Cellular Uptake and Intracellular Pathways of PLL-g-PEG-DNA

Aug 22, 2008 - Tessa Lühmann,†,‡ Markus Rimann,†,‡ Anne Greet Bittermann,§ and Heike Hall*,‡. Department of Materials, ETH Zurich, Zurich,...
5 downloads 0 Views 2MB Size
Bioconjugate Chem. 2008, 19, 1907–1916

1907

Cellular Uptake and Intracellular Pathways of PLL-g-PEG-DNA Nanoparticles Tessa Lu¨hmann,†,‡ Markus Rimann,†,‡ Anne Greet Bittermann,§ and Heike Hall*,‡ Department of Materials, ETH Zurich, Zurich, Switzerland, and ZMB, Center for Microscopy and Image Analysis, University of Zurich, Switzerland. Received May 20, 2008; Revised Manuscript Received July 23, 2008

Polycationic molecules form condensates with DNA and are used for gene therapy as an alternative to viral vectors. As clinical efficacy corresponds to cellular uptake, intracellular stability of the condensates, and bioavailability of the DNA, it is crucial to analyze uptake mechanisms and trafficking pathways. Here, a detailed study of uptake, stability, and localization of PLL-g-PEG-DNA nanoparticles within COS-7 cells is presented, using FACS analysis to assess the involvement of different uptake mechanisms, colocalization studies with markers indicative for different endocytotic pathways, and immunofluorescence staining to analyze colocalization with intracellular compartments. PLL-g-PEG-DNA nanoparticles were internalized in an energy-dependent manner after 2 h and accumulated in the perinuclear region after >6 h. The nanoparticles were found to be stable within the cytoplasm for at least 24 h and did not colocalize with the endosomal pathway. Nanoparticle uptake was ∼50% inhibited by genistein, an inhibitor of the caveolae-mediated pathway. However, genistein did not inhibit gene expression, and PLL-g-PEG-DNA nanoparticles were not colocalized with caveolin-1 indicating that caveolaemediated endocytosis is not decisive for DNA delivery. Clathrin-mediated endocytosis and macropinocytosis pathways were reduced by 17 and 24%, respectively, in the presence of the respective inhibitors. When cells were transfected in the presence of double and triple inhibitors, transfection efficiencies were increasingly reduced by 40 and 70%, respectively; however, no differences were found between the different uptake mechanisms. These findings suggest that PLL-g-PEG-DNA nanoparticles enter by several pathways and might therefore be an efficient and versatile tool to deliver therapeutic DNA.

INTRODUCTION Gene therapy has gained a lot of attention for the treatment of chronic diseases, cancer, and genetic disorders (1-4). It is also considered as a valuable alternative for conventional protein therapy since it overcomes inherent problems that are associated with the administration of protein drugs in terms of bioavailability, systemic toxicity, in ViVo clearance rate, and manufacturing costs (2, 4, 5). For this reason, safe and efficient delivery systems for therapeutic DNA are developed. Polycationic substances such as poly-L-lysine (PLL), poly-L-ornithine, or polyethylenimine (PEI) alone or as graft- and copolymers with different amounts of poly(ethylene glycol) (PEG) have been shown to form complexes with DNA and are widely used as an attractive alternative to viral vectors in gene therapy (2, 6-11). However, not only the DNA-delivery vehicle itself but also cellular uptake, intracellular stability, and bioavailability of the therapeutic DNA within the cell are crucial for clinical relevance. Therefore, uptake mechanisms of DNA-containing vehicles and trafficking pathways within the cells were analyzed (3, 12-14). Unfortunately, (nano)-particle uptake and intracellular pathways seem to be dependent on the analyzed cell type (15) as well as on the particular size, shape, charge, and chemistry of the delivery vehicle (16, 17). Especially, when certain cell types or specific uptake pathways are targeted, the uptake mecha* Corresponding author. ETH Zurich, Department of Materials, HCI E415, Cells and BioMaterials, Wolfgang-Pauli-Strasse 10, CH-8093 Zu¨rich, Switzerland Phone: +41 44 633 69 75. Fax: +41 44 632 10 73. E-mail: [email protected]. † These authors contributed equally to this work and should be regarded as first authors. ‡ ETH Zurich. § University of Zurich, Switzerland.

nism(s) and intracellular pathways need to be analyzed carefully (1, 3, 4, 17). Therefore, rational design of PLL-gPEG-DNA nanoparticles might benefit from the knowledge raised about the uptake pathways involved and the release mechanisms of therapeutic DNA in order to optimize their performance. The analysis of gene delivery vehicles is still a complex challenge, as endocytosis may involve several pathways (3, 18-21). The first barrier is to overcome the plasma membrane, followed by cytoplasmatic pathways prior to DNA delivery into the nucleus where gene expression takes place. So far, five major uptake mechanisms are distinguished (14, 18, 21, 22) including phagocytosis that is performed only by specialized cells such as macrophages, neutrophils, or monocytes to eliminate debris and pathogens. In this study, four pinocytotic pathways are studied to assess cellular uptake and intracellular pathway(s) of PLL-g-PEG-DNA nanoparticles: (i) macropinocytosis that is stimulated by growth factors or phorbol esters, which cause actin-driven formation of membrane ruffles that fuse with the plasma membrane forming large endocytotic vesicles known as macropinosomes (14, 23). They either fuse with lysosomes or are recycled back to the cell surface. Large particles up to 1-5 µm in diameter can be internalized by macropinocytosis (14, 22). (ii) Clathrin-mediated endocytosis is a process involving specific receptors that recognize and internalize cargo (24). Ligand-bound receptors are internalized via clathrin-coated vesicles involving GTPase activity and are transformed to early endosomes. Receptor-ligand complexes are separated by low pH generated by vacuolar proton ATPases (25). The ligands are transferred into late endosomes for potential degradation in lysosomes, whereas the receptors are recycled and transported back to the plasma membrane (19). (iii) Caveolae-mediated endocytosis is characterized by the evolution of caveolae, which

10.1021/bc800206r CCC: $40.75  2008 American Chemical Society Published on Web 08/22/2008

1908 Bioconjugate Chem., Vol. 19, No. 9, 2008

are associated with caveolin-1/caveolin-2 (or caveolin-3 in the muscle) (26). Caveolae derive from subdomains of sphingolipidand cholesterol-rich cell membrane fractions called lipid rafts (27). Caveolae are considered to be static organelles, whose internalization is induced by specific ligands such as simian virus-40 (SV40) or cholera toxin (18, 28, 29). After internalization, those ligands are either transferred into caveosomes having neutral pH that translocate to the endoplasmic reticulum or the Golgi complex (30) or they enter the endosomal pathway (31). (iv) This pathway is described as clathrin- and caveolaeindependent, transporting cargo to the GPI-anchored proteinenriched early endosomal compartment (GEEC) and ending in the Golgi complex or in recycling endosomes (18). Glycosylphophatidylinositol-anchored proteins (GPI-AP) can be internalized via 40-50 nm lipid rafts (18). But regulatory mechanisms for this pathway are still unknown. A recent report from our group showed evidence that PLLg-PEG-DNA nanoparticles are a promising tool for effective transport and delivery of therapeutic DNA as they show longterm stability, a hydrodynamic diameter of 80-90 nm, and high transfection efficiency at a N/P ratio 3.125 of ∼40% combined with low cytotoxicity (>95% of cell viability) in COS-7 cells (11). In order to understand the uptake mechanism(s) and intracellular pathway(s) of PLL-g-PEG-DNA nanoparticles, PLL-g-PEG and plasmid DNA were differentially fluorescently labeled, and uptake into COS-7 cells was determined by uptakeinhibition analysis. Furthermore, transfected COS-7 cells were analyzed after different time points for colocalization with markers for individual uptake mechanisms, and immunofluorescence staining of cellular compartments was performed. This study may be helpful in the future design of PEGylated PLLpolymers for gene delivery approaches to identify the most efficient uptake and delivery mechanisms.

