol-siRNA complexes in the following experiments. 3.3. Particle size, zeta potential and morphology of complexes DLS measurement revealed that the particle size of Lipos/CholsiRNA complexes was w60 nm (polydispersity index ¼ 0.115) and that of rHDL/Chol-siRNA complexes was w90 nm (polydispersity index ¼ 0.145) (Fig. 2A). This might confirm the formation of rHDL/ Chol-siRNA complexes composed of a shell of amphiphilic phospholipid monolayer with apoA-I and a core of hydrophobic lipids. Also, the particle size of rHDL/Chol-siRNA complexes was kept w90 nm during 24 h (Fig. 2B), suggesting that the complexes were of great stability in the buffer and could be taken up into tumor cells. According to the previous report, it was possible for these small particles to exhibit accumulation in tumors via the “leaky” vasculature through the enhanced permeation and retention (EPR) effect [35]. Furthermore, rHDL/Chol-siRNA complexes possessed a neutral charge (zeta potential, -4.2 mV).

3.2. Determination of Chol-siRNA-loading efficiency in rHDL The amount of Chol-siRNA (or unmodified siRNA) loaded into rHDL was determined by spectrophotometrically measuring the optical density (O.D.) of the supernatant obtained after centrifugation of rHDL/Chol-siRNA (or unmodified siRNA) complexes. The amount of Chol-siRNA (or unmodified siRNA) in the supernatant was calculated taking 1 O.D. at 260 nm equal to 50 mg of siRNA. Fig. 1B showed the loading efficiency of different formulations at various volume ratios (lipid: Chol-siRNA or lipid: unmodified siRNA). Compared with unmodified siRNA group, rHDL showed significantly higher Chol-siRNA-loading efficiency. Therefore, it was concluded that the covalent conjugation of Chol to siRNA would help the formation of rHDL/Chol-siRNA complexes. Moreover, when the volume ratio (lipid: Chol-siRNA) was beyond 20, the loading efficiency in rHDL/Chol-siRNA complexes was found to be w99%, and the same loading efficiency (w99%) was achieved in

Fig. 4. The serum stability and Chol-siRNA release of rHDL/Chol-siRNA complexes at the optimal volume ratio (lipid: Chol-siRNA) of 20:1. (A) Serum stability assay of naked Chol-siRNA and rHDL/Chol-siRNA complexes. Free Chol-siRNA (0.5 mg) and rHDL/CholsiRNA (0.5 mg) complexes were separately incubated in 50% serum-containing media at 37  C for the indicated durations and degradation of Chol-siRNA was investigated by 2.0% agarose gel electrophoresis. (B) The release profile of Chol-siRNA from rHDL/CholsiRNA complexes in vitro. RHDL/Chol-siRNA complexes (100 mL) were prepared in 0.1 M PBS buffer (pH 7.4), and gently shaken at 37  C for 7 days in the water bath at 100 rpm. The Chol-siRNA plus rHDL solution was set as a control. The error bars in the graph represent standard deviations (n ¼ 3). (C) The stability of rHDL/Chol-siRNA complexes during the release test was determined by 2.0% agarose gel electrophoresis.

Y. Ding et al. / Biomaterials 33 (2012) 8893e8905

Fig. 3 showed AFM images of Lipos/Chol-siRNA and rHDL/CholsiRNA complexes. Both of them were almost spherical in shape, and the particle size of rHDL/Chol-siRNA complexes was larger than Lipos/Chol-siRNA complexes (Fig. 3B). The results demonstrated that rHDL/Chol-siRNA complexes might be a Lipos/protein structure with apoA-I inserted in the surface of Lipos/Chol-siRNA complexes. Noticeably, the particle size measured from AFM was larger than that from DLS, which might be attributed to lipid

8899

spreading or flattening of complexes onto the mica surface and fusion process during sample preparation [36]. The height of Lipos/ Chol-siRNA complexes measured from AFM images was lower than that of rHDL/Chol-siRNA complexes (18 nm versus 23 nm), and the X/Z curve of Lipos/Chol-siRNA complexes appeared smooth, but that of rHDL/Chol-siRNA complexes was crude, which indicated that rHDL/Chol-siRNA complexes were well-formed by incubation of Lipos/Chol-siRNA complexes with apoA-I.

