Dextran (Mw∼40kDa), stearyl amine (99% pure), cystamine, pyridine, sodium periodate (NaIO₄), sodium cyanoborohydride (NaCNBH₃), potassium sulfate (K₂SO₄₎ and azo-bis-isobutyronitrile (AIBN) were obtained from Sigma-Aldrich Chemical Co (St. Louis, MO). Dithiol-modified poly(ethylene glycol) (PEG-(SH)₂, M.W. 2,000) was purchased from SunBio, Inc. (Seoul, South Korea). Anhydrous lithium chloride (LiCl) was obtained from Fisher Scientific (Philadelphia, PA). Dehydrated dimethylformamide (DMF) and dimethylsulfoxide (DMSO) with molecular sieves were obtained from Acros Organics (Parsipanny, NJ). Acryloyl chloride, pyridine and other reagents and solvents were from Sigma-Aldrich and were used as received without further purification.

The siRNA sequence targeting ABCB1 (Genbank accession no. NM_000927) corresponded to coding regions of these genes. Four target sequences were selected for the ABCB1 gene. Sense sequences for each siRNA were 5′ GAGCUUAACA CC CGA CUUAUU 3′, 5′ GAAAGUAUACCUCCAGUUUUU 3′, 5′ GAC CAUAAAU GUAAG GUUUUU 3′, and 5′ CCAGGUAUGCCUAUUAUUAUU 3′. Synthetic siRNA duplexes were obtained from Dharmacon Inc. (Layfayette, CO). The siRNAs were dissolved by adding 1 mL of the buffer [100 mmol/L potassium acetate, 30 mmol/L HEPES-KOH, and 2 mmol/L magnesium acetate (pH 7.4)] to each tube, and stored at −20°C until the following transfection experiment. Non-specific siRNA which has no significant homology to any known gene sequences from mouse, rat, or human was obtained from Ambion (Austin, Tx).

The MDR1 siRNA containing dextran nanoparticles were prepared in a similar fashion as reported recently for the preparation of doxorubicin loaded dextran nanosystems [18]. A stock solution of 5 mg/ml (∼2.5 mM) dextran thiol, 5 mg/ml (∼2.5 mM) dextran stearylamine, 5 mg/ml (125 µM) PEG-thiol, and 2 µM MDR1 siRNA was prepared. For obtaining a 100 nM siRNA concentration, 120 µl stock solution of MDR1 siRNA was mixed with 40 µl of dextran-thiol stock solution using a vortex shaker. It was then incubated for 5 minutes. To this mixture, 40 µl dextran-stearylamine derivative was then added and incubated for another 5 mins. Finally 40 µl of PEG-(SH)₂ was added and incubated for another 15 minutes to form the hydrophilic shell of the nanoparticles. This method of sequential addition was used so that the siRNA is entrapped in the interpenetrating (IPN) dextran hydrogel network. Further, a PEG-thiol derivative was used to provide stealth character to the nanoparticles. The final mixture was subsequently diluted with the medium and applied at the designated concentrations. Nanoparticles containing 100 nM non-specific siRNA was used as negative controls in some experiments.

The particle size and zeta potentials of MDR1 siRNA loaded nanoparticles were performed with a Brookhaven Zeta PALS Instrument (Holtsville, NY). For light scattering experiments, the samples were measured at fixed angle of 90° at 25°C. The scattering intensity was adjusted in the range of 50–500 kcps by diluting the samples with deionised water. For zeta potentials, default parameters of dielectric constant, refractive index, and viscosity of water were used based on the electrophoretic mobility of the nanoparticles.

Dr. Efstathios S. Gonos (National Hellenic research Foundation, Athens, Greece) kindly provided the human osteosarcoma cell line KHOS and the multidrug resistant MDR1 (P-gp) expressing cell lines KHOS_R2 and U-2OS_R2. All cell lines were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum, 100 U/ml penicillin, and 100 ug/ml streptomycin (all obtained from invitrogen, Carlsbad, CA). Cells were incubated at 37°C in 5% CO2-95% air atmosphere and passaged when near confluent monolayers were achieved using trypsin-EDTA solution. Resistant cell lines were continuously cultured in 0.1uM doxorubicin. Doxorubicin was obtained as unused residual clinical material at the Massachusetts General Hospital. The P-gp1 monoclonal antibody C219 was purchased from Signet (Dedham, MA). The human β-actin monoclonal antibody and the MTT reagents were purchased from Sigma-Aldrich (St. Louis, MO).