MATERIALS AND METHODS Polymer Synthesis. The polymers were synthesized using poly-L-lysine hydrobromide with a molecular weight of 20 kDa (Sigma), referred to as PLL20 and an N-hydroxysuccinimidyl ester of methoxy-terminated poly(ethylene glycol) (Nektar, USA, MW 5 kDa) as described previously (11, 32). FITClabeled PLL-g-PEG polymer was synthesized in an analogous manner, using 20% FITC-terminated poly(ethylene glycol) N-hydroxy succinimidyl ester (Nektar, USA, MW 5 kDa) and 80% N-hydroxysuccinimidyl ester of methoxy-terminated poly(ethylene glycol). Grafting (g) was calculated using the absolute number of PEG-moieties per lysine of the PLL backbone and was about 5% for all the polymers. PLL20-g5-PEG5 or PLL20-g5-PEG5-FITC was used for all experiments and is referred to as PLL-g-PEG or PLL-g-PEG-FITC, respectively. Reaction rates were determined by 1H NMR (D2O) (β, γ, and δ CH2 peaks of lysine at 1-2 ppm, CH2CH2O of PEG at 3.3-3.9 ppm, and the FITC-group at 6.9-7.2 ppm). The polymers were stored in PBS at -20 °C until further use. Formulation of PLL-g-PEG-DNA Condensates. PLL-gPEG-DNA condensates were formed using a plasmid expressing enhanced green fluorescent protein (pEGFP-N1, Clontech) at a final DNA concentration of 1 µg/150 µL () 6.7 µg/mL) in PBS (0.01 M phosphate buffered saline, pH 7.4, Sigma) and a PLLg-PEG concentration according to the NH4+/PO43- (amino group in PLL/phosphate group in DNA; N/P) charge ratio of 3.125. For a 24-well plate, PLL-g-PEG was provided in 75 µL of PBS, and the DNA dissolved in 75 µL of PBS was added dropwise into the polymer solution. The complexes were equilibrated for 30 min at room temperature (RT) and were always freshly prepared prior to use. Plasmid Propagation. The vector pEGFP-N1 used in this study was transformed into E. coli strain Top10 (Invitrogen)

Lu¨hmann et al.

and grown in LB medium supplemented with 30 µg/mL kanamycin. The plasmid DNA was purified using a Maxiprep kit (Qiagen), and the concentration was determined spectrophotometrically at 260 nm. DNA Labeling. Plasmid DNA pEGFP-N1 was covalently labeled using LabelIT nucleic acid labeling kits (Mirus, Cat. No.: MIR3100 (CX-rhodamine) and MIR3700 (Cy5), respectively) according to the manufacturer’s protocol. For colocalization studies, the DNA was labeled with CX-rhodamine (DNA-CX-rh), and for flow cytometry analysis, the plasmid was labeled with Cy5 (DNA-Cy5). Cells and Cell Culture. COS-7 (SV 40 transformed kidney cells of African green monkey, ATCC CRL 1651, ATCC Rocksville, MA) cells were cultured as exponentially growing subconfluent monolayers at 37 °C and 5% CO2. The cell culture was kept in 25 cm2 culture flasks in DMEM (Cat. No.: 21885, low glucose, Glutamax I, GibcoBRL) supplemented with 10% heat inactivated FBS (Cat. No.: F7524, Sigma) and 1% ABAM (Cat. No.: 15240, GibcoBRL). For experiments, cells were used 24 h post seeding. Temperature-Dependent Nanoparticle Uptake. In order to determine energy-dependency of nanoparticle uptake, COS-7 cells were transfected with DNA nanoparticles at either 37 or 4 °C. COS-7 cells were seeded 24 h prior to transfection in 24 well plates (10,000 cells/well). The growth medium then was replaced with 150 µL of PLL-g-PEG-DNA nanoparticles in PBS (per well: 1 µg plasmid pEGFP-N1, N/P ratio ) 3.125), and the cells were incubated for 30 min, either at 37 °C or at 4 °C (here, in all the subsequent steps, the solutions were prechilled to 4 °C). Then, 400 µL of 2% serum-containing DMEM was added, and the cells were again incubated at either 37 or 4 °C for an additional 3 h. Finally, the cells were extensively washed 3× with PBS at the corresponding temperatures before DMEM supplemented with 10% FBS at 37 °C was added. After 48 h of incubation at 37 °C for both conditions, the cells were analyzed for GFP expression. To identify the total cell number, Hoechst 33342 nuclear stain (Invitrogen) was used, and ethidium bromide (AppliChem) visualized potential cytotoxicity. Briefly, the cells were incubated for 2 min in PBS containing 4 µg/mL Hoechst 33342 and 10 µg/mL ethidium bromide. After rinsing the cells three times with PBS, the transfection efficiency was analyzed by counting the GFP-expressing cells in 3 nonoverlapping areas of 500 µm2 in relation to the total cell number as determined by the Hoechst nuclear stain. In order to analyze the data, ImageJ 1.38 was used. In order to visualize nanoparticle uptake at 4 and 37 °C, 8000 COS-7 cells were seeded in 4 well chamber slides (Lab-Tek Chambered #1.0 Borosilicate Cover glass System, Nunc) and cultured for 16 h in DMEM containing 10% FBS. Complexes were formed by using PLL-g-PEG-FITC and DNA-CX-rh (1 µg/well) at N/P ratio 3.125 and were incubated with the cells for 3.5 h at either 4 °C or 37 °C as mentioned above. After 3× PBS washing, the cells were Hoechst-stained. Finally, cells were fixed with 1% (w/v) paraformaldehyde (PFA, Sigma) in PBS for 30 min at RT and analyzed using a high-resolution TCSSP5 laser scanning confocal microscope (Leica Microsystems, Mannheim, Germany). To avoid cross talk, emission signals were collected independently. Image processing was performed using IMARIS software, version 6.0.0 (Bitplane AG, Zurich, Switzerland). Colocalization Studies of Nanoparticles Formed by PLLg-PEG-FITC Polymer and DNA-CX-rh. PLL-g-PEG-FITC was first confirmed to have transfection efficiency similar to that of nonlabeled PLL-g-PEG. Eight thousand COS-7 cells were seeded in 4 well chamber slides and cultured for 16 h in DMEM containing 10% FBS as indicated above. Complexes were formed by using PLL-g-PEG-FITC and DNA-CX-rh (1 µg/well)