Fig. 5. Cellular uptake and intracellular location of FAM-Chol-siRNA from different formulations by HepG2 cells. FACS analysis for cellular uptake of FAM-Chol-siRNA (A) and the percentage of siRNA positive cells (B) after treatment with FAM-Chol-siRNA alone, Lipos/FAM-Chol-siRNA, rHDL/FAM-Chol-siRNA or LipofectamineÔ 2000/FAM-Chol-siRNA complexes for 6 h. Results were expressed as means  S.D. (n ¼ 3). ***P < 0.001, relative to Lipos/FAM-Chol-siRNA group. (C) Confocal microscopic observations for intracellular location of FAM-Chol-siRNA in HepG2 cells. For each panel, images from left to right showed the green fluorescence of FAM-Chol-siRNA in cytoplasm (or nucleus), the blue fluorescence of Hoechst 33258 in nucleus and the merged fluorescence of FAM-Chol-siRNA and Hoechst 33258. Scale bars correspond to 50 mm in all the images.

8900

Y. Ding et al. / Biomaterials 33 (2012) 8893e8905

3.4. Protection of Chol-siRNA from serum degradation by rHDL The role of rHDL in protecting Chol-siRNA from serum degradation was investigated. As shown in Fig. 4A, free Chol-siRNA band was faint at 3 h, and nearly disappeared after 3 h, suggesting that naked Chol-siRNA was prone to be degraded by serum. In contrast, Chol-siRNA in rHDL was protected well in serum, and remained relatively intact even at 12 h. These results demonstrated that rHDL provided a favorable protection of Chol-siRNA from serum degradation compared with naked Chol-siRNA group. Meanwhile, it was worth noting that the ability of rHDL/Chol-siRNA complexes against serum degradation was especially essential for a successful siRNA delivery in vivo [9]. 3.5. Release of Chol-siRNA from rHDL in vitro

green fluorescence of rHDL/FAM-Chol-siRNA group was exclusively observed in cytoplasm as indicated by the red arrows after 6 h transfection. FAM-Chol-siRNA alone group gave almost no detectable fluorescence in cells, and the fluorescence was slightly detectable in cells treated with Lipos/FAM-Chol-siRNA complexes. High fluorescence level of LipofectamineÔ 2000/FAM-Chol-siRNA group was observed in both cytoplasm and nucleus. The results from confocal images revealed that rHDL could specifically deliver FAM-Chol-siRNA into cytoplasm, whereas FAM-Chol-siRNA delivered by LipofectamineÔ 2000 was found to be in cytoplasm as well as in nucleus. Accumulation of Chol-siRNA in cytoplasm where its target mRNA located (rather than nucleus) provided an advantage for the developed rHDL over LipofectamineÔ 2000 for Chol-siRNA delivery [38]. The specific cytoplasmic delivery of FAM-Chol-siRNA mediated by rHDL might be attributed to the general consensus that hydrophobic molecules encapsulated in HDL were selectively

To further investigate the effect of the release of Chol-siRNA from complexes upon the efficiency of gene silencing in vivo, we performed a 7-day Chol-siRNA-releasing experiment in PBS buffer (pH 7.4) in order to simulate in vivo biological environment, by comparing two formulations of Chol-siRNA plus rHDL solution and rHDL/Chol-siRNA complexes. As shown in Fig. 4B, the Chol-siRNA released fairly rapidly in Chol-siRNA plus rHDL solution. Within 24 h, approximately 98% Chol-siRNA was released from the solution. But for the rHDL/Chol-siRNA complexes, a steady continuedrelease effect of Chol-siRNA was observed. For instance, in vitro release of Chol-siRNA from complexes was negligible (<1%) within 24 h. These results showed that the rHDL/Chol-siRNA complexes displayed the sustained-release profile and relative stability for a long time. It could be concluded that the sustained-release profile was attributed to the formulation of complexes. This was because Chol-siRNA was incorporated into the phospholipid monolayer of rHDL and its release was controlled by its dissociation from the surface of rHDL. Whether the Chol-siRNA remained stable in rHDL during the release test in vitro was also investigated using 2.0% agarose gel electrophoresis, as evidenced by the different gel shifts of rHDL/Chol-siRNA complexes and naked Chol-siRNA. The results showed almost no degradation or migration of Chol-siRNA in the lanes of rHDL/Chol-siRNA complexes (Fig. 4C), thus indicating that the complexes were physically stable and intact after 7-day release at 37  C in PBS buffer, and might provide efficient delivery of CholsiRNA to cancer cells for gene silencing in vitro and in vivo. 3.6. Cellular uptake and intracellular location of Chol-siRNA One bottleneck in achieving efficient transfection with siRNA was the delivery of siRNA across cell membrane. We therefore investigated rHDL-mediated uptake of FAM-Chol-siRNA into HepG2 cells. The fluorescent tag FAM enabled the cellular uptake and the intracellular localization of Chol-siRNA to be monitored. The results from FACS analysis showed rHDL mediated the highly efficient uptake of FAM-Chol-siRNA into HepG2 cells. The percentage of siRNA positive cells in the rHDL/FAM-Chol-siRNA group exceeded 50%, which was slightly lower than that achieved by LipofectamineÔ 2000/FAM-Chol-siRNA complexes (w55%), but significantly higher than that of Lipos/FAM-Chol-siRNA complexes (w9.0%) and FAM-Chol-siRNA alone (only w2.0%) (Fig. 5A and B). The rHDL-facilitated uptake of FAM-Chol-siRNA could be attributed to the interaction of apoA-I as a ligand with the receptor SR-BI on the HepG2 cells [37], which was inferred from the significantly increased cell uptake of FAM-Chol-siRNA after incubation of Lipos/ FAM-Chol-siRNA complexes with apoA-I. Also, this might be due to better protection of FAM-Chol-siRNA by rHDL during transport or a combination of these effects. The cellular uptake was also confirmed with confocal microscopy observation (Fig. 5C). Clear

