High-Yield Nontoxic Gene Transfer through Conjugation of the CM18

Article pubs.acs.org/molecularpharmaceutics

High-Yield Nontoxic Gene Transfer through Conjugation of the CM18Tat11 Chimeric Peptide with Nanosecond Electric Pulses Fabrizio Salomone,*,†,‡ Marie Breton,§ Isabelle Leray,§,∥,⊥ Francesco Cardarelli,‡ Claudia Boccardi,‡ Daniel Bonhenry,# Mounir Tarek,# Lluis M. Mir,§,∥,⊥ and Fabio Beltram†,‡ †

NEST, Scuola Normale Superiore and Istituto Nanoscienze-CNR, Piazza San Silvestro 12, 56127 Pisa, Italy Center for Nanotechnology Innovation @NEST, Istituto Italiano di Tecnologia, Piazza San Silvestro 12, 56127 Pisa, Italy § CNRS, Laboratoire de Vectorologie et Thérapeutiques Anticancéreuses, UMR 8203, Orsay F-91405, France ∥ Université Paris-Sud, Laboratoire de Vectorologie et Thérapeutiques Anticancéreuses, UMR 8203, Orsay F-91405, France ⊥ Institute Gustave-Roussy, Laboratoire de Vectorologie et Thérapeutiques Anticancéreuses, UMR 8203, Villejuif F-94805, France # Université de Lorraine, UMR 7565, Structure et Réactivité des Systèmes Moléculaires Complexes, CNRS, Nancy F-54003, France ‡

S Supporting Information *

ABSTRACT: We report a novel nontoxic, high-yield, gene delivery system based on the synergistic use of nanosecond electric pulses (NPs) and nanomolar doses of the recently introduced CM18-Tat11 chimeric peptide (sequence of KWKLFKKIGAVLKVLTTGYGRKKRRQRRR, residues 1−7 of cecropin-A, 2−12 of melittin, and 47−57 of HIV-1 Tat protein). This combined use makes it possible to drastically reduce the required CM18-Tat11 concentration and confines stable nanopore formation to vesicle membranes followed by DNA release, while no detectable perturbation of the plasma membrane is observed. Two different experimental assays are exploited to quantitatively evaluate the details of NPs and CM18-Tat11 cooperation: (i) cytofluorimetric analysis of the integrity of synthetic 1,2-dioleoyl-sn-glycero-3-phosphocholine giant unilamellar vesicles exposed to CM18-Tat11 and NPs and (ii) the in vitro transfection efficiency of a green fluorescent protein-encoding plasmid conjugated to CM18-Tat11 in the presence of NPs. Data support a model in which NPs induce membrane perturbation in the form of transient pores on all cellular membranes, while the peptide stabilizes membrane defects selectively within endosomes. Interestingly, atomistic molecular dynamics simulations show that the latter activity can be specifically attributed to the CM18 module, while Tat11 remains essential for cargo binding and vector subcellular localization. We argue that this result represents a paradigmatic example that can open the way to other targeted delivery protocols. KEYWORDS: gene delivery, nanopulses, cell-penetrating peptides, antimicrobial peptides, electropermeabilization, electroporation, nanoporation

■

because of electrophoretic effects.3−7 In all these applications, pulses must be administered carefully to prevent excessive perturbation and even permanent cell damage.8 The development of devices that can administer pulses in the kilovolts per

INTRODUCTION

Low-intensity microsecond and millisecond electric field pulses administered to cells promote plasma membrane electroporation (EP), i.e., the reversible promotion of membrane permeability changes,1 and thus promote the uptake of molecules that display poor membrane crossing abilities.2 This approach was extensively exploited for gene transfer applications because besides permeabilizing the cell membrane, electric pulses can also drive DNA toward the transient pores © XXXX American Chemical Society

Received: March 25, 2014 Revised: May 10, 2014 Accepted: May 27, 2014

A

dx.doi.org/10.1021/mp500223t | Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics

Article

mL. DOPE-rhodamine was added at a 1% molar concentration, and the solution was stored at −20 °C. Vesicles were prepared at 6 °C using the electroformation protocol described by Angelova.25 Fifteen microliters of the lipid solution was deposited on the conducting side of two glass slides coated with indium tin oxide (Sigma, St. Louis, MO). Slides were then kept under vacuum for 2 h in a desiccator to remove all traces of organic solvent. A chamber was assembled with the slides spaced by a 1.5 mm silicone isolator (Sigma). The chamber was filled with a sucrose solution (240 mM sucrose, 1 mM NaCl, and 1 mM KH2PO4/K2HPO4). The slides were connected to a function/arbitrary waveform generator (HP Agilent 33120A, Agilent, Santa Clara, CA), and a peak to peak sinusoidal voltage of 25 mV and 8 Hz was applied. The voltage was increased by 100 mV steps every 5 min to a value of 1225 mV and maintained under these conditions overnight. Finally, a squarewave dc field of the same amplitude was applied at 4 Hz for 1 h to detach the GUVs from the slides. Exposure of Vesicles to Electric Pulses and/or CM18Tat11. Twenty microliters of the GUV solution was mixed with different amounts of a peptide (CM18-Tat11, CM18, or Tat11) solution in sucrose buffer in a final volume of 200 μL. A control solution was also prepared with 20 μL of the GUV solution and 180 μL of sucrose buffer. These preparations were transferred in conventional electroporation cuvettes (STD, Dutcher, Issy les Moulineaux, France). The distance between the two planar electrodes of the cuvette (d) was 0.1 cm. A number of 10 ns pulses (between 1 and 100) were then applied to the cuvette. A commercial generator purchased from FID (model FPG 10ISM10, FID GmbH, Burbach, Germany) with an output impedance of 50 Ω was used. It generates trapezoidal monopolar pulses with a full width at half-maximum of 10 ns. The following output voltages were applied: U = 0.5, 2, and 6 kV. This setup was previously described by Silve et al.21 GUV Stability Qualitative Assay. GUV populations exposed to 1 μM dye-labeled CM18-Tat11, CM18, or Tat11 with or without 10 ns NPs were analyzed qualitatively by depositing the solution inside a chamber consisting of a glass coverslip mounted onto a glass slide with heated parafilm. All images were obtained with an inverted confocal microscope (Zeiss LSM510, Carl Zeiss, Jena, Germany) equipped with a Plan Apochromat 63×/ON 1.4 objective. For the red channel (rhodamine-labeled vesicles), the excitation wavelength was 543 nm and the emission filter was a 560 nm long-pass type. For the green channel (atto-495 peptide), the excitation wavelength was 488 nm and the emission filter was a 500−530 band-pass type. GUV Stability Quantitative Assay. For the quantitative assays, the vesicle solution after pulse delivery or peptide exposure only was examined by flow cytometry on a BD Accuri C6 flow cytometer (BD Biosciences) to evaluate the number of stable GUVs through the analysis of event numbers inside a selected region of interest (ROI, obtained from a stable vesicle solution) in the vesicle distribution graph (y FSC-A, size; x FL2-A, rhodamine signal) using the BD Accuri CFlow Plus software. Destabilized, collapsed vesicles were detected as debris components (falling out of the ROI) by cytometer analysis. Thus, to evaluate the percentage of stable GUVs in a sample, we measured the ratio between the ROI event number of treated and control samples. GUVs exposed to a 0.005% solution of commercial surfactant Triton X-100 (Sigma-Aldrich, Buchs, Switzerland) were used as a positive control.