The Stat3 and enhanced green fluorescent protein (EGFP) fusion protein expression vector pCORON1000 EGFP-Stat3 (abbreviated as pEGFP-Stat3) was purchased from Amersham Biosciences (Buckinghamshire, United Kingdom). This pEGFP-Stat3 vector was generated by fusing Stat3 to the COOH terminus of EGFP. A hamster kidney cell line (BHK-21) was stably transfected with pEGFP-Stat3 through selection with G418 (Invitrogen, Carlsbad, CA). EGFP expressing cells were seeded at a density of 4000 cells per well in 96-well plates with addition of increased concentration of EGFP siRNA and subsequently incubated for 48 hr. EGFP siRNA was purchased from Ambion (Austin, Tx). siPORT™ NeoFX™Transfection Agent was used as a positive control (Ambion). Images were acquired by Nikon Eclipse Ti-U fluorescence microscope (Nikon Corp.) equipped with a SPOT RT digital camera (Diagnostic Instruments, Inc., Sterling Heights, MI).

P-gp1 was analyzed in total cell lysates. Protein lysates from cells were generated through lysis with 1× RIPA Lysis Buffer (Upstate Biotechnology, Charlottesville, VA). The concentration of the protein was determined by Protein Assay reagents (Bio-Rad, Hercules, CA) and a spectrophotometer (Beckman DU-640, Beckman Instruments, Inc., Columbia, MD). 25 µg of total protein was processed on Nu-Page 4–12% Bis-Tris Gel (Invitrogen) and transferred to a pure nitrocellulose membrane (Bio-Rad Laboratories, Hecules, CA). Primary antibodies were incubated in Tris-buffered saline, pH 7.4, with 0.1% Tween 20 overnight at 4°C. Signal was generated through incubation with horseradish peroxidase-conjugated secondary antibodies (Bio-Rad, Hercules, CA) incubated in Tris-buffered saline, pH 7.4, with 5% nonfat milk and 0.1% Tween 20 at 1 ∶ 2000 dilution for 1h at room temperature. Positive immunoreactions were detected by using Super Signal ®West Pico Chemiluminescent Substrate (Pierce, Rockford, IL).

Bands were semi-quantified by reverse image scanning densitometry with PhotoShop 7.0 (Adobe, San Jose, CA). An area of the gel image that was devoid of signal was assigned to be the background value. Then each band of the protein representing P-gp from KHOS_R2 or U2OS_R2 treated with various concentrations of MDR1 siRNA nanoparticles was analyzed for density beyond the background level. To ensure that the loading of the protein was equal and differences were not being observed because of one specimen having more protein than another, the band corresponding to β-actin was determined for each protein. The density of the protein band was normalized to the β-actin band for the protein and the ratio of the P-gp was normalized by dividing by the ratio of the β-actin corresponding to each P-gp.

For comparison of the duration of MDR1 inhibition by either MDR1 siRNA with siPORT™ NeoFX™ Transfection Agent or MDR1 siRNA loaded nanoparticles, 1×10⁵ KHOS_R2 cells/well were incubated with each MDR1 siRNA combinations for 5 days. The expression of P-gp was determined by western blot analysis as described above.

The Vybrant™ multi-drug resistance assay kit (Invitrogen/Molecular Probes) was used to measure the drug efflux properties of the resistant cell lines. This assay utilizes the fluorogenic dye calcein acetoxymethyl ester (calcein AM) as a substrate for efflux activity of P-gp or other membrane pump ABC proteins. Calcein AM is taken up by cells and hydrolyzed by cytoplasmic esterases into fluorescent calcein. Calcein is well retained in the cytosol. However, multidrug resistant cells expressing high levels of P-gp rapidly extrude non-fluorescent calcein AM from the plasma membrane, reducing accumulation of fluorescent calcein in the cytosol. Drug resistant cells (1×10⁵) were cultured in 96-well plates with either increasing concentrations of MDR1 siRNA loaded nanoparticles or with media alone. After 48 hours, they were incubated with 0.25 µM calcein AM in 150 µl total volume. Ten µM Verapamil was used as a positive control. PBS and nanoparticles loaded with 100 nM nonspecific siRNA were used as negative controls. After 30 minutes, the cells were washed and centrifuged twice with 200 µl cold RPMI1640 culture medium, and cell fluorescence was measured at a wavelength of 490 nm (A₄₉₀) on a SPECTRAmax® Microplate Spectrofluorometer (Molecular Devices). For visualization of the intracellular retention of calcein AM, images were acquired by Nikon Eclipse Ti-U fluorescence microscope (Nikon Corp.) equipped with a SPOT RT digital camera (Diagnostic Instruments, Inc., Sterling Heights, MI).

For cellular uptake studies, KHOS and KHOS_R2 cells were seeded at densities of 5×10⁵ cells/well in 6 well plates. MDR1 siRNA was applied to a well of KHOS_R2 and incubated for 48 h. Following the incubation, doxorubicin was added to each well and was incubated for additional 3 hours. After incubation, the cells were washed, suspended in fresh RPMI 1640, then visualized on a Nikon Eclipse Ti-U fluorescence microscope (Nikon Corp.) equipped with a SPOT RT digital camera (Diagnostic Instruments, Inc., Sterling Heights, MI). Fluorescence intensity and cellular localization was analyzed at a wavelength of 488 nm in triplicate in random, different fields.

