Surface-engineered polyethyleneimine-modified liposomes as novel carrier of siRNA and chemotherapeutics for combination treatment of drug-resistant cancers

Abstract

Modification of nanoparticle surfaces with PEG has been widely considered the gold standard for many years. However, PEGylation presents controversial and serious challenges including lack of functionality, hindered cellular interaction, allergic reactions, and stimulation of IgM production after repetitive dosing that accelerates blood clearance of the nanoparticles. We report the development of novel liposomal formulations surface-modified with a low molecular weight, branched polyethyleneimine (bPEI)-lipid conjugate for use as an alternative to PEG. The formulations had very good stability characteristics in ion- and protein-rich mediums. Protein adsorption onto the liposomal surface did not interfere with the cellular interaction. bPEI-modified liposomes (PEIPOS) showed enhanced association with three different cell lines by up to 75 times compared to plain or PEGylated liposomes and were without carrier toxicity. They also penetrated the deeper layers of 3D spheroids. Encapsulating paclitaxel (PTX) into PEIPOS did not change its main mechanism of action. PEIPOS complexed and intracellularly delivered siRNAs and downregulated resistance-associated proteins. Finally, tumor growth inhibition was observed in a mouse ovarian xenograft tumor model, without signs of toxicity, in animals treated with the siRNA/PTX co-loaded formulation. These complex-in-nature but simple-in-design novel liposomal formulations constitute viable and promising alternatives with added functionality to their PEGylated counterparts.

Keywords: Drug delivery; combination cancer therapy; liposomes; multidrug resistance; polyethyleneimine; siRNA delivery; surface-modification.

Figure 1.

Physical–chemical characterization of PEIPOS. (a–d) Stability of PEIPOS upon incubation in 10% (v/v) FBS in PBS pH 7.4 at 37 °C for over 24h. (a) ZP (mV), (b) particle size (nm), and (c) PdI values at given time points. Error bars represent the standard deviation of three independent sample measurements. (d) Proteins adsorbed in liposomes incubated in 10% FBS at 37 °C overnight. Lane 1 shows FBS diluted to 2.5% (v/v), lanes 2 and 3 are 0% PEIPOS and 0.5% PEIPOS and lane 4 is a PEGylated liposomal formulation (3 mol% PEG2000-PE). (e) Cumulative in vitro release of PTX from bPEI-coated liposomes at different pH under sink conditions. Free PTX groups were used as controls to evaluate the diffusion rate through the release membrane.

Figure 2.

Cellular association of coated and noncoated PEIPOS with monolayers of three different cancer cell lines; (a) HeLa, (b) A2780-ADR, and (c) SKOV-3TR. Lipid concentration used for each cell line is indicated below each graph and incubation time was 4h. Control groups are nontreated cells. * indicates difference from control, whereas # indicates differences within the groups. Mean ± SD, n = 3, **p < .001, ****p < .0001. One-way ANOVA with Tukey’s multiple comparisons test. (d) Mode of internalization by A2780-ADR cells by flow cytometry. Lipid concentration was 100 µg/mL and incubation time was 90 min. PEGylated liposomes were modified with 3 mol% PEG2000-PE. Results are given as the fold-increase over the nonfluorescently labeled control formulations.

Figure 3.

Kinetics of cellular association and penetration of formulations in 3D HeLa spheroids. (a) Rhodamine intensity increase as the indicator of cellular association of liposomes with HeLa cells after disassociation of spheroids into single cells, obtained by flow cytometry analysis. Control groups are nontreated cells, while control liposomes are cells treated with nonlabeled liposomes. n = 3, a total of 15 spheroids, mean ± SD, one-way ANOVA with Tukey’s multiple comparison tests, ****p < .0001. (b) Corrected integrated pixel density values of rhodamine vs. optical section depth as a representation of liposome distribution throughout the spheroids. n = 5, mean ± SD, two-way ANOVA with Sidak’s multiple comparisons, ****p < .0001. (c) A representative HeLa 3D spheroid incubated with 0.5% PEIPOS formulation, imaged by CLSM PMT. Scale bar indicates 500 µm. The 3D reconstruction of the same spheroid using rhodamine intensity confirmed the spheroidal shape of the 3D cell model. (d) Rhodamine-labeled liposome penetration into the spheroids at different layers of depth. The Z-projection was obtained using maximum pixel intensity collected from each layer of the spheroid. MFI, mean fluorescence intensity. Scale bars = 500 µm.