Fig. 6. (A) Cytotoxicity of PEI and rHDL at various concentrations (0.05e1000 mg/mL) against HepG2 cells after 72 h incubation. ***P < 0.001, relative to PEI group. (B) Cytotoxicity of rHDL/Chol-siRNA complexes, Chol-siRNA-Pokemon alone and rHDL/ Chol-siRNA-Pokemon complexes against HepG2 cells at the same Chol-siRNA concentration of 100 nM for different durations of incubation. *P < 0.05, **P < 0.01 and ***P < 0.001, versus rHDL/Chol-siRNA control group. Results were expressed as means  S.D. (n ¼ 3).

Y. Ding et al. / Biomaterials 33 (2012) 8893e8905

8901

Fig. 7. RHDL/Chol-siRNA-Pokemon complexes induced HepG2 cell apoptosis and necrosis in vitro. (A) Morphologic changes of cells observed under an inverted light microscope (  200) after treatment with rHDL, Chol-siRNA-Pokenmon alone or rHDL/Chol-siRNA-Pokemon complexes. (B) Quantification of the apoptotic and necrotic cells by annexin V/PI double-staining assay. Results were expressed as means  S.D. (n ¼ 3). *P < 0.05 and ***P < 0.001, versus Chol-siRNA-Pokemon alone group.

taken up via SR-BI receptor into cancer cells, and this trafficking was mediated through a non-endocytic pathway [39]. LipofectamineÔ 2000 with high charge density would efficiently condense FAM-Chol-siRNA to form tightly compact complexes, which remain relatively intact after cellular internalization, and might then enter the nucleus together with its cargo of FAM-CholsiRNA. However, due to lower charge density, rHDL could form loosely compact complexes with FAM-Chol-siRNA, which might dissociate from the complexes after the recognition of SR-BI receptor by apoA-I, limiting its localization to cytoplasm. One of the major obstacles to effective siRNA delivery is the cytotoxicity of non-viral vectors. Light microscopic observation of cells treated with LipofectamineÔ 2000/FAM-Chol-siRNA complexes revealed gross morphological alteration, cell disruption and reduction in viability, while rHDL/FAM-Chol-siRNA complexes had no cytotoxicity and good toleration in HepG2 cells (data not shown). Taking these advantages of rHDL into consideration, our observations raised the possibility that rHDL was a safe and efficient alternative vector for Chol-siRNA delivery. 3.7. In vitro cytotoxicity studies The cytotoxicity of PEI, rHDL, rHDL/Chol-siRNA complexes, CholsiRNA-Pokemon alone and rHDL/Chol-siRNA-Pokemon complexes was evaluated against HepG2 cells by MTT assay. Cells treated with various concentrations of rHDL ranging from 0.05 to 1000 mg/mL yielded an almost similar viability like control (untreated cells). In contrast, significant inhibitory effects on cells treated with PEI were observed (Fig. 6A). The cytotoxicity of PEI was related to its considerable surface positive charges [40], while the remarkable decrease in cytotoxicity of rHDL might be attributed to its neutral charge. As shown in Fig. 6B, a little decrease of cell viability was observed when Chol-siRNA-Pokemon alone was applied, and significant inhibitory

Fig. 8. Expression of Pokemon and Bcl-2 proteins in HepG2 cells at 48 h after rHDLmediated Chol-siRNA-Pokemon transfection assayed by Western-blot. Representative immunoblots were shown as (A). Lane 1: Control; Lane 2: Chol-siRNA-Pokemon alone; Lane 3: rHDL/Chol-siRNA-Pokemon complexes. Results were expressed as means  S.D. (n ¼ 3). *P < 0.05, compared with control (B).