Intracellular Pathways of PLL-g-PEG Nanoparticles

at N/P ratio 3.125 as described above and were applied on the cells for 30 min and 2, 6, and 24 h. DMEM with 2% FBS was added 30 min after nanoparticle addition to guarantee cell viability. After the indicated time points, noninternalized nanoparticles were removed by rinsing the cells with PBS. Hoechst 33342 nuclear stain was performed and cells were rinsed 3× with PBS. Finally, cells were fixed with 1% (w/v) paraformaldehyde (PFA, Sigma) in PBS for 30 min at RT and analyzed by CLSM. Optimization of Endocytosis Inhibitor Concentrations. In order to determine the most efficient concentrations of several endocytosis inhibitors without decreasing cell viability, a WST-1 proliferation assay (Roche) was performed in 48 well plates (TPP, Switzerland). The following inhibitors have been used: 150 µM-10 mM methyl-β-cyclodextrin (Sigma, Cat. No.: C4555), 200 µM genistein (Sigma Cat. No.: 91955), 2 µM-60 µM chlorpromazine (Sigma, Cat. No.: C8138), and 50 nM-20 µM wortmannin (Sigma, Cat. No.: 95455). As positive control, cells were treated with PLL-g-PEG-DNA nanoparticles in the absence of any inhibitors. Eight thousand COS-7 cells/48 well were seeded 24 h prior to the addition of inhibitors. Afterwards, the cell culture medium was replaced with 270 µL of serum free DMEM containing the different inhibitors, and the cells were incubated at 37 °C for 1 h. Then, freshly prepared 30 µL PLL-g-PEG-DNA nanoparticles in PBS (per well: 0.4 µg plasmid DNA; N/P ratio ) 3.125) were added and incubated for another 3 h. Finally, the cells were washed 3× with PBS before they were incubated in 300 µL of growth medium for 48 h. Subsequently, 30 µL of WST-1 reagent was applied to each well, and the cells were incubated for another hour at 37 °C and 5% CO2. Absorption at 440 nm was measured using a microplate reader Infinite M200 (Tecan). Each condition was performed in quadruple and repeated three times. DNA Nanoparticle Uptake in the Presence of Endocytosis Inhibitors (FACS Analysis). Twenty-four hours prior to the experiment, 50,000 COS-7 cells were seeded into a 6 well plate and cultured as describes above. The growth medium was replaced with 1 mL of serum free medium containing the different inhibitors using the optimized concentrations. The cells were incubated for 1 h. Subsequently, the medium was replaced with 500 µL of PBS containing PLL-g-PEG-DNA-Cy5 nanoparticles (per well: 1 µg DNA-Cy5; N/P ratio ) 3.125) in addition to the inhibitors as described above. The cells were incubated for 30 min, and then an additional 1.5 mL serum free DMEM medium containing the inhibitors was added. After a total inhibitor incubation time of 4 h, the cells were rinsed twice with PBS before they were detached with 500 µL of trypsinEDTA (0.25%, GibcoBRL), resuspended, and incubated for 10 min in PBS containing 10 mg of heparin (H9399, Sigma) at RT to detach cell-surface-bound DNA-containing nanoparticles. Subsequently, the cells were washed twice with PBS before they were transferred into tubes (REF 352054, Falcon) for flow cytometry analysis. The analysis was carried out with a FACSCalibur instrument and CellGate software (BD Biosciences). Dead cells and debris were gated out by forward and side scatter spectra. The mean fluorescence intensity of 10,000 individual cells was determined. All experiments have been performed in duplicate and were repeated three times. As controls, COS-7 cells were transfected with DNA-Cy5-nanoparticles in the absence of inhibitors, and naked DNA-Cy5 was used. Transfection Studies with Endocytosis Inhibitors. Eight thousand COS-7 cells were seeded for 24 h in 24 well plates and incubated for 1 h with inhibitors dissolved in 200 µL of serum free DMEM. The optimized inhibitor concentrations were used as follows: methyl-β-cyclodextrin (5 mM), genistein (200 µM), chlorpromazine (10 µM), and wortmannin (10 µM). Later,

Bioconjugate Chem., Vol. 19, No. 9, 2008 1909

the inhibitor solutions were removed, and 150 µL of freshly prepared PLL-g-PEG-DNA nanoparticles in PBS (per well: 1 µg pEGFP-N1; N/P ratio ) 3.125) containing the same inhibitor concentrations as those mentioned above were added and further incubated for 30 min. Then, 400 µL of serum free medium supplemented with inhibitors was added and the incubation time extended for an additional 2.5 h. Subsequently, the cells were washed 3× with PBS before 10% serum-containing medium was added to further incubate the cells for 48 h. Double and triple inhibitor transfection experiments were performed in an analogous manner. In order to determine cell viability of the multiple inhibitors used, a WST-1 proliferation assay was carried out in parallel as described above. Finally, the cells were analyzed for GFP expression by fluorescence microscopy as mentioned before. As positive control, DNA nanoparticles without inhibitors were used to transfect the cells. Nanoparticle Distribution in the Presence of Colocalization Markers. Eight thousand COS-7 cells were seeded in 4 well chamber slides (Lab-Tek Chambered #1.0 Borosilicate Cover glass System, Nunc) and cultured as described before. After washing the cells with PBS, the condensates, consisting of 1 µg of plasmid DNA (pEGFP-N1) complexed with PLLg-PEG-FITC at N/P ratio 3.125, were coincubated with either cholera toxin subunit B (10 µg/mL) labeled with Alexa Fluor 594 (Molecular Probes, AF-ctB) or Texas Red-labeled transferrin (50 µg/mL: Molecular Probes, TR-tf) for 30 min. DMEM supplemented with 2% FBS was added to a final volume of 500 µL per chamber slide. Additionally, cells were first incubated with FITC-labeled condensates for 1.5 h before adding AF-ctB (10 µg/mL) or TR-tf (50 µg/mL). For both experimental setups, coincubation of the nanoparticles with the colocalization markers or when colocalization markers were added after 1.5 h, the nuclei were Hoechst-stained after 2 h total incubation time. Then, cells were washed 3× with PBS and fixed in 1% (w/v) PFA in PBS for 30 min at RT. Cells were washed again with PBS and analyzed by CLSM as described above. Immunofluorescence of TFR, Caveolin-1, GM1, and EEA-1. Eight thousand COS-7 cells were seeded into 4 well chamber slides and cultured for 16 h prior to incubation with PLL-g-PEG-DNA-CX-rh (DNA-CX-rh ) 1 µg/well) condensates at N/P ) 3.125. Cells were stained with antibodies against transferrin receptor (TFR), caveolin-1, and GM1 after 30 min and 45 min, and 1 and 2 h. The cultures were stained against early endosome antigen-1 (EEA-1) after 1, 2, 4, and 6 h. After the indicated time points, the cell culture medium was removed, the cells were washed 3× with PBS and fixed in 1% (w/v) PFA in PBS for 30 min at RT followed by rinsing 3× with PBS. Cells were permeabilized with 0.1% Triton X-100 in PBS for 10 min and washed with PBS. Subsequently, the cells were blocked with 5% donkey serum (Chemicon) and 2% BSA (Sigma) in PBS for at least 1 h. Polyclonal rabbit antibodies against TFR (Chemicon), caveolin-1 (Upstate), ganglioside GM1 (Calbiochem), and mouse against EEA-1 (early endosome antigen-1, BD-Transduction Laboratories) were diluted 1:100 in 1% BSA in PBS and either incubated for 1 h at RT or overnight at 4 °C. Each well was washed intensively 3× with PBS, before secondary antibodies donkey antimouse-FITC and donkey anti-rabbit-FITC (both from Chemicon) were added for 1 h (dilution 1:200 in 5% donkey serum in PBS). Cell nuclei were stained with DAPI (1 mg/mL, dilution 1:1000, Molecular Probes) for 10 min. The cells were washed 3× with PBS and analyzed by CLSM as described above. Statistical Analysis. The mean values were compared by oneway ANOVA analysis using Origin 7.5 software. Statistical

