RESEARCH ARTICLE Cytoskeleton, October 2014 71:587–594 (doi: 10.1002/cm.21194) C V
2014 Wiley Periodicals, Inc.
AFM Sensing Cortical Actin Cytoskeleton Destabilization During Plasma Membrane Electropermeabilization Chopinet Louise,1,2,3 Dague Etienne,1,3,4* and Rols Marie-Pierre2,3* 1
Centre National de la Recherche Scientifique, Laboratoire d’Analyse et d’Architecture des Syste`mes (LAAS), NanoBioSystem Group, Toulouse, F-31400, France 2 Centre National de la Recherche Scientifique, Institut de Pharmacologie et de Biologie Structurale (IPBS), Cell Biophysics Group, UMR 5089, BP64182, F-31077, Toulouse Cedex 4, France 3 Universite de Toulouse, UPS, INSA, INP, ISAE, UT1, UTM, LAAS, ITAV, F-31077, Toulouse Cedex 4, France 4 Centre National de la Recherche Scientifique, Institut des Technologies Avancees en Sciences du Vivant (ITAV), Nanocardiology and Cancer Cell Biophysics Group, USR3505, F31106, Toulouse, France
Received 17 April 2014; Revised 29 September 2014; Accepted 3 October 2014 Monitoring Editor: Manuel Thery
Electropermeabilization is a physical method that uses electric field pulses to deliver molecules into cells and tissues. Despite its increasing interest in clinics, little is known about plasma membrane destabilization process occurring during electropermeabilization. Having previously demonstrated the usefulness of Atomic Force Microscopy to study electropermeabilization effect on plasma membrane, we further investigated the plasma membrane destabilization process. We got mainly interested in the cytoskeleton role in stiffness of the plasma membrane, and thus in the effect of electric field on the cortical actin network. We show here that cortical actin is destabilized by electric pulses and that this effect is not directly related to the electropermeabilization of the plasma membrane. V 2014 Wiley Periodicals, Inc. C
Key Words:
electroporation; atomic force microscope; living cells; actin cytoskeleton; force spectroscopy; quantitative imaging
Additional Supporting Information may be found in the online version of this article. *Address correspondence to: Rols Marie-Pierre, Centre National de la Recherche Scientifique, Institut de Pharmacologie et de Biologie Structurale (IPBS), UMR 5089, BP64182, F-31077 Toulouse Cedex 4, France. E-mail: [email protected] correspondence to: Dague Etienne, Centre National de la Recherche Scientifique, Laboratoire d’Analyse et d’Architecture des Syste`mes (LAAS), NanoBioSystem Group, Toulouse, F-31400, France. E-mail: [email protected] Abbreviations used: AFM, atomic force microscopy; CHO, Chinese hamster ovary cell; EP, electropermeabilization or electroporation; LatB, Latrunculine B treatment; QITM, Quantitative ImagingTM; YM, Young modulus. Published online 11 October 2014 in Wiley Online Library (wileyonlinelibrary.com).
Introduction
E
lectropermeabilization (EP), or electroporation, is a non-viral vectorization technique based on electric field pulses application [Neumann and Rosenheck, 1972]. It is now used in clinics for cancer treatment. Electrochemotherapy is a standard treatment in many cancer centers around Europe that potentiates the delivery of cytotoxic drugs by local electric pulses delivery (http://guidance.nice.org.uk/ IPG446) [Mir, 2006; Caraco et al., 2013]. EP can also be used for gene transfer, and its most famous application resides in biotechnology for bacterial transformation. Moreover, it presents great promises in gene therapy and vaccination; clinical trials are undergoing [Heller et al., 2002; Daud et al., 2008; Gehl, 2008; Low et al., 2009; Chiarella et al., 2013]. Nevertheless, there is still a lack of knowledge about how electric fields are acting on cell components (mainly membrane) to permit entry of molecules and/or macromolecules (such as DNA) into cell cytoplasm [Teissie et al., 2005]. Most of the studies focusing on plasma membrane permeabilization description have been conducted using fluorescent dyes and only gave partial results based on indirect visualization and measurement of the phenomena [Escoffre et al., 2009]. Recently, we exploited Atomic Force Microscopy to investigate membrane permeabilization induced by EP at the nanoscale. We imaged and measured locally by force spectroscopy the consequences of EP on living cells plasma membrane elasticity [Chopinet et al., 2013b]. These previous results have shown that AFM allows visualizing membrane permeabilization with a different time scale than propidium iodide (PI) and other molecules entry. Indeed, we showed that the membrane elasticity decreases slower than the PI enter the cell after electric pulses application. This mean that even if membrane permeabilization is locally induced in specific regions of the cell and is transient (as seen by fluorescence imaging), its 587 䊏
consequences on stiffness are global and still present when membrane resealing has occurred. Moreover, this membrane destabilization has been observed independently to the cell orientation toward the electric field, in the contrary of molecules uptake (e.g., DNA) that takes place only at the face of the cell facing the cathode (which is due to electrophoresis processes and probably not to an orientated membrane destabilization). These results pointed out the ability of AFM to investigate electropermeabilization effects, and more specifically, we showed that AFM senses an impact of electric field different than plasma membrane permeabilization. Our results obtained both on fixed and living CHO cells (Chinese Hamster Ovary cells) give evidence of an inner effect affecting the entire cell surface that may be related to cytoskeleton destabilization. In 2000, Rotsch and Radmacher have shown that decrease of young modulus can be related specifically to the actin cytoskeleton destabilization, and not microtubule one [Rotsch and Radmacher, 2000]. The actin cytoskeleton network is composed of globular monomer of actin polymerized in filaments. These filaments are interacting with plasma membrane directly to lipid or through transmembrane proteins [Lodish et al., 2000]. Two filaments types are described: microfilaments that are composed of several filaments interconnected thanks to linking proteins, and cortical actin network which is located underneath plasma membrane in a sustaining dense network composed of solely filaments [Alberts et al., 2002]. The cortical actin network has been shown to be detected in the first 100 nm of indentation [Fels et al., 2012]. Very few articles get interested in EP effects on actin. In 1992, Rols and Teissie have shown that cytoskeleton was involved in the resealing process [Rols and Teissie, 1992]. They showed that CHO cells with depolymerized actin are resealing slower than control cells. In 1994, same authors gave further description for the implication of cytoskeleton in permeabilization phenomenon [Teissie and Rols, 1994]. Other studies performed on the cytoskeleton of endothelial cells have shown a profound disruption of microfilament and microtubule cytoskeletal networks after EP, paralleled by a rapid increase in endothelial monolayer permeability [Kanthou et al., 2006]. These results showing the remodeling of the endothelial cytoskeleton and changes in endothelial barrier function provide an insight into putative mechanisms responsible for the observed increase in permeability and cessation of blood flow in vivo. More recently, actin has been shown to play an important role in plasmid internalization and intracellular traffic during electrotransfer, as actin is polymerizing under DNA spots at the membrane [Rosazza et al., 2011, 2013]. There are thus evidences for the implication of actin in plasma membrane electropermeabilization but without any direct correlation. AFM allows physical measurement of actin destabilization on living cell along time and can thus enhance the mechanical characterization and understanding
䊏 588
Louise et al.
of permeabilization process. This is why in the present article we chose to use this instrument to investigate the implication of cortical actin in the measured increase in elasticity after EP, and performed experiments on cells treated with the drug Latrunculine B which affects actin polymerization (LatB treatment) [Coue et al., 1987; Rosazza et al., 2011]. Imaging and measuring plasma membrane have been performed following the same protocol as in Chopinet et al. [2013b]. Force volume (FV) acquisition has been realized on 4 areas on each cell. Images have been acquired thanks to the multiparametric mode “Quantitative Imaging” from JPK [Chopinet et al., 2013a; Dufr^ene et al., 2013]. We show that AFM technology is of great interest to resolve biological questions such as deciphering between two treatments effect namely electropermeabilization and drug treatment by Latrunculine B. Especially, it provides results about inner component of the cell, namely cortical actin. We demonstrate in this article that electropermeabilization is affecting cortical actin network interaction with plasma membrane through the lipid bilayer destabilization by electric field thanks to the unique use of AFM.