8902

Y. Ding et al. / Biomaterials 33 (2012) 8893e8905

Fig. 9. (A) Whole-body imaging of tumor-bearing nude mice administrated with Lipos/Cy5-Chol-siRNA complexes at 6 h (a1) and 24 h (a3) and rHDL/Cy5-Chol-siRNA complexes at 6 h (a2) and 24 h (a4), respectively. The NIR fluorescence images and X-ray images were fused together with Kodak molecular imaging systems software V5.0.1. (B) Quantification of the ex vivo tumor uptake characteristics of complexes in tumor-bearing nude mice after 24 h injection. Uptake expressed as photo flux per mm2 of tumor. Results were expressed as the mean  S.D. (n ¼ 6). ***P < 0.001, versus Lipos/Cy5-Chol-siRNA group. Representative ex vivo NIR fluorescence images of tumors collected at 24 h post-injection of complexes (the upper right corner of the figure).

effects on cells treated with rHDL/Chol-siRNA-Pokemon complexes were observed compared to cells treated with rHDL/Chol-siRNA complexes. Moreover, the viability of cells exposed to rHDL/CholsiRNA-Pokemon complexes gradually decreased over time after transfection. The effective cell growth inhibition by rHDL/CholsiRNA-Pokemon complexes could be due to receptor-mediated uptake of Chol-siRNA-Pokemon and the decrease of Pokemon and Bcl-2 protein expression in HepG2 cells (see Fig. 8). 3.8. Induction of apoptosis and necrosis in HepG2 cells in vitro After treatment with rHDL/Chol-siRNA-Pokemon complexes, HepG2 cells grew slowly, turned round in shape and were even severely distorted. The control group displayed normal, healthy growth as demonstrated by the clear skeletons observed by

inverted microscopy (Fig. 7A). To further assess whether rHDLmediated Chol-siRNA-Pokemon transfer could induce apoptosis of HepG2 cells, cells were stained by annexin V/PI and the quantification of the apoptotic and necrotic cells was detected. Cells labeled with annexin Vþ/PI were early apoptotic cells. In control and rHDL groups, little apoptosis was found in 1.12% and 1.38% of the total cell population, respectively. The rHDL/Chol-siRNA-Pokemon group could significantly induce more apoptosis than the Chol-siRNAPokemon alone group, and the percentages of apoptotic cells were 2.73% (Chol-siRNA-Pokemon alone) and 33.69% (rHDL/CholsiRNA-Pokemon complexes), respectively (Fig. 7B). The rHDL/CholsiRNA-Pokemon group also had great ability to induce more necrosis than the other groups. This obvious apoptotic and necrotic occurrence in HepG2 cells was consistent with the effective cell growth inhibition observed by MTT assay.

Y. Ding et al. / Biomaterials 33 (2012) 8893e8905

3.9. Expression of Pokemon and Bcl-2 proteins assayed by Westernblot As shown in Fig. 8, in untreated cells, a high-level expression of Pokemon protein was detected. At 48 h after transfection with rHDL/ Chol-siRNA-Pokemon complexes, the expression level of Pokemon protein decreased dramatically, whereas that of cells treated with Chol-siRNA-Pokemon alone had no significant changes compared with the control. Besides, it was interesting to note that the transfection of rHDL/Chol-siRNA-Pokemon complexes also downregulated the expression of apoptosis-related protein Bcl-2. Higher transfection of rHDL/Chol-siRNA-Pokemon complexes in HepG2 cells might be responsible for the lower expression of Pokemon mRNA (data not shown). Moreover, the reduced mRNA level of Pokemon could decrease the protein level of Pokemon in the cells transfected with rHDL/Chol-siRNA-Pokemon complexes. Pokemon, an upstream oncogene, plays a key role in malignant transformation of tumor and may regulate some downstream target genes (e.g. Bcl-2 gene) to inhibit apoptosis of tumor cells. Bcl-2 is an apoptosis suppressor gene in human tumors. The protein encoded by Bcl-2 gene is implicated in the prolongation of cell survival by blocking programmed cell death, i.e. apoptosis. Therefore, after treatment with rHDL/Chol-siRNA-Pokemon complexes, the reduction of Pokemon protein level might result in the downregulation of Bcl-2 protein thereby significantly inducing apoptosis and inhibiting growth of HepG2 cells. In our future work, further studies should be performed to explore apoptosis the relation of occurrence and changes of other downstream proteins following the silence of Pokemon gene in HepG2 cells transfected with rHDL/Chol-siRNAPokemon complexes. 3.10. Whole-body imaging and ex vivo imaging RHDL/Chol-siRNA-Pokemon complexes have shown considerable gene silencing ability in vitro. Subsequently, the in vivo biodistribution of rHDL/Cy5-Chol-siRNA complexes in tumor-bearing nude mice was investigated using a non-invasive NIR optical