1910 Bioconjugate Chem., Vol. 19, No. 9, 2008

Lu¨hmann et al.

Figure 1. Temperature-dependent nanoparticle uptake into COS-7 cells. (A) Representative confocal images of COS-7 cells, which were transfected with PLL-g-PEG-FITC and CX-rhodamine-labeled pEGFP-N1 (DNA-CX-rh) nanoparticles for 3.5 h at 37 °C (a, b, and c) or 4 °C (d, e, and f). b and e display overlay images of all sections, and c and f show a section view. Uptake of PLL-g-PEG-FITC without DNA cargo at 37 and 4 °C, respectively, is demonstrated in a and d. Blue indicates Hoechst-stained nuclei, red shows CX-rhodamine-labeled DNA, and green displays PLLg-PEG-FITC polymer. Scale bar is 10 µm. (B) PLL-g-PEG-DNA (plasmid DNA: pEGFP-N1) nanoparticles were applied on COS-7 cells and incubated for 3.5 h at either 37 or 4 °C. Transfection efficiency was analyzed after 48 h by counting transfected cells in relation to the total number of cells. The experiments indicate mean values ( standard deviation of three independent experiments carried out in triplicate. The values are normalized to particle uptake at 37 °C.

significance was accepted for p < 0.05 after comparing the mean values by Bonferroni test.

RESULTS PLL-g-PEG-DNA Nanoparticles are Internalized by Endocytosis and Remain Stable within the Cytoplasm. As nonviral gene delivery systems are promising tools for gene therapy approaches, it is important to analyze nanoparticle internalization and trafficking pathways within the cell. This is especially important when comparing to viral gene delivery systems that benefit from their inherent high cell transduction efficiency. Information on particle uptake mechanisms and intracellular pathways will lead to improved design of PLL-gPEG-DNA nanoparticles for further applications. As PLL-g-PEG-DNA nanoparticle stability is maintained by ionic interactions between the PLL backbone of positively charged amines in PLL-g-PEG to negatively charged phosphate groups of the DNA, it is crucial to gain insight into nanoparticle uptake kinetics and intracellular stability. Therefore, PLL-gPEG and DNA were covalently labeled with two different fluorescent labels. An FITC-labeled PEG-moiety was incorporated during PLL-g-PEG synthesis to obtain PLL-g-PEG-FITC. As determined by NMR, 1 FITC per PLL-g-PEG polymer was attached per PLL-g-PEG-FITC (data not shown). The plasmid DNA harboring the sequence of green fluorescent protein (pEGFP-N1) was covalently labeled at amino groups with CXrhodamine to maintain condensate formation. PLL-g-PEG-FITCDNA-CX-rh nanoparticles that were transfected into COS-7 cells showed similar transfection efficiency compared to that of the unlabeled condensates (data not shown). Initial experiments were performed to determine the energydependent dependency of PLL-g-PEG-DNA nanoparticle uptake as endocytosis can be strongly inhibited by lowering the temperature (19). For this purpose, fluorescent PLL-g-PEGDNA nanoparticles were prepared, and COS-7 cells were transfected at 4 and 37 °C (Figure 1). Representative confocal images taken 3.5 h after nanoparticle exposure at 4 and 37 °C show that at 37 °C nanoparticles were found in the cytoplasm of the cell (Figure 1A, b and c). In c, an optical cross section indicated that the nanoparticles were indeed found within the cytoplasm. When only fluorescently labeled PLL-g-PEG was incubated with the cells (Figure 1A, a), it was also found within the cytoplasm close to the nucleus. In contrast, lowering the temperature strongly reduced nanoparticle uptake as they

remained at the cell membrane (Figure 1A, e and f). This behavior was also seen for PLL-g-PEG-FITC (Figure 1A, d). These findings were quantified; at 4 °C, transfection efficiency was found to be only 25% as compared to transfection efficiency at 37 °C. These findings suggest that PLL-g-PEG enters the cytoplasm by energy-dependent uptake processes and that it enters independent of the DNA-cargo. Close colocalization of PLL-g-PEG-FITC and DNA-CX-rh resulted in yellow nanoparticles that could be detected. Uptake studies were performed and COS-7 cells were analyzed after 30 min and 2, 6, and 24 h of nanoparticle incubation. Figure 2A-D shows representative confocal images of fluorescentlabeled nanoparticles (DNA-CX-rh, PLL-g-PEG-FITC). After 30 min of nanoparticle incubation, the condensates were poorly incorporated into COS-7 cells as only few fluorescent particles were observed in association with the plasma membrane. After 2 and 6 h, nanoparticle uptake was enhanced indicated by the increase in the number of yellow particles. Initially, the nanoparticles were homogenously distributed within the cytoplasm (2 h), whereas after more than 6 h, nanoparticles accumulated in the perinuclear region of the cell. Optical crosssections of transfected cells (Figure 2E and F) show that after 2 h, the nanoparticles were clearly localized within the cytoplasm and remained closely associated with the nucleus for more than 24 h. When COS-7 cells were exposed to DNA-CXrh or PLL-g-PEG-FITC for 2 h, DNA-CX-rh did not enter the cells, whereas PLL-g-PEG-FITC was found to be distributed within the entire cytoplasm (Figure 2G and H, respectively). Analysis of PLL-g-PEG-DNA Nanoparticle Uptake into COS-7 Cells. Uptake of PLL-g-PEG-DNA nanoparticles was studied by using different endocytosis inhibitors. Figure 3A lists endocytosis inhibitors that were used to inhibit their respective target pathways. As the effective range of inhibitors is very narrow, cell viability assays were performed 48 h after inhibitor treatment to determine efficient but not toxic inhibitor concentrations (Figure 3C). The inhibitor concentrations used here showed more than 85% cell viability compared to untreated COS-7 cells. To quantify nanoparticle uptake into COS-7 cells, FACS analysis was carried out (Figure 3B). PLL-g-PEG-DNACy5 nanoparticles were individually coincubated with different substances, and inhibition of nanoparticle uptake was determined. As controls, fluorescent-labeled nanoparticles and DNACy5 in the absence of inhibitors were used. Genistein, an inhibitor of caveolae-mediated endocytosis, reduced nanoparticle