centimeter magnitude range on the nanosecond time scale (4− 600 ns electric pulses, NPs) opened new exciting possibilities.9−11 These NPs are actively researched12,13 as they appear to target preferentially malignant cells14 with reduced unwanted thermal effects, thus minimizing the damage to biological tissues.9 In addition, by modulation of pulse parameters (magnitude, duration, frequency of repetition, and total number),15,16 NPs showed the ability to perturb not only the plasma membrane but also membranes of internal organelles (e.g., endosomal vesicles, endoplasmic reticulum, storage vacuoles, etc.17,18), thus suggesting new opportunities for targeted membrane electroporation.19 To date, the reduced size and lifetime of nanopores together with the intrinsic lack of target site selectivity limit the application of NPs to the transfer of small- to medium-sized cargoes (e.g., bleomycin and siRNA) through the plasma membrane.20−22 In this context, some of the properties of the recently proposed CM18-Tat11 chimeric peptide (CM18, KWKLFKKIGAVLKVLTTG, residues 1−7 of cecropin-A and 2−12 of melittin antimicrobial peptides; Tat11, YGRKKRRQRRR, residues 47−57 of HIV-1 Tat protein)23,24 deserve attention. In particular, CM18-Tat11 was successfully exploited as a delivery vector for plasmidic DNA (pDNA). The CM18-Tat11−pDNA complex is readily taken up by cells, accumulates within endocytic vesicles, and promotes vesicle membrane destabilization followed by cargo (i.e., DNA) cytoplasmic release. This requires, however, a relatively high peptide concentration within vesicles, typically >10 μM. It was reported that these levels can be reached within vesicles even starting from much lower externally delivered concentrations that do not appear to perturb the cell membrane,23,24 but the lowest possible peptide concentration is desired. Here we demonstrate the possibility of using synergistically NPs and the peptide vector. We show that CM18-Tat11−pDNA vectors administered at very low concentrations (not active per se) lead to the accumulation of pDNA molecules in endosomal compartments and can lead to an efficient transfer of pDNA into the cytoplasm following NP-driven membrane destabilization. Two experimental platforms were used to evaluate the cooperative action of NPs and CM18-Tat11: (i) cytofluorimetric evaluation of the integrity of synthetic 1,2-dioleoyl-sn-glycero-3phosphocholine (DOPC) giant unilamellar vesicles (GUVs) exposed to CM18-Tat11 and NPs and (ii) the in vitro transfection efficiency of a green fluorescent protein (GFP)encoding plasmid conjugated to CM18-Tat11 in the presence of NPs. GUV stability data highlighted the membrane perturbing capabilities of 10 ns electric pulses conjugated to CM18-Tat11. The in vitro transfection assay performed in HeLa cells confirmed the cooperative effect, yielding GFP expression efficiency levels comparable to gold standards of transfection. Finally, we investigated the molecular details of this cooperative process by atomistic molecular dynamics (MD) simulations of planar POPC bilayers exposed to the various peptides. MD data suggest a specific role of the CM18 module in fixing the membrane defect (nanopore) activated by the electric pulses, preventing complete resealing of the bilayer.

■

EXPERIMENTAL SECTION Preparation of Giant Unilamellar Vesicles (GUVs). 1,2Dioleoyl-sn-glycero-3-phosphocholine (DOPC) and 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-(lissamine rhodamine B sulfonyl) ammonium salt (DOPE-rhodamine) were purchased from Avanti Polar Lipids (Alabaster, AL). DOPC was dissolved in chloroform at a mass concentration of 0.5 mg/ B