Results Drug efficiency
The efficiency of the treatment by LatB has been assessed by confocal microscopy thanks to cells transfected for actinGFP protein 24 h prior the treatment (Fig. 1) [Rosazza et al., 2011]. The labeled actin is uniformly distributed in control cells, and aggregated in treated cells. Latrunculin effects are directly visible on cells by phase contrast as shown by changes in the shape of the cells (Figs. 1A0 and 1B0 ). Thus during AFM experiments, drug efficiency has been checked only by phase contrast, avoiding the use of actin-GFP transfected cells that may create bias related to this procedure and the production of the GFP coupled actin protein (Fig. 1C). Correlation Between EP and Latrunculine Effect
To validate that the decrease in elasticity observed in our previous work is related to the destabilization of the cortical actin network [Chopinet et al., 2013b], we compared elasticity in electropermeabilized cells and in cells treated with Latraculine B (LatB). The same experimental procedure as in Chopinet et al. [2013b] was used. The analysis of cell elasticity in the 50 first nanometers of indentation before and after LatB treatment shows a decrease in elasticity of around 30 kPa (Fig. 2A, Supporting Information Table I). Measurement of cells treated and placed in pulsation buffer at t 5 0 min show a recovery of elasticity along time (Fig. 2B). The difference between elasticity at time 25 min (when actin is completely depolymerized) and 33 min (when actin has repolymerized) is the same as the ones observed before and after electric field application (around CYTOSKELETON
Fig. 1. LatB effect on actin polymerization. (A) Confocal imaging of control untreated cells expressing a GFP labeled actin protein, fluorescence image (A) and merge with phase contrast image (A’). (B) Confocal imaging of cells treated with 0.1 mM Latrunculine B during 1 h at 37 C, fluorescence (B) and merge with contrast phase (B’). (C) A cell recovery after removing the drug followed by phase contrast imaging during an AFM experiment.
10 kPa, [Chopinet et al., 2013b]). However if electric fields are applied on cells treated with LatB (always at time 5 0 when the drug is removed and the cells are placed in pulsation buffer), cells are not recovering their initial elasticity (Fig. 2C and Supporting Information Table I). Imaging with QI mode reveals the de novo formation of actin filaments in cells when the drug is removed. Height images show the elongation of cells and production of actin filament in the longitudinal way (Fig. 3A). However on cell submitted to electric fields, even if the cell still seems to spread along time, it is not possible to distinguish filaments underneath plasma membrane (Fig. 4A). The force curve analysis performed with OpenFovea software [Roduit et al., 2012] has been realized on several segments of 50 nm of indentation every 50 nm along the force curve following the point of contact between the tip and the sample. Actin filaments are visible on segments located at 100 nm depth as they are localized underneath the membrane. The elasticity maps are showing these filaments that are stiffer than the others cell components (Fig. 3B). Elasticity maps are showing the same differences that the ones observed on height images: LatB-treated cell submitted to electric field are unable to produce actin filament de novo. Cells are able to do so along time only when no electric fields are applied. Cortical Actin or Microfilament?
To go deeper in the interpretation of the phenomenon observed on cortical actin, we performed experiment combining electric field application and drug treatment. As presented in Fig. 5, we first applied electric field, acquire CYTOSKELETON
elasticity measurement, then added the drug to the media, or, we first added the drug, and then applied the electric field. Results show that when EP is applied before the drug is added, a supplementary decrease in elasticity is measured (Fig. 5A, P value < 0.05). In the contrary, when the drug is added before EP, there is no such decrease observed in elasticity (Fig. 5B). These results show that even if EP and LatB treatment result in a decrease in elasticity, they do not act similarly on it. LatB treatment has a stronger impact on elasticity than EP, and thus provokes a supplementary decrease in elasticity when the drug is added after pulse electric field application.
Discussion Understanding the mechanism governing molecules electrotransfer in cells by electric pulses is a prerequisite for the development of more efficient clinical protocols based on this approach. This concerns both the use of the method as an efficient way to deliver anticancer drugs as well as its use for DNA vaccination or gene therapy. Therefore, future clinical use requires a much better knowledge of the processes involved at the cellular level. Our interest was focused on unraveling the role of the actin cytoskeleton in stiffness of the plasma membrane and in the effect of electric pulses on cortical actin network. We previously observed, on cells submitted to EP, a decrease in elasticity of the 50 first nm of indentation followed by a gradual recovery along the 35 min following electric field
AFM Sensing Actin Cytoskeleton During Electropermeabilization
589 䊏
Fig. 2. Membrane elasticity along time following LatB treatment. (A) Cell treated, the drug is added at time 0 and corresponding force curves (n 5 3 cells). Before treatment t 5 27 min: YM 5 15.47 kPa 6 1.8 (SE) (1024 force curves); after treatment t 5 33 min: YM 5 10.65 kPa 6 0.66 (SE) (1024 force curves). (B) Recovery of elasticity after drug removal at time 0 (n 5 3 cells). Before drug removal t 5 25 min: YM 5 11.33 kPa 6 0.28 (SE) (1024 force curves); after drug removal t 5 7 min: YM 5 12.59 kPa 6 0.64 (SE) (1024 force curves); after drug removal t 5 33 min: YM 5 17.33 kPa 6 1.66 (SE) (1024 force curves). (C) No recovery in elasticity after drug removal and electric field application at time 0 and corresponding force curves (n 5 3 cells). Before drug removal and electric field application t 5 25 min: YM 5 8.62 kPa 6 0.69 (SE) (1024 force curves); after drug removal and electric field application t 5 32 min: YM 5 9.22 kPa 6 0.57 (SE) (1024 force curves). Unpaired t test relative to the point before t 5 0: *P value < 0.05, **P value < 0.005, ***P value < 0.0001. Detailed data in Supporting Information.
application. Cells treated with LatB are displaying the same decrease in elasticity and the same recovery (Fig. 2). This similitude between EP and drug treatment reinforces the idea that pulsed electric field have an effect on actin cytoskeleton. QI height images allowed going further in the investigation. On the one hand, cells submitted only to EP do not display actin microfilament [Chopinet et al., 2013b]. This event occurs together with cell swelling, in absence of any change in cell shape which can indicate that cytoplasmic
䊏 590
Louise et al.
actin (actin microfilament, in the contrary of cortical actin which is located at the membrane) is not destabilized by electric pulses. Indeed, as visible on cells treated with LatB, actin depolymerization is affecting cell shape (Figs. 3 and 4). However after EP, the plasma membrane shows ripples that can point out an impact on cortical actin [Chopinet et al., 2013b]. On the other hand, images of cells treated with LatB show de novo formation of stress fibers after drug removal. When electric field pulses are applied simultaneously with drug removal, no microfilament are imaged, but cells spread gradually along time, leading to the hypothesis that cytoplasmic actin (microfilament linked to focal adhesion points) can be able to polymerize, while cortical actin is affected by electric pulses. On the whole, these observations lead to consider that electric field pulses are affecting cortical actin and not actin filaments. Indeed, as a first observation, cells are swelling which can indicate an effect on membrane/cytoskeleton interaction. Second, the impossibility to image microfilament after EP leads to the interpretation that what is interacting with plasma membrane is destabilized (directly by electric field, or indirectly by the plasma membrane electrodestabilization itself ). This event is first decreasing elasticity in the first 50 nm of indentation, and second is avoiding the imaging of inner microfilaments by releasing membrane from its anchoring. This interpretation can be supported by previous work using the same electric parameters and the same cells, that showed unaffected actin microfilaments [Rosazza et al., 2011]. Moreover, it has recently been shown that the 100 firsts nm of indentation display elasticity linked to cortical actin network [Fels et al., 2012]. Thus, altogether, these observations intend to show that the measured elasticity after EP is linked to cortical actin cytoskeleton destabilization. At this point, two hypotheses can be formulated: (1) electric pulses are affecting directly cortical actin filaments and destabilizing their structure leading to depolymerization or rupture. This effect could be involved in the permeable state of membrane. It is known that one EP side effect is the release of ATP after membrane permeabilization [Delteil et al., 2000] and can explain an effect on actin polymerization which is ATP-dependent. (2) Effect of electric pulses on plasma membrane, i.e. lipid bilayer destabilization leading to permeabilization [Lopez et al., 1988], can provoke a rupture of the interaction between cortical actin network and plasma membrane, which can release the membrane and cancel any resistance at the membrane level leading to a decrease in elasticity. The last experiment consisting in combining EP and drug exposure helps to go further. As shown in Fig. 4A, the Latrunculin drug provokes a supplementary effect when added after the electric pulses application, while applying electric pulses on cells already exposed to the drug does not have any supplementary effect (Fig. 5B). These results show CYTOSKELETON
Fig. 3. Recovery of actin cytoskeleton after 1-h treatment with LatrunculinB 0.1 mM at 37 C. t0 correspond to the change from media containing LatrunculinB to media free of any drug. Cells have been kept at 37 C in HEPES media during all the QITM measurements (75 3 75 pixels). (A) Young modulus map at 100 nm of depth along time in minutes, colorscale: from 0 (dark blue) to 60 kPa (dark red), one segment equals 8 kPa. (B) Corresponding height images, scale bar: 5 mm, color scale: 0–4 mm. Extra cells visible in Supporting Information.
