Acta Biomaterialia 10 (2014) 2663–2673

Contents lists available at ScienceDirect

Acta Biomaterialia journal homepage: www.elsevier.com/locate/actabiomat

Storage stability of optimal liposome–polyethylenimine complexes (lipopolyplexes) for DNA or siRNA delivery Alexander Ewe a, Andreas Schaper b, Sabine Barnert c, Rolf Schubert c, Achim Temme d, Udo Bakowsky e, Achim Aigner a,⇑ a

Rudolf-Boehm-Institute for Pharmacology and Toxicology, Clinical Pharmacology, University of Leipzig, Leipzig, Germany Philipps University Marburg, Materials Science Center, EM&Mlab, Marburg, Germany Institute of Pharmaceutical Sciences, Department of Pharmaceutical Technology and Biopharmacy, Albert Ludwig University Freiburg, Germany d Department of Neurosurgery, University Hospital Carl Gustav Carus, TU Dresden, Germany e Philipps University Marburg, Department of Pharmaceutical Technology and Biopharmaceutics, Marburg, Germany b c

a r t i c l e

i n f o

Article history: Received 15 October 2013 Received in revised form 10 February 2014 Accepted 21 February 2014 Available online 1 March 2014 Keywords: Polyethylenimine PEI Liposomes Lipopolyplexes siRNA

a b s t r a c t The delivery of nucleic acids such as DNA or siRNA still represents a major hurdle, especially with regard to possible therapeutic applications in vivo. Much attention has been focused on the development of nonviral gene delivery vectors, including liposomes or cationic polymers. Among them, polyethylenimines (PEIs) have been widely explored for the delivery of nucleic acids and show promising results. The combination of cationic polymers and liposomes (lipopolyplexes) for gene delivery may further improve their efficacy and biocompatibility, by combining the favourable properties of lipid systems (high stability, efficient cellular uptake, low cytotoxicity) and PEIs (nucleic acid condensation, facilitated endosomal release). In this study, we systematically analyse various conditions for the preparation of liposome– polyethylenimine-based lipopolyplexes with regard to biological activity (DNA transfection efficacy, siRNA knockdown efficacy) and physicochemical properties (size, zeta potential, stability). This includes the exploration of lipopolyplex compositions containing different liposomes and different relevant branched or linear low-molecular-weight PEIs. We establish optimal parameters for lipopolyplex generation, based on various PEIs, N/P ratios, lipids, lipid/PEI ratios and preparation conditions. Importantly, we also demonstrate that certain lipopolyplexes retain their biological activity and physicochemical integrity upon prolonged storage, even at 37 °C and/or in the presence of serum, thus providing formulations with considerably higher stability as compared to polyplexes. In conclusion, we establish optimal liposome– polyethylenimine lipopolyplexes that allow storage under ambient conditions. This is the basis and an essential prerequisite for novel, promising and easy-to-handle formulations for possible therapeutic applications. Ó 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

1. Introduction The use of therapeutic nucleic acids, such as DNA and siRNA, has been a promising strategy for the treatment of genetic disorders and cancer. One of the great challenges is the delivery of nucleic acids in vivo, including their protection from nuclease degradation and their cellular uptake and intracellular release. For this purpose, viral and non-viral vectors are used. Recombinant viruses have been shown to infect cells efficiently, but revealed some disadvantages with regard to immunogenicity, inflamma-

⇑ Corresponding author. Tel.: +49 341 97 24661; fax: +49 341 97 24669. E-mail address: [email protected] (A. Aigner). http://dx.doi.org/10.1016/j.actbio.2014.02.037 1742-7061/Ó 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

tion, quality control and limitations in gene size and large scale manufacturing [1,2]. To overcome these problems, much attention has been focused on the development of non-viral gene delivery vectors such as cationic liposomes, cationic polymers, dendrimers, peptides and inorganic compounds [3–5]. Commonly used cationic liposomes/lipids or cationic polymers form self-assembled complexes (‘‘polyplexes’’) based on electrostatic attraction with negatively charged nucleic acids, and hence protect them from degradation [6]. These non-viral systems can be divided into organic or inorganic nanoparticles (nanoplexes) [7,8]. One of the major hurdles associated with liposomes (lipoplexes) or polycationic polymers (polyplexes) is their escape from endosomal/lysosomal vesicles, where nucleases and acidic pH digest nucleic acids, into the desired cellular compartments

2664

A. Ewe et al. / Acta Biomaterialia 10 (2014) 2663–2673

(nucleus for DNA and cytoplasm for siRNA). It is assumed that lipoplexes release the nucleic acids into the cytosol via membrane fusion or destabilization by interactions of the cationic liposomes with negatively charged membrane lipids. While many cationic liposomes show high biocompatibility, they exhibit some disadvantages such as high zeta potentials and low transfection efficiencies due to inefficient protection against lysosomal nucleic acid degradation [9,10]. Therefore, many efforts have been made to synthesize new lipids and investigate the physicochemical properties of liposomes to improve the transfection efficiency [11,12]. Neutral helper lipids, e.g. DOPE, are often used as components in cationic liposomal compositions in order to neutralize high surface charges and thus reduce cytotoxicity. Furthermore, DOPE is known to destabilize lipid bilayers and favours endosomal disruption [13]. Polyethylenimine (PEI) is a widely explored cationic polymer for the delivery of nucleic acids [14–17]. PEIs are synthetic, water-soluble branched or linear polymers available in a broad range of molecular weights (0.8–800 kDa) and possess a high cationic charge density at physiological pH due to protonable amino groups in every third position [18–20]. PEI protects nucleic acids from nuclease digestion, mediates the endosomal/lysosomal release of nanoscale complexes due to the so-called ‘‘proton sponge effect’’ [21,22] and was shown to facilitate the DNA entry into the nucleus [23,24]. Still, transfection efficacy and cytotoxicity strongly depend on the molecular structure and molecular weight [18,25]. In general, cytotoxicity and efficiency are higher with increasing molecular weight, while linear PEIs have shown to be less toxic and more efficient in the case of DNA transfections [26,27]. Furthermore, based on their surface charge, cationic vectors display low colloidal stability and tend to aggregate in the presence of salts and serum proteins, thus hampering biological activity. For further improvement of non-viral gene vectors, particularly PEI-based formulations, different strategies have been pursued. Chemical modifications of PEI include the grafting with poly(ethylene glycol) (PEG) or other polymers like chitosan, dextran, hydroxyethyl starch or carbohydrates [28–33], or the chemical coupling of hydrophobic chains, fatty acid residues, cholesterol, hyaluronic acid, polyglycerol, amino acids or peptides [34–46]. For the successful establishment of non-viral nanoparticulate systems with regard to possible therapeutic applications, the further improvement of biological activities and pharmacokinetics are still major goals. Beyond this, storage stability is another very important aspect. So far, the performance of many systems, including polymeric nanoparticles, in this regard is rather poor. More specifically, polymeric nanoparticles have a strong tendency to aggregate, leading to the complete loss of bioactivity in relatively short time. This requires the establishment and further refinement of novel or existing nanoparticle systems. In recent years, some studies have been published on the combination of cationic polymers and liposomes (lipopolyplexes) for gene delivery. Some groups could demonstrate that lipopolyplexes, consisting of PEI and cationic liposomes, showed enhanced in vitro transfection efficiencies and improved serum stability [10,47–51]. Lipopolyplexes containing neutral, anionic or PEG-modified (phospho-) lipids, however, are more promising candidates. In vitro experiments have shown higher transfection efficiencies, lower cytotoxicities and relatively high colloidal stability against serum proteins and physiological salt concentrations. Additionally, in vivo studies indicated prolonged clearance properties, altered biodistribution and no significant toxicity [52–55]. Previously, we have introduced the 4–10 kDa branched PEI F25LMW as a potent non-viral vector for the delivery of pDNA/siRNA in vitro and in vivo [25,56–58]. We have also established lipopolyplexes, comprising PEI F25-LMW and the neutral phospholipid

