Solutions and Reagents

All reagents were purchased from Sigma-Aldrich (Steinheim, Germany) if not mentioned otherwise. During experiments the cells were continuously perfused with HEPES buffered solutions (in mM: 140 NaCl, 5 KCl, 1 MgCl₂, 1 CaCl₂, 5 glucose and 1 L-arginine adjusted to pH 7.4 at RT). CD buffer was supplemented with 50 nM cytochalasin D (1 mM stock in 100% ethanol). High extracellular K⁺ was achieved by isoosmotically increasing KCl from 5 to 12 mM. In some experiments 140 mM NaCl was replaced by sodium gluconate (140 mM) to reduce chloride concentration. All control buffers were supplemented with the respective amount of solvent or were isoosmotically adjusted with mannitol if required.

Cell Culture

Bovine aortic endothelial cells (GM7373) [21] were cultured as described previously [4]. In brief, cells were cultured in minimal essential medium supplemented with 1% non-essential amino acids, 1% MEM vitamins (Invitrogen Corp., Karlsruhe, Germany) and 20% fetal bovine serum (FCS, PAA Clone, Coelbe, Germany). T₂₅ culture flasks were used to culture the cells in an incubator at 37°C, 5% CO₂ and 100% humidity.

Lifeact-eGFP Transfection

GM7373 cells were transfected with Lifeact-eGFP to trace F-actin dynamics in living cells. Lifeact consists of a 17 amino acids long actin binding motif, derived from the saccharomyces cerevisiae protein Abp140 [22]. The peptide sequence has been ligated to the sequence of the enhanced green fluorescent protein eGFP, which allows tracing actin dynamics in living cells without interference of other signalling pathways or modification of actin expression and polymerization [19]. Cells used for measurements of Lifeact-eGFP fluorescence were cultured as described above. Three days prior to experiments, the cells were trypsinized, centrifuged and resuspended in OptiMEM medium (Invitrogen) supplemented with 10% FCS. The transfection reagent, using FuGENE6 transfection reagent (Roche, Mannheim, Germany) was prepared as described in the manufacturers protocol. In brief, 230 µl FCS-free OptiMEM Medium was mixed with 10 µl FuGENE6 Transfection Reagent and incubated for 10 minutes at room temperature (RT). Then, 5 µg Lifeact-eGFP DNA (in pEGFP-N1 vector, kindly provided by Dr. Roland Wedlich-Soldner, Max Planck Institute of Biochemistry, Martinsried, Germany) was added and the solution was gently mixed. After a second period of incubation (10 minutes at RT), the cells were added, mixed gently and then seeded on Ø 40 mm glass bottom dishes (WillCo Wells B.V., Amsterdam, Netherlands). To improve the transfection efficiency, the cells were subsequently incubated on a shaker (180 rpm) for another 40 minutes in the cell culture incubator. At the day of experiment, the cell monolayer was 100% confluent. To verify the specificity of Lifeact to label F-actin, GM7373 cells were stained with phalloidin-tetramethylrhodamine B isothiocyanate (phalloidin-TRITC). For this, transfected Lifeact-eGFP cells were fixed with 4% formaldehyde in phosphate buffered solution (PBS) for 15 minutes at RT. Subsequently, the cells were washed with PBS and incubated (30 min, RT) in 10% goat-normal-serum supplemented with phalloidin-TRITC (final 5 units/ml) and 4′,6-diamidino-2-phenylindole (DAPI, Invitrogen, final 0.1 µg/ml). Figure 1 shows a fluorescence image of a phalloidin-stained Lifeact-eGFP-expressing endothelial cell. The overlay of the Lifeact-eGFP and phalloidin-TRITC image verifies that both dyes match and approves Lifeact as a tool to visualize F-actin in endothelial cells.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Lifeact verification.

