Inhibition of murine breast cancer growth and metastasis by survivin-targeted siRNA using disulfide cross-linked linear PEI

European Journal of Pharmaceutical Sciences 82 (2016) 171–182

Contents lists available at ScienceDirect

European Journal of Pharmaceutical Sciences journal homepage: www.elsevier.com/locate/ejps

Inhibition of murine breast cancer growth and metastasis by survivin-targeted siRNA using disulfide cross-linked linear PEI Shan Liu a,c,1, Wei Huang a,b,1, Ming-Ji Jin a,b, Bo Fan a,b, Gui-Min Xia c, Zhong-Gao Gao a,b,⁎ a State Key Laboratory of Bioactive Substance and Function of Natural Medicines, Institute of Materia Medica, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing 100050, PR China b Department of Pharmaceutics, Institute of Materia Medica, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing 100050, PR China c Department of Pharmaceutics, Institute of Medicinal Biotechnology, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing 100050, PR China

a r t i c l e

i n f o

Article history: Received 14 July 2015 Received in revised form 11 October 2015 Accepted 6 November 2015 Available online 7 November 2015 Keywords: Polyethylenimine Disulfide bond Survivin Breast cancer SiRNA

a b s t r a c t Biodegradable disulfide-containing polyethyleneimine (PEI) derivatives showed great potential as siRNA vectors for the treatment of cancer due to the reduction-sensitive property. In this study, we developed and characterized a hyperbranched disulfide cross-linked PEI (lPEI-SS) based on linear PEI (lPEI) by ring-opening reaction of propylene sulfide. We evaluated the efficiency of lPEI-SS as a siRNA vector in vitro with luciferase reporter gene system, and investigated the anti-tumor efficacy of survivin-targeted siRNA (siRNAsur) on 4T1 murine breast cancer model using lPEI-SS synthesized here. Results from cytotoxicity and hemolysis assay proved that lPEI-SS showed favorable cell and blood compatibility. lPEI-SS/siRNA polyplexes prepared under the optimized condition were compact spherical particles with the average size of 229.0 nm and zeta potential of 42.67 mV. Cellular uptake of lPEI-SS/siRNA polyplexes was significantly improved due to the higher branching degree of lPEI-SS over the parent lPEI. lPEI-SS/siRNAsur exhibited great anti-proliferation effect on 4T1 cell line, which was found to be caused by the induction of apoptosis. Most importantly, results of tumor volume, tumor weight and histological observation demonstrated that lPEI-SS/siRNAsur polyplexes effectively inhibited the tumor growth and metastasis of 4T1 murine breast cancer model. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Breast carcinoma is the leading cause of cancer death in women, and the second mostly diagnosed cancer after lung cancer(Jemal et al., 2011). Currently, using gene therapy to treat breast cancer has aroused great interest and been proven to be quite effective. Survivin is a member of the inhibitors of apoptosis protein (IAP) family, and is highly overexpressed in breast cancer (Jha et al., 2012). Moreover, the overexpression of survivin has been proved to play an important role in the diagnosis, treatment and prognosis of breast cancer(Lv et al., 2010). Therefore, survivin has been considered as a potential pharmacological target for breast cancer therapy. Several reports have shown that targeting survivin could effectively inhibit the growth as well as metastasis of breast cancer (James et al., 2014; Li et al., 2013; Peng et al., 2008b; Yang et al., 2013; Yin et al., 2013), which is a promising strategy for breast cancer therapy. Developing non-viral gene vectors able to transfer plasmid DNA or siRNA into cells safely and efficiently is the essence of gene therapy (O'Connor and Glynn, 2010). Among the various non-viral vectors, polyethyleneimine (PEI) has an obvious advantage over the other ⁎ Corresponding author at: 1 Xian Nong Tan Street, 100050 Beijing, PR China. E-mail address: [email protected] (Z.-G. Gao). 1 These authors contributed equally to this work.

http://dx.doi.org/10.1016/j.ejps.2015.11.009 0928-0987/© 2015 Elsevier B.V. All rights reserved.

cationic polymers due to its high charge density and so called “proton sponge effect”. However, the high efficiency of PEI is usually accompanied with high toxicity which hinders its clinical application, and the high toxicity of PEI was proved to be partially due to nondegradability (Luten et al., 2008). Hence, many biodegradable PEI derivatives based on low molecular weight (LMW) PEI have been developed to acquire the high efficiency and limit the cytotoxicity as well. Among the many degradable linkages such as ester(Luu et al., 2012; Park et al., 2012), amide (Xiong et al., 2007), ketal (Shim and Kwon, 2008), and disulfide (Bauhuber et al., 2012), disulfide is of great superiority to fabricate bioreducible PEI derivatives for DNA or siRNA delivery, because (1) the bond strength of disulfide is appropriate, (2) degradation kinetics of disulfide is quick, and most importantly, (3) disulfide is reduction-sensitive. Different concentrations of glutathione (GSH) between the intracellular (3–10 mM) and extracellular regions (∼2.8 μM) lead to the extra-and intracellular redox potential gradient (Ottaviano et al., 2008), thus disulfide could utilize such redox potential gradient to tightly condense DNA or siRNA outside the cells and easily release them inside the cells, which solves the contradiction between tight polyplexes formation outside cells and the easy dissociation of polyplexes inside cells(Wang et al., 2006). Therefore, integrating disulfide bond into PEI polymers is supposed to provide not only biodegradable PEI derivatives with low toxicity but also favorable reductionsensitive polymers.