dipalmitoyl-phosphatidyl-choline (DPPC) with various colipids, for enhanced DNA and siRNA delivery, reduced toxicity and altered physicochemical properties [51]. In this study, we systematically analyse various lipopolyplex preparation conditions with regard to biological activity and physicochemical properties, and extend lipopolyplex compositions towards relevant linear low-molecular weight PEIs, linPEI and linPEImax. We thus combine the favourable properties of optimal PEIs with those of lipid systems. Importantly, we also demonstrate that lipopolyplex formation retains the biological activity and physicochemical integrity of the nanoparticles upon prolonged storage, even at 37 °C and/or in the presence of serum, thus providing more stable formulations as compared to polyplexes. 2. Materials and methods 2.1. Materials PEI F25-LMW was prepared as described previously [25]. Branched PEI 25 kDa, 1.8–2 kDa and 0.6–0.8 kDa were obtained from Sigma (Taufkirchen, Germany), branched PEI 10 kDa, linear PEI 25 kDa (‘‘linPEI’’) and linear PEI max (‘‘linPEImax’’) were purchased from Polysciences (Eppelheim, Germany). All PEIs were diluted in sterile distilled H2O to a final concentration of 1 mg ml–1, with the linear PEIs being adjusted to pH 4 with hydrochloric acid for improved solubility as suggested by the vendor. DPPC, N-[1-(2,3-dioleoyloxy)]-N,N,N-trimethylammonium-propane methylsulfate (DOTAP) and dipalmitoyl-phosphatidyl-ethanolaminen-polyethylene glycol 5000 (DPPEmPEG5k) were obtained from Avanti Polar Lipids (Alabaster, USA). The luciferase expression plasmid pGL3 (Promega, Mannheim, Germany) and the b-galactosidase expression plasmid (pSELECT-zeo-LacZ, InvivoGen) were propagated in Escherichia coli DH5a. Plasmid DNA was isolated from an overnight culture using the Midi KIT from Macherey and Nagel (Düren, Germany) according to the manufacturer’s protocol. Chemically synthesized siRNA duplexes directed against luciferase siLuc3 (sense: 50 -CUUACGCUGAGUACUUCGAdTdT-30 , antisense: 30 -dTdTGAAUGCGACUCAUGAAGCU-50 ) and siLuc2 as negative control (sense: 50 -CGUACGCGGAAUACUUCGATTdTdT-30 ; antisense: 30 -dTdTGCAUGCGCCUUAUGAAGCU-50 ) were purchased from MWG (Ebersberg, Germany). Wildtype cell lines (SKOV-3 ovarian carcinoma, SW620 colon carcinoma, MCF-7 mamma carcinoma and PC-3 prostate carcinoma cells; see the ATCC website for extensive documentation) were obtained from the American Type Culture Collection (ATCC/LGC Promochem, Wesel, Germany) and constitutive luciferase expressing SKOV-3-LUC cells were described previously [25]. 2.2. Liposome and PEI complex preparation Three different liposomal formulations were prepared using the thin lipid film method; DPPC, DPPC/DOTAP (92:8 molar ratio) and DPPC/DPPEmPEG5k (95:5 molar ratio). To this end, 5 mg of phospholipid or lipid formulation dissolved in chloroform/methanol (2:1, v/v) was mixed in a 5 ml round-bottom flask. The solvent was evaporated at 55 °C on a rotary evaporator by a programmable vacuum pump and well-defined time/pressure steps (0 s/ 1000 mbar, 30 s/800 mbar, 5 min/500 mbar, 30 min/0 mbar). The thin lipid film was hydrated with 1 ml sterile dH2O and incubated for 2 min at 55 °C in an ultrasound bath sonicator. Subsequently, the liposome suspension was extruded 11 times through a 200 nm polycarbonate membrane using a preheated Mini-Extruder (Avanti Polar Lipids, Alabaster, USA). The PEI polyplexes were prepared as described previously [25] at PEI/nucleic acid ratios as indicated in the figure and the text.

A. Ewe et al. / Acta Biomaterialia 10 (2014) 2663–2673

Unless stated otherwise, PEI F25-LMW complexes were at a mass ratio of 5. N/P ratios were calculated according to the following equation: N/P = (mass of the polymer  330 g mol–1)/(43 g mol–1  mass of the nucleic acid). Generally, 0.5 lg DNA or 0.8 lg siRNA were dissolved in 12.5 ll HN-buffer (10 mM HEPES, 150 mM NaCl, pH 7.4) and in a second vial, the desired amount of PEI was dissolved in 12.5 ll of the same buffer. The PEI solution was then added to the nucleic acid solution, the mixture was briefly vortexed and incubated for 30 min at room temperature. 2.3. Lipopolyplex generation For the preparation of lipopolyplexes, the PEI/nucleic acid complexes were incubated with preformed liposomes. To this end, appropriate amounts of liposomes (see text and figures for details) were diluted in 25 ll dH2O and pipetted to 25 ll of the polyplex solution containing 0.5 lg DNA or 0.8 lg siRNA. Unless indicated otherwise, the mixture was vigorously pipetted, vortexed and incubated for 60 min at room temperature prior to use or subsequent storage. For serum stability experiments, different amounts of fetal calf serum (FCS) were added as indicated in the figure and complexes/lipopolyplexes were incubated at 4 °C.

2665

125 ll of HN-buffer. For lipopolyplex preparation indicated amounts of liposomes diluted in 125 ll pure water were added to 125 ll of polyplex solution. For the measurements the nanoparticle formulations were further diluted to a total volume of 1.5 ml pure water. 2.6. NanoSight Size determination by nanoparticle tracking was performed with a NanoSight LM-10HS apparatus equipped with a 640 nm sCMOS camera and a temperature controlled sample chamber. For capturing and analysing the data, the NTA 2.3 software was used in the advanced mode. The complexes were prepared as described above and diluted with pure water to an optimal concentration (PEI/DNA complexes: 10 lg PEI F25-LMW/2 lg DNA in 500 ll dH2O; lipopolyplexes: 50 lg DPPC/10 lg PEI F25-LMW/ 1 lg DNA in 1000 ll dH2O; DPPC liposomes: 1 ll stock solution (5 mg ml–1) in 2000 ll dH2O). Typically, 300 ll of the sample was injected into the sample chamber using sterile 1 ml syringes. All measurements were performed at 25 °C; the samples were measured for 60 s with manual shutter and gain adjustments. Four measurements of each sample were taken and the mean size was calculated by the NTA software.

2.4. Cell culture All tumour cell lines were cultivated under standard conditions (37 °C, 5% CO2 in a humid atmosphere) in IMDM medium (PAA, Cölbe, Germany), supplemented with 10% fetal calf serum (Gibco, Karlsruhe, Germany). For transfection experiments, cells were seeded at 3.5  104 cells per well in 24-well plate 24 h prior the experiment. Complexes or lipopolyplexes were prepared according to the above protocol and amounts corresponding to 0.5 lg DNA or 0.8 lg siRNA, respectively, were added per well. The experiments were conducted in IMDM supplemented with 10% FCS and penicillin/ streptomycin without a medium exchange. The determination of luciferase activity was performed 72 h after transfection using the Beetle-Juice kit from PJK (Kleinblittersdorf, Germany). Briefly, the medium was aspirated and the cells were lysed in 100 ll lysis buffer (Promega). In an appropriate tube, 25 ll luciferin substrate was mixed with 10 ll lysate and chemoluminescence was immediately determined in a luminometer (Berthold, Bad Wildbad, Germany). The b-galactosidase expression in glutaraldehyde-fixed cells was visualized using the b-Gal Staining Kit (Life Technologies, Darmstadt, Germany). Cell viability was determined using a colorimetric assay (Cell Proliferation Reagent WST-1; Roche Molecular Biochemicals, Mannheim, Germany) according to the manufacturer’s protocol and as described previously [56], with 7500 cells per well (96-well plate) and 0.1 lg DNA for transfection. 2.5. Zetasizer Hydrodynamic diameters and zeta potentials were measured by dynamic light scattering using a Brookhaven ZetaPALS system and calculated with the manufacturer’s software. The viscosity and refractive index of pure water at 25 °C were chosen for analysis. For size measurements, data were collected from five runs, with a run duration of 1 min each. Polyplex sizes are expressed as effective diameters based on intensity weighted analysis, assuming a lognormal distribution. Selected data were also shown by multimodal size distribution using the non-negatively constrained least squares algorithm analysis. Zeta potentials were determined in five runs, with each run comprising ten cycles, by applying the Smoluchowski model. Complex formation was done as described above, with 2.5 lg pDNA and the corresponding amount of PEI in a total volume of