Double staining of F-actin in GM7373 bovine aortic endothelial cells using Lifeact-eGFP and TRITC-phalloidin. DAPI is used to stain nuclei. The merge image verifies that Lifeact-eGFP and TRITC-phalloidin are equally applicable to detect F-actin in endothelial cells.

https://doi.org/10.1371/journal.pone.0041520.g001

Simultaneous Measurements of Cortical Stiffness and Fluorescence Intensity

AFM-based analysis of living cells allows the assessment of cell topography [23] as well as the analysis of protein interaction and stoichiometry [24], [25], [26]. Additionally, the AFM is the tool of choice to analyse regional stiffness of living cells [27]. A hybrid setup consisting of an AFM integrated into an epifluorescence microscope allows the simultaneous measurements of К_cortex and the actin dynamics of the cell cortex (Figure 2a). The Bioscope Catalyst AFM (Bruker, Karlsruhe, Germany) was used to perform К_cortex and К_bulk measurements ( = force tomography) as described previously [28], [29]. In brief, AFM-cantilevers (MLCT contact microlever, Bruker, spring constant approx. 0.01 pN/nm), equipped with 10 Ø µm colloidal tips (Novascan, Ames, IA, USA, for high resolution SEM image see [5]), were used to periodically indent the cells. A laser beam reflected from the backside of the cantilever is detected by a photo diode and allowing measurements of cantilever deflection. The large size of the spherical cantilever tip allowed measurements at very low forces (range: 200 pN) and thereby does not disturb cell function. By the determination of the spring constant as well as by measuring the deflection sensitivity of the cantilever, the force acting on the cantilever (an in turn the force exerted by the cantilever to the sample) can be calculated:

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. Stiffness measurements.

a) Combined setup of an atomic force microscope integrated into an inverse epifluorescence microscope (mod. after [20]). b) Two force-indentation curves acquired on a single endothelial cell before (black) and after (red) cortical softening, induced by 50 nM cytochalasin D. Cells with a soft cortex may exhibit an indifferent zone of nonlinear stiffness in the area of the cortex-bulk-transition. c) Zoom-in of the approach and cortical indentation part of the force-indentation curves. Both curves are linear (no force needed) when the colloidal tip is approaching the cell. Indentation forces differ in soft-cortex and stiff-cortex cells. d) Force-indentation curves corresponding to the bulk phase of the cell. The curves show that bulk stiffness is unaffected by changes in the nanomechanics of the cell cortex.

https://doi.org/10.1371/journal.pone.0041520.g002

Additionally, by taking the piezo displacement into account, the indentation (deformation) of the sample can be determined:

Subsequently, force-indentation-curves (F/I-curve), generated by plotting the force against the indentation, can be used to analyse the stiffness of different cell regions (cell depths). The stiffness (К) is an indicator of the mechanical resistance of a sample against a defined deformation (i.e. indentation). If the slope of the F/I-curve correlates to a linear fit, К can be estimated by:

Figure 2b displays two typical F/I-curves, the first one taken from a cell with a stiff cortex, the second one with a soft cortex. During the vertical approach, the cantilever does not bend since no force acts on the cantilever. Therefore, the stiffness of the cell (slope of both curves) tends to be ∼0 pN/nm (Figure 2b). As soon as the cantilever gets into contact to the cell surface and starts to indent the cell, force is exerted on the cantilever which results in cantilever bending. Hence, the slope of the F/I-curves changes to positive values. Interestingly, a typical F/I-curve shows a linear slope within the first 50 to 100 nm. This implicates the linear fit model to be appropriate for stiffness analysis (Figure 2c). A further indentation induces a shift in the F/I-curve, which shows again an almost linear behaviour. The second linear slope of the F/I-curves corresponds to the stiffness of the bulk phase of the cell (Figure 2d). It should be mentioned that the transition from the cortex to the bulk phase is not in any event characterized by a distinct shift of the slope of F/I-curves. As shown in Figure 2b, cells exhibiting a very soft cell cortex may show non-linear stiffness behaviour in the region of the cortex-bulk-transition. However, the cortex itself as well as the bulk region generally shows a F/I-curve with linear characteristics (R²≥0.96). Hence, to detect cortical stiffness, the first linear 50 nm of indentation were analysed whereas bulk stiffness was analysed in the linear region of 350–500 nm of indentation.

