Article pubs.acs.org/bc
SiRNA Delivery Systems Based on Neutral Cross-Linked Dendrimers Jie Liu,† Jihan Zhou,‡ and Ying Luo*,† †
Department of Biomedical Engineering, College of Engineering, Peking University, Room 206, Fangzheng Building, 298 Chengfu Road, Haidian District, Beijing, China 100871 ‡ Beijing National Laboratory for Molecular Sciences and the Key Laboratory of Polymer Chemistry and Physics of Ministry of Education, College of Chemistry and Molecular Engineering, Peking University, Beijing, China, 100871 S Supporting Information *
ABSTRACT: A neutral cross-linked dendritic system is described for use of delivering small interfering RNA (siRNA) for targeted gene silencing. By replacing the terminal amines with hydrazide groups and N-acetylgalactosamine (GalNAc) ligands, cationic polyamidoamine (PAMAM) dendrimers were transformed into neutral glycosylated carriers for siRNA delivery. Mainly owing to the pH sensitivity and the proton-absorption capability of the tertiary amines within the interior branches, these PAMAM derivatives showing neutrality under physiological conditions (pH 7.4) formed complexes with siRNA through electrostatic interactions at pH 5. Cross-linking procedures via reactions with glutaraldehyde were established, and cytocompatible cross-linked systems loaded with siRNA obtained, with the particulate properties variable with the cross-linking condition and the GalNAc level in the dendritic carrier. In vitro cellular uptake and RNAi experiments showed that the cross-linked dendritic systems with an appropriate level of GalNAc composition effectively interacted with HepG2 cells and inhibited the expression of a luciferase reporter gene. Neutral cross-linked dendritic systems provide a new paradigm for designing siRNA delivery systems with biocompatibility and targeting capability.
■
INTRODUCTION RNA interference (RNAi) is a fundamental cellular pathway through which the post-transcriptional gene expression is down-regulated via small RNA molecules. As an effective exogenous RNAi inducer, chemically synthesized small interfering RNA (siRNA), with 21−23 nucleotides in length, has not only provided powerful tools in cell culture for fundamental research,1 but also ushered in opportunities for new clinical therapeutics with great potential.2−7 However, as siRNA is susceptible to degradation by RNase and by itself can hardly cross the cell membrane due to its polyanionic nature and relatively large molecular weight, successful translation of siRNAs into clinical products is contingent on safe and efficient delivery systems. Numerous materials and methods have been investigated to create siRNA-containing nanoparticles, with a few promising examples advancing to nonhuman primate preclinical and further clinical studies in recent years.8−12 As illustrated by these studies regarding the design criteria for siRNA delivery systems, minimizing the toxicity and side effects of delivery materials (e.g., immunogenicity), and in the meantime ensuring the stability and transfection efficiency of siRNA-loaded systems are the most critical problems remaining to be addressed for optimal in vivo results; investigating strategies that can provide an integral solution is therefore of great importance. As in the development of delivery systems for DNA-based therapy, investigations on nonviral siRNA delivery have mainly centered on cationic carrier materials, typically prepared from © 2012 American Chemical Society
polymers, lipids, and peptides which can electrostatically complex with siRNA under physiological aqueous conditions.13 The ease and convenience have allowed the development of formulations and processing methods aiming at controlling the various properties of delivery systems with respect to particle size, stability, toxicity, and targeting capability. However, it is notable that the cationic materials as the carrier for nucleic acids may bear inherent disadvantages attributable to their positively charged components. First, the cytotoxicity of materials is often associated with cations such as the protonated primary amines, which have been found to cause membrane damage, apoptosis pathway activation, and medium depletion.14−17 Second, the electrostatic forces could be unstable in binding materials and siRNA together, which could cause dissociation or deformation of polyelectrolyte complexes. For example, it was reported that the particles made from siRNA and a cationic polypropyleneimine dendrimer exhibited apparent alterations in size after 48 h in PBS buffer;18 environmental factors such as ionic strength and shear stress may also change the structure and property of polyelectrolyte delivery systems.19−22 Third, the release of siRNA from polyelectrolyte complexes is an uncontrollable process, since no specific conditions or signals can be defined for breaking the electrostatic interaction. Received: August 8, 2011 Revised: December 6, 2011 Published: January 31, 2012 174
dx.doi.org/10.1021/bc200433s | Bioconjugate Chem. 2012, 23, 174−183