8903

imaging technique, which allowed for monitoring of nanoparticle tumor penetration with high spatial and temporal resolution. The tumor-bearing nude mice were injected with Lipos/Cy5-CholsiRNA (50 mg) or rHDL/Cy5-Chol-siRNA (50 mg) complexes, respectively. Whole-body fluorescent images were taken at 6 and 24 h after i.v. administration, respectively. As shown in Fig. 9A, the fluorescence intensity was increased in tumor, liver and kidney at 6 h after administration of rHDL/Cy5-Chol-siRNA complexes. This pattern of distribution was entirely consistent with the distribution of SR-BI receptors, which were expressed mainly in tumor or liver and, to a lesser extent, in kidney or normal ovarian tissues [41e43], and might be responsible for the receptor-mediated targeting of rHDL/Cy5-Chol-siRNA complexes. However, compared with Lipos/Cy5-Chol-siRNA group, rHDL/Cy5-CholsiRNA complexes accumulated in tumor were obviously increased in the nude mice. Moreover, a signal enhancement localized in the regions corresponding to the tumor at 24 h after administration of rHDL/Cy5-Chol-siRNA complexes was still observed. In vivo biodistribution and targeting efficiency of rHDL/Cy5-Chol-siRNA complexes indicated that rHDL could assist Chol-siRNA-Pokemon targeting to the tumors. This high targeting efficiency of rHDL/ Cy5-Chol-siRNA complexes in tumors might be due to a combination of an EPR effect and receptor-mediated uptake mechanism. As shown in Fig. 9B, which gave the ex vivo images, a strong NIR dye Cy5 fluorescent signal was present in the isolated tumor of nude mice treated with rHDL/Cy5-Chol-siRNA complexes at 24 h after injection. However, much less Cy5 fluorescent signal was detected in the tumor of Lipos/Cy5-Chol-siRNA group. In the quantitative analyses, the fluorescent intensity from the tumor of rHDL/Cy5-Chol-siRNA group was about 4.22 times ([31.71  1.23]  105 versus [7.51  1.68]  105) greater than that of Lipos/Cy5-Chol-siRNA group. This result further confirmed that rHDL could effectively distribute into the tumors, reduce the elimination of complexes and prolong their retention in tumor site. Thus, further studies should be performed to investigate in vivo antitumor effects of rHDL-mediated targeted Chol-siRNAPokemon delivery.

Fig. 10. Growth suppression and protein expression of HepG2 xenograft tumor in BALB/c nude mice. (A) Tumor volume analysis of tumor-bearing nude mice after i.v. administration of rHDL/Chol-siRNA-Pokemon complexes. Starting from day 0, tumor volumes were measured every other day for 2 weeks. Results were expressed as means  S.D. (n ¼ 6). (B) Representative immunoblots of Pokemon and Bcl-2 protein expression in HepG2 cells in vivo at day 14 after i.v. treatments assayed by Western-blot. Lane 1: Control; Lane 2: CholsiRNA-Pokemon alone; Lane 3: rHDL/Chol-siRNA-Pokemon complexes. (C) Results from Western-blot analysis were expressed as means  S.D. (n ¼ 3). *P < 0.05, compared with control group.