2.7. Cryo-scanning electron microscopy (SEM) and cryo-transmission electron microscopy (TEM) The SEM observations were performed using a JSM-7500F (Jeol Ltd, Tokyo, Japan) with a cold-field emission gun (FEG). Our JSM-7500F is equipped with an ALTO-2500 liquid nitrogen (LN2) cryo-transfer system (Gatan, Pleasanton, CA, USA) which allows us to prepare freeze-fracture surfaces of the lipo-polyplex samples and SEM observations at a specimen temperature of 130 °C. The fracture surfaces were sputtered directly in the chamber of the ALTO system with a thin platinum layer to avoid charging. The SEM was operated at an accelerating voltage of 5.0 kV. Cryo-TEM was performed according to Ref. [59]. Sample preparation steps were done in a climate-controlled room using a CryoBox 340719 (Carl Zeiss, Oberkochen, Germany). Briefly, 3 ll of the sample was applied on a 400  100 mesh QuantifoilÒ S7/2 holey carbon film on copper grids (Quantifoil Micro Tools, Jena, Germany). After removing excess liquid with filter paper, the grid was immediately shock-frozen by injecting it into liquid ethane. The subsequent fixation of the grid on the sample rod (626-DH, Gatan, Pleasanton, CA) and transfer of the rod into the TEM (Leo 912 X-megaÒ, Leo, Oberkochen, Germany) was performed under a nitrogen atmosphere at a temperature of 90 K (–183 °C). The instrument was operated at 120 kV, and images were taken at a 6300- to 12 500-fold magnification. 2.8. Statistical analysis All data are presented as mean ± standard deviation (SD), obtained from triplicates of n = 3 independent experiments and as indicated in the figure legends. 3. Results 3.1. Comparison of different PEIs and lipids for lipopolyplex formation For transfection, only certain PEIs can be employed due to cytotoxicity of high-molecular-weight PEIs and poor bioactivity of very-low-molecular-weight PEIs. The branched, 4–10 kDa PEI F25-LMW has been established previously for the formation of efficient and biocompatible PEI complexes at a N/P ratio of 38 [25].

A. Ewe et al. / Acta Biomaterialia 10 (2014) 2663–2673

80 60 40

lipopolyplex

C

SW620

PC-3

100

lipopolyplex

SKOV-3

DPPC/ DPPEmPEG5k

DPPC

DPPC/DOTAP

linPEImax

0

DPPC/ DPPEmPEG5k

20 DPPC/DOTAP

DPPC/ DPPEmPEG5k

DPPC

DPPC/DOTAP

linPEImax

0

100

lipopolyplex

lipopolyplex

MCF-7

160

DPPC/ DPPEmPEG5k

DPPC/DOTAP

linPEI

lipopolyplex

DPPC

DPPC/DOTAP DPPC/ DPPEmPEG5k

F25

20

50

120

linPEI

40

100

140

DPPC

60

150

PC-3

160

DPPC/DOTAP DPPC/ DPPEmPEG5k

80

RLU (% of polyplex +/- SD)

100

F25

120

200

F25/DPPC

Number of β-gal positive cells (% of polyplex +/- SD)

140

0

lipopolyplex

MCF-7

SW620

140 Viable cells in WST-1 (% of control +/- SD)

RLU (% of polyplex +/- SD)

B

SKOV-3

160

DPPC

RLU (% of polyplex +/- SD)

A

better transfection results and was used in subsequent experiments. Next, lipopolyplexes based on different liposomes (DPPC, DPPC/ DOTAP, DPPC/DPPEmPEG5k) and the three bioactive PEIs (linPEI, linPEImax, PEI F25-LMW) were compared. The optimal lipid/PEI ratio (5 for PEI F25-LMW, 10 for linPEI and linPEImax) was employed, and transfection efficacies were determined in SKOV-3 cells and compared to their parent PEI/DNA polyplex. While all lipopolyplexes showed transfection efficacy, some differences were noted (Fig. 1A, left). DPPC-based lipopolyplexes performed best for all

F25

Additionally, two linear 25 kDa PEIs were selected and tested at different complexation ratios. For linPEI as well as for its completely deacetylated derivative, linPEImax, the N/P ratio of 19 was optimal (Supplementary Fig. S1, left) and used in subsequent experiments. For lipopolyplex formation of the linear PEIs, DPPC at different lipid/PEI mass ratios (equal amounts up to 20-fold lipid excess) was explored with regard to transfection efficacy (Supplementary Fig. S1, right). Notably, only minor differences were observed, indicating sufficient amounts of lipid being present already at 1:1 ratios. Still, a 10-fold DPPC excess showed slightly

DPPC

2666

120 100 80 60 40

80 60 40 20

F25/DPPC

F25

F25

F25/DPPC

F25

F25/DPPC

lP/DPPC

linPEI

lPM/DPPC

linPEImax

F25

F25/DPPC

0

DPPC/ DPPEmPEG5k

DPPC/DOTAP

F25

lipopolyplex

DPPC

DPPC/ DPPEmPEG5k

DPPC/DOTAP

F25

0

DPPC

20

lipopolyplex

D RLU (% of PEI F25 +/- SD)

100 80 60 40 20

0.1 0 PEI/DNA ratio

F25

5 10 20 100 0.6 bPEI

5 10 20 100 2 bPEI

5 10 20 100 10 bPEI

5 10 20 25 bPEI

F25

5 1 0.6 bPEI

5 1 2 bPEI

2 1 10 bPEI

2 1 25 bPEI

Lipid/PEI ratio

Fig. 1. (A) Transfection efficacies of various lipopolyplexes comprising PEI F25-LMW, linPEI or linPEImax and different liposomes in SKOV-3 cells (left). Results are compared to luciferase activities upon transfection of the respective fresh complex which are set to 100%. Lipopolyplex formation of PEI F25-LMW complexes also leads to an increase in the number of transfected cells, as determined by b-galactosidase positivity (right). (B) Transfection efficacies of various lipopolyplexes in other cell lines. (C) Cell viabilities of the various cell lines upon transfection with PEI-based polyplexes or lipopolyplexes, as determined by a WST-1 assay for viable cells. (D) Transfection efficacies of polyplexes based on a broader set of different ‘‘non-optimal’’ PEIs, with polyplexes prepared at various N/P ratios (left). Best N/P ratios were used for lipopolyplex formation at different lipid/PEI mass ratios and reveal an only slight improvement over polyplexes (right). Luciferase activities are compared to fresh polyplex (‘‘F25’’; left bars). Data represent the mean ± SD of n = 3 independent experiments in SKOV-3 cells, unless indicated otherwise.