Bioconjugate Chemistry
Article
evaporation. The resulting G4.5 PAMAM was dried under vacuum and stored at 4 °C. To introduce hydrazide groups to the dendrimer surface, hydrazine hydrate in 10-fold molar excess of the methyl esters of PAMAM was added to a G4.5 PAMAM solution (5 mg/mL in ethanol). The reaction was run under reflux at 55 °C for 24 h, before the ethanol was removed by rotary evaporation. The crude product was purified by extensive dialysis against deionized water using a cellulose membrane (10 kDa MWCO, Union Carbide, NY). The final product was lyophilized and stored at −20 °C for future use. To conjugate GalNAc ligands, 10 mg/mL PAMAM-HYD in PBS at pH 5.0, was reacted with GalNAc of 0.5, 1, or 1.5 mol equiv of the hydrazide groups. The reaction was kept at 50 °C for 24 h, before the GPH products were purified through extensive dialysis against deionized water using the 10 kDa MWCO cellulose membrane and dried via lyophilization. The GalNAc modification level in each product was analyzed by 1H NMR. Complexation of siRNA with Dendritic Materials and Preparation of Cross-Linked Delivery Systems. 10× siRNA stock solutions were prepared at 20 μM in DEPC treated water. To study the siRNA−dendrimer complex, siRNA in PBS (2 μM) was added with an equal volume of a dendrimer solution with a prescribed concentration between 1.3 μM and 6.6 μM. The complexation experiments were investigated under both acidic and physiological neutral conditions, with the pH value set at 5.0 and 7.4, respectively. To prepare cross-linked systems, siRNA (2 μM), the selected dendrimer material (3.3 μM) and glutaraldehyde were separately dissolved in PBS at pH 5.0 and mixed at 1:1:1 volume ratio. Typically, glutaraldehyde and siRNA were first mixed well before the third part containing dendrimers was added. The glutaraldehyde concentration was specified below in each experiment. After the ternary blends were incubated at 37 °C for 1 h, an excessive amount of ADH was added to terminate the cross-linking reaction. The pH was adjusted to 7 using a 0.2 M NaOH solution, and cross-linked dendritic systems containing siRNA were dialyzed against PBS at pH 7.4 for 2 h using Slide-A-Lyzer MINI Dialysis Devices (3500 MWCO, Pierce, Rockford, IL). The resulting products were used directly for characterization and cellular studies. Gel Retardation Analysis. To evaluate whether siRNA can form complexes with the dendritic carrier material or be trapped in the cross-linked system, gel retardation experiments were performed to visually analyze the unbound siRNA in solution. The ternary systems comprising siRNA, dendrimer, and glutaraldehyde were loaded in 3.5% agarose gel in a pH 8.5 TBE buffer, and imaged by a Tanon-1600 Gel Documentation System (Tanon, Shanghai, China). To prepare cross-linked systems, siRNA, PAMAM-HYD, and glutaraldehyde were mixed at equal volumes and cross-linked at pH 5.0 as described above; the glutaraldehyde concentration was varied from 0.17 mM to 83 mM. To understand the effect of pH conditions, the siRNA, PAMAM-HYD and glutaraldehyde (1.7 mM to 8.3 mM) solutions were also prepared and mixed at pH 7.4, and then electrophoresed in the agarose gel. To study the ability of GPH cross-linked systems to trap siRNA, siRNA, dendrimer, and glutaraldehyde (25 mM) were mixed and cross-linked under acidic conditions as described above. The samples were then electrophoresed; naked siRNA and the siRNA-PAMAM-HYD mixture without glutaraldehyde were control systems for comparison.
In this study, we intend to devise a neutral cross-linked system which can provide stability, safety, and controllability for siRNA delivery, so as to overcome some of the limitations inherent to cationic polymers or lipids. Toward this goal, we focus on transforming a cationic dendritic polymer material polyamidoamine (PAMAM) which has been used for gene deliveryinto neutral targeted carriers for siRNA. In particular, G4.0 PAMAMs with the surface primary amines substituted by neutral cross-linkable hydrazide groups and carbohydratetargeting ligands, N-acetyl-galactosamine (GalNAc), were investigated. These glycosylated dendritic materials, which are neutral at physiological conditions, can become protonated when pH is lowered, owing to the pH sensitivity and the proton-adsorption capability of the interior tertiary amine within the PAMAM scaffolds.23 This property allows the neutral dendritic materials to form complexes with siRNA at pH 5, which can be cross-linked via the reaction with glutaraldehyde for the preparation of a new type of stable siRNA delivery system. Based on the cross-linking methodology, we investigated variables affecting the properties of the cross-linked dendritic delivery systems. Formulations that could most effectively induce RNAi in HepG2 cells were studied and discussed. In summary, cross-linked delivery systems based on neutral dendritic materials may offer advantages over conventional delivery systems and hold promise for development of new nucleic acid-based therapeutics.