In vitro cytotoxity assays were performed by MTT assay as previously described. Briefly, 3×10 ³ cells per well were plated in a 96-well plate during this process. After 48 hours of incubation with MDR1 siRNA loaded nanocarriers, nonspecific siRNA loaded nanocarriers or with medium alone, increasing concentrations of doxorubicin were applied. After culturing for 5 days, 10 µl of MTT (5 mg/ml in PBS) was added to each well and incubated for 3 hr. After dissolving the resulting formazan product with acid isopropnol, the absorbance (A490) was read on a SPECTRA max Microplate Spectrophotometer (Molecular Devices) at a wavelength of 490 nm. Each experiment was performed in triplicate.

Student's t-test was used to compare the differences between groups (GraphPad PRISM® 4 software, GraphPad Software, San Diego, CA). Results are given as mean ±SD and results with p<0.05 were considered statistically significant.

We have recently developed a self assembled nanosystem based on dextran nanoparticles to stably encapsulate doxorubicin [18]. We used a similar sequential addition method with some modifications to encapsulate MDR1 siRNA. The seqential addition method resulted in high encapsulation efficiency of >95% into the nanoparticles as observed from a FITC-labeled scrambled sequence siRNA encapslation and release study (data not shown). The mean particle size of the MDR1 siRNA loaded nanoparticles as determined by dynamic light scattering (DLS) measurement was 104.4±3.7 nm and the zeta potential was almost neutral (−0.19±1.13 mV), indicative of the entrapment of the siRNA within the interpenetrating dextran hydrogel network. Further, the particles were stable at room temperature or at 4°C and there was minimal change in the particle size upon storage (for 1 week at 4°C) as determined by DLS measurements.

To assess the transfection efficacy of siRNA loaded nanoparticles on cell lines, we first utilized EGFP expressing BHK-21 cells to test the effects. The EGFP siRNA loaded nanoparticles were non-toxic at the concentrations utilized in this study (Fig. S2). The EGFP siRNA was efficiently incorporated into cells and effectively inhibited the expression of EGFP in a dose dependent manner, but reached a plateau at approximately 100 nM (Fig. 2).

Western blotting was performed to estimate the effect of MDR1 siRNA loaded nanoparticle on P-gp expression. P-gp expression has been previously confirmed in the two drug resistant cell lines KHOS_R2 and U-2OS_R2 (data not shown). MDR1 siRNA loaded nanoparticle inhibited the expression of P-gp at a concentration of as low as 30 nM (Fig. 3A, B). siRNA transfected with siPORT™ NeoFX™ Transfection Agent was able to suppress P-gp expression for 48 hours (Fig. 4A). siRNA loaded nanoparticles were slower in achieving the suppression of P-gp, but were able to maintain suppression for 96 hours (Fig. 4B).

Reversal of MDR is usually manifested as an increased intracellular accumulation of chemotherapeutics, which can be achieved by disturbing P-gp-mediated drug uptake and efflux. Therefore, the effect of MDR1 siRNA loaded nanoparticles was examined for the uptake and efflux of a P-gp substrate, calcein AM, in KHOS_R2. Cells treated with MDR1 siRNA were shown to decrease calcein AM efflux in a dose-dependent manner as determined by image analysis, and confirmed by microplate spectrofluorometer analysis (Fig. 5).

Using fluorescent microscopy, subcellular distribution of doxorubicin in KHOS and KHOS_R2 was analyzed. After a 3 hour incubation period with free doxorubicin in drug resistant osteosarcoma cells, the drug was primarily concentrated in the cytoplasm with a very low level of fluorescence observed in the nucleus (Fig. 6A). When doxorubicin was administered after treatment with MDR1 siRNA loaded nanoparticle to drug resistant cell lines, an increase in fluorescence was observed in the nucleus and cytoplasm (Fig. 6B). This subcellular distribution mimicked that of the drug sensitive variant when treated with doxorubicin (Fig. 6C).

The dextran nanoparticles are not cytotoxic by themselves at the dose used in this study (data not shown). After treatment with MDR1 siRNA loaded nanoparticles, we found that doxorubicin showed an increased amount of anti-proliferative activity in drug resistant osteosarcoma cell lines in a dose dependent manner (Fig. 7A, B). With the delivery of nanoparticles loaded with 100 nM MDR1 siRNA, growth inhibition with doxorubicin was substantially more marked than with the administration of 100 fold higher amounts of free drugs. For example, in the KHOS_R2 cell line the IC₅₀ of doxorubicin alone was 10 µM which was reduced to 0.1 µM when the cells were co-treated with MDR1 siRNA loaded nanoparticles. Likewise, the drug resistant cell line U-2OS_R2 displayed a similar trend; the IC₅₀ for doxorubicin alone was 6 µM which was reduced to 0.06 µM when co-treated with the MDR1 siRNA loaded nanoparticles.