Article

Cell Culture, Peptides, and DNA Plasmids. HeLa cells were purchased from ATCC and cultured following the manufacturer’s instructions. Cells were maintained at 37 °C in a humidified 5% CO2 atmosphere. All peptides were prepared by solid-phase synthesis using Fmoc chemistry on an automatic Liberty-12-Channel Automated Peptide Synthesizer. The crude peptides were purified by reverse-phase highperformance liquid chromatography (RP-HPLC) on a Jupiter 4m Proteo 90 A column [250 mm × 10 mm (Phenomenex)]. HPLC analysis and purification were performed on a Dionex Ultimate 3000 PLC system with an autosampler. The correct purified product was confirmed by electrospray mass spectroscopy. The purity was >95% as determined by analytical highperformance liquid chromatography. The cysteine residue added to the C-terminus of each peptide provided a sulfhydryl group for further ligation to the atto-495-maleimide fluorophore. The labeling of purified peptides was performed by incubating the peptide for 3 h with a 3-fold molar excess of atto-495-maleimide (ATTO-TEC GmbH), 150 mM PBS buffer, and TCEP (pH 7.4). Finally, atto-495-labeled peptides were purified by HPLC (see above) and then lyophilized overnight. The molecular weight of all conjugated peptides was confirmed by electrospray mass spectroscopy, and the concentration of each peptide stock solution was verified by UV−vis absorbance. The electrospray ionization mass spectrometry spectra of the peptides were obtained with an API 3200 QTRAP, a Hybrid Triple Quadrupole/Linear Ion Trap (ABSciex, Foster City, CA). Peptides were stored at −80 °C. The pCMV-GFP plasmid [3.5 kb (Plasmid Factory GmbH & Co., Bielefeld, Germany)] was used for the transfection efficiency assay. A label IT Cy3 plasmid delivery control [2.7 kb, 1730 kDa (Mirus Bio Corp., Madison, WI)] was used for the determination of the peptide/pDNA intracellular location. Transfection Vectors. Peptide−pDNA binary complexes were prepared as follows: 0.35 μg of plasmidic pCMV-GFP DNA was dissolved in 200 μL of a PBS solution, and different volumes of the peptide solution were added to obtain the desired 4:1 N:P (nucleotide:phosphate) molar ratio. The mixture was then incubated at room temperature for 30 min. Finally, the transfection medium was added until a total volume of 1 mL was reached (i.e., in the 4:1 CM18-Tat11−DNA complex, concentrations are 0.3 μM peptide and approximately 0.1 nM pDNA; these are the concentrations employed for all the transfection experiments). In Vitro DNA Transfection and Transfection Efficiency Evaluation. Cells were plated in WillCo dishes at 60% confluence and cultivated in growth medium with 10% FBS. After 24 h, the growth medium in each well was replaced with transfection medium containing peptide−pDNA binary complexes. Incubation with the cells lasted 6 h at 37 °C in 5% (v/v) CO2. Cells were subsequently washed three times in PBS before trypsination for 1 min in a 0.25% trypsin solution (Invitrogen, Stockholm, Sweden). Detached cells were resuspended in growth medium and transferred into conventional electroporation cuvettes (STD, Dutcher). The distance between the two planar electrodes of the cuvette (d) was 0.1 cm. For the generation of NPs, a spark gap pulse generator designed by Europulse (Cressensac, France) was used. Cells were exposed in suspension in classical electroporation cuvettes (Cell project, Sutton Valence, U.K.). The electric field was measured during exposure with a D-dot sensor inserted into the ground electrode. The europulse generator has an output impedance of 50 Ω and an output voltage magnitude between

10 and 20 kV. It generates trapezoidal monopolar pulses of 10, 20, 40, 80, or 100 ns. After pulse delivery, cells were plated again, incubated for 24 h in growth medium before a new trypsin detachment, and finally resuspended in growth medium. The transfection efficiency (TE) was evaluated by flow cytometry analysis of 15000 detached cells per sample. An average GFP intensity per cell was obtained for each sample using the BD Accuri CFlow Plus software. As a control, commercial polymer formulation Jetprime reagent (Polyplus, New York, NY) and lipid transfection Lipofectamine (Invitrogen, Carlsbad, CA) were used according to the manufacturer’s instructions. For confocal live scanning microscopy (CLSM), cells were plated onto 35 mm glassbottom Petri dishes (WillCo, dish GWSt-3522) the day before the experiment so that the cells reached 70% confluence. After 24 h, growth cell medium was replaced with 1 mL of transfection medium containing CM18-Tat11(Tat11)-atto633− Cy3-pDNA binary complexes at 4:1 N:P ratio (see Transfection Vectors). Incubation with the cells lasted 6 h at 37 °C. Finally, we replaced transfection medium with growth medium. Cytotoxicity Evaluation and YO-PRO-1 Membrane Permeabilization Analysis. To test the impact of the delivery treatments (e.g., peptide−pDNA binary complexes or commercial transfection reagents with NPs) on cell metabolism, we used the Wst-8 (water-soluble tetrazolium) assay (Sigma-Aldrich). In detail, 24 h after the pulse delivery, cells plated in 96-wells plates were exposed to the Wst-8 reagent according to the manufacturer’s protocol. The absorbance at 450 nm was measured 2 h later on a Synergy HT multiplate reader (Bio Tek instruments, Winooski, VT). Untreated cells were defined as 100% viable, while cells exposed to 20% dimethyl sulfoxide (DMSO) for 1 h were used as a positive control. Cells exposed to the transfection protocol were also tested for membrane permeabilization using the YO-PRO-1 organic dye. YO-PRO-1 (5 μM) was administered to the medium of detached cells after transfection, and its average fluorescence intensity per cell was evaluated by flow cytometry from 10000 cells per sample 10 min after NP delivery. Statistical Analysis. Data are expressed as means ± standard deviations (SDs). Mean values and SDs were calculated for each sample examined in at least three independent experiments. Statistical comparisons between the control and experimental groups were performed using the Student’s t test. The level of statistical significance was set at p < 0.05. MD Simulation Procedures. System. The membrane model used for this study is an equilibrated fully hydrated palmitoyloleoylphosphocholine (POPC) bilayer. Because of the availability of well-equilibrated membranes composed of POPC molecules, we decided to employ this setup. Both DOPC and POPC carry phosphatidylcholine headgroups. The presence of a second oleyl chain in DOPC has been here considered to have no effect on the overall behavior of the bilayer with respect to electroporation. The model consists of 256 lipids units and roughly 22400 water molecules organized in two lamellae above and below the lipids. At the temperature set for the study, i.e., 300 K, the bilayer is in the biologically relevant liquid-crystal La phase. Three different systems were considered involving (a) CM18-Tat11, (b) only CM18, or (c) only Tat11. An initial conformation of CM18-Tat11 was taken from a previous study,23 and the peptide was set in two different configurations: either just above the top membrane−solution interface or with the amphipathic CM18 fragment already adsorbed in the phosphate C