172

S. Liu et al. / European Journal of Pharmaceutical Sciences 82 (2016) 171–182

Up to date, many reducible disulfide-containing PEI derivatives have been reported, and their design, development and application as nucleic acid vector have been extensively reviewed (Cho, 2012; Ryu and Kim, 2014; Son et al., 2012). There have been two approaches to introduce disulfide into polymers(Son et al., 2012), one is using disulfide-containing cross-linkers, the other is processed by initial prethiolation of polymers (polymer-SH) and subsequent oxidation of the thiolated polymers (polymer-SS). Because prethiolation strategy could facilitate further modification of polymer-SH with thiol-containing bioactive moieties such as targeting peptides (Son et al., 2010, 2011), it was more popular to fabricate disulfide-containing polymers. The prethiolation method has been used to synthesize both hyperbranched PEI-SS based on branched LMW PEI (M.W.: 800 (Xia et al., 2012); 1200 (Son et al., 2010, 2011); 1800 (Liu et al., 2010); 2000 (Xiao et al., 2013b)) and linear PEI-SS based on linear LMW PEI (M.W.: 2600 (Breunig et al., 2007, 2008); 3100(Breunig et al., 2007); 4600 (Breunig et al., 2007)). So far, however, branched PEI-SS based on linear LMW PEI via the prethiolation method has not been reported yet. It is generally considered that branched PEI (bPEI) has more advantage over linear PEI (lPEI) of the same molecular weight in binding with nucleic acid and enhancing cellular uptake (Breunig et al., 2008); whereas lPEI is more efficient than bPEI in releasing/dissecting nucleic acid. Therefore, we assumed that constructing branched PEI-SS by cross-linking linear PEI might be able to integrate the respective advantage of bPEI and lPEI and address the contradiction between extracellular tight binding and intracellular easy dissection of nucleic acid (Wang et al., 2006). Therefore, in this study, we synthesized branched disulfide-bonded PEI (lPEI2200-SS) based on linear LMW PEI (M.W.: 2200), studied the efficiency of lPEI2200-SS as a siRNA vector, and evaluated the anti-tumor efficacy of survivin-targeted siRNA loaded with the bioreducible PEI developed here. 2. Materials and methods 2.1. Materials 5 kDa poly (2-ethyl-2-oxazoline) s (PEOZs) was from Alfa Aesar (Tianjing, China). Propylene sulfide, N-(2-hydroxyethyl) piperazine-N ′-ethanesulfonic acid (HEPES), 5-diphenyltetrazolium bromide (MTT), ethidium bromide (EtBr), and dithiothreitol (DTT) were from SigmaAldrich (St. Louis, MO, USA). Annexin V-FITC Apoptosis detection kit was provided by KeyGEN Biosciences Company (Nanjing, China). Luciferase assay kit was from Promega Corporation (Madison, WI, USA). Trypsin, fetal bovine serum (FBS), DMEM and RPMI 1640 medium were all provided by Hyclone Company (Logan, Utah, USA). FAMlabeled siRNA was from Shanghai GenePharma Co., Ltd. (Shanghai, China). 2.2. Synthesis of siRNA Two kinds of siRNAs which target luciferase and survivin gene respectively, namely siRNAluc and siRNAsur, were synthesized by Shanghai GenePharma Co., Ltd. (Shanghai, China). The sequences of siRNAluc and siRNAsur were as follows: siRNAluc: 5′-CUUACGCUGAGUACUUCGATT-3′. siRNAsur: 5′-GAACAUCAUCAUCCAGGAC-3′. 2.3. Synthesis and structural confirmation of lPEI2200 and lPEI2200-SS Linear PEI with a molecular weight of 2200 (lPEI2200) was synthesized from commercial poly (2-ethyl-2-oxazoline)s (PEOZs) (M.W. = 5 kDa) by acid-catalyzed hydrolysis, as reported previously (Thomas et al., 2005). lPEI2200-SS was synthesized following previous method(Peng et al., 2008a) with some modifications. Typically, 0.45 mmol lPEI2200 obtained by the aforementioned method was

dissolved in the mixture of methanol (10 mL) and triethylamine (2 mL) in a flask, and the flask was purged with nitrogen for three times. Then 20 mL of methanol containing propylene sulfide (2.25 mmol) was added by syringe. The mixture was kept in 60 °C oil bath for 24 h while stirring. The resulted solution was evaporated to dryness under reduced pressure, dissolved in 30 mL DMSO, kept in dark and stirred for 48 h at room temperature (r.t.). Finally, lPEI2200-SS was obtained by dialyzing the mixture against distilled water (MWCO 1000) and lyophilization. lPEI2200-SS was structurally confirmed by 1H NMR (D2O, 400 MHz).

2.4. Cell line and cell culture Mouse mammary tumor cell line 4T1 and human breast adenocarcinoma cell line MCF-7, which were labeled with luciferase reporter gene and can stably express the firefly luciferase, were obtained from the Department of Pathology in Institute of Medicinal Biotechnology in Peking Union Medical College. MCF-7-luc and 4T1-luc cells were cultured in humidified atmosphere containing 5% CO2 at 37 °C in DMEM and RPMI 1640 medium containing 10% FBS, respectively. Cells in logarithmic growth phase were used to conduct all cell experiments in this study.

2.5. Cytotoxicity of lPEI2200-SS Cytotoxicity of lPEI2200-SS polymers was evaluated by MTT assay with both 4T1-luc and MCF-7-luc cells. Briefly, 4T1-luc and MCF-7-luc cells were seeded onto 96-well plates at a density of 4 × 103/well and 6 × 103/well, respectively. After incubated overnight for adherence, the previous medium was replaced with FBS-supplemented medium containing various concentrations of lPEI2200-SS or lPEI2200. After 4 h of incubation, the medium was removed and 100 μL fresh medium containing 0.5 mg/mL MTT was added to each well. Cells were further incubated for 4 h, and medium was replaced with 150 μL DMSO to solubilize the converted formazan. Absorbance was measured at 570 nm with a microplate photometer (Molecular Devices, USA). Cells without exposure to the polymers were used to represent 100% cell viability. Experiments were performed in triplicate.

2.6. pH titration pH titrations were performed as reported before(Thomas et al., 2005). Shortly, 2 mL solution of lPEI2200-SS or lPEI2200 (0.4 mg/mL) was adjusted to pH 12 with NaOH. Sequential 5 μL of 1 M HCl was added, and the pH after each addition was measured with a pH meter till the pH was reached the value of 2.5.

2.7. Hemolysis assay Rat erythrocytes were collected from heparin-treated blood, washed with 0.01 M isotonic PBS (pH 7.4) for 4 times (700 g for 10 min at 4 °C). The erythrocyte pellet was diluted 10-fold with 0.01 M isotonic PBS (pH 7.4) to a concentration of 109/mL. A 75 μL aliquot of erythrocyte suspension was added to a 96-well plate containing 75 μL serial solutions of lPEI2200 or lPEI2200-SS. The final concentrations of lPEI2200 and lPEI2200-SS varied from 6 to 25 μg/mL. After incubated for 1 h at 37 °C with constant shaking, unlysed erythrocytes were removed by centrifugation (700 g for 10 min at 4 °C), then 80 μL supernatant was transferred to a new 96-well plate and hemoglobin absorption was measured at 450 nm using SpectraMax 190 Absorbance Microplate Reader (Molecular Devices, USA). 1% Triton-X 100 was used as the positive control (100% lysis). Experiments were performed in triplicate.

S. Liu et al. / European Journal of Pharmaceutical Sciences 82 (2016) 171–182

2.8. Preparation and characterization of lPEI2200-SS/siRNA polyplexes To prepare lPEI2200-SS/siRNA polyplexes at different polymer/ siRNA (w/w) ratios, 1 mg/mL lPEI2200-SS was diluted with HBG buffer (HBG: 20 mM Hepes, pH 7.4, 5 wt.% glucose) to certain concentration, siRNA stock solution was diluted with RNase-free water to a concentration of 1 pmol/μL, then lPEI2200-SS/siRNA polyplexes were prepared by mixing the diluted lPEI2200-SS and siRNA solutions at equal volume, vortex for 5 s, and incubation at r.t. for 20 min. Size distribution and zeta potential of lPEI2200-SS/siRNA polyplexes after appropriate dilution were measured by Nicomp380/ZLS analyzer (Particle Sizing System, USA). The morphology of polyplexes was visualized by transmission electron microscopy (TEM). Briefly, 10 μL of the diluted polyplexes solution was dropped onto a copper grid coated with carbon membrane, and then copper grid was air-dried and negatively stained with 1% sodium phosphotungstate. Subsequently, the morphology of polyplexes was observed using a transmission electron microscope (Hitachi H-7650, Japan) with an acceleration voltage of 80 kV. 2.9. Gel retardation assay Agarose gel retardation assays were conducted with 4% agarose gel electrophoresis. Varying concentrations of lPEI2200-SS and an equal volume (10 μL) of siRNA solution (concentration kept at 100 ng/μL) were mixed in HBG buffer, incubated at r.t. for 20 min, then mixed with 4 μL of 6 × loading buffer, loaded into 4.0% agarose gel, and run for 20 min at 120 V. After staining the agarose gel with 0.5 μg/mL ethidium bromide for 30 min, the plasmid DNA bands were visualized with an UV gel image system (SIM135A, SIMON). Similarly, 10 μL of lPEI2200-SS/siRNA polyplexes prepared at weight ratio of 7.5:1 was mixed respectively with 10 μL HBG buffer containing different concentrations of DTT, incubated at r.t. for 1.5 h, mixed with loading buffer,