To verify that the first event of indentation definitely corresponds to the cell cortex (and not to e.g. membrane ruffles, glycocalyx or microvilli), we performed high resolution imaging of the cell surface. As shown in figure S1a, confluent GM7373 cells were fixed with 0.5% glutaraldehyde and imaged with MLCT cantilever in fluid (Bruker, spring constant 0.01 pN/nm). The magnified regions displayed in suppl. figure S1b show that the surface of endothelial cells is rather smooth and holds no deep ruffles or microvilli. Instead, the submembraneous actin filaments (arrowheads) as well as tiny invaginations (arrows), possibly corresponding to caveolae, can be imaged. Even an initial detection of the endothelial glycocalyx can be excluded since much lower forces and ramp parameters are necessary to detect the endothelial surface layer. Hence it is evident that the detected first slope of the F/I-curve matches to the cell cortex. To prevent mechanical alteration of the cell, a low indentation velocity of the cantilever (max. 1 µm/second) was chosen for experiments. Force-distance curves were acquired using the Nanoscope v7.30 software (Bruker) and data analysis was performed with Protein Unfolding and Nano-Indentation Software (PUNIAS, http://punias.voila.net/).

Simultaneously to the AFM measurements, the integrated DMI 6000 B inverted epifluorescence microscope (Leica, Microsystems, Wetzlar, Germany) was used for detection of Lifeact-eGFP fluorescence intensity. Epifluorescence is a eligible method since other fluorescence techniques do have lower time resolutions (e.g. confocal microscopy) or do not allow an illumination of the apical surface during constant perfusion (e.g. TIRF). The fluorophore was excited at 480/40 nm and emission was detected at 527/30 nm (Figure 2a). To obtain a high resolution, a 100× oil immersion objective was used for acquisition with an exposure time of 200 ms. Several regions of interest were selected in the peripheral regions of the transfected cells. Background fluorescence was subtracted as appropriate. The frequency of image (and force-distance curve) acquisition was set to 20 seconds in order to prevent photo bleaching. Fluorescence imaging and data analysis was performed using LAS AS 1.7.0 software (Leica). All experiments were performed at room temperature. Stiffness and fluorescence intensities were recorded for several minutes in control buffer allowing the system to equilibrate. All detected parameters 10 minutes prior to the change of the perfusion buffer (i.e. control conditions) where averaged and set as 100%. During experiments, the cells were perfused using a home-built perfusion chamber to guarantee a continuous and noise-free exchange of the respective perfusion buffer and a physiological shear stress [20].

AFM Imaging and Profile Analysis

The AFM is an ideal method to visualize cell surface, height and volume [30], [23]. To generate 3D topography maps, subconfluent GM7373 (∼90%) were scanned with cantilever (MSCT, Bruker). The scan size was set to 100 µm × 100 µm (256×256 pixels) with a scan rate of 0.375 Hz. These settings guarantee an almost noise-free measurement of cell height without disturbance of the cell topography by the indentation force [31].

Analysis of eNOS Activity

eNOS activity was measured by analysis of DAF-FM DA (Merck, Darmstadt, Germany) fluorescence intensity, a specific fluorescent probe for NO [32], [8]. The interaction of actin and eNOS was inhibited using the recently introduced peptide P326TAT (Metabion, Martinsried, Germany) [11]. The inhibitor peptide consist of the 8 amino acid (aa) sequence of the eNOS binding site for actin (P326) tagged to the 10 aa TAT sequence (derived from the transduction domain of HIV TAT) which facilitates the uptake of this peptide into the cells. Thus P326TAT acts as an antagonist for the eNOS/actin interaction. For these experiments, GM7373 cells were seeded on 96-well plates (Becton Dickinson Labware, France) and cultured for 48 h until they reached confluence. Then, culture medium was replaced by the corresponding DAF-FM DA solution (final 10 µM) containing control buffer, 12 mM K⁺ buffer or 50 nM CD buffer each supplemented with a control peptide or P326TAT, respectively. The plate was incubated at room temperature on a shaker (240 rpm) for 3 h. Incubation was followed by a fluorescence read out using a Fluoroskan II plate reader (Thermo Labsystems). Fluorescence intensities of control values (control buffer + control peptide) were averaged and set to 100%. All other values were analysed relative to control conditions.