Article

Figure 1. Exposure of GUVs to CM18-Tat11 with or without 10 ns electric pulse(s). (a) Confocal fluorescence images of a rhodamine-labeled GUV (red) exposed to 1 μM atto-495-labeled CM18-Tat11 (green). (b) GUV as in panel a with a 10 ns electric pulse. The scale bar is 10 μm. (c) Flow cytometry analysis of the stable GUVs after exposure to 1 μM CM18-Tat11 (black column) and 1 μM CM18-Tat11 with 1, 10, or 100 electric pulses (10 ns each) (red columns). (d) Flow cytometry analysis of the stable GUVs after exposure to 1 μM Tat11 (left black column) or CM18 (left gray column) and 100 electric pulses (right black and gray colmns). Significant differences are determined at the p < 0.05 level of between control and NP treatments. Data are presented as means ± SDs from three independent experiments for each treatment.

version for proteins28 that treats protein dynamics more accurately. Simulation Protocols. The system was first equilibrated at a constant temperature (300 K) and a constant pressure (1 atm), using a full three-dimensional (3D) periodic boundary. Further simulations were performed using the standard “electric field method” for which a constant electric field normal to the bilayer was applied to each particle of the system bearing a charge qi in the form of a force F = qiE, using 3D periodic boundary conditions. Electric fields were applied for configurations a−c. For configurations a and b, the configuration with the CM18 anchored to the membrane was used as a starting point. In MD simulations, the transmembrane voltage induced by an electric field is ∼Um = E × Lz, where Lz is the size of the simulation box. In our setup, the value of the electric field that allows poration of the membrane is ∼0.2 V/nm. Because of the length of the simulation box (120 Å), the modeled bilayer is under these conditions subject to a transmembrane voltage of ∼2.4 V, consistent with previous investigations.22,26,29 In simulations, it is possible to mimic the experimental conditions of NPs by modeling the application of an external electric field E (with the same duration) in the form of an external force F = qE acting on each particle carrying a charge q. Because of the use of periodic boundary conditions, the effect of an electric field on a bilayer, in particular the induced transmembrane voltage (and therefore the ability to electroporate the membrane), depends not only on field strength E but also on the size of the system. In our system, because of the setup, simulations at a field in the same magnitude range of the experiments (2 kV/mm) yield a voltage too low for the electroporation of the membrane, or even the stabilization of a preformed pore. The field magnitude that is able to generate a transmembrane voltage of ∼1.6 V and therefore yield the membrane electroporation was considered as a threshold value to use in simulations. In all our simulations, we have therefore used electric fields comparable to this threshold magnitude as previously found in refs 22, 26, and 29.

glycerol region with the hydrophobic part looking toward the hydrophobic core and the hydrophilic moieties in contact with the solvent. This second configuration is based on the evidence presented in the literature30 regarding the membrane insertion and localization of CM peptides, and it allows a great reduction of the duration of computation. For the anchored configurations, seven lipid molecules were removed to avoid bad contacts and the system was equilibrated. The same protocols were followed with the simulations involving thCM18 alone. To model the effect of high CM18 peptide concentrations, we have also modeled a lipid patch with three peptides (i.e., increasing by a factor of 3 the CM18:lipid ratio). The simulations with Tat11 all started with the peptide placed in the solution near the choline of the lipid headgroups. To counterbalance the positive charge of the molecule, 13 chloride anions were added randomly to the solution for configuration a, 5 chloride anions for configuration b, and 7 chloride anions for configuration c. For configuration a, a simulation with an ionic concentration of 0.15 M NaCl was set up and included 127 sodium cations and 140 chloride anions. When the electric field was applied, only counterions were present, following the recommendations of Delemotte et al.26 The final dimensions of the system were 89 Å × 88 Å × 120 Å, and the total number of atoms was roughly equal to 100000 for all configurations. Simulation Parameters. The MD simulations presented were conducted using the program NAMD2 targeted for massively parallel architectures.27 The systems were examined at constant pressure and constant temperature (1 atm and 300 K) employing the Langevin dynamics and Langevin piston method. The equations of motion were integrated using the multiple-time step algorithm. A time step of 2.0 fs was employed. Short- and long-range forces were calculated every two and four time steps, respectively. Chemical bonds between hydrogen and heavy atoms were constrained to their equilibrium value. Long-range, electrostatics forces were taken into account using a fast implementation of the particle mesh Ewald (PME) method, with a direct space sum tolerance of 10−6 and a spherical truncation of 12 Å with a modified interaction potential through a switching distance of 10 Å. The water molecules were described using the TIP3P model. The parameters are employed as implemented in the CHARMM36 version for the POPC lipid molecules and the CHARMM36

■

RESULTS Cooperative GUV Destabilization Action of CM18-Tat11 and 10 ns NPs. To investigate the possible cooperative action of CM18-Tat11 and NPs on membrane integrity, we studied the D