173

then subjected to gel electrophoresis and visualized with the above method. The amount of siRNA in each lane was 1 μg, naked siRNA was used as control. 2.10. Cellular uptake studies 4T1-luc cells were seeded onto a 12-well plate and incubated overnight for attachment. Then 4T1 cells were rinsed thrice with serumlacked medium, and incubated with polyplexes prepared with FAMlabeled siRNA for 4 h at 37 °C in serum-free medium. At the determined time point, cells were washed thrice with PBS, detached by trypsin, centrifuged for 5 min, suspended in 0.5 mL PBS buffer, and analyzed with FACSCaliber flow cytometer (Becton Dickinson, Franklin Lake, NJ, USA). 2.11. Luciferase gene silencing with lPEI2200-SS/siRNAluc polyplexes 4T1-luc cells were seeded into a 96-well plate at a density of 4 × 103/well. After 24-h incubation, the previous medium was replaced with serum-free medium containing various concentrations of polyplexes prepared at different weight ratios and in different medium such as HBG buffer, RPMI 1640, 10% FBS-containing RPMI 1640 and HBS buffer ((10 mM HEPES, 150 mM NaCl, pH 7.4)), respectively. lPEI2200/siRNAluc polyplexes were taken as the negative control, lipofectamine 2000/siRNA luc lipoplexes was chosen as the positive control. After 4 h of incubation, the medium was removed and 200 μL fresh medium containing 10% FBS was added to each well. After another 20 h or 44 h culture, the medium was removed and the cells were washed twice with PBS. Then cells were treated according to the protocol of luciferase assay kit from Promega Corporation (Madison, WI, USA), and the luciferase activity of each well was measured by GLOMAX 20/20 Luminometer (Promega, Madison, WI, USA). Experiments were performed in triplicate.

Fig. 1. Synthesis route of lPEI2200 (A) and lPEI2200-SS (B).

174

S. Liu et al. / European Journal of Pharmaceutical Sciences 82 (2016) 171–182

Fig. 2. 1H NMR spectra of lPEI2200 (A) and lPEI2200-SS (B) at 400 MHz in D2O.

2.12. Anti-proliferation assay CCK-8 assay was used to evaluate the anti-proliferation efficacy of lPEI2200-SS/siRNAsur polyplexes on 4T1-luc cells. 4T1-luc cells were seeded into a 96-well plate at a density of 4 × 103/well and cultured 24 h for adherence. Then the medium was replaced with 180 μL fresh

serum-free medium, and 20 μL of lPEI2200-SS/siRNAsur polyplexes solution was added to each well. Naked siRNAsur and lPEI2200/siRNAsur polyplexes were used as the controls. The amount of siRNAsur in each well was 5 pmol. After 4 h incubation, the medium was substituted with medium containing 10% FBS and cultured for 44 h. Then the cell viability of each well was calculated based on the optical density (OD)

Fig. 3. Characterization of lPEI2200-SS. (A) pH titration profiles of lPEI2200 and lPEI2200-SS. (B) Cytotoxicity of lPEI2200-SS and lPEI2200 on 4T1-luc cell lines, n = 3. (C) Cytotoxicity of lPEI2200-SS and lPEI2200 on MCF-7-luc cell lines, n = 3. (D) Hemolysis assay of lPEI2200-SS and lPEI2200 at pH 7.4, n = 3. Rat erythrocytes were incubated with various concentrations of polymers at 37 °C for 1 h while stirring, and the released amount of hemoglobin was measured to represent lysis efficiency.

S. Liu et al. / European Journal of Pharmaceutical Sciences 82 (2016) 171–182

175

value at 450 nm detected with CCK-8 assay kit (Dojindo, Japan) according to the standard protocol. The viability of cells without any treatment was set as 100%.

3. Results and discussion

2.13. Wound healing assay

Disulfide bond, due to its relatively high bond strength and readily quick cleavage under reductive conditions, is a popular degradable moiety to fabricate environment-sensitive non-viral gene vectors. Herein, we utilized the prethiolation strategy to construct hyperbranched lPEI2200-SS based on linear LMW PEI (M.W.: 2200) with propylene sulfide (Fig. 1), which was a bit different from the previously reported method (Lee et al., 2007). In the previously reported study (Lee et al., 2007), linear PEI-SS polymer was synthesized by prethiolating the terminal hydroxyl group of linear PEI, thus generating a linear PEI-SS. Whereas we fabricated hyperbranched PEI-SS based on linear LMW PEI (M.W.: 2200) by prethiolation of the middle secondary amines of

Wound healing assay was conducted to investigate the antiproliferation efficacy of lPEI2200-SS/siRNAsur polyplexes on 4T1-luc cells. Briefly, 4T1-luc cells were seeded into a 24-well plate at a density of 1 × 105/well and cultured for 24 h. Then one linear scratch wound was made with a 10 μL pipette tip in each well. Subsequently, each well was washed with serum-free medium to remove cell debris, and added with 480 μL serum-free medium. Then 20 μL of lPEI2200-SS/ siRNAsur polyplexes was added to each tested well, and wells that added with lPEI2200/siRNAsur polyplexes were taken as the control. The amount of siRNAsur in each well was 20 pmol. After 4 h incubation, medium in each well was replaced with fresh medium containing 10% FBS. The subsequent proliferation of cells was observed with an inverted microscope.

3.1. Synthesis and characteristics of lPEI2200-SS

2.14. Apoptosis assay 4T1-luc cells were seeded onto a 12-well plate at a density of 1 × 105 cells/well and cultured overnight for attachment. Then the previous medium was replaced with fresh medium, and 40 μL of lPEI2200-SS/siRNAsur polyplexes was added to each tested well, wells that added with lPEI2200/siRNAsur polyplexes were taken as the control. The amount of siRNAsur in each well was 40 pmol. After 4 h incubation, medium in each well was replaced with fresh medium containing 10% FBS. After cultured for 20 h, cells were washed twice with warm PBS, detached by trypsin without EDTA, collected, centrifuged, suspended in the binding buffer and further stained with PI and Annexin V-FITC for 15 min at r.t. in the dark. The apoptosis of cells was analyzed by FACSCaliber flow cytometer equipped with the CellQuest software (Becton Dickinson, Franklin Lake, NJ, USA). Cells incubated only with medium were set as the control. 2.15. In vivo anti-tumor effects of lPEI2200-SS/siRNAsur polyplexes Animal experiment was adhered to the Principles of Laboratory Animal Ethics Committee in the Institute of Materia Medica in Peking Union Medical College, and conducted with necessary humane care. Malignant breast tumor model was established by inoculating 1 × 105 4T1-luc cells at the fourth mammary fat pad of the 6–8 week old female BALB/c mice. A week after the inoculation, tumor volumes were monitored with a dial caliper every other day. When the average tumor volume reached about 230 mm3 (on the 16th day after inoculation), mice were randomly divided into three groups (n = 4): HBG, naked siRNA and lPEI2200-SS/siRNAsur polyplexes. Mice of each group were intratumorally injected with either naked siRNA or lPEI2200-SS/ siRNAsur polyplexes at a dose of 0.3 mg siRNAsur/kg bodyweight on alternate days for a total of 5 times, and mice of HBG group were administered with HBG of the equal volume. On the 4th day after the final administration, mice were sacrificed, lung and liver tissues were fixed in Bouin's solution to observe the tumor metastasis. Also, tumor tissues and livers were fixed with 4% neutral formalin to perform the histopathological examination. 2.16. Statistical analysis Results were expressed as mean ± S.D. of at least three independent experiments. Student's t test was used to evaluate the difference between two groups. One-way analysis of variance (ANOVA) with Bonferroni's post hoc test was used to compare the statistical significance among the groups, and P less than 0.05 is considered significant.