Statistics

All correlations of stiffness and Lifeact-eGFP fluorescence intensity were performed in paired experiments. The obtained data were analysed using paired t-test if the values were parametrically distributed. If the results were non-parametric, they were analysed by Wilcoxon test. Data in all experiments are given as mean values ± standard error of the mean (SEM) of n experiments. NO detection was performed in unpaired experiments and analysed by ANOVA (+ Bonferroni test) due to parametric distribution of acquired data. Values are given as mean ± SEM with N numbers of used 96-well plates and n number of analysed wells. Results of all experiments were considered statistically significant if p<0.05 (indicated by *).

Destabilization of the Cortical Cytoskeleton by Low Doses of Cytochalasin D

Analysis of cortical fluorescence intensity by whole-cell epifluorescence microscopy faces the problem that bulk fluorescence, emitted by fluorophores from the cell interior, is hardly distinguishable from the fluorescence of cortically located fluorophores. This impedes the analysis of cortical actin using fluorescent probes and epifluorescence microscopy. On the other hand, high resolution microscopy techniques that might resolve the apical cell cortex (e.g. TIRF) cannot be matched with AFM-based stiffness measurements. However, the natural cell shape might allow measurements of cortical fluorescence by quantitative fluorescence analysis in the peripheral zones of the cell. The cell periphery is rather flat and thus dominated by the cell cortex. Hence, peripheral fluorescence is expected to be significantly determined by cortical fluorophores. The peripheral morphology can be investigated by AFM topography mapping. To do so, we performed AFM surface scans (Fig. 3a) of confluent monolayer of bovine aortic endothelial cells (GM7373). The analysis of the height profile of the cells (Fig. 3a, dashed line) demonstrates that the cells are, as expected, much flatter in the region of the cell periphery than in the cell centre (Fig 3b). Quantification of peripheral cell heights indicates an average height of 415±28 nm (n = 33). Figure 3c displays 4 representative surface scans (100×100 µm) of endothelial cells (peripheral regions marked by white frames). Due to the rather low height of the cell periphery, the fluorescence emitted by cortical (Lifeact-)fluorophores holds a higher fraction of the total intensity compared to the bulk fluorescence in the respective “far-away” regions. Hence, in a Lifeact-eGFP expressing cell, any depolymerisation of cortical F-actin induces a diffusion of the freed Lifeact fluorophores from the cortex into the bulk of the cell, resulting in a decrease of peripheral fluorescence intensity. In contrast, any polymerization may recruit new Lifeact molecules, which leads to an increase of peripheral intensity (Fig. 3d). This enables the investigation of dynamic processes in the cell cortex of living endothelium by epifluorescence microscopy. We analysed the fluorescence intensity in the peripheral regions of Lifeact-eGFP transfected cells to detect changes in cortical F-actin, as depicted in figure 3e (circled regions, same cell as in Fig. 3a, 3c). As a proof of principle that cortical actin dynamics determines cortical stiffness, we destabilized the cortical cytoskeleton with cytochalasin D (CD). Large doses of CD, commonly used to destabilize the actin cytoskeleton (∼ 1 µM CD), induce a non-selective depolymerisation F-actin [13], [33]. As shown in figure 4a, whole cell F-actin of a Lifeact-eGFP transfected cell (sitting in a non-transfected - and thus non-fluorescent - monolayer) is found depolymerized 18 minutes after CD application. The close-up views of the cell periphery illustrate that cortical F-actin (indicated by #) as well as actin stress fibres (*) are destabilized. However, as shown in figure 4b, application of a 20-times low dose of CD (50 nM) selectively destabilizes cortical F-actin (#) while classical stress fibres (*) are unaffected. In average, peripheral Lifeact fluorescence, and thereby cortical F-actin, was reduced to 86.7±3.8% of control after 16 minutes of CD treatment (Fig. 4c, black circles). The selective destabilization of the cortical actin can be simultaneously analysed by the AFM-based nanoindentation technique [29], [34]. This technique allows performing one-dimensional stiffness tomography of living cells using force-indentation curves (F/I-curve, for details see methods part). Analysis of the slope of these F/I-curves allows a quantification of stiffness in different layers (compartments) of the cell. Both parameters, cortical and bulk stiffness, of living endothelial cells are constant under control conditions (Fig. 4c, n = 7). The minor variations of up to 2% are due physiological oscillations of К_cortex induced by actin dynamics [35]. Immediately after applying 50 nM CD, К_cortex decreases (Fig. 4c, red circles). The effect is significant after 4 minutes (p<0.05) and reaches a new plateau after approx. 16 minutes at 63.6±5.10% of control (1.9±0.16 pN/nm to 1.2±0.11 pN/nm). In contrast to the cortex, the К_bulk is not affected by these low concentrations of CD and still exhibits a stiffness of 98.7±6.0% of control (19.1±1.12 pN/nm to 18.9±1.07 pN/nm, Fig. 4c, red triangle). Hence it is possible to perform an optical and nanomechanical sectioning of the cell while selectively influencing the thin cell cortex. These results indicate that К_cortex is an expression of cortical actin dynamics. The described decrease in fluorescence intensity correlates directly with the softening of the cell cortex, since changes in К_cortex and cortical F-actin show an almost linear dependence (R² = 0.85, Fig. 4d). Modifications in К_bulk and cortical F-actin, however, show no correlation at all (R² = 0.12, Fig. 4e).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. Cortical fluorescence in the cell periphery.