Article

ratio, which reflects onto the vector surface charge and its consequent ability to enter cells and accumulate within endosomes. In other words, peptide−pDNA formulations leading to intravesicular peptide concentrations below its critical endosomolytic threshold are not competent for gene transfer. As an example of this condition, we selected CM18Tat11−pDNA complexes formed at a 4:1 N:P ratio. When administered to cells, these complexes are not active per se;24 i.e., they are trapped inside the endosomal compartments (as shown in Figure 3a of the Supporting Information for the cy3labeled complex), and pDNA molecules are not able to reach the nucleus. As schematically shown in Figure 3b of the Supporting Information, we set up an experimental protocol to measure the transfection efficiency (TE) in HeLa cells upon the combined administration of peptide−pDNA vectors and NPs. Notably, as shown in Figure 3c of the Supporting Information, 10 ns pulses are not sufficient to significantly destabilize internal membranes in live cells; i.e., they do not result in acceptable TE levels (black colums). By contrast, longer pulses (40 ns) appear to be effective for transfecting cells at levels comparable to gold standards (red columns). In particular, administration of 20 NPs of 40 ns each is sufficient to increase CM18-Tat11-driven TE by more than 1 order of magnitude, making it similar to commercial standards (Figure 2a, red columns). Interestingly, the same treatment does not boost the TE levels reached by isolated Tat11, Lipo2000, or Jetprime. As expected, if NPs are not administered, the TE elicited by a total CM 18 -Tat 11 concentration of 0.3 μM is on the same order of magnitude as that of naked pDNA (i.e., more than 1 order of magnitude lower than those of commercial transfectants) (Figure 2a, black columns). We concluded this analysis by performing a complementary cytotoxicity assay [WST-8 (see the Experimental Section)] on the same transfected cells. Figure 2b shows that classical lipofection, in the absence of pulses, yields more than 20% viability reduction, while no significant cell toxicity is observed for CM18-Tat11 and Jetprime (black columns). The administration of NPs causes an additional decrease in viability in Lipo2000-treated cells but does not affect CM18-Tat11- or Jetprime-treated cells (Figure 2b, red columns). Additional tests on membrane permeabilization were conducted using the YO-PRO-1 assay (see the Experimental Section and Figure 4 of the Supporting Information). Cells exposed to CM18-Tat11 only do not show any increment in the magnitude of the YO-PRO-1 signal compared to that of untreated cells, confirming that the plasma membrane is not perturbed. On the other hand, the large increase in the magnitude of the intracellular YO-PRO-1 signal after the administration of NPs (in the presence or absence of CM18Tat11) indicates the formation of reversible nanopores at the level of the plasma membrane. Combined with TE data and the WST-8 viability test, these assays confirm that NP−peptide cooperation leads to extensive membrane destabilization only within endosomes, although NP-induced transient perturbation is still present at the plasma membrane. Atomistic MD Simulations and Molecular Mechanism of NP−Peptide Cooperative Effects. The results reported above indicate a synergy between NPs and membraneperturbing peptides that can lead to membrane perturbation in a biological environment. However, the mechanism by which this cooperation takes place in the lipid bilayer needed to be clarified. We used atomistic MD simulations to investigate at the molecular level the effect of NPs on membrane integrity in

effect of their co-administration on GUVs. Rhodamine-labeled GUVs were prepared as described in the Experimental Section. The resulting solution contained GUVs with diameters ranging from 1 to 100 μm, with a majority having diameters in the range of 10−20 μm. When GUVs are exposed to atto-495labeled CM18-Tat11, accumulation of the peptide on the surface of the intact spherical vesicles in suspension can be readily observed by confocal microscopy imaging, as shown in Figure 1a. The same accumulation was observed with isolated modules, CM18 and Tat11 (Figure 1a,b of the Supporting Information). A peptide concentration of at least 1 μM was needed to obtain a detectable signal from the vesicle membrane. If vesicles loaded with 1 μM CM18-Tat11 are exposed to even a single 10 ns electric pulse, several destabilized “ghost” vesicles are suddenly deposited on the slide surface, as shown in the example of Figure 1b. By contrast, similar administration of NPs to peptide-free GUVs does not cause any visible vesicle destabilization (they appear as in the DOPC-rhodamine panel of Figure 1a). Visual inspection of NP-exposed GUVs treated with the two isolated modules, CM18 and Tat11, at 1 μM reveals that the former is able to produce vesicle disruption while the latter is not (Figure 1c,d of the Supporting Information). To obtain a more quantitative picture of the process, we analyzed GUV solutions under different conditions (varying peptide concentration, NP number, and/or intensity) by flow cytometry. In such a destabilized system, collapsed vesicles appear preferentially within the debris fraction. Initially, we exposed the rhodamine-labeled GUV population to sets of electric pulses different in number and intensity (1, 10, or 100 pulses; 0.5, 2, or 6.5 kV/mm), in the absence of the peptide. As reported in Figure 2a of the Supporting Information, none of the tested pulse conditions produced a significant decrease in GUV stability. Then we administered different amounts of unlabeled CM18-Tat11 peptide (from 0.1 to 10 μM) to the GUV solution, using commercial surfactant Triton X-100 as a positive control. Figure 2b of the Supporting Information shows that CM18-Tat11 concentrations of up to 3 μM do not destabilize the vesicle membrane, in keeping with the visual inspection described above. Once the nondestabilizing conditions for NPs and peptide-only treatments had been defined, we tested the combination of 1 μM CM18-Tat11 and 2 kV/mm electric nanopulses. Notably, a single pulse is already sufficient to destabilize almost 25% of the vesicles. The perturbation efficacy increases, as expected, with the number of pulses and reaches 70% with 100 pulses (Figure 1c, red columns). The same quantitative evaluation was conducted on GUVs treated with the two isolated control peptides. It is worth noting that delivery of even 100 NPs at 2 kV/mm to GUVs loaded with 1 μM isolated Tat 11 cannot promote vesicle membrane perturbation (Figure 1d, black columns). By contrast, the same treatment of GUVs loaded with 1 μM CM18 produces the destabilization of more than 50% of the vesicles (Figure 1d, gray columns). These results point to a specific role of the CM18 module in cooperation with NPs to produce membrane destabilization (see below). Cooperative Action of CM18-Tat11 and 40 ns NPs on Live Cell Endosomal Vesicles. A very suitable experimental platform for probing the practical exploitability of the cooperative effects of CM18-Tat11 and NPs in live cells can be found in a recent report by some of us.24 We showed that the ability of CM18-Tat11 to drive efficient pDNA transfection in vitro is strictly dependent on the peptide:DNA stoichiometric E



Article

As shown in panels a and b of Figure 3, once the bilayer patch is exposed to an electric field above the electroporative

Figure 3. MD simulation of the molecular interaction of CM18-Tat11 with the membrane bilayer during the administration of NPs. Panels a−c represent snapshots along a 10 ns simulation run of the POPC bilayer−CM18-Tat11 system during the application of a high electric field, while the system configurations after the E field has been switched off are shown in panels d−f. The peptide is represented as van der Waals spheres. The lipid headgroups are shown as light purple and gray spheres and the hydrophobic tails as dark purple spheres. Water was omitted for the sake of clarity. The two moieties of the peptide are pictured, CM18 (yellow) and Tat11 (orange).

