Ethics Statement

All animals were maintained in our animal facilities in compliance with the guidelines of the UFMG (Universidade Federal de Minas Gerais) ethics committee for the use of laboratory animals (protocol 45/2009 approved by CETEA-UFMG) and are in accordance with CONCEA, the Brazilian institution that regulates animal husbandry.

Cell Cultures

Most of the experiments were performed using a cloned mouse fibroblast cell line (WTCL3). These fibroblasts were derived from wild type (WT) C57BL/6J embryos and generated in Dr. Paul Saftig’s laboratory (University of Kiel, Germany) by spontaneous immortalization of primary embryonic fibroblasts in culture, around passages 10–20 [22]. Cells were maintained in culture by consecutive passages in high glucose DMEM (Invitrogen, Carlsbad, CA, USA) supplemented with 10% FBS, 1% penicillin-streptomycin and 1% glutamine, as described previously [23]. For siRNA assays we used NRK (Natural Rat Kidney) cells, which were also maintained in DMEM supplemented with 10% FBS, 1% penicillin-streptomycin and 1% glutamine. Primary cardiomyocyte cultures, used in the lysosomal exocytosis assay in the presence of BAPTA-AM, were prepared as described previously [15]. Briefly, 15 newborn mice were killed by decaptation and had their hearts removed aseptically. Those hearts were minced and submitted to trypsin and collagenase digestion. In the end of isolation procedure, enzymes were inactivated and cells were plated in 24 well plates at a density of 1,0×10⁵ cells/well. Experiments were performed after 72 hours of culture.

For all experiments, except for primary cardiomyocyte cultures, cells were plated 24 hours in advance. For lysosomal exocytosis assays or actin labeling, the fibroblasts were plated on 24 well plates containing 13 mm round coverslips at a density of 5.0×10⁴ cells/well. For tether extraction or defocusing microscopy experiments, fibroblasts were plated on 35 mm glass-bottom dishes at a density of 2.0×10⁵ cells/well. For cholesterol quantification and Rho activation assays fibroblasts were plated on 6 well plates at a density of 5.0×10⁴ cells/well or on 60 mm culture dishes at a density of 2.5×10 ⁶ cells/well, respectively. For knockdown experiments, NRK cells were seeded on 6 well plates at a density of 1.25×10⁵ cells/well.

Drug Treatments

All reagents used were acquired from Sigma-Aldrich (St. Louis, MO, USA) unless otherwise stated. For plasma membrane cholesterol sequestration, cells were first washed three times with Hank’s Balanced Salt Solution supplemented with 1.8 mM of CaCl₂.2H₂O (HBSS +), and then incubated with 5 or 10 mM methyl-beta cyclodextrin (MβCD) solutions diluted in DMEM high glucose (Invitrogen, Carlsbad, CA, USA), in the absence of serum, for 45 minutes at 37°C and 5% CO₂. After treatment, cells were washed three times with HBSS+ and subjected to experimental assays. For F-actin cytoskeleton disruption assays, fibroblasts were incubated with 95 nM Latrunculin-A (Lat-A) for 1 hour at 37°C and 5% CO₂, washed three times with HBSS+ and incubated with DMEM alone (without serum) or in the presence of 5 or 10 mM of MβCD, for 45 minutes.

Quantification of Plasma Membrane Cholesterol

In order to quantify the amount of cholesterol in control and MβCD treated cells, fibroblasts were first washed with HBSS+ and then incubated with phosphate buffered saline containing 0.02 mM of ethylenediamine tetraacetic acid (PBS- EDTA) for 5 minutes at 37°C and 5% CO₂. The detached fibroblasts were collected, counted using a Neubauer chamber and submitted to lysis for 1.5 h in a 4°C ice-cold buffer solution containing 100 mM Tris-HCl (pH 8.0), 150 mM NaCl, 2 mM MgCl₂, 1% Triton X-100, 5 mM iodoacetamide, 0.025% NaN₃, 1 mM PMSF, 1 mM di-isopropylfluorophosphate and 0.02 U/mL of aprotinin, with gentle stirring. After lysis, small aliquots of the cellular extracts were collected for protein quantification using the Bradford method (Bio-rad, Brazil) [24]. The remaining lysates were submitted to the Folch lipid extraction method [25]. Briefly, the extracts were diluted in a 8∶4:3 chloroform:methanol:sample ratio, the upper aqueous phase was removed and the organic phase was dried under N₂. Dried non-polar lipids were subjected to cholesterol quantification using an Amplex Red Cholesterol Assay Kit (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions. Untreated and cholesterol-sequestered fibroblasts apolar lipids were suspended in 500 µL of methanol and 5 µL of these solutions were diluted in 45 µL of the reaction buffer (1∶10) and added to a 96 well black microplate. After that, 50 µL of a solution containing Amplex Red Reagent/Horseradish Peroxidase (HRP)/cholesterol oxidase/cholesterol esterase were added to every well and the reaction was incubated for 30 minutes, at 37°C, protected from light. The reaction buffer (without cholesterol) alone or containing 10 mM of H₂O₂ were used as negative and positive controls, respectively. The fluorescence levels were measured in a fluorimeter (Gemini XPS- Molecular Devices, Sunnyvale, CA, USA) using of 560 nm for excitation and 590 nm for emission. The amount of cholesterol was quantified for all the experimental conditions.