Fig. 4. Luciferase gene silencing by siRNAluc delivered into 4T1-luc cells using lPEI2200-SS/ siRNAluc polyplexes. (A) Effect of weight ratio (polymer to siRNA) on luciferase gene silencing efficiency, all polyplexes were prepared in HBG buffer, n = 3, *P ﹤ 0.05 vs. weight ratio of 1:1, ***P ﹤ 0.001 vs. weight ratio of 1:1. (B) Effect of complexing medium on luciferase gene silencing efficiency, all polyplexes were prepared at the same weight ratio of 5:1, n = 3, **P ﹤ 0.01 vs. 10% FBS. (C) Luciferase gene silencing efficiency of lPEI2200-SS/ siRNAluc prepared under the optimal condition. Lipofectamine2000 (lipo2000) and lPEI2200 were used as controls. Cells without any treatment were used as blank. Luciferase gene silencing efficiency was represented by the relative luciferase activity compared to that of blank (expressed as 100). n = 3, **P ﹤ 0.01 vs. lipo2000, ***P ﹤ 0.001 vs. lipo2000.

176

S. Liu et al. / European Journal of Pharmaceutical Sciences 82 (2016) 171–182

lPEI, and obtained PEI-SS with a higher branching degree, which was supposed to be a more superior siRNA vector. We synthesized linear LMW PEI with a molecular weight of 2200 (lPEI2200) by previously reported acid-catalyzed hydrolysis method (Thomas et al., 2005). The structure of lPEI2200 was confirmed by 1H NMR in D2O, The presence of the characteristic singlet at 3.54 ppm 1 (− CH2–CH2–NH+ 2 –) in the H NMR spectrum of lPEI2200 indicated the successful synthesis of lPEI2200 (Thomas et al., 2005) (Fig. 2A). Since lPEI2200 was synthesized by acid-catalyzed hydrolysis method, and the obtained product (lPEI2200) was directly subjected to NMR analysis after air-dry without adjusting the pH of the product to 7.0, so lPEI2200 that used for NMR analysis was largely protonated, thus leading to a shift of resonance signal from high field (~2.8 ppm, –CH2– CH2–NH-) to a relatively low field (3.54 ppm, –CH2–CH2–NH+ 2 -). lPEI2200-SS was synthesized by initial prethiolation of the secondary amines via the ring-opening reaction of propylene sulfide and the following oxidation of -SH in DMSO. The resulted lPEI2200-SS was structurally confirmed with 1H NMR in D2O. In the 1H NMR spectrum of lPEI2200-SS (Fig. 2B), the chemical shifts (δ) of 3.30–2.55 were corresponded to –NCH2CH2N- and –NCH2CHS-, and chemical shifts (δ) of 1.2–1.5 were caused by –CH3, indicating the successful synthesis of lPEI2200-SS. In this study, the reacted molar ratio of lPEI2200 and propylene sulfide was 5:1, and the thiolation degree of the resulted lPEI2200-SH was determined to be 4.4% (quantity ratio of -SH group to lPEI2200) by Ellman's method(Ellman, 1959). It is well known that there were only secondary amine groups in linear PEI. As depicted in the synthetic scheme (Fig. 1), cross-linking lPEI2200 with propylene disulfide would turn partial secondary amines (pKa around 8) into tertiary amines (pKa around 6–7) (Godbey et al., 1999; Wang et al., 2006), so the total pKa slightly changed after the cross-linking. Additionally, it has been proved that cross-linking PEI with propylene sulfide would not change the acid–base property of PEI significantly(Peng et al., 2008a). Because the molecular weight of propylene sulfide (M.W.: 74.14) was quite low, after cross-linking, the

mass fraction of the parent LMW PEI in PEI-SS was nearly 100%, so PEI-SS showed similar acid–base property to the parent LMW PEI. Based on the above information, it was assumed that the buffering capacity of lPEI2200-SS would be quite similar to that of lPEI2200. Such assumption was proved by the result of pH titration (Fig. 3A). As depicted in Fig. 3A, the pH titration profile of lPEI2200-SS was almost the same as that of lPEI2200, both lPEI2200 and lPEI2200-SS maintained a moderate buffering capacity in the physiological pH range of 5.1–7.4. 3.2. Cytotoxicity and blood compatibility of lPEI2200-SS Cytotoxicity is one of the major concerns for developing non-viral gene vectors. High molecular weight (HMW) PEIs show high nucleic acid delivery efficacy, but they cannot be metabolized in vivo and would cause severe toxicity due to accumulation, thus hinders their clinical application. However, due to the intracellular reducing environment, it was assumed that lPEI2200-SS synthesized here could be degraded into LMW PEI which would undergo facile excretion from the body without inducing significant cytotoxicity(Son et al., 2012). Therefore, we investigated the cytotoxicity of lPEI2200-SS with 4T1-luc and MCF-7-luc cells by MTT method. As illustrated in Fig. 3B and C, the cytotoxicity of lPEI2200-SS was as low as lPEI2200. Under the tested concentrations ranging from 5 to 20 μg/mL, lPEI2200-SS exerted almost no cytotoxicity to either 4T1-luc or MCF-7-luc cells, both cell lines maintained nearly 90% viability. Blood compatibility is a prerequisite for the in vivo application of nucleic acid vectors. It has been proved that the high positive charge density of HMW PEI would destabilize negatively charged cell membranes and cause biomembrane lysis (Fischer et al., 2003), so systemic administration of HMW PEI would arouse lysis of erythrocyte and cause hemolysis. Since LMW PEI showed no membrane lytic activity (Kichler et al., 2001), we performed hemolysis assay to test whether disulfide cross-linked LMW PEI (lPEI2200-SS) still maintained the low membrane lytic activity. Various concentrations of lPEI2200-SS were

Fig. 5. Characterization of lPEI2200-SS/siRNA. (A) Gel retardation assay of lPEI2200-SS/siRNA prepared at different weight ratios in the absence of DTT or at weight ratio of 7.5:1 in the presence of varying concentrations of DTT. (B) Typical TEM image of lPEI2200-SS/siRNA polyplexes prepared at weight ratio of 7.5:1 in HBG buffer. (C) Cellular uptake of naked siRNA (black), lPEI2200/siRNA polyplexes (red), and lPEI2200-SS/siRNA polyplexes (yellow) by flow cytometry.