that the drug has a more dramatic effect than EP does, and thus encourages the hypothesis (1). Indeed, as Latrunculin is known to destabilize actin filaments, it means that it will affect every filament, and provokes the decrease in elasticity by a global effect. However, EP has a more spared effect that can thus be considered as the destabilization of cortical actin network interaction with plasma membrane rather than the actin depolymerization itself. To address the remaining question about the fact that EP and LatB treatment have the same effect in term of elasticity, it is necessary to consider that cortical actin depolymerization and disruption of actin-membrane interaction will lead to the same result: a decrease in membrane elasticity in the 50 first nm, even if the mechanisms responsible for this decrease are different. Indeed, the height images of cells confirm these considerations. LatB treated cells still spread
on the substrate (meaning that actin filaments are polymerizing) after EP application but no actin filaments are imaged or measured underneath the membrane. This observation is consistent with the fact that EP may affect the actin-membrane interaction rather than the actin polymerization itself. This effect avoids imaging actin filament after EP when membrane is totally weak and floppy. Moreover, we have shown here that elasticity is not recovered when EP is applied after drug removal, which also implies that the plasma membrane remains floppy even if the cells are able to repolymerize their actin filaments, once again confirming that only the interaction between actin and membrane is not recovered. The two effects (of EP and of LatB) are clearly different as it can be seen on Fig. 5A where the addition of the two treatments creates two different steps in elasticity decrease.
Fig. 4. Absence of recovery of actin cytoskeleton after 1-h treatment with LatrunculinB 0.1 mM and pulse application at t0. T0 correspond to the change from media containing LatrunculinB to media free of any drug combined to pulse application. Cells have been kept at 37 C in HEPES media during all the QITM measurements (75 3 75 pixels). (A) Young modulus map at 100 nm of depth along time in minutes, colorscale: from 0 (dark blue) to 60 kPa (dark red), one segment equals 8 kPa. (B) Corresponding height images, scale bar: 5 mm, color scale: 0–4 mm. Extra cells visible in Supporting Information.
CYTOSKELETON
AFM Sensing Actin Cytoskeleton During Electropermeabilization
591 䊏
Fig. 5. Effect of combination of the two treatment. (A) Supplementary effect on elasticity when the drug is added at time 10 min after pulse application at time 0 and corresponding force curves (n 5 3 cells). Before electric field application t 5 210 min: YM 5 14.83 kPa 6 1.3 (SE); after electric field application t 5 5 min: YM 5 10.85 kPa 6 1.62 (SE) (1024 force curves); after drug treatment t 5 15min: YM 5 8.28 kPa 6 0.32 (SE) (1024 force curves). (B) No combined effect when pulses are applied at time 12 after drug is added at time 0 and corresponding force curves (n 5 3 cells). Before drug treatment t 5 26 min: YM 5 14.67 kPa 6 0.97 (SE) (1024 force curves); after drug treatment t 5 6 min: YM 5 9.66 kPa 6 0.5 (SE) (1024 force curves); after electric field application t 5 20 min: YM 5 10.84 kPa 6 1.02 (SE) (1024 force curves. Unpaired t test relative to the point before t 5 0: *P value < 0.05, **P value < 0.005, ***P value < 0.0001. Detailed data in Supporting Information.
It thus confirms that EP has a different role than Latrunculin in actin destabilization. Moreover, it is known from the literature [Rols and Teissie, 1992] that Cytochalasin treated cells, as well as Table I.
Latrunculin treated cells stay permeable to propidium iodide. This information makes a direct link between the EP effect on membrane, namely its permeabilization, and the role of actin in plasma membrane properties. As actin depolymerization is responsible for the persistent permeabilization state of plasma membrane, and as we shown that LatB-treated cells exposed to EP do not recover their elasticity, we can thus conclude that EP is only affecting actinmembrane interaction through the destabilization of lipid bilayer. To finish, we can say that EP never affects cell spreading; meaning EP does not provoke actin depolymerization. Table I provided below helps organizing the effects observed, and thus supports the hypothesis that EP is affecting cortical actin through the destabilization of lipid bilayer and disruption of actin/membrane interactions. To conclude, we show in this article that AFM provides a new approach to study biophysical mechanisms, and specifically the analysis of electropermeabilization effect on internal component like cytoskeleton. We demonstrated the ability of AFM to discriminate between two treatments effects and thus to investigate, at the molecular level, the effects of electric fields on membrane permeability properties.
Materials and Methods Sample Preparation
Nearly 75,000 CHO cells were grown in Petri dish during 24 h in the same cell culture conditions described in Chopinet et al. [2013b]. Before measurements, classical MEM medium was replaced by MEM-HEPES medium (CM1MEM46-6U, Eurobio, France) supplemented with 8% fetal calf serum (Lonza Group, Switzerland) and cells were placed in the PetriDishHeater (JPK) that maintained
Sum up of the Main Results Presented in the Article and in Chopinet et al 2013b Elasticity
Recovery
Filament
Spreading
-
1
-
5
Lipid destabilization
1LatB -LatB 1LatB1EP
/ 1
1 -
1 -
1EP1LatB
–
-
-
Actin depolymerization Actin repolymerization Actin depolymerization 1 lipid destabilzation Lipid destabilization 1 actin depolymerization
-LatB1EP
-
-
-
1EP
䊏 592
Louise et al.
Known effect
Actin depolymerization 1 lipid destabilzation
Assumed effect
Disruption of actin-membrane interaction
Not possible to reform the interaction 1. Actin-membrane interaction is disrupted; 2. Actin is depolymerized Actin is already depolmerized, thus no disruption. However: no new interaction with membrane because no actin available, which means that elasticity remains low.
CYTOSKELETON
30 C during all the experiment to slow down the membrane activity and process to monitor. Prior to electrical pulse application, cells were washed three times with phosphate-buffered saline (PBS) 13 (Invitrogen, USA). About 1 mL of pulsation buffer (PB) at 4 C (10 mM K2HPO-4/KH2PO42buffer, 1 mM MgCl2, 250 mM sucrose, pH 7.4) was added and electrodes were placed in contact with Petri dish bottom. Pulsed were delivered and AFM head was replaced.
was used for force curve analysis with the following settings on 50 nm of indentation: Model: cone; Tip size: 0.62 rad; Poisson ratio: 0.5; Method: raw. Frequency distributions were obtained with OriginPro 8 software, and histograms were plotted with GraphPad Prism software, that was used for statistical analysis too (unpaired t-test, *P value < 0.05, **P value < 0.005, ***P value < 0.0001), results are presented as boxplots with min and max value, and median. See Chopinet et al. [2013b] for further details.