A. Ewe et al. / Acta Biomaterialia 10 (2014) 2663–2673

PEIs and thus DPPC was used for subsequent experiments. In contrast, previous findings of DOTAP being favourable for gene delivery did not apply for this DPPC/DOTAP combination for generating our lipopolyplexes. Our results of slightly improved transfection efficacy upon DPPC-based lipopolyplex formation of PEI complexes (leftmost bars in Fig. 1A, left, and Ref. [25]) were also confirmed in b-galactosidase (b-gal) transfection experiments. Upon staining of the fixed cells for b-gal expression, a 1.5-fold increase in the number of positive cells was observed (Fig. 1A, right). Transfection experiments were also extended towards other cell lines derived from different tumour entities (SW620, colon carcinoma; PC-3, prostate carcinoma; MCF-7, breast carcinoma) and known for different transfection efficacies, thus aiming at a more general assessment of the biological properties of the lipopolyplexes (Fig. 1B). While minor differences between the cell lines were observed, these results confirmed the increase in transfection efficacy upon DPPC lipopolyplex formation, especially in the case of PEI F25-LMW, and thus the selection of DPPC for subsequent experiments. Cell viability experiments on PEI F25-LMW-based complexes vs. DPPC lipopolyplexes revealed in some cell lines a slight decrease in the number of viable cells (Fig. 1C). This effect, however, was in a maximum range of 20% (PC-3 cells) and was not observed in all cell lines (MCF-7, SW620) or with the other PEIs, thus demonstrating the overall high biocompatibility of the lipopolyplexes. Next, we asked the question whether DPPC lipopolyplex formation would enhance transfection efficacies of poor performing PEIs. To this end, four branched PEIs with molecular weight below (0.6 kDa, 2 kDa) or above (10 kDa, 25 kDa) PEI F25-LMW were selected. The best N/P ratios for transfection were determined to be very high (N/P = 100) for the very low molecular weight PEIs, while the higher molecular weight 10 kDa and 25 kDa PEIs showed maximum efficacy at N/P = 5. Overall, transfection efficacy increased with higher molecular weight, but was never above 30% of the PEI F25-LMW standard (Fig. 1D, left). When 0.6 kDa or 2 kDa PEI complexes at optimal N/P ratio (N/P = 100) were subjected to DPPC lipopolyplex formation, no increase in transfection efficacy was observed (Fig. 1D, right). In contrast, 1.5–2-fold higher biological activities were obtained for the 10 kDa or 25 kDa PEI complexes at a lipid/PEI ratio of 2 (Fig. 1D, right). The use of a > 2-fold excess (w/w) of DPPC did not further improve transfection efficacy (not shown). From these results we conclude that lipopolyplex formation leads to a slight improvement of transfection efficacy in the case of higher-molecular-weight PEIs, but not the very-low-molecularweight PEIs. However, this effect is not sufficiently profound and biological activities always remained below PEI F25-LMW, which made us stick to the optimal PEIs (linPEI, linPEImax, PEI F25-LMW).

3.2. Lipopolyplex efficacy is independent of preparation conditions Lipopolyplex formation will rely on the interaction of PEI-based complexes with the respective lipid. To address the effects of different preparation conditions, lipopolyplexes were generated either over 20 min at room temperature +5 min in a 50 °C ultrasound bath, with or without subsequent cooling in 4 °C water, or by vortexing, vigorous pipetting or slow mixing of the components at 4 °C or at 37 °C. Furthermore, prior to ultrasound the liposomes were pre-incubated for 30 min at 4 °C or 37 °C, and in some cases the lipopolyplex formation step was extended by an incubation for 1 h at room temperature (Supplementary Fig. S2A, left panel). The determination of transfection efficacies revealed slightly higher luciferase activities over fresh complexes, with no major differences between the various modes of preparation (Supplementary Fig. S2A, right panel).

2667

3.3. Lipopolyplex formation inhibits complex aggregation upon storage A marked problem when using PEI complexes, independent of molecular weight and biological activities, is their tendency to aggregate, thus leading to larger, inactive particles. Indeed, when analysing effective diameters of PEI F25-LMW/DNA complexes by zetasizer measurements, a 1.5 h storage at room temperature led already to an increase in complex size from 320 nm, which is well within the range reported previously [57], to 600 nm (Fig. 2A). This was also true for complexes based on other PEIs, and aggregation further proceeded towards the formation of micrometre-scale particles upon prolonged storage. The more detailed analysis of zetasizer histograms revealed a peak between 250 and 450 nm with a maximum at 320 nm (Fig. 2C, upper left). Upon storage for 6 h, most of the PEI/DNA complexes were around or above 1 lm (Fig. 2C, lower right). These results were confirmed by NanoSight measurements (right panels in Fig. 2C), with fresh complexes showing a peak at 250 nm (Fig. 2C, upper left). Since the NanoSight is unable to detect particles above 1 lm, storage at 6 h led to an overall loss of signal with only residual levels of the nanosized complexes (Fig. 2C, lower right, right panel), which is in good agreement with aggregate formation. In sharp contrast, the corresponding DPPC lipopolyplexes were initially smaller and remained at the same size upon storage (Fig. 2A). This size stability was independent of the storage temperature (4 °C, room temperature or 37 °C) and was also true for storage further prolonged to least up to 7 days (Fig. 2B). Again, these findings were confirmed by zetasizer histogram analyses and independently by NanoSight measurements (Fig. 2C, upper right). Cryo-SEM of DPPC-based lipopolyplexes showed similar lipopolyplex sizes and confirmed some size heterogeneity (Fig. 2D). A more detailed analysis by cryo-TEM, which was again in agreement with the above size measurements, furthermore revealed that most lipopolyplexes are not ideally round-shaped, but rather appear in a polygonal structure with flattened membrane areas (Fig. 2E), due to the rigidity of DPPC below the phase transition temperature of 41 °C. This was true for DPPC as well as for DPPC/ DOTAP-based lipopolyplexes (see upper and lower panels in Fig. 2E, respectively). Finally, the determination of the lipopolyplex zeta potential revealed that a 1:1 mass ratio was, at early time points, below the optimum and only led to a partial shielding of the polyplexes. More specifically, the zeta potential of the resulting lipopolyplexes was between the polyplex and the liposome (35 mV and –2 mV; upper and lower arrows in Fig. 2F, respectively). In contrast, higher lipid/PEI ratios (i.e., mass ratios 5 or 10) led to lipopolyplexes with zeta potentials of 0 mV and thus in the range of the corresponding liposome. The zeta potential at mass ratio 5 was only very slightly above the one at ratio 10, indicating complete lipid saturation. In both cases, the surface charge remained largely unchanged upon storage. In contrast, however, in the case of the 1:1 mass ratio, the zeta potential decreased over time, finally approaching the surface charge of optimal lipopolyplexes after 1 day of storage. This suggests a gradual increase in lipid association with the nanoparticles or other alterations in the lipopolyplex structure over time that will require further analysis, and also supports the relevance of the lipids for lipopolyplex stability over time. These findings were independent of the storage, as shown here by the comparison of lipopolyplexes being stored at room temperature vs. 37 °C. 3.4. Lipopolyplex formation preserves biological activity upon storage To assess the more crucial aspect, the preservation of biological activity, lipopolyplexes based on the different liposomes and PEIs (Fig. 3A) were stored at different temperatures (Fig. 3B) for up to 2 weeks and analysed for transfection efficacies in SKOV-3 cells.

2668

A. Ewe et al. / Acta Biomaterialia 10 (2014) 2663–2673

polyplex lipopolyplex RT

500 400 300 200 100

400 300 200 100 0

1.5 6 Duration of storage (h)

24 72 168 Duration of storage (h) Lipopolyplex t=24 h

F25/DNA t=0 h

Diameter (nm)

8 6 4

Diameter (nm)

DPPC liposomes

Diameter (nm)

F25/DNA t=6 h

100

0.2

Diameter (nm)

Diameter (nm)

Diameter (nm)

0.1

1000

0

750

0.05

500

1400

1200

1000

800

750

1000

0 500

0

0 100 200 300

2

0.15

250

50

0

4

106 particles / mL

Intensity (%)

6

0

400

300

100

200

0

0

8

250

106 particles / mL

Intensity (%)

10

50

1000

0

750

2 500

200

100

750

0

1000

500

0

0

0.5

Diameter (nm)

100

50

10

0

1.0

12 106 particles / mL

Intensity (%)

1.5

250

106 particles / mL

2.0

400

300

200

100

0

50

0

100

2.5

0

100

400

0

500

300

0

lipopolyplex 4°C lipopolyplex RT lipopolyplex 37°C

600

250

600

C Intensity (%)

Effective diameter (nm) +/-SEM

B Effective diameter (nm) +/-SEM

A

Diameter (nm)

Fig. 2. (A) Polyplex or lipopolyplex sizes, fresh vs. upon storage at room temperature, as determined by zetasizer. (B) Lipopolyplex sizes after prolonged storage at various temperatures. (C) More detailed analysis of sizes from zetasizer histograms (left panels) and NanoSight measurements (right panels), done for fresh (upper left) or stored polyplexes (lower right), fresh liposomes (lower left) and lipopolyplexes stored for 24 h (upper right). (D) Cryo-SEM images of DPPC-based lipopolyplexes. (E) Cryo-TEM images of DPPC and DPPC/DOTAP-based lipopolyplexes. (F) Zeta potentials of fresh polyplexes and DPPC liposomes (arrows), compared to lipopolyplexes prepared at different lipid/PEI ratios and measured freshly or upon storage as indicated. Representative data from one experiment, each with at least 4–5 runs or measurements (A–D and F), are shown.