S. Liu et al. / European Journal of Pharmaceutical Sciences 82 (2016) 171–182

incubated with erythrocytes in neutral buffer (pH 7.4) for 1 h at 37 °C, and the released amount of hemoglobin was measured to represent lysis efficiency. As shown in Fig. 3D, lPEI2200 caused no membrane lysis, and lPEI2200-SS at concentrations below 25 μg/mL aroused less than 10% lysis of erythrocyte, which can be regarded as negligible. Results from the hemolysis assay indicated that lPEI2200-SS showed favorable blood compatibility and can be applied for intravenous administration. 3.3. Preparation and characterization of lPEI2200-SS/siRNA polyplexes Preparation process (weight ratio, complexing medium, etc.) is highly related to the properties and the in vitro and in vivo transfection efficiency of PEI/siRNA polyplexes. Using 4T1-luc cell lines which could stably express the luciferase gene, we screened the optimal preparation condition of lPEI2200-SS/siRNA polyplexes using luciferase gene silencing efficiency as the evaluation criterion. As depicted in Fig. 4A,

177

when the weight ratio of polymer to siRNA increased from 1:1 to 20:1, the gene silencing efficiency elevated accordingly. During the transfection, the final concentration of lPEI2200-SS at weight ratio of 20:1 was 6.68 μg/mL; lPEI2200-SS at such concentration showed no cytotoxicity to 4T1-luc cells as indicated by the cytotoxicity studies (Fig. 3B). So the reduced luciferase activity was definitely caused by gene silencing efficiency rather than loss of cell viability. Among the weight ratios from 1:1 to 20:1, the dramatic increase of gene silencing efficiency occurred within the range from 5:1 to 10:1, so the following polyplexes were prepared within this weight ratio range (5:1–10:1). Complexing medium is another important factor for the fabrication of polyplexes. It was proved that complexing medium with low ionic strength was beneficial to form polyplexes with small size and high stability (van Gaal et al., 2011). As shown in Fig. 4B, among the screened complexing medium, HBG which is a low ionic strength buffer was found to be the best medium for polyplexes fabrication, polyplexes prepared in 10% FBS-containing RPMI 1640 showed the lowest gene

Fig. 6. Anti-proliferation effects of lPEI2200-SS/siRNAsur polyplexes on 4T1-luc cells. (A) Cell viability of 4T1-luc cells after transfected with lPEI2200-SS/siRNAsur polyplexes for 48 h, n = 5. Naked siRNAsur and lPEI2200/siRNAsur were taken as the controls. Cells without any treatment were used as blank. *P ﹤ 0.05 vs. naked siRNA and lPEI2200/siRNAsur, respectively. (B) Wound healing assay of lPEI2200/siRNAsur (b1) and lPEI2200-SS/siRNAsur polyplexes (b2) on 4T1-luc cells. The healing situation of scratch wound was observed 24 h and 48 h after scratching using an inverted microscope. (C) Apoptosis assay of 4T1-luc cells induced by lPEI2200/siRNAsur and lPEI2200-SS/siRNAsur polyplexes. Cells without any treatment were used as blank.

178

S. Liu et al. / European Journal of Pharmaceutical Sciences 82 (2016) 171–182

silencing efficacy, which might be due to the instability caused by serum. Finally, 48-h gene silencing efficacy of lPEI2200-SS/siRNAluc polypexes prepared under the optimal condition (weight ratios from 5:1 to 10:1; complexed in HBG buffer) was evaluated in 4T1-luc cells, using lipofectamine and the parent lPEI2200 as the positive and negative controls, respectively (Fig. 4C). As shown in Fig. 4C, the gene silencing efficacy of lPEI2200-SS/siRNAluc at all three weight ratios were significantly higher than that of lipofectamine2000/siRNAluc lipoplexes (P ﹤ 0.01) and lPEI2200/siRNAluc polyplexes (P ﹤ 0.001). LPEI2200-SS/ siRNAluc polyplexes prepared at weight ratio of 7.5:1 silenced the expression of luciferase gene by 87.83%. Based on the above results, the following lPEI2200-SS/siRNA polyplexes were prepared at weight ratio of 7.5:1 in HBG buffer, which was found to be the optimal preparation condition for lPEI2200-SS/siRNA polyplexes. siRNA is a double-stranded RNA with a length of 20–25 bp, it is short in length and its charge density is quite low compared with that of plasmid DNA. As a result, it is relatively more difficult for cationic polymers to condense siRNA and form stable polyplexes. Therefore, many structural modifications have been employed to enhance the complexing capacity of polymers with siRNA. Disulfide cross-linking is one of the most popular modification strategies, because introducing disulfide into polymers could achieve not only the extracellular tight condensation but also the intracellular efficient dissociation of polyplexes due to extra- and intra-cellular redox potential gradient. Here we evaluated the complexing capacity of lPEI2200-SS with siRNA by gel retardation assay (Fig. 5A). As shown in Fig. 5A, lPEI2200-SS completely condensed siRNA at the weight ratio of 1:1. Moreover, when exposing lPEI2200-SS/ siRNA polyplexes (prepared at weight ratio of 7.5:1) to various concentrations of DDT for 1.5 h, siRNA could be sufficiently released out from the polyplexes, suggesting that disulfide bond in lPEI2200-SS could be cleaved inside the cells to facilitate the intracellular release of siRNA (Xia et al., 2012). The morphology of lPEI2200-SS/siRNA polyplexes was observed with transmission electron microscopy (TEM), the result was shown in Fig. 5B. In Fig. 5B, we can see that lPEI2200-SS/siRNA polyplexes displayed a compact spherical structure with the size of around 250 nm. Result from dynamic light scattering (DLS) indicated that the average size of lPEI2200-SS/siRNA polyplexes prepared at weight ratio of 7.5 was 229.0 nm, and the polydispersity index was 0.284. Zeta potential of lPEI2200-SS/siRNA polyplexes prepared at weight ratio of 7.5 was found to be around 42.67 mV by electrophoretic light scattering (ELS). The surficial positive charge would enhance the affinity of lPEI2200-SS/siRNA polyplexes with negatively charged cell membranes and facilitate the cellular entry of polyplexes. FAM-labeled siRNA was used to prepare polyplexes to study the cellular uptake of lPEI2200-SS/FAM-siRNA by flow cytometry, the results were shown in Fig. 5C. As indicated in Fig. 5C, in comparison with naked FAM-siRNA (black), both of the fluorescence intensity profiles of lPEI2200/FAMsiRNA (red) and lPEI2200-SS/FAM-siRNA (yellow) shifted to right, indicating that the cellular entry of naked siRNA itself was extremely inefficient. Among lPEI2200-SS/FAM-siRNA, lPEI2200/FAM-siRNA and naked FAM-siRNA, lPEI2200-SS/FAM-siRNA showed the highest cellular entry. Moreover, the right-shift of the fluorescence intensity peak of lPEI2200SS/FAM-siRNA was more significant than that of lPEI2200/FAM-siRNA, suggesting that cross-linking enhanced the cellular uptake of lPEI2200SS/FAM-siRNA. The reason of the enhanced cellular uptake of lPEI2200SS/FAM-siRNA was mainly that lPEI2200-SS possessed higher branch degree than lPEI2200, and the cellular uptake of siRNA was proved to improve with the increasing branch degree of polymer (Breunig et al., 2008).

cancer, and targeting survivin has been proved to be an effective strategy to treat breast cancer (Jha et al., 2012), in this study, we attempted to investigate the anti-proliferation effect of survivin-targeted siRNA (siRNAsur) on breast cancer cells using lPEI2200-SS as a vector. CCK-8 assay was performed to study the anti-proliferation effect of lPEI2200-SS/siRNAsur polyplexes on 4T1-luc mouse mammary tumor cells, the transfection was lasted for 48 h; naked siRNAsur and lPEI2200/siRNAsur were taken as the controls. As shown in Fig. 6A, the viability of cells transfected with lPEI2200-SS/siRNAsur polyplexes (33.67%) was significantly lower than that of naked siRNAsur (80.58%) and lPEI2200/siRNAsur polyplexes (68.76%) (P ﹤ 0.05), suggesting that siRNAsur when loaded with lPEI2200-SS could effectively inhibit the proliferation of 4T1-luc cells. Moreover, such high anti-proliferation efficiency of lPEI2200-SS/siRNAsur polyplexes was also confirmed by wound healing assay (Fig. 6B). As shown in Fig. 6B, in the lPEI2200/ siRNAsur group (b1), due to the proliferation of cells, the previous wound became narrow with the increase of time and was barely visible at the culture time point of 48 h. Whereas, in the lPEI2200-SS/siRNAsur group (b2), the previous wound showed no obvious change during the 48 h culture. Moreover, with the increment of culture time, cells