Latrunculine Treatment
Acknowledgments
A 0.1 mM of Latrunculine B (Sigma–Aldrich, L5288) was added to the cells 1 h before experiments or after pulses application (mentioned in the text when drug is added before or after EP to allow measurement of the effects of both treatment). The maximum volume of DMSO in which the drug is solubilized was 2& of the total volume of media. The drug is removed by taking out the medium of the petri dish, washing twice with PBS, and then adding a drug free media (pulsation buffer or HEPES media).
The authors thank Marie-Charline Blatche for helping in cell sample preparation, Direction Generale de l’Armement for 3 years PhD grant to Louise Chopinet, ANR Young Scientist Program “AFMyst” (#30024332) and ANR Astrid “PIERGEN” (#ANR2122ASTR20039201) that provided financial support. This report is the result of networking efforts of COST Action TD1104 (http://www. electroporation.net) and was conducted in the scope of the EBAM European Associated Laboratory.
Electropermeabilization
References
The generator used for this study was a Jouan (Herberlain, France). Electrical parameters for gene transfer were used: 8 square-wave electric pulses of 5 ms duration at 400 V cm21 applied at a frequency of 1 Hz through stainless steel parallel electrodes directly on the petri dish in PB [Rosazza et al., 2011].
Alberts B, Johnson A, Lewis J. 2002. Molecular Biology of the Cell. New York: Garland Science.
AFM Measurements
Caraco C, Mozzillo N, Marone U, Simeone E, Benedetto L, Di Monta G, Di Cecilia ML, Botti G, Ascierto PA. 2013. Long-lasting response to electrochemotherapy in melanoma patients with cutaneous metastasis. BMC Cancer 13:564. Chiarella P, Fazio VM, Signori E. 2013. Electroporation in DNA vaccination protocols against cancer. Curr Drug Metab 14:291– 299.
An AFM Nanowizard 3 (JPK Instrument, Berlin, Germany) was used in the same conditions than previously stated [Chopinet et al., 2013b]. We used MLCT cantilevers (Bruker probes, USA, angle 17.5 ) with spring constant ranging from 0.028 to 0.042 N m21. QITM and force mapping settings used were the following: Z-length 5 lm; applied force 4 nN; speed: 166 lm s21 for QITM imaging on fixed cells, 1000 lm s21 for living cells; 24.98 lm s21 for force mapping. Before each experiment, sensitivity, and spring constant (thermal noise method) of cantilever were calibrated. Experiments were carried out with the same parameters then in Chopinet et al. [2013b] and the same procedure regarding data acquisition; either concerning elasticity recording or imaging.
Chopinet L, Formosa C, Rols MP, Duval RE, Dague E. 2013a. Imaging living cells surface and quantifying its properties at high resolution using AFM in QITM mode. Micron 48:26–33.
Data Analysis
Dufr^ene YF, Martınez-Martın D, Medalsy I, Alsteens D, M€uller DJ. 2013. Multiparametric imaging of biological systems by forcedistance curve-based AFM. Nat Methods 10:847–854.
JPKDataProcessing software was used for image processing. Images were flattened (order 1) and a 3D projection was made. The Hertz model gives the force F as a function of the indentation (d) and of the Young modulus (YM). The opening angle (a) of this sort of tip was 17.5 and we arbitrary choose a Poisson ratio (m) of 0.5. F 5 ((2.E.tana)/ (p.(1 2 m2)).d2. OpenFovea [Roduit et al., 2012; download at http://www.freesbi.ch/fr/openfovea] 0.1a152 software CYTOSKELETON
Chopinet L, Roduit C, Rols M-P, Dague. 2013b. Destabilization induced by electropermeabilization analyzed by atomic force microscopy. Biochim Biophys Acta BBA Biomembr 1828:2223– 2229. Coue M, Brenner SL, Spector I, Korn ED. 1987. Inhibition of actin polymerization by latrunculin A. FEBS Lett 213:316–318. Daud AI, DeConti RC, Andrews S, Urbas P, Riker AI, Sondak VK, Munster PN, Sullivan DM, Ugen KE, Messina JL, Heller R. 2008. Phase I trial of interleukin-12 plasmid electroporation in patients with metastatic melanoma. J Clin Oncol 26:5896–5903. Delteil C, Teissie J, Rols MP. 2000. Effect of serum on in vitro electrically mediated gene delivery and expression in mammalian cells. Biochim Biophys Acta 1467:362–368.
Escoffre J-M, Portet T, Wasungu L, Teissie J, Dean D, Rols M.-P. 2009. What is (still not) known of the mechanism by which electroporation mediates gene transfer and expression in cells and tissues. Mol Biotechnol 41:286–295. Fels J, Jeggle P, Kusche-Vihrog K, Oberleithner H. 2012. Cortical actin nanodynamics determines nitric oxide release in vascular endothelium. PloS One 7:e41520.
AFM Sensing Actin Cytoskeleton During Electropermeabilization
593 䊏
Gehl J. 2008. Electroporation for drug and gene delivery in the clinic: Doctors go electric. Methods Mol Biol Clifton NJ 423:351–359. Heller LC, Ingram SF, Lucas ML, Gilbert RA, Heller R. 2002. Effect of electrically mediated intratumor and intramuscular delivery of a plasmid encoding IFN alpha on visible B16 mouse melanomas. Technol Cancer Res Treat 1:205–209. Kanthou C, Kranjc S, Sersa G, Tozer G, Zupanic A, Cemazar M. 2006. The endothelial cytoskeleton as a target of electroporationbased therapies. Mol Cancer Ther 5:3145–3152. Lodish H, Berk A, Zipursky SL. 2000. The actin cytoskeleton. In: Molecular Cell Biology. New York: W. H. Freeman. Section 18.1. Available at: http://www.ncbi.nlm.nih.gov/books/NBK21493/. Lopez A, Rols MP, Teissie J. 1988. 31P NMR analysis of membrane phospholipid organization in viable, reversibly electropermeabilized Chinese hamster ovary cells. Biochemistry (Mosc.) 27:1222–1228. Low L, Mander A, McCann K, Dearnaley D, Tjelle T, Mathiesen I, Stevenson F, Ottensmeier CH. 2009. DNA vaccination with electroporation induces increased antibody responses in patients with prostate cancer. Hum Gene Ther 20:1269–1278. Mir LM. 2006. Bases and rationale of the electrochemotherapy. Eur J Cancer Suppl 4:38–44. Neumann E, Rosenheck K. 1972. Permeability changes induced by electric impulses in vesicular membranes. J Membr Biol 10:279–290.
䊏 594
Louise et al.
Roduit C, Saha B, Alonso-Sarduy L, Volterra A, Dietler G, Kasas S. 2012. OpenFovea: Open-source AFM data processing software. Nat Methods 9:774–775. Rols MP, Teissie J. 1992. Experimental evidence for the involvement of the cytoskeleton in mammalian cell electropermeabilization. Biochim Biophys Acta 1111:45–50. Rosazza C, Escoffre J-M, Zumbusch A, Rols M-P. 2011. The actin cytoskeleton has an active role in the electrotransfer of plasmid DNA in mammalian cells. Mol Ther 19:913–921. Rosazza C, Buntz A, Rieß T, W€oll D, Zumbusch A, Rols MP. 2013. Intracellular tracking of single plasmid DNA-particles after delivery by electroporation. Mol Ther 12:2217–2226. Rotsch C, Radmacher M. 2000. Drug-induced changes of cytoskeletal structure and mechanics in fibroblasts: An atomic force microscopy study. Biophys J 78:520–535. Teissie J, Rols MP. 1994. Manipulation of cell cytoskeleton affects the lifetime of cell membrane electropermeabilization. Ann N Y Acad Sci 720:98–110. Teissie J, Golzio M, Rols MP. 2005. Mechanisms of cell membrane electropermeabilization: A minireview of our present (lack of?) knowledge. Biochim Biophys Acta BBA—Gen Subj 1724: 270–280.