a) 3D topography map of a confluent endothelial monolayer generated by an AFM surface scan. Cell height is illustrated by the colour code, shown at the right. The dashed line indicates the section for the height profile (scale bar = 20 µm). b) Height profile of the transfected endothelial cell (see also a). c) Height analysis of cell periphery. Four representative surface scans of endothelial cells are shown. Values indicate mean heights of the respective cell periphery. Quantification of all analysed cells (n = 33) is displayed. d) Proposed model of Lifeact-eGFP binding to actin filaments. Actin depolymerisation of cortical filaments leads to a dissociation of actin and Lifeact which results in a diffusion of fluorophores into the cell centre and a decrease of peripheral fluorescence intensity. e) Epifluorescence of the Lifeact-eGFP transfected cell (see also e). The white frames indicate the region of interest used for analysis of peripheral fluorescence intensity.

https://doi.org/10.1371/journal.pone.0041520.g003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 4. Proof of principle.

a) Fluorescence images of a Lifeact-eGFP transfected cell (grown in a confluent monolayer of non-transfected cells) before and after treatment with 1 µM CD. Cellular as well as cortical F-actin depolymerizes upon treatment with CD. Highlighted details illustrate the time course of the CD effect. Cortical F-actin (#) as well as F-actin stress fibres (*) are destabilized. b) Lifeact-eGFP transfected cell (grown in a confluent monolayer of untransfected cells) treated with 50 nM CD. Application of low doses of CD destabilizes cortical F-actin (#) and spares bulk stress fibres (*). c) Simultaneous analysis of cortical F-actin ( = peripheral Lifeact-eGFP fluorescence intensity, black circles), cortical stiffness (К_cortex, red circles) and bulk stiffness (К_bulk, red triangles). Application of 50 nM cytochalasin D (CD) decreases cortical F-actin and К_cortex significantly (after 4 minutes, p<0.05) whereas К_bulk is unaffected (n = 7). d) Correlation of CD-induced changes of К_cortex and cortical F-actin. The correlation coefficient, R² = 0.85, indicates a linear correlation. e) Analysis of the correlation between К_bulk and cortical F-actin indicates no correlation (R² = 0.1).