As expected, PEI complexes without lipids rapidly lost their biological activity, with transfection efficacy below 2% at the earliest time point of analysis (3 days). This effect was independent of the PEI and the storage temperature (Fig. 3C). In contrast, DPPC lipopolyplex formation largely (linPEI, PEI F25-LMW) or entirely (linPEImax) preserved biological activity, even upon storage of the lipopolyplexes for up to 14 days (Fig. 3A, upper panel). Notably, this was not true for other lipids, with little (DPPC/DOTAP) or almost no (DPPC/DPPEmPEG5k) activity remaining after 3 days or thereafter (Fig. 3A, centre and lower panel). We therefore focused on DPPC and analysed transfection efficacies upon the more relevant storage at temperatures above 4 °C, i.e. room temperature or 37 °C. At room temperature, lipopolyplexes remained active, with the preservation of transfection efficacies being only slightly below the results of the lipopolyplexes stored at 4 °C (Fig. 3B, left). With 80% biological activity after 2 weeks, linPEImax-based lipopolyplexes showed again best results. This was also true for storage at 37 °C. While overall preservation of transfection efficacy was somewhat lower, linPEImax lipopolyplexes were still in the range of 60% after 2 weeks. Notably, the shape of the curves was different in the 37 °C incubation groups, with a more rapid decline within the first 3 days and little or no decrease thereafter (Fig. 3B, right).

To validate that the observed (partial) preservation of biological activities is indeed critically dependent on the presence of lipid, we analysed lipopolyplexes prepared at two lipid/PEI ratios, i.e. ratio 1 and ratio 0.01, which is well below the optimum. While at ratio 1, lipopolyplex activity was still at 50% after 1 month and only slightly below after 2 months, the suboptimal ratio 0.01 was insufficient and only slowed down the loss of transfection efficacy within the first 3–7 days (Fig. 3D). To confirm the results on the temperature-dependent protection of transfection efficacies upon lipopolyplex formation, experiments were again extended towards other cell lines (Fig. 3E). Despite some variations, the increase in the transfection efficacy of DPPC-based lipopolyplexes over polyplexes as well as the at least partial preservation of biological activity over the 7 days of analysis was again observed. Additionally, the transfection with stored lipopolyplexes did not impair the viability of any of the cell lines (Fig. 3F). From these data, we conclude that DPPC-based lipopolyplex formation is optimal for the protection of biological activity, allowing the storage of lipopolyplexes for months and to keep them even at physiological temperatures (37 °C). In line with the above data demonstrating similar transfection efficacies upon different conditions of complex generation, the temperature or the use of ultrasound during lipopolyplex

2669

A. Ewe et al. / Acta Biomaterialia 10 (2014) 2663–2673

E

D

DPPC lipopolyplexes

200 nm

500 nm

400 nm

DPPC/DOTAP lipopolyplexes

300 nm

F

40

Zeta-potential (mV)

300 nm

30

200 nm

500 nm

polyplex DPPC liposomes lipopolyplex 1/1 lipopolyplex 1/1 lipopolyplex 5/1 lipopolyplex 10/1

RT 37°C RT RT

20

10

0 0

6 Duration of storage (h)

24

Fig. 2 (continued)

preparation did not affect the preservation of biological lipopolyplex activity (Supplementary Fig. S2B). 3.5. Storage stability also occurs in the presence of serum Critical for in vivo applications is the question whether the lipid-mediated stabilization of polyplexes also occurs in the presence of serum. We therefore stored lipopolyplexes for 1 day or 4 days in medium supplemented with 10% or 50% FCS (Fig. 4). As compared to 0% FCS, the preservation of biological activity was only slightly lower under serum conditions, with transfection efficacies being >80% as compared to day 0. In contrast, nonlipid-modified polyplexes showed again a rapid loss of biological activity. No differences were observed between lipopolyplex storage under 10% vs. 50% serum conditions, and results were similar for all three PEIs (Fig. 4). 3.6. Lipopolyplexes for siRNA delivery Beyond DNA transfection, PEIs have also been established and widely used for siRNA delivery in vitro and in vivo. We thus aimed at identifying lipopolyplex compositions which are optimal with regard to knockdown efficacy and its preservation upon storage. PEI F25-LMW-based lipopolyplexes prepared with DPPC led to a 50% knockdown of luciferase expression at 3 days after transfection of stably luciferase expressing SKOV-3 cells (Fig. 5A; compare grey vs. black bars), which is in the range of knockdown efficacies observed in this cell line when using PEI-based complexes. Comparable target gene reductions were observed over a wide range of lipid/PEI ratios for lipopolyplex generation, i.e. between 0.1 and 10 (Fig. 5A, left). Similar results were obtained when using DPPC/DOTAP as lipid, with the exception of decreased knockdown at the highest lipid/PEI

ratio 10 (Fig. 5A, centre). Fewer efficacies with a maximum 30% knockdown were achieved in the case of DPPC/DPPEmPEG5 k lipopolyplexes (Fig. 5A, right). In this case, biological activities were completely lost after 3 days or 17 days of storage (Fig. 5B, right panels). In contrast, DPPC or DPPC/DOTAP lipopolyplexes completely preserved knockdown efficacies (Fig. 5B, left and center panels, respectively). Sufficiently high lipid/PEI ratios were important, as indicated by the loss of biological activity at day 17 in the lipopolyplexes prepared at a 0.01 ratio. This was particularly critical for DPPC where knockdown efficacy was already lost after 3 days, while in DPPC/DOTAP lipopolyplexes loss of biological activity only occurred at the later time point and at the highest lipid/PEI ratio 10 a somewhat diminished knockdown was observed.

4. Discussion While liposomes as well as polymeric nanoparticles have been extensively explored for nucleic acid therapy, leading to optimized systems for in vitro and in vivo use, their combination is expected to be particularly attractive. This approach has been explored previously, e.g., for the generation of dendrosomes consisting of PAMAM dendrimer-based dendriplexes which were subsequently incubated with a dispersion of PC/DOPE/cholesterol liposomes [60]. The interaction of DNA/polylysine complexes with negatively charged lipid vesicles also indicated a stabilization of the particles by lipid coating [61], and the lipid encapsulation of DNA polyplexes has been shown to alter their pharmacologic properties [52]. Vice versa, PEI for coating of anionic liposomes has been explored as well [62]. Combining the favourable properties of PEI (nucleic acid condensation, facilitated endosomal release) and of lipid systems (high stability, efficient cellular uptake, low cytotoxicity) is appealing. This is particularly true for certain low-molecular-weight branched

A. Ewe et al. / Acta Biomaterialia 10 (2014) 2663–2673

B DPPC_storage: 4°C

120 100 80 60 40

F25 linPEI linPEImax

20

DPPC_storage: rt

120

RLU (% of polyplex +/- SD)

140

RLU (% of polyplex +/- SD)

A

RLU (% of polyplex +/- SD)