CYTOSKELETON
2014 Wiley Periodicals, Inc.
AFM Sensing Cortical Actin Cytoskeleton Destabilization During Plasma Membrane Electropermeabilization Chopinet Louise,1,2,3 Dague Etienne,1,3,4* and Rols Marie-Pierre2,3* 1
Centre National de la Recherche Scientifique, Laboratoire d’Analyse et d’Architecture des Syste`mes (LAAS), NanoBioSystem Group, Toulouse, F-31400, France 2 Centre National de la Recherche Scientifique, Institut de Pharmacologie et de Biologie Structurale (IPBS), Cell Biophysics Group, UMR 5089, BP64182, F-31077, Toulouse Cedex 4, France 3 Universite de Toulouse, UPS, INSA, INP, ISAE, UT1, UTM, LAAS, ITAV, F-31077, Toulouse Cedex 4, France 4 Centre National de la Recherche Scientifique, Institut des Technologies Avancees en Sciences du Vivant (ITAV), Nanocardiology and Cancer Cell Biophysics Group, USR3505, F31106, Toulouse, France
Received 17 April 2014; Revised 29 September 2014; Accepted 3 October 2014 Monitoring Editor: Manuel Thery
Electropermeabilization is a physical method that uses electric field pulses to deliver molecules into cells and tissues. Despite its increasing interest in clinics, little is known about plasma membrane destabilization process occurring during electropermeabilization. Having previously demonstrated the usefulness of Atomic Force Microscopy to study electropermeabilization effect on plasma membrane, we further investigated the plasma membrane destabilization process. We got mainly interested in the cytoskeleton role in stiffness of the plasma membrane, and thus in the effect of electric field on the cortical actin network. We show here that cortical actin is destabilized by electric pulses and that this effect is not directly related to the electropermeabilization of the plasma membrane. V 2014 Wiley Periodicals, Inc. C
Key Words:
electroporation; atomic force microscope; living cells; actin cytoskeleton; force spectroscopy; quantitative imaging
Additional Supporting Information may be found in the online version of this article. *Address correspondence to: Rols Marie-Pierre, Centre National de la Recherche Scientifique, Institut de Pharmacologie et de Biologie Structurale (IPBS), UMR 5089, BP64182, F-31077 Toulouse Cedex 4, France. E-mail: [email protected] correspondence to: Dague Etienne, Centre National de la Recherche Scientifique, Laboratoire d’Analyse et d’Architecture des Syste`mes (LAAS), NanoBioSystem Group, Toulouse, F-31400, France. E-mail: [email protected] Abbreviations used: AFM, atomic force microscopy; CHO, Chinese hamster ovary cell; EP, electropermeabilization or electroporation; LatB, Latrunculine B treatment; QITM, Quantitative ImagingTM; YM, Young modulus. Published online 11 October 2014 in Wiley Online Library (wileyonlinelibrary.com).
Introduction
E
lectropermeabilization (EP), or electroporation, is a non-viral vectorization technique based on electric field pulses application [Neumann and Rosenheck, 1972]. It is now used in clinics for cancer treatment. Electrochemotherapy is a standard treatment in many cancer centers around Europe that potentiates the delivery of cytotoxic drugs by local electric pulses delivery (http://guidance.nice.org.uk/ IPG446) [Mir, 2006; Caraco et al., 2013]. EP can also be used for gene transfer, and its most famous application resides in biotechnology for bacterial transformation. Moreover, it presents great promises in gene therapy and vaccination; clinical trials are undergoing [Heller et al., 2002; Daud et al., 2008; Gehl, 2008; Low et al., 2009; Chiarella et al., 2013]. Nevertheless, there is still a lack of knowledge about how electric fields are acting on cell components (mainly membrane) to permit entry of molecules and/or macromolecules (such as DNA) into cell cytoplasm [Teissie et al., 2005]. Most of the studies focusing on plasma membrane permeabilization description have been conducted using fluorescent dyes and only gave partial results based on indirect visualization and measurement of the phenomena [Escoffre et al., 2009]. Recently, we exploited Atomic Force Microscopy to investigate membrane permeabilization induced by EP at the nanoscale. We imaged and measured locally by force spectroscopy the consequences of EP on living cells plasma membrane elasticity [Chopinet et al., 2013b]. These previous results have shown that AFM allows visualizing membrane permeabilization with a different time scale than propidium iodide (PI) and other molecules entry. Indeed, we showed that the membrane elasticity decreases slower than the PI enter the cell after electric pulses application. This mean that even if membrane permeabilization is locally induced in specific regions of the cell and is transient (as seen by fluorescence imaging), its 587 䊏
consequences on stiffness are global and still present when membrane resealing has occurred. Moreover, this membrane destabilization has been observed independently to the cell orientation toward the electric field, in the contrary of molecules uptake (e.g., DNA) that takes place only at the face of the cell facing the cathode (which is due to electrophoresis processes and probably not to an orientated membrane destabilization). These results pointed out the ability of AFM to investigate electropermeabilization effects, and more specifically, we showed that AFM senses an impact of electric field different than plasma membrane permeabilization. Our results obtained both on fixed and living CHO cells (Chinese Hamster Ovary cells) give evidence of an inner effect affecting the entire cell surface that may be related to cytoskeleton destabilization. In 2000, Rotsch and Radmacher have shown that decrease of young modulus can be related specifically to the actin cytoskeleton destabilization, and not microtubule one [Rotsch and Radmacher, 2000]. The actin cytoskeleton network is composed of globular monomer of actin polymerized in filaments. These filaments are interacting with plasma membrane directly to lipid or through transmembrane proteins [Lodish et al., 2000]. Two filaments types are described: microfilaments that are composed of several filaments interconnected thanks to linking proteins, and cortical actin network which is located underneath plasma membrane in a sustaining dense network composed of solely filaments [Alberts et al., 2002]. The cortical actin network has been shown to be detected in the first 100 nm of indentation [Fels et al., 2012]. Very few articles get interested in EP effects on actin. In 1992, Rols and Teissie have shown that cytoskeleton was involved in the resealing process [Rols and Teissie, 1992]. They showed that CHO cells with depolymerized actin are resealing slower than control cells. In 1994, same authors gave further description for the implication of cytoskeleton in permeabilization phenomenon [Teissie and Rols, 1994]. Other studies performed on the cytoskeleton of endothelial cells have shown a profound disruption of microfilament and microtubule cytoskeletal networks after EP, paralleled by a rapid increase in endothelial monolayer permeability [Kanthou et al., 2006]. These results showing the remodeling of the endothelial cytoskeleton and changes in endothelial barrier function provide an insight into putative mechanisms responsible for the observed increase in permeability and cessation of blood flow in vivo. More recently, actin has been shown to play an important role in plasmid internalization and intracellular traffic during electrotransfer, as actin is polymerizing under DNA spots at the membrane [Rosazza et al., 2011, 2013]. There are thus evidences for the implication of actin in plasma membrane electropermeabilization but without any direct correlation. AFM allows physical measurement of actin destabilization on living cell along time and can thus enhance the mechanical characterization and understanding
䊏 588
Louise et al.
of permeabilization process. This is why in the present article we chose to use this instrument to investigate the implication of cortical actin in the measured increase in elasticity after EP, and performed experiments on cells treated with the drug Latrunculine B which affects actin polymerization (LatB treatment) [Coue et al., 1987; Rosazza et al., 2011]. Imaging and measuring plasma membrane have been performed following the same protocol as in Chopinet et al. [2013b]. Force volume (FV) acquisition has been realized on 4 areas on each cell. Images have been acquired thanks to the multiparametric mode “Quantitative Imaging” from JPK [Chopinet et al., 2013a; Dufr^ene et al., 2013]. We show that AFM technology is of great interest to resolve biological questions such as deciphering between two treatments effect namely electropermeabilization and drug treatment by Latrunculine B. Especially, it provides results about inner component of the cell, namely cortical actin. We demonstrate in this article that electropermeabilization is affecting cortical actin network interaction with plasma membrane through the lipid bilayer destabilization by electric field thanks to the unique use of AFM.