https://doi.org/10.1371/journal.pone.0041520.g004

Physiological Depolymerisation of Cortical F-actin

Although CD is a useful agent to depolymerise actin, it is a rather unphysiological stimulus. To verify that actin dynamics is a major determinant of К_cortex, we induced a change in К_cortex by varying the external potassium concentration within a physiological range. It is known that, in the healthy human organism, physiological plasma potassium concentrations vary around 3.7–5.3 mM. However, locally, extracellular potassium ([K⁺]_e) can rise up to 12 mM due to increased muscle or nerve cell activity [36], [37]. Previous data from our lab demonstrated that high [K⁺]_e induces a significant softening of the cortex paralleled by an increase of NO bioavailability [5]. Hence, if cortical actin dynamics is indeed a functional component of К_cortex, an increase in [K⁺]_e should depolymerize cortical actin and thereby decrease К_cortex. Therefore, we analysed the effect of high [K⁺]_e on К_cortex and cortical actin dynamics. As shown in figure 5a, both parameters, К_cortex (red circles) and cortical F-actin concentration (black circles), show an immediate decrease as soon as [K⁺]_e is isoosmotically increased from 5 to 12 mM, which is significant after 6 minutes (n = 12, p<0.05). After 14 minutes К_cortex reaches a new plateau at 84.4±2.3% of control (1.6±0.06 pN/nm to 1.4±0.05 pN/nm). Simultaneously cortical F-actin decreases to 95.2±0.7% of control. The correlation coefficient of R² = 0.79 indicates an almost linear response (Fig. 5b). In contrast, К_bulk is not affected by [K⁺]_e as is remains constant and shows a stiffness of 97.9±2.40% of control (19.1±0.84 pN/nm to 18.6±0.80 pN/nm). Similar to the CD-dependent effect, there is, due to a correlation coefficient of R² = 0.06, no correlation of cortical F-actin and К_bulk (Fig. 5c). Taken together, a physiological trigger, e.g. [K⁺]_e, depolymerizes the submembranous actin web and thus explains the decrease of К_cortex.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 5. Physiological relevance.

a) Simultaneous measurement of cortical F-actin intensity (black circles), cortical stiffness (К_cortex, red circles) and bulk stiffness (К_bulk, red triangles). After an increase of [K⁺]_e from 5 to 12 mM cortical F-actin and К_cortex decrease. The effect is significant after 6 minutes (n = 12). К_bulk is not affected by an increased [K⁺]_e. b) Changes in К_cortex and Lifeact-eGFP intensity show an almost linear correlation (R² = 0.79), c) whereas the correlation of K⁺-induced changes in cortical F-actin and К_bulk do not correlate (R² = 0.06).

https://doi.org/10.1371/journal.pone.0041520.g005

Lifeact Fluorescence and Cell Volume

The described method to analyze cortical actin dynamics based on measurements of fluorescence intensity in the cell periphery. However, it is known that CD and [K⁺]_e influence cell volume [5] and cell volume regulation [38]. Therefore, it cannot be excluded, that the described effects on cortical Lifeact-eGFP fluorescence are due to a swelling-induced dilution of fluorophores in the analysed regions and not, as postulated, to an expression of F-actin dynamics. A swelling-induced dilution of fluorophores may lead to diminished fluorescence intensity and thereby to a misinterpretation regarding the cytoskeletal organization. To eliminate this possibility, we treated the transfected endothelial cells with low extracellular chloride ([Cl^-]_e) and simultaneously analysed К_cortex and Lifeact-eGFP fluorescence intensity. An isoosmotic decrease in [Cl^-]_e reduces cell stiffness but is, in contrast to the effect of high [K⁺]_e, accompanied by a significant decrease in cell volume [31]. If the effect of high [K⁺]_e on cortical Lifeact-eGFP fluorescence was only due to a dilution of fluorophores, based on an increase of cell volume, a Cl^–induced decrease in cell volume should exert the opposite effect. Accordingly, low [Cl^-]_e should increase fluorescence intensity because of cell shrinkage despite cortical softening. However, as shown in figure 6, this is not the case. A change to low [Cl^-]_e conditions induces a decrease in cortical Lifeact-eGFP fluorescence intensity (n = 8). After 10 minutes, Lifeact-eGFP intensity (black circles) and К_cortex (red circles) decline to 95.4±2.8% and 86.5±5.8% (1.8±0.10 pN/nm to 1.6±0.08 pN/nm), respectively (Fig. 6a). The effect of low [Cl^-]_e is significant after 4 minutes (p<0.05) and shows an almost linear correlation with a correlation coefficient of R² = 0.76 (Fig. 6b). Interestingly, the induced cell shrinkage significantly affected К_bulk as it increases after application of low [Cl^-]_e conditions (Fig. 6c). This effect may be due to the higher concentration of cell mass (i.e. proteins, organelles, vesicles etc.) in the cytoplasm. Taken together, these findings indicate that the observed correlation of cortical stiffness and Lifeact-eGFP intensity cannot be explained by changes in cell volume. Hence cortical Lifeact-eGFP intensity indeed represents submembranous F-actin dynamics. Furthermore, the simultaneous analysis of К_cortex and К_bulk underlines the independent role of the cell cortex as a functional compartment within the cell.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 6. Exclusion of volume artefacts.