2670

100 80 60 40

F25 linPEI linPEImax

20

100 80 60 40 20 0

0

0 0

2

4

6

8

10

12

0

14

2

4

6

8

10

12

DPPC_storage: 37°C

120

14

0

2

4

6

8

10

12

14

Days of storage

Days of storage

Days of storage DPPC/DOTAP_storage: 4°C

C

120

100 100

RLU (% of fresh polyplex +/- SD)

RLU (% of polyplex +/- SD)

140

80 60 40

PEI F25-LMW

80

80

60

60

40

40

20

20

20

0

0

60

RT 37 C

40

0

0 2

4

6

8

10

12

2

4

6

8 10 12 14

0 0

2

Days of storage

14

Linear PEI Max

100

80

20 0

Linear PEI

100 4 C

4

6

8 10 12 14

0

2

Days of storage

4

6

8 10 12 14

Days of storage

Days of storage

D

DPPC/DPPEmPEG5k_storage: 4°C 120

RLU (% of PEI F25 +/- SD)

100 80 60 40 20

lipid/PEI ratio 1 lipid/PEI ratio 0.01

150

100

50

0

0 8

10

12

14

0

3

Days of storage

200 0

31

MCF-7

SW620

PC-3

150 0

RT 150 0

150 0

150

100 0

0 100

100

0 50

0 50

50

37°C

0

0 0

2

4

6

Days of storage

60

F

200

0 200

4°C

14

8

Viable cells in WST-1 (% of control +/- SD)

RLU (% of polyplex +/- SD)

E

7

Days of storage

0 0

2

4

6

8

100 0

0 50

0 0

Days of storage

2

4

6

Days of storage

8

0 SW-620

6

4

MCF-7 7

2

3 SKOV-3

0

PC-3 3

RLU (% of polyplex +/- SD)

140

Cell line

Fig. 3. Biological activities of lipopolyplexes or polyplexes upon storage under various conditions. Experiments in SKOV-3 cells are shown unless stated otherwise. (A) Lipopolyplexes containing different lipids and different PEIs as indicated, freshly prepared or stored for up to 14 days at 4 °C. (B) Comparison of DPPC-based lipopolyplexes stored at room temperature (left) vs. 37 °C (right). (C) Loss of transfection efficacy upon storage of PEI complexes. (D) Critical importance of a sufficient lipid/PEI ratio for the preservation of DPPC-based lipopolyplex activity upon storage. Data represent the mean ± SD of n = 3 independent experiments. (E) Transfection efficacies of DPPC-based lipopolyplexes, upon their storage as indicated, in MCF-7, SW620 and PC-3 cells. (F) Effects of lipopolyplexes, stored for 3 days, on cell viabilities.

or linear PEIs, which show high efficacy and relatively low cytotoxicity and are explored in this paper in lipopolyplexes. Despite the usefulness of PEI complexes for in vitro transfection and also for in vivo applications, their poor size stability upon storage or exposure to physiological conditions (37 °C, high protein concentrations) poses major limitations. Indeed, we demonstrate in this paper that the formation of lipopolyplexes leads to stabilization over weeks. This is true for their physicochemical properties as well as, more importantly, for their biological activity, and considerably extends their possible use towards longer-term delivery (infusion, Alzet pump) or storage at ambient temperatures. So far, storage was limited to freezing, which is only applicable for certain PEI complexes [57], or lyophilisation as a somewhat more tedious process, requiring certain buffer conditions and in some cases being accompanied by a partial loss of biological activity

[25,63]. Thus, the lipopolyplexes described here provide an important avenue towards the in vivo use of non-viral polymeric nanoparticles without the requirement of fresh preparation or storage under defined, and rather impractical, conditions. Notably, the protective effects described here were found to be largely independent of the precise mode of lipopolyplex generation, thus indicating that no sophisticated and hard-to-reproduce setting is necessary for their preparation. While a certain minimum amount of lipid necessary for generating stable lipopolyplexes was to be expected from the literature [64,65], it should also be noted that the range of optimal lipid/PEI ratios above this threshold is rather broad, again indicating that no particular precision is required here. Interestingly, within this optimal range the properties of the lipopolyplexes not only depend on the lipid, but are also still influenced by the PEI, as indicated previously [64]. While the

2671

A. Ewe et al. / Acta Biomaterialia 10 (2014) 2663–2673

1 d storage 4 d storage

100 80 60 40 20

PEI F25-LMW

1 d storage 4 d storage

120

RLU (% of polyplex +/- SD)

120

Linear PEI RLU (% of polyplex +/- SD)

RLU (% of polyplex +/- SD)

Linear PEI Max

100 80 60 40 20

100

0

0 fresh complex 0% FCS 10% FCS complex stored

50% FCS

80 60 40 20 0

fresh complex 0% FCS 10% FCS 50% FCS complex stored lipopolyplex

lipopolyplex

1 d storage 4 d storage

120

fresh complex 0% FCS 10% FCS 50% FCS complex stored lipopolyplex

Fig. 4. Protection of lipopolyplex transfection efficacies upon storage under serum-free, 10% FCS or 50% FCS conditions for 1 day or 4 days, in comparison to polyplexes (0% FCS). Results are normalized to fresh complexes (set to 100%) and shown for linPEImax (left), linPEI (centre) and PEI F25-LMW (right). Data represent the mean ± SD of n = 3 independent experiments in SKOV-3 cells.

Neg. ctrl. siRNA Luc siRNA

120 100 80 60 40 20

DPPC

0.1

0.01

1

5

0.01

0.1

1

5

10

DPPC/ DPPEmPEG5k

DPPC/DOTAP

17 days

3 days

B

Neg. ctrl. siRNA Luc siRNA

Neg. ctrl. siRNA Luc siRNA

120 100 80 60 40 20 10 5 1 0.1 0.01

10 5 1 0.1 0.01

10 5 1 0.1 0.01

10 5 1 0.1 0.01

10 5 1 0.1 0.01

0 10 5 1 0.1 0.01

Luciferase activity (% of negative ctrl. siRNA +/- SD)

10

0.01

1

0.1

5

0 10

Luciferase activity (% of negative ctrl. siRNA +/- SD)

A

DPPC

DPPC/DOTAP

DPPC/ DPPEmPEG5k

DPPC

DPPC/DOTAP

DPPC/ DPPEmPEG5k

Fig. 5. SiRNA-mediated knockdown efficacies of various lipopolyplexes prepared at different lipid/PEI ratios and using different lipids. The luciferase activities upon transfection of the corresponding lipopolyplexes containing negative control siRNAs are set to 100% (black bars), and differences between black and grey bars represent luciferase knockdown. Lipopolyplexes were (A) freshly prepared or (B) stored for 3 days (left) or 17 days (right). Data represent the mean ± SD of n = 3 independent experiments in stable luciferase-expressing SKOV-3 cells.

branched PEI F25-LMW has been explored in a former study [51], we also extend the set towards two linear PEIs. The finding that the linPEImax is more efficient than linPEI also applies to lipopolyplexes, again indicating that their properties are determined by the lipid as well as by the PEI. In contrast, the nucleic acid is not a major issue, since our lipopolyplexes work for large plasmid DNA as well as for small siRNAs. Despite the fact that PEI/nucleic acid binding cooperativity is lost in siRNAs, the small double-stranded siRNAs are considerably more rigid than large plasmids and the intracellular mode of action is different between both, thus requiring different intracellular localizations.