Results Drug efficiency
The efficiency of the treatment by LatB has been assessed by confocal microscopy thanks to cells transfected for actinGFP protein 24 h prior the treatment (Fig. 1) [Rosazza et al., 2011]. The labeled actin is uniformly distributed in control cells, and aggregated in treated cells. Latrunculin effects are directly visible on cells by phase contrast as shown by changes in the shape of the cells (Figs. 1A0 and 1B0 ). Thus during AFM experiments, drug efficiency has been checked only by phase contrast, avoiding the use of actin-GFP transfected cells that may create bias related to this procedure and the production of the GFP coupled actin protein (Fig. 1C). Correlation Between EP and Latrunculine Effect
To validate that the decrease in elasticity observed in our previous work is related to the destabilization of the cortical actin network [Chopinet et al., 2013b], we compared elasticity in electropermeabilized cells and in cells treated with Latraculine B (LatB). The same experimental procedure as in Chopinet et al. [2013b] was used. The analysis of cell elasticity in the 50 first nanometers of indentation before and after LatB treatment shows a decrease in elasticity of around 30 kPa (Fig. 2A, Supporting Information Table I). Measurement of cells treated and placed in pulsation buffer at t 5 0 min show a recovery of elasticity along time (Fig. 2B). The difference between elasticity at time 25 min (when actin is completely depolymerized) and 33 min (when actin has repolymerized) is the same as the ones observed before and after electric field application (around CYTOSKELETON
Fig. 1. LatB effect on actin polymerization. (A) Confocal imaging of control untreated cells expressing a GFP labeled actin protein, fluorescence image (A) and merge with phase contrast image (A’). (B) Confocal imaging of cells treated with 0.1 mM Latrunculine B during 1 h at 37 C, fluorescence (B) and merge with contrast phase (B’). (C) A cell recovery after removing the drug followed by phase contrast imaging during an AFM experiment.
10 kPa, [Chopinet et al., 2013b]). However if electric fields are applied on cells treated with LatB (always at time 5 0 when the drug is removed and the cells are placed in pulsation buffer), cells are not recovering their initial elasticity (Fig. 2C and Supporting Information Table I). Imaging with QI mode reveals the de novo formation of actin filaments in cells when the drug is removed. Height images show the elongation of cells and production of actin filament in the longitudinal way (Fig. 3A). However on cell submitted to electric fields, even if the cell still seems to spread along time, it is not possible to distinguish filaments underneath plasma membrane (Fig. 4A). The force curve analysis performed with OpenFovea software [Roduit et al., 2012] has been realized on several segments of 50 nm of indentation every 50 nm along the force curve following the point of contact between the tip and the sample. Actin filaments are visible on segments located at 100 nm depth as they are localized underneath the membrane. The elasticity maps are showing these filaments that are stiffer than the others cell components (Fig. 3B). Elasticity maps are showing the same differences that the ones observed on height images: LatB-treated cell submitted to electric field are unable to produce actin filament de novo. Cells are able to do so along time only when no electric fields are applied. Cortical Actin or Microfilament?
To go deeper in the interpretation of the phenomenon observed on cortical actin, we performed experiment combining electric field application and drug treatment. As presented in Fig. 5, we first applied electric field, acquire CYTOSKELETON
elasticity measurement, then added the drug to the media, or, we first added the drug, and then applied the electric field. Results show that when EP is applied before the drug is added, a supplementary decrease in elasticity is measured (Fig. 5A, P value < 0.05). In the contrary, when the drug is added before EP, there is no such decrease observed in elasticity (Fig. 5B). These results show that even if EP and LatB treatment result in a decrease in elasticity, they do not act similarly on it. LatB treatment has a stronger impact on elasticity than EP, and thus provokes a supplementary decrease in elasticity when the drug is added after pulse electric field application.
Discussion Understanding the mechanism governing molecules electrotransfer in cells by electric pulses is a prerequisite for the development of more efficient clinical protocols based on this approach. This concerns both the use of the method as an efficient way to deliver anticancer drugs as well as its use for DNA vaccination or gene therapy. Therefore, future clinical use requires a much better knowledge of the processes involved at the cellular level. Our interest was focused on unraveling the role of the actin cytoskeleton in stiffness of the plasma membrane and in the effect of electric pulses on cortical actin network. We previously observed, on cells submitted to EP, a decrease in elasticity of the 50 first nm of indentation followed by a gradual recovery along the 35 min following electric field
AFM Sensing Actin Cytoskeleton During Electropermeabilization
589 䊏
Fig. 2. Membrane elasticity along time following LatB treatment. (A) Cell treated, the drug is added at time 0 and corresponding force curves (n 5 3 cells). Before treatment t 5 27 min: YM 5 15.47 kPa 6 1.8 (SE) (1024 force curves); after treatment t 5 33 min: YM 5 10.65 kPa 6 0.66 (SE) (1024 force curves). (B) Recovery of elasticity after drug removal at time 0 (n 5 3 cells). Before drug removal t 5 25 min: YM 5 11.33 kPa 6 0.28 (SE) (1024 force curves); after drug removal t 5 7 min: YM 5 12.59 kPa 6 0.64 (SE) (1024 force curves); after drug removal t 5 33 min: YM 5 17.33 kPa 6 1.66 (SE) (1024 force curves). (C) No recovery in elasticity after drug removal and electric field application at time 0 and corresponding force curves (n 5 3 cells). Before drug removal and electric field application t 5 25 min: YM 5 8.62 kPa 6 0.69 (SE) (1024 force curves); after drug removal and electric field application t 5 32 min: YM 5 9.22 kPa 6 0.57 (SE) (1024 force curves). Unpaired t test relative to the point before t 5 0: *P value < 0.05, **P value < 0.005, ***P value < 0.0001. Detailed data in Supporting Information.
application. Cells treated with LatB are displaying the same decrease in elasticity and the same recovery (Fig. 2). This similitude between EP and drug treatment reinforces the idea that pulsed electric field have an effect on actin cytoskeleton. QI height images allowed going further in the investigation. On the one hand, cells submitted only to EP do not display actin microfilament [Chopinet et al., 2013b]. This event occurs together with cell swelling, in absence of any change in cell shape which can indicate that cytoplasmic
䊏 590
Louise et al.
actin (actin microfilament, in the contrary of cortical actin which is located at the membrane) is not destabilized by electric pulses. Indeed, as visible on cells treated with LatB, actin depolymerization is affecting cell shape (Figs. 3 and 4). However after EP, the plasma membrane shows ripples that can point out an impact on cortical actin [Chopinet et al., 2013b]. On the other hand, images of cells treated with LatB show de novo formation of stress fibers after drug removal. When electric field pulses are applied simultaneously with drug removal, no microfilament are imaged, but cells spread gradually along time, leading to the hypothesis that cytoplasmic actin (microfilament linked to focal adhesion points) can be able to polymerize, while cortical actin is affected by electric pulses. On the whole, these observations lead to consider that electric field pulses are affecting cortical actin and not actin filaments. Indeed, as a first observation, cells are swelling which can indicate an effect on membrane/cytoskeleton interaction. Second, the impossibility to image microfilament after EP leads to the interpretation that what is interacting with plasma membrane is destabilized (directly by electric field, or indirectly by the plasma membrane electrodestabilization itself ). This event is first decreasing elasticity in the first 50 nm of indentation, and second is avoiding the imaging of inner microfilaments by releasing membrane from its anchoring. This interpretation can be supported by previous work using the same electric parameters and the same cells, that showed unaffected actin microfilaments [Rosazza et al., 2011]. Moreover, it has recently been shown that the 100 firsts nm of indentation display elasticity linked to cortical actin network [Fels et al., 2012]. Thus, altogether, these observations intend to show that the measured elasticity after EP is linked to cortical actin cytoskeleton destabilization. At this point, two hypotheses can be formulated: (1) electric pulses are affecting directly cortical actin filaments and destabilizing their structure leading to depolymerization or rupture. This effect could be involved in the permeable state of membrane. It is known that one EP side effect is the release of ATP after membrane permeabilization [Delteil et al., 2000] and can explain an effect on actin polymerization which is ATP-dependent. (2) Effect of electric pulses on plasma membrane, i.e. lipid bilayer destabilization leading to permeabilization [Lopez et al., 1988], can provoke a rupture of the interaction between cortical actin network and plasma membrane, which can release the membrane and cancel any resistance at the membrane level leading to a decrease in elasticity. The last experiment consisting in combining EP and drug exposure helps to go further. As shown in Fig. 4A, the Latrunculin drug provokes a supplementary effect when added after the electric pulses application, while applying electric pulses on cells already exposed to the drug does not have any supplementary effect (Fig. 5B). These results show CYTOSKELETON
Fig. 3. Recovery of actin cytoskeleton after 1-h treatment with LatrunculinB 0.1 mM at 37 C. t0 correspond to the change from media containing LatrunculinB to media free of any drug. Cells have been kept at 37 C in HEPES media during all the QITM measurements (75 3 75 pixels). (A) Young modulus map at 100 nm of depth along time in minutes, colorscale: from 0 (dark blue) to 60 kPa (dark red), one segment equals 8 kPa. (B) Corresponding height images, scale bar: 5 mm, color scale: 0–4 mm. Extra cells visible in Supporting Information.