Labeling of Polymerized Actin

After drug treatment, fibroblasts were washed 3 times with HBSS+ and fixed with paraformaldehyde (PFA) 4% in PBS for 15 minutes at room temperature. Cells were rinsed 3 times with PBS and permeabilized with Triton X-100 0.02% for 5 minutes before incubation with PBS/BSA 5% for 30 minutes. In sequence, cells were incubated with 50 µg/mL of Phalloidin-FITC in PBS for 2 hours. Following, monolayers were washed 3 times with PBS, incubated with 4′,6-Diamidino-2-Phenylindole, Dihydrochloride (DAPI) for 5 minutes, washed 4 times with PBS and mounted in glass slides using FluorSave™ Reagent (Calbiochem-Millipore, Billerica, MA, USA). Images were captured using a TCS-SP5 II confocal microscope (Leica Microsystems, Germany) equipped with LAS AF 2.2.0 Software (Leica Microsystems, Germany).

Rho-GTP Pull-down Assay

In order to quantify Rho-GTP activation upon MβCD treatment, fibroblasts were incubated with MβCD 10 mM for 20 or 45 minutes. After incubation, cells were washed with ice-cold TBS (25 mM Tris- HCl, pH 7.5 and 150 mM NaCl) and lysed using the lysis buffer provided by Active Rho Pull-Down and Detection Kit (Thermo Scientific Waltham, MA, USA), together with a protease inhibitor cocktail (Aprotinin 0.8 µM; 2 mM PMSF, 0.01 mM Pepstatin-A, 20 µM Leupeptin and 5 mM EDTA pH 8.0). Cell extracts were then collected and centrifuged at 16.000 g for 15 minutes and used for the pull down assay according to the manufacturer’s instructions, as follows. As an internal control, a small fraction of each cellular extract (20 µL of each group) was collected in order to measure the total amount of Rho present in each sample. The extracts were transferred to collection tubes coupled to spin cups, each tube containing 100 µL of glutathione resin slurry and 80 µL of GST-Rhotekin-RBD. Cell extracts for incubation with GDP were also prepared as negative controls for the pull-down assay. After transference, the tubes were incubated at 4°C for one hour, centrifuged at 6.000 g for 30 seconds and washed with lysis/binding/washing buffer. Then the eluates and cell extracts received 50 µL of sample buffer SDS 2X containing β-mercaptoethanol (provided by the kit) and were boiled for 5 minutes. The samples (together with a molecular marker, Bio-rad Brazil) were run on a 10% SDS-PAGE acrylamide gel and transferred to a PVDF membrane (Millipore, Billerica, Massachusetts, USA). The membrane was blocked with TBS, supplemented with 3% of bovine serum albumin (BSA), for 2 hours, at room temperature and incubated, at 4°C, with 10 mL of TBST with 3% of BSA and 15 µL anti-Rho primary antibody overnight. The membrane was then washed 5 times with TBST and incubated for one hour at room temperature with horseradish peroxidase-conjugated anti-rabbit antibody (1∶2,000 diluted in TBST 5% non-fat dry milk). After washes, immunoreactive GTP-bound Rho was detected by enhanced chemiluminescence (ECL) according to the manufacturer’s instructions (GE Healthcare, Pittsburgh, PA, USA). Densitometric analysis was performed with ImageJ software (National Institutes of Health, Bethesda, MD, USA).

Membrane Tether Extraction Experiments with Optical Tweezers

Fibroblasts were treated with 5 or 10 mM MβCD, washed 3 times with HBSS+, replenished with fresh media, and submitted to tether extraction using optical tweezers (OT). The OT system used employed an infrared Nd:YVO₄ Osprey laser with a wavelength of 1.064 µm (Quantronix, East Setauket, NY, USA). The laser has a Gaussian intensity profile (TEM₀₀ mode) and a beam half width of 2.3±0.2 mm at the back focal plane of the objective lens. The infrared laser is coupled to an Eclipse TE300 inverted microscope (Nikon, Melville, NY, USA) equipped with a PLAN APO 100X 1.4 N.A. DIC H objective that generates the optical trap. The OT calibration and tether extraction experiments were performed as described previously [26], [27]. Briefly, WTCL3 fibroblasts treated with 5 or 10 mM MβCD received a 1 µL solution of uncoated polystyrene beads of 1.52±0.2 µm radius (Polysciences Inc, Warrington, PA, USA) and each dish was placed on the microscope. One bead was captured with the OT and pressed against the membrane of a chosen cell for 5 seconds, in order to attach to it. The motorized microscope stage (Prior Scientific, Rockland, MA, USA) was set to move with a controlled velocity of 1 µm/s. Images of the entire process were collected by a charge-couple device (CCD) C2400 camera (Hamamatsu, Japan) and digitalized by a Scion FG7 frame grabber (Scion, Torrance, CA, USA) using a frame rate of 10 frames/second. Variations in the trapped bead equilibrium position as a function of the stage displacement was determined using ImageJ software (National Institutes of Health, Bethesda, MD, USA). Using the trap calibration, the displacement of the bead’s center of mass was converted to measured force. All the experiments were performed in an environmental chamber that maintained 37°C temperature and 5% CO₂. The data were analyzed using Kaleidagraph software (Synergy Software, Essex Junction, VT, USA).