a) Effect of low [Cl^-]_e on cortical F-actin (black circles) and К_cortex (red circles) of endothelial cells are detected simultaneously. Changing to low [Cl^-]_e decreases stiffness and F-actin in the cell cortex. b) Correlation of low [Cl^-]_e-induced changes between cortical F-actin and К_cortex. The correlation coefficient of R² = 0.76 indicates a linear dependence (n = 8). c) Effect of low [Cl^-]_e on К_bulk (triangles) and cortical F-actin (black circles). While cortical F-actin depolymerizes (see also d), К_bulk increases simultaneously.

https://doi.org/10.1371/journal.pone.0041520.g006

Actin-associated К_cortex and eNOS/actin Interaction

After verifying that cortical actin dynamics is a major determinant of cortical stiffness, we addressed the question whether a decrease in cortical F-actin, and thus a larger G-actin content, also causes increased eNOS activity. It has been shown before that eNOS activity is stimulated by G-actin [11]. Hence, elevation in cortical G-actin should lead to increased G-actin/eNOS association and thereby to NO release. To confirm this presumption, we analysed the effect of [K⁺]_e and CD on endothelial NO release, with and without inhibition of the eNOS/actin interaction. eNOS activity was measured using DAF-FM DA, a fluorescent probe which detects intracellular released NO [32], [8]. To inhibit the eNOS/actin interaction, P326TAT was applied to the cells. The 18 amino acid peptide P326TAT exhibits an eNOS-homologue binding domain for actin which “masks” actin and thereby inhibits eNOS/actin association [11]. After loading the cells with DAF-FM DA, eNOS activity was analysed in the presence of P326TAT or a control peptide (N = 5, n = 30). eNOS activity under control conditions was averaged and set to 100%. As shown in figure 7, P326TAT significantly decreases eNOS activity to 79.8±3.1% of control conditions. 12 mM [K⁺]_e as well as 50 nM CD significantly increases NO production to 111.0±2.9% and 127.0±3.4% of control (p<0.05). In both cases, endothelial NO release is significantly decreased when the eNOS/actin interaction is inhibited by P326TAT. The inhibiting peptide reduced eNOS activity to 78.4±2.5% of control under high [K⁺]_e. CD-stimulated eNOS activity is lowered by P326TAT to 96.2±3.8% of control (Fig. 7). The data indicate that cortical NO release is indeed mediated by a depolymerisation of F-actin in the cell cortex.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 7. eNOS actin interaction.

Effect of high [K⁺]_e and CD on endothelial NO release. Cells loaded with the fluorescent NO indicator DAF-FM DA were incubated in control buffer, high [K⁺]_e or CD in the presence of control peptide or P326TAT, respectively (N = 5, n = 30). DAF-FM fluorescence intensity under control conditions was averaged and set to 100%. Respective values are given as % of control ± SEM. High [K⁺]_e and CD significantly increase NO release relative to control. In all approaches P326TAT inhibits the eNOS/actin interaction and thereby the NO release.

https://doi.org/10.1371/journal.pone.0041520.g007