that the drug has a more dramatic effect than EP does, and thus encourages the hypothesis (1). Indeed, as Latrunculin is known to destabilize actin filaments, it means that it will affect every filament, and provokes the decrease in elasticity by a global effect. However, EP has a more spared effect that can thus be considered as the destabilization of cortical actin network interaction with plasma membrane rather than the actin depolymerization itself. To address the remaining question about the fact that EP and LatB treatment have the same effect in term of elasticity, it is necessary to consider that cortical actin depolymerization and disruption of actin-membrane interaction will lead to the same result: a decrease in membrane elasticity in the 50 first nm, even if the mechanisms responsible for this decrease are different. Indeed, the height images of cells confirm these considerations. LatB treated cells still spread
on the substrate (meaning that actin filaments are polymerizing) after EP application but no actin filaments are imaged or measured underneath the membrane. This observation is consistent with the fact that EP may affect the actin-membrane interaction rather than the actin polymerization itself. This effect avoids imaging actin filament after EP when membrane is totally weak and floppy. Moreover, we have shown here that elasticity is not recovered when EP is applied after drug removal, which also implies that the plasma membrane remains floppy even if the cells are able to repolymerize their actin filaments, once again confirming that only the interaction between actin and membrane is not recovered. The two effects (of EP and of LatB) are clearly different as it can be seen on Fig. 5A where the addition of the two treatments creates two different steps in elasticity decrease.
Fig. 4. Absence of recovery of actin cytoskeleton after 1-h treatment with LatrunculinB 0.1 mM and pulse application at t0. T0 correspond to the change from media containing LatrunculinB to media free of any drug combined to pulse application. Cells have been kept at 37 C in HEPES media during all the QITM measurements (75 3 75 pixels). (A) Young modulus map at 100 nm of depth along time in minutes, colorscale: from 0 (dark blue) to 60 kPa (dark red), one segment equals 8 kPa. (B) Corresponding height images, scale bar: 5 mm, color scale: 0–4 mm. Extra cells visible in Supporting Information.
CYTOSKELETON
AFM Sensing Actin Cytoskeleton During Electropermeabilization
591 䊏
Fig. 5. Effect of combination of the two treatment. (A) Supplementary effect on elasticity when the drug is added at time 10 min after pulse application at time 0 and corresponding force curves (n 5 3 cells). Before electric field application t 5 210 min: YM 5 14.83 kPa 6 1.3 (SE); after electric field application t 5 5 min: YM 5 10.85 kPa 6 1.62 (SE) (1024 force curves); after drug treatment t 5 15min: YM 5 8.28 kPa 6 0.32 (SE) (1024 force curves). (B) No combined effect when pulses are applied at time 12 after drug is added at time 0 and corresponding force curves (n 5 3 cells). Before drug treatment t 5 26 min: YM 5 14.67 kPa 6 0.97 (SE) (1024 force curves); after drug treatment t 5 6 min: YM 5 9.66 kPa 6 0.5 (SE) (1024 force curves); after electric field application t 5 20 min: YM 5 10.84 kPa 6 1.02 (SE) (1024 force curves. Unpaired t test relative to the point before t 5 0: *P value < 0.05, **P value < 0.005, ***P value < 0.0001. Detailed data in Supporting Information.
It thus confirms that EP has a different role than Latrunculin in actin destabilization. Moreover, it is known from the literature [Rols and Teissie, 1992] that Cytochalasin treated cells, as well as Table I.
Latrunculin treated cells stay permeable to propidium iodide. This information makes a direct link between the EP effect on membrane, namely its permeabilization, and the role of actin in plasma membrane properties. As actin depolymerization is responsible for the persistent permeabilization state of plasma membrane, and as we shown that LatB-treated cells exposed to EP do not recover their elasticity, we can thus conclude that EP is only affecting actinmembrane interaction through the destabilization of lipid bilayer. To finish, we can say that EP never affects cell spreading; meaning EP does not provoke actin depolymerization. Table I provided below helps organizing the effects observed, and thus supports the hypothesis that EP is affecting cortical actin through the destabilization of lipid bilayer and disruption of actin/membrane interactions. To conclude, we show in this article that AFM provides a new approach to study biophysical mechanisms, and specifically the analysis of electropermeabilization effect on internal component like cytoskeleton. We demonstrated the ability of AFM to discriminate between two treatments effects and thus to investigate, at the molecular level, the effects of electric fields on membrane permeability properties.
Materials and Methods Sample Preparation
Nearly 75,000 CHO cells were grown in Petri dish during 24 h in the same cell culture conditions described in Chopinet et al. [2013b]. Before measurements, classical MEM medium was replaced by MEM-HEPES medium (CM1MEM46-6U, Eurobio, France) supplemented with 8% fetal calf serum (Lonza Group, Switzerland) and cells were placed in the PetriDishHeater (JPK) that maintained
Sum up of the Main Results Presented in the Article and in Chopinet et al 2013b Elasticity
Recovery
Filament
Spreading
-
1
-
5
Lipid destabilization
1LatB -LatB 1LatB1EP
/ 1
1 -
1 -
1EP1LatB
–
-
-
Actin depolymerization Actin repolymerization Actin depolymerization 1 lipid destabilzation Lipid destabilization 1 actin depolymerization
-LatB1EP
-
-
-
1EP
䊏 592
Louise et al.
Known effect
Actin depolymerization 1 lipid destabilzation
Assumed effect
Disruption of actin-membrane interaction
Not possible to reform the interaction 1. Actin-membrane interaction is disrupted; 2. Actin is depolymerized Actin is already depolmerized, thus no disruption. However: no new interaction with membrane because no actin available, which means that elasticity remains low.
CYTOSKELETON
30 C during all the experiment to slow down the membrane activity and process to monitor. Prior to electrical pulse application, cells were washed three times with phosphate-buffered saline (PBS) 13 (Invitrogen, USA). About 1 mL of pulsation buffer (PB) at 4 C (10 mM K2HPO-4/KH2PO42buffer, 1 mM MgCl2, 250 mM sucrose, pH 7.4) was added and electrodes were placed in contact with Petri dish bottom. Pulsed were delivered and AFM head was replaced.
was used for force curve analysis with the following settings on 50 nm of indentation: Model: cone; Tip size: 0.62 rad; Poisson ratio: 0.5; Method: raw. Frequency distributions were obtained with OriginPro 8 software, and histograms were plotted with GraphPad Prism software, that was used for statistical analysis too (unpaired t-test, *P value < 0.05, **P value < 0.005, ***P value < 0.0001), results are presented as boxplots with min and max value, and median. See Chopinet et al. [2013b] for further details.