8904

Y. Ding et al. / Biomaterials 33 (2012) 8893e8905

3.11. In vivo tumor growth suppression and protein expression To further reveal the potential of rHDL-based Chol-siRNAPokemon delivery system in cancer therapy, HepG2 cells were xenografted subcutaneously in BALB/c nude mice. As indicated in Fig. 10A, almost no inhibition of tumor growth was observed in Chol-siRNA-Pokemon group compared with control group. In contrast, the group treated with rHDL/Chol-siRNA-Pokemon complexes showed significant growth inhibition of established xenograft tumor in nude mice with HepG2 cells. As shown in Fig. 10B and C, the relative expression levels of Pokemon and Bcl-2 protein in tumor tissue at day 14 after injection of rHDL/CholsiRNA-Pokemon complexes were markedly reduced in comparison with those of control group. The results indicated that the tumor-suppressive effect was particularly due to the Chol-siRNA for the silencing of Pokemon, and the suppressive effect of Chol-siRNA carried by rHDL was more powerful. Because Pokemon proved to be an essential factor for oncogenesis in vivo, rHDL-mediated Chol-siRNA targeting-silencing of Pokemon could obviously inhibit the growth of HepG2 cells in BALB/c nude mice. When exploring the anti-tumor role of CholsiRNA-Pokemon in vivo, we were particularly interested in a HepG2 xenograft tumor model with i.v. administration of rHDL/ Chol-siRNA-Pokemon complexes for the following reasons. Firstly, previous studies indicated that Pokemon was highly expressed in many human cancers, including hepatocellular carcinoma. Secondly, Pokemon was critical for oncogenic transformation of HepG2 cells and was able to act as a proto-oncogene in cooperation with other classic oncogenes [24]. Lastly, rHDL facilitated efficient systemic delivery of siRNA in vivo, mediated by the SR-BI receptor overexpressed in the HepG2 cells. In summary, the efficient systemic delivery ability of Chol-siRNA mediated by the rHDL might offer the selective and effective delivery of siRNA into targeted cells, and provide a promising strategy to bypass the challenge of endolysosomal trafficking, thus making the rHDL a useful tool of targeted Chol-siRNA delivery for tumor gene therapy. 4. Conclusions In this study, we successfully developed a biomimetic nanovector rHDL for targeted Chol-siRNA delivery and for Pokemon gene silencing therapy. The well-formed rHDL/Chol-siRNA complexes showed high Chol-siRNA-loading efficiency, desirable size distribution, excellent serum and storage stability. The most important advantage of rHDL was that it could provide highly efficient transfection, direct cytosolic delivery and superior tumor targetability, which decreased the expression levels of Pokemon and Bcl-2 proteins thereby significantly improving the antitumor efficacy of Chol-siRNA-Pokemon against HepG2 cells in vitro and in vivo. Therefore, rHDL should be a promising non-viral Chol-siRNA vector, and rHDL-mediated Chol-siRNA-Pokemon transfer might have the potential for clinical applications of hepatocellular carcinoma gene therapy. Acknowledgments This study was technically supported by the staff from Center of Biomedical Analysis, Department of Pharmaceutics, China Pharmaceutical University, and financially supported by the National Natural Science Foundation of China (No. 81102398), the Natural Science Foundation of Jiangsu Province (No. BK2011624), the Natural Science Fund for Colleges and Universities in Jiangsu Province (No. 10KJB310007), the Fundamental Research Funds for the Central Universities (No. JKP2011007), the Open Project