Membrane Tether Radius Measurements by Scanning Electron Microscopy (SEM)

In order to measure the tether radius extracted from control and MβCD treated cells, immediately after extracting the tethers, cells were fixed using a solution containing 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) for 1 hour, at room temperature. Fixed fibroblasts were rinsed with 0.1 M cacodylate buffer and incubated with OsO₄ 1% in cacodylate buffer 0.1 M, for 40 minutes. Cells were then dehydrated using an ethanol series ranging from 10 to 100%. After dehydration, samples were removed from culture dishes, dried at a CPD 030 Critical Point Drier (BAL-TEC, Liechtenstein) and mounted on specimen stubs. The next step was to metalize the samples using a SCD 050 Sputter Coater (BAL-TEC, Liechtenstein) and observe the tethers in a FEI QUANTA 250 scanning electron microscope (FEI Company, Hillsboro, OR, USA). In order to measure the tether radius, we used the methodology previously established [27]. After measuring the tether radius, R, and the tether force, F₀, both cell surface tension, σ, and bending modulus, κ, were determined using the following equations:(1)(2)

Defocusing Microscopy

Defocusing microscopy (DM) technique was used to analyze structural modifications in both plasma membrane and cytoskeleton of fibroblasts during treatment with 5 or 10 mM of MβCD, in comparison to control untreated cells. With this method, it is possible to turn a phase-transparent object (like adhered cells) into a visible object by introducing defocalization in the microscope [28], [29]. We recorded control fibroblasts by DM in serum free DMEM High Glucose for 10 minutes, then added 5 or 10 mM of MβCD to the medium and continued recording this cell for another 45 minutes. All experiments were analyzed using ImageJ Plugins that correct for background and calculate the temporal autocorrelation function among frames. These temporal autocorrelation functions were adjusted in KaleidaGraph Software (Synergy Software, Essex Junction, VT, USA) using single exponential decay curves that carry information about their amplitude and time characteristics. The experiments were performed in a Nikon Eclipse TI inverted microscope equipped with a 530 nm wavelength green filter, a stage-heated oil-immersion objective Nikon Apo Tirf 100X, NA 1.49 (Nikon, Japan), and a an environmental chamber (model Chamlide IC- CU:109, Live Cell Instrument, Nowan-gu, Korea) which provides a 5% CO₂ atmosphere, 37°C temperature and 50% humidity. The images were captured using a 12 bit Uniq camera (model 1800 CL) (Epix Inc, Buffalo Grove, IL, USA, 4096 gray levels and 0.064 µm of pixel square side), with a gain of 11.04 db and a capture rate of 1 frame per second. The focal distance was controlled during the entire experiment by a Nikon Perfect Focus System (PFS) apparatus and the camera gray level calibration was performed as previously described [28].

Lysosomal Exocytosis Assay

In order to quantify the amount of lysosomal exocytosis upon cholesterol removal and/or cytoskeleton disruption by treatment with MβCD or Lat-A, respectively, a time and dose-dependent β-hexosaminidase (β-hex) secretion assay was performed as previously described [15]. Briefly, cells previously treated or not with Lat-A were incubated with 5 or 10 mM of MβCD for 10, 20 or 40 minutes. Untreated cells and ionomycin treated cells (10 µM, 10 minutes) were used, respectively, as negative and positive controls. For the exocytosis assays in primary cardiomyocyte cultures, cells were treated with 10 mM of MβCD in the presence or absence of 1 mM of BAPTA-AM. After treatment, 350 µL of the cell supernatant media were collected and the cells were lysed using 1% Triton X-100 in PBS. The recovered media and 50 µL of the cellular lysates were incubated with 6 mM of β-hex substrate (4-methylumbelliferyl-N-acetyl-B-D-glucosaminide) dissolved in sodium citrate-phosphate buffer, pH 4.5. The reactions were stopped by adding to each sample 100 µL of the stop solution (2 M Na₂CO₃.H₂O and 1.1 M glycin) followed by reading in a spectrofluorimeter at 365 nm excitation and 450 nm emission (Synergy 2, Biotek, Winooski, VT, USA). The results were expressed as secreted β-hex as a percentage of the total enzyme content in the cells.

Lysosomal Dispersion Analysis

Lysosomal dispersion analysis was performed as described previously [15]. Briefly, control fibroblasts and fibroblasts treated with either MβCD alone or after Lat-A were fixed and immunostained with anti-LAMP-1 and DAPI as described previously [15]. Images were taken using a 100× objective with N.A. of 1.30 at Olympus BX51 fluorescence microscope (Shinjuku, Tokyo, Japan). To each image (associated to a specific drug treatment), we computed the mean center position and the mean radius (R) of each nucleus. A computational code has been written using the IDL (Interactive Data Language) programming language in order to calculate the lysosomal dispersion distribution within the cells from the different experimental groups. Subsequently, an average distance (D) was computed by using the distances of each lysosome relative to the cell center. Lysosomes farther than 1.5 radii from cell center were excluded from the computation of this average value since the majority of lysosomes were found nearer than 1.5 in relation to the cell nuclei. Moreover, when increasing D values, lysosomes from neighboring cells could be taken into account in the dispersion calculation. Finally, the mean lysosome distance (D) relative to the mean nucleus’ radius (R) was defined as the ratio D/R. Such procedure was repeated for the maximum number of isolated cells available from the image sets of each drug treatment. The results of this analysis are a distribution of D/R values associated to each drug treatment, and are represented as histograms.

siRNA Silencing of Synaptotagmin VII

In order to investigate the role of synaptotagmin VII (Syt-VII) in lysosomal exocytosis triggered by cholesterol removal, we performed siRNA knockdown experiments with NRK cells. Syt-VII knockdown was performed using Lipofectamine RNAiMAX Reagent (Invitrogen, Carlsbad, CA, USA), according to manufacturer’s instruction with the syt-VII siRNA oligo (Invitrogen, Carlsbad, CA, USA, Syt7RSS339874 5′-GCCAACUCCAUCAUCGUGAACAUCA-3′). As a control, a medium GC siRNA oligo (random oligo) (Invitrogen, Carlsbad, CA, USA) was used in parallel. After transfection, cells were harvested either for lysosomal exocytosis experiments, as described above, or used for RT-PCR to assess level of Syt-VII silencing.