Figure 2. TE and cytotoxicity of the vector−pDNA complex with or without 40 ns electric pulses in vitro. (a) Transgene expression is detected 24 h after transfection by flow cytometry analysis of 15000 detached HeLa cells per sample. An average GFP intensity per cell was obtained for each sample. The gray column represents the mean value obtained for untreated cells, while red and black columns are for the vector−pDNA complex with and without 20 electric pulses (40 ns each), respectively. The reported average RLU values represent the means ± SDs of three independent measurements, each performed in triplicate. Significant differences are determined at the p < 0.05 level between control and transfection treatments. (b) Wst-8 assay for evaluating cell metabolic activity. The wst-8 reagent was added for 2 h to cells treated as described for the transfection assay and absorbance at 450 nm measured. Untreated cells are defined as 100% viable, while cells exposed to 20% dimethyl sulfoxide (gray column) are used as a positive control for decreased metabolic activity. As described previously, red and black columns are for the vector−pDNA complex with and without 20 electric pulses (40 ns each), respectively. Significant differences are determined at the p < 0.05 level between control and Lipo2000 or DMSO treatments. Data are presented as means ± SDs from three independent experiments for each treatment.

threshold, a hydrophilic pore forms in the bilayer within a few nanoseconds, in agreement with analogous simulations performed by others.22,26,31 It is worth noting that the peptide α-helical structure is not affected by the exposure to an electric field. Also, there is no evidence of the peptide pulling out or in from the bilayer under these conditions, probably because of the strong interaction between the amphipathic part of the CM18-Tat11 peptide and the phospholipids. When the bilayer is subject to sufficiently high electric fields, the pore may expand (Figure 3c). If it is located close enough to a pore mouth, the peptide is dragged electrophoretically into the hydrophilic pore (see Figure 3c). During the translocation process, the CM18 αhelix remains embedded under the lipid headgroup area while the highly charged unstructured Tat11 fragment remains exposed to the solvent. In contrast to the case of siRNA,22 where complete translocation could occur within a 10 ns long pulse, here, sliding of the peptide along the pore was severely slowed by the strong interaction of the CM18 fragment with the membrane bilayer. As we are modeling the effect of short pulses, the simulations were extended beyond the first 10 ns with the electric field switched off. As expected, the release of the electric stress led to the collapse of the nanopore22,26 (Figure 3d), enclosing the peptide altogether. One expects evidently that the presence of the CM18-Tat11 peptide in the collapsed pore impairs the rapid reconstitution of the bilayer (Figure 3e,f) constituting a “weak spot” in the membrane. In a second series of simulations, we have considered separately the interplay of the CM18 and Tat11 peptides with the membrane patch during electric field exposure. As shown in Figure 5 of the Supporting Information, CM18 can be dragged into the pore and reoriented along the pore axis (Figure 5 of the Supporting Information, top panels), even if its motion is sensibly slower

the presence of CM18-Tat11 and control peptides. Because of obvious computational limitations, we modeled only a fraction of the vesicle surface, i.e., the portion that is perpendicular to the applied field. We took advantage of the available secondary structure in the hydrophobic medium of the CM18-Tat11 peptide previously determined23 and considered planar POPC bilayer patches in which a single peptide is placed with its unstructured portion extruding from the membrane surface and its α-helical portion ∼5 Å below the membrane−solution interface and parallel to the membrane surface (following the work performed by Bhargava et al.30 on the CM15 analogue) (see the Experimental Section for further simulation details). Such localization is well-suited to burying nonpolar side chains in the hydrophobic core of the membrane, while positioning lysine residues to interact with lipid phosphate. F



Article

Figure 4. Schematic model of CM18-Tat11 and NP cooperative endosomal vesicle perturbation and pDNA release. The CM18 α-helix is colored red, the Tat11 unstructured coil blue, and pDNA green. The model pictures the fate of the peptide−pDNA complex after cell internalization by endocytosis from the external medium. In the first panel, the endosomal vesicle shows the accumulation of the peptide on the vesicle internal bilayer surface at low concentrations after disassembly of the complex. In the second panel is pictured the stochastic formation of nanopores after the administration of a single electric pulse to the cell, while the third panel displays how peptides are located on the edge of a nanopore defect in the bilayer after the electric field is switched off. The final outcome of the administration of multiple pulses is exhibited in the last vesicle picture where the proximity of several membrane defects leads to local bilayer micellization and release of pDNA molecules into the cytosolic lumen.

Interestingly, the co-administration of CM18-Tat11 and NPs shows no increase in toxicity (Figure 2b). Although NPs are able to form pores both on external and on internal membranes, CM18-Tat11 stabilizes only the latter and avoids any permanent damage at the level of the plasma membrane (see Figure 4 of the Supporting Information). Finally, on the basis of atomistic MD simulations of planar POPC bilayer patches subjected to 10 ns pulses in the presence of the various peptides (Figure 3 and Figures 5 and 6 of the Supporting Information), we were able to shed light on the molecular mechanism at the basis of the cooperative perturbing action observed in our experiments. In fact, MD data show that the CM18 module acts as a catalyst to stabilize the NP-elicited membrane defects, either alone or fused to Tat11. The peptide, if located close enough to the pore mouth, can be dragged electrophoretically along the field direction into the hydrophilic pore while keeping its structural stability. As soon as the electric field is switched off, the nanopore starts to collapse but the complete reconstitution of the membrane is impaired by the persistency of the CM18-Tat11 peptide along the defect created in the bilayer. This scenario is not encountered by the isolated Tat11 module, whose high electrophoretic mobility during NP administration (probably because of its high positive charge density) and most importantly low binding activity to the lipid membrane lead to rapid and complete translocation within a single 10 ns pulse. On the basis of the data presented here, as schematically represented in Figure 4, we propose a mechanism in which CM18-Tat11 first accumulates on the internal surface of the endocytic vesicle bilayer. Then, NP-induced nanopores randomly form in different locations on the vesicle membrane. Each time a CM18-Tat11 module falls close to the pore mouth, it can be electrophoretically dragged into the pore lumen, stabilizing therefore the defect created by the NP. Multiple events eventually lead to local bilayer micellization, comporting the unsealing of the vesicle wall and the release of pDNA molecules in the cytosolic lumen. Altogether, our results, besides providing insights onto peptide−NP cooperation, enlighten the crucial role played by CM18 in the membrane destabilization process observed also in previous reports.23,24 This study has demonstrated for the first time that electric nanopulses can be successfully exploited for gene transfer applications through the co-administration of CM18-Tat11 at nanomolar concentrations. We believe that this result opens the way to further studies that combine external stimuli, like NPs, to smartly designed vectors that act as catalysts for targeted delivery.