Intracellular Pathways of PLL-g-PEG Nanoparticles

Bioconjugate Chem., Vol. 19, No. 9, 2008 1911

Figure 2. Colocalization of PLL-g-PEG-FITC and CX-rhodamine-labeled pEGFP-N1 (DNA-CX-rh) in COS-7 cells. Nanoparticles were prepared and applied on COS-7 cells for 30 min and 2, 6, and 24 h. After the indicated time points, cells were fixed and analyzed by confocal microscopy. Images A-D show COS-7 cells incubated with nanoparticles for the indicated times, whereas images E and F display section views of cells incubated for 2 and 24 h, respectively. G and H indicate uptake of PLL-g-PEG-FITC and DNA-CX-rh, respectively. Blue indicates Hoechst-stained nuclei, green shows PLL-g-PEG-FITC, red displays DNA-CX-rh, and yellow indicates tight association of labeled polymer and plasmid DNA. Scale bar is 10 µm.

uptake to 43 ( 5%. Whereas all other inhibitors reduced nanoparticle uptake only a little, wortmannin by 17 ( 3% and chlorpromazine by 3 ( 2%, indicating the involvement of several pathways, 74 ( 2% of the nanoparticles were internalized when methyl-β-cyclodextrin was applied. As nanoparticle uptake is not indicative for gene expression, transfection experiments in the presence of single inhibitors were performed (Figure 3D). PLL-g-PEG-DNA nanoparticles were coincubated with the respective inhibitors, and GFP expression was analyzed. In the presence of wortmannin and methyl-βcyclodextrin, transfection efficiency was reduced by 35 ( 18% and 30 ( 19%, respectively, whereas in the presence of genistein or chlorpromazine, transfection efficiency was not reduced significantly. In order to further analyze if multiple pathways are involved in PLL-g-PEG nanoparticle uptake, double and triple inhibitor experiments were performed. Cell viability with double inhibitors was found to be above 90% (Figure 3E), whereas triple inhibitors decreased cell viability by ∼20% (Figure 3G). Double inhibitors showed a reduction in transfection efficiency to ∼70% (Figure 3F). The transfection efficiency with triple inhibitors was reduced to ∼30% in comparison to that of control cells (Figure 3H). Interestingly, no inhibitor combination used was able to completely block nanoparticle uptake and gene expression, indicating that PLL-g-PEG-DNA nanoparticles might use several uptake mechanisms. Analysis of PLL-g-PEG-DNA Nanoparticle Pathways within COS-7 Cells. To visualize intracellular pathways of the PLL-g-PEG-DNA nanoparticle within COS-7 cells, PLL-g-PEG-

DNA nanoparticles were coincubated with known endocytosis ligands cholera toxin B and transferrin. The nontoxic subunit B of cholera toxin was used to study lipid raft-mediated endocytosis as cholera toxin B binds to membrane sphingolipid GM1 to enter a cell (29, 33). Transferrin is internalized by receptor-mediated endocytosis forming clathrin-coated pits and is widely used as a tracker for clathrin-mediated endocytosis. COS-7 cells were incubated for 2 h with PLL-g-PEG-FITCDNA nanoparticles and endocytosis markers Alexa Fluor-594labeled cholera toxin B (AF-ctB) and Texas Red-labeled transferrin (TR-tf) were either coincubated with nanoparticles for 2 h or COS-7 cells were preincubated for 1.5 h with nanoparticles prior to the addition of transferrin or cholera toxin B for 30 min (Figure 4). When COS-7 cells were preincubated for 1.5 h with the nanoparticles prior to the addition of cholera toxin B or transferrin (Figure 4A and C), cholera toxin B was found attached to the plasma membrane, whereas transferrin was already internalized and localized close to the nucleus. PLLg-PEG-FITC-DNA nanoparticles were internalized, but only few particles colocalized with the endocytosis markers as indicated by a few yellow particles. When COS-7 cells were coincubated with PLL-g-PEG-FITC-DNA nanoparticles and endocytosis markers for 2 h (Figure 4B and D), cholera toxin B was found in the nucleus not colocalized with PLL-g-PEG-FITC-DNA nanoparticles. Also transferrin and PLL-g-PEG-FITC-DNA nanoparticles were not colocated within COS-7 cells. To support the findings obtained with colocalization studies, immunofluorescence stainings with endocytosis markers GM1,

1912 Bioconjugate Chem., Vol. 19, No. 9, 2008

Lu¨hmann et al.

Figure 3. Inhibition of PLL-g-PEG-DNA nanoparticle uptake. (A) Endocytosis inhibitors and their targets used. (B) FACS analysis of COS-7 cells incubated with nanoparticles in the presence of four different endocytosis inhibitors. The graph displays mean fluorescence intensities of one of three independent experiments performed in duplicate. As controls, PLL-g-PEG-DNA-Cy5 without inhibitors, plasmid DNA only, and untreated COS-7 cells were used (not shown). Cell viability as determined by the WST-1 proliferation assay performed in the presence of single (C), double (E), and triple (G) endocytosis inhibitors. Transfection efficiencies after coincubation with single (D), double (F), and triple (H) endocytosis inhibitors. In D, F, and H, unlabeled pEGFP-N1 was used for nanoparticle production. The values were normalized to transfection efficiencies in the absence of any inhibitor (control). The values shown here represent mean values ( standard deviation from two to five independent experiments performed in duplicate. * refers to values that are statistically different from the control (p < 0.05).

Intracellular Pathways of PLL-g-PEG Nanoparticles

Figure 4. Colocalization studies with PLL-g-PEG-FITC-DNA nanoparticles and endocytosis markers Alexa Fluor-594-labeled cholera toxin B (AF-ctB) and Texas Red-labeled transferrin (TR-tf). In A and C, COS-7 cells were preincubated with PLL-g-PEG-FITC-DNA nanoparticles for 1.5 h, before cholera toxin B or transferrin was added for 30 min. After a total of 2 h, the cells were analyzed. In B and D, cholera toxin B and transferrin were coapplied with the nanoparticles for an entire 2 h prior to analysis. The cells were analyzed by CLSM. Blue indicates Hoechst-stained nuclei, green shows PLL-g-PEG-FITC, and red displays colocalization markers for cholera toxin B (in A and B) and transferrin (in C and D). Scale bar is 10 µm.