For RT-PCR, total RNA was isolated using Trizol reagent (Invitrogen, Carlsbad, CA) according to manufacturer’s instructions. cDNA was prepared from total RNA using the SuperScript III kit (Invitrogen, Carlsbad, CA, USA) in accordance with the company’s instructions. The RT-PCR reaction was performed to verify the presence or the absence of syt-VII cDNA using the following primers: forward 5′-GGG TTT CCC TAT GAG AAA GTG GT-3′ and reverse 5′-CCT TCC AGA AGG TCT GCA TCT GG-3′. Alternatively, as an experimental control, we also amplified actin, a housekeeping gene, using the following primers: forward 5′-CCG TAA AGA CCT CTA TGC CA-3′ and reverse 5′-CAT CTG CTG GAA GGT GGA CA-3′. The PCR reaction was performed with a final volume of 10 µL containing Taq DNA Polymerase buffer (Tris-HCL 10 mM pH 8.4, KCl 50 mM, 0,1% Triton X-100, MgCl2 1.5 mM), 2 pmol of each specific primer, 2,5 units of Taq DNA Polymerase (Phoneutria, Brazil) and 1 µL of cDNA. The mixture was submitted to 35 cycles of 1 minute at 95°C for denaturation, 1 minute at 55°C for annealing and 1 minute at 72°C for extension in an Eppendorf, termocycler (Hamburg, Germany). In the last cycle extension was performed for 10 minutes. PCR products were submitted to electrophoresis in agarose gel 0,1% (wt/vol) and stained with ethidium bromide. The intensities of the bands on the agarose gel were measured using ImageJ software (National Institutes of Health, Bethesda, MD, USA).

Statistical Analysis

All data were represented as mean ± standard error. Student t-test or One-way ANOVA followed by Neuman-Keuls post hoc comparison test were performed in order to calculate statistical differences between control and treated groups. All statistical analyses were performed using GraphPad Prism 4 software (La Jolla, CA, USA). For histogram distributions, the cumulative frequency was calculated and analyzed using the Kolmogorov-Smirnov statistical test, performed using IDL. Values of p<0.05 were considered statistically significant.

Fibroblasts Treated with MβCD had their Cholesterol Content Efficiently Sequestered

In order to evaluate the effectiveness of MβCD treatment in sequestering cellular cholesterol from immortalized fibroblasts, we performed a quantitative assay to measure the amount of this lipid in control (non-treated) and in cells treated with 5 or 10 mM MβCD, using the Amplex Red Cholesterol Assay Kit (Invitrogen, Carlsbad, CA, USA). This quantitative assay confirmed that cholesterol was efficiently sequestered from fibroblasts treated with either 5 or 10 mM of MβCD. Cells incubated with 5 mM MβCD presented 4.5 times less cholesterol whereas cells treated with 10 mM of MβCD showed 6.2 times less cholesterol, when compared to control fibroblasts (Table 1).

Download:

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

Table 1. Incubation with MβCD diminished cholesterol content in fibroblasts.

https://doi.org/10.1371/journal.pone.0082988.t001

Cholesterol Removal Alters Organization of the Actin Cytoskeleton via Rho Activation

Proteomic studies have demonstrated that many cytoskeleton-binding proteins are located in plasma membrane cholesterol-enriched platforms [17], [30]. These findings, together with additional data from the literature, have demonstrated the importance of cholesterol-enriched microdomains in establishing connections between the plasma membrane and the cytoskeleton (reviewed by Chichili and Rodgers, 2009 [31]). Therefore, we decided to examine the organization of the actin cytoskeleton upon cholesterol removal from the plasma membrane in our fibroblast cultures. Figure 1A–C, shows the profile of filamentous actin (F-actin) distribution in control fibroblasts. F-actin was visualized homogenously distributed throughout the entire cell. In contrast, when cells were treated with 5 (Fig. 1D–F) or 10 mM (Fig. 1G–I) MβCD, there was F-actin redistribution with more evident stress fibers when compared to untreated fibroblasts. These results revealed that cholesterol removal from plasma membrane changes the actin cytoskeleton organization in fibroblasts, confirming previous results from Qi and co workers (2009) [21]. These authors also demonstrated that the actin cytoskeleton reorganization and stress fiber formation occurred via Rho activation. In order to verify whether the effect on actin observed in our system could also be a consequence of Rho activation, we performed a Rho-GTP pull-down assay. Eluates obtained from fibroblasts cell lysates after the pull-down assay (Fig. 2A, left side), treated or not with MβCD, were run on a gel and transferred to a PVDF membrane and detected with an Anti-Rho GTP antibody. Simultaneously, cell extracts (containing the total amount of Rho) were also run on the gel and transferred to a PVDF membrane (Fig. 2A, right side). These inputs were used to normalize the values, obtained by the densitometry analysis, for the eluates. We observed that after 40 minutes of exposure to MβCD there was an almost 2 fold increase in the amount of activated Rho (Rho-GTP) when compared to control untreated cells (Fig. 2A and B). Rho activation was also measured in cell extracts incubated with GDP, as a negative control (Fig. 2A). Figure 2A also shows the cellular extracts, indicating that the amount of protein loaded in each lane was similar. In agreement with and also reinforcing previous results, we showed that cholesterol sequestration, due to MβCD treatment, induces Rho activation, which is probably responsible for the observed actin reorganization pattern.