Program of State Key Laboratory of Natural Medicines, China Pharmaceutical University (No. SKLNMKF201215) and the National Found for Fostering Talents of Basic Science (No. J0630858). References [1] de Fougerolles A, Vornlocher HP, Maraganore J, Lieberman J. Interfering with disease: a progress report on siRNA-based therapeutics. Nat Rev Drug Discov 2007;6:443e53. [2] Iorns E, Lord CJ, Turner N, Ashworth A. Utilizing RNA interference to enhance cancer drug discovery. Nat Rev Drug Discov 2007;6:556e68. [3] Kim D, Rossi J. RNAi mechanisms and applications. Biotechniques 2008;44: 613e6. [4] Zhang N, Tan C, Cai P, Zhang P, Zhao Y, Jiang Y. RNA interference in mammalian cells by siRNAs modified with morpholino nucleoside analogues. Bioorg Med Chem 2009;17:2441e6. [5] Wolfrum C, Shi S, Jayaprakash KN, Jayaraman M, Wang G, Pandey RK, et al. Mechanisms and optimization of in vivo delivery of lipophilic siRNAs. Nat Biotechnol 2007;25:1149e57. [6] Fattal E, Bochot A. State of the art and perspectives for the delivery of antisense oligonucleotides and siRNA by polymeric nanocarriers. Int J Pharm 2008;364:237e48. [7] Chiu YL, Ali A, Chu CY, Cao H, Rana TM. Visualizing a correlation between siRNA localization, cellular uptake, and RNAi in living cells. Chem Biol 2004; 11:1165e75. [8] Song E, Zhu P, Lee SK, Chowdhury D, Kussman S, Dykxhoorn DM, et al. Antibody mediated in vivo delivery of small interfering RNAs via cell-surface receptors. Nat Biotechnol 2005;23:709e17. [9] Soutschek J, Akinc A, Bramlage B, Charisse K, Constien R, Donoghue M, et al. Therapeutic silencing of an endogenous gene by systemic administration of modified siRNAs. Nature 2004;432:173e8. [10] Morrissey DV, Lockridge JA, Shaw L, Blanchard K, Jensen K, Breen W, et al. Potent and persistent in vivo anti-HBV activity of chemically modified siRNAs. Nat Biotechnol 2005;23:1002e7. [11] Lorenz C, Hadwiger P, John M, Vornlocher HP, Unverzagt C. Steroid and lipid conjugates of siRNAs to enhance cellular uptake and gene silencing in liver cells. Bioorg Med Chem Lett 2004;14:4975e7. [12] Sioud M, Sorensen DR. Cationic liposome-mediated delivery of siRNAs in adult mice. Biochem Biophys Res Commun 2003;312:1220e5. [13] Meade BR, Dowdy SF. Exogenous siRNA delivery using peptide transduction domains/cell penetrating peptides. Adv Drug Deliv Rev 2007;59:134e40. [14] Rigotti A, Miettinen HE, Krieger M. The role of the high-density lipoprotein receptor SR-BI in the lipid metabolism of endocrine and other tissues. Endocr Rev 2003;24:357e87. [15] Rensen PC, de Vrueh RL, Kuiper J, Bijsterbosch MK, Biessen EA, van Berkel TJ. Recombinant lipoproteins: lipoprotein-like lipid particles for drug targeting. Adv Drug Deliv Rev 2001;47:251e76. [16] Connelly MA, Williams DL. SR-BI and HDL cholesteryl ester metabolism. Endocr Res 2004;30:697e703. [17] Connelly MA, Williams DL. Scavenger receptor BI: a scavenger receptor with a mission to transport high density lipoprotein lipids. Curr Opin Lipidol 2004; 15:287e95. [18] Zannis VI, Chroni A, Krieger M. Role of apoA-I, ABCA1, LCAT, and SR-BI in the biogenesis of HDL. J Mol Med (Berl) 2006;84:276e94. [19] Pussinen PJ, Lindner H, Glatter O, Reicher H, Kostner GM, Wintersperger A, et al. Lipoprotein-associated alpha-tocopheryl-succinate inhibits cell growth and induces apoptosis in human MCF-7 and HBL-100 breast cancer cells. Biochim Biophys Acta 2000;1485:129e44. [20] Lacko AG, Nair M, Paranjape S, Johnso S, McConathy WJ. High density lipoprotein complexes as delivery vehicles for anticancer drugs. Anticancer Res 2002;22:2045e9. [21] Lou B, Liao XL, Wu MP, Cheng PF, Yin CY, Fei Z. High-density lipoprotein as a potential carrier for delivery of a lipophilic antitumoral drug into hepatoma cells. World J Gastroenterol 2005;11:954e9. [22] McConathy WJ, Nair MP, Paranjape S, Mooberry L, Lacko AG. Evaluation of synthetic/reconstituted high-density lipoproteins as delivery vehicles for paclitaxel. Anticancer Drugs 2008;19:183e8. [23] Maeda T, Hobbs RM, Merghoub T, Guernah I, Zelent A, Cordon-Cardo C, et al. Role of the proto-oncogene Pokemon in cellular transformation and ARF repression. Nature 2005;433:278e85. [24] Maeda T, Hobbs RM, Pandolfi PP. The transcription factor Pokemon: a new key player in cancer pathogenesis. Cancer Res 2005;65:8575e8. [25] Agrawal A, Yang J, Murphy RF, Agrawal DK. Regulation of the p14ARF-Mdm2p53 pathway: an overview in breast cancer. Exp Mol Pathol 2006;81:115e22. [26] Apostolopoulou K, Pateras IS, Evangelou K, Tsantoulis PK, Liontos M, Kittas C, et al. Gene amplification is a relatively frequent event leading to ZBTB7A (Pokemon) overexpression in non-small cell lung cancer. J Pathol 2007;213: 294e302. [27] Cui J, Meng X, Gao X, Tan G. Curcumin decreases the expression of Pokemon by suppressing the binding activity of the Sp1 protein in human lung cancer cells. Mol Biol Rep 2010;37:1627e32. [28] Jeon BN, Yoo JY, Choi WI, Lee CE, Yoon HG, Hur MW. Proto-oncogene FBI-1 (Pokemon/ZBTB7A) represses transcription of the tumor suppressor Rb gene