3.4. Anti-proliferation effects of lPEI2200-SS/siRNAsur polyplexes Based on the results of cytotoxicity assay and luciferase reporter gene silencing experiment, lPEI2200-SS has been proved to be a safe and highly efficient vector for siRNA delivery. Fatherly, we investigated the potential of lPEI2200-SS in delivering therapeutic siRNA for cancer therapy. Because survivin was found to be over-expressed in breast

Fig. 7. In vivo anti-tumor effect of lPEI2200-SS/siRNAsur polyplexes on 4T1-bearing breast cancer mouse models. (A) Tumor volume of mice treated with HBG buffer, naked siRNAsur, and lPEI2200-SS/siRNAsur polyplexes. (B) Tumor weight of mice treated with HBG buffer, naked siRNAsur, and lPEI2200-SS/siRNAsur polyplexes on the 4th day after 5 times' administration. (C) Representative picture of tumor tissues of mice treated with HBG buffer (c1), naked siRNAsur (c2), and lPEI2200-SS/siRNAsur polyplexes (c3) on the 4th day after 5 times' administration. *P ﹤ 0.05 vs. HBG buffer and naked siRNA, respectively.

S. Liu et al. / European Journal of Pharmaceutical Sciences 82 (2016) 171–182

on both sides of the linear wound became scarce, which was owing to the cell death caused by lPEI2200-SS/siRNAsur. Taken the results of CCK-8 and wound healing assay together, we concluded that lPEI2200-SS/siRNAsur exhibited great potential in inhibiting the proliferation of 4T1-luc cells. Since survivin is a member of IAP family, the anti-proliferation effect of siRNAsur probably results from induction of cell apoptosis. Such assumption was proved by the results of apoptosis assay (Fig. 6C). As illustrated in Fig. 6C, after 24 h transfection, 35.85% of cells in the lPEI2200-SS/siRNAsur group underwent apoptosis, and there were almost no apoptotic cells in the group of lPEI2200/siRNAsur. This result was in agreement with that of CCK-8 and wound healing assay. 3.5. In vivo anti-tumor effects of lPEI2200-SS/siRNAsur polyplexes Although lPEI2200-SS possessed high effectiveness for in vitro siRNA delivery, the in vivo siRNA delivery efficiency of lPEI2200-SS cannot be inferred accordingly. Since the successful in vivo application is the ultimate goal of developing non-viral gene vectors, we tested the in vivo siRNAsur delivery efficiency of lPEI2200-SS and studied the anti-tumor effect of lPEI2200-SS/siRNAsur polyplexes with breast cancer mouse models. 16 days after the inoculation of 4T1 cells, the average tumor volume was reached around 230 mm3. Then mice were randomly divided and administrated with HBG buffer, naked siRNAsur, and lPEI2200-SS/ siRNAsur, respectively, and the tumor volume was monitored every other day (Fig. 7A). As shown in Fig. 7A, tumor volume profiles of mice in HBG and naked siRNAsur groups were almost overlapped with each other, indicating that the anti-tumor effect of naked siRNAsur was negligible. Whereas, tumor volume of mice in lPEI2200-SS/siRNAsur group exhibited a gradual profile, suggesting that tumor growth of mice in lPEI2200-SS/siRNAsur group was quite slow and lPEI2200-SS/

179

siRNAsur effectively inhibited the growth of tumor. 4 days after the 5th administration, all mice were sacrificed and the dissected tumor tissues were weighted (Fig. 7B). In Fig. 7B, we can see that the dissected tumor tissues of HBG and naked siRNAsur groups weighted equally, while the average tumor weight of lPEI2200-SS/siRNAsur group was significantly lower than those of HBG and naked siRNAsur groups (P ﹤ 0.05). Difference of the tumor size among the three groups can also be vividly observed in Fig. 7C, which was in accordance with the results of tumor volume and tumor weight. Based on the above results, we concluded that lPEI2200-SS/siRNAsur showed great potential in inhibiting tumor growth of 4T1-bearing mouse models. As mammary carcinoma 4T1 cells were proved to be highly metastatic even at early stage (Gao et al., 2011), we studied the tumor metastasis of mice in the three groups besides monitoring the tumor growth. Since breast cancer primarily metastasizes to the lung, liver, bone, etc.(Minn et al., 2005; Weigelt et al., 2005), in this study, we focused mainly on the lung and liver metastasis aroused by primary breast cancer. After fixation with Bouin's solution, many white nodules were observed on the surficial lung lobes in groups of both HBG and naked siRNAsur (Fig. 8B and C), which implied that lung metastasis had occurred in mice of these two groups (Panigrahy et al., 2012). Whereas, the surface of lungs in lPEI2200-SS/siRNAsur group was smooth and showed no nodules (Fig. 8D), indicating that mice of lPEI2200-SS/ siRNAsur group were free from lung metastasis. Additionally, the fixed lung and liver tissues were performed with hematoxylin and eosin (H&E) staining for histopathological examination. Results of liver histopathological examination were displayed in Fig. 9. As shown in Fig. 9, many scattered hepatic metastatic focuses were observed in the liver tissues of both HBG (Fig. 9B) and naked siRNAsur (Fig. 9C) groups, and the liver specimen (Fig. 9D) from lPEI2200-SS/siRNAsur group was

Fig. 8. Typical lung tissue photos of 4T1-bearing mice respectively treated with HBG buffer (B), naked siRNAsur (C), and PEI2200-SS/siRNAsur polyplexes (D) on the 4th day after the final administration. Lung tissue of normal mouse was taken as control (A).

180

S. Liu et al. / European Journal of Pharmaceutical Sciences 82 (2016) 171–182

Fig. 9. Representative hematoxylin and eosin-stained liver sections of 4T1-bearing mice respectively treated with HBG buffer (B), naked siRNAsur (C), and PEI2200-SS/siRNAsur polyplexes (D) on the 4th day after the final administration. Liver tissue of normal mouse was taken as control (A).