Download:

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

Figure 1. Cholesterol removal changed actin cytoskeleton organization.

Representative images of actin cytoskeleton from fibroblasts untreated (A–C) or treated with either 5 (D–F) or 10 (G–I) mM of MβCD for 45 minutes at 37°C, fixed and labeled with Phalloidin-FITC and DAPI (cell nuclei). Three independent experiments were performed in triplicates and at least 50 different cells were observed per group. Each selected image represents the most predominant actin cytoskeleton cell morphology of that experimental condition. Scale bars are all 10 µm.

https://doi.org/10.1371/journal.pone.0082988.g001

Download:

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

Figure 2. Cholesterol sequestration was able to activate Rho upon treatment with MβCD.

Fibroblasts, treated with MβCD 10 mM, for 20 or 40 minutes, were submitted to Rho-GTP pull down assay. In (A), left side, we have the western-blotting result of the pull-down eluates showing that after 20 or 40 minutes of cholesterol removal Rho is activated in fibroblasts. In (A), right side, we have cell lysates that were not submitted to the pull-down assay showing the total amount of Rho present in each experimental group. Control untreated fibroblasts and cell extract incubated with GDP were used as controls of the Rho activation and of the pull-down technique, respectively. (B) Graph showing the densitometric analysis of activated Rho seen in the autoradiogram of the upper panel (A) normalized by the corresponding band at the cell lysates. The data are mean ± standard error of two independent experiments. Data are expressed in arbitrary units (folds in relation to control).

https://doi.org/10.1371/journal.pone.0082988.g002

Cholesterol Removal Induces “de novo” Formation of Actin Filaments

The results described above confirmed that cholesterol removal changes actin cytoskeleton organization in fibroblasts and activates Rho GTPase, which is known to induce actin polymerization. However, whether the stress fibers induced by cholesterol sequestration resulted from existing F-actin or from de novo formation of new F-actin filaments was not known. To examine this issue, we pre-treated cells with 95 nM Latrunculin-A (Lat-A), a drug that disrupts F-actin cytoskeleton by sequestering actin monomers, for 1 hour before treatment with MβCD. As controls, fibroblasts treated with Lat-A were washed and re-incubated in fresh media without the drug for an additional 45 minutes. Figure 3A–C show control cells treated with 95 nM Lat-A for 1 hour followed by 45 minutes of incubation with fresh medium. As expected, these fibroblasts exhibited extensive disruption of F-actin cytoskeleton. Figures 3D–F show cells previously treated with 95 nM Lat-A for 1 hour followed by 45 minutes incubation with 5 mM MβCD. In these cells it was possible to observe a complete reorganization of the F-actin, where most of the cells presented cytoskeleton organization similar to control non-treated cells (Fig. 1A–C). These results demonstrate that cholesterol extraction with MβCD treatment is able to induce de novo actin polymerization. Treatment of cells with 10 mM MβCD after 95 nM Lat-A incubation led to a full recovery of the actin cytoskeleton organization, with all cells in the field showing normal distribution of F-actin structures (Fig. 3G–I). These results demonstrate that cholesterol removal not only promotes actin filament reorganization but can also induce its de novo polymerization, which is more prominent in the case when the highest concentration of MβCD was used.

Download:

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

Figure 3. Cholesterol sequestration induces de novo actin polymerization.

Fibroblasts were submitted to the following treatments: 95 nM of Latrunculin-A (Lat A) for one hour followed by washing with HBSS+ and incubation with fresh DMEM, without serum, for 45 minutes (A–C); previous incubation with Lat-A followed by washing with HBSS+ and incubation with 5 (D–F) or 10 (G–I) mM of MβCD diluted in fresh DMEM, without serum, for 45 minutes. The cells were fixed simultaneously and processed for F-actin cytoskeleton labeling using Phalloidin-FITC. Cell nuclei were stained with DAPI. Three independent experiments were performed in triplicates and at least 50 different cells were observed per group. Each selected image represents the most predominant actin cytoskeleton cell morphology of that experimental condition. Scale bars are all 10 µm.

https://doi.org/10.1371/journal.pone.0082988.g003

Cholesterol Removal Alters Membrane-cytoskeleton Mechanical Properties

Once we determined that cholesterol removal was able not only to reorganize, but to induce de novo formation of F-actin filaments, we proceeded to test the effects of these membrane-cytoskeleton modifications on the mechanical properties of the cells. For this, both control and MβCD treated fibroblasts were submitted to tether extraction using the OT technique. Once the tether is formed (Fig. 4A), the force F₀ required to extract it is uniform and appears as a plateau in the force versus displacement curve (Fig. 4B). We measured F₀ for control and cholesterol sequestered fibroblasts and the results showed that the lower the cholesterol content in cell membranes, the higher F₀ forces were needed to pull the membrane tether (Fig. 4C).

Download:

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

Figure 4. Cholesterol sequestration and tether extraction.