Y. Ding et al. / Biomaterials 33 (2012) 8893e8905

[29]

[30]

[31]

[32]

[33] [34]

[35] [36]

via binding competition with Sp1 and recruitment of co-repressors. J Biol Chem 2008;283:33199e210. Zhao ZH, Wang SF, Yu L, Wang J, Chang H, Yan WL, et al. Overexpression of Pokemon in non-small cell lung cancer and foreshowing tumor biological behavior as well as clinical results. Lung Cancer 2008;62:113e9. Choi WI, Jeon BN, Yun CO, Kim PH, Kim SE, Choi KY, et al. Proto-oncogene FBI1 represses transcription of p21CIP1 by inhibition of transcription activation by p53 and Sp1. J Biol Chem 2009;284:12633e44. Zhang WL, Xiao Y, Liu JP, Wu ZM, Gu X, Xu YM, et al. Structure and remodeling behavior of drug-loaded high density lipoproteins and their atherosclerotic plaque targeting mechanism in foam cell model. Int J Pharm 2011;419:314e 21. Schiffelers RM, Ansari A, Xu J, Zhou Q, Tang Q, Storm G, et al. Cancer siRNA therapy by tumor selective delivery with ligand-targeted sterically stabilized nanoparticle. Nucleic Acids Res 2004;32:e149. Baker M. RNA interference: Homing in on delivery. Nature 2010;464:1225e8. Urban-Klein B, Werth S, Abuharbeid S, Czubayko F, Aigner A. RNAi-mediated gene-targeting through systemic application of polyethylenimine (PEI)-complexed siRNA in vivo. Gene Ther 2005;12:461e6. Li W, Szoka Jr FC. Lipid-based nanoparticles for nucleic acid delivery. Pharm Res 2007;24:438e49. Sriamornsak P, Thirawong N, Nunthanid J, Puttipipatkhachorn S, Thongborisute J, Takeuchi H. Atomic force microscopy imaging of novel

[37]

[38] [39] [40]

[41]

[42]

[43]

8905

self-assembling pectineliposome nanocomplexes. Carbohydr Polym 2008;71: 324e9. Collet X, Tall AR, Serajuddin H, Guendouzi K, Royer L, Oliveira H, et al. Remodeling of HDL by CETP in vivo and by CETP and hepatic lipase in vitro results in enhanced uptake of HDL CE by cells expressing scavenger receptor B-I. J Lipid Res 1999;40:1185e93. Song JJ, Smith SK, Hannon GJ, Joshua-Tor L. Crystal structure of Argonaute and its implications for RISC slicer activity. Science 2004;305:1434e7. Ng KK, Lovell JF, Zheng G. Lipoprotein-inspired nanoparticles for cancer theranostics. Acc Chem Res 2011;44:1105e13. Zhong Z, Feijen J, Lok MC, Hennink WE, Christensen LV, Yockman JW, et al. Low molecular weight linear polyethylenimine-b-poly(ethylene glycol)-bpolyethylenimine triblock copolymers: synthesis, characterization, and in vitro gene transfer properties. Biomacromolecules 2005;6:3440e8. Feng M, Cai Q, Huang H, Zhou P. Liver targeting and anti-HBV activity of reconstituted HDL-acyclovir palmitate complex. Eur J Pharm Biopharm 2008; 68:688e93. Feng M, Cai Q, Shi X, Huang H, Zhou P, Guo X. Recombinant high-density lipoprotein complex as a targeting system of nosiheptide to liver cells. J Drug Target 2008;16:502e8. Mooberry LK, Nair M, Paranjape S, McConathy WJ, Lacko AG. Receptor mediated uptake of paclitaxel from a synthetic high density lipoprotein nanocarrier. J Drug Target 2010;18:53e8.