■
MATERIALS AND METHODS Materials. Water was distilled and deionized at 18 MΩ resistance (Gelante Pure Water, Shijiazhuang, China). Generation 4.0 and 5.0 PAMAMs, methyl acrylate, GalNAc, adipic acid dihydrazide (ADH), and diethyl pyrocarbonate (DEPC) were purchased from Sigma-Aldrich (Milwaukee, WI) and hydrazine hydrate from Alfa Aesar (Ward Hill, MA). GeneFinder nucleic acid stain was obtained from Bio-V (Xiamen, China). NHS-fluorescein and BCA protein assay kit were purchased from Pierce (Rockford, IL). Firefly luciferase reporter vector pGL4.51[luc2/CMV/Neo], Glo Lysis Buffer, Bright-Glo luciferase assay kits, and CellTiter 96 AQueous One Solution Cell Proliferation Assay (MTS) systems were from Promega (Madison, WI). Cell culture reagents and Lipofectamine 2000 were obtained from Invitrogen (Carlsbad, CA) unless otherwise specified. SiRNAs against firefly luciferase with or without Cy3 labeling and the scrambled control siRNA were synthesized by RiboBio (Guangzhou, China). The siRNA sequences for luciferase were 5′-UGAAGAGCCUGAUCAAAUA dTdT-3′ (sense) and 3′-dTdT ACUUCUCGGACUAGUUUAU-5′ (antisense), and the scrambled siRNA sequences were 5′-UUCUCCGAACGUGUCACGU dTdT (sense) and 3′-dTdT AAGAGGCUUGCACAGUGCA-5′ (antisense). Suppliers of other chemicals, biological reagents, and equipment are specified below. Synthesis of PAMAM-Hydrazide (PAMAM-HYD) and GalNAc Modified PAMAM-HYD (GPH). PAMAM-HYD and GPH dendrimers were synthesized via a protocol established in our laboratory.24 To synthesize PAMAM-HYD bearing hydrazide terminal groups, G4.5 PAMAM was first prepared starting from commercially obtained G4.0 PAMAM.25 Briefly, a 50 mg/mL G4.0 PAMAM solution was reacted with methyl acrylate in 10-fold molar excess of the primary amines on the PAMAM surface at 37 °C for 48 h. The solvent and the unreacted methyl acrylate were removed at 65 °C by rotary 175
dx.doi.org/10.1021/bc200433s | Bioconjugate Chem. 2012, 23, 174−183
Bioconjugate Chemistry
Article
Scheme 1a
a
(A) The structure of GalNAc and hydrazide decorated PAMAM dendrimer; (B) schematic process of preparing cross-linked dendritic systems for siRNA delivery.
Fluorescent Dye Exclusion Assay. To quantitatively determine the unbound siRNA in binary mixtures of siRNA and dendrimer materials, a nucleic acid stain, GeneFinder, was used. Briefly, GeneFinder was added to the siRNA solution (2 μM) at 1:50 volume ratio, and the siRNA was then blended with a dendrimer solution as described above. The fluorescence signals (excitation: 488 nm; emission: 522 nm) resulting from GeneFinder intercalating with siRNA were measured by a SpectraMax M2 microplate reader (Molecular Devices, CA). To study the siRNA loading efficiency in cross-linked systems, a 2 μM siRNA solution was used to prepare cross-linked system with PAMAM-HYD or GPH materials as described above. After pH adjustment, 20 μL of each sample was diluted with 80 μL PBS (pH 7.4), before the GeneFinder solution was added to the sample solutions at 1:100 volume ratio and the fluorescence signals measured. Zeta-Potential and Particle Size Analysis. To characterize the zeta-potentials of un-cross-linked siRNA−dendrimer complexes and cross-linked systems, samples were prepared as described in the section, Complexation of siRNA with Dendritic Materials and Preparation of Cross-linked Delivery Systems. To prepare PAMAM-HYD cross-linked system, the concentration of glutaraldehyde was varied from 0.83 to 833 mM. For GPH cross-linked systems, the glutaraldehyde concentration was fixed at 8.3 mM. To perform zeta-potential measurements, all final mixed or cross-linked systems were diluted 10 times using deionized water and analyzed through electrophoretic light scattering experiments by a ZetaPALS (Brookhaven Instruments, NY). To determine the size of siRNA-loaded systems, samples were analyzed through dynamic light scattering (DLS) experiments at an angle of 90° by the ZetaPALS instrument. DLS experiments were also conducted in Dr. Dehai Liang’s Laboratory at the College of Chemistry and Molecular Engineering of Peking University, using a commercialized spectrometer (Brookhaven Inc., Holtsville, NY) equipped with