Latrunculine Treatment
Acknowledgments
A 0.1 mM of Latrunculine B (Sigma–Aldrich, L5288) was added to the cells 1 h before experiments or after pulses application (mentioned in the text when drug is added before or after EP to allow measurement of the effects of both treatment). The maximum volume of DMSO in which the drug is solubilized was 2& of the total volume of media. The drug is removed by taking out the medium of the petri dish, washing twice with PBS, and then adding a drug free media (pulsation buffer or HEPES media).
The authors thank Marie-Charline Blatche for helping in cell sample preparation, Direction Generale de l’Armement for 3 years PhD grant to Louise Chopinet, ANR Young Scientist Program “AFMyst” (#30024332) and ANR Astrid “PIERGEN” (#ANR2122ASTR20039201) that provided financial support. This report is the result of networking efforts of COST Action TD1104 (http://www. electroporation.net) and was conducted in the scope of the EBAM European Associated Laboratory.
Electropermeabilization
References
The generator used for this study was a Jouan (Herberlain, France). Electrical parameters for gene transfer were used: 8 square-wave electric pulses of 5 ms duration at 400 V cm21 applied at a frequency of 1 Hz through stainless steel parallel electrodes directly on the petri dish in PB [Rosazza et al., 2011].
Alberts B, Johnson A, Lewis J. 2002. Molecular Biology of the Cell. New York: Garland Science.
AFM Measurements
Caraco C, Mozzillo N, Marone U, Simeone E, Benedetto L, Di Monta G, Di Cecilia ML, Botti G, Ascierto PA. 2013. Long-lasting response to electrochemotherapy in melanoma patients with cutaneous metastasis. BMC Cancer 13:564. Chiarella P, Fazio VM, Signori E. 2013. Electroporation in DNA vaccination protocols against cancer. Curr Drug Metab 14:291– 299.
An AFM Nanowizard 3 (JPK Instrument, Berlin, Germany) was used in the same conditions than previously stated [Chopinet et al., 2013b]. We used MLCT cantilevers (Bruker probes, USA, angle 17.5 ) with spring constant ranging from 0.028 to 0.042 N m21. QITM and force mapping settings used were the following: Z-length 5 lm; applied force 4 nN; speed: 166 lm s21 for QITM imaging on fixed cells, 1000 lm s21 for living cells; 24.98 lm s21 for force mapping. Before each experiment, sensitivity, and spring constant (thermal noise method) of cantilever were calibrated. Experiments were carried out with the same parameters then in Chopinet et al. [2013b] and the same procedure regarding data acquisition; either concerning elasticity recording or imaging.
Chopinet L, Formosa C, Rols MP, Duval RE, Dague E. 2013a. Imaging living cells surface and quantifying its properties at high resolution using AFM in QITM mode. Micron 48:26–33.
Data Analysis
Dufr^ene YF, Martınez-Martın D, Medalsy I, Alsteens D, M€uller DJ. 2013. Multiparametric imaging of biological systems by forcedistance curve-based AFM. Nat Methods 10:847–854.
JPKDataProcessing software was used for image processing. Images were flattened (order 1) and a 3D projection was made. The Hertz model gives the force F as a function of the indentation (d) and of the Young modulus (YM). The opening angle (a) of this sort of tip was 17.5 and we arbitrary choose a Poisson ratio (m) of 0.5. F 5 ((2.E.tana)/ (p.(1 2 m2)).d2. OpenFovea [Roduit et al., 2012; download at http://www.freesbi.ch/fr/openfovea] 0.1a152 software CYTOSKELETON
Chopinet L, Roduit C, Rols M-P, Dague. 2013b. Destabilization induced by electropermeabilization analyzed by atomic force microscopy. Biochim Biophys Acta BBA Biomembr 1828:2223– 2229. Coue M, Brenner SL, Spector I, Korn ED. 1987. Inhibition of actin polymerization by latrunculin A. FEBS Lett 213:316–318. Daud AI, DeConti RC, Andrews S, Urbas P, Riker AI, Sondak VK, Munster PN, Sullivan DM, Ugen KE, Messina JL, Heller R. 2008. Phase I trial of interleukin-12 plasmid electroporation in patients with metastatic melanoma. J Clin Oncol 26:5896–5903. Delteil C, Teissie J, Rols MP. 2000. Effect of serum on in vitro electrically mediated gene delivery and expression in mammalian cells. Biochim Biophys Acta 1467:362–368.
Escoffre J-M, Portet T, Wasungu L, Teissie J, Dean D, Rols M.-P. 2009. What is (still not) known of the mechanism by which electroporation mediates gene transfer and expression in cells and tissues. Mol Biotechnol 41:286–295. Fels J, Jeggle P, Kusche-Vihrog K, Oberleithner H. 2012. Cortical actin nanodynamics determines nitric oxide release in vascular endothelium. PloS One 7:e41520.
AFM Sensing Actin Cytoskeleton During Electropermeabilization
593 䊏
Gehl J. 2008. Electroporation for drug and gene delivery in the clinic: Doctors go electric. Methods Mol Biol Clifton NJ 423:351–359. Heller LC, Ingram SF, Lucas ML, Gilbert RA, Heller R. 2002. Effect of electrically mediated intratumor and intramuscular delivery of a plasmid encoding IFN alpha on visible B16 mouse melanomas. Technol Cancer Res Treat 1:205–209. Kanthou C, Kranjc S, Sersa G, Tozer G, Zupanic A, Cemazar M. 2006. The endothelial cytoskeleton as a target of electroporationbased therapies. Mol Cancer Ther 5:3145–3152. Lodish H, Berk A, Zipursky SL. 2000. The actin cytoskeleton. In: Molecular Cell Biology. New York: W. H. Freeman. Section 18.1. Available at: http://www.ncbi.nlm.nih.gov/books/NBK21493/. Lopez A, Rols MP, Teissie J. 1988. 31P NMR analysis of membrane phospholipid organization in viable, reversibly electropermeabilized Chinese hamster ovary cells. Biochemistry (Mosc.) 27:1222–1228. Low L, Mander A, McCann K, Dearnaley D, Tjelle T, Mathiesen I, Stevenson F, Ottensmeier CH. 2009. DNA vaccination with electroporation induces increased antibody responses in patients with prostate cancer. Hum Gene Ther 20:1269–1278. Mir LM. 2006. Bases and rationale of the electrochemotherapy. Eur J Cancer Suppl 4:38–44. Neumann E, Rosenheck K. 1972. Permeability changes induced by electric impulses in vesicular membranes. J Membr Biol 10:279–290.
䊏 594
Louise et al.
Roduit C, Saha B, Alonso-Sarduy L, Volterra A, Dietler G, Kasas S. 2012. OpenFovea: Open-source AFM data processing software. Nat Methods 9:774–775. Rols MP, Teissie J. 1992. Experimental evidence for the involvement of the cytoskeleton in mammalian cell electropermeabilization. Biochim Biophys Acta 1111:45–50. Rosazza C, Escoffre J-M, Zumbusch A, Rols M-P. 2011. The actin cytoskeleton has an active role in the electrotransfer of plasmid DNA in mammalian cells. Mol Ther 19:913–921. Rosazza C, Buntz A, Rieß T, W€oll D, Zumbusch A, Rols MP. 2013. Intracellular tracking of single plasmid DNA-particles after delivery by electroporation. Mol Ther 12:2217–2226. Rotsch C, Radmacher M. 2000. Drug-induced changes of cytoskeletal structure and mechanics in fibroblasts: An atomic force microscopy study. Biophys J 78:520–535. Teissie J, Rols MP. 1994. Manipulation of cell cytoskeleton affects the lifetime of cell membrane electropermeabilization. Ann N Y Acad Sci 720:98–110. Teissie J, Golzio M, Rols MP. 2005. Mechanisms of cell membrane electropermeabilization: A minireview of our present (lack of?) knowledge. Biochim Biophys Acta BBA—Gen Subj 1724: 270–280.
CYTOSKELETON