transferrin receptor (TFR), and caveolin-1 were performed. Ganglioside GM1, is mainly found in lipid rafts indicative for clathrin-independent pathways, TFR is involved in clathrinmediated endocytosis, and caveolin-1 represents the main component in caveolae-mediated endocytosis. As shown in Figure 2, the majority of nanoparticles were located within the cytoplasm after 2 h; therefore, colocalization of PLL-g-PEGDNA nanoparticles with GM1, TFR, and caveolin-1 was analyzed after 30 and 45 min, and 1 and 2 h. In addition, it was important to find out whether PLL-g-PEG-DNA nanoparticles would enter the lysosomal pathway. Therefore, colocalization with EEA-1 (early endosome antigen-1) was analyzed after 1, 2, 4, and 6 h to guarantee sufficient time for sorting into the endosomal pathway. Most PLL-g-PEG-DNA nanoparticles visualized by DNA-CX-rh did not colocalize with GM1 (Figure 5A-D). They also did not colocalize with transferrin receptors (Figure 5E-H) nor with caveolin-1 (Figure 5I-L). Moreover, PLL-g-PEG-DNA nanoparticles were not localized in early endosomal compartments (Figure 5M-P). These findings essentially confirm the inhibition of uptake and colocalization experiments in that internalized PLL-g-PEG-DNA nanoparticles do not strictly follow one of the main uptake mechanisms. Moreover, they avoid endosomal degradation within COS-7 cells that might lead to fast clearance and inefficient transfection.

DISCUSSION Knowledge about uptake mechanisms and intracellular pathways of nonviral gene delivery vehicles is of great interest for the design of high performance delivery systems for therapeutic DNA. As PLL-g-PEG-DNA nanoparticles were previously identified as potent gene delivery vehicles in COS-7 cells that combine small size (80-90 nm), long storage stability, and low cytotoxicity with high transfection rates (11), this study aimed at elucidating the uptake mechanisms and intracellular pathways of these particles. Cargo uptake is usually energy-dependent

Bioconjugate Chem., Vol. 19, No. 9, 2008 1913

and leads to the formation of membranous vesicles (3, 4, 14, 18, 22). PLL-g-PEG-DNA nanoparticles were shown to enter COS-7 cells by energy-dependent mechanisms as reduction of the temperature to 4 °C reduced nanoparticle uptake as shown by confocal imaging. Consequently, transfection efficiency was decreased by 75% at 4 °C. The nanoparticles were shown to enter the cytoplasm within the first 2 h of transfection; later, PLL-g-PEG-DNA nanoparticles accumulate in the perinuclear region preceding nuclear uptake. Furthermore, PLL-g-PEGDNA nanoparticles were found within the cytoplasm at least for 24 h, and no colocalization with endosomal compartments, as indicated by fluorescence staining against early endosome antigen-1, was observed. These experiments indicate that PLLg-PEG-DNA nanoparticles translocate efficiently to the nucleus and eventually enter the nucleus, e.g., during mitosis, to express the gene of interest. Most studies in the field of particle uptake focused on pathways via caveolin- or clathrin-coated vesicles and cholesterol- and ganglioside GM1-enriched microdomainmediated endocytosis (14, 34-36). Therefore, our studies were based on such methodology to classify the intracellular fate of PLL-g-PEG-DNA nanoparticles. Inhibition of uptake experiments were performed in the presence of optimized single inhibitor concentrations of wortmannin (macropinocytosis), chlorpromazine (clathrin-mediated), genistein (caveolae-mediated), and methyl-β-cyclodextrin (caveolae-mediated and caveolae- and clathrin- independent pathways) and with double and triple inhibitor combinations. These experiments revealed that only genistein was able to substantially inhibit (∼55%) PLL-g-PEG-DNA nanoparticle uptake into COS-7 cells. However, transfection efficiency performed in the presence of genistein did not show any reduction. Also, colocalization studies with antibodies against caveolin-1 indicate no correlation between caveolin-1 and PLL-g-PEG-DNA nanoparticles. In addition, caveolae-mediated endocytosis was described to be very fast (>20 min), and the vesicles produced are usually smaller (50-80 nm diameter) (22) than the diameter of PLLg-PEG-DNA nanoparticles. Therefore, all findings suggest that these nanoparticels do not follow caveolae-mediated pathways and that the inhibitory effects of genistein might be the result of genistein side reactions. It is known that genistein acts as an ATPase inhibitor (37) as well as a hormone analogue (38) that might affect nanoparticle uptake in one way or another. Similar findings with respect to genistein activity were shown when COS-7 cells were transfected with cationic polymer pDMAEMA and PEI nanoparticles (21). Clathrin-mediated pathways might also be an option for PLLg-PEG-DNA nanoparticle uptake as particles up to 120 nm in diameter can be internalized (22). Here, no correlation between nanoparticle uptake and markers indicative for clathrin-mediated pathways could be demonstrated. Nanoparticle uptake was not inhibited in the presence of chlorpromazine, nor was transfection efficiency in the presence of chlorpromazine impaired. Moreover, no colocalization was found for PLL-g-PEG-DNA nanoparticles with transferrin, transferrin receptors, and EEA-1. Also macropinocytotic and caveolae- and clathrin-independent uptake mechanisms could not be assigned as primary uptake pathways for PLL-g-PEG-DNA nanoparticles. Particle uptake and transfection could be inhibited in the presence of wortmannin or methyl-β-cyclodextrin by 17 and 24%, respectively, whereas colocalization between cholera toxin subunit B and PLL-g-PEG-DNA nanoparticles as well as the membrane ganglioside GM1 with the nanoparticles failed. The conclusion is that these PLL-g-PEG-DNA nanoparticles do not strictly follow one of the classical endocytotic pathways, although some inhibition of endocytosis can be observed for several pathways, and transfection efficiencies were reduced when combinations of inhibitors were used (Figure 6). These

1914 Bioconjugate Chem., Vol. 19, No. 9, 2008

Lu¨hmann et al.

Figure 5. Colocalization studies with PLL-g-PEG-DNA-CX-rh-labeled nanoparticles and markers indicative for different endocytosis pathways. PLL-g-PEG-DNA-CX-rh nanoparticles were added to COS-7 cells, and at indicated time points, cells were fixed and immunostained against GM1 (ganglioside GM1, A-D), transferrin receptor (TFR, E-H), caveolin-1 (I-L), and early endosome antigen-1 (M-P). The cells were analyzed by CLSM. Blue indicates Hoechst-stained nuclei, red shows CX-rhodamine-labeled DNA, and green displays different endocytosis markers. Scale bar is 10 µm.