Figure 4.

Cell viability and investigation of PTX course of action on HeLa cells. (a) Cytotoxicity profile of HeLa cells after continuous treatment with the formulations for 48h or 4 + 44h in serum-complete medium. Empty PEIPOS consists of formulations without PTX used at the same lipid concentration as those containing PTX. Data shown indicate triplicate mean ± SD from a blinded experiment. *p < .05, ***p < .0005, ****p < .0001, two-way ANOVA with Tukey’s multiple comparisons test. (b) Immunofluorescent detection of liposomal PTX-mediated? -tubulin polymerization on HeLa cells. Nuclei of the cells were stained with Hoechst (magenta), and? -tubulin structures stained green. (c-f) LSC analysis of HeLa cells treated with different formulations and for different time points. (c) and (d) Cell cycle distribution depending on the gating outlined in figure S2a for different time points, treatments, and PTX concentrations. (e) and (f) Analysis of the cells at their different stages indicated by the nuclei staining with Hoechst, early apoptotic cell staining with YoPro and necrotic cell staining with PI. Scale bars = 20 µm.

Figure 5.

PEIPOS efficacy against MDR ovarian cancer cells and siRNA complexation of surface-modified PEIPOS liposomal formulations. (a) P-gp characterization of the A2780 and A2780-ADR cells using fluorescently-labeled P-gp antibody. The mean fluorescence intensity (MFI) was normalized by isotype control. n = 3, mean ± SD. (b) Cytotoxicity of noncoated and coated PEIPOS formulations on A2780-ADR P-gp overexpressing human ovarian cancer cell line following 4 + 44h treatment scheme in serum-complete media. Data shown indicate triplicate mean ± SD from blinded experiment. *p < .05, ***p < .0005, ****p < .0001, two-way ANOVA with Tukey’s multiple comparisons test. (c) Ethidium bromide stained agarose gel electrophoresis of free siRNA and its complexes with 0.5%PEIPOS at different N/P ratios. (d) P-gp downregulation on A2780-ADR cells after treatment with 0.5%PEIPOS formulation with an N/P 13 ratio. The downregulated protein determination was carried out at the end of 48h incubation following 4h of treatment in serum-complete media. n = 3 independent trials, mean ± SD.

Figure 6.

In vivo evaluation of the formulation efficacy in nude athymic mice bearing A2780-ADR resistant human ovarian tumor xenografts. (a) Antitumor efficacy of liposomal and nonencapsulated PTX formulations on nude athymic mice bearing A2780-ADR resistant human ovarian tumor xenografts. After tumors were established, mice were treated every other day (first injection on day 0) with saline (•), PTX solubilized in Cremophor (?), 0.5%PEIPOS/PTX (?) or 0.5%PEIPOS/PTX/siMDR1(?) at 5.5 mg/kg of PTX and 0.8 mg/kg of siMDR1. (n = 4) Each point represents a mean ± S.D., two-way ANOVA with Dunnett’s multiple comparisons. ****p < .0001 (b) The survival curve of tumor-bearing animals. Mice were treated as indicated in (a) throughout the study period. The survival end-point was when tumors reached 1500 mm³. (c) No significant differences were observed in body weight of the animals. (d), Effect of different treatments on liver function evaluated by quantification of alanine aminotransferase (ALT) levels in blood samples collected from the animals before euthanasia. Only a mild increase in ALT levels in animals treated with 0.5% PEIPOS/PTX/siMDR1 was observed, but no statistical difference was identified among the groups.