similar with that of normal mice (Fig. 9A), showing only normal hepatic cells. Results of lung histopathological examination were displayed in Fig. 10. As depicted in Fig. 10, we can see that big metastatic focus appeared in the lung specimens of both HBG (Fig. 10B) and naked siRNAsur (Fig. 10C) groups, while the lung specimen from lPEI2200-SS/siRNAsur group was similar with that of normal mice (Fig. 10A), displaying only normal lung structure and alveolar septum (Fig. 10D). Results from liver and lung pathological examinations suggested that breast cancerbearing mice in HBG and naked siRNAsur groups underwent lung and liver metastasis, whereas mice in lPEI2200-SS/siRNAsur group were free from lung and liver metastasis, indicating that lPEI2200-SS/ siRNAsur could effectively inhibit the pulmonary and hepatic metastasis of 4T1-bearing mouse models. Recently, several PEI based bioreducible non-viral vectors have shown great promise for the treatment of diseases such as cancer, inflammation, myocardial infarction, and diabetes(Ryu and Kim, 2014). For example, biodegradable Tween 85-SS-bPEI (bPEI2000) (Xiao et al., 2013b), SS-bPEI (bPEI800) (Xia et al., 2012) and p(CBA-bPEI)-PEGMannose (bPEI800) (Xiao et al., 2013a) were all demonstrated to be promising non-viral siRNA vectors for the treatment of breast cancer, liver cancer, and inflammatory bowel disease, respectively. All these reported disulfide-bonded PEI polymers were synthesized by crosslinking PEI with disulfide-containing moieties such as N, N′-cystamine bisacrylamide (CBA) and 3, 3′-dithiodipropionic acid (DTPA), and the disulfides were located outside the parent PEI molecules. While lPEI2200-SS developed in this study was synthesized by initial prethiolation of the secondary amines of lPEI2200 (lPEI2200-SH) with propylene sulfide and the following oxidation, the resulted lPEI2200SS maintained a hyperbranched structure (Fig. 1) which would be beneficial for compacting siRNA. Depending on the ratio of lPEI2200 and

propylene sulfide, the branch degree of lPEI2200-SS was tunable, thus the binding and release of siRNA was controllable. Extracellularly, hyperbranched lPEI2200-SS would condense siRNA tightly forming compact spherical particles (Fig. 5B); while intracellularly, due to the reductive cytosolic enzymes, hyperbranched lPEI2200-SS would be cleaved into linear PEI which was supposed to release siRNA quite easily (Fig. 5A). Based on the above results and discussion, lPEI2200-SS developed here was favorable for siRNA delivery and different from the previously reported bioreducible PEI derivatives. Targeting survivin showed great potential in inhibiting the growth and metastasis of breast cancer (Abdraboh et al., 2011; Xu et al., 2012). Recently, it has been demonstrated that breast cancer with multidrug resistance (MDR) could even be reversed by simultaneous delivery of doxorubicin and survivin-targeted shRNA(Yin et al., 2013). In this study, we complexed siRNAsur with bioreducible lPEI200-SS and proved the successful application of lPEI200-SS/siRNAsur in inhibiting tumor growth and metastasis using a highly malignant 4T1 tumor-bearing mouse models. Results obtained here demonstrated the efficiency of disulfide-bonded linear PEI (lPEI2200-SS) as a siRNA vector for cancer gene therapy. 4. Conclusion In closing, based on linear LMW PEI (lPEI2200), we successfully developed a biodegradable disulfide-containing PEI polymer (lPEI2200SS) using ring-opening reaction of propylene sulfide, which proved to be highly efficient for siRNA delivery in vitro and in vivo. lPEI2200-SS showed low cytotoxicity and favorable blood compatibility, and could condense siRNA forming nano-sized compact polyplexes. Moreover, lPE2200-SS/siRNAsur polyplexes exerted significant anti-proliferation

S. Liu et al. / European Journal of Pharmaceutical Sciences 82 (2016) 171–182

181

Fig. 10. Typical hematoxylin and eosin-stained lung tissues of 4T1-bearing mice respectively treated with HBG buffer (B), naked siRNAsur (C), and PEI2200-SS/siRNAsur polyplexes (D) on the 4th day after the final administration. Lung tissue of normal mouse was taken as control (A).

effect on 4T1 cells, and possessed great inhibition of tumor growth and metastasis on 4T1 murine breast cancer models. Therefore, lPEI2200-SS developed in this study was a promising siRNA vector for cancer gene therapy. In the future, we would focus on the fabrication of multifunctional PEI polymers by introducing tumor-homing moieties into PEI-SS to enhance tumor targeting and broaden its application for the treatment of various diseases.

Acknowledgments This work was financially supported by Beijing Natural Science Foundation (7142114, 2141004), National Natural Science Foundation of China (81373342).

References Abdraboh, M.E., Gaur, R.L., Hollenbach, A.D., Sandquist, D., Raj, M.H., Ouhtit, A., 2011. Survivin is a novel target of CD44-promoted breast tumor invasion. Am. J. Pathol. 179, 555–563. Bauhuber, S., Liebl, R., Tomasetti, L., Rachel, R., Goepferich, A., Breunig, M., 2012. A library of strictly linear poly(ethylene glycol)-poly(ethylene imine) diblock copolymers to perform structure-function relationship of non-viral gene carriers. J. Control. Release 162, 446–455. Breunig, M., Lungwitz, U., Liebl, R., Goepferich, A., 2007. Breaking up the correlation between efficacy and toxicity for nonviral gene delivery. Proc. Natl. Acad. Sci. U. S. A. 104, 14454–14459. Breunig, M., Hozsa, C., Lungwitz, U., Watanabe, K., Umeda, I., Kato, H., Goepferich, A., 2008. Mechanistic investigation of poly(ethylene imine)-based siRNA delivery: disulfide bonds boost intracellular release of the cargo. J. Control. Release 130, 57–63. Cho, C.-S., 2012. Design and Development of Degradable Polyethylenimines for Delivery of DNA and Small Interfering RNA: An Updated Review. ISRN Materials Science. Ellman, G.L., 1959. Tissue sulfhydryl groups. Arch. Biochem. Biophys. 82, 70–77.

Fischer, D., Li, Y., Ahlemeyer, B., Krieglstein, J., Kissel, T., 2003. In vitro cytotoxicity testing of polycations: influence of polymer structure on cell viability and hemolysis. Biomaterials 24, 1121–1131. Gao, Z.G., Tian, L., Hu, J., Park, I.S., Bae, Y.H., 2011. Prevention of metastasis in a 4T1 murine breast cancer model by doxorubicin carried by folate conjugated pH sensitive polymeric micelles. J. Control. Release 152, 84–89. Godbey, W.T., Wu, K.K., Mikos, A.G., 1999. Size matters: molecular weight affects the efficiency of poly(ethylenimine) as a gene delivery vehicle. J. Biomed. Mater. Res. 45, 268–275. Jemal, A., Bray, F., Center, M.M., Ferlay, J., Ward, E., Forman, D., 2011. Global cancer statistics. CA Cancer J. Clin. 61, 69–90. James, H.P., Rijo, J., Anju, A., Anoop, K.R., 2014. Smart polymers for the controlled delivery of drugs-a concise overview. Acta Pharm. Sin. B. Jha, K., Shukla, M., Pandey, M., 2012. Survivin expression and targeting in breast cancer. Surg. Oncol. 21, 125–131. Kichler, A., Leborgne, C., Coeytaux, E., Danos, O., 2001. Polyethylenimine-mediated gene delivery: a mechanistic study. J. Gene Med. 3, 135–144. Lee, Y., Mo, H., Koo, H., Park, J.Y., Cho, M.Y., Jin, G.W., Park, J.S., 2007. Visualization of the degradation of a disulfide polymer, linear poly(ethylenimine sulfide), for gene delivery. Bioconjug. Chem. 18, 13–18. Li, Z., Yin, P.H., Yang, S.S., Li, Q.Y., Chang, T., Fang, L., Shi, L.X., Fang, G.E., 2013. Recombinant attenuated Salmonella typhimurium carrying a plasmid co-expressing ENDOVEGI151 and survivin siRNA inhibits the growth of breast cancer in vivo. Mol. Med. Rep. 7, 1215–1222. Liu, J., Jiang, X., Xu, L., Wang, X., Hennink, W.E., Zhuo, R., 2010. Novel reduction-responsive cross-linked polyethylenimine derivatives by click chemistry for nonviral gene delivery. Bioconjug. Chem. 21, 1827–1835. Luten, J., van Nostrum, C.F., De Smedt, S.C., Hennink, W.E., 2008. Biodegradable polymers as non-viral carriers for plasmid DNA delivery. J. Control. Release 126, 97–110. Luu, Q.P., Shin, J.Y., Kim, Y.K., Islam, M.A., Kang, S.K., Cho, M.H., Choi, Y.J., Cho, C.S., 2012. High gene transfer by the osmotic polysorbitol-mediated transporter through the selective caveolae endocytic pathway. Mol. Pharm. 9, 2206–2218. Lv, Y.G., Yu, F., Yao, Q., Chen, J.H., Wang, L., 2010. The role of survivin in diagnosis, prognosis and treatment of breast cancer. J. Thorac. Dis. 2, 100–110. Minn, A.J., Gupta, G.P., Siegel, P.M., Bos, P.D., Shu, W., Giri, D.D., Viale, A., Olshen, A.B., Gerald, W.L., Massague, J., 2005. Genes that mediate breast cancer metastasis to lung. Nature 436, 518–524. O'Connor, L., Glynn, B., 2010. Recent advances in the development of nucleic acid diagnostics. Expert Rev. Med. Devices 7, 529–539.