Figure 6. Summary of PLL-g-PEG-DNA nanoparticle uptake mechanisms and trafficking pathways. The schematic summarizes endocytosis pathways that have been investigated. The numbers in brackets in the figure represent the diameters of particles that have been found to be internalized through that pathway.

findings might have been due to a general cellular inhibition, although overall cell viability was not impaired. Nanoparticle

uptake has been described for pDMAEMA and PEI nanoparticles to follow the clathrin- and caveolae-dependent pathways

Intracellular Pathways of PLL-g-PEG Nanoparticles

in COS-7 cells, respectively (21). Another study pointed out that COS-7 cells were very efficiently transfected with alginate-chitosan nanoparticles that were internalized predominantly by caveolae-mediated endocytosis (15). Pegylated polyL-lysine (C1K30-PEG) compacted DNA transfected into COS-7 cells also did not show colocalization with early endosome antigen-1 (EEA-1) or transferrin receptor (TFR), indicating cytosomal stability and no involvement of clathrin-mediated uptake pathways (3). As C1K30-PEG-DNA condensates displayed a rod-like shape with a length of up to 300 nm, involvement of macropinocytotic uptake could clearly be identified (3). However, in our study the PLL-g-PEG-DNA nanoparticles display a spherical shape with homogeneous hydrodynamic diameter of 80-90 nm (11). One explanation for the failure to assign one prominent uptake pathway into COS-7 cells might be that PLL-g-PEG-DNA nanoparticles are formed by noncovalent bonds between the amino groups of poly-L-lysine and phosphate groups of the plasmid DNA. As the structure of such condensates is not known and might also be flexible in watery solution, it might well be that dangling ends of poly-L-lysine extend to the surface of the condensates and serve as cell-penetrating peptide sequences helping to internalize PLL-g-PEG-DNA nanoparticles (Figure 6). Cellpenetrating peptides cross biological membranes in a nondisruptive way without apparent toxicity and can therefore deliver macromolecular cargo into target cells via endocytotic-dependent and -independent ways (14, 17, 39, 40). They form stable complexes with proteins and nucleic acids, and the interaction is either covalent or noncovalent (40, 41). Cell-penetrating peptides may consist of Pep and MPG families that are short amphipathic peptides consisting of three domains: a variable N-terminal hydrophobic motif, responsible for membrane fusion; a hydrophilic lysine-rich domain, which is required for the interactions with DNA, intracellular trafficking of the cargo, and solubility of the peptide vector; and a linker domain separating the two domains (40, 42, 43). Structural and mechanistic investigations have revealed that the flexibility between the two domains, which is maintained by the linker sequence between the fusion and the DNA-binding motifs, is crucial for macromolecule delivery (42-45). Uptake mechanisms of condensates formed between cell penetration peptides and their cargo are usually energy-dependent, independent of the endosomal pathway, and are directly correlated to the size of particles and the nature of cargoes (46, 47). Here, PLL-gPEG-DNA nanoparticles might function as cell-penetrating condensates where surface-exposed lysines in combination with flexible, noncharged PEG moieties are able to deliver plasmid DNA into the cytoplasm (14) and thereby circumvent the endocytotic machinery. In favor of such a hypothesis might be the fact that PLL-g-PEG enters into the cytoplasm of COS-7 cells independently of the presence of plasmid DNA (Figures 1A, a and 2G). This hypothesis might also be an explanation for the fast and efficient delivery of plasmid DNA combined with the low cytotoxicity of PLL-g-PEG-DNA nanoparticles used in this study. In future studies, this feature needs to be explored for the delivery of therapeutic DNA in the context of gene therapy.

ACKNOWLEDGMENT This study was supported by Gebert Ru¨f Foundation (GRS053/05), Switzerland and by CCMX, Competence Center for Materials Science and Technology, Switzerland. We thank Annina Maria Hafner for the introduction in the FACSCalibur system.

Bioconjugate Chem., Vol. 19, No. 9, 2008 1915

LITERATURE CITED (1) Benns, J. M., and Kim, S. W. (2000) Tailoring new gene delivery designs for specific targets. J. Drug Targetting 8, 1– 12. (2) Park, T. G., Jeong, J. H., and Kim, S. W. (2006) Current status of polymeric gene delivery systems. AdV. Drug DeliVery ReV. 58, 467–486. (3) Walsh, M., Tangney, M., O′Neill, M. J., Larkin, J. O., Soden, D. M., McKenna, S. L., Darcy, R., O’Sullivan, G. C., and O′Driscoll, C. M. (2006) Evaluation of cellular uptake and gene transfer efficiency of pegylated poly-L-lysine compacted DNA: implications for cancer gene therapy. Mol. Pharmaceutics 3, 644– 653. (4) Torchilin, V. P. (2006) Recent approaches to intracellular delivery of drugs and DNA and organelle targeting. Annu. ReV. Biomed. Eng. 8, 1.11.33. (5) Ledley, F. D. (1996) Pharmaceutical approach to somatic gene therapy. Pharm. Res. 13, 1595–1614. (6) Dincer, S., Turk, M., and Piskin, E. (2005) Intelligent polymers as nonviral vectors. Gene Ther. 12 (Suppl 1), S139–S45. (7) Lee, M., and Kim, S. W. (2005) Polyethylene glycol-conjugated copolymers for plasmid DNA delivery. Pharm. Res. 22, 1–10. (8) Rosenecker, J., Huth, S., and Rudolph, C. (2006) Gene therapy for cystic fibrosis lung disease: current status and future perspectives. Curr. Opin. Mol. Ther. 8, 439–445. (9) Pietersz, G. A., Tang, C. K., and Apostolopoulos, V. (2006) Structure and design of polycationic carriers for gene delivery. Mini ReV. Med. Chem. 6, 1285–1298. (10) Romano, G. (2007) Current development of nonviral-mediated gene transfer. Drug News Perspect. 20, 227–231. (11) Rimann, M., Lu¨hmann, T., Textor, M., Guerino, B., Ogier, J., and Hall, H. (2008) Characterization of PLL-g-PEG-DNA nanoparticles for the delivery of therapeutic DNA. Bioconjugate Chem. 19, 548–557. (12) Panyam, J., Sahoo, S. K., Prabha, S., Bargar, T., and Labhasetwar, V. (2003) Fluorescence and electron microscopy probes for cellular and tissue uptake of poly(D,L-lactide-coglycolide) nanoparticles. Int. J. Pharm. 262, 1–11. (13) Feng, M., Lee, D., and Li, P. (2006) Intracellular uptake and release of poly(ethyleneimine)-co-poly(methyl methacrylate) nanoparticle/pDNA complexes for gene delivery. Int. J. Pharm. 311, 209–214. (14) Patel, L. N., Zaro, J. L., and Shen, W. C. (2007) Cell penetrating peptides: Intracellular pathways and pharmaceutical perspectives. Pharm. Res. 24, 1977–1992. (15) Douglas, K. L., Piccirillo, C. A., and Tabrizian, M. (2008) Cell line-dependent internalization pathways and intracellular trafficking determine transfection efficiency of nanoparticle vectors. Eur. J. Pharm. Biopharm. 68, 676–687. (16) Chithrami, B. D., and Chan, W. C. W. (2007) Elucidating the mechanism of cellular uptake and the removal of proteincoated gold nanoparticles of different sizes and shapes. Nano Letters 7, 1542–1550. (17) Harush-Frenkel, O., Debotton, N., Benita, S., and Altschuler, Y. (2007) Targetting nanoparticles to the clathrin-mediated endocytotic pathway. BBRC 353, 26–32. (18) Perret, E., Lakkaraju, A., Deborde, S., Schreiner, R., and Rodriguez-Boulan, E. (2005) Evolving endosomes: how many varieties and why? Curr. Opin. Cell Biol. 17, 423–434. (19) Khalil, I. A., Kogure, K., Akita, H., and Harashima, H. (2006) Uptake pathways and subsequent intracellular trafficking in nonviral gene delivery. Pharmacol. ReV. 58, 32–45. (20) van der Aa, M. A., Huth, U. S., Ha¨fele, S. Y., Schubert, R., Oosting, R. S., Mastrobattista, E., Hennink, W. E., Peschka-Suss, R., Koning, G. A., and Crommelin, D. J. (2007) Cellular uptake of cationic polymer-DNA complexes via caveolin plays a pivotal role in gene transfection in COS-7 cells. Pharm. Res. 24, 1590– 1598. (21) van der Aa, M. A., Huth, U. S., Hafele, S. Y., Schubert, R., Oosting, R. S., Mastrobattista, E., Hennink, W. E., Peschka-Suss,