than that of CM18-Tat11, probably because of its lower content of positive charges. As the electric field switches off (Figure 5 of the Supporting Information, bottom panels), CM18 behaves like the chimeric peptide and introduces a defect into the bilayer. On the other hand, the Tat11 module (placed in the proximity of the lipidic bilayer according to the report of Herce et al.32) does interact with phosphate negative charges and is strongly affected by the electric field because of its high positive charge density. As a result, it is not able to linger on the hydrophilic nanopore long enough to impair the complete reversibility of the electroporation process (see Figure 6 of the Supporting Information).

■

DISCUSSION We propose herein a new gene transfer methodology based on the combination of NPs and a recently introduced endosomolytic peptide, CM18-Tat11. The rationale was to test the possibility of using synergistically the two approaches to lower the required CM18-Tat11 concentration and, at the same time, to confine NP-induced nanopore formation to vesicles with negligible perturbation to the plasma membrane. Preliminary in cuvette tests of GUVs proved that our hypothesis was viable: 10 ns electric pulses administered to CM18- or CM18-Tat11-loaded GUVs lead to a significant (more than 66%) decrease in the number of stable vesicles (Figure 1c,d). This is not the case for Tat11-loaded GUVs (Figure 1d). We then tested the same procedure on endosomal vesicles of living cells. These are much more complex membrane systems and represent at the moment one of the most important obstacles to overcome for gene transfer by nonviral vectors.33−35 Electrostatic interaction between pDNA molecules and CM18-Tat11 peptides makes it possible to produce a series of transfectant nanoparticles with different N:P ratios (i.e., charge ratios).24 Our previously published data show that the 16:1 N:P ratio complex administered at micromolar concentrations in the medium represents a very efficient in vitro transfectant system.24 On the other hand, the 4:1 complex, despite its identical stability, does not exhibit a significant gene transfer activity at any tested concentration, most likely because of the accumulation of CM18 within vesicles below its threshold activity. Notably, the TE of the 4:1 complex administered at nanomolar concentrations can be increased at levels comparable with transfection gold standards by the concomitant exposure of cells to 40 ns pulses. The 4-fold higher pulse length required for membrane bilayer disruption in endosomal vesicles compared to GUVs prompts us to speculate that a distinct composition and/or environment (e.g., presence of the plasma membrane and actin skeleton22) might hamper the effectiveness of the electroporative action. G



■

Article

translocation and caspase activation: Differential sensitivities of Jurkat T lymphoblasts and rat glioma C6 cells. IEEE Trans. Dielectr. Electr. Insul. 2003, 10, 795−809. (15) Vernier, P. T.; Sun, Y.; Gundersen, M. A. Nanoelectropulsedriven membrane perturbation and small molecule permeabilization. BMC Cell Biol. 2006, 7, 37. (16) Ibey, B. L.; Xiao, S.; Schoenbach, K. H.; Murphy, M. R.; Pakhomov, A. G. Plasma membrane permeabilization by 60- and 600ns electric pulses is determined by the absorbed dose. Bioelectromagnetics 2009, 30 (2), 92−99. (17) Pakhomov, A. G.; Kolb, J. F.; White, J. A.; Joshi, R. P.; Xiao, S.; Schoenbach, K. H. Long-lasting plasma membrane permeabilization in mammalian cells by nanosecond pulsed electric field (nsPEF). Bioelectromagnetics 2007, 28 (8), 655−663. (18) Vernier, P. T.; Sun, Y.; Marcu, L.; Salemi, S.; Craft, C. M.; Gundersen, M. A. Calcium bursts induced by nanosecond electric pulses. Biochem. Biophys. Res. Commun. 2003, 310 (2), 286−295. (19) Napotnik, T. B.; Rebersek, M.; Kotnik, T.; Lebrasseur, E.; Cabodevila, G.; Miklavcic, D. Electropermeabilization of endocytotic vesicles in B16 F1 mouse melanoma cells. Med. Biol. Eng. Comput. 2010, 48 (5), 407−413. (20) Chen, N.; Schoenbach, K. H.; Kolb, J. F.; James Swanson, R.; Garner, A. L.; Yang, J.; Joshi, R. P.; Beebe, S. J. Leukemic cell intracellular responses to nanosecond electric fields. Biochem. Biophys. Res. Commun. 2004, 317 (2), 421−427. (21) Silve, A.; Leray, I.; Mir, L. M. Demonstration of cell membrane permeabilization to medium-sized molecules caused by a single 10 ns electric pulse. Bioelectrochemistry 2011, 87, 260−264. (22) Breton, M.; Delemotte, L.; Silve, A.; Mir, L. M.; Tarek, M. Transport of siRNA through lipid membranes driven by nanosecond electric pulses: An experimental and computational study. J. Am. Chem. Soc. 2012, 134 (34), 13938−13941. (23) Salomone, F.; Cardarelli, F.; Di Luca, M.; Boccardi, C.; Nifosi, R.; Bardi, G.; Di Bari, L.; Serresi, M.; Beltram, F. A novel chimeric cellpenetrating peptide with membrane-disruptive properties for efficient endosomal escape. J. Controlled Release 2012, 163 (3), 293−303. (24) Salomone, F.; Cardarelli, F.; Signore, G.; Boccardi, C.; Beltram, F. In Vitro Efficient Transfection by CM18-Tat11 Hybrid Peptide: A New Tool for Gene-Delivery Applications. PLoS One 2013, 8 (7), e70108. (25) Angelova, I. A.; Dimitrov, S. D. Liposome electroformation. Faraday Discuss. Chem. Soc. 1986, 1986 (81), 303−311. (26) Delemotte, L.; Tarek, M. Molecular dynamics simulations of lipid membrane electroporation. J. Membr. Biol. 2012, 245 (9), 531− 543. (27) Kale, L.; Skeel, R.; Bhandarkar, M.; Brunner, R.; Gursoy, A.; Krawetz, N.; Phillips, J.; Shinozaki, A.; Varadarajan, K.; Schulten, K. NAMD2: Greater scalability for parallel molecular dynamics. J. Comput. Phys. 1999, 151 (1), 283−312. (28) Huang, J.; MacKerell, A. D., Jr. CHARMM36 all-atom additive protein force field: Validation based on comparison to NMR data. J. Comput. Chem. 2013, 34 (25), 2135−2145. (29) Levine, Z. A.; Vernier, P. T. Life cycle of an electropore: Fielddependent and field-independent steps in pore creation and annihilation. J. Membr. Biol. 2010, 236 (1), 27−36. (30) Bhargava, K.; Feix, J. B. Membrane binding, structure, and localization of cecropin-mellitin hybrid peptides: A site-directed spinlabeling study. Biophys. J. 2004, 86 (Part 1), 329−336. (31) Tieleman, D. P. The molecular basis of electroporation. BMC Biochem. 2004, 5, 10. (32) Herce, H. D.; Garcia, A. E. Molecular dynamics simulations suggest a mechanism for translocation of the HIV-1 TAT peptide across lipid membranes. Proc. Natl. Acad. Sci. U.S.A. 2007, 104 (52), 20805−20810. (33) Medina-Kauwe, L. K.; Xie, J.; Hamm-Alvarez, S. Intracellular trafficking of nonviral vectors. Gene Ther. 2005, 12 (24), 1734−1751. (34) Fonseca, S. B.; Pereira, M. P.; Kelley, S. O. Recent advances in the use of cell-penetrating peptides for medical and biological applications. Adv. Drug Delivery Rev. 2009, 61 (11), 953−964.