(A) Typical image of a tether extracted from a fibroblast, indicated with a white arrowhead. Scale bar is 5 µm. (B) Typical force versus displacement plot showing the plateau corresponding to F₀. (C) Plot representing the mean values of the force F₀ for the control, 5 and 10 mM of MβCD. (D) Representative image of a tether obtained by scanning electron microscopy. Scale bar is 1 µm. (E) Plot representing the mean values of the radii obtained for control 5 and 10 mM of MβCD. Three independent experiments were performed and at least 20 different cells were analyzed per each group. The plotted results represent the mean ± standard error of all the measurements. ns means no significant difference and ** means p<0.01 in student t-test.

https://doi.org/10.1371/journal.pone.0082988.g004

We also measured tethers’ radii (R) for each experimental condition by using scanning electron microscopy (Fig. 4D–E) and used both F₀ and R measurements in equations 3 and 4 to determine the surface tension (σ) and the bending modulus (κ) of cells, two cellular mechanical properties (Table 2). We observed that cholesterol sequestration, induced by 5 and 10 mM MβCD treatment, significantly increased the surface tension values when compared to untreated cells. Bending modulus was not altered upon treatment of cells with 5 mM of the drug, but an increase could be observed when cells were treated with 10 mM of MβCD when compared to control non-treated cells.

Download:

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

Table 2. Cholesterol depletion increased surface tension and bending modulus.

https://doi.org/10.1371/journal.pone.0082988.t002

These results demonstrate that cholesterol sequestration affects surface tension and bending modulus. Additionally it also indicates that in scenarios where cholesterol is reduced, cells become more rigid.

Cholesterol Removal changes Membrane-cytoskeleton Dynamics

The results obtained so far indicated that upon cholesterol extraction actin polymerizes and these membrane-cytoskeleton alterations change the mechanical properties of the fibroblasts, resulting in an increase in cell rigidity. In order to evaluate how mechanical properties of the cytoskeleton-membrane interface were affected by treatment with MβCD we used, for the first time in cholesterol sequestering experiments, a technique called defocusing microscopy (DM). Using this technique we were able to visualize the changes in cellular mechanics along the process of cholesterol sequestration induced by treatment with MβCD. For these experiments, fibroblasts were plated in individual wells and single cell images were taken continuously at 1 frame per second. We analyzed 10 minutes movie cuts from the whole experiment. During the first 10 minutes images were acquired from untreated fibroblasts. After this, MβCD (either at 5 or 10 mM) was added and images continued to be collected for another 45 minutes. Representative frames are shown in Figure 5, showing pre-treatment control images (Fig. 5A and 5B 0′) and images from 10–45 minutes after addition of 10 mM MβCD (Fig. 5C–G). Temporal correlation fits are shown below each of the frames (Fig. 5B–G). The data was fitted using a temporal autocorrelation function [29], [32] so that the values of the relaxation time (τ) and amplitude of curvature (A) could be determined (Fig. 6). Analysis of the relaxation time observed during cholesterol sequestration with 5 mM of MβCD showed a more defined behavior when compared to the data obtained from cells treated with 10 mM MβCD. Nonetheless, in both cases it was possible to observe that the relaxation time increased after MβCD addition and then decayed during the final 5 minutes of the experiment. The highest relaxation time values observed after treatment with 5 mM MβCD occurred between 10–30 minutes after addition of the drug, whereas for cells treated with 10 mM MβCD the highest values occurred 10–40 minutes after treatment (Fig. 6A and B). Additionally, we also analyzed the amplitudes of the exponential decay curves. According to DM theory, these amplitudes indicate variations in cell curvature [29]. Both treatments, with 5 and 10 mM MβCD, exhibited a reduction in the amplitude of membrane-cytoskeleton fluctuation, which occurred immediately after the addition of the drug and continued to the subsequent times (Fig. 6C and D). In agreement with the results obtained from OT, these experiments showed a clear alteration in cellular mechanical properties, indicating that after cholesterol extraction cells became stiffer when compared to control conditions. Since we were able to test this parameter along the treatment with the drug we were able to show that the process is fast and happens in the first ten minutes of exposure to the drug.

Download:

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

Figure 5. Defocusing microscopy of cells during MβCD treatment.

(A) Representative DM image showing a control untreated fibroblast. Scale bar: 20 µM. (B) The region of the inset in (A) shows control case. (C–G) Same region of the inset in (B) at different stages of cyclodextrin treatment. Temporal correlation functions for each corresponding time interval are also displayed under each representative images. Scale bar: 10 µM.

https://doi.org/10.1371/journal.pone.0082988.g005

Download:

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

Figure 6. Treatment with MβCD alters relaxation time and amplitude of curvature of cellular membranes.

Images of untreated cells were collected during 10 minutes. Following it, cyclodextrin was added and images continued to be collected for the subsequent 45 minutes. Images were analyzed and the relaxation time τ (A and B) and the amplitude of curvature A (C and D) of fibroblasts treated either with 5 (A and C) or 10 mM (B and D) of MβCD was obtained according to the equation above graphs. Five independent experiments were performed and at least three regions of each one of the five cells were analyzed. The plotted data represent a mean ± standard error of these five independent experiments. Different letters indicate statistically significant differences (p<0,05 using One-way ANOVA and Newman-Keuls post-test).

https://doi.org/10.1371/journal.pone.0082988.g006

MβCD Treatment alone Triggers Lysosomal Exocytosis, Independently of the Calcium Sensor Protein Synaptotagmin VII