1916 Bioconjugate Chem., Vol. 19, No. 9, 2008 R., Koning, G. A., and Crommelin, D. J. (2007) Cellular uptake of cationic polymer-DNA complexes via caveolae plays a pivotal role in gene transfection in COS-7 cells. Pharm. Res. 24, 1590– 1598. (22) Conner, S. D., and Schmid, S. L. (2003) Regulated portals of entry into the cell. Nature 422, 37–44. (23) Swanson, J. A., and Watts, C. (1995) Macropinocytosis. Trends Cell Biol. 5, 424–428. (24) Takei, K., and Haucke, V. (2001) Clathrin-mediated endocytosis: membrane factors pull the trigger. Trends Cell Biol. 11, 385–391. (25) Maxfield, F. R., and McGraw, T. E. (2004) Endocytic recycling. Nat. ReV. Mol. Cell Biol. 5, 121–32. (26) Thomsen, P., Roepstorff, K., Stahlhut, M., and van Deurs, B. (2002) Caveolae are highly immobile plasma membrane microdomains, which are not involved in constitutive endocytic trafficking. Mol. Biol. Cell 13, 238–250. (27) Parton, R. G., and Richards, A. A. (2003) Lipid rafts and caveolae as portals for endocytosis: new insights and common mechanisms. Traffic 4, 724–738. (28) Pelkmans, L., and Helenius, A. (2002) Endocytosis via caveolae. Traffic 3, 311–320. (29) Neu, U., Woellner, K., Gauglitz, G., and Stehle, T. (2008) Structural basis of GM1 ganglioside recognition by simian virus 40. PNAS 105, 5219–5224. (30) Pelkmans, L., Kartenbeck, J., and Helenius, A. (2001) Caveolar endocytosis of simian virus 40 reveals a new two-step vesicular-transport pathway to the ER. Nat. Cell Biol. 3, 473– 483. (31) Parton, R. G., and Simons, K. (2007) The multiple faces of caveolae. Nat. ReV. Mol. Cell Biol. 8, 185–94. (32) Pasche, S., Textor, M., Meagher, L., Spencer, N. D., and Griesser, H. J. (2005) Relationship between interfacial forces measured by colloid-probe atomic force microscopy and protein resistance of poly(ethylene glycol)-grafted poly(L-lysine) adlayers on niobia surfaces. Langmuir 21, 6508–6520. (33) Saslowski, D. E., and Lancer, W. I. (2008) Conversion of apical plasma membrane sphingomyelin to ceramide attenuates the intoxication of host cells by cholera toxin. Cell Microbiol. 10, 67–80. (34) Foerg, C., Ziegler, U., Fernandez-Carneado, J., Giralt, E., Rennert, R., Beck-Sickinger, A. G., and Merkle, H. P. (2005) Decoding the entry of two novel cell-penetrating peptides in HeLa cells: lipid raft-mediated endocytosis and endosomal escape. Biochemistry 44, 72–81. (35) Vendeville, A., Rayne, F., Bonhoure, A., Bettache, N., Montcourrier, P., and Beaumelle, B. (2004) HIV-1 Tat enters T cells using coated pits before translocating from acidified

Lu¨hmann et al. endosomes and eliciting biological responses. Mol. Biol. Cell 15, 2347–2360. (36) Schmid, S. L. (1997) Clathrin-coated vesicle formation and protein sorting: an integrated process. Annu. ReV. Biochem. 66, 511–548. (37) Zheng, J., and Ramirez, V. D. (2000) Inhibition of mitochondrial proton F0F1-ATPase/ATP synthase by polyphenolic phytochemicals. Br. J. Pharmacol. 130, 1115–1123. (38) Kuiper, G. G. J. M., Lemmen, J. G., Carlsson, B., Corton, J. C., Safe, S. H., Van der Saag, P. T., Van der Burg, B., and Gustavsson, J. A. (1998) Interaction of estrogenic chemicals and phytoestrogens with estrogen receptor β. Biotechnology 139, 4252–4263. (39) Amand, H. L., Fant, K., Norden, B., and Esbjo¨rner, E. K. (2008) Stimulated endocytosis in penetratin uptake: effect of arginine and lysine. Biochem. Biophys. Res. Commun. 371, 621– 625. (40) Morris, M. C., Deshayes, S., Heitz, F., and Divita, G. (2008) Cell-penetrating peptides: from molecular mechanisms to therapeutics. Biol. Cell 100, 201–217. (41) Foged, C., and Nielsen, H. M. (2008) Cell-penetrating peptides for drug delivery across membrane barriers. Expert Opin. Drug DeliVery 5, 105–117. (42) Morris, M. C., Chaloin, L., Me´ry, J., Heitz, F., and Divita, G. 1999a A novel potent strategy for gene delivery using a single peptide vector as a carrier. Nucleic Acids Res. 27, 3510–3517. (43) Simeoni, F., Morris, M. C., Heitz, F., and Divita, G. (2003) Insight into the mechanism of the peptide-based gene delivery system MPG: implications for delivery of siRNA into mammalian cells. Nucleic Acids Res. 31, 2717–2724. (44) Deshayes, S., Gerbal-Chaloin, S., Morris, M. C., AldrianHerrada, G., Charnet, P., Divita, G., and Heitz, F. (2004) On the mechanism of non-endosomal peptide-mediated cellular delivery of nucleic acids. Biochim. Biophys. Acta 1667, 141– 147. (45) Simeoni, F., Morris, M. C., Heitz, F., and Divita, G. (2005) Peptide-based strategy for siRNA delivery into mammalian cells. Methods Mol. Biol. 309, 251–260. (46) Gros, E., Deshayes, S., Morris, M. C., Aldrian-Herrada, G., Depollier, J., F., H., and Divita, G. (2006) A non-covalent peptide-based strategy for protein and peptide nucleic acid transduction. Biochim. Biophys. Acta 1758, 384–393. (47) Mun˜oz-Morris, M. A., Heitz, F., Divita, G., and Morris, M. C. (2007) The ep-1 forms biologically efficient nanoparticle complexes. Biochem. Biophys. Res. Commun. 355, 877–882. BC800206R