a BI-200SM Goniometer and a BI-TurboCo Digital Correlator. A solid-state laser (100 mW, 532 nm, Changchun, China) polarized at the vertical direction was used as the light source and the scattering angles were from 30° to 120°. By using a Laplace inversion program, CONTIN, the normalized distribution function of the characteristic line width which could be further converted into the hydrodynamic radius R h distribution was obtained (Supporting Information S1). Cellular Studies of Dendritic Materials and CrossLinked siRNA Delivery Systems. To conduct cellular studies, HepG2 cells were seeded in 96-well plates at a density of 1.5 × 104 cells per well and allowed to attach to surface and grow for 24 h before being used for further experiments. SiRNA-loaded cross-linked systems made from PAMAM-HYD or GPH dendrimers were prepared by siRNA (2 μM), dendrimer (3.3 μM), and glutaraldehyde (25 mM) solutions as described above. The cross-linked systems were diluted in the DMEM medium to appropriate concentrations for cellular experiments. Cytotoxicity Assay. Dendritic material solutions of a series of concentrations or cross-linked delivery systems were prepared in DMEM containing 10% FBS. The samples of dendrimer materials and cross-linked delivery systems were then incubated with HepG2 cells cultured with 10% serum for 24 and 48 h, respectively. To analyze cytotoxicity, the cell culture supernatants were replaced by 100 μL of fresh DMEM medium and 20 μL of MTS reagent. After incubation at 37 °C with 5% CO2 for 1 h, the light absorbance at 490 nm of each well was recorded by the SpectraMax M2 microplate reader. RNAi Experiments. HepG2 cells were transfected with firefly luciferase reporter vector pGL4.51 using Lipofectamine 2000 following the manufacturer’s guide. After 24 h, siRNA (2 μM) against the firefly luciferase gene or with scrambled sequences was loaded in cross-linked systems, and reconstituted at 1:10 volume ratio in DMEM supplemented with 10% FBS. 200 μL of the solutions with approximate 10 pmol siRNA were 176
dx.doi.org/10.1021/bc200433s | Bioconjugate Chem. 2012, 23, 174−183
Bioconjugate Chemistry
Article
dispensed into each well containing HepG2 cells expressing firefly luciferase. After incubation for 48 h, the medium was removed, and the cells were treated with Glo Lysis Buffer for 20 min. The luciferase protein expression was measured by a Bright-Glo luciferase assay system. To normalize the gene silencing efficiency, the total protein was measured via a BCA protein assay kit. Fluorescence Microscopy of Cellular Uptake of Dendritic Materials and Cross-Linked Delivery Systems. For microscopic cellular uptake studies, dendritic materials were labeled with a fluorescein dye. Specifically, NHSfluorescein (0.1 mg/mL) and dendrimer (2 mg/mL) solutions were mixed at 2:1 molar ratio, with DPBS made up volume to 50 μL and reacted for 3 h at 37 °C. The labeled dendrimer solutions were diluted in DMEM to a concentration of 1.6 μM, and 100 μL of sample solutions were added to HepG2 cells. After 24 h incubation, the medium was removed, and the cells were washed with DPBS for three times. Fluorescence images were captured by an IX71 fluorescence microscope (Olympus, Japan), and cellular uptake of materials were compared. To investigate siRNA delivery, cross-linked systems containing Cy3-labeled siRNA were prepared and diluted in DMEM with siRNA concentration at 50 nM. Samples were incubated with HepG2 cells for 12 h before imaging studies. Data Analysis and Statistics. Quantitative results were presented as means ± standard error of measurements (S.E.M.). Statistical comparisons were performed by one-way analysis of variance (ANOVA) followed by the Newman-Keuls tests to compare selected data pairs using SigmaStat 3.5. The level of significance was denoted by * and **, with P values set at