As we and others recently reported, cholesterol sequestration with β-cyclodextrins treatment induces lysosomal exocytosis in fibroblasts, cardiomyocytes and epithelial cells [13], [14], [15]. However, the need of intracellular calcium signaling and/or extra stimuli for MβCD induced exocytic events has remained unclear. By using wild type BALB/c and NPC1 mice, Chen et al. (2010) reported that 2-hydroxypropyl-beta-cyclodextrin (HPβC) alone was capable of inducing lysosomal exocytosis without the concomitant use of calcium ionophores. However, this study also proposed that, for the cells analyzed, these exocytic events were dependent on extracellular calcium [13]. Xu et al. (2012), on the other hand, reported no effect of cyclodextrin treatment alone in inducing lysosomal exocytosis in MDCK epithelial cells [14]. In our experience with cardiomyocytes, treatment with MβCD alone was sufficient to induce exocytosis of lysosomes, and the process was independent of extracellular calcium [15]. In order to determine whether in fibroblasts MβCD treatment alone was capable of triggering lysosome exocytic events without the participation of calcium, we measured the secretion of the lysosomal enzyme β-hexosaminidase (β-hex). As a positive control we used ionomycin, a calcium ionophore, which is known to induce lysosomal secretion. The results are shown in Fig. 7A. As expected, treatment with ionomycin led to an increase in β-hex activity in the supernanatant of control, untreated cells. Notably, 10, 20 and 40 minutes of incubation with MβCD also induced enhancement in β-hex exocytosis in relation to untreated cells. These results demonstrate that cholesterol sequestration by MβCD is sufficient to induce lysosomal exocytosis in fibroblasts, as previously observed for cardiomyocytes [15]. Since cell viability was not altered by MβCD treatment (data not shown), we also did not expect membrane injuries in this condition, which would lead to extracellular calcium influx, suggesting that lysosome exocytosis induced by cholesterol removal was also independent of intracellular calcium increase.

Download:

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

Figure 7. Lysosome exocytosis due to cholesterol removal is syt-VII independent and mobilizes a specific lysosome pool.

(A) Lysosomal exocytosis assay of fibroblasts exposed to 10 mM of MβCD for 10, 20 or 40 minutes. Ionomycin was used as a positive control for lysosomal exocytosis. Asterisks (*) indicate statistically significant differences between control and treated groups (p<0.05 using student t-test). (B) Knocking down of syt-VII in NRK cells was confirmed by RT-PCR, using sytVII or actin specific primers. Left: Control non-transfected NRK (control), NRK transfected with medium GC siRNA – negative control oligo (NC oligo) and NRK transfected with syt-VII siRNA oligo (siRNA sytVII). Right: quantification of the intensity of the gel bands (C) Lysosomal exocytosis assay of non-transfected NRK (control cells), NRK transfected with medium GC siRNA – negative control oligo (NC oligo) and NRK transfected with syt-VII siRNA oligo (siRNA sytVII). Asterisk (*), hashtag (#) and diamond (υ) indicate statistically significant differences between MβCD treated groups and their respective controls (p<0.05 using student t-test). (D) Lysosomal exocytosis assay of fibroblasts pre-treated or not with Lat-A 95 nM, for one hour, followed or not by incubation with MβCD 10 mM for 10, 20 or 40 minutes. Asterisks (*) and hashtags (#) indicate statistically significant differences between control and MβCD treated groups (*) or Lat-A and Lat-A+MβCD treatment (#) (p<0.05 using student t-test). (A, C and D) The results are expressed as the ratio between β-hex activity in the cell supernatant and total β-hex activity and are shown as mean of triplicates ± standard error. Three independent experiments were performed in triplicates. Pooled data are shown as mean of all results ± standard error.

https://doi.org/10.1371/journal.pone.0082988.g007

In cardiomyocytes, we previously observed that cholesterol sequestration triggered the exocytosis of peripheral lysosomes, suggesting that there might be two functionally independent sets of these organelles [15]. It had also been shown in the literature that peripherally-docked lysosomes represent the population that fuses with the plasma membrane upon intracellular calcium elevation, under the regulatory effect of synaptotagmin VII (syt-VII) [33], [34]. Syt-VII is a calcium sensor protein present in lysosomes, which is able to detect calcium in the intracellular milieu and control the triggering and secretion of those organelles [12]. Based on these data, we decided to investigate whether peripheral lysosome secretion induced by MβCD treatment required Syt-VII. In order to deplete cells in Syt-VII, siRNA knock-down experiments were performed using NRK cells. Cells were either transfected with an oligo directed to Syt-VII (siRNA SytVII - Syt7RSS339874– Invitrogen, Carlsbad, CA, USA) or a negative control siRNA (medium GC oligo), as an experimental control. After confirming the success of the Syt-VII knockdown by RT-PCR (Fig. 7B) we performed a lysosomal exocytosis assay in control or MβCD treated cells (Fig. 7C). We found that the absence of Syt-VII did not abolish lysosomal exocytic events triggered by MβCD treatment, demonstrating that the exocytosis of peripheral lysosomes induced by cholesterol sequestration does not require the calcium sensor Syt-VII (Fig. 7C), and therefore is independent of intracellular calcium signaling. We have also shown that MβCD induced lysosomal exocytosis in primary cardiomyocyte (CM) cultures was independent of calcium by performing the β-hex assay in CM cultures treated with 10 mM of MβCD in the presence of 1 mM of BAPTA-AM, an intracellular calcium chelating agent (Fig. S1).

Lysosomal Exocytosis Triggered by MβCD may be Controlled by Actin Polymerization

There is significant evidence in the literature that the actin cytoskeleton plays a role in exocytic events. While some authors demonstrated that the actin cortical cytoskeleton works as a barrier to vesicle secretion [35], [36], [37], others have shown that actin can act as a facilitator of exocytic events through the formation of a readily releasable pool of vesicles, near the plasma membrane and/or by controlling vesicle fusion of pre-docked vesicles [38], [39].