182

S. Liu et al. / European Journal of Pharmaceutical Sciences 82 (2016) 171–182

Ottaviano, F.G., Handy, D.E., Loscalzo, J., 2008. Redox regulation in the extracellular environment. Circ. J. 72, 1–16. Panigrahy, D., Edin, M.L., Lee, C.R., Huang, S., Bielenberg, D.R., Butterfield, C.E., Barnes, C.M., Mammoto, A., Mammoto, T., Luria, A., Benny, O., Chaponis, D.M., Dudley, A.C., Greene, E.R., Vergilio, J.A., Pietramaggiori, G., Scherer-Pietramaggiori, S.S., Short, S.M., Seth, M., Lih, F.B., Tomer, K.B., Yang, J., Schwendener, R.A., Hammock, B.D., Falck, J.R., Manthati, V.L., Ingber, D.E., Kaipainen, A., D'Amore, P.A., Kieran, M.W., Zeldin, D.C., 2012. Epoxyeicosanoids stimulate multiorgan metastasis and tumor dormancy escape in mice. J. Clin. Invest. 122, 178–191. Park, T.E., Kang, B., Kim, Y.K., Zhang, Q., Lee, W.S., Islam, M.A., Kang, S.K., Cho, M.H., Choi, Y.J., Cho, C.S., 2012. Selective stimulation of caveolae-mediated endocytosis by an osmotic polymannitol-based gene transporter. Biomaterials 33, 7272–7281. Peng, Q., Zhong, Z., Zhuo, R., 2008a. Disulfide cross-linked polyethylenimines (PEI) prepared via thiolation of low molecular weight PEI as highly efficient gene vectors. Bioconjug. Chem. 19, 499–506. Peng, X.C., Yang, L., Yang, L.P., Mao, Y.Q., Yang, H.S., Liu, J.Y., Zhang, D.M., Chen, L.J., Wei, Y.Q., 2008b. Efficient inhibition of murine breast cancer growth and metastasis by gene transferred mouse survivin Thr34– N ala mutant. J. Exp. Clin. Cancer Res. 27, 46. Ryu, K., Kim, T.I., 2014. Therapeutic gene delivery using bioreducible polymers. Arch. Pharm. Res. 37, 31–42. Shim, M.S., Kwon, Y.J., 2008. Controlled delivery of plasmid DNA and siRNA to intracellular targets using ketalized polyethylenimine. Biomacromolecules 9, 444–455. Son, S., Singha, K., Kim, W.J., 2010. Bioreducible BPEI-SS-PEG-cNGR polymer as a tumor targeted nonviral gene carrier. Biomaterials 31, 6344–6354. Son, S., Hwang do, W., Singha, K., Jeong, J.H., Park, T.G., Lee, D.S., Kim, W.J., 2011. RVG peptide tethered bioreducible polyethylenimine for gene delivery to brain. J. Control. Release 155, 18–25. Son, S., Namgung, R., Kim, J., Singha, K., Kim, W.J., 2012. Bioreducible polymers for gene silencing and delivery. Acc. Chem. Res. 45, 1100–1112. Thomas, M., Lu, J.J., Ge, Q., Zhang, C., Chen, J., Klibanov, A.M., 2005. Full deacylation of polyethylenimine dramatically boosts its gene delivery efficiency and specificity to mouse lung. Proc. Natl. Acad. Sci. U. S. A. 102, 5679–5684.

van Gaal, E.V., van Eijk, R., Oosting, R.S., Kok, R.J., Hennink, W.E., Crommelin, D.J., Mastrobattista, E., 2011. How to screen non-viral gene delivery systems in vitro? J. Control. Release 154, 218–232. Wang, Y., Chen, P., Shen, J., 2006. The development and characterization of a glutathionesensitive cross-linked polyethylenimine gene vector. Biomaterials 27, 5292–5298. Weigelt, B., Peterse, J.L., van't Veer, L.J., 2005. Breast cancer metastasis: markers and models. Nat. Rev. Cancer 5, 591–602. Xia, W., Wang, P., Lin, C., Li, Z., Gao, X., Wang, G., Zhao, X., 2012. Bioreducible polyethylenimine-delivered siRNA targeting human telomerase reverse transcriptase inhibits HepG2 cell growth in vitro and in vivo. J. Control. Release 157, 427–436. Xiao, B., Laroui, H., Ayyadurai, S., Viennois, E., Charania, M.A., Zhang, Y., Merlin, D., 2013a. Mannosylated bioreducible nanoparticle-mediated macrophage-specific TNF-alpha RNA interference for IBD therapy. Biomaterials 34, 7471–7482. Xiao, J., Duan, X., Yin, Q., Miao, Z., Yu, H., Chen, C., Zhang, Z., Wang, J., Li, Y., 2013b. The inhibition of metastasis and growth of breast cancer by blocking the NF-kappaB signaling pathway using bioreducible PEI-based/p65 shRNA complex nanoparticles. Biomaterials 34, 5381–5390. Xiong, M.P., Forrest, M.L., Ton, G., Zhao, A., Davies, N.M., Kwon, G.S., 2007. Poly(aspartateg-PEI800), a polyethylenimine analogue of low toxicity and high transfection efficiency for gene delivery. Biomaterials 28, 4889–4900. Xu, G.C., Zhang, P., Leng, F., Pan, L., Li, Z.Y., Yu, D.D., Shan, Y., Yuan, Q.Z., Wen, Y., Mu, B., Shi, H.S., Chen, X., Wang, C.T., 2012. Inhibition of lymphatic metastases by a survivin dominant-negative mutant. Oncol. Res. 20, 579–587. Yang, F., Huang, W., Li, Y., Liu, S., Jin, M., Wang, Y., Jia, L., Gao, Z., 2013. Anti-tumor effects in mice induced by survivin-targeted siRNA delivered through polysaccharide nanoparticles. Biomaterials 34, 5689–5699. Yin, Q., Shen, J., Zhang, Z., Yu, H., Chen, L., Gu, W., Li, Y., 2013. Multifunctional nanoparticles improve therapeutic effect for breast cancer by simultaneously antagonizing multiple mechanisms of multidrug resistance. Biomacromolecules 14, 2242–2252.