ASSOCIATED CONTENT

S Supporting Information *

Additional observations and figures. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Piazza San Silvestro 12, 56127 Pisa, Italy. Telephone: 0039050509124. Fax: 0039050509417. E-mail: fabrizio. [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This research was performed in the scope of the EBAM European Associated Laboratory and of COST Action TD1104 and is partly funded by the CNRS, the University of Paris-Sud, Gustave Roussy, the Fondation EDF, and the French National Agency for Research (ANR) through the Intcell ANR-10BLAN-916, IPSIOAT ANR-11-BS09-0031, and Memove ANR11-BS01-006 grants.

■

REFERENCES

(1) Neumann, E.; Rosenheck, K. Permeability changes induced by electric impulses in vesicular membranes. J. Membr. Biol. 1972, 10 (3), 279−290. (2) Teissie, J.; Eynard, N.; Gabriel, B.; Rols, M. P. Electropermeabilization of cell membranes. Adv. Drug Delivery Rev. 1999, 35 (1), 3−19. (3) Escoffre, J. M.; Teissie, J.; Rols, M. P. Gene transfer: How can the biological barriers be overcome? J. Membr. Biol. 2010, 236 (1), 61−74. (4) Golzio, M.; Teissie, J.; Rols, M. P. Direct visualization at the single-cell level of electrically mediated gene delivery. Proc. Natl. Acad. Sci. U.S.A. 2002, 99 (3), 1292−1297. (5) Paganin-Gioanni, A.; Bellard, E.; Escoffre, J. M.; Rols, M. P.; Teissie, J.; Golzio, M. Direct visualization at the single-cell level of siRNA electrotransfer into cancer cells. Proc. Natl. Acad. Sci. U.S.A. 2011, 108 (26), 10443−10447. (6) Satkauskas, S.; Bureau, M. F.; Puc, M.; Mahfoudi, A.; Scherman, D.; Miklavcic, D.; Mir, L. M. Mechanisms of in vivo DNA electrotransfer: Respective contributions of cell electropermeabilization and DNA electrophoresis. Mol. Ther. 2002, 5 (2), 133−140. (7) Mir, L. M. Nucleic acids electrotransfer-based gene therapy (electrogenetherapy): Past, current, and future. Mol. Biotechnol. 2009, 43 (2), 167−176. (8) Gehl, J.; Mir, L. M. Determination of optimal parameters for in vivo gene transfer by electroporation, using a rapid in vivo test for cell permeabilization. Biochem. Biophys. Res. Commun. 1999, 261 (2), 377− 380. (9) Schoenbach, K. H.; Beebe, S. J.; Buescher, E. S. Intracellular effect of ultrashort electrical pulses. Bioelectromagnetics 2001, 22 (6), 440− 448. (10) Schoenbach, K. H.; Joshi, R.; Kolb, J.; Buescher, S.; Beebe, S. Subcellular effects of nanosecond electrical pulses. Conference proceedings: Annual International Conference of the IEEE Engineering in Medicine and Biology Society 2004, 7, 5447−5450. (11) Kolb, J. F.; Kono, S.; Schoenbach, K. H. Nanosecond pulsed electric field generators for the study of subcellular effects. Bioelectromagnetics 2006, 27 (3), 172−187. (12) Sundararajan, R. Nanosecond electroporation: Another look. Mol. Biotechnol. 2009, 41 (1), 69−82. (13) Breton, M.; Mir, L. M. Microsecond and nanosecond electric pulses in cancer treatments. Bioelectromagnetics 2011, DOI: 10.1002/ bem.20692. (14) Vernier, P. T.; Li, A.; Marcu, L.; Craft, C. M.; Gundersen, M. A. Ultrashort pulsed electric fields induce membrane phospholipids H



Article

(35) El-Sayed, A.; Futaki, S.; Harashima, H. Delivery of macromolecules using arginine-rich cell-penetrating peptides: Ways to overcome endosomal entrapment. AAPS J. 2009, 11 (1), 13−22.

I


High-Yield Nontoxic Gene Transfer through Conjugation of the CM18

Recommend Documents