The results shown here as well as published (see above) and unpublished data demonstrate that PEI-based lipopolyplex formation is possible for a wide range of lipids, independent of their charge, rigidity and other physicochemical properties. While some differences are observed with regard to lipopolyplex stability, major differences become obvious with regard to in vitro transfection efficacy and cytotoxicity, as shown here and previously, with surface charge and endosomal release being rate-limiting steps in cell transfection [51,66]. Although this is very relevant for their in vitro application, it should also be noted that the highest performer in cell culture may not be the best candidate in vivo. In fact, from a

2672

A. Ewe et al. / Acta Biomaterialia 10 (2014) 2663–2673

large set of oligomaltose-modified PEIs it was shown recently that, despite their relatively low in vitro efficacy, some grafted PEIs showed marked changes in their distribution profile in vivo [33], thus leading to some tissue specificity without targeted, ligandmediated delivery. It is intriguing to speculate that this may also apply to lipopolyplexes and opens the possibility of durable, tailor-made, albeit relatively easy to produce nanoparticles for therapeutic nucleic acid delivery in vivo. Acknowledgments This work was supported by Grants from the Deutsche For schungsgemeinschaft (AI 24/6-1 and AI 24/9-1) and the SMWK (Saxonian ministry for science and art) to A.A., and from the Deutsche Krebshilfe (Grant 110184) to A.A. and A.T. We are grateful to Bärbel Obst and Andrea Wüstenhagen for expert technical assistance, and to Michael Hellwig for the cryo-SEM analyses. Appendix A. Figures with essential colour discrimination Certain figures in this article, particularly Fig. 3 is difficult to interpret in black and white. The full colour images can be found in the online version, at http://dx.doi.org/10.1016/j.actbio.2014. 02.037. Appendix B. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.actbio.2014.02. 037. References [1] Kay MA. State-of-the-art gene-based therapies: the road ahead. Nat Rev Genet 2011;12:316–28. [2] Itaka K, Kataoka K. Recent development of nonviral gene delivery systems with virus-like structures and mechanisms. Eur J Pharm Biopharm 2009;71:475–83. [3] Piskin E, Dincer S, Turk M. Gene delivery: intelligent but just at the beginning. J Biomater Sci 2004;15:1181–202. [4] Basarkar A, Singh J. Nanoparticulate systems for polynucleotide delivery. Int J Nanomed 2007;2:353–60. [5] Aigner A. Cellular delivery in vivo of siRNA-based therapeutics. Curr Pharm Des 2008;14:3603–19. [6] Kodama K, Katayama Y, Shoji Y, Nakashima H. The features and shortcomings for gene delivery of current non-viral carriers. Curr Med Chem 2006;13:2155–61. [7] Ravi Kumar MN, Sameti M, Mohapatra SS, Kong X, Lockey RF, Bakowsky U, et al. Cationic silica nanoparticles as gene carriers: synthesis, characterization and transfection efficiency in vitro and in vivo. J Nanosci Nanotechnol 2004;4:876–81. [8] Kneuer C, Ehrhardt C, Bakowsky H, Kumar MN, Oberle V, Lehr CM, et al. The influence of physicochemical parameters on the efficacy of non-viral DNA transfection complexes: a comparative study. J Nanosci Nanotechnol 2006;6:2776–82. [9] Xu Y, Szoka Jr FC. Mechanism of DNA release from cationic liposome/DNA complexes used in cell transfection. Biochemistry 1996;35:5616–23. [10] Pelisek J, Gaedtke L, DeRouchey J, Walker GF, Nikol S, Wagner E. Optimized lipopolyplex formulations for gene transfer to human colon carcinoma cells under in vitro conditions. J Gene Med 2006;8:186–97. [11] Semple SC, Akinc A, Chen J, Sandhu AP, Mui BL, Cho CK, et al. Rational design of cationic lipids for siRNA delivery. Nat Biotechnol 2010;28:172–6. [12] Balazs DA, Godbey W. Liposomes for use in gene delivery. J Drug Deliv 2011;2011:326497. [13] Zhang S, Xu Y, Wang B, Qiao W, Liu D, Li Z. Cationic compounds used in lipoplexes and polyplexes for gene delivery. J Controlled Release 2004;100:165–80. [14] Boussif O, Lezoualc’h F, Zanta MA, Mergny MD, Scherman D, Demeneix B, et al. A versatile vector for gene and oligonucleotide transfer into cells in culture and in vivo: polyethylenimine. Proc Natl Acad Sci USA 1995;92:7297–301. [15] Neu M, Fischer D, Kissel T. Recent advances in rational gene transfer vector design based on poly(ethylene imine) and its derivatives. J Gene Med 2005;7:992–1009. [16] Lai WF. In vivo nucleic acid delivery with PEI and its derivatives: current status and perspectives. Exp Rev Med Dev 2011;8:173–85.

[17] Hobel S, Aigner A. Polyethylenimines for siRNA and miRNA delivery in vivo. Wiley Interdiscip Rev Nanomed Nanobiotechnol 2013. [18] Godbey WT, Wu KK, Mikos AG. Size matters: molecular weight affects the efficiency of poly(ethylenimine) as a gene delivery vehicle. J Biomed Mater Res 1999;45:268–75. [19] Tang MX, Szoka FC. The influence of polymer structure on the interactions of cationic polymers with DNA and morphology of the resulting complexes. Gene Ther 1997;4:823–32. [20] Lungwitz U, Breunig M, Blunk T, Gopferich A. Polyethylenimine-based nonviral gene delivery systems. Eur J Pharm Biopharm 2005;60:247–66. [21] Zuber G, Dauty E, Nothisen M, Belguise P, Behr JP. Towards synthetic viruses. Adv Drug Deliv Rev 2001;52:245–53. [22] Behr JP. The proton sponge: a trick to enter cells the viruses did not exploit. Chimia 1997;51:34–6. [23] Godbey WT, Wu KK, Mikos AG. Tracking the intracellular path of poly(ethylenimine)/DNA complexes for gene delivery. Proc Natl Acad Sci USA 1999;96:5177–81. [24] Pollard H, Remy JS, Loussouarn G, Demolombe S, Behr JP, Escande D. Polyethylenimine but not cationic lipids promotes transgene delivery to the nucleus in mammalian cells. J Biol Chem 1998;273:7507–11. [25] Werth S, Urban-Klein B, Dai L, Hobel S, Grzelinski M, Bakowsky U, et al. A low molecular weight fraction of polyethylenimine (PEI) displays increased transfection efficiency of DNA and siRNA in fresh or lyophilized complexes. J Controlled Release 2006;112:257–70. [26] Kwok A, Hart SL. Comparative structural and functional studies of nanoparticle formulations for DNA and siRNA delivery. Nanomed Nanotechnol Biol Med 2011;7:210–9. [27] Wiseman JW, Goddard CA, McLelland D, Colledge WH. A comparison of linear and branched polyethylenimine (PEI) with DCChol/DOPE liposomes for gene delivery to epithelial cells in vitro and in vivo. Gene Ther 2003;10:1654–62. [28] Malek A, Czubayko F, Aigner A. PEG grafting of polyethylenimine (PEI) exerts different effects on DNA transfection and siRNA-induced gene targeting efficacy. J Drug Target 2008;16:124–39. [29] Mao S, Neu M, Germershaus O, Merkel O, Sitterberg J, Bakowsky U, et al. Influence of polyethylene glycol chain length on the physicochemical and biological properties of poly(ethylene imine)-graft-poly(ethylene glycol) block copolymer/SiRNA polyplexes. Bioconjugate Chem 2006;17:1209–18. [30] Nguyen HK, Lemieux P, Vinogradov SV, Gebhart CL, Guerin N, Paradis G, et al. Evaluation of polyether-polyethyleneimine graft copolymers as gene transfer agents. Gene Ther 2000;7:126–38. [31] Erbacher P, Bettinger T, Belguise-Valladier P, Zou S, Coll JL, Behr JP, et al. Transfection and physical properties of various saccharide, poly(ethylene glycol), and antibody-derivatized polyethylenimines (PEI). J Gene Med 1999;1:210–22. [32] Bauhuber S, Liebl R, Tomasetti L, Rachel R, Goepferich A, Breunig M. A library of strictly linear poly(ethylene glycol)–poly(ethylene imine) diblock copolymers to perform structure–function relationship of non-viral gene carriers. J Controlled Release 2012;162:446–55. [33] Hobel S, Loos A, Appelhans D, Schwarz S, Seidel J, Voit B, et al. Maltose- and maltotriose-modified, hyperbranched poly(ethylene imine)s (OM-PEIs): physicochemical and biological properties of DNA and siRNA complexes. J Controlled Release 2011;149:146–58. [34] Jiang HL, Kim TH, Kim YK, Park IY, Cho MH, Cho CS. Efficient gene delivery using chitosan-polyethylenimine hybrid systems. Biomed Mater (Bristol, England) 2008;3:25013. [35] Ito T, Yoshihara C, Hamada K, Koyama Y. DNA/polyethyleneimine/hyaluronic acid small complex particles and tumor suppression in mice. Biomaterials 2010;31:2912–8. [36] Zhang L, Hu CH, Cheng SX, Zhuo RX. PEI grafted hyperbranched polymers with polyglycerol as a core for gene delivery. Colloids Surf 2010;76:427–33. [37] Noga M, Edinger D, Rodl W, Wagner E, Winter G, Besheer A. Controlled shielding and deshielding of gene delivery polyplexes using hydroxyethyl starch (HES) and alpha-amylase. J Controlled Release 2012;159:92–103. [38] Schroeder A, Dahlman JE, Sahay G, Love KT, Jiang S, Eltoukhy AA, et al. Alkanemodified short polyethyleneimine for siRNA delivery. J Controlled Release 2012;160:172–6. [39] Fortune JA, Novobrantseva TI, Klibanov AM. Highly effective gene transfection in vivo by alkylated polyethylenimine. J Drug Deliv 2011;2011:204058. [40] Oskuee RK, Dehshahri A, Shier WT, Ramezani M. Alkylcarboxylate grafting to polyethylenimine: a simple approach to producing a DNA nanocarrier with low toxicity. J Gene Med 2009;11:921–32. [41] Jiang D, Salem AK. Optimized dextran-polyethylenimine conjugates are efficient non-viral vectors with reduced cytotoxicity when used in serum containing environments. Int J Pharm 2012;427:71–9. [42] Tripathi SK, Goyal R, Kumar P, Gupta KC. Linear polyethylenimine-graftchitosan copolymers as efficient DNA/siRNA delivery vectors in vitro and in vivo. Nanomed Nanotechnol Biol Med 2012;8:337–45. [43] Wang X, Yao J, Zhou JP, Lu Y, Wang W. Synthesis and evaluation of chitosangraft-polyethylenimine as a gene vector. Die Pharm 2010;65:572–9. [44] Hsu CY, Hendzel M, Uludag H. Improved transfection efficiency of an aliphatic lipid substituted 2 kDa polyethylenimine is attributed to enhanced nuclear association and uptake in rat bone marrow stromal cell. J Gene Med 2011;13:46–59. [45] Alshamsan A, Haddadi A, Incani V, Samuel J, Lavasanifar A, Uludag H. Formulation and delivery of siRNA by oleic acid and stearic acid modified polyethylenimine. Mol Pharm 2009;6:121–33.