Since cholesterol removal in our assays led to an increase in lysosomal exocytosis, without the apparent participation of calcium, we set out to investigate whether actin polymerization and stress fiber formation induced by cholesterol sequestration could be playing a role in this process. In order to answer this question, fibroblasts were pre-treated or not with Lat-A and submitted to incubation with 10 mM of MβCD for 10, 20 or 40 minutes. β-hex activity was measured for each condition, including control cells and cells treated only with Lat-A (Fig. 7D). Notably, cells treated only with Lat-A exhibited a considerable level of β-hex release when compared with control cells (Fig. 7B). Moreover, Lat-A incubation and subsequent 10 mM MβCD treatments, for 20 and 40 minutes, induced higher levels of lysosomal exocytosis when compared to the exocytosis in cells only exposed to Lat-A (Fig. 7D).

We also performed qualitative and quantitative analysis of lysosomal distribution in control and MβCD treated cells previously exposed or not to Lat-A (Fig. 8). Fibroblasts treated only with Lat-A were not analyzed since these cells present a great variation in their morphology and cell area in comparison to control and the other treatments (see Fig. 3C). Figure 8A–C shows representative images of lysosomal distribution by immunofluorescence in control fibroblasts (Fig. 8A), fibroblasts treated with 10 mM MβCD alone (Fig. 8B) or after Lat-A 95 nM incubation (Fig. 8C). Qualitatively, we can notice that fibroblasts with less cholesterol content in their membranes (Fig. 8B) exhibit, in average, lysosomes nearer their respective nuclei in comparison to untreated cells (Fig. 8A). Interestingly, fibroblasts treated first with Lat-A and then with MβCD 10 mM (Fig. 8C) showed lysosomes even closer to cell nuclei in comparison to cells treated only with MβCD (Fig. 8B). In order to precisely determine these differences, the same images were used to perform a quantitative assay of lysosomal dispersion (Fig. 8D and E). First, for each isolated nucleus it was calculated the mean radius (R). The next step was to select each lysosome associated with its respective nucleus and to measure the mean distance between a lysosome and cell center (D). Lysosomes farther than 1.5 radii were not considered in the calculation to avoid inclusion of lysosomes from neighboring cells. Finally, the mean lysosome distance (D) relative to the mean nucleus’ radius (R) was defined as the ratio D/R, where values closer to one indicate lysosomes are nearer the perinuclear region whereas the opposite indicates lysosomes are more frequent at cell borders. This ratio D/R was measured for several groups of lysosomes associated with each nucleus in the different treatments. The results of this analysis are distributions of D/R values associated to each drug treatment, and are represented as histograms (Fig. 8D and E). Gaussian fits from control cells show that the majority of lysosomes are preferentially localized at ratio 1.07 from cell center whereas the peak of Gaussian fits from 10 mM MβCD treated cells show a ratio of 1.00 (Fig. 8D). Gaussian distribution from fibroblasts treated with Lat-A and MβCD indicates that lysosomes are preferentially located at ratio 0.93 (Fig. 8E). Cumulative frequencies of lysosomes (Fig. 8F) from the histograms of figures 8D and E were plotted and analyzed using Kolmogorov-Smirnov statistical test. Statistically significant differences were observed among all groups. In order to prove that lysosomal dispersion had not been altered due to changes in cell area, we also quantified the mean cell surface area for the different conditions and no statistically significant differences was detected among groups (Fig. 8G).

Download:

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

Figure 8. Cholesterol sequestration mobilizes a different lysosomal pool in comparison to cytoskeleton disruption.

(A–C) Representative images of control, MβCD 10 mM treated and Lat-A+MβCD 10 mM treated cells, fixed and immunolabeled for DAPI and LAMP-1. PC: phase contrast. Scale bar: 10 µm. (D–E) Quantitative analysis of lysosomal distribution, relative to cell nuclei, in MβCD 10 mM (D) and Lat-A 95 nM+MβCD 10 mM (E) in comparison to control non-treated cells. The mean distance between a lysosome and its respective cell center is represented by letter D and the mean lysosome distance relative to the mean nucleus’ radius R was defined as the ratio D/R. The histogram for the control (non-treated) lysosomal distribution is represented with a red solid thick line, MβCD 10 mM is represented with a blue dashed line and Lat-A+MβCD combined treatment is represented with a green dashed line. Overlaid Gaussian fits to each distribution were added in order to highlight the main statistical trend of each diagram. The x-axis of each graph represents the ratio D/R whereas the y-axis represents the percentage of analyzed nuclei from each treatment normalized by total cell number. (F) Cumulative frequency of the histograms. Plus sign blue line represents MβCD 10 mM treated cells, dashed green line represent Lat-A+MβCD 10 mM treated cells and red continuous line represent control untreated cells. The Kolmogorov- Smirnov (KS) statistical test was performed to compare the cumulative frequency distributions. Statistically significant differences were obtained for MβCD 10 mM 0.25% (p = 0.0025), Lat-A+MβCD 10 mM 0.000091% (p = 0.00000091) treated cells in relation to control, and for MβCD 10 mM in relation to Lat-A+MβCD 10 mM 4,3% (p = 0.043). (G) Cell area comparison between control and treated groups. No statistical differences were detected among them (p>0.05 student T-test). Three independent experiments were performed in triplicates and at least 20 cells from each condition were analyzed. Data are shown as mean of at least 20 cells from a representative experiment ± standard error.

https://doi.org/10.1371/journal.pone.0082988.g008

The results obtained for lysosomal secretion and distribution upon treatment with each drug, Lat A and MβCD, alone or in combination, strongly suggest that fibroblasts have at least two distinct pools of lysosomes, which respond differently to exocytosis stimulus triggered by cholesterol sequestration or by actin cytoskeleton disruption.