A. Ewe et al. / Acta Biomaterialia 10 (2014) 2663–2673 [46] Aldawsari H, Edrada-Ebel R, Blatchford DR, Tate RJ, Tetley L, Dufes C. Enhanced gene expression in tumors after intravenous administration of arginine-, lysine- and leucine-bearing polypropylenimine polyplex. Biomaterials 2011;32:5889–99. [47] Garcia L, Bunuales M, Duzgunes N, Tros de Ilarduya C. Serum-resistant lipopolyplexes for gene delivery to liver tumour cells. Eur J Pharm Biopharm 2007;67:58–66. [48] Lee CH, Ni YH, Chen CC, Chou C, Chang FH. Synergistic effect of polyethylenimine and cationic liposomes in nucleic acid delivery to human cancer cells. Biochim Biophys Acta 2003;1611:55–62. [49] Gaedtke L, Pelisek J, Lipinski KS, Wrighton CJ, Wagner E. Transcriptionally targeted nonviral gene transfer using a beta-catenin/TCF-dependent promoter in a series of different human low passage colon cancer cells. Mol Pharm 2007;4:129–39. [50] Hanzlikova M, Soininen P, Lampela P, Mannisto PT, Raasmaja A. The role of PEI structure and size in the PEI/liposome-mediated synergism of gene transfection. Plasmid 2009;61:15–21. [51] Schafer J, Hobel S, Bakowsky U, Aigner A. Liposome-polyethylenimine complexes for enhanced DNA and siRNA delivery. Biomaterials 2010. [52] Heyes J, Palmer L, Chan K, Giesbrecht C, Jeffs L, MacLachlan I. Lipid encapsulation enables the effective systemic delivery of polyplex plasmid DNA. Mol Ther 2007;15:713–20. [53] Ko YT, Bhattacharya R, Bickel U. Liposome encapsulated polyethylenimine/ ODN polyplexes for brain targeting. J Controlled Release 2009;133:230–7. [54] Egle R, Milek M, Mlinaric-Rascan I, Fahr A, Kristl J. A novel gene delivery system for stable transfection of thiopurine-S-methyltransferase gene in versatile cell types. Eur J Pharm Biopharm 2008;69:23–30. [55] Navarro G, Sawant RR, Essex S, Tros de Ilarduya C, Torchilin VP. Phospholipid– polyethylenimine conjugate-based micelle-like nanoparticles for siRNA delivery. Drug Deliv Transl Res 2011;1:25–33. [56] Hobel S, Koburger I, John M, Czubayko F, Hadwiger P, Vornlocher HP, et al. Polyethylenimine/small interfering RNA-mediated knockdown of vascular endothelial growth factor in vivo exerts anti-tumor effects synergistically with Bevacizumab. J Gene Med 2010;12:287–300.

2673

[57] Hobel S, Prinz R, Malek A, Urban-Klein B, Sitterberg J, Bakowsky U, et al. Polyethylenimine PEI F25-LMW allows the long-term storage of frozen complexes as fully active reagents in siRNA-mediated gene targeting and DNA delivery. Eur J Pharm Biopharm 2008;70:29–41. [58] Ibrahim AF, Weirauch U, Thomas M, Grunweller A, Hartmann RK, Aigner A. MicroRNA replacement therapy for miR-145 and miR-33a is efficacious in a model of colon carcinoma. Cancer Res 2011;71:5214–24. [59] Holzer M, Barnert S, Momm J, Schubert R. Preparative size exclusion chromatography combined with detergent removal as a versatile tool to prepare unilamellar and spherical liposomes of highly uniform size distribution. J Chromatogr A 2009;1216:5838–48. [60] Movassaghian S, Moghimi HR, Shirazi FH, Koshkaryev A, Trivedi MS, Torchilin VP. Efficient down-regulation of PKC-alpha gene expression in A549 lung cancer cells mediated by antisense oligodeoxynucleotides in dendrosomes. Int J Pharm 2013;441:82–91. [61] Zintchenko A, Konak C. Interaction of DNA/polycation complexes with phospholipids: stabilizing strategy for gene delivery. Macromol Biosci 2005;5:1169–74. [62] Opanasopit P, Paecharoenchai O, Rojanarata T, Ngawhirunpat T, Ruktanonchai U. Type and composition of surfactants mediating gene transfection of polyethylenimine-coated liposomes. Int J Nanomed 2011;6:975–83. [63] Brus C, Kleemann E, Aigner A, Czubayko F, Kissel T. Stabilization of oligonucleotide–polyethylenimine complexes by freeze-drying: physicochemical and biological characterization. J Controlled Release 2004;95:119–31. [64] Tros de Ilarduya C, Garcia L, Duzgunes N. Liposomes and lipopolymeric carriers for gene delivery. J Microencapsul 2010;27:602–8. [65] Guo W, Lee RJ. Efficient gene delivery using anionic liposome-complexed polyplexes (LPDII). Biosci Rep 2000;20:419–32. [66] Luo X, Feng M, Pan S, Wen Y, Zhang W, Wu C. Charge shielding effects on gene delivery of polyethylenimine/DNA complexes: PEGylation and phospholipid coating. J Mater Sci Mater Med 2012